diff options
Diffstat (limited to 'drivers')
197 files changed, 33499 insertions, 4245 deletions
diff --git a/drivers/gpu/drm/amd/amdgpu/Kconfig b/drivers/gpu/drm/amd/amdgpu/Kconfig index f3cb69de0c44..53cf3971dfc3 100644 --- a/drivers/gpu/drm/amd/amdgpu/Kconfig +++ b/drivers/gpu/drm/amd/amdgpu/Kconfig @@ -1,3 +1,10 @@ +config DRM_AMDGPU_SI + bool "Enable amdgpu support for SI parts" + depends on DRM_AMDGPU + help + Choose this option if you want to enable experimental support + for SI asics. + config DRM_AMDGPU_CIK bool "Enable amdgpu support for CIK parts" depends on DRM_AMDGPU diff --git a/drivers/gpu/drm/amd/amdgpu/Makefile b/drivers/gpu/drm/amd/amdgpu/Makefile index 21dd7c00da15..dc6df075bafc 100644 --- a/drivers/gpu/drm/amd/amdgpu/Makefile +++ b/drivers/gpu/drm/amd/amdgpu/Makefile @@ -30,6 +30,8 @@ amdgpu-$(CONFIG_DRM_AMDGPU_CIK)+= cik.o cik_ih.o kv_smc.o kv_dpm.o \ ci_smc.o ci_dpm.o dce_v8_0.o gfx_v7_0.o cik_sdma.o uvd_v4_2.o vce_v2_0.o \ amdgpu_amdkfd_gfx_v7.o +amdgpu-$(CONFIG_DRM_AMDGPU_SI)+= si.o gmc_v6_0.o gfx_v6_0.o si_ih.o si_dma.o dce_v6_0.o si_dpm.o si_smc.o + amdgpu-y += \ vi.o diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h index 3cc2629eb158..ee45d9f7f3dc 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h @@ -64,6 +64,7 @@ extern int amdgpu_modeset; extern int amdgpu_vram_limit; extern int amdgpu_gart_size; +extern int amdgpu_moverate; extern int amdgpu_benchmarking; extern int amdgpu_testing; extern int amdgpu_audio; @@ -94,6 +95,7 @@ extern unsigned amdgpu_pg_mask; extern char *amdgpu_disable_cu; extern int amdgpu_sclk_deep_sleep_en; extern char *amdgpu_virtual_display; +extern unsigned amdgpu_pp_feature_mask; #define AMDGPU_WAIT_IDLE_TIMEOUT_IN_MS 3000 #define AMDGPU_MAX_USEC_TIMEOUT 100000 /* 100 ms */ @@ -108,7 +110,7 @@ extern char *amdgpu_virtual_display; #define AMDGPU_MAX_RINGS 16 #define AMDGPU_MAX_GFX_RINGS 1 #define AMDGPU_MAX_COMPUTE_RINGS 8 -#define AMDGPU_MAX_VCE_RINGS 2 +#define AMDGPU_MAX_VCE_RINGS 3 /* max number of IP instances */ #define AMDGPU_MAX_SDMA_INSTANCES 2 @@ -318,6 +320,10 @@ struct amdgpu_ring_funcs { /* note usage for clock and power gating */ void (*begin_use)(struct amdgpu_ring *ring); void (*end_use)(struct amdgpu_ring *ring); + void (*emit_switch_buffer) (struct amdgpu_ring *ring); + void (*emit_cntxcntl) (struct amdgpu_ring *ring, uint32_t flags); + unsigned (*get_emit_ib_size) (struct amdgpu_ring *ring); + unsigned (*get_dma_frame_size) (struct amdgpu_ring *ring); }; /* @@ -618,6 +624,7 @@ void amdgpu_gart_unbind(struct amdgpu_device *adev, unsigned offset, int amdgpu_gart_bind(struct amdgpu_device *adev, unsigned offset, int pages, struct page **pagelist, dma_addr_t *dma_addr, uint32_t flags); +int amdgpu_ttm_recover_gart(struct amdgpu_device *adev); /* * GPU MC structures, functions & helpers @@ -963,6 +970,7 @@ struct amdgpu_ctx { spinlock_t ring_lock; struct fence **fences; struct amdgpu_ctx_ring rings[AMDGPU_MAX_RINGS]; + bool preamble_presented; }; struct amdgpu_ctx_mgr { @@ -1222,11 +1230,16 @@ struct amdgpu_cs_parser { struct fence *fence; uint64_t bytes_moved_threshold; uint64_t bytes_moved; + struct amdgpu_bo_list_entry *evictable; /* user fence */ struct amdgpu_bo_list_entry uf_entry; }; +#define AMDGPU_PREAMBLE_IB_PRESENT (1 << 0) /* bit set means command submit involves a preamble IB */ +#define AMDGPU_PREAMBLE_IB_PRESENT_FIRST (1 << 1) /* bit set means preamble IB is first presented in belonging context */ +#define AMDGPU_HAVE_CTX_SWITCH (1 << 2) /* bit set means context switch occured */ + struct amdgpu_job { struct amd_sched_job base; struct amdgpu_device *adev; @@ -1235,9 +1248,10 @@ struct amdgpu_job { struct amdgpu_sync sync; struct amdgpu_ib *ibs; struct fence *fence; /* the hw fence */ + uint32_t preamble_status; uint32_t num_ibs; void *owner; - uint64_t ctx; + uint64_t fence_ctx; /* the fence_context this job uses */ bool vm_needs_flush; unsigned vm_id; uint64_t vm_pd_addr; @@ -1686,6 +1700,7 @@ struct amdgpu_vce { unsigned harvest_config; struct amd_sched_entity entity; uint32_t srbm_soft_reset; + unsigned num_rings; }; /* @@ -1703,6 +1718,10 @@ struct amdgpu_sdma_instance { struct amdgpu_sdma { struct amdgpu_sdma_instance instance[AMDGPU_MAX_SDMA_INSTANCES]; +#ifdef CONFIG_DRM_AMDGPU_SI + //SI DMA has a difference trap irq number for the second engine + struct amdgpu_irq_src trap_irq_1; +#endif struct amdgpu_irq_src trap_irq; struct amdgpu_irq_src illegal_inst_irq; int num_instances; @@ -1819,6 +1838,9 @@ struct amdgpu_asic_funcs { int (*set_vce_clocks)(struct amdgpu_device *adev, u32 evclk, u32 ecclk); /* query virtual capabilities */ u32 (*get_virtual_caps)(struct amdgpu_device *adev); + /* static power management */ + int (*get_pcie_lanes)(struct amdgpu_device *adev); + void (*set_pcie_lanes)(struct amdgpu_device *adev, int lanes); }; /* @@ -1993,6 +2015,8 @@ struct amdgpu_device { spinlock_t pcie_idx_lock; amdgpu_rreg_t pcie_rreg; amdgpu_wreg_t pcie_wreg; + amdgpu_rreg_t pciep_rreg; + amdgpu_wreg_t pciep_wreg; /* protects concurrent UVD register access */ spinlock_t uvd_ctx_idx_lock; amdgpu_rreg_t uvd_ctx_rreg; @@ -2033,6 +2057,14 @@ struct amdgpu_device { atomic64_t num_evictions; atomic_t gpu_reset_counter; + /* data for buffer migration throttling */ + struct { + spinlock_t lock; + s64 last_update_us; + s64 accum_us; /* accumulated microseconds */ + u32 log2_max_MBps; + } mm_stats; + /* display */ bool enable_virtual_display; struct amdgpu_mode_info mode_info; @@ -2101,6 +2133,10 @@ struct amdgpu_device { /* link all shadow bo */ struct list_head shadow_list; struct mutex shadow_list_lock; + /* link all gtt */ + spinlock_t gtt_list_lock; + struct list_head gtt_list; + }; bool amdgpu_device_is_px(struct drm_device *dev); @@ -2133,6 +2169,8 @@ void amdgpu_mm_wdoorbell(struct amdgpu_device *adev, u32 index, u32 v); #define REG_GET(FIELD, v) (((v) << FIELD##_SHIFT) & FIELD##_MASK) #define RREG32_PCIE(reg) adev->pcie_rreg(adev, (reg)) #define WREG32_PCIE(reg, v) adev->pcie_wreg(adev, (reg), (v)) +#define RREG32_PCIE_PORT(reg) adev->pciep_rreg(adev, (reg)) +#define WREG32_PCIE_PORT(reg, v) adev->pciep_wreg(adev, (reg), (v)) #define RREG32_SMC(reg) adev->smc_rreg(adev, (reg)) #define WREG32_SMC(reg, v) adev->smc_wreg(adev, (reg), (v)) #define RREG32_UVD_CTX(reg) adev->uvd_ctx_rreg(adev, (reg)) @@ -2223,6 +2261,9 @@ amdgpu_get_sdma_instance(struct amdgpu_ring *ring) #define amdgpu_asic_set_uvd_clocks(adev, v, d) (adev)->asic_funcs->set_uvd_clocks((adev), (v), (d)) #define amdgpu_asic_set_vce_clocks(adev, ev, ec) (adev)->asic_funcs->set_vce_clocks((adev), (ev), (ec)) #define amdgpu_asic_get_virtual_caps(adev) ((adev)->asic_funcs->get_virtual_caps((adev))) +#define amdgpu_get_pcie_lanes(adev) (adev)->asic_funcs->get_pcie_lanes((adev)) +#define amdgpu_set_pcie_lanes(adev, l) (adev)->asic_funcs->set_pcie_lanes((adev), (l)) +#define amdgpu_asic_get_gpu_clock_counter(adev) (adev)->asic_funcs->get_gpu_clock_counter((adev)) #define amdgpu_asic_read_disabled_bios(adev) (adev)->asic_funcs->read_disabled_bios((adev)) #define amdgpu_asic_read_bios_from_rom(adev, b, l) (adev)->asic_funcs->read_bios_from_rom((adev), (b), (l)) #define amdgpu_asic_read_register(adev, se, sh, offset, v)((adev)->asic_funcs->read_register((adev), (se), (sh), (offset), (v))) @@ -2244,9 +2285,13 @@ amdgpu_get_sdma_instance(struct amdgpu_ring *ring) #define amdgpu_ring_emit_gds_switch(r, v, db, ds, wb, ws, ab, as) (r)->funcs->emit_gds_switch((r), (v), (db), (ds), (wb), (ws), (ab), (as)) #define amdgpu_ring_emit_hdp_flush(r) (r)->funcs->emit_hdp_flush((r)) #define amdgpu_ring_emit_hdp_invalidate(r) (r)->funcs->emit_hdp_invalidate((r)) +#define amdgpu_ring_emit_switch_buffer(r) (r)->funcs->emit_switch_buffer((r)) +#define amdgpu_ring_emit_cntxcntl(r, d) (r)->funcs->emit_cntxcntl((r), (d)) #define amdgpu_ring_pad_ib(r, ib) ((r)->funcs->pad_ib((r), (ib))) #define amdgpu_ring_init_cond_exec(r) (r)->funcs->init_cond_exec((r)) #define amdgpu_ring_patch_cond_exec(r,o) (r)->funcs->patch_cond_exec((r),(o)) +#define amdgpu_ring_get_emit_ib_size(r) (r)->funcs->get_emit_ib_size((r)) +#define amdgpu_ring_get_dma_frame_size(r) (r)->funcs->get_dma_frame_size((r)) #define amdgpu_ih_get_wptr(adev) (adev)->irq.ih_funcs->get_wptr((adev)) #define amdgpu_ih_decode_iv(adev, iv) (adev)->irq.ih_funcs->decode_iv((adev), (iv)) #define amdgpu_ih_set_rptr(adev) (adev)->irq.ih_funcs->set_rptr((adev)) @@ -2402,6 +2447,8 @@ void amdgpu_gtt_location(struct amdgpu_device *adev, struct amdgpu_mc *mc); void amdgpu_ttm_set_active_vram_size(struct amdgpu_device *adev, u64 size); u64 amdgpu_ttm_get_gtt_mem_size(struct amdgpu_device *adev); int amdgpu_ttm_global_init(struct amdgpu_device *adev); +int amdgpu_ttm_init(struct amdgpu_device *adev); +void amdgpu_ttm_fini(struct amdgpu_device *adev); void amdgpu_program_register_sequence(struct amdgpu_device *adev, const u32 *registers, const u32 array_size); @@ -2434,8 +2481,8 @@ void amdgpu_driver_postclose_kms(struct drm_device *dev, struct drm_file *file_priv); void amdgpu_driver_preclose_kms(struct drm_device *dev, struct drm_file *file_priv); -int amdgpu_suspend_kms(struct drm_device *dev, bool suspend, bool fbcon); -int amdgpu_resume_kms(struct drm_device *dev, bool resume, bool fbcon); +int amdgpu_device_suspend(struct drm_device *dev, bool suspend, bool fbcon); +int amdgpu_device_resume(struct drm_device *dev, bool resume, bool fbcon); u32 amdgpu_get_vblank_counter_kms(struct drm_device *dev, unsigned int pipe); int amdgpu_enable_vblank_kms(struct drm_device *dev, unsigned int pipe); void amdgpu_disable_vblank_kms(struct drm_device *dev, unsigned int pipe); @@ -2481,6 +2528,7 @@ static inline void amdgpu_acpi_fini(struct amdgpu_device *adev) { } struct amdgpu_bo_va_mapping * amdgpu_cs_find_mapping(struct amdgpu_cs_parser *parser, uint64_t addr, struct amdgpu_bo **bo); +int amdgpu_cs_sysvm_access_required(struct amdgpu_cs_parser *parser); #include "amdgpu_object.h" #endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v7.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v7.c index 362bedc9e507..1a0a5f7cccbc 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v7.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v7.c @@ -103,11 +103,11 @@ static bool kgd_hqd_is_occupied(struct kgd_dev *kgd, uint64_t queue_address, uint32_t pipe_id, uint32_t queue_id); static int kgd_hqd_destroy(struct kgd_dev *kgd, uint32_t reset_type, - unsigned int timeout, uint32_t pipe_id, + unsigned int utimeout, uint32_t pipe_id, uint32_t queue_id); static bool kgd_hqd_sdma_is_occupied(struct kgd_dev *kgd, void *mqd); static int kgd_hqd_sdma_destroy(struct kgd_dev *kgd, void *mqd, - unsigned int timeout); + unsigned int utimeout); static int kgd_address_watch_disable(struct kgd_dev *kgd); static int kgd_address_watch_execute(struct kgd_dev *kgd, unsigned int watch_point_id, @@ -437,11 +437,12 @@ static bool kgd_hqd_sdma_is_occupied(struct kgd_dev *kgd, void *mqd) } static int kgd_hqd_destroy(struct kgd_dev *kgd, uint32_t reset_type, - unsigned int timeout, uint32_t pipe_id, + unsigned int utimeout, uint32_t pipe_id, uint32_t queue_id) { struct amdgpu_device *adev = get_amdgpu_device(kgd); uint32_t temp; + int timeout = utimeout; acquire_queue(kgd, pipe_id, queue_id); WREG32(mmCP_HQD_PQ_DOORBELL_CONTROL, 0); @@ -452,9 +453,8 @@ static int kgd_hqd_destroy(struct kgd_dev *kgd, uint32_t reset_type, temp = RREG32(mmCP_HQD_ACTIVE); if (temp & CP_HQD_ACTIVE__ACTIVE_MASK) break; - if (timeout == 0) { - pr_err("kfd: cp queue preemption time out (%dms)\n", - temp); + if (timeout <= 0) { + pr_err("kfd: cp queue preemption time out.\n"); release_queue(kgd); return -ETIME; } @@ -467,12 +467,13 @@ static int kgd_hqd_destroy(struct kgd_dev *kgd, uint32_t reset_type, } static int kgd_hqd_sdma_destroy(struct kgd_dev *kgd, void *mqd, - unsigned int timeout) + unsigned int utimeout) { struct amdgpu_device *adev = get_amdgpu_device(kgd); struct cik_sdma_rlc_registers *m; uint32_t sdma_base_addr; uint32_t temp; + int timeout = utimeout; m = get_sdma_mqd(mqd); sdma_base_addr = get_sdma_base_addr(m); @@ -485,7 +486,7 @@ static int kgd_hqd_sdma_destroy(struct kgd_dev *kgd, void *mqd, temp = RREG32(sdma_base_addr + mmSDMA0_RLC0_CONTEXT_STATUS); if (temp & SDMA0_STATUS_REG__RB_CMD_IDLE__SHIFT) break; - if (timeout == 0) + if (timeout <= 0) return -ETIME; msleep(20); timeout -= 20; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v8.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v8.c index 04b744d64b57..6697612239c2 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v8.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v8.c @@ -62,10 +62,10 @@ static bool kgd_hqd_is_occupied(struct kgd_dev *kgd, uint64_t queue_address, uint32_t pipe_id, uint32_t queue_id); static bool kgd_hqd_sdma_is_occupied(struct kgd_dev *kgd, void *mqd); static int kgd_hqd_destroy(struct kgd_dev *kgd, uint32_t reset_type, - unsigned int timeout, uint32_t pipe_id, + unsigned int utimeout, uint32_t pipe_id, uint32_t queue_id); static int kgd_hqd_sdma_destroy(struct kgd_dev *kgd, void *mqd, - unsigned int timeout); + unsigned int utimeout); static void write_vmid_invalidate_request(struct kgd_dev *kgd, uint8_t vmid); static int kgd_address_watch_disable(struct kgd_dev *kgd); static int kgd_address_watch_execute(struct kgd_dev *kgd, @@ -349,11 +349,12 @@ static bool kgd_hqd_sdma_is_occupied(struct kgd_dev *kgd, void *mqd) } static int kgd_hqd_destroy(struct kgd_dev *kgd, uint32_t reset_type, - unsigned int timeout, uint32_t pipe_id, + unsigned int utimeout, uint32_t pipe_id, uint32_t queue_id) { struct amdgpu_device *adev = get_amdgpu_device(kgd); uint32_t temp; + int timeout = utimeout; acquire_queue(kgd, pipe_id, queue_id); @@ -363,9 +364,8 @@ static int kgd_hqd_destroy(struct kgd_dev *kgd, uint32_t reset_type, temp = RREG32(mmCP_HQD_ACTIVE); if (temp & CP_HQD_ACTIVE__ACTIVE_MASK) break; - if (timeout == 0) { - pr_err("kfd: cp queue preemption time out (%dms)\n", - temp); + if (timeout <= 0) { + pr_err("kfd: cp queue preemption time out.\n"); release_queue(kgd); return -ETIME; } @@ -378,12 +378,13 @@ static int kgd_hqd_destroy(struct kgd_dev *kgd, uint32_t reset_type, } static int kgd_hqd_sdma_destroy(struct kgd_dev *kgd, void *mqd, - unsigned int timeout) + unsigned int utimeout) { struct amdgpu_device *adev = get_amdgpu_device(kgd); struct cik_sdma_rlc_registers *m; uint32_t sdma_base_addr; uint32_t temp; + int timeout = utimeout; m = get_sdma_mqd(mqd); sdma_base_addr = get_sdma_base_addr(m); @@ -396,7 +397,7 @@ static int kgd_hqd_sdma_destroy(struct kgd_dev *kgd, void *mqd, temp = RREG32(sdma_base_addr + mmSDMA0_RLC0_CONTEXT_STATUS); if (temp & SDMA0_STATUS_REG__RB_CMD_IDLE__SHIFT) break; - if (timeout == 0) + if (timeout <= 0) return -ETIME; msleep(20); timeout -= 20; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_atombios.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_atombios.c index 1b621160b52e..59961db9c390 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_atombios.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_atombios.c @@ -978,6 +978,48 @@ int amdgpu_atombios_get_clock_dividers(struct amdgpu_device *adev, return -EINVAL; switch (crev) { + case 2: + case 3: + case 5: + /* r6xx, r7xx, evergreen, ni, si. + * TODO: add support for asic_type <= CHIP_RV770*/ + if (clock_type == COMPUTE_ENGINE_PLL_PARAM) { + args.v3.ulClockParams = cpu_to_le32((clock_type << 24) | clock); + + amdgpu_atom_execute_table(adev->mode_info.atom_context, index, (uint32_t *)&args); + + dividers->post_div = args.v3.ucPostDiv; + dividers->enable_post_div = (args.v3.ucCntlFlag & + ATOM_PLL_CNTL_FLAG_PLL_POST_DIV_EN) ? true : false; + dividers->enable_dithen = (args.v3.ucCntlFlag & + ATOM_PLL_CNTL_FLAG_FRACTION_DISABLE) ? false : true; + dividers->whole_fb_div = le16_to_cpu(args.v3.ulFbDiv.usFbDiv); + dividers->frac_fb_div = le16_to_cpu(args.v3.ulFbDiv.usFbDivFrac); + dividers->ref_div = args.v3.ucRefDiv; + dividers->vco_mode = (args.v3.ucCntlFlag & + ATOM_PLL_CNTL_FLAG_MPLL_VCO_MODE) ? 1 : 0; + } else { + /* for SI we use ComputeMemoryClockParam for memory plls */ + if (adev->asic_type >= CHIP_TAHITI) + return -EINVAL; + args.v5.ulClockParams = cpu_to_le32((clock_type << 24) | clock); + if (strobe_mode) + args.v5.ucInputFlag = ATOM_PLL_INPUT_FLAG_PLL_STROBE_MODE_EN; + + amdgpu_atom_execute_table(adev->mode_info.atom_context, index, (uint32_t *)&args); + + dividers->post_div = args.v5.ucPostDiv; + dividers->enable_post_div = (args.v5.ucCntlFlag & + ATOM_PLL_CNTL_FLAG_PLL_POST_DIV_EN) ? true : false; + dividers->enable_dithen = (args.v5.ucCntlFlag & + ATOM_PLL_CNTL_FLAG_FRACTION_DISABLE) ? false : true; + dividers->whole_fb_div = le16_to_cpu(args.v5.ulFbDiv.usFbDiv); + dividers->frac_fb_div = le16_to_cpu(args.v5.ulFbDiv.usFbDivFrac); + dividers->ref_div = args.v5.ucRefDiv; + dividers->vco_mode = (args.v5.ucCntlFlag & + ATOM_PLL_CNTL_FLAG_MPLL_VCO_MODE) ? 1 : 0; + } + break; case 4: /* fusion */ args.v4.ulClock = cpu_to_le32(clock); /* 10 khz */ @@ -1122,6 +1164,32 @@ void amdgpu_atombios_set_engine_dram_timings(struct amdgpu_device *adev, amdgpu_atom_execute_table(adev->mode_info.atom_context, index, (uint32_t *)&args); } +void amdgpu_atombios_get_default_voltages(struct amdgpu_device *adev, + u16 *vddc, u16 *vddci, u16 *mvdd) +{ + struct amdgpu_mode_info *mode_info = &adev->mode_info; + int index = GetIndexIntoMasterTable(DATA, FirmwareInfo); + u8 frev, crev; + u16 data_offset; + union firmware_info *firmware_info; + + *vddc = 0; + *vddci = 0; + *mvdd = 0; + + if (amdgpu_atom_parse_data_header(mode_info->atom_context, index, NULL, + &frev, &crev, &data_offset)) { + firmware_info = + (union firmware_info *)(mode_info->atom_context->bios + + data_offset); + *vddc = le16_to_cpu(firmware_info->info_14.usBootUpVDDCVoltage); + if ((frev == 2) && (crev >= 2)) { + *vddci = le16_to_cpu(firmware_info->info_22.usBootUpVDDCIVoltage); + *mvdd = le16_to_cpu(firmware_info->info_22.usBootUpMVDDCVoltage); + } + } +} + union set_voltage { struct _SET_VOLTAGE_PS_ALLOCATION alloc; struct _SET_VOLTAGE_PARAMETERS v1; @@ -1129,6 +1197,52 @@ union set_voltage { struct _SET_VOLTAGE_PARAMETERS_V1_3 v3; }; +int amdgpu_atombios_get_max_vddc(struct amdgpu_device *adev, u8 voltage_type, + u16 voltage_id, u16 *voltage) +{ + union set_voltage args; + int index = GetIndexIntoMasterTable(COMMAND, SetVoltage); + u8 frev, crev; + + if (!amdgpu_atom_parse_cmd_header(adev->mode_info.atom_context, index, &frev, &crev)) + return -EINVAL; + + switch (crev) { + case 1: + return -EINVAL; + case 2: + args.v2.ucVoltageType = SET_VOLTAGE_GET_MAX_VOLTAGE; + args.v2.ucVoltageMode = 0; + args.v2.usVoltageLevel = 0; + + amdgpu_atom_execute_table(adev->mode_info.atom_context, index, (uint32_t *)&args); + + *voltage = le16_to_cpu(args.v2.usVoltageLevel); + break; + case 3: + args.v3.ucVoltageType = voltage_type; + args.v3.ucVoltageMode = ATOM_GET_VOLTAGE_LEVEL; + args.v3.usVoltageLevel = cpu_to_le16(voltage_id); + + amdgpu_atom_execute_table(adev->mode_info.atom_context, index, (uint32_t *)&args); + + *voltage = le16_to_cpu(args.v3.usVoltageLevel); + break; + default: + DRM_ERROR("Unknown table version %d, %d\n", frev, crev); + return -EINVAL; + } + + return 0; +} + +int amdgpu_atombios_get_leakage_vddc_based_on_leakage_idx(struct amdgpu_device *adev, + u16 *voltage, + u16 leakage_idx) +{ + return amdgpu_atombios_get_max_vddc(adev, VOLTAGE_TYPE_VDDC, leakage_idx, voltage); +} + void amdgpu_atombios_set_voltage(struct amdgpu_device *adev, u16 voltage_level, u8 voltage_type) @@ -1349,6 +1463,50 @@ static ATOM_VOLTAGE_OBJECT_V3 *amdgpu_atombios_lookup_voltage_object_v3(ATOM_VOL return NULL; } +int amdgpu_atombios_get_svi2_info(struct amdgpu_device *adev, + u8 voltage_type, + u8 *svd_gpio_id, u8 *svc_gpio_id) +{ + int index = GetIndexIntoMasterTable(DATA, VoltageObjectInfo); + u8 frev, crev; + u16 data_offset, size; + union voltage_object_info *voltage_info; + union voltage_object *voltage_object = NULL; + + if (amdgpu_atom_parse_data_header(adev->mode_info.atom_context, index, &size, + &frev, &crev, &data_offset)) { + voltage_info = (union voltage_object_info *) + (adev->mode_info.atom_context->bios + data_offset); + + switch (frev) { + case 3: + switch (crev) { + case 1: + voltage_object = (union voltage_object *) + amdgpu_atombios_lookup_voltage_object_v3(&voltage_info->v3, + voltage_type, + VOLTAGE_OBJ_SVID2); + if (voltage_object) { + *svd_gpio_id = voltage_object->v3.asSVID2Obj.ucSVDGpioId; + *svc_gpio_id = voltage_object->v3.asSVID2Obj.ucSVCGpioId; + } else { + return -EINVAL; + } + break; + default: + DRM_ERROR("unknown voltage object table\n"); + return -EINVAL; + } + break; + default: + DRM_ERROR("unknown voltage object table\n"); + return -EINVAL; + } + + } + return 0; +} + bool amdgpu_atombios_is_voltage_gpio(struct amdgpu_device *adev, u8 voltage_type, u8 voltage_mode) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_atombios.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_atombios.h index 15dd43ec38bb..17356151db38 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_atombios.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_atombios.h @@ -208,5 +208,19 @@ void amdgpu_atombios_scratch_regs_save(struct amdgpu_device *adev); void amdgpu_atombios_scratch_regs_restore(struct amdgpu_device *adev); void amdgpu_atombios_copy_swap(u8 *dst, u8 *src, u8 num_bytes, bool to_le); - +int amdgpu_atombios_get_max_vddc(struct amdgpu_device *adev, u8 voltage_type, + u16 voltage_id, u16 *voltage); +int amdgpu_atombios_get_leakage_vddc_based_on_leakage_idx(struct amdgpu_device *adev, + u16 *voltage, + u16 leakage_idx); +void amdgpu_atombios_get_default_voltages(struct amdgpu_device *adev, + u16 *vddc, u16 *vddci, u16 *mvdd); +int amdgpu_atombios_get_clock_dividers(struct amdgpu_device *adev, + u8 clock_type, + u32 clock, + bool strobe_mode, + struct atom_clock_dividers *dividers); +int amdgpu_atombios_get_svi2_info(struct amdgpu_device *adev, + u8 voltage_type, + u8 *svd_gpio_id, u8 *svc_gpio_id); #endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c index bc0440f7a31d..f1c53a2b09c6 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c @@ -616,7 +616,7 @@ static int amdgpu_cgs_irq_put(struct cgs_device *cgs_device, unsigned src_id, un return amdgpu_irq_put(adev, adev->irq.sources[src_id], type); } -int amdgpu_cgs_set_clockgating_state(struct cgs_device *cgs_device, +static int amdgpu_cgs_set_clockgating_state(struct cgs_device *cgs_device, enum amd_ip_block_type block_type, enum amd_clockgating_state state) { @@ -637,7 +637,7 @@ int amdgpu_cgs_set_clockgating_state(struct cgs_device *cgs_device, return r; } -int amdgpu_cgs_set_powergating_state(struct cgs_device *cgs_device, +static int amdgpu_cgs_set_powergating_state(struct cgs_device *cgs_device, enum amd_ip_block_type block_type, enum amd_powergating_state state) { @@ -848,6 +848,12 @@ static int amdgpu_cgs_query_system_info(struct cgs_device *cgs_device, case CGS_SYSTEM_INFO_GFX_SE_INFO: sys_info->value = adev->gfx.config.max_shader_engines; break; + case CGS_SYSTEM_INFO_PCIE_SUB_SYS_ID: + sys_info->value = adev->pdev->subsystem_device; + break; + case CGS_SYSTEM_INFO_PCIE_SUB_SYS_VENDOR_ID: + sys_info->value = adev->pdev->subsystem_vendor; + break; default: return -ENODEV; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c index d80e5d3a4add..b8412bcbad2a 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c @@ -91,6 +91,7 @@ static int amdgpu_cs_user_fence_chunk(struct amdgpu_cs_parser *p, uint32_t *offset) { struct drm_gem_object *gobj; + unsigned long size; gobj = drm_gem_object_lookup(p->filp, data->handle); if (gobj == NULL) @@ -101,6 +102,11 @@ static int amdgpu_cs_user_fence_chunk(struct amdgpu_cs_parser *p, p->uf_entry.tv.bo = &p->uf_entry.robj->tbo; p->uf_entry.tv.shared = true; p->uf_entry.user_pages = NULL; + + size = amdgpu_bo_size(p->uf_entry.robj); + if (size != PAGE_SIZE || (data->offset + 8) > size) + return -EINVAL; + *offset = data->offset; drm_gem_object_unreference_unlocked(gobj); @@ -235,56 +241,115 @@ free_chunk: return ret; } -/* Returns how many bytes TTM can move per IB. +/* Convert microseconds to bytes. */ +static u64 us_to_bytes(struct amdgpu_device *adev, s64 us) +{ + if (us <= 0 || !adev->mm_stats.log2_max_MBps) + return 0; + + /* Since accum_us is incremented by a million per second, just + * multiply it by the number of MB/s to get the number of bytes. + */ + return us << adev->mm_stats.log2_max_MBps; +} + +static s64 bytes_to_us(struct amdgpu_device *adev, u64 bytes) +{ + if (!adev->mm_stats.log2_max_MBps) + return 0; + + return bytes >> adev->mm_stats.log2_max_MBps; +} + +/* Returns how many bytes TTM can move right now. If no bytes can be moved, + * it returns 0. If it returns non-zero, it's OK to move at least one buffer, + * which means it can go over the threshold once. If that happens, the driver + * will be in debt and no other buffer migrations can be done until that debt + * is repaid. + * + * This approach allows moving a buffer of any size (it's important to allow + * that). + * + * The currency is simply time in microseconds and it increases as the clock + * ticks. The accumulated microseconds (us) are converted to bytes and + * returned. */ static u64 amdgpu_cs_get_threshold_for_moves(struct amdgpu_device *adev) { - u64 real_vram_size = adev->mc.real_vram_size; - u64 vram_usage = atomic64_read(&adev->vram_usage); + s64 time_us, increment_us; + u64 max_bytes; + u64 free_vram, total_vram, used_vram; - /* This function is based on the current VRAM usage. + /* Allow a maximum of 200 accumulated ms. This is basically per-IB + * throttling. * - * - If all of VRAM is free, allow relocating the number of bytes that - * is equal to 1/4 of the size of VRAM for this IB. + * It means that in order to get full max MBps, at least 5 IBs per + * second must be submitted and not more than 200ms apart from each + * other. + */ + const s64 us_upper_bound = 200000; - * - If more than one half of VRAM is occupied, only allow relocating - * 1 MB of data for this IB. - * - * - From 0 to one half of used VRAM, the threshold decreases - * linearly. - * __________________ - * 1/4 of -|\ | - * VRAM | \ | - * | \ | - * | \ | - * | \ | - * | \ | - * | \ | - * | \________|1 MB - * |----------------| - * VRAM 0 % 100 % - * used used - * - * Note: It's a threshold, not a limit. The threshold must be crossed - * for buffer relocations to stop, so any buffer of an arbitrary size - * can be moved as long as the threshold isn't crossed before - * the relocation takes place. We don't want to disable buffer - * relocations completely. + if (!adev->mm_stats.log2_max_MBps) + return 0; + + total_vram = adev->mc.real_vram_size - adev->vram_pin_size; + used_vram = atomic64_read(&adev->vram_usage); + free_vram = used_vram >= total_vram ? 0 : total_vram - used_vram; + + spin_lock(&adev->mm_stats.lock); + + /* Increase the amount of accumulated us. */ + time_us = ktime_to_us(ktime_get()); + increment_us = time_us - adev->mm_stats.last_update_us; + adev->mm_stats.last_update_us = time_us; + adev->mm_stats.accum_us = min(adev->mm_stats.accum_us + increment_us, + us_upper_bound); + + /* This prevents the short period of low performance when the VRAM + * usage is low and the driver is in debt or doesn't have enough + * accumulated us to fill VRAM quickly. * - * The idea is that buffers should be placed in VRAM at creation time - * and TTM should only do a minimum number of relocations during - * command submission. In practice, you need to submit at least - * a dozen IBs to move all buffers to VRAM if they are in GTT. + * The situation can occur in these cases: + * - a lot of VRAM is freed by userspace + * - the presence of a big buffer causes a lot of evictions + * (solution: split buffers into smaller ones) * - * Also, things can get pretty crazy under memory pressure and actual - * VRAM usage can change a lot, so playing safe even at 50% does - * consistently increase performance. + * If 128 MB or 1/8th of VRAM is free, start filling it now by setting + * accum_us to a positive number. + */ + if (free_vram >= 128 * 1024 * 1024 || free_vram >= total_vram / 8) { + s64 min_us; + + /* Be more aggresive on dGPUs. Try to fill a portion of free + * VRAM now. + */ + if (!(adev->flags & AMD_IS_APU)) + min_us = bytes_to_us(adev, free_vram / 4); + else + min_us = 0; /* Reset accum_us on APUs. */ + + adev->mm_stats.accum_us = max(min_us, adev->mm_stats.accum_us); + } + + /* This returns 0 if the driver is in debt to disallow (optional) + * buffer moves. */ + max_bytes = us_to_bytes(adev, adev->mm_stats.accum_us); - u64 half_vram = real_vram_size >> 1; - u64 half_free_vram = vram_usage >= half_vram ? 0 : half_vram - vram_usage; - u64 bytes_moved_threshold = half_free_vram >> 1; - return max(bytes_moved_threshold, 1024*1024ull); + spin_unlock(&adev->mm_stats.lock); + return max_bytes; +} + +/* Report how many bytes have really been moved for the last command + * submission. This can result in a debt that can stop buffer migrations + * temporarily. + */ +static void amdgpu_cs_report_moved_bytes(struct amdgpu_device *adev, + u64 num_bytes) +{ + spin_lock(&adev->mm_stats.lock); + adev->mm_stats.accum_us -= bytes_to_us(adev, num_bytes); + spin_unlock(&adev->mm_stats.lock); } static int amdgpu_cs_bo_validate(struct amdgpu_cs_parser *p, @@ -297,15 +362,10 @@ static int amdgpu_cs_bo_validate(struct amdgpu_cs_parser *p, if (bo->pin_count) return 0; - /* Avoid moving this one if we have moved too many buffers - * for this IB already. - * - * Note that this allows moving at least one buffer of - * any size, because it doesn't take the current "bo" - * into account. We don't want to disallow buffer moves - * completely. + /* Don't move this buffer if we have depleted our allowance + * to move it. Don't move anything if the threshold is zero. */ - if (p->bytes_moved <= p->bytes_moved_threshold) + if (p->bytes_moved < p->bytes_moved_threshold) domain = bo->prefered_domains; else domain = bo->allowed_domains; @@ -317,17 +377,67 @@ retry: p->bytes_moved += atomic64_read(&bo->adev->num_bytes_moved) - initial_bytes_moved; - if (unlikely(r)) { - if (r != -ERESTARTSYS && domain != bo->allowed_domains) { - domain = bo->allowed_domains; - goto retry; - } + if (unlikely(r == -ENOMEM) && domain != bo->allowed_domains) { + domain = bo->allowed_domains; + goto retry; } return r; } -int amdgpu_cs_list_validate(struct amdgpu_cs_parser *p, +/* Last resort, try to evict something from the current working set */ +static bool amdgpu_cs_try_evict(struct amdgpu_cs_parser *p, + struct amdgpu_bo_list_entry *lobj) +{ + uint32_t domain = lobj->robj->allowed_domains; + int r; + + if (!p->evictable) + return false; + + for (;&p->evictable->tv.head != &p->validated; + p->evictable = list_prev_entry(p->evictable, tv.head)) { + + struct amdgpu_bo_list_entry *candidate = p->evictable; + struct amdgpu_bo *bo = candidate->robj; + u64 initial_bytes_moved; + uint32_t other; + + /* If we reached our current BO we can forget it */ + if (candidate == lobj) + break; + + other = amdgpu_mem_type_to_domain(bo->tbo.mem.mem_type); + + /* Check if this BO is in one of the domains we need space for */ + if (!(other & domain)) + continue; + + /* Check if we can move this BO somewhere else */ + other = bo->allowed_domains & ~domain; + if (!other) + continue; + + /* Good we can try to move this BO somewhere else */ + amdgpu_ttm_placement_from_domain(bo, other); + initial_bytes_moved = atomic64_read(&bo->adev->num_bytes_moved); + r = ttm_bo_validate(&bo->tbo, &bo->placement, true, false); + p->bytes_moved += atomic64_read(&bo->adev->num_bytes_moved) - + initial_bytes_moved; + + if (unlikely(r)) + break; + + p->evictable = list_prev_entry(p->evictable, tv.head); + list_move(&candidate->tv.head, &p->validated); + + return true; + } + + return false; +} + +static int amdgpu_cs_list_validate(struct amdgpu_cs_parser *p, struct list_head *validated) { struct amdgpu_bo_list_entry *lobj; @@ -351,9 +461,15 @@ int amdgpu_cs_list_validate(struct amdgpu_cs_parser *p, binding_userptr = true; } - r = amdgpu_cs_bo_validate(p, bo); + if (p->evictable == lobj) + p->evictable = NULL; + + do { + r = amdgpu_cs_bo_validate(p, bo); + } while (r == -ENOMEM && amdgpu_cs_try_evict(p, lobj)); if (r) return r; + if (bo->shadow) { r = amdgpu_cs_bo_validate(p, bo); if (r) @@ -481,6 +597,9 @@ static int amdgpu_cs_parser_bos(struct amdgpu_cs_parser *p, p->bytes_moved_threshold = amdgpu_cs_get_threshold_for_moves(p->adev); p->bytes_moved = 0; + p->evictable = list_last_entry(&p->validated, + struct amdgpu_bo_list_entry, + tv.head); r = amdgpu_cs_list_validate(p, &duplicates); if (r) { @@ -494,6 +613,8 @@ static int amdgpu_cs_parser_bos(struct amdgpu_cs_parser *p, goto error_validate; } + amdgpu_cs_report_moved_bytes(p->adev, p->bytes_moved); + fpriv->vm.last_eviction_counter = atomic64_read(&p->adev->num_evictions); @@ -524,8 +645,12 @@ static int amdgpu_cs_parser_bos(struct amdgpu_cs_parser *p, } } - if (p->uf_entry.robj) - p->job->uf_addr += amdgpu_bo_gpu_offset(p->uf_entry.robj); + if (!r && p->uf_entry.robj) { + struct amdgpu_bo *uf = p->uf_entry.robj; + + r = amdgpu_ttm_bind(uf->tbo.ttm, &uf->tbo.mem); + p->job->uf_addr += amdgpu_bo_gpu_offset(uf); + } error_validate: if (r) { @@ -735,6 +860,14 @@ static int amdgpu_cs_ib_fill(struct amdgpu_device *adev, if (r) return r; + if (ib->flags & AMDGPU_IB_FLAG_PREAMBLE) { + parser->job->preamble_status |= AMDGPU_PREAMBLE_IB_PRESENT; + if (!parser->ctx->preamble_presented) { + parser->job->preamble_status |= AMDGPU_PREAMBLE_IB_PRESENT_FIRST; + parser->ctx->preamble_presented = true; + } + } + if (parser->job->ring && parser->job->ring != ring) return -EINVAL; @@ -874,7 +1007,7 @@ static int amdgpu_cs_submit(struct amdgpu_cs_parser *p, } job->owner = p->filp; - job->ctx = entity->fence_context; + job->fence_ctx = entity->fence_context; p->fence = fence_get(&job->base.s_fence->finished); cs->out.handle = amdgpu_ctx_add_fence(p->ctx, ring, p->fence); job->uf_sequence = cs->out.handle; @@ -1040,3 +1173,29 @@ amdgpu_cs_find_mapping(struct amdgpu_cs_parser *parser, return NULL; } + +/** + * amdgpu_cs_sysvm_access_required - make BOs accessible by the system VM + * + * @parser: command submission parser context + * + * Helper for UVD/VCE VM emulation, make sure BOs are accessible by the system VM. + */ +int amdgpu_cs_sysvm_access_required(struct amdgpu_cs_parser *parser) +{ + unsigned i; + int r; + + if (!parser->bo_list) + return 0; + + for (i = 0; i < parser->bo_list->num_entries; i++) { + struct amdgpu_bo *bo = parser->bo_list->array[i].robj; + + r = amdgpu_ttm_bind(bo->tbo.ttm, &bo->tbo.mem); + if (unlikely(r)) + return r; + } + + return 0; +} diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c index c38dc47cd767..3ddae5ff41bb 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c @@ -41,6 +41,9 @@ #include "atom.h" #include "amdgpu_atombios.h" #include "amd_pcie.h" +#ifdef CONFIG_DRM_AMDGPU_SI +#include "si.h" +#endif #ifdef CONFIG_DRM_AMDGPU_CIK #include "cik.h" #endif @@ -52,6 +55,11 @@ static int amdgpu_debugfs_regs_init(struct amdgpu_device *adev); static void amdgpu_debugfs_regs_cleanup(struct amdgpu_device *adev); static const char *amdgpu_asic_name[] = { + "TAHITI", + "PITCAIRN", + "VERDE", + "OLAND", + "HAINAN", "BONAIRE", "KAVERI", "KABINI", @@ -1027,7 +1035,7 @@ static void amdgpu_switcheroo_set_state(struct pci_dev *pdev, enum vga_switchero /* don't suspend or resume card normally */ dev->switch_power_state = DRM_SWITCH_POWER_CHANGING; - amdgpu_resume_kms(dev, true, true); + amdgpu_device_resume(dev, true, true); dev->pdev->d3_delay = d3_delay; @@ -1037,7 +1045,7 @@ static void amdgpu_switcheroo_set_state(struct pci_dev *pdev, enum vga_switchero printk(KERN_INFO "amdgpu: switched off\n"); drm_kms_helper_poll_disable(dev); dev->switch_power_state = DRM_SWITCH_POWER_CHANGING; - amdgpu_suspend_kms(dev, true, true); + amdgpu_device_suspend(dev, true, true); dev->switch_power_state = DRM_SWITCH_POWER_OFF; } } @@ -1231,6 +1239,18 @@ static int amdgpu_early_init(struct amdgpu_device *adev) if (r) return r; break; +#ifdef CONFIG_DRM_AMDGPU_SI + case CHIP_VERDE: + case CHIP_TAHITI: + case CHIP_PITCAIRN: + case CHIP_OLAND: + case CHIP_HAINAN: + adev->family = AMDGPU_FAMILY_SI; + r = si_set_ip_blocks(adev); + if (r) + return r; + break; +#endif #ifdef CONFIG_DRM_AMDGPU_CIK case CHIP_BONAIRE: case CHIP_HAWAII: @@ -1347,6 +1367,9 @@ static int amdgpu_late_init(struct amdgpu_device *adev) for (i = 0; i < adev->num_ip_blocks; i++) { if (!adev->ip_block_status[i].valid) continue; + if (adev->ip_blocks[i].type == AMD_IP_BLOCK_TYPE_UVD || + adev->ip_blocks[i].type == AMD_IP_BLOCK_TYPE_VCE) + continue; /* enable clockgating to save power */ r = adev->ip_blocks[i].funcs->set_clockgating_state((void *)adev, AMD_CG_STATE_GATE); @@ -1490,6 +1513,7 @@ int amdgpu_device_init(struct amdgpu_device *adev, { int r, i; bool runtime = false; + u32 max_MBps; adev->shutdown = false; adev->dev = &pdev->dev; @@ -1513,6 +1537,8 @@ int amdgpu_device_init(struct amdgpu_device *adev, adev->smc_wreg = &amdgpu_invalid_wreg; adev->pcie_rreg = &amdgpu_invalid_rreg; adev->pcie_wreg = &amdgpu_invalid_wreg; + adev->pciep_rreg = &amdgpu_invalid_rreg; + adev->pciep_wreg = &amdgpu_invalid_wreg; adev->uvd_ctx_rreg = &amdgpu_invalid_rreg; adev->uvd_ctx_wreg = &amdgpu_invalid_wreg; adev->didt_rreg = &amdgpu_invalid_rreg; @@ -1549,12 +1575,22 @@ int amdgpu_device_init(struct amdgpu_device *adev, spin_lock_init(&adev->didt_idx_lock); spin_lock_init(&adev->gc_cac_idx_lock); spin_lock_init(&adev->audio_endpt_idx_lock); + spin_lock_init(&adev->mm_stats.lock); INIT_LIST_HEAD(&adev->shadow_list); mutex_init(&adev->shadow_list_lock); - adev->rmmio_base = pci_resource_start(adev->pdev, 5); - adev->rmmio_size = pci_resource_len(adev->pdev, 5); + INIT_LIST_HEAD(&adev->gtt_list); + spin_lock_init(&adev->gtt_list_lock); + + if (adev->asic_type >= CHIP_BONAIRE) { + adev->rmmio_base = pci_resource_start(adev->pdev, 5); + adev->rmmio_size = pci_resource_len(adev->pdev, 5); + } else { + adev->rmmio_base = pci_resource_start(adev->pdev, 2); + adev->rmmio_size = pci_resource_len(adev->pdev, 2); + } + adev->rmmio = ioremap(adev->rmmio_base, adev->rmmio_size); if (adev->rmmio == NULL) { return -ENOMEM; @@ -1562,8 +1598,9 @@ int amdgpu_device_init(struct amdgpu_device *adev, DRM_INFO("register mmio base: 0x%08X\n", (uint32_t)adev->rmmio_base); DRM_INFO("register mmio size: %u\n", (unsigned)adev->rmmio_size); - /* doorbell bar mapping */ - amdgpu_doorbell_init(adev); + if (adev->asic_type >= CHIP_BONAIRE) + /* doorbell bar mapping */ + amdgpu_doorbell_init(adev); /* io port mapping */ for (i = 0; i < DEVICE_COUNT_RESOURCE; i++) { @@ -1660,6 +1697,14 @@ int amdgpu_device_init(struct amdgpu_device *adev, adev->accel_working = true; + /* Initialize the buffer migration limit. */ + if (amdgpu_moverate >= 0) + max_MBps = amdgpu_moverate; + else + max_MBps = 8; /* Allow 8 MB/s. */ + /* Get a log2 for easy divisions. */ + adev->mm_stats.log2_max_MBps = ilog2(max(1u, max_MBps)); + amdgpu_fbdev_init(adev); r = amdgpu_ib_pool_init(adev); @@ -1764,7 +1809,8 @@ void amdgpu_device_fini(struct amdgpu_device *adev) adev->rio_mem = NULL; iounmap(adev->rmmio); adev->rmmio = NULL; - amdgpu_doorbell_fini(adev); + if (adev->asic_type >= CHIP_BONAIRE) + amdgpu_doorbell_fini(adev); amdgpu_debugfs_regs_cleanup(adev); amdgpu_debugfs_remove_files(adev); } @@ -1774,7 +1820,7 @@ void amdgpu_device_fini(struct amdgpu_device *adev) * Suspend & resume. */ /** - * amdgpu_suspend_kms - initiate device suspend + * amdgpu_device_suspend - initiate device suspend * * @pdev: drm dev pointer * @state: suspend state @@ -1783,7 +1829,7 @@ void amdgpu_device_fini(struct amdgpu_device *adev) * Returns 0 for success or an error on failure. * Called at driver suspend. */ -int amdgpu_suspend_kms(struct drm_device *dev, bool suspend, bool fbcon) +int amdgpu_device_suspend(struct drm_device *dev, bool suspend, bool fbcon) { struct amdgpu_device *adev; struct drm_crtc *crtc; @@ -1796,7 +1842,8 @@ int amdgpu_suspend_kms(struct drm_device *dev, bool suspend, bool fbcon) adev = dev->dev_private; - if (dev->switch_power_state == DRM_SWITCH_POWER_OFF) + if (dev->switch_power_state == DRM_SWITCH_POWER_OFF || + dev->switch_power_state == DRM_SWITCH_POWER_DYNAMIC_OFF) return 0; drm_kms_helper_poll_disable(dev); @@ -1851,6 +1898,10 @@ int amdgpu_suspend_kms(struct drm_device *dev, bool suspend, bool fbcon) /* Shut down the device */ pci_disable_device(dev->pdev); pci_set_power_state(dev->pdev, PCI_D3hot); + } else { + r = amdgpu_asic_reset(adev); + if (r) + DRM_ERROR("amdgpu asic reset failed\n"); } if (fbcon) { @@ -1862,7 +1913,7 @@ int amdgpu_suspend_kms(struct drm_device *dev, bool suspend, bool fbcon) } /** - * amdgpu_resume_kms - initiate device resume + * amdgpu_device_resume - initiate device resume * * @pdev: drm dev pointer * @@ -1870,32 +1921,37 @@ int amdgpu_suspend_kms(struct drm_device *dev, bool suspend, bool fbcon) * Returns 0 for success or an error on failure. * Called at driver resume. */ -int amdgpu_resume_kms(struct drm_device *dev, bool resume, bool fbcon) +int amdgpu_device_resume(struct drm_device *dev, bool resume, bool fbcon) { struct drm_connector *connector; struct amdgpu_device *adev = dev->dev_private; struct drm_crtc *crtc; int r; - if (dev->switch_power_state == DRM_SWITCH_POWER_OFF) + if (dev->switch_power_state == DRM_SWITCH_POWER_OFF || + dev->switch_power_state == DRM_SWITCH_POWER_DYNAMIC_OFF) return 0; - if (fbcon) { + if (fbcon) console_lock(); - } + if (resume) { pci_set_power_state(dev->pdev, PCI_D0); pci_restore_state(dev->pdev); - if (pci_enable_device(dev->pdev)) { + r = pci_enable_device(dev->pdev); + if (r) { if (fbcon) console_unlock(); - return -1; + return r; } } /* post card */ - if (!amdgpu_card_posted(adev)) - amdgpu_atom_asic_init(adev->mode_info.atom_context); + if (!amdgpu_card_posted(adev) || !resume) { + r = amdgpu_atom_asic_init(adev->mode_info.atom_context); + if (r) + DRM_ERROR("amdgpu asic init failed\n"); + } r = amdgpu_resume(adev); if (r) @@ -2163,6 +2219,11 @@ retry: } if (!r) { amdgpu_irq_gpu_reset_resume_helper(adev); + if (need_full_reset && amdgpu_need_backup(adev)) { + r = amdgpu_ttm_recover_gart(adev); + if (r) + DRM_ERROR("gart recovery failed!!!\n"); + } r = amdgpu_ib_ring_tests(adev); if (r) { dev_err(adev->dev, "ib ring test failed (%d).\n", r); @@ -2600,7 +2661,7 @@ static ssize_t amdgpu_debugfs_regs_smc_read(struct file *f, char __user *buf, while (size) { uint32_t value; - value = RREG32_SMC(*pos >> 2); + value = RREG32_SMC(*pos); r = put_user(value, (uint32_t *)buf); if (r) return r; @@ -2631,7 +2692,7 @@ static ssize_t amdgpu_debugfs_regs_smc_write(struct file *f, const char __user * if (r) return r; - WREG32_SMC(*pos >> 2, value); + WREG32_SMC(*pos, value); result += 4; buf += 4; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c index 7c911d0be2b3..596362624610 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c @@ -55,13 +55,15 @@ * - 3.3.0 - Add VM support for UVD on supported hardware. * - 3.4.0 - Add AMDGPU_INFO_NUM_EVICTIONS. * - 3.5.0 - Add support for new UVD_NO_OP register. + * - 3.6.0 - kmd involves use CONTEXT_CONTROL in ring buffer. */ #define KMS_DRIVER_MAJOR 3 -#define KMS_DRIVER_MINOR 5 +#define KMS_DRIVER_MINOR 6 #define KMS_DRIVER_PATCHLEVEL 0 int amdgpu_vram_limit = 0; int amdgpu_gart_size = -1; /* auto */ +int amdgpu_moverate = -1; /* auto */ int amdgpu_benchmarking = 0; int amdgpu_testing = 0; int amdgpu_audio = -1; @@ -93,6 +95,7 @@ unsigned amdgpu_cg_mask = 0xffffffff; unsigned amdgpu_pg_mask = 0xffffffff; char *amdgpu_disable_cu = NULL; char *amdgpu_virtual_display = NULL; +unsigned amdgpu_pp_feature_mask = 0xffffffff; MODULE_PARM_DESC(vramlimit, "Restrict VRAM for testing, in megabytes"); module_param_named(vramlimit, amdgpu_vram_limit, int, 0600); @@ -100,6 +103,9 @@ module_param_named(vramlimit, amdgpu_vram_limit, int, 0600); MODULE_PARM_DESC(gartsize, "Size of PCIE/IGP gart to setup in megabytes (32, 64, etc., -1 = auto)"); module_param_named(gartsize, amdgpu_gart_size, int, 0600); +MODULE_PARM_DESC(moverate, "Maximum buffer migration rate in MB/s. (32, 64, etc., -1=auto, 0=1=disabled)"); +module_param_named(moverate, amdgpu_moverate, int, 0600); + MODULE_PARM_DESC(benchmark, "Run benchmark"); module_param_named(benchmark, amdgpu_benchmarking, int, 0444); @@ -172,6 +178,9 @@ module_param_named(powerplay, amdgpu_powerplay, int, 0444); MODULE_PARM_DESC(powercontainment, "Power Containment (1 = enable (default), 0 = disable)"); module_param_named(powercontainment, amdgpu_powercontainment, int, 0444); + +MODULE_PARM_DESC(ppfeaturemask, "all power features enabled (default))"); +module_param_named(ppfeaturemask, amdgpu_pp_feature_mask, int, 0444); #endif MODULE_PARM_DESC(sclkdeepsleep, "SCLK Deep Sleep (1 = enable (default), 0 = disable)"); @@ -196,6 +205,80 @@ MODULE_PARM_DESC(virtual_display, "Enable virtual display feature (the virtual_d module_param_named(virtual_display, amdgpu_virtual_display, charp, 0444); static const struct pci_device_id pciidlist[] = { +#ifdef CONFIG_DRM_AMDGPU_SI + {0x1002, 0x6780, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_TAHITI}, + {0x1002, 0x6784, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_TAHITI}, + {0x1002, 0x6788, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_TAHITI}, + {0x1002, 0x678A, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_TAHITI}, + {0x1002, 0x6790, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_TAHITI}, + {0x1002, 0x6791, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_TAHITI}, + {0x1002, 0x6792, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_TAHITI}, + {0x1002, 0x6798, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_TAHITI}, + {0x1002, 0x6799, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_TAHITI}, + {0x1002, 0x679A, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_TAHITI}, + {0x1002, 0x679B, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_TAHITI}, + {0x1002, 0x679E, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_TAHITI}, + {0x1002, 0x679F, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_TAHITI}, + {0x1002, 0x6800, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_PITCAIRN|AMD_IS_MOBILITY}, + {0x1002, 0x6801, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_PITCAIRN|AMD_IS_MOBILITY}, + {0x1002, 0x6802, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_PITCAIRN|AMD_IS_MOBILITY}, + {0x1002, 0x6806, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_PITCAIRN}, + {0x1002, 0x6808, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_PITCAIRN}, + {0x1002, 0x6809, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_PITCAIRN}, + {0x1002, 0x6810, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_PITCAIRN}, + {0x1002, 0x6811, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_PITCAIRN}, + {0x1002, 0x6816, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_PITCAIRN}, + {0x1002, 0x6817, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_PITCAIRN}, + {0x1002, 0x6818, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_PITCAIRN}, + {0x1002, 0x6819, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_PITCAIRN}, + {0x1002, 0x6600, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_OLAND|AMD_IS_MOBILITY}, + {0x1002, 0x6601, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_OLAND|AMD_IS_MOBILITY}, + {0x1002, 0x6602, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_OLAND|AMD_IS_MOBILITY}, + {0x1002, 0x6603, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_OLAND|AMD_IS_MOBILITY}, + {0x1002, 0x6604, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_OLAND|AMD_IS_MOBILITY}, + {0x1002, 0x6605, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_OLAND|AMD_IS_MOBILITY}, + {0x1002, 0x6606, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_OLAND|AMD_IS_MOBILITY}, + {0x1002, 0x6607, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_OLAND|AMD_IS_MOBILITY}, + {0x1002, 0x6608, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_OLAND}, + {0x1002, 0x6610, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_OLAND}, + {0x1002, 0x6611, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_OLAND}, + {0x1002, 0x6613, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_OLAND}, + {0x1002, 0x6617, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_OLAND|AMD_IS_MOBILITY}, + {0x1002, 0x6620, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_OLAND|AMD_IS_MOBILITY}, + {0x1002, 0x6621, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_OLAND|AMD_IS_MOBILITY}, + {0x1002, 0x6623, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_OLAND|AMD_IS_MOBILITY}, + {0x1002, 0x6631, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_OLAND}, + {0x1002, 0x6820, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VERDE|AMD_IS_MOBILITY}, + {0x1002, 0x6821, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VERDE|AMD_IS_MOBILITY}, + {0x1002, 0x6822, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VERDE|AMD_IS_MOBILITY}, + {0x1002, 0x6823, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VERDE|AMD_IS_MOBILITY}, + {0x1002, 0x6824, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VERDE|AMD_IS_MOBILITY}, + {0x1002, 0x6825, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VERDE|AMD_IS_MOBILITY}, + {0x1002, 0x6826, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VERDE|AMD_IS_MOBILITY}, + {0x1002, 0x6827, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VERDE|AMD_IS_MOBILITY}, + {0x1002, 0x6828, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VERDE}, + {0x1002, 0x6829, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VERDE}, + {0x1002, 0x682A, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VERDE|AMD_IS_MOBILITY}, + {0x1002, 0x682B, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VERDE|AMD_IS_MOBILITY}, + {0x1002, 0x682C, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VERDE}, + {0x1002, 0x682D, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VERDE|AMD_IS_MOBILITY}, + {0x1002, 0x682F, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VERDE|AMD_IS_MOBILITY}, + {0x1002, 0x6830, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VERDE|AMD_IS_MOBILITY}, + {0x1002, 0x6831, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VERDE|AMD_IS_MOBILITY}, + {0x1002, 0x6835, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VERDE}, + {0x1002, 0x6837, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VERDE}, + {0x1002, 0x6838, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VERDE}, + {0x1002, 0x6839, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VERDE}, + {0x1002, 0x683B, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VERDE}, + {0x1002, 0x683D, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VERDE}, + {0x1002, 0x683F, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VERDE}, + {0x1002, 0x6660, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_HAINAN|AMD_IS_MOBILITY}, + {0x1002, 0x6663, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_HAINAN|AMD_IS_MOBILITY}, + {0x1002, 0x6664, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_HAINAN|AMD_IS_MOBILITY}, + {0x1002, 0x6665, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_HAINAN|AMD_IS_MOBILITY}, + {0x1002, 0x6667, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_HAINAN|AMD_IS_MOBILITY}, + {0x1002, 0x666F, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_HAINAN|AMD_IS_MOBILITY}, +#endif #ifdef CONFIG_DRM_AMDGPU_CIK /* Kaveri */ {0x1002, 0x1304, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_KAVERI|AMD_IS_MOBILITY|AMD_IS_APU}, @@ -393,32 +476,72 @@ amdgpu_pci_remove(struct pci_dev *pdev) drm_put_dev(dev); } +static void +amdgpu_pci_shutdown(struct pci_dev *pdev) +{ + struct drm_device *dev = pci_get_drvdata(pdev); + struct amdgpu_device *adev = dev->dev_private; + + /* if we are running in a VM, make sure the device + * torn down properly on reboot/shutdown + */ + if (adev->virtualization.is_virtual) + amdgpu_pci_remove(pdev); +} + static int amdgpu_pmops_suspend(struct device *dev) { struct pci_dev *pdev = to_pci_dev(dev); + struct drm_device *drm_dev = pci_get_drvdata(pdev); - return amdgpu_suspend_kms(drm_dev, true, true); + return amdgpu_device_suspend(drm_dev, true, true); } static int amdgpu_pmops_resume(struct device *dev) { struct pci_dev *pdev = to_pci_dev(dev); struct drm_device *drm_dev = pci_get_drvdata(pdev); - return amdgpu_resume_kms(drm_dev, true, true); + + /* GPU comes up enabled by the bios on resume */ + if (amdgpu_device_is_px(drm_dev)) { + pm_runtime_disable(dev); + pm_runtime_set_active(dev); + pm_runtime_enable(dev); + } + + return amdgpu_device_resume(drm_dev, true, true); } static int amdgpu_pmops_freeze(struct device *dev) { struct pci_dev *pdev = to_pci_dev(dev); + struct drm_device *drm_dev = pci_get_drvdata(pdev); - return amdgpu_suspend_kms(drm_dev, false, true); + return amdgpu_device_suspend(drm_dev, false, true); } static int amdgpu_pmops_thaw(struct device *dev) { struct pci_dev *pdev = to_pci_dev(dev); + + struct drm_device *drm_dev = pci_get_drvdata(pdev); + return amdgpu_device_resume(drm_dev, false, true); +} + +static int amdgpu_pmops_poweroff(struct device *dev) +{ + struct pci_dev *pdev = to_pci_dev(dev); + + struct drm_device *drm_dev = pci_get_drvdata(pdev); + return amdgpu_device_suspend(drm_dev, true, true); +} + +static int amdgpu_pmops_restore(struct device *dev) +{ + struct pci_dev *pdev = to_pci_dev(dev); + struct drm_device *drm_dev = pci_get_drvdata(pdev); - return amdgpu_resume_kms(drm_dev, false, true); + return amdgpu_device_resume(drm_dev, false, true); } static int amdgpu_pmops_runtime_suspend(struct device *dev) @@ -436,7 +559,7 @@ static int amdgpu_pmops_runtime_suspend(struct device *dev) drm_kms_helper_poll_disable(drm_dev); vga_switcheroo_set_dynamic_switch(pdev, VGA_SWITCHEROO_OFF); - ret = amdgpu_suspend_kms(drm_dev, false, false); + ret = amdgpu_device_suspend(drm_dev, false, false); pci_save_state(pdev); pci_disable_device(pdev); pci_ignore_hotplug(pdev); @@ -469,7 +592,7 @@ static int amdgpu_pmops_runtime_resume(struct device *dev) return ret; pci_set_master(pdev); - ret = amdgpu_resume_kms(drm_dev, false, false); + ret = amdgpu_device_resume(drm_dev, false, false); drm_kms_helper_poll_enable(drm_dev); vga_switcheroo_set_dynamic_switch(pdev, VGA_SWITCHEROO_ON); drm_dev->switch_power_state = DRM_SWITCH_POWER_ON; @@ -523,8 +646,8 @@ static const struct dev_pm_ops amdgpu_pm_ops = { .resume = amdgpu_pmops_resume, .freeze = amdgpu_pmops_freeze, .thaw = amdgpu_pmops_thaw, - .poweroff = amdgpu_pmops_freeze, - .restore = amdgpu_pmops_resume, + .poweroff = amdgpu_pmops_poweroff, + .restore = amdgpu_pmops_restore, .runtime_suspend = amdgpu_pmops_runtime_suspend, .runtime_resume = amdgpu_pmops_runtime_resume, .runtime_idle = amdgpu_pmops_runtime_idle, @@ -606,6 +729,7 @@ static struct pci_driver amdgpu_kms_pci_driver = { .id_table = pciidlist, .probe = amdgpu_pci_probe, .remove = amdgpu_pci_remove, + .shutdown = amdgpu_pci_shutdown, .driver.pm = &amdgpu_pm_ops, }; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c index bf033b58056c..107fbb2d2847 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c @@ -25,6 +25,7 @@ */ #include <linux/module.h> #include <linux/slab.h> +#include <linux/pm_runtime.h> #include <drm/drmP.h> #include <drm/drm_crtc.h> @@ -47,8 +48,35 @@ struct amdgpu_fbdev { struct amdgpu_device *adev; }; +static int +amdgpufb_open(struct fb_info *info, int user) +{ + struct amdgpu_fbdev *rfbdev = info->par; + struct amdgpu_device *adev = rfbdev->adev; + int ret = pm_runtime_get_sync(adev->ddev->dev); + if (ret < 0 && ret != -EACCES) { + pm_runtime_mark_last_busy(adev->ddev->dev); + pm_runtime_put_autosuspend(adev->ddev->dev); + return ret; + } + return 0; +} + +static int +amdgpufb_release(struct fb_info *info, int user) +{ + struct amdgpu_fbdev *rfbdev = info->par; + struct amdgpu_device *adev = rfbdev->adev; + + pm_runtime_mark_last_busy(adev->ddev->dev); + pm_runtime_put_autosuspend(adev->ddev->dev); + return 0; +} + static struct fb_ops amdgpufb_ops = { .owner = THIS_MODULE, + .fb_open = amdgpufb_open, + .fb_release = amdgpufb_release, .fb_check_var = drm_fb_helper_check_var, .fb_set_par = drm_fb_helper_set_par, .fb_fillrect = drm_fb_helper_cfb_fillrect, diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c index f5810f700668..4127e7ceace0 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c @@ -124,7 +124,8 @@ int amdgpu_ib_schedule(struct amdgpu_ring *ring, unsigned num_ibs, bool skip_preamble, need_ctx_switch; unsigned patch_offset = ~0; struct amdgpu_vm *vm; - uint64_t ctx; + uint64_t fence_ctx; + uint32_t status = 0, alloc_size; unsigned i; int r = 0; @@ -135,10 +136,10 @@ int amdgpu_ib_schedule(struct amdgpu_ring *ring, unsigned num_ibs, /* ring tests don't use a job */ if (job) { vm = job->vm; - ctx = job->ctx; + fence_ctx = job->fence_ctx; } else { vm = NULL; - ctx = 0; + fence_ctx = 0; } if (!ring->ready) { @@ -151,7 +152,10 @@ int amdgpu_ib_schedule(struct amdgpu_ring *ring, unsigned num_ibs, return -EINVAL; } - r = amdgpu_ring_alloc(ring, 256 * num_ibs); + alloc_size = amdgpu_ring_get_dma_frame_size(ring) + + num_ibs * amdgpu_ring_get_emit_ib_size(ring); + + r = amdgpu_ring_alloc(ring, alloc_size); if (r) { dev_err(adev->dev, "scheduling IB failed (%d).\n", r); return r; @@ -174,13 +178,22 @@ int amdgpu_ib_schedule(struct amdgpu_ring *ring, unsigned num_ibs, /* always set cond_exec_polling to CONTINUE */ *ring->cond_exe_cpu_addr = 1; - skip_preamble = ring->current_ctx == ctx; - need_ctx_switch = ring->current_ctx != ctx; + skip_preamble = ring->current_ctx == fence_ctx; + need_ctx_switch = ring->current_ctx != fence_ctx; + if (job && ring->funcs->emit_cntxcntl) { + if (need_ctx_switch) + status |= AMDGPU_HAVE_CTX_SWITCH; + status |= job->preamble_status; + amdgpu_ring_emit_cntxcntl(ring, status); + } + for (i = 0; i < num_ibs; ++i) { ib = &ibs[i]; /* drop preamble IBs if we don't have a context switch */ - if ((ib->flags & AMDGPU_IB_FLAG_PREAMBLE) && skip_preamble) + if ((ib->flags & AMDGPU_IB_FLAG_PREAMBLE) && + skip_preamble && + !(status & AMDGPU_PREAMBLE_IB_PRESENT_FIRST)) continue; amdgpu_ring_emit_ib(ring, ib, job ? job->vm_id : 0, @@ -209,7 +222,9 @@ int amdgpu_ib_schedule(struct amdgpu_ring *ring, unsigned num_ibs, if (patch_offset != ~0 && ring->funcs->patch_cond_exec) amdgpu_ring_patch_cond_exec(ring, patch_offset); - ring->current_ctx = ctx; + ring->current_ctx = fence_ctx; + if (ring->funcs->emit_switch_buffer) + amdgpu_ring_emit_switch_buffer(ring); amdgpu_ring_commit(ring); return 0; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ih.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ih.c index 5ebb3f43feb6..3ab4c65ecc8b 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ih.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ih.c @@ -119,8 +119,6 @@ int amdgpu_ih_ring_init(struct amdgpu_device *adev, unsigned ring_size, */ void amdgpu_ih_ring_fini(struct amdgpu_device *adev) { - int r; - if (adev->irq.ih.use_bus_addr) { if (adev->irq.ih.ring) { /* add 8 bytes for the rptr/wptr shadows and @@ -132,17 +130,9 @@ void amdgpu_ih_ring_fini(struct amdgpu_device *adev) adev->irq.ih.ring = NULL; } } else { - if (adev->irq.ih.ring_obj) { - r = amdgpu_bo_reserve(adev->irq.ih.ring_obj, false); - if (likely(r == 0)) { - amdgpu_bo_kunmap(adev->irq.ih.ring_obj); - amdgpu_bo_unpin(adev->irq.ih.ring_obj); - amdgpu_bo_unreserve(adev->irq.ih.ring_obj); - } - amdgpu_bo_unref(&adev->irq.ih.ring_obj); - adev->irq.ih.ring = NULL; - adev->irq.ih.ring_obj = NULL; - } + amdgpu_bo_free_kernel(&adev->irq.ih.ring_obj, + &adev->irq.ih.gpu_addr, + (void **)&adev->irq.ih.ring); amdgpu_wb_free(adev, adev->irq.ih.wptr_offs); amdgpu_wb_free(adev, adev->irq.ih.rptr_offs); } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c index 6674d40eb3ab..8c5807994073 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c @@ -91,7 +91,7 @@ void amdgpu_job_free_resources(struct amdgpu_job *job) amdgpu_ib_free(job->adev, &job->ibs[i], f); } -void amdgpu_job_free_cb(struct amd_sched_job *s_job) +static void amdgpu_job_free_cb(struct amd_sched_job *s_job) { struct amdgpu_job *job = container_of(s_job, struct amdgpu_job, base); @@ -124,7 +124,7 @@ int amdgpu_job_submit(struct amdgpu_job *job, struct amdgpu_ring *ring, return r; job->owner = owner; - job->ctx = entity->fence_context; + job->fence_ctx = entity->fence_context; *f = fence_get(&job->base.s_fence->finished); amdgpu_job_free_resources(job); amd_sched_entity_push_job(&job->base); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c index b78e74048f3d..c2c7fb140338 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c @@ -296,7 +296,7 @@ static int amdgpu_info_ioctl(struct drm_device *dev, void *data, struct drm_file break; case AMDGPU_HW_IP_VCE: type = AMD_IP_BLOCK_TYPE_VCE; - for (i = 0; i < AMDGPU_MAX_VCE_RINGS; i++) + for (i = 0; i < adev->vce.num_rings; i++) ring_mask |= ((adev->vce.ring[i].ready ? 1 : 0) << i); ib_start_alignment = AMDGPU_GPU_PAGE_SIZE; ib_size_alignment = 1; @@ -542,12 +542,16 @@ int amdgpu_driver_open_kms(struct drm_device *dev, struct drm_file *file_priv) return r; fpriv = kzalloc(sizeof(*fpriv), GFP_KERNEL); - if (unlikely(!fpriv)) - return -ENOMEM; + if (unlikely(!fpriv)) { + r = -ENOMEM; + goto out_suspend; + } r = amdgpu_vm_init(adev, &fpriv->vm); - if (r) - goto error_free; + if (r) { + kfree(fpriv); + goto out_suspend; + } mutex_init(&fpriv->bo_list_lock); idr_init(&fpriv->bo_list_handles); @@ -556,12 +560,9 @@ int amdgpu_driver_open_kms(struct drm_device *dev, struct drm_file *file_priv) file_priv->driver_priv = fpriv; +out_suspend: pm_runtime_mark_last_busy(dev->dev); pm_runtime_put_autosuspend(dev->dev); - return 0; - -error_free: - kfree(fpriv); return r; } @@ -600,6 +601,9 @@ void amdgpu_driver_postclose_kms(struct drm_device *dev, kfree(fpriv); file_priv->driver_priv = NULL; + + pm_runtime_mark_last_busy(dev->dev); + pm_runtime_put_autosuspend(dev->dev); } /** @@ -614,6 +618,7 @@ void amdgpu_driver_postclose_kms(struct drm_device *dev, void amdgpu_driver_preclose_kms(struct drm_device *dev, struct drm_file *file_priv) { + pm_runtime_get_sync(dev->dev); } /* diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c index b17734e0ecc8..428aa00025e4 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c @@ -38,8 +38,6 @@ #include "amdgpu_trace.h" -int amdgpu_ttm_init(struct amdgpu_device *adev); -void amdgpu_ttm_fini(struct amdgpu_device *adev); static u64 amdgpu_get_vis_part_size(struct amdgpu_device *adev, struct ttm_mem_reg *mem) @@ -287,6 +285,35 @@ error_free: return r; } +/** + * amdgpu_bo_free_kernel - free BO for kernel use + * + * @bo: amdgpu BO to free + * + * unmaps and unpin a BO for kernel internal use. + */ +void amdgpu_bo_free_kernel(struct amdgpu_bo **bo, u64 *gpu_addr, + void **cpu_addr) +{ + if (*bo == NULL) + return; + + if (likely(amdgpu_bo_reserve(*bo, false) == 0)) { + if (cpu_addr) + amdgpu_bo_kunmap(*bo); + + amdgpu_bo_unpin(*bo); + amdgpu_bo_unreserve(*bo); + } + amdgpu_bo_unref(bo); + + if (gpu_addr) + *gpu_addr = 0; + + if (cpu_addr) + *cpu_addr = NULL; +} + int amdgpu_bo_create_restricted(struct amdgpu_device *adev, unsigned long size, int byte_align, bool kernel, u32 domain, u64 flags, @@ -646,6 +673,11 @@ int amdgpu_bo_pin_restricted(struct amdgpu_bo *bo, u32 domain, dev_err(bo->adev->dev, "%p pin failed\n", bo); goto error; } + r = amdgpu_ttm_bind(bo->tbo.ttm, &bo->tbo.mem); + if (unlikely(r)) { + dev_err(bo->adev->dev, "%p bind failed\n", bo); + goto error; + } bo->pin_count = 1; if (gpu_addr != NULL) @@ -692,7 +724,7 @@ int amdgpu_bo_unpin(struct amdgpu_bo *bo) bo->adev->vram_pin_size -= amdgpu_bo_size(bo); if (bo->flags & AMDGPU_GEM_CREATE_NO_CPU_ACCESS) bo->adev->invisible_pin_size -= amdgpu_bo_size(bo); - } else { + } else if (bo->tbo.mem.mem_type == TTM_PL_TT) { bo->adev->gart_pin_size -= amdgpu_bo_size(bo); } @@ -918,8 +950,11 @@ void amdgpu_bo_fence(struct amdgpu_bo *bo, struct fence *fence, u64 amdgpu_bo_gpu_offset(struct amdgpu_bo *bo) { WARN_ON_ONCE(bo->tbo.mem.mem_type == TTM_PL_SYSTEM); + WARN_ON_ONCE(bo->tbo.mem.mem_type == TTM_PL_TT && + !amdgpu_ttm_is_bound(bo->tbo.ttm)); WARN_ON_ONCE(!ww_mutex_is_locked(&bo->tbo.resv->lock) && !bo->pin_count); + WARN_ON_ONCE(bo->tbo.mem.start == AMDGPU_BO_INVALID_OFFSET); return bo->tbo.offset; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h index b6a27390ef88..8255034d73eb 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h @@ -31,6 +31,8 @@ #include <drm/amdgpu_drm.h> #include "amdgpu.h" +#define AMDGPU_BO_INVALID_OFFSET LONG_MAX + /** * amdgpu_mem_type_to_domain - return domain corresponding to mem_type * @mem_type: ttm memory type @@ -128,6 +130,8 @@ int amdgpu_bo_create_kernel(struct amdgpu_device *adev, unsigned long size, int align, u32 domain, struct amdgpu_bo **bo_ptr, u64 *gpu_addr, void **cpu_addr); +void amdgpu_bo_free_kernel(struct amdgpu_bo **bo, u64 *gpu_addr, + void **cpu_addr); int amdgpu_bo_kmap(struct amdgpu_bo *bo, void **ptr); void amdgpu_bo_kunmap(struct amdgpu_bo *bo); struct amdgpu_bo *amdgpu_bo_ref(struct amdgpu_bo *bo); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_pll.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_pll.c index d15314957732..8e67c1210d7c 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_pll.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_pll.c @@ -25,6 +25,7 @@ #include "amdgpu.h" #include "atom.h" #include "atombios_encoders.h" +#include "amdgpu_pll.h" #include <asm/div64.h> #include <linux/gcd.h> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_powerplay.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_powerplay.c index 545074479e1f..1e7f160f23d8 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_powerplay.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_powerplay.c @@ -30,6 +30,7 @@ #include "amdgpu_pm.h" #include <drm/amdgpu_drm.h> #include "amdgpu_powerplay.h" +#include "si_dpm.h" #include "cik_dpm.h" #include "vi_dpm.h" @@ -52,10 +53,6 @@ static int amdgpu_powerplay_init(struct amdgpu_device *adev) pp_init->chip_family = adev->family; pp_init->chip_id = adev->asic_type; pp_init->device = amdgpu_cgs_create_device(adev); - pp_init->rev_id = adev->pdev->revision; - pp_init->sub_sys_id = adev->pdev->subsystem_device; - pp_init->sub_vendor_id = adev->pdev->subsystem_vendor; - ret = amd_powerplay_init(pp_init, amd_pp); kfree(pp_init); #endif @@ -63,6 +60,15 @@ static int amdgpu_powerplay_init(struct amdgpu_device *adev) amd_pp->pp_handle = (void *)adev; switch (adev->asic_type) { +#ifdef CONFIG_DRM_AMDGPU_SI + case CHIP_TAHITI: + case CHIP_PITCAIRN: + case CHIP_VERDE: + case CHIP_OLAND: + case CHIP_HAINAN: + amd_pp->ip_funcs = &si_dpm_ip_funcs; + break; +#endif #ifdef CONFIG_DRM_AMDGPU_CIK case CHIP_BONAIRE: case CHIP_HAWAII: diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c index 242ba04bfde6..777f11b63b4c 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c @@ -252,28 +252,17 @@ int amdgpu_ring_init(struct amdgpu_device *adev, struct amdgpu_ring *ring, */ void amdgpu_ring_fini(struct amdgpu_ring *ring) { - int r; - struct amdgpu_bo *ring_obj; - - ring_obj = ring->ring_obj; ring->ready = false; - ring->ring = NULL; - ring->ring_obj = NULL; amdgpu_wb_free(ring->adev, ring->cond_exe_offs); amdgpu_wb_free(ring->adev, ring->fence_offs); amdgpu_wb_free(ring->adev, ring->rptr_offs); amdgpu_wb_free(ring->adev, ring->wptr_offs); - if (ring_obj) { - r = amdgpu_bo_reserve(ring_obj, false); - if (likely(r == 0)) { - amdgpu_bo_kunmap(ring_obj); - amdgpu_bo_unpin(ring_obj); - amdgpu_bo_unreserve(ring_obj); - } - amdgpu_bo_unref(&ring_obj); - } + amdgpu_bo_free_kernel(&ring->ring_obj, + &ring->gpu_addr, + (void **)&ring->ring); + amdgpu_debugfs_ring_fini(ring); } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c index e36bfc47dceb..b63969d7887c 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c @@ -89,10 +89,10 @@ int amdgpu_ttm_global_init(struct amdgpu_device *adev) global_ref->init = &amdgpu_ttm_mem_global_init; global_ref->release = &amdgpu_ttm_mem_global_release; r = drm_global_item_ref(global_ref); - if (r != 0) { + if (r) { DRM_ERROR("Failed setting up TTM memory accounting " "subsystem.\n"); - return r; + goto error_mem; } adev->mman.bo_global_ref.mem_glob = @@ -103,26 +103,30 @@ int amdgpu_ttm_global_init(struct amdgpu_device *adev) global_ref->init = &ttm_bo_global_init; global_ref->release = &ttm_bo_global_release; r = drm_global_item_ref(global_ref); - if (r != 0) { + if (r) { DRM_ERROR("Failed setting up TTM BO subsystem.\n"); - drm_global_item_unref(&adev->mman.mem_global_ref); - return r; + goto error_bo; } ring = adev->mman.buffer_funcs_ring; rq = &ring->sched.sched_rq[AMD_SCHED_PRIORITY_KERNEL]; r = amd_sched_entity_init(&ring->sched, &adev->mman.entity, rq, amdgpu_sched_jobs); - if (r != 0) { + if (r) { DRM_ERROR("Failed setting up TTM BO move run queue.\n"); - drm_global_item_unref(&adev->mman.mem_global_ref); - drm_global_item_unref(&adev->mman.bo_global_ref.ref); - return r; + goto error_entity; } adev->mman.mem_global_referenced = true; return 0; + +error_entity: + drm_global_item_unref(&adev->mman.bo_global_ref.ref); +error_bo: + drm_global_item_unref(&adev->mman.mem_global_ref); +error_mem: + return r; } static void amdgpu_ttm_global_fini(struct amdgpu_device *adev) @@ -197,6 +201,7 @@ static void amdgpu_evict_flags(struct ttm_buffer_object *bo, .lpfn = 0, .flags = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM }; + unsigned i; if (!amdgpu_ttm_bo_is_amdgpu_bo(bo)) { placement->placement = &placements; @@ -208,10 +213,25 @@ static void amdgpu_evict_flags(struct ttm_buffer_object *bo, rbo = container_of(bo, struct amdgpu_bo, tbo); switch (bo->mem.mem_type) { case TTM_PL_VRAM: - if (rbo->adev->mman.buffer_funcs_ring->ready == false) + if (rbo->adev->mman.buffer_funcs_ring->ready == false) { amdgpu_ttm_placement_from_domain(rbo, AMDGPU_GEM_DOMAIN_CPU); - else + } else { amdgpu_ttm_placement_from_domain(rbo, AMDGPU_GEM_DOMAIN_GTT); + for (i = 0; i < rbo->placement.num_placement; ++i) { + if (!(rbo->placements[i].flags & + TTM_PL_FLAG_TT)) + continue; + + if (rbo->placements[i].lpfn) + continue; + + /* set an upper limit to force directly + * allocating address space for the BO. + */ + rbo->placements[i].lpfn = + rbo->adev->mc.gtt_size >> PAGE_SHIFT; + } + } break; case TTM_PL_TT: default: @@ -257,8 +277,12 @@ static int amdgpu_move_blit(struct ttm_buffer_object *bo, new_start = new_mem->start << PAGE_SHIFT; switch (old_mem->mem_type) { - case TTM_PL_VRAM: case TTM_PL_TT: + r = amdgpu_ttm_bind(bo->ttm, old_mem); + if (r) + return r; + + case TTM_PL_VRAM: old_start += bo->bdev->man[old_mem->mem_type].gpu_offset; break; default: @@ -266,8 +290,12 @@ static int amdgpu_move_blit(struct ttm_buffer_object *bo, return -EINVAL; } switch (new_mem->mem_type) { - case TTM_PL_VRAM: case TTM_PL_TT: + r = amdgpu_ttm_bind(bo->ttm, new_mem); + if (r) + return r; + + case TTM_PL_VRAM: new_start += bo->bdev->man[new_mem->mem_type].gpu_offset; break; default: @@ -312,7 +340,7 @@ static int amdgpu_move_vram_ram(struct ttm_buffer_object *bo, placement.num_busy_placement = 1; placement.busy_placement = &placements; placements.fpfn = 0; - placements.lpfn = 0; + placements.lpfn = adev->mc.gtt_size >> PAGE_SHIFT; placements.flags = TTM_PL_MASK_CACHING | TTM_PL_FLAG_TT; r = ttm_bo_mem_space(bo, &placement, &tmp_mem, interruptible, no_wait_gpu); @@ -359,7 +387,7 @@ static int amdgpu_move_ram_vram(struct ttm_buffer_object *bo, placement.num_busy_placement = 1; placement.busy_placement = &placements; placements.fpfn = 0; - placements.lpfn = 0; + placements.lpfn = adev->mc.gtt_size >> PAGE_SHIFT; placements.flags = TTM_PL_MASK_CACHING | TTM_PL_FLAG_TT; r = ttm_bo_mem_space(bo, &placement, &tmp_mem, interruptible, no_wait_gpu); @@ -521,6 +549,7 @@ struct amdgpu_ttm_tt { spinlock_t guptasklock; struct list_head guptasks; atomic_t mmu_invalidations; + struct list_head list; }; int amdgpu_ttm_tt_get_user_pages(struct ttm_tt *ttm, struct page **pages) @@ -638,7 +667,6 @@ static int amdgpu_ttm_backend_bind(struct ttm_tt *ttm, struct ttm_mem_reg *bo_mem) { struct amdgpu_ttm_tt *gtt = (void*)ttm; - uint32_t flags = amdgpu_ttm_tt_pte_flags(gtt->adev, ttm, bo_mem); int r; if (gtt->userptr) { @@ -648,7 +676,7 @@ static int amdgpu_ttm_backend_bind(struct ttm_tt *ttm, return r; } } - gtt->offset = (unsigned long)(bo_mem->start << PAGE_SHIFT); + gtt->offset = (u64)bo_mem->start << PAGE_SHIFT; if (!ttm->num_pages) { WARN(1, "nothing to bind %lu pages for mreg %p back %p!\n", ttm->num_pages, bo_mem, ttm); @@ -659,14 +687,62 @@ static int amdgpu_ttm_backend_bind(struct ttm_tt *ttm, bo_mem->mem_type == AMDGPU_PL_OA) return -EINVAL; + return 0; +} + +bool amdgpu_ttm_is_bound(struct ttm_tt *ttm) +{ + struct amdgpu_ttm_tt *gtt = (void *)ttm; + + return gtt && !list_empty(>t->list); +} + +int amdgpu_ttm_bind(struct ttm_tt *ttm, struct ttm_mem_reg *bo_mem) +{ + struct amdgpu_ttm_tt *gtt = (void *)ttm; + uint32_t flags; + int r; + + if (!ttm || amdgpu_ttm_is_bound(ttm)) + return 0; + + flags = amdgpu_ttm_tt_pte_flags(gtt->adev, ttm, bo_mem); r = amdgpu_gart_bind(gtt->adev, gtt->offset, ttm->num_pages, ttm->pages, gtt->ttm.dma_address, flags); if (r) { - DRM_ERROR("failed to bind %lu pages at 0x%08X\n", - ttm->num_pages, (unsigned)gtt->offset); + DRM_ERROR("failed to bind %lu pages at 0x%08llX\n", + ttm->num_pages, gtt->offset); return r; } + spin_lock(>t->adev->gtt_list_lock); + list_add_tail(>t->list, >t->adev->gtt_list); + spin_unlock(>t->adev->gtt_list_lock); + return 0; +} + +int amdgpu_ttm_recover_gart(struct amdgpu_device *adev) +{ + struct amdgpu_ttm_tt *gtt, *tmp; + struct ttm_mem_reg bo_mem; + uint32_t flags; + int r; + + bo_mem.mem_type = TTM_PL_TT; + spin_lock(&adev->gtt_list_lock); + list_for_each_entry_safe(gtt, tmp, &adev->gtt_list, list) { + flags = amdgpu_ttm_tt_pte_flags(gtt->adev, >t->ttm.ttm, &bo_mem); + r = amdgpu_gart_bind(adev, gtt->offset, gtt->ttm.ttm.num_pages, + gtt->ttm.ttm.pages, gtt->ttm.dma_address, + flags); + if (r) { + spin_unlock(&adev->gtt_list_lock); + DRM_ERROR("failed to bind %lu pages at 0x%08llX\n", + gtt->ttm.ttm.num_pages, gtt->offset); + return r; + } + } + spin_unlock(&adev->gtt_list_lock); return 0; } @@ -674,6 +750,9 @@ static int amdgpu_ttm_backend_unbind(struct ttm_tt *ttm) { struct amdgpu_ttm_tt *gtt = (void *)ttm; + if (!amdgpu_ttm_is_bound(ttm)) + return 0; + /* unbind shouldn't be done for GDS/GWS/OA in ttm_bo_clean_mm */ if (gtt->adev->gart.ready) amdgpu_gart_unbind(gtt->adev, gtt->offset, ttm->num_pages); @@ -681,6 +760,10 @@ static int amdgpu_ttm_backend_unbind(struct ttm_tt *ttm) if (gtt->userptr) amdgpu_ttm_tt_unpin_userptr(ttm); + spin_lock(>t->adev->gtt_list_lock); + list_del_init(>t->list); + spin_unlock(>t->adev->gtt_list_lock); + return 0; } @@ -717,6 +800,7 @@ static struct ttm_tt *amdgpu_ttm_tt_create(struct ttm_bo_device *bdev, kfree(gtt); return NULL; } + INIT_LIST_HEAD(>t->list); return >t->ttm.ttm; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h index 72f6bfc15d8f..3ee825f4de28 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h @@ -26,13 +26,13 @@ #include "gpu_scheduler.h" -#define AMDGPU_PL_GDS TTM_PL_PRIV0 -#define AMDGPU_PL_GWS TTM_PL_PRIV1 -#define AMDGPU_PL_OA TTM_PL_PRIV2 +#define AMDGPU_PL_GDS (TTM_PL_PRIV + 0) +#define AMDGPU_PL_GWS (TTM_PL_PRIV + 1) +#define AMDGPU_PL_OA (TTM_PL_PRIV + 2) -#define AMDGPU_PL_FLAG_GDS TTM_PL_FLAG_PRIV0 -#define AMDGPU_PL_FLAG_GWS TTM_PL_FLAG_PRIV1 -#define AMDGPU_PL_FLAG_OA TTM_PL_FLAG_PRIV2 +#define AMDGPU_PL_FLAG_GDS (TTM_PL_FLAG_PRIV << 0) +#define AMDGPU_PL_FLAG_GWS (TTM_PL_FLAG_PRIV << 1) +#define AMDGPU_PL_FLAG_OA (TTM_PL_FLAG_PRIV << 2) #define AMDGPU_TTM_LRU_SIZE 20 @@ -77,4 +77,7 @@ int amdgpu_fill_buffer(struct amdgpu_bo *bo, struct fence **fence); int amdgpu_mmap(struct file *filp, struct vm_area_struct *vma); +bool amdgpu_ttm_is_bound(struct ttm_tt *ttm); +int amdgpu_ttm_bind(struct ttm_tt *ttm, struct ttm_mem_reg *bo_mem); + #endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c index 5cc95f1a7dab..7a05f79818f1 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c @@ -247,35 +247,28 @@ int amdgpu_ucode_init_bo(struct amdgpu_device *adev) const struct common_firmware_header *header = NULL; err = amdgpu_bo_create(adev, adev->firmware.fw_size, PAGE_SIZE, true, - AMDGPU_GEM_DOMAIN_GTT, 0, NULL, NULL, bo); + AMDGPU_GEM_DOMAIN_GTT, 0, NULL, NULL, bo); if (err) { dev_err(adev->dev, "(%d) Firmware buffer allocate failed\n", err); - err = -ENOMEM; goto failed; } err = amdgpu_bo_reserve(*bo, false); if (err) { - amdgpu_bo_unref(bo); dev_err(adev->dev, "(%d) Firmware buffer reserve failed\n", err); - goto failed; + goto failed_reserve; } err = amdgpu_bo_pin(*bo, AMDGPU_GEM_DOMAIN_GTT, &fw_mc_addr); if (err) { - amdgpu_bo_unreserve(*bo); - amdgpu_bo_unref(bo); dev_err(adev->dev, "(%d) Firmware buffer pin failed\n", err); - goto failed; + goto failed_pin; } err = amdgpu_bo_kmap(*bo, &fw_buf_ptr); if (err) { dev_err(adev->dev, "(%d) Firmware buffer kmap failed\n", err); - amdgpu_bo_unpin(*bo); - amdgpu_bo_unreserve(*bo); - amdgpu_bo_unref(bo); - goto failed; + goto failed_kmap; } amdgpu_bo_unreserve(*bo); @@ -290,10 +283,16 @@ int amdgpu_ucode_init_bo(struct amdgpu_device *adev) fw_offset += ALIGN(le32_to_cpu(header->ucode_size_bytes), PAGE_SIZE); } } + return 0; +failed_kmap: + amdgpu_bo_unpin(*bo); +failed_pin: + amdgpu_bo_unreserve(*bo); +failed_reserve: + amdgpu_bo_unref(bo); failed: - if (err) - adev->firmware.smu_load = false; + adev->firmware.smu_load = false; return err; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c index cc766cc53a87..25dd58a65905 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c @@ -249,22 +249,13 @@ int amdgpu_uvd_sw_init(struct amdgpu_device *adev) int amdgpu_uvd_sw_fini(struct amdgpu_device *adev) { - int r; - kfree(adev->uvd.saved_bo); amd_sched_entity_fini(&adev->uvd.ring.sched, &adev->uvd.entity); - if (adev->uvd.vcpu_bo) { - r = amdgpu_bo_reserve(adev->uvd.vcpu_bo, false); - if (!r) { - amdgpu_bo_kunmap(adev->uvd.vcpu_bo); - amdgpu_bo_unpin(adev->uvd.vcpu_bo); - amdgpu_bo_unreserve(adev->uvd.vcpu_bo); - } - - amdgpu_bo_unref(&adev->uvd.vcpu_bo); - } + amdgpu_bo_free_kernel(&adev->uvd.vcpu_bo, + &adev->uvd.gpu_addr, + (void **)&adev->uvd.cpu_addr); amdgpu_ring_fini(&adev->uvd.ring); @@ -891,6 +882,10 @@ int amdgpu_uvd_ring_parse_cs(struct amdgpu_cs_parser *parser, uint32_t ib_idx) return -EINVAL; } + r = amdgpu_cs_sysvm_access_required(parser); + if (r) + return r; + ctx.parser = parser; ctx.buf_sizes = buf_sizes; ctx.ib_idx = ib_idx; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c index da52af2a935a..2c9ea9b50f48 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c @@ -634,7 +634,11 @@ int amdgpu_vce_ring_parse_cs(struct amdgpu_cs_parser *p, uint32_t ib_idx) uint32_t allocated = 0; uint32_t tmp, handle = 0; uint32_t *size = &tmp; - int i, r = 0, idx = 0; + int i, r, idx = 0; + + r = amdgpu_cs_sysvm_access_required(p); + if (r) + return r; while (idx < ib->length_dw) { uint32_t len = amdgpu_get_ib_value(p, ib_idx, idx); @@ -799,6 +803,18 @@ void amdgpu_vce_ring_emit_fence(struct amdgpu_ring *ring, u64 addr, u64 seq, amdgpu_ring_write(ring, VCE_CMD_END); } +unsigned amdgpu_vce_ring_get_emit_ib_size(struct amdgpu_ring *ring) +{ + return + 4; /* amdgpu_vce_ring_emit_ib */ +} + +unsigned amdgpu_vce_ring_get_dma_frame_size(struct amdgpu_ring *ring) +{ + return + 6; /* amdgpu_vce_ring_emit_fence x1 no user fence */ +} + /** * amdgpu_vce_ring_test_ring - test if VCE ring is working * @@ -850,8 +866,8 @@ int amdgpu_vce_ring_test_ib(struct amdgpu_ring *ring, long timeout) struct fence *fence = NULL; long r; - /* skip vce ring1 ib test for now, since it's not reliable */ - if (ring == &ring->adev->vce.ring[1]) + /* skip vce ring1/2 ib test for now, since it's not reliable */ + if (ring != &ring->adev->vce.ring[0]) return 0; r = amdgpu_vce_get_create_msg(ring, 1, NULL); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.h index 63f83d0d985c..12729d2852df 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.h @@ -42,5 +42,7 @@ int amdgpu_vce_ring_test_ring(struct amdgpu_ring *ring); int amdgpu_vce_ring_test_ib(struct amdgpu_ring *ring, long timeout); void amdgpu_vce_ring_begin_use(struct amdgpu_ring *ring); void amdgpu_vce_ring_end_use(struct amdgpu_ring *ring); +unsigned amdgpu_vce_ring_get_emit_ib_size(struct amdgpu_ring *ring); +unsigned amdgpu_vce_ring_get_dma_frame_size(struct amdgpu_ring *ring); #endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c index bf56f1814437..bd5af328154f 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c @@ -1163,7 +1163,8 @@ int amdgpu_vm_bo_update(struct amdgpu_device *adev, } flags = amdgpu_ttm_tt_pte_flags(adev, bo_va->bo->tbo.ttm, mem); - gtt_flags = (adev == bo_va->bo->adev) ? flags : 0; + gtt_flags = (amdgpu_ttm_is_bound(bo_va->bo->tbo.ttm) && + adev == bo_va->bo->adev) ? flags : 0; spin_lock(&vm->status_lock); if (!list_empty(&bo_va->vm_status)) diff --git a/drivers/gpu/drm/amd/amdgpu/atombios_crtc.c b/drivers/gpu/drm/amd/amdgpu/atombios_crtc.c index 49a39b1a0a96..f7d236f95e74 100644 --- a/drivers/gpu/drm/amd/amdgpu/atombios_crtc.c +++ b/drivers/gpu/drm/amd/amdgpu/atombios_crtc.c @@ -497,7 +497,13 @@ void amdgpu_atombios_crtc_set_disp_eng_pll(struct amdgpu_device *adev, * SetPixelClock provides the dividers */ args.v6.ulDispEngClkFreq = cpu_to_le32(dispclk); - args.v6.ucPpll = ATOM_EXT_PLL1; + if (adev->asic_type == CHIP_TAHITI || + adev->asic_type == CHIP_PITCAIRN || + adev->asic_type == CHIP_VERDE || + adev->asic_type == CHIP_OLAND) + args.v6.ucPpll = ATOM_PPLL0; + else + args.v6.ucPpll = ATOM_EXT_PLL1; break; default: DRM_ERROR("Unknown table version %d %d\n", frev, crev); diff --git a/drivers/gpu/drm/amd/amdgpu/atombios_i2c.c b/drivers/gpu/drm/amd/amdgpu/atombios_i2c.c index bc56c8a181e6..b374653bd6cf 100644 --- a/drivers/gpu/drm/amd/amdgpu/atombios_i2c.c +++ b/drivers/gpu/drm/amd/amdgpu/atombios_i2c.c @@ -27,6 +27,7 @@ #include "amdgpu.h" #include "atom.h" #include "amdgpu_atombios.h" +#include "atombios_i2c.h" #define TARGET_HW_I2C_CLOCK 50 diff --git a/drivers/gpu/drm/amd/amdgpu/ci_dpm.c b/drivers/gpu/drm/amd/amdgpu/ci_dpm.c index a0d63a293bb0..1d8c375a3561 100644 --- a/drivers/gpu/drm/amd/amdgpu/ci_dpm.c +++ b/drivers/gpu/drm/amd/amdgpu/ci_dpm.c @@ -5396,7 +5396,7 @@ static void ci_dpm_disable(struct amdgpu_device *adev) amdgpu_irq_put(adev, &adev->pm.dpm.thermal.irq, AMDGPU_THERMAL_IRQ_HIGH_TO_LOW); - ci_dpm_powergate_uvd(adev, false); + ci_dpm_powergate_uvd(adev, true); if (!amdgpu_ci_is_smc_running(adev)) return; @@ -6036,7 +6036,7 @@ static int ci_dpm_init(struct amdgpu_device *adev) pi->caps_dynamic_ac_timing = true; - pi->uvd_power_gated = false; + pi->uvd_power_gated = true; /* make sure dc limits are valid */ if ((adev->pm.dpm.dyn_state.max_clock_voltage_on_dc.sclk == 0) || @@ -6179,8 +6179,6 @@ static int ci_dpm_late_init(void *handle) if (ret) return ret; - ci_dpm_powergate_uvd(adev, true); - return 0; } diff --git a/drivers/gpu/drm/amd/amdgpu/cik_sdma.c b/drivers/gpu/drm/amd/amdgpu/cik_sdma.c index e71cd12104b3..e6d7bf9520a0 100644 --- a/drivers/gpu/drm/amd/amdgpu/cik_sdma.c +++ b/drivers/gpu/drm/amd/amdgpu/cik_sdma.c @@ -847,6 +847,22 @@ static void cik_sdma_ring_emit_vm_flush(struct amdgpu_ring *ring, amdgpu_ring_write(ring, (0xfff << 16) | 10); /* retry count, poll interval */ } +static unsigned cik_sdma_ring_get_emit_ib_size(struct amdgpu_ring *ring) +{ + return + 7 + 4; /* cik_sdma_ring_emit_ib */ +} + +static unsigned cik_sdma_ring_get_dma_frame_size(struct amdgpu_ring *ring) +{ + return + 6 + /* cik_sdma_ring_emit_hdp_flush */ + 3 + /* cik_sdma_ring_emit_hdp_invalidate */ + 6 + /* cik_sdma_ring_emit_pipeline_sync */ + 12 + /* cik_sdma_ring_emit_vm_flush */ + 9 + 9 + 9; /* cik_sdma_ring_emit_fence x3 for user fence, vm fence */ +} + static void cik_enable_sdma_mgcg(struct amdgpu_device *adev, bool enable) { @@ -1220,6 +1236,8 @@ static const struct amdgpu_ring_funcs cik_sdma_ring_funcs = { .test_ib = cik_sdma_ring_test_ib, .insert_nop = cik_sdma_ring_insert_nop, .pad_ib = cik_sdma_ring_pad_ib, + .get_emit_ib_size = cik_sdma_ring_get_emit_ib_size, + .get_dma_frame_size = cik_sdma_ring_get_dma_frame_size, }; static void cik_sdma_set_ring_funcs(struct amdgpu_device *adev) diff --git a/drivers/gpu/drm/amd/amdgpu/cz_dpm.c b/drivers/gpu/drm/amd/amdgpu/cz_dpm.c index 794c5f36ca68..f80a0834e889 100644 --- a/drivers/gpu/drm/amd/amdgpu/cz_dpm.c +++ b/drivers/gpu/drm/amd/amdgpu/cz_dpm.c @@ -44,6 +44,7 @@ static void cz_dpm_powergate_uvd(struct amdgpu_device *adev, bool gate); static void cz_dpm_powergate_vce(struct amdgpu_device *adev, bool gate); +static void cz_dpm_fini(struct amdgpu_device *adev); static struct cz_ps *cz_get_ps(struct amdgpu_ps *rps) { @@ -350,6 +351,8 @@ static int cz_parse_power_table(struct amdgpu_device *adev) ps = kzalloc(sizeof(struct cz_ps), GFP_KERNEL); if (ps == NULL) { + for (j = 0; j < i; j++) + kfree(adev->pm.dpm.ps[j].ps_priv); kfree(adev->pm.dpm.ps); return -ENOMEM; } @@ -409,11 +412,11 @@ static int cz_dpm_init(struct amdgpu_device *adev) ret = amdgpu_get_platform_caps(adev); if (ret) - return ret; + goto err; ret = amdgpu_parse_extended_power_table(adev); if (ret) - return ret; + goto err; pi->sram_end = SMC_RAM_END; @@ -467,23 +470,26 @@ static int cz_dpm_init(struct amdgpu_device *adev) ret = cz_parse_sys_info_table(adev); if (ret) - return ret; + goto err; cz_patch_voltage_values(adev); cz_construct_boot_state(adev); ret = cz_parse_power_table(adev); if (ret) - return ret; + goto err; ret = cz_process_firmware_header(adev); if (ret) - return ret; + goto err; pi->dpm_enabled = true; pi->uvd_dynamic_pg = false; return 0; +err: + cz_dpm_fini(adev); + return ret; } static void cz_dpm_fini(struct amdgpu_device *adev) @@ -672,17 +678,12 @@ static void cz_reset_ap_mask(struct amdgpu_device *adev) struct cz_power_info *pi = cz_get_pi(adev); pi->active_process_mask = 0; - } static int cz_dpm_download_pptable_from_smu(struct amdgpu_device *adev, void **table) { - int ret = 0; - - ret = cz_smu_download_pptable(adev, table); - - return ret; + return cz_smu_download_pptable(adev, table); } static int cz_dpm_upload_pptable_to_smu(struct amdgpu_device *adev) @@ -822,9 +823,9 @@ static void cz_init_sclk_limit(struct amdgpu_device *adev) pi->sclk_dpm.hard_min_clk = 0; cz_send_msg_to_smc(adev, PPSMC_MSG_GetMaxSclkLevel); level = cz_get_argument(adev); - if (level < table->count) + if (level < table->count) { clock = table->entries[level].clk; - else { + } else { DRM_ERROR("Invalid SLCK Voltage Dependency table entry.\n"); clock = table->entries[table->count - 1].clk; } @@ -850,9 +851,9 @@ static void cz_init_uvd_limit(struct amdgpu_device *adev) pi->uvd_dpm.hard_min_clk = 0; cz_send_msg_to_smc(adev, PPSMC_MSG_GetMaxUvdLevel); level = cz_get_argument(adev); - if (level < table->count) + if (level < table->count) { clock = table->entries[level].vclk; - else { + } else { DRM_ERROR("Invalid UVD Voltage Dependency table entry.\n"); clock = table->entries[table->count - 1].vclk; } @@ -878,9 +879,9 @@ static void cz_init_vce_limit(struct amdgpu_device *adev) pi->vce_dpm.hard_min_clk = table->entries[0].ecclk; cz_send_msg_to_smc(adev, PPSMC_MSG_GetMaxEclkLevel); level = cz_get_argument(adev); - if (level < table->count) + if (level < table->count) { clock = table->entries[level].ecclk; - else { + } else { /* future BIOS would fix this error */ DRM_ERROR("Invalid VCE Voltage Dependency table entry.\n"); clock = table->entries[table->count - 1].ecclk; @@ -907,9 +908,9 @@ static void cz_init_acp_limit(struct amdgpu_device *adev) pi->acp_dpm.hard_min_clk = 0; cz_send_msg_to_smc(adev, PPSMC_MSG_GetMaxAclkLevel); level = cz_get_argument(adev); - if (level < table->count) + if (level < table->count) { clock = table->entries[level].clk; - else { + } else { DRM_ERROR("Invalid ACP Voltage Dependency table entry.\n"); clock = table->entries[table->count - 1].clk; } @@ -934,7 +935,6 @@ static void cz_init_sclk_threshold(struct amdgpu_device *adev) struct cz_power_info *pi = cz_get_pi(adev); pi->low_sclk_interrupt_threshold = 0; - } static void cz_dpm_setup_asic(struct amdgpu_device *adev) @@ -1207,7 +1207,7 @@ static int cz_enable_didt(struct amdgpu_device *adev, bool enable) int ret; if (pi->caps_sq_ramping || pi->caps_db_ramping || - pi->caps_td_ramping || pi->caps_tcp_ramping) { + pi->caps_td_ramping || pi->caps_tcp_ramping) { if (adev->gfx.gfx_current_status != AMDGPU_GFX_SAFE_MODE) { ret = cz_disable_cgpg(adev); if (ret) { @@ -1281,7 +1281,7 @@ static void cz_apply_state_adjust_rules(struct amdgpu_device *adev, ps->force_high = false; ps->need_dfs_bypass = true; pi->video_start = new_rps->dclk || new_rps->vclk || - new_rps->evclk || new_rps->ecclk; + new_rps->evclk || new_rps->ecclk; if ((new_rps->class & ATOM_PPLIB_CLASSIFICATION_UI_MASK) == ATOM_PPLIB_CLASSIFICATION_UI_BATTERY) @@ -1339,7 +1339,6 @@ static int cz_dpm_enable(struct amdgpu_device *adev) } cz_reset_acp_boot_level(adev); - cz_update_current_ps(adev, adev->pm.dpm.boot_ps); return 0; @@ -1669,7 +1668,6 @@ static void cz_dpm_post_set_power_state(struct amdgpu_device *adev) struct amdgpu_ps *ps = &pi->requested_rps; cz_update_current_ps(adev, ps); - } static int cz_dpm_force_highest(struct amdgpu_device *adev) @@ -2201,7 +2199,6 @@ static int cz_update_vce_dpm(struct amdgpu_device *adev) /* Stable Pstate is enabled and we need to set the VCE DPM to highest level */ if (pi->caps_stable_power_state) { pi->vce_dpm.hard_min_clk = table->entries[table->count-1].ecclk; - } else { /* non-stable p-state cases. without vce.Arbiter.EcclkHardMin */ /* leave it as set by user */ /*pi->vce_dpm.hard_min_clk = table->entries[0].ecclk;*/ diff --git a/drivers/gpu/drm/amd/amdgpu/cz_smc.c b/drivers/gpu/drm/amd/amdgpu/cz_smc.c index ac7fee7b7eca..95887e484c51 100644 --- a/drivers/gpu/drm/amd/amdgpu/cz_smc.c +++ b/drivers/gpu/drm/amd/amdgpu/cz_smc.c @@ -29,6 +29,8 @@ #include "cz_smumgr.h" #include "smu_ucode_xfer_cz.h" #include "amdgpu_ucode.h" +#include "cz_dpm.h" +#include "vi_dpm.h" #include "smu/smu_8_0_d.h" #include "smu/smu_8_0_sh_mask.h" @@ -48,7 +50,7 @@ static struct cz_smu_private_data *cz_smu_get_priv(struct amdgpu_device *adev) return priv; } -int cz_send_msg_to_smc_async(struct amdgpu_device *adev, u16 msg) +static int cz_send_msg_to_smc_async(struct amdgpu_device *adev, u16 msg) { int i; u32 content = 0, tmp; @@ -140,7 +142,7 @@ int cz_read_smc_sram_dword(struct amdgpu_device *adev, u32 smc_address, return 0; } -int cz_write_smc_sram_dword(struct amdgpu_device *adev, u32 smc_address, +static int cz_write_smc_sram_dword(struct amdgpu_device *adev, u32 smc_address, u32 value, u32 limit) { int ret; diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c new file mode 100644 index 000000000000..d3512f381e53 --- /dev/null +++ b/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c @@ -0,0 +1,3160 @@ +/* + * Copyright 2015 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + */ +#include "drmP.h" +#include "amdgpu.h" +#include "amdgpu_pm.h" +#include "amdgpu_i2c.h" +#include "atom.h" +#include "amdgpu_atombios.h" +#include "atombios_crtc.h" +#include "atombios_encoders.h" +#include "amdgpu_pll.h" +#include "amdgpu_connectors.h" +#include "si/si_reg.h" +#include "si/sid.h" + +static void dce_v6_0_set_display_funcs(struct amdgpu_device *adev); +static void dce_v6_0_set_irq_funcs(struct amdgpu_device *adev); + +static const u32 crtc_offsets[6] = +{ + SI_CRTC0_REGISTER_OFFSET, + SI_CRTC1_REGISTER_OFFSET, + SI_CRTC2_REGISTER_OFFSET, + SI_CRTC3_REGISTER_OFFSET, + SI_CRTC4_REGISTER_OFFSET, + SI_CRTC5_REGISTER_OFFSET +}; + +static const uint32_t dig_offsets[] = { + SI_CRTC0_REGISTER_OFFSET, + SI_CRTC1_REGISTER_OFFSET, + SI_CRTC2_REGISTER_OFFSET, + SI_CRTC3_REGISTER_OFFSET, + SI_CRTC4_REGISTER_OFFSET, + SI_CRTC5_REGISTER_OFFSET, + (0x13830 - 0x7030) >> 2, +}; + +static const struct { + uint32_t reg; + uint32_t vblank; + uint32_t vline; + uint32_t hpd; + +} interrupt_status_offsets[6] = { { + .reg = DISP_INTERRUPT_STATUS, + .vblank = DISP_INTERRUPT_STATUS__LB_D1_VBLANK_INTERRUPT_MASK, + .vline = DISP_INTERRUPT_STATUS__LB_D1_VLINE_INTERRUPT_MASK, + .hpd = DISP_INTERRUPT_STATUS__DC_HPD1_INTERRUPT_MASK +}, { + .reg = DISP_INTERRUPT_STATUS_CONTINUE, + .vblank = DISP_INTERRUPT_STATUS_CONTINUE__LB_D2_VBLANK_INTERRUPT_MASK, + .vline = DISP_INTERRUPT_STATUS_CONTINUE__LB_D2_VLINE_INTERRUPT_MASK, + .hpd = DISP_INTERRUPT_STATUS_CONTINUE__DC_HPD2_INTERRUPT_MASK +}, { + .reg = DISP_INTERRUPT_STATUS_CONTINUE2, + .vblank = DISP_INTERRUPT_STATUS_CONTINUE2__LB_D3_VBLANK_INTERRUPT_MASK, + .vline = DISP_INTERRUPT_STATUS_CONTINUE2__LB_D3_VLINE_INTERRUPT_MASK, + .hpd = DISP_INTERRUPT_STATUS_CONTINUE2__DC_HPD3_INTERRUPT_MASK +}, { + .reg = DISP_INTERRUPT_STATUS_CONTINUE3, + .vblank = DISP_INTERRUPT_STATUS_CONTINUE3__LB_D4_VBLANK_INTERRUPT_MASK, + .vline = DISP_INTERRUPT_STATUS_CONTINUE3__LB_D4_VLINE_INTERRUPT_MASK, + .hpd = DISP_INTERRUPT_STATUS_CONTINUE3__DC_HPD4_INTERRUPT_MASK +}, { + .reg = DISP_INTERRUPT_STATUS_CONTINUE4, + .vblank = DISP_INTERRUPT_STATUS_CONTINUE4__LB_D5_VBLANK_INTERRUPT_MASK, + .vline = DISP_INTERRUPT_STATUS_CONTINUE4__LB_D5_VLINE_INTERRUPT_MASK, + .hpd = DISP_INTERRUPT_STATUS_CONTINUE4__DC_HPD5_INTERRUPT_MASK +}, { + .reg = DISP_INTERRUPT_STATUS_CONTINUE5, + .vblank = DISP_INTERRUPT_STATUS_CONTINUE5__LB_D6_VBLANK_INTERRUPT_MASK, + .vline = DISP_INTERRUPT_STATUS_CONTINUE5__LB_D6_VLINE_INTERRUPT_MASK, + .hpd = DISP_INTERRUPT_STATUS_CONTINUE5__DC_HPD6_INTERRUPT_MASK +} }; + +static const uint32_t hpd_int_control_offsets[6] = { + DC_HPD1_INT_CONTROL, + DC_HPD2_INT_CONTROL, + DC_HPD3_INT_CONTROL, + DC_HPD4_INT_CONTROL, + DC_HPD5_INT_CONTROL, + DC_HPD6_INT_CONTROL, +}; + +static u32 dce_v6_0_audio_endpt_rreg(struct amdgpu_device *adev, + u32 block_offset, u32 reg) +{ + DRM_INFO("xxxx: dce_v6_0_audio_endpt_rreg ----no impl!!!!\n"); + return 0; +} + +static void dce_v6_0_audio_endpt_wreg(struct amdgpu_device *adev, + u32 block_offset, u32 reg, u32 v) +{ + DRM_INFO("xxxx: dce_v6_0_audio_endpt_wreg ----no impl!!!!\n"); +} + +static bool dce_v6_0_is_in_vblank(struct amdgpu_device *adev, int crtc) +{ + if (RREG32(EVERGREEN_CRTC_STATUS + crtc_offsets[crtc]) & EVERGREEN_CRTC_V_BLANK) + return true; + else + return false; +} + +static bool dce_v6_0_is_counter_moving(struct amdgpu_device *adev, int crtc) +{ + u32 pos1, pos2; + + pos1 = RREG32(EVERGREEN_CRTC_STATUS_POSITION + crtc_offsets[crtc]); + pos2 = RREG32(EVERGREEN_CRTC_STATUS_POSITION + crtc_offsets[crtc]); + + if (pos1 != pos2) + return true; + else + return false; +} + +/** + * dce_v6_0_wait_for_vblank - vblank wait asic callback. + * + * @crtc: crtc to wait for vblank on + * + * Wait for vblank on the requested crtc (evergreen+). + */ +static void dce_v6_0_vblank_wait(struct amdgpu_device *adev, int crtc) +{ + unsigned i = 0; + + if (crtc >= adev->mode_info.num_crtc) + return; + + if (!(RREG32(EVERGREEN_CRTC_CONTROL + crtc_offsets[crtc]) & EVERGREEN_CRTC_MASTER_EN)) + return; + + /* depending on when we hit vblank, we may be close to active; if so, + * wait for another frame. + */ + while (dce_v6_0_is_in_vblank(adev, crtc)) { + if (i++ % 100 == 0) { + if (!dce_v6_0_is_counter_moving(adev, crtc)) + break; + } + } + + while (!dce_v6_0_is_in_vblank(adev, crtc)) { + if (i++ % 100 == 0) { + if (!dce_v6_0_is_counter_moving(adev, crtc)) + break; + } + } +} + +static u32 dce_v6_0_vblank_get_counter(struct amdgpu_device *adev, int crtc) +{ + if (crtc >= adev->mode_info.num_crtc) + return 0; + else + return RREG32(CRTC_STATUS_FRAME_COUNT + crtc_offsets[crtc]); +} + +static void dce_v6_0_pageflip_interrupt_init(struct amdgpu_device *adev) +{ + unsigned i; + + /* Enable pflip interrupts */ + for (i = 0; i <= adev->mode_info.num_crtc; i++) + amdgpu_irq_get(adev, &adev->pageflip_irq, i); +} + +static void dce_v6_0_pageflip_interrupt_fini(struct amdgpu_device *adev) +{ + unsigned i; + + /* Disable pflip interrupts */ + for (i = 0; i <= adev->mode_info.num_crtc; i++) + amdgpu_irq_put(adev, &adev->pageflip_irq, i); +} + +/** + * dce_v6_0_page_flip - pageflip callback. + * + * @adev: amdgpu_device pointer + * @crtc_id: crtc to cleanup pageflip on + * @crtc_base: new address of the crtc (GPU MC address) + * + * Does the actual pageflip (evergreen+). + * During vblank we take the crtc lock and wait for the update_pending + * bit to go high, when it does, we release the lock, and allow the + * double buffered update to take place. + * Returns the current update pending status. + */ +static void dce_v6_0_page_flip(struct amdgpu_device *adev, + int crtc_id, u64 crtc_base, bool async) +{ + struct amdgpu_crtc *amdgpu_crtc = adev->mode_info.crtcs[crtc_id]; + + /* flip at hsync for async, default is vsync */ + WREG32(EVERGREEN_GRPH_FLIP_CONTROL + amdgpu_crtc->crtc_offset, async ? + EVERGREEN_GRPH_SURFACE_UPDATE_H_RETRACE_EN : 0); + /* update the scanout addresses */ + WREG32(EVERGREEN_GRPH_PRIMARY_SURFACE_ADDRESS_HIGH + amdgpu_crtc->crtc_offset, + upper_32_bits(crtc_base)); + WREG32(EVERGREEN_GRPH_PRIMARY_SURFACE_ADDRESS + amdgpu_crtc->crtc_offset, + (u32)crtc_base); + + /* post the write */ + RREG32(EVERGREEN_GRPH_PRIMARY_SURFACE_ADDRESS + amdgpu_crtc->crtc_offset); +} + +static int dce_v6_0_crtc_get_scanoutpos(struct amdgpu_device *adev, int crtc, + u32 *vbl, u32 *position) +{ + if ((crtc < 0) || (crtc >= adev->mode_info.num_crtc)) + return -EINVAL; + *vbl = RREG32(EVERGREEN_CRTC_V_BLANK_START_END + crtc_offsets[crtc]); + *position = RREG32(EVERGREEN_CRTC_STATUS_POSITION + crtc_offsets[crtc]); + + return 0; + +} + +/** + * dce_v6_0_hpd_sense - hpd sense callback. + * + * @adev: amdgpu_device pointer + * @hpd: hpd (hotplug detect) pin + * + * Checks if a digital monitor is connected (evergreen+). + * Returns true if connected, false if not connected. + */ +static bool dce_v6_0_hpd_sense(struct amdgpu_device *adev, + enum amdgpu_hpd_id hpd) +{ + bool connected = false; + + switch (hpd) { + case AMDGPU_HPD_1: + if (RREG32(DC_HPD1_INT_STATUS) & DC_HPDx_SENSE) + connected = true; + break; + case AMDGPU_HPD_2: + if (RREG32(DC_HPD2_INT_STATUS) & DC_HPDx_SENSE) + connected = true; + break; + case AMDGPU_HPD_3: + if (RREG32(DC_HPD3_INT_STATUS) & DC_HPDx_SENSE) + connected = true; + break; + case AMDGPU_HPD_4: + if (RREG32(DC_HPD4_INT_STATUS) & DC_HPDx_SENSE) + connected = true; + break; + case AMDGPU_HPD_5: + if (RREG32(DC_HPD5_INT_STATUS) & DC_HPDx_SENSE) + connected = true; + break; + case AMDGPU_HPD_6: + if (RREG32(DC_HPD6_INT_STATUS) & DC_HPDx_SENSE) + connected = true; + break; + default: + break; + } + + return connected; +} + +/** + * dce_v6_0_hpd_set_polarity - hpd set polarity callback. + * + * @adev: amdgpu_device pointer + * @hpd: hpd (hotplug detect) pin + * + * Set the polarity of the hpd pin (evergreen+). + */ +static void dce_v6_0_hpd_set_polarity(struct amdgpu_device *adev, + enum amdgpu_hpd_id hpd) +{ + u32 tmp; + bool connected = dce_v6_0_hpd_sense(adev, hpd); + + switch (hpd) { + case AMDGPU_HPD_1: + tmp = RREG32(DC_HPD1_INT_CONTROL); + if (connected) + tmp &= ~DC_HPDx_INT_POLARITY; + else + tmp |= DC_HPDx_INT_POLARITY; + WREG32(DC_HPD1_INT_CONTROL, tmp); + break; + case AMDGPU_HPD_2: + tmp = RREG32(DC_HPD2_INT_CONTROL); + if (connected) + tmp &= ~DC_HPDx_INT_POLARITY; + else + tmp |= DC_HPDx_INT_POLARITY; + WREG32(DC_HPD2_INT_CONTROL, tmp); + break; + case AMDGPU_HPD_3: + tmp = RREG32(DC_HPD3_INT_CONTROL); + if (connected) + tmp &= ~DC_HPDx_INT_POLARITY; + else + tmp |= DC_HPDx_INT_POLARITY; + WREG32(DC_HPD3_INT_CONTROL, tmp); + break; + case AMDGPU_HPD_4: + tmp = RREG32(DC_HPD4_INT_CONTROL); + if (connected) + tmp &= ~DC_HPDx_INT_POLARITY; + else + tmp |= DC_HPDx_INT_POLARITY; + WREG32(DC_HPD4_INT_CONTROL, tmp); + break; + case AMDGPU_HPD_5: + tmp = RREG32(DC_HPD5_INT_CONTROL); + if (connected) + tmp &= ~DC_HPDx_INT_POLARITY; + else + tmp |= DC_HPDx_INT_POLARITY; + WREG32(DC_HPD5_INT_CONTROL, tmp); + break; + case AMDGPU_HPD_6: + tmp = RREG32(DC_HPD6_INT_CONTROL); + if (connected) + tmp &= ~DC_HPDx_INT_POLARITY; + else + tmp |= DC_HPDx_INT_POLARITY; + WREG32(DC_HPD6_INT_CONTROL, tmp); + break; + default: + break; + } +} + +/** + * dce_v6_0_hpd_init - hpd setup callback. + * + * @adev: amdgpu_device pointer + * + * Setup the hpd pins used by the card (evergreen+). + * Enable the pin, set the polarity, and enable the hpd interrupts. + */ +static void dce_v6_0_hpd_init(struct amdgpu_device *adev) +{ + struct drm_device *dev = adev->ddev; + struct drm_connector *connector; + u32 tmp = DC_HPDx_CONNECTION_TIMER(0x9c4) | + DC_HPDx_RX_INT_TIMER(0xfa) | DC_HPDx_EN; + + list_for_each_entry(connector, &dev->mode_config.connector_list, head) { + struct amdgpu_connector *amdgpu_connector = to_amdgpu_connector(connector); + + if (connector->connector_type == DRM_MODE_CONNECTOR_eDP || + connector->connector_type == DRM_MODE_CONNECTOR_LVDS) { + /* don't try to enable hpd on eDP or LVDS avoid breaking the + * aux dp channel on imac and help (but not completely fix) + * https://bugzilla.redhat.com/show_bug.cgi?id=726143 + * also avoid interrupt storms during dpms. + */ + continue; + } + switch (amdgpu_connector->hpd.hpd) { + case AMDGPU_HPD_1: + WREG32(DC_HPD1_CONTROL, tmp); + break; + case AMDGPU_HPD_2: + WREG32(DC_HPD2_CONTROL, tmp); + break; + case AMDGPU_HPD_3: + WREG32(DC_HPD3_CONTROL, tmp); + break; + case AMDGPU_HPD_4: + WREG32(DC_HPD4_CONTROL, tmp); + break; + case AMDGPU_HPD_5: + WREG32(DC_HPD5_CONTROL, tmp); + break; + case AMDGPU_HPD_6: + WREG32(DC_HPD6_CONTROL, tmp); + break; + default: + break; + } + dce_v6_0_hpd_set_polarity(adev, amdgpu_connector->hpd.hpd); + amdgpu_irq_get(adev, &adev->hpd_irq, amdgpu_connector->hpd.hpd); + } + +} + +/** + * dce_v6_0_hpd_fini - hpd tear down callback. + * + * @adev: amdgpu_device pointer + * + * Tear down the hpd pins used by the card (evergreen+). + * Disable the hpd interrupts. + */ +static void dce_v6_0_hpd_fini(struct amdgpu_device *adev) +{ + struct drm_device *dev = adev->ddev; + struct drm_connector *connector; + + list_for_each_entry(connector, &dev->mode_config.connector_list, head) { + struct amdgpu_connector *amdgpu_connector = to_amdgpu_connector(connector); + + switch (amdgpu_connector->hpd.hpd) { + case AMDGPU_HPD_1: + WREG32(DC_HPD1_CONTROL, 0); + break; + case AMDGPU_HPD_2: + WREG32(DC_HPD2_CONTROL, 0); + break; + case AMDGPU_HPD_3: + WREG32(DC_HPD3_CONTROL, 0); + break; + case AMDGPU_HPD_4: + WREG32(DC_HPD4_CONTROL, 0); + break; + case AMDGPU_HPD_5: + WREG32(DC_HPD5_CONTROL, 0); + break; + case AMDGPU_HPD_6: + WREG32(DC_HPD6_CONTROL, 0); + break; + default: + break; + } + amdgpu_irq_put(adev, &adev->hpd_irq, amdgpu_connector->hpd.hpd); + } +} + +static u32 dce_v6_0_hpd_get_gpio_reg(struct amdgpu_device *adev) +{ + return SI_DC_GPIO_HPD_A; +} + +static bool dce_v6_0_is_display_hung(struct amdgpu_device *adev) +{ + DRM_INFO("xxxx: dce_v6_0_is_display_hung ----no imp!!!!!\n"); + + return true; +} + +static u32 evergreen_get_vblank_counter(struct amdgpu_device* adev, int crtc) +{ + if (crtc >= adev->mode_info.num_crtc) + return 0; + else + return RREG32(CRTC_STATUS_FRAME_COUNT + crtc_offsets[crtc]); +} + +static void dce_v6_0_stop_mc_access(struct amdgpu_device *adev, + struct amdgpu_mode_mc_save *save) +{ + u32 crtc_enabled, tmp, frame_count; + int i, j; + + save->vga_render_control = RREG32(VGA_RENDER_CONTROL); + save->vga_hdp_control = RREG32(VGA_HDP_CONTROL); + + /* disable VGA render */ + WREG32(VGA_RENDER_CONTROL, 0); + + /* blank the display controllers */ + for (i = 0; i < adev->mode_info.num_crtc; i++) { + crtc_enabled = RREG32(EVERGREEN_CRTC_CONTROL + crtc_offsets[i]) & EVERGREEN_CRTC_MASTER_EN; + if (crtc_enabled) { + save->crtc_enabled[i] = true; + tmp = RREG32(EVERGREEN_CRTC_BLANK_CONTROL + crtc_offsets[i]); + + if (!(tmp & EVERGREEN_CRTC_BLANK_DATA_EN)) { + dce_v6_0_vblank_wait(adev, i); + WREG32(EVERGREEN_CRTC_UPDATE_LOCK + crtc_offsets[i], 1); + tmp |= EVERGREEN_CRTC_BLANK_DATA_EN; + WREG32(EVERGREEN_CRTC_BLANK_CONTROL + crtc_offsets[i], tmp); + WREG32(EVERGREEN_CRTC_UPDATE_LOCK + crtc_offsets[i], 0); + } + /* wait for the next frame */ + frame_count = evergreen_get_vblank_counter(adev, i); + for (j = 0; j < adev->usec_timeout; j++) { + if (evergreen_get_vblank_counter(adev, i) != frame_count) + break; + udelay(1); + } + + /* XXX this is a hack to avoid strange behavior with EFI on certain systems */ + WREG32(EVERGREEN_CRTC_UPDATE_LOCK + crtc_offsets[i], 1); + tmp = RREG32(EVERGREEN_CRTC_CONTROL + crtc_offsets[i]); + tmp &= ~EVERGREEN_CRTC_MASTER_EN; + WREG32(EVERGREEN_CRTC_CONTROL + crtc_offsets[i], tmp); + WREG32(EVERGREEN_CRTC_UPDATE_LOCK + crtc_offsets[i], 0); + save->crtc_enabled[i] = false; + /* ***** */ + } else { + save->crtc_enabled[i] = false; + } + } +} + +static void dce_v6_0_resume_mc_access(struct amdgpu_device *adev, + struct amdgpu_mode_mc_save *save) +{ + u32 tmp; + int i, j; + + /* update crtc base addresses */ + for (i = 0; i < adev->mode_info.num_crtc; i++) { + WREG32(EVERGREEN_GRPH_PRIMARY_SURFACE_ADDRESS_HIGH + crtc_offsets[i], + upper_32_bits(adev->mc.vram_start)); + WREG32(EVERGREEN_GRPH_SECONDARY_SURFACE_ADDRESS_HIGH + crtc_offsets[i], + upper_32_bits(adev->mc.vram_start)); + WREG32(EVERGREEN_GRPH_PRIMARY_SURFACE_ADDRESS + crtc_offsets[i], + (u32)adev->mc.vram_start); + WREG32(EVERGREEN_GRPH_SECONDARY_SURFACE_ADDRESS + crtc_offsets[i], + (u32)adev->mc.vram_start); + } + + WREG32(EVERGREEN_VGA_MEMORY_BASE_ADDRESS_HIGH, upper_32_bits(adev->mc.vram_start)); + WREG32(EVERGREEN_VGA_MEMORY_BASE_ADDRESS, (u32)adev->mc.vram_start); + + /* unlock regs and wait for update */ + for (i = 0; i < adev->mode_info.num_crtc; i++) { + if (save->crtc_enabled[i]) { + tmp = RREG32(EVERGREEN_MASTER_UPDATE_MODE + crtc_offsets[i]); + if ((tmp & 0x7) != 3) { + tmp &= ~0x7; + tmp |= 0x3; + WREG32(EVERGREEN_MASTER_UPDATE_MODE + crtc_offsets[i], tmp); + } + tmp = RREG32(EVERGREEN_GRPH_UPDATE + crtc_offsets[i]); + if (tmp & EVERGREEN_GRPH_UPDATE_LOCK) { + tmp &= ~EVERGREEN_GRPH_UPDATE_LOCK; + WREG32(EVERGREEN_GRPH_UPDATE + crtc_offsets[i], tmp); + } + tmp = RREG32(EVERGREEN_MASTER_UPDATE_LOCK + crtc_offsets[i]); + if (tmp & 1) { + tmp &= ~1; + WREG32(EVERGREEN_MASTER_UPDATE_LOCK + crtc_offsets[i], tmp); + } + for (j = 0; j < adev->usec_timeout; j++) { + tmp = RREG32(EVERGREEN_GRPH_UPDATE + crtc_offsets[i]); + if ((tmp & EVERGREEN_GRPH_SURFACE_UPDATE_PENDING) == 0) + break; + udelay(1); + } + } + } + + /* Unlock vga access */ + WREG32(VGA_HDP_CONTROL, save->vga_hdp_control); + mdelay(1); + WREG32(VGA_RENDER_CONTROL, save->vga_render_control); + +} + +static void dce_v6_0_set_vga_render_state(struct amdgpu_device *adev, + bool render) +{ + if (!render) + WREG32(R_000300_VGA_RENDER_CONTROL, + RREG32(R_000300_VGA_RENDER_CONTROL) & C_000300_VGA_VSTATUS_CNTL); + +} + +static void dce_v6_0_program_fmt(struct drm_encoder *encoder) +{ + + struct drm_device *dev = encoder->dev; + struct amdgpu_device *adev = dev->dev_private; + struct amdgpu_encoder *amdgpu_encoder = to_amdgpu_encoder(encoder); + struct drm_connector *connector = amdgpu_get_connector_for_encoder(encoder); + struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(encoder->crtc); + int bpc = 0; + u32 tmp = 0; + enum amdgpu_connector_dither dither = AMDGPU_FMT_DITHER_DISABLE; + + if (connector) { + struct amdgpu_connector *amdgpu_connector = to_amdgpu_connector(connector); + bpc = amdgpu_connector_get_monitor_bpc(connector); + dither = amdgpu_connector->dither; + } + + /* LVDS FMT is set up by atom */ + if (amdgpu_encoder->devices & ATOM_DEVICE_LCD_SUPPORT) + return; + + if (bpc == 0) + return; + + + switch (bpc) { + case 6: + if (dither == AMDGPU_FMT_DITHER_ENABLE) + /* XXX sort out optimal dither settings */ + tmp |= (FMT_FRAME_RANDOM_ENABLE | FMT_HIGHPASS_RANDOM_ENABLE | + FMT_SPATIAL_DITHER_EN); + else + tmp |= FMT_TRUNCATE_EN; + break; + case 8: + if (dither == AMDGPU_FMT_DITHER_ENABLE) + /* XXX sort out optimal dither settings */ + tmp |= (FMT_FRAME_RANDOM_ENABLE | FMT_HIGHPASS_RANDOM_ENABLE | + FMT_RGB_RANDOM_ENABLE | + FMT_SPATIAL_DITHER_EN | FMT_SPATIAL_DITHER_DEPTH); + else + tmp |= (FMT_TRUNCATE_EN | FMT_TRUNCATE_DEPTH); + break; + case 10: + default: + /* not needed */ + break; + } + + WREG32(FMT_BIT_DEPTH_CONTROL + amdgpu_crtc->crtc_offset, tmp); +} + +/** + * cik_get_number_of_dram_channels - get the number of dram channels + * + * @adev: amdgpu_device pointer + * + * Look up the number of video ram channels (CIK). + * Used for display watermark bandwidth calculations + * Returns the number of dram channels + */ +static u32 si_get_number_of_dram_channels(struct amdgpu_device *adev) +{ + u32 tmp = RREG32(MC_SHARED_CHMAP); + + switch ((tmp & MC_SHARED_CHMAP__NOOFCHAN_MASK) >> MC_SHARED_CHMAP__NOOFCHAN__SHIFT) { + case 0: + default: + return 1; + case 1: + return 2; + case 2: + return 4; + case 3: + return 8; + case 4: + return 3; + case 5: + return 6; + case 6: + return 10; + case 7: + return 12; + case 8: + return 16; + } +} + +struct dce6_wm_params { + u32 dram_channels; /* number of dram channels */ + u32 yclk; /* bandwidth per dram data pin in kHz */ + u32 sclk; /* engine clock in kHz */ + u32 disp_clk; /* display clock in kHz */ + u32 src_width; /* viewport width */ + u32 active_time; /* active display time in ns */ + u32 blank_time; /* blank time in ns */ + bool interlaced; /* mode is interlaced */ + fixed20_12 vsc; /* vertical scale ratio */ + u32 num_heads; /* number of active crtcs */ + u32 bytes_per_pixel; /* bytes per pixel display + overlay */ + u32 lb_size; /* line buffer allocated to pipe */ + u32 vtaps; /* vertical scaler taps */ +}; + +/** + * dce_v6_0_dram_bandwidth - get the dram bandwidth + * + * @wm: watermark calculation data + * + * Calculate the raw dram bandwidth (CIK). + * Used for display watermark bandwidth calculations + * Returns the dram bandwidth in MBytes/s + */ +static u32 dce_v6_0_dram_bandwidth(struct dce6_wm_params *wm) +{ + /* Calculate raw DRAM Bandwidth */ + fixed20_12 dram_efficiency; /* 0.7 */ + fixed20_12 yclk, dram_channels, bandwidth; + fixed20_12 a; + + a.full = dfixed_const(1000); + yclk.full = dfixed_const(wm->yclk); + yclk.full = dfixed_div(yclk, a); + dram_channels.full = dfixed_const(wm->dram_channels * 4); + a.full = dfixed_const(10); + dram_efficiency.full = dfixed_const(7); + dram_efficiency.full = dfixed_div(dram_efficiency, a); + bandwidth.full = dfixed_mul(dram_channels, yclk); + bandwidth.full = dfixed_mul(bandwidth, dram_efficiency); + + return dfixed_trunc(bandwidth); +} + +/** + * dce_v6_0_dram_bandwidth_for_display - get the dram bandwidth for display + * + * @wm: watermark calculation data + * + * Calculate the dram bandwidth used for display (CIK). + * Used for display watermark bandwidth calculations + * Returns the dram bandwidth for display in MBytes/s + */ +static u32 dce_v6_0_dram_bandwidth_for_display(struct dce6_wm_params *wm) +{ + /* Calculate DRAM Bandwidth and the part allocated to display. */ + fixed20_12 disp_dram_allocation; /* 0.3 to 0.7 */ + fixed20_12 yclk, dram_channels, bandwidth; + fixed20_12 a; + + a.full = dfixed_const(1000); + yclk.full = dfixed_const(wm->yclk); + yclk.full = dfixed_div(yclk, a); + dram_channels.full = dfixed_const(wm->dram_channels * 4); + a.full = dfixed_const(10); + disp_dram_allocation.full = dfixed_const(3); /* XXX worse case value 0.3 */ + disp_dram_allocation.full = dfixed_div(disp_dram_allocation, a); + bandwidth.full = dfixed_mul(dram_channels, yclk); + bandwidth.full = dfixed_mul(bandwidth, disp_dram_allocation); + + return dfixed_trunc(bandwidth); +} + +/** + * dce_v6_0_data_return_bandwidth - get the data return bandwidth + * + * @wm: watermark calculation data + * + * Calculate the data return bandwidth used for display (CIK). + * Used for display watermark bandwidth calculations + * Returns the data return bandwidth in MBytes/s + */ +static u32 dce_v6_0_data_return_bandwidth(struct dce6_wm_params *wm) +{ + /* Calculate the display Data return Bandwidth */ + fixed20_12 return_efficiency; /* 0.8 */ + fixed20_12 sclk, bandwidth; + fixed20_12 a; + + a.full = dfixed_const(1000); + sclk.full = dfixed_const(wm->sclk); + sclk.full = dfixed_div(sclk, a); + a.full = dfixed_const(10); + return_efficiency.full = dfixed_const(8); + return_efficiency.full = dfixed_div(return_efficiency, a); + a.full = dfixed_const(32); + bandwidth.full = dfixed_mul(a, sclk); + bandwidth.full = dfixed_mul(bandwidth, return_efficiency); + + return dfixed_trunc(bandwidth); +} + +/** + * dce_v6_0_dmif_request_bandwidth - get the dmif bandwidth + * + * @wm: watermark calculation data + * + * Calculate the dmif bandwidth used for display (CIK). + * Used for display watermark bandwidth calculations + * Returns the dmif bandwidth in MBytes/s + */ +static u32 dce_v6_0_dmif_request_bandwidth(struct dce6_wm_params *wm) +{ + /* Calculate the DMIF Request Bandwidth */ + fixed20_12 disp_clk_request_efficiency; /* 0.8 */ + fixed20_12 disp_clk, bandwidth; + fixed20_12 a, b; + + a.full = dfixed_const(1000); + disp_clk.full = dfixed_const(wm->disp_clk); + disp_clk.full = dfixed_div(disp_clk, a); + a.full = dfixed_const(32); + b.full = dfixed_mul(a, disp_clk); + + a.full = dfixed_const(10); + disp_clk_request_efficiency.full = dfixed_const(8); + disp_clk_request_efficiency.full = dfixed_div(disp_clk_request_efficiency, a); + + bandwidth.full = dfixed_mul(b, disp_clk_request_efficiency); + + return dfixed_trunc(bandwidth); +} + +/** + * dce_v6_0_available_bandwidth - get the min available bandwidth + * + * @wm: watermark calculation data + * + * Calculate the min available bandwidth used for display (CIK). + * Used for display watermark bandwidth calculations + * Returns the min available bandwidth in MBytes/s + */ +static u32 dce_v6_0_available_bandwidth(struct dce6_wm_params *wm) +{ + /* Calculate the Available bandwidth. Display can use this temporarily but not in average. */ + u32 dram_bandwidth = dce_v6_0_dram_bandwidth(wm); + u32 data_return_bandwidth = dce_v6_0_data_return_bandwidth(wm); + u32 dmif_req_bandwidth = dce_v6_0_dmif_request_bandwidth(wm); + + return min(dram_bandwidth, min(data_return_bandwidth, dmif_req_bandwidth)); +} + +/** + * dce_v6_0_average_bandwidth - get the average available bandwidth + * + * @wm: watermark calculation data + * + * Calculate the average available bandwidth used for display (CIK). + * Used for display watermark bandwidth calculations + * Returns the average available bandwidth in MBytes/s + */ +static u32 dce_v6_0_average_bandwidth(struct dce6_wm_params *wm) +{ + /* Calculate the display mode Average Bandwidth + * DisplayMode should contain the source and destination dimensions, + * timing, etc. + */ + fixed20_12 bpp; + fixed20_12 line_time; + fixed20_12 src_width; + fixed20_12 bandwidth; + fixed20_12 a; + + a.full = dfixed_const(1000); + line_time.full = dfixed_const(wm->active_time + wm->blank_time); + line_time.full = dfixed_div(line_time, a); + bpp.full = dfixed_const(wm->bytes_per_pixel); + src_width.full = dfixed_const(wm->src_width); + bandwidth.full = dfixed_mul(src_width, bpp); + bandwidth.full = dfixed_mul(bandwidth, wm->vsc); + bandwidth.full = dfixed_div(bandwidth, line_time); + + return dfixed_trunc(bandwidth); +} + +/** + * dce_v6_0_latency_watermark - get the latency watermark + * + * @wm: watermark calculation data + * + * Calculate the latency watermark (CIK). + * Used for display watermark bandwidth calculations + * Returns the latency watermark in ns + */ +static u32 dce_v6_0_latency_watermark(struct dce6_wm_params *wm) +{ + /* First calculate the latency in ns */ + u32 mc_latency = 2000; /* 2000 ns. */ + u32 available_bandwidth = dce_v6_0_available_bandwidth(wm); + u32 worst_chunk_return_time = (512 * 8 * 1000) / available_bandwidth; + u32 cursor_line_pair_return_time = (128 * 4 * 1000) / available_bandwidth; + u32 dc_latency = 40000000 / wm->disp_clk; /* dc pipe latency */ + u32 other_heads_data_return_time = ((wm->num_heads + 1) * worst_chunk_return_time) + + (wm->num_heads * cursor_line_pair_return_time); + u32 latency = mc_latency + other_heads_data_return_time + dc_latency; + u32 max_src_lines_per_dst_line, lb_fill_bw, line_fill_time; + u32 tmp, dmif_size = 12288; + fixed20_12 a, b, c; + + if (wm->num_heads == 0) + return 0; + + a.full = dfixed_const(2); + b.full = dfixed_const(1); + if ((wm->vsc.full > a.full) || + ((wm->vsc.full > b.full) && (wm->vtaps >= 3)) || + (wm->vtaps >= 5) || + ((wm->vsc.full >= a.full) && wm->interlaced)) + max_src_lines_per_dst_line = 4; + else + max_src_lines_per_dst_line = 2; + + a.full = dfixed_const(available_bandwidth); + b.full = dfixed_const(wm->num_heads); + a.full = dfixed_div(a, b); + + b.full = dfixed_const(mc_latency + 512); + c.full = dfixed_const(wm->disp_clk); + b.full = dfixed_div(b, c); + + c.full = dfixed_const(dmif_size); + b.full = dfixed_div(c, b); + + tmp = min(dfixed_trunc(a), dfixed_trunc(b)); + + b.full = dfixed_const(1000); + c.full = dfixed_const(wm->disp_clk); + b.full = dfixed_div(c, b); + c.full = dfixed_const(wm->bytes_per_pixel); + b.full = dfixed_mul(b, c); + + lb_fill_bw = min(tmp, dfixed_trunc(b)); + + a.full = dfixed_const(max_src_lines_per_dst_line * wm->src_width * wm->bytes_per_pixel); + b.full = dfixed_const(1000); + c.full = dfixed_const(lb_fill_bw); + b.full = dfixed_div(c, b); + a.full = dfixed_div(a, b); + line_fill_time = dfixed_trunc(a); + + if (line_fill_time < wm->active_time) + return latency; + else + return latency + (line_fill_time - wm->active_time); + +} + +/** + * dce_v6_0_average_bandwidth_vs_dram_bandwidth_for_display - check + * average and available dram bandwidth + * + * @wm: watermark calculation data + * + * Check if the display average bandwidth fits in the display + * dram bandwidth (CIK). + * Used for display watermark bandwidth calculations + * Returns true if the display fits, false if not. + */ +static bool dce_v6_0_average_bandwidth_vs_dram_bandwidth_for_display(struct dce6_wm_params *wm) +{ + if (dce_v6_0_average_bandwidth(wm) <= + (dce_v6_0_dram_bandwidth_for_display(wm) / wm->num_heads)) + return true; + else + return false; +} + +/** + * dce_v6_0_average_bandwidth_vs_available_bandwidth - check + * average and available bandwidth + * + * @wm: watermark calculation data + * + * Check if the display average bandwidth fits in the display + * available bandwidth (CIK). + * Used for display watermark bandwidth calculations + * Returns true if the display fits, false if not. + */ +static bool dce_v6_0_average_bandwidth_vs_available_bandwidth(struct dce6_wm_params *wm) +{ + if (dce_v6_0_average_bandwidth(wm) <= + (dce_v6_0_available_bandwidth(wm) / wm->num_heads)) + return true; + else + return false; +} + +/** + * dce_v6_0_check_latency_hiding - check latency hiding + * + * @wm: watermark calculation data + * + * Check latency hiding (CIK). + * Used for display watermark bandwidth calculations + * Returns true if the display fits, false if not. + */ +static bool dce_v6_0_check_latency_hiding(struct dce6_wm_params *wm) +{ + u32 lb_partitions = wm->lb_size / wm->src_width; + u32 line_time = wm->active_time + wm->blank_time; + u32 latency_tolerant_lines; + u32 latency_hiding; + fixed20_12 a; + + a.full = dfixed_const(1); + if (wm->vsc.full > a.full) + latency_tolerant_lines = 1; + else { + if (lb_partitions <= (wm->vtaps + 1)) + latency_tolerant_lines = 1; + else + latency_tolerant_lines = 2; + } + + latency_hiding = (latency_tolerant_lines * line_time + wm->blank_time); + + if (dce_v6_0_latency_watermark(wm) <= latency_hiding) + return true; + else + return false; +} + +/** + * dce_v6_0_program_watermarks - program display watermarks + * + * @adev: amdgpu_device pointer + * @amdgpu_crtc: the selected display controller + * @lb_size: line buffer size + * @num_heads: number of display controllers in use + * + * Calculate and program the display watermarks for the + * selected display controller (CIK). + */ +static void dce_v6_0_program_watermarks(struct amdgpu_device *adev, + struct amdgpu_crtc *amdgpu_crtc, + u32 lb_size, u32 num_heads) +{ + struct drm_display_mode *mode = &amdgpu_crtc->base.mode; + struct dce6_wm_params wm_low, wm_high; + u32 dram_channels; + u32 pixel_period; + u32 line_time = 0; + u32 latency_watermark_a = 0, latency_watermark_b = 0; + u32 priority_a_mark = 0, priority_b_mark = 0; + u32 priority_a_cnt = PRIORITY_OFF; + u32 priority_b_cnt = PRIORITY_OFF; + u32 tmp, arb_control3; + fixed20_12 a, b, c; + + if (amdgpu_crtc->base.enabled && num_heads && mode) { + pixel_period = 1000000 / (u32)mode->clock; + line_time = min((u32)mode->crtc_htotal * pixel_period, (u32)65535); + priority_a_cnt = 0; + priority_b_cnt = 0; + + dram_channels = si_get_number_of_dram_channels(adev); + + /* watermark for high clocks */ + if (adev->pm.dpm_enabled) { + wm_high.yclk = + amdgpu_dpm_get_mclk(adev, false) * 10; + wm_high.sclk = + amdgpu_dpm_get_sclk(adev, false) * 10; + } else { + wm_high.yclk = adev->pm.current_mclk * 10; + wm_high.sclk = adev->pm.current_sclk * 10; + } + + wm_high.disp_clk = mode->clock; + wm_high.src_width = mode->crtc_hdisplay; + wm_high.active_time = mode->crtc_hdisplay * pixel_period; + wm_high.blank_time = line_time - wm_high.active_time; + wm_high.interlaced = false; + if (mode->flags & DRM_MODE_FLAG_INTERLACE) + wm_high.interlaced = true; + wm_high.vsc = amdgpu_crtc->vsc; + wm_high.vtaps = 1; + if (amdgpu_crtc->rmx_type != RMX_OFF) + wm_high.vtaps = 2; + wm_high.bytes_per_pixel = 4; /* XXX: get this from fb config */ + wm_high.lb_size = lb_size; + wm_high.dram_channels = dram_channels; + wm_high.num_heads = num_heads; + + if (adev->pm.dpm_enabled) { + /* watermark for low clocks */ + wm_low.yclk = + amdgpu_dpm_get_mclk(adev, true) * 10; + wm_low.sclk = + amdgpu_dpm_get_sclk(adev, true) * 10; + } else { + wm_low.yclk = adev->pm.current_mclk * 10; + wm_low.sclk = adev->pm.current_sclk * 10; + } + + wm_low.disp_clk = mode->clock; + wm_low.src_width = mode->crtc_hdisplay; + wm_low.active_time = mode->crtc_hdisplay * pixel_period; + wm_low.blank_time = line_time - wm_low.active_time; + wm_low.interlaced = false; + if (mode->flags & DRM_MODE_FLAG_INTERLACE) + wm_low.interlaced = true; + wm_low.vsc = amdgpu_crtc->vsc; + wm_low.vtaps = 1; + if (amdgpu_crtc->rmx_type != RMX_OFF) + wm_low.vtaps = 2; + wm_low.bytes_per_pixel = 4; /* XXX: get this from fb config */ + wm_low.lb_size = lb_size; + wm_low.dram_channels = dram_channels; + wm_low.num_heads = num_heads; + + /* set for high clocks */ + latency_watermark_a = min(dce_v6_0_latency_watermark(&wm_high), (u32)65535); + /* set for low clocks */ + latency_watermark_b = min(dce_v6_0_latency_watermark(&wm_low), (u32)65535); + + /* possibly force display priority to high */ + /* should really do this at mode validation time... */ + if (!dce_v6_0_average_bandwidth_vs_dram_bandwidth_for_display(&wm_high) || + !dce_v6_0_average_bandwidth_vs_available_bandwidth(&wm_high) || + !dce_v6_0_check_latency_hiding(&wm_high) || + (adev->mode_info.disp_priority == 2)) { + DRM_DEBUG_KMS("force priority to high\n"); + priority_a_cnt |= PRIORITY_ALWAYS_ON; + priority_b_cnt |= PRIORITY_ALWAYS_ON; + } + if (!dce_v6_0_average_bandwidth_vs_dram_bandwidth_for_display(&wm_low) || + !dce_v6_0_average_bandwidth_vs_available_bandwidth(&wm_low) || + !dce_v6_0_check_latency_hiding(&wm_low) || + (adev->mode_info.disp_priority == 2)) { + DRM_DEBUG_KMS("force priority to high\n"); + priority_a_cnt |= PRIORITY_ALWAYS_ON; + priority_b_cnt |= PRIORITY_ALWAYS_ON; + } + + a.full = dfixed_const(1000); + b.full = dfixed_const(mode->clock); + b.full = dfixed_div(b, a); + c.full = dfixed_const(latency_watermark_a); + c.full = dfixed_mul(c, b); + c.full = dfixed_mul(c, amdgpu_crtc->hsc); + c.full = dfixed_div(c, a); + a.full = dfixed_const(16); + c.full = dfixed_div(c, a); + priority_a_mark = dfixed_trunc(c); + priority_a_cnt |= priority_a_mark & PRIORITY_MARK_MASK; + + a.full = dfixed_const(1000); + b.full = dfixed_const(mode->clock); + b.full = dfixed_div(b, a); + c.full = dfixed_const(latency_watermark_b); + c.full = dfixed_mul(c, b); + c.full = dfixed_mul(c, amdgpu_crtc->hsc); + c.full = dfixed_div(c, a); + a.full = dfixed_const(16); + c.full = dfixed_div(c, a); + priority_b_mark = dfixed_trunc(c); + priority_b_cnt |= priority_b_mark & PRIORITY_MARK_MASK; + } + + /* select wm A */ + arb_control3 = RREG32(DPG_PIPE_ARBITRATION_CONTROL3 + amdgpu_crtc->crtc_offset); + tmp = arb_control3; + tmp &= ~LATENCY_WATERMARK_MASK(3); + tmp |= LATENCY_WATERMARK_MASK(1); + WREG32(DPG_PIPE_ARBITRATION_CONTROL3 + amdgpu_crtc->crtc_offset, tmp); + WREG32(DPG_PIPE_LATENCY_CONTROL + amdgpu_crtc->crtc_offset, + (LATENCY_LOW_WATERMARK(latency_watermark_a) | + LATENCY_HIGH_WATERMARK(line_time))); + /* select wm B */ + tmp = RREG32(DPG_PIPE_ARBITRATION_CONTROL3 + amdgpu_crtc->crtc_offset); + tmp &= ~LATENCY_WATERMARK_MASK(3); + tmp |= LATENCY_WATERMARK_MASK(2); + WREG32(DPG_PIPE_ARBITRATION_CONTROL3 + amdgpu_crtc->crtc_offset, tmp); + WREG32(DPG_PIPE_LATENCY_CONTROL + amdgpu_crtc->crtc_offset, + (LATENCY_LOW_WATERMARK(latency_watermark_b) | + LATENCY_HIGH_WATERMARK(line_time))); + /* restore original selection */ + WREG32(DPG_PIPE_ARBITRATION_CONTROL3 + amdgpu_crtc->crtc_offset, arb_control3); + + /* write the priority marks */ + WREG32(PRIORITY_A_CNT + amdgpu_crtc->crtc_offset, priority_a_cnt); + WREG32(PRIORITY_B_CNT + amdgpu_crtc->crtc_offset, priority_b_cnt); + + /* save values for DPM */ + amdgpu_crtc->line_time = line_time; + amdgpu_crtc->wm_high = latency_watermark_a; +} + +/* watermark setup */ +static u32 dce_v6_0_line_buffer_adjust(struct amdgpu_device *adev, + struct amdgpu_crtc *amdgpu_crtc, + struct drm_display_mode *mode, + struct drm_display_mode *other_mode) +{ + u32 tmp, buffer_alloc, i; + u32 pipe_offset = amdgpu_crtc->crtc_id * 0x8; + /* + * Line Buffer Setup + * There are 3 line buffers, each one shared by 2 display controllers. + * DC_LB_MEMORY_SPLIT controls how that line buffer is shared between + * the display controllers. The paritioning is done via one of four + * preset allocations specified in bits 21:20: + * 0 - half lb + * 2 - whole lb, other crtc must be disabled + */ + /* this can get tricky if we have two large displays on a paired group + * of crtcs. Ideally for multiple large displays we'd assign them to + * non-linked crtcs for maximum line buffer allocation. + */ + if (amdgpu_crtc->base.enabled && mode) { + if (other_mode) { + tmp = 0; /* 1/2 */ + buffer_alloc = 1; + } else { + tmp = 2; /* whole */ + buffer_alloc = 2; + } + } else { + tmp = 0; + buffer_alloc = 0; + } + + WREG32(DC_LB_MEMORY_SPLIT + amdgpu_crtc->crtc_offset, + DC_LB_MEMORY_CONFIG(tmp)); + + WREG32(PIPE0_DMIF_BUFFER_CONTROL + pipe_offset, + DMIF_BUFFERS_ALLOCATED(buffer_alloc)); + for (i = 0; i < adev->usec_timeout; i++) { + if (RREG32(PIPE0_DMIF_BUFFER_CONTROL + pipe_offset) & + DMIF_BUFFERS_ALLOCATED_COMPLETED) + break; + udelay(1); + } + + if (amdgpu_crtc->base.enabled && mode) { + switch (tmp) { + case 0: + default: + return 4096 * 2; + case 2: + return 8192 * 2; + } + } + + /* controller not enabled, so no lb used */ + return 0; +} + + +/** + * + * dce_v6_0_bandwidth_update - program display watermarks + * + * @adev: amdgpu_device pointer + * + * Calculate and program the display watermarks and line + * buffer allocation (CIK). + */ +static void dce_v6_0_bandwidth_update(struct amdgpu_device *adev) +{ + struct drm_display_mode *mode0 = NULL; + struct drm_display_mode *mode1 = NULL; + u32 num_heads = 0, lb_size; + int i; + + if (!adev->mode_info.mode_config_initialized) + return; + + amdgpu_update_display_priority(adev); + + for (i = 0; i < adev->mode_info.num_crtc; i++) { + if (adev->mode_info.crtcs[i]->base.enabled) + num_heads++; + } + for (i = 0; i < adev->mode_info.num_crtc; i += 2) { + mode0 = &adev->mode_info.crtcs[i]->base.mode; + mode1 = &adev->mode_info.crtcs[i+1]->base.mode; + lb_size = dce_v6_0_line_buffer_adjust(adev, adev->mode_info.crtcs[i], mode0, mode1); + dce_v6_0_program_watermarks(adev, adev->mode_info.crtcs[i], lb_size, num_heads); + lb_size = dce_v6_0_line_buffer_adjust(adev, adev->mode_info.crtcs[i+1], mode1, mode0); + dce_v6_0_program_watermarks(adev, adev->mode_info.crtcs[i+1], lb_size, num_heads); + } +} +/* +static void dce_v6_0_audio_get_connected_pins(struct amdgpu_device *adev) +{ + int i; + u32 offset, tmp; + + for (i = 0; i < adev->mode_info.audio.num_pins; i++) { + offset = adev->mode_info.audio.pin[i].offset; + tmp = RREG32_AUDIO_ENDPT(offset, + AZ_F0_CODEC_PIN_CONTROL_RESPONSE_CONFIGURATION_DEFAULT); + if (((tmp & PORT_CONNECTIVITY_MASK) >> PORT_CONNECTIVITY_SHIFT) == 1) + adev->mode_info.audio.pin[i].connected = false; + else + adev->mode_info.audio.pin[i].connected = true; + } + +} + +static struct amdgpu_audio_pin *dce_v6_0_audio_get_pin(struct amdgpu_device *adev) +{ + int i; + + dce_v6_0_audio_get_connected_pins(adev); + + for (i = 0; i < adev->mode_info.audio.num_pins; i++) { + if (adev->mode_info.audio.pin[i].connected) + return &adev->mode_info.audio.pin[i]; + } + DRM_ERROR("No connected audio pins found!\n"); + return NULL; +} + +static void dce_v6_0_afmt_audio_select_pin(struct drm_encoder *encoder) +{ + struct amdgpu_device *adev = encoder->dev->dev_private; + struct amdgpu_encoder *amdgpu_encoder = to_amdgpu_encoder(encoder); + struct amdgpu_encoder_atom_dig *dig = amdgpu_encoder->enc_priv; + u32 offset; + + if (!dig || !dig->afmt || !dig->afmt->pin) + return; + + offset = dig->afmt->offset; + + WREG32(AFMT_AUDIO_SRC_CONTROL + offset, + AFMT_AUDIO_SRC_SELECT(dig->afmt->pin->id)); + +} + +static void dce_v6_0_audio_write_latency_fields(struct drm_encoder *encoder, + struct drm_display_mode *mode) +{ + DRM_INFO("xxxx: dce_v6_0_audio_write_latency_fields---no imp!!!!!\n"); +} + +static void dce_v6_0_audio_write_speaker_allocation(struct drm_encoder *encoder) +{ + DRM_INFO("xxxx: dce_v6_0_audio_write_speaker_allocation---no imp!!!!!\n"); +} + +static void dce_v6_0_audio_write_sad_regs(struct drm_encoder *encoder) +{ + DRM_INFO("xxxx: dce_v6_0_audio_write_sad_regs---no imp!!!!!\n"); + +} +*/ +static void dce_v6_0_audio_enable(struct amdgpu_device *adev, + struct amdgpu_audio_pin *pin, + bool enable) +{ + DRM_INFO("xxxx: dce_v6_0_audio_enable---no imp!!!!!\n"); +} + +static const u32 pin_offsets[7] = +{ + (0x1780 - 0x1780), + (0x1786 - 0x1780), + (0x178c - 0x1780), + (0x1792 - 0x1780), + (0x1798 - 0x1780), + (0x179d - 0x1780), + (0x17a4 - 0x1780), +}; + +static int dce_v6_0_audio_init(struct amdgpu_device *adev) +{ + return 0; +} + +static void dce_v6_0_audio_fini(struct amdgpu_device *adev) +{ + +} + +/* +static void dce_v6_0_afmt_update_ACR(struct drm_encoder *encoder, uint32_t clock) +{ + DRM_INFO("xxxx: dce_v6_0_afmt_update_ACR---no imp!!!!!\n"); +} +*/ +/* + * build a HDMI Video Info Frame + */ +/* +static void dce_v6_0_afmt_update_avi_infoframe(struct drm_encoder *encoder, + void *buffer, size_t size) +{ + DRM_INFO("xxxx: dce_v6_0_afmt_update_avi_infoframe---no imp!!!!!\n"); +} + +static void dce_v6_0_audio_set_dto(struct drm_encoder *encoder, u32 clock) +{ + DRM_INFO("xxxx: dce_v6_0_audio_set_dto---no imp!!!!!\n"); +} +*/ +/* + * update the info frames with the data from the current display mode + */ +static void dce_v6_0_afmt_setmode(struct drm_encoder *encoder, + struct drm_display_mode *mode) +{ + DRM_INFO("xxxx: dce_v6_0_afmt_setmode ----no impl !!!!!!!!\n"); +} + +static void dce_v6_0_afmt_enable(struct drm_encoder *encoder, bool enable) +{ + struct drm_device *dev = encoder->dev; + struct amdgpu_device *adev = dev->dev_private; + struct amdgpu_encoder *amdgpu_encoder = to_amdgpu_encoder(encoder); + struct amdgpu_encoder_atom_dig *dig = amdgpu_encoder->enc_priv; + + if (!dig || !dig->afmt) + return; + + /* Silent, r600_hdmi_enable will raise WARN for us */ + if (enable && dig->afmt->enabled) + return; + if (!enable && !dig->afmt->enabled) + return; + + if (!enable && dig->afmt->pin) { + dce_v6_0_audio_enable(adev, dig->afmt->pin, false); + dig->afmt->pin = NULL; + } + + dig->afmt->enabled = enable; + + DRM_DEBUG("%sabling AFMT interface @ 0x%04X for encoder 0x%x\n", + enable ? "En" : "Dis", dig->afmt->offset, amdgpu_encoder->encoder_id); +} + +static void dce_v6_0_afmt_init(struct amdgpu_device *adev) +{ + int i; + + for (i = 0; i < adev->mode_info.num_dig; i++) + adev->mode_info.afmt[i] = NULL; + + /* DCE8 has audio blocks tied to DIG encoders */ + for (i = 0; i < adev->mode_info.num_dig; i++) { + adev->mode_info.afmt[i] = kzalloc(sizeof(struct amdgpu_afmt), GFP_KERNEL); + if (adev->mode_info.afmt[i]) { + adev->mode_info.afmt[i]->offset = dig_offsets[i]; + adev->mode_info.afmt[i]->id = i; + } + } +} + +static void dce_v6_0_afmt_fini(struct amdgpu_device *adev) +{ + int i; + + for (i = 0; i < adev->mode_info.num_dig; i++) { + kfree(adev->mode_info.afmt[i]); + adev->mode_info.afmt[i] = NULL; + } +} + +static const u32 vga_control_regs[6] = +{ + AVIVO_D1VGA_CONTROL, + AVIVO_D2VGA_CONTROL, + EVERGREEN_D3VGA_CONTROL, + EVERGREEN_D4VGA_CONTROL, + EVERGREEN_D5VGA_CONTROL, + EVERGREEN_D6VGA_CONTROL, +}; + +static void dce_v6_0_vga_enable(struct drm_crtc *crtc, bool enable) +{ + struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc); + struct drm_device *dev = crtc->dev; + struct amdgpu_device *adev = dev->dev_private; + u32 vga_control; + + vga_control = RREG32(vga_control_regs[amdgpu_crtc->crtc_id]) & ~1; + if (enable) + WREG32(vga_control_regs[amdgpu_crtc->crtc_id], vga_control | 1); + else + WREG32(vga_control_regs[amdgpu_crtc->crtc_id], vga_control); +} + +static void dce_v6_0_grph_enable(struct drm_crtc *crtc, bool enable) +{ + struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc); + struct drm_device *dev = crtc->dev; + struct amdgpu_device *adev = dev->dev_private; + + if (enable) + WREG32(EVERGREEN_GRPH_ENABLE + amdgpu_crtc->crtc_offset, 1); + else + WREG32(EVERGREEN_GRPH_ENABLE + amdgpu_crtc->crtc_offset, 0); +} + +static int dce_v6_0_crtc_do_set_base(struct drm_crtc *crtc, + struct drm_framebuffer *fb, + int x, int y, int atomic) +{ + struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc); + struct drm_device *dev = crtc->dev; + struct amdgpu_device *adev = dev->dev_private; + struct amdgpu_framebuffer *amdgpu_fb; + struct drm_framebuffer *target_fb; + struct drm_gem_object *obj; + struct amdgpu_bo *rbo; + uint64_t fb_location, tiling_flags; + uint32_t fb_format, fb_pitch_pixels, pipe_config; + u32 fb_swap = EVERGREEN_GRPH_ENDIAN_SWAP(EVERGREEN_GRPH_ENDIAN_NONE); + u32 viewport_w, viewport_h; + int r; + bool bypass_lut = false; + + /* no fb bound */ + if (!atomic && !crtc->primary->fb) { + DRM_DEBUG_KMS("No FB bound\n"); + return 0; + } + + if (atomic) { + amdgpu_fb = to_amdgpu_framebuffer(fb); + target_fb = fb; + } + else { + amdgpu_fb = to_amdgpu_framebuffer(crtc->primary->fb); + target_fb = crtc->primary->fb; + } + + /* If atomic, assume fb object is pinned & idle & fenced and + * just update base pointers + */ + obj = amdgpu_fb->obj; + rbo = gem_to_amdgpu_bo(obj); + r = amdgpu_bo_reserve(rbo, false); + if (unlikely(r != 0)) + return r; + + if (atomic) + fb_location = amdgpu_bo_gpu_offset(rbo); + else { + r = amdgpu_bo_pin(rbo, AMDGPU_GEM_DOMAIN_VRAM, &fb_location); + if (unlikely(r != 0)) { + amdgpu_bo_unreserve(rbo); + return -EINVAL; + } + } + + amdgpu_bo_get_tiling_flags(rbo, &tiling_flags); + amdgpu_bo_unreserve(rbo); + + switch (target_fb->pixel_format) { + case DRM_FORMAT_C8: + fb_format = (EVERGREEN_GRPH_DEPTH(EVERGREEN_GRPH_DEPTH_8BPP) | + EVERGREEN_GRPH_FORMAT(EVERGREEN_GRPH_FORMAT_INDEXED)); + break; + case DRM_FORMAT_XRGB4444: + case DRM_FORMAT_ARGB4444: + fb_format = (EVERGREEN_GRPH_DEPTH(EVERGREEN_GRPH_DEPTH_16BPP) | + EVERGREEN_GRPH_FORMAT(EVERGREEN_GRPH_FORMAT_ARGB4444)); +#ifdef __BIG_ENDIAN + fb_swap = EVERGREEN_GRPH_ENDIAN_SWAP(EVERGREEN_GRPH_ENDIAN_8IN16); +#endif + break; + case DRM_FORMAT_XRGB1555: + case DRM_FORMAT_ARGB1555: + fb_format = (EVERGREEN_GRPH_DEPTH(EVERGREEN_GRPH_DEPTH_16BPP) | + EVERGREEN_GRPH_FORMAT(EVERGREEN_GRPH_FORMAT_ARGB1555)); +#ifdef __BIG_ENDIAN + fb_swap = EVERGREEN_GRPH_ENDIAN_SWAP(EVERGREEN_GRPH_ENDIAN_8IN16); +#endif + break; + case DRM_FORMAT_BGRX5551: + case DRM_FORMAT_BGRA5551: + fb_format = (EVERGREEN_GRPH_DEPTH(EVERGREEN_GRPH_DEPTH_16BPP) | + EVERGREEN_GRPH_FORMAT(EVERGREEN_GRPH_FORMAT_BGRA5551)); +#ifdef __BIG_ENDIAN + fb_swap = EVERGREEN_GRPH_ENDIAN_SWAP(EVERGREEN_GRPH_ENDIAN_8IN16); +#endif + break; + case DRM_FORMAT_RGB565: + fb_format = (EVERGREEN_GRPH_DEPTH(EVERGREEN_GRPH_DEPTH_16BPP) | + EVERGREEN_GRPH_FORMAT(EVERGREEN_GRPH_FORMAT_ARGB565)); +#ifdef __BIG_ENDIAN + fb_swap = EVERGREEN_GRPH_ENDIAN_SWAP(EVERGREEN_GRPH_ENDIAN_8IN16); +#endif + break; + case DRM_FORMAT_XRGB8888: + case DRM_FORMAT_ARGB8888: + fb_format = (EVERGREEN_GRPH_DEPTH(EVERGREEN_GRPH_DEPTH_32BPP) | + EVERGREEN_GRPH_FORMAT(EVERGREEN_GRPH_FORMAT_ARGB8888)); +#ifdef __BIG_ENDIAN + fb_swap = EVERGREEN_GRPH_ENDIAN_SWAP(EVERGREEN_GRPH_ENDIAN_8IN32); +#endif + break; + case DRM_FORMAT_XRGB2101010: + case DRM_FORMAT_ARGB2101010: + fb_format = (EVERGREEN_GRPH_DEPTH(EVERGREEN_GRPH_DEPTH_32BPP) | + EVERGREEN_GRPH_FORMAT(EVERGREEN_GRPH_FORMAT_ARGB2101010)); +#ifdef __BIG_ENDIAN + fb_swap = EVERGREEN_GRPH_ENDIAN_SWAP(EVERGREEN_GRPH_ENDIAN_8IN32); +#endif + /* Greater 8 bpc fb needs to bypass hw-lut to retain precision */ + bypass_lut = true; + break; + case DRM_FORMAT_BGRX1010102: + case DRM_FORMAT_BGRA1010102: + fb_format = (EVERGREEN_GRPH_DEPTH(EVERGREEN_GRPH_DEPTH_32BPP) | + EVERGREEN_GRPH_FORMAT(EVERGREEN_GRPH_FORMAT_BGRA1010102)); +#ifdef __BIG_ENDIAN + fb_swap = EVERGREEN_GRPH_ENDIAN_SWAP(EVERGREEN_GRPH_ENDIAN_8IN32); +#endif + /* Greater 8 bpc fb needs to bypass hw-lut to retain precision */ + bypass_lut = true; + break; + default: + DRM_ERROR("Unsupported screen format %s\n", + drm_get_format_name(target_fb->pixel_format)); + return -EINVAL; + } + + if (AMDGPU_TILING_GET(tiling_flags, ARRAY_MODE) == ARRAY_2D_TILED_THIN1) { + unsigned bankw, bankh, mtaspect, tile_split, num_banks; + + bankw = AMDGPU_TILING_GET(tiling_flags, BANK_WIDTH); + bankh = AMDGPU_TILING_GET(tiling_flags, BANK_HEIGHT); + mtaspect = AMDGPU_TILING_GET(tiling_flags, MACRO_TILE_ASPECT); + tile_split = AMDGPU_TILING_GET(tiling_flags, TILE_SPLIT); + num_banks = AMDGPU_TILING_GET(tiling_flags, NUM_BANKS); + + fb_format |= EVERGREEN_GRPH_NUM_BANKS(num_banks); + fb_format |= EVERGREEN_GRPH_ARRAY_MODE(EVERGREEN_GRPH_ARRAY_2D_TILED_THIN1); + fb_format |= EVERGREEN_GRPH_TILE_SPLIT(tile_split); + fb_format |= EVERGREEN_GRPH_BANK_WIDTH(bankw); + fb_format |= EVERGREEN_GRPH_BANK_HEIGHT(bankh); + fb_format |= EVERGREEN_GRPH_MACRO_TILE_ASPECT(mtaspect); + } else if (AMDGPU_TILING_GET(tiling_flags, ARRAY_MODE) == ARRAY_1D_TILED_THIN1) + fb_format |= EVERGREEN_GRPH_ARRAY_MODE(EVERGREEN_GRPH_ARRAY_1D_TILED_THIN1); + + pipe_config = AMDGPU_TILING_GET(tiling_flags, PIPE_CONFIG); + fb_format |= SI_GRPH_PIPE_CONFIG(pipe_config); + + dce_v6_0_vga_enable(crtc, false); + + /* Make sure surface address is updated at vertical blank rather than + * horizontal blank + */ + WREG32(EVERGREEN_GRPH_FLIP_CONTROL + amdgpu_crtc->crtc_offset, 0); + + WREG32(EVERGREEN_GRPH_PRIMARY_SURFACE_ADDRESS_HIGH + amdgpu_crtc->crtc_offset, + upper_32_bits(fb_location)); + WREG32(EVERGREEN_GRPH_SECONDARY_SURFACE_ADDRESS_HIGH + amdgpu_crtc->crtc_offset, + upper_32_bits(fb_location)); + WREG32(EVERGREEN_GRPH_PRIMARY_SURFACE_ADDRESS + amdgpu_crtc->crtc_offset, + (u32)fb_location & EVERGREEN_GRPH_SURFACE_ADDRESS_MASK); + WREG32(EVERGREEN_GRPH_SECONDARY_SURFACE_ADDRESS + amdgpu_crtc->crtc_offset, + (u32) fb_location & EVERGREEN_GRPH_SURFACE_ADDRESS_MASK); + WREG32(EVERGREEN_GRPH_CONTROL + amdgpu_crtc->crtc_offset, fb_format); + WREG32(EVERGREEN_GRPH_SWAP_CONTROL + amdgpu_crtc->crtc_offset, fb_swap); + + /* + * The LUT only has 256 slots for indexing by a 8 bpc fb. Bypass the LUT + * for > 8 bpc scanout to avoid truncation of fb indices to 8 msb's, to + * retain the full precision throughout the pipeline. + */ + WREG32_P(EVERGREEN_GRPH_LUT_10BIT_BYPASS_CONTROL + amdgpu_crtc->crtc_offset, + (bypass_lut ? EVERGREEN_LUT_10BIT_BYPASS_EN : 0), + ~EVERGREEN_LUT_10BIT_BYPASS_EN); + + if (bypass_lut) + DRM_DEBUG_KMS("Bypassing hardware LUT due to 10 bit fb scanout.\n"); + + WREG32(EVERGREEN_GRPH_SURFACE_OFFSET_X + amdgpu_crtc->crtc_offset, 0); + WREG32(EVERGREEN_GRPH_SURFACE_OFFSET_Y + amdgpu_crtc->crtc_offset, 0); + WREG32(EVERGREEN_GRPH_X_START + amdgpu_crtc->crtc_offset, 0); + WREG32(EVERGREEN_GRPH_Y_START + amdgpu_crtc->crtc_offset, 0); + WREG32(EVERGREEN_GRPH_X_END + amdgpu_crtc->crtc_offset, target_fb->width); + WREG32(EVERGREEN_GRPH_Y_END + amdgpu_crtc->crtc_offset, target_fb->height); + + fb_pitch_pixels = target_fb->pitches[0] / (target_fb->bits_per_pixel / 8); + WREG32(EVERGREEN_GRPH_PITCH + amdgpu_crtc->crtc_offset, fb_pitch_pixels); + + dce_v6_0_grph_enable(crtc, true); + + WREG32(EVERGREEN_DESKTOP_HEIGHT + amdgpu_crtc->crtc_offset, + target_fb->height); + x &= ~3; + y &= ~1; + WREG32(EVERGREEN_VIEWPORT_START + amdgpu_crtc->crtc_offset, + (x << 16) | y); + viewport_w = crtc->mode.hdisplay; + viewport_h = (crtc->mode.vdisplay + 1) & ~1; + + WREG32(EVERGREEN_VIEWPORT_SIZE + amdgpu_crtc->crtc_offset, + (viewport_w << 16) | viewport_h); + + /* set pageflip to happen anywhere in vblank interval */ + WREG32(EVERGREEN_MASTER_UPDATE_MODE + amdgpu_crtc->crtc_offset, 0); + + if (!atomic && fb && fb != crtc->primary->fb) { + amdgpu_fb = to_amdgpu_framebuffer(fb); + rbo = gem_to_amdgpu_bo(amdgpu_fb->obj); + r = amdgpu_bo_reserve(rbo, false); + if (unlikely(r != 0)) + return r; + amdgpu_bo_unpin(rbo); + amdgpu_bo_unreserve(rbo); + } + + /* Bytes per pixel may have changed */ + dce_v6_0_bandwidth_update(adev); + + return 0; + +} + +static void dce_v6_0_set_interleave(struct drm_crtc *crtc, + struct drm_display_mode *mode) +{ + struct drm_device *dev = crtc->dev; + struct amdgpu_device *adev = dev->dev_private; + struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc); + + if (mode->flags & DRM_MODE_FLAG_INTERLACE) + WREG32(EVERGREEN_DATA_FORMAT + amdgpu_crtc->crtc_offset, + EVERGREEN_INTERLEAVE_EN); + else + WREG32(EVERGREEN_DATA_FORMAT + amdgpu_crtc->crtc_offset, 0); +} + +static void dce_v6_0_crtc_load_lut(struct drm_crtc *crtc) +{ + + struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc); + struct drm_device *dev = crtc->dev; + struct amdgpu_device *adev = dev->dev_private; + int i; + + DRM_DEBUG_KMS("%d\n", amdgpu_crtc->crtc_id); + + WREG32(NI_INPUT_CSC_CONTROL + amdgpu_crtc->crtc_offset, + (NI_INPUT_CSC_GRPH_MODE(NI_INPUT_CSC_BYPASS) | + NI_INPUT_CSC_OVL_MODE(NI_INPUT_CSC_BYPASS))); + WREG32(NI_PRESCALE_GRPH_CONTROL + amdgpu_crtc->crtc_offset, + NI_GRPH_PRESCALE_BYPASS); + WREG32(NI_PRESCALE_OVL_CONTROL + amdgpu_crtc->crtc_offset, + NI_OVL_PRESCALE_BYPASS); + WREG32(NI_INPUT_GAMMA_CONTROL + amdgpu_crtc->crtc_offset, + (NI_GRPH_INPUT_GAMMA_MODE(NI_INPUT_GAMMA_USE_LUT) | + NI_OVL_INPUT_GAMMA_MODE(NI_INPUT_GAMMA_USE_LUT))); + + + + WREG32(EVERGREEN_DC_LUT_CONTROL + amdgpu_crtc->crtc_offset, 0); + + WREG32(EVERGREEN_DC_LUT_BLACK_OFFSET_BLUE + amdgpu_crtc->crtc_offset, 0); + WREG32(EVERGREEN_DC_LUT_BLACK_OFFSET_GREEN + amdgpu_crtc->crtc_offset, 0); + WREG32(EVERGREEN_DC_LUT_BLACK_OFFSET_RED + amdgpu_crtc->crtc_offset, 0); + + WREG32(EVERGREEN_DC_LUT_WHITE_OFFSET_BLUE + amdgpu_crtc->crtc_offset, 0xffff); + WREG32(EVERGREEN_DC_LUT_WHITE_OFFSET_GREEN + amdgpu_crtc->crtc_offset, 0xffff); + WREG32(EVERGREEN_DC_LUT_WHITE_OFFSET_RED + amdgpu_crtc->crtc_offset, 0xffff); + + WREG32(EVERGREEN_DC_LUT_RW_MODE + amdgpu_crtc->crtc_offset, 0); + WREG32(EVERGREEN_DC_LUT_WRITE_EN_MASK + amdgpu_crtc->crtc_offset, 0x00000007); + + WREG32(EVERGREEN_DC_LUT_RW_INDEX + amdgpu_crtc->crtc_offset, 0); + for (i = 0; i < 256; i++) { + WREG32(EVERGREEN_DC_LUT_30_COLOR + amdgpu_crtc->crtc_offset, + (amdgpu_crtc->lut_r[i] << 20) | + (amdgpu_crtc->lut_g[i] << 10) | + (amdgpu_crtc->lut_b[i] << 0)); + } + + WREG32(NI_DEGAMMA_CONTROL + amdgpu_crtc->crtc_offset, + (NI_GRPH_DEGAMMA_MODE(NI_DEGAMMA_BYPASS) | + NI_OVL_DEGAMMA_MODE(NI_DEGAMMA_BYPASS) | + NI_ICON_DEGAMMA_MODE(NI_DEGAMMA_BYPASS) | + NI_CURSOR_DEGAMMA_MODE(NI_DEGAMMA_BYPASS))); + WREG32(NI_GAMUT_REMAP_CONTROL + amdgpu_crtc->crtc_offset, + (NI_GRPH_GAMUT_REMAP_MODE(NI_GAMUT_REMAP_BYPASS) | + NI_OVL_GAMUT_REMAP_MODE(NI_GAMUT_REMAP_BYPASS))); + WREG32(NI_REGAMMA_CONTROL + amdgpu_crtc->crtc_offset, + (NI_GRPH_REGAMMA_MODE(NI_REGAMMA_BYPASS) | + NI_OVL_REGAMMA_MODE(NI_REGAMMA_BYPASS))); + WREG32(NI_OUTPUT_CSC_CONTROL + amdgpu_crtc->crtc_offset, + (NI_OUTPUT_CSC_GRPH_MODE(0) | + NI_OUTPUT_CSC_OVL_MODE(NI_OUTPUT_CSC_BYPASS))); + /* XXX match this to the depth of the crtc fmt block, move to modeset? */ + WREG32(0x1a50 + amdgpu_crtc->crtc_offset, 0); + + +} + +static int dce_v6_0_pick_dig_encoder(struct drm_encoder *encoder) +{ + struct amdgpu_encoder *amdgpu_encoder = to_amdgpu_encoder(encoder); + struct amdgpu_encoder_atom_dig *dig = amdgpu_encoder->enc_priv; + + switch (amdgpu_encoder->encoder_id) { + case ENCODER_OBJECT_ID_INTERNAL_UNIPHY: + if (dig->linkb) + return 1; + else + return 0; + break; + case ENCODER_OBJECT_ID_INTERNAL_UNIPHY1: + if (dig->linkb) + return 3; + else + return 2; + break; + case ENCODER_OBJECT_ID_INTERNAL_UNIPHY2: + if (dig->linkb) + return 5; + else + return 4; + break; + case ENCODER_OBJECT_ID_INTERNAL_UNIPHY3: + return 6; + break; + default: + DRM_ERROR("invalid encoder_id: 0x%x\n", amdgpu_encoder->encoder_id); + return 0; + } +} + +/** + * dce_v6_0_pick_pll - Allocate a PPLL for use by the crtc. + * + * @crtc: drm crtc + * + * Returns the PPLL (Pixel PLL) to be used by the crtc. For DP monitors + * a single PPLL can be used for all DP crtcs/encoders. For non-DP + * monitors a dedicated PPLL must be used. If a particular board has + * an external DP PLL, return ATOM_PPLL_INVALID to skip PLL programming + * as there is no need to program the PLL itself. If we are not able to + * allocate a PLL, return ATOM_PPLL_INVALID to skip PLL programming to + * avoid messing up an existing monitor. + * + * + */ +static u32 dce_v6_0_pick_pll(struct drm_crtc *crtc) +{ + struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc); + struct drm_device *dev = crtc->dev; + struct amdgpu_device *adev = dev->dev_private; + u32 pll_in_use; + int pll; + + if (ENCODER_MODE_IS_DP(amdgpu_atombios_encoder_get_encoder_mode(amdgpu_crtc->encoder))) { + if (adev->clock.dp_extclk) + /* skip PPLL programming if using ext clock */ + return ATOM_PPLL_INVALID; + else + return ATOM_PPLL0; + } else { + /* use the same PPLL for all monitors with the same clock */ + pll = amdgpu_pll_get_shared_nondp_ppll(crtc); + if (pll != ATOM_PPLL_INVALID) + return pll; + } + + /* PPLL1, and PPLL2 */ + pll_in_use = amdgpu_pll_get_use_mask(crtc); + if (!(pll_in_use & (1 << ATOM_PPLL2))) + return ATOM_PPLL2; + if (!(pll_in_use & (1 << ATOM_PPLL1))) + return ATOM_PPLL1; + DRM_ERROR("unable to allocate a PPLL\n"); + return ATOM_PPLL_INVALID; +} + +static void dce_v6_0_lock_cursor(struct drm_crtc *crtc, bool lock) +{ + struct amdgpu_device *adev = crtc->dev->dev_private; + struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc); + uint32_t cur_lock; + + cur_lock = RREG32(EVERGREEN_CUR_UPDATE + amdgpu_crtc->crtc_offset); + if (lock) + cur_lock |= EVERGREEN_CURSOR_UPDATE_LOCK; + else + cur_lock &= ~EVERGREEN_CURSOR_UPDATE_LOCK; + WREG32(EVERGREEN_CUR_UPDATE + amdgpu_crtc->crtc_offset, cur_lock); +} + +static void dce_v6_0_hide_cursor(struct drm_crtc *crtc) +{ + struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc); + struct amdgpu_device *adev = crtc->dev->dev_private; + + WREG32_IDX(EVERGREEN_CUR_CONTROL + amdgpu_crtc->crtc_offset, + EVERGREEN_CURSOR_MODE(EVERGREEN_CURSOR_24_8_PRE_MULT) | + EVERGREEN_CURSOR_URGENT_CONTROL(EVERGREEN_CURSOR_URGENT_1_2)); + + +} + +static void dce_v6_0_show_cursor(struct drm_crtc *crtc) +{ + struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc); + struct amdgpu_device *adev = crtc->dev->dev_private; + + WREG32(EVERGREEN_CUR_SURFACE_ADDRESS_HIGH + amdgpu_crtc->crtc_offset, + upper_32_bits(amdgpu_crtc->cursor_addr)); + WREG32(EVERGREEN_CUR_SURFACE_ADDRESS + amdgpu_crtc->crtc_offset, + lower_32_bits(amdgpu_crtc->cursor_addr)); + + WREG32_IDX(EVERGREEN_CUR_CONTROL + amdgpu_crtc->crtc_offset, + EVERGREEN_CURSOR_EN | + EVERGREEN_CURSOR_MODE(EVERGREEN_CURSOR_24_8_PRE_MULT) | + EVERGREEN_CURSOR_URGENT_CONTROL(EVERGREEN_CURSOR_URGENT_1_2)); + +} + +static int dce_v6_0_cursor_move_locked(struct drm_crtc *crtc, + int x, int y) +{ + struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc); + struct amdgpu_device *adev = crtc->dev->dev_private; + int xorigin = 0, yorigin = 0; + + int w = amdgpu_crtc->cursor_width; + + /* avivo cursor are offset into the total surface */ + x += crtc->x; + y += crtc->y; + DRM_DEBUG("x %d y %d c->x %d c->y %d\n", x, y, crtc->x, crtc->y); + + if (x < 0) { + xorigin = min(-x, amdgpu_crtc->max_cursor_width - 1); + x = 0; + } + if (y < 0) { + yorigin = min(-y, amdgpu_crtc->max_cursor_height - 1); + y = 0; + } + + WREG32(EVERGREEN_CUR_POSITION + amdgpu_crtc->crtc_offset, (x << 16) | y); + WREG32(EVERGREEN_CUR_HOT_SPOT + amdgpu_crtc->crtc_offset, (xorigin << 16) | yorigin); + WREG32(EVERGREEN_CUR_SIZE + amdgpu_crtc->crtc_offset, + ((w - 1) << 16) | (amdgpu_crtc->cursor_height - 1)); + + amdgpu_crtc->cursor_x = x; + amdgpu_crtc->cursor_y = y; + return 0; +} + +static int dce_v6_0_crtc_cursor_move(struct drm_crtc *crtc, + int x, int y) +{ + int ret; + + dce_v6_0_lock_cursor(crtc, true); + ret = dce_v6_0_cursor_move_locked(crtc, x, y); + dce_v6_0_lock_cursor(crtc, false); + + return ret; +} + +static int dce_v6_0_crtc_cursor_set2(struct drm_crtc *crtc, + struct drm_file *file_priv, + uint32_t handle, + uint32_t width, + uint32_t height, + int32_t hot_x, + int32_t hot_y) +{ + struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc); + struct drm_gem_object *obj; + struct amdgpu_bo *aobj; + int ret; + + if (!handle) { + /* turn off cursor */ + dce_v6_0_hide_cursor(crtc); + obj = NULL; + goto unpin; + } + + if ((width > amdgpu_crtc->max_cursor_width) || + (height > amdgpu_crtc->max_cursor_height)) { + DRM_ERROR("bad cursor width or height %d x %d\n", width, height); + return -EINVAL; + } + + obj = drm_gem_object_lookup(file_priv, handle); + if (!obj) { + DRM_ERROR("Cannot find cursor object %x for crtc %d\n", handle, amdgpu_crtc->crtc_id); + return -ENOENT; + } + + aobj = gem_to_amdgpu_bo(obj); + ret = amdgpu_bo_reserve(aobj, false); + if (ret != 0) { + drm_gem_object_unreference_unlocked(obj); + return ret; + } + + ret = amdgpu_bo_pin(aobj, AMDGPU_GEM_DOMAIN_VRAM, &amdgpu_crtc->cursor_addr); + amdgpu_bo_unreserve(aobj); + if (ret) { + DRM_ERROR("Failed to pin new cursor BO (%d)\n", ret); + drm_gem_object_unreference_unlocked(obj); + return ret; + } + + amdgpu_crtc->cursor_width = width; + amdgpu_crtc->cursor_height = height; + + dce_v6_0_lock_cursor(crtc, true); + + if (hot_x != amdgpu_crtc->cursor_hot_x || + hot_y != amdgpu_crtc->cursor_hot_y) { + int x, y; + + x = amdgpu_crtc->cursor_x + amdgpu_crtc->cursor_hot_x - hot_x; + y = amdgpu_crtc->cursor_y + amdgpu_crtc->cursor_hot_y - hot_y; + + dce_v6_0_cursor_move_locked(crtc, x, y); + + amdgpu_crtc->cursor_hot_x = hot_x; + amdgpu_crtc->cursor_hot_y = hot_y; + } + + dce_v6_0_show_cursor(crtc); + dce_v6_0_lock_cursor(crtc, false); + +unpin: + if (amdgpu_crtc->cursor_bo) { + struct amdgpu_bo *aobj = gem_to_amdgpu_bo(amdgpu_crtc->cursor_bo); + ret = amdgpu_bo_reserve(aobj, false); + if (likely(ret == 0)) { + amdgpu_bo_unpin(aobj); + amdgpu_bo_unreserve(aobj); + } + drm_gem_object_unreference_unlocked(amdgpu_crtc->cursor_bo); + } + + amdgpu_crtc->cursor_bo = obj; + return 0; +} + +static void dce_v6_0_cursor_reset(struct drm_crtc *crtc) +{ + struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc); + + if (amdgpu_crtc->cursor_bo) { + dce_v6_0_lock_cursor(crtc, true); + + dce_v6_0_cursor_move_locked(crtc, amdgpu_crtc->cursor_x, + amdgpu_crtc->cursor_y); + + dce_v6_0_show_cursor(crtc); + + dce_v6_0_lock_cursor(crtc, false); + } +} + +static int dce_v6_0_crtc_gamma_set(struct drm_crtc *crtc, u16 *red, u16 *green, + u16 *blue, uint32_t size) +{ + struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc); + int i; + + /* userspace palettes are always correct as is */ + for (i = 0; i < size; i++) { + amdgpu_crtc->lut_r[i] = red[i] >> 6; + amdgpu_crtc->lut_g[i] = green[i] >> 6; + amdgpu_crtc->lut_b[i] = blue[i] >> 6; + } + dce_v6_0_crtc_load_lut(crtc); + + return 0; +} + +static void dce_v6_0_crtc_destroy(struct drm_crtc *crtc) +{ + struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc); + + drm_crtc_cleanup(crtc); + kfree(amdgpu_crtc); +} + +static const struct drm_crtc_funcs dce_v6_0_crtc_funcs = { + .cursor_set2 = dce_v6_0_crtc_cursor_set2, + .cursor_move = dce_v6_0_crtc_cursor_move, + .gamma_set = dce_v6_0_crtc_gamma_set, + .set_config = amdgpu_crtc_set_config, + .destroy = dce_v6_0_crtc_destroy, + .page_flip_target = amdgpu_crtc_page_flip_target, +}; + +static void dce_v6_0_crtc_dpms(struct drm_crtc *crtc, int mode) +{ + struct drm_device *dev = crtc->dev; + struct amdgpu_device *adev = dev->dev_private; + struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc); + unsigned type; + + switch (mode) { + case DRM_MODE_DPMS_ON: + amdgpu_crtc->enabled = true; + amdgpu_atombios_crtc_enable(crtc, ATOM_ENABLE); + amdgpu_atombios_crtc_blank(crtc, ATOM_DISABLE); + /* Make sure VBLANK and PFLIP interrupts are still enabled */ + type = amdgpu_crtc_idx_to_irq_type(adev, amdgpu_crtc->crtc_id); + amdgpu_irq_update(adev, &adev->crtc_irq, type); + amdgpu_irq_update(adev, &adev->pageflip_irq, type); + drm_vblank_post_modeset(dev, amdgpu_crtc->crtc_id); + dce_v6_0_crtc_load_lut(crtc); + break; + case DRM_MODE_DPMS_STANDBY: + case DRM_MODE_DPMS_SUSPEND: + case DRM_MODE_DPMS_OFF: + drm_vblank_pre_modeset(dev, amdgpu_crtc->crtc_id); + if (amdgpu_crtc->enabled) + amdgpu_atombios_crtc_blank(crtc, ATOM_ENABLE); + amdgpu_atombios_crtc_enable(crtc, ATOM_DISABLE); + amdgpu_crtc->enabled = false; + break; + } + /* adjust pm to dpms */ + amdgpu_pm_compute_clocks(adev); +} + +static void dce_v6_0_crtc_prepare(struct drm_crtc *crtc) +{ + /* disable crtc pair power gating before programming */ + amdgpu_atombios_crtc_powergate(crtc, ATOM_DISABLE); + amdgpu_atombios_crtc_lock(crtc, ATOM_ENABLE); + dce_v6_0_crtc_dpms(crtc, DRM_MODE_DPMS_OFF); +} + +static void dce_v6_0_crtc_commit(struct drm_crtc *crtc) +{ + dce_v6_0_crtc_dpms(crtc, DRM_MODE_DPMS_ON); + amdgpu_atombios_crtc_lock(crtc, ATOM_DISABLE); +} + +static void dce_v6_0_crtc_disable(struct drm_crtc *crtc) +{ + + struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc); + struct drm_device *dev = crtc->dev; + struct amdgpu_device *adev = dev->dev_private; + struct amdgpu_atom_ss ss; + int i; + + dce_v6_0_crtc_dpms(crtc, DRM_MODE_DPMS_OFF); + if (crtc->primary->fb) { + int r; + struct amdgpu_framebuffer *amdgpu_fb; + struct amdgpu_bo *rbo; + + amdgpu_fb = to_amdgpu_framebuffer(crtc->primary->fb); + rbo = gem_to_amdgpu_bo(amdgpu_fb->obj); + r = amdgpu_bo_reserve(rbo, false); + if (unlikely(r)) + DRM_ERROR("failed to reserve rbo before unpin\n"); + else { + amdgpu_bo_unpin(rbo); + amdgpu_bo_unreserve(rbo); + } + } + /* disable the GRPH */ + dce_v6_0_grph_enable(crtc, false); + + amdgpu_atombios_crtc_powergate(crtc, ATOM_ENABLE); + + for (i = 0; i < adev->mode_info.num_crtc; i++) { + if (adev->mode_info.crtcs[i] && + adev->mode_info.crtcs[i]->enabled && + i != amdgpu_crtc->crtc_id && + amdgpu_crtc->pll_id == adev->mode_info.crtcs[i]->pll_id) { + /* one other crtc is using this pll don't turn + * off the pll + */ + goto done; + } + } + + switch (amdgpu_crtc->pll_id) { + case ATOM_PPLL1: + case ATOM_PPLL2: + /* disable the ppll */ + amdgpu_atombios_crtc_program_pll(crtc, amdgpu_crtc->crtc_id, amdgpu_crtc->pll_id, + 0, 0, ATOM_DISABLE, 0, 0, 0, 0, 0, false, &ss); + break; + default: + break; + } +done: + amdgpu_crtc->pll_id = ATOM_PPLL_INVALID; + amdgpu_crtc->adjusted_clock = 0; + amdgpu_crtc->encoder = NULL; + amdgpu_crtc->connector = NULL; +} + +static int dce_v6_0_crtc_mode_set(struct drm_crtc *crtc, + struct drm_display_mode *mode, + struct drm_display_mode *adjusted_mode, + int x, int y, struct drm_framebuffer *old_fb) +{ + struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc); + + if (!amdgpu_crtc->adjusted_clock) + return -EINVAL; + + amdgpu_atombios_crtc_set_pll(crtc, adjusted_mode); + amdgpu_atombios_crtc_set_dtd_timing(crtc, adjusted_mode); + dce_v6_0_crtc_do_set_base(crtc, old_fb, x, y, 0); + amdgpu_atombios_crtc_overscan_setup(crtc, mode, adjusted_mode); + amdgpu_atombios_crtc_scaler_setup(crtc); + dce_v6_0_cursor_reset(crtc); + /* update the hw version fpr dpm */ + amdgpu_crtc->hw_mode = *adjusted_mode; + + return 0; +} + +static bool dce_v6_0_crtc_mode_fixup(struct drm_crtc *crtc, + const struct drm_display_mode *mode, + struct drm_display_mode *adjusted_mode) +{ + + struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc); + struct drm_device *dev = crtc->dev; + struct drm_encoder *encoder; + + /* assign the encoder to the amdgpu crtc to avoid repeated lookups later */ + list_for_each_entry(encoder, &dev->mode_config.encoder_list, head) { + if (encoder->crtc == crtc) { + amdgpu_crtc->encoder = encoder; + amdgpu_crtc->connector = amdgpu_get_connector_for_encoder(encoder); + break; + } + } + if ((amdgpu_crtc->encoder == NULL) || (amdgpu_crtc->connector == NULL)) { + amdgpu_crtc->encoder = NULL; + amdgpu_crtc->connector = NULL; + return false; + } + if (!amdgpu_crtc_scaling_mode_fixup(crtc, mode, adjusted_mode)) + return false; + if (amdgpu_atombios_crtc_prepare_pll(crtc, adjusted_mode)) + return false; + /* pick pll */ + amdgpu_crtc->pll_id = dce_v6_0_pick_pll(crtc); + /* if we can't get a PPLL for a non-DP encoder, fail */ + if ((amdgpu_crtc->pll_id == ATOM_PPLL_INVALID) && + !ENCODER_MODE_IS_DP(amdgpu_atombios_encoder_get_encoder_mode(amdgpu_crtc->encoder))) + return false; + + return true; +} + +static int dce_v6_0_crtc_set_base(struct drm_crtc *crtc, int x, int y, + struct drm_framebuffer *old_fb) +{ + return dce_v6_0_crtc_do_set_base(crtc, old_fb, x, y, 0); +} + +static int dce_v6_0_crtc_set_base_atomic(struct drm_crtc *crtc, + struct drm_framebuffer *fb, + int x, int y, enum mode_set_atomic state) +{ + return dce_v6_0_crtc_do_set_base(crtc, fb, x, y, 1); +} + +static const struct drm_crtc_helper_funcs dce_v6_0_crtc_helper_funcs = { + .dpms = dce_v6_0_crtc_dpms, + .mode_fixup = dce_v6_0_crtc_mode_fixup, + .mode_set = dce_v6_0_crtc_mode_set, + .mode_set_base = dce_v6_0_crtc_set_base, + .mode_set_base_atomic = dce_v6_0_crtc_set_base_atomic, + .prepare = dce_v6_0_crtc_prepare, + .commit = dce_v6_0_crtc_commit, + .load_lut = dce_v6_0_crtc_load_lut, + .disable = dce_v6_0_crtc_disable, +}; + +static int dce_v6_0_crtc_init(struct amdgpu_device *adev, int index) +{ + struct amdgpu_crtc *amdgpu_crtc; + int i; + + amdgpu_crtc = kzalloc(sizeof(struct amdgpu_crtc) + + (AMDGPUFB_CONN_LIMIT * sizeof(struct drm_connector *)), GFP_KERNEL); + if (amdgpu_crtc == NULL) + return -ENOMEM; + + drm_crtc_init(adev->ddev, &amdgpu_crtc->base, &dce_v6_0_crtc_funcs); + + drm_mode_crtc_set_gamma_size(&amdgpu_crtc->base, 256); + amdgpu_crtc->crtc_id = index; + adev->mode_info.crtcs[index] = amdgpu_crtc; + + amdgpu_crtc->max_cursor_width = CURSOR_WIDTH; + amdgpu_crtc->max_cursor_height = CURSOR_HEIGHT; + adev->ddev->mode_config.cursor_width = amdgpu_crtc->max_cursor_width; + adev->ddev->mode_config.cursor_height = amdgpu_crtc->max_cursor_height; + + for (i = 0; i < 256; i++) { + amdgpu_crtc->lut_r[i] = i << 2; + amdgpu_crtc->lut_g[i] = i << 2; + amdgpu_crtc->lut_b[i] = i << 2; + } + + amdgpu_crtc->crtc_offset = crtc_offsets[amdgpu_crtc->crtc_id]; + + amdgpu_crtc->pll_id = ATOM_PPLL_INVALID; + amdgpu_crtc->adjusted_clock = 0; + amdgpu_crtc->encoder = NULL; + amdgpu_crtc->connector = NULL; + drm_crtc_helper_add(&amdgpu_crtc->base, &dce_v6_0_crtc_helper_funcs); + + return 0; +} + +static int dce_v6_0_early_init(void *handle) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + adev->audio_endpt_rreg = &dce_v6_0_audio_endpt_rreg; + adev->audio_endpt_wreg = &dce_v6_0_audio_endpt_wreg; + + dce_v6_0_set_display_funcs(adev); + dce_v6_0_set_irq_funcs(adev); + + switch (adev->asic_type) { + case CHIP_TAHITI: + case CHIP_PITCAIRN: + case CHIP_VERDE: + adev->mode_info.num_crtc = 6; + adev->mode_info.num_hpd = 6; + adev->mode_info.num_dig = 6; + break; + case CHIP_OLAND: + adev->mode_info.num_crtc = 2; + adev->mode_info.num_hpd = 2; + adev->mode_info.num_dig = 2; + break; + default: + /* FIXME: not supported yet */ + return -EINVAL; + } + + return 0; +} + +static int dce_v6_0_sw_init(void *handle) +{ + int r, i; + bool ret; + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + for (i = 0; i < adev->mode_info.num_crtc; i++) { + r = amdgpu_irq_add_id(adev, i + 1, &adev->crtc_irq); + if (r) + return r; + } + + for (i = 8; i < 20; i += 2) { + r = amdgpu_irq_add_id(adev, i, &adev->pageflip_irq); + if (r) + return r; + } + + /* HPD hotplug */ + r = amdgpu_irq_add_id(adev, 42, &adev->hpd_irq); + if (r) + return r; + + adev->mode_info.mode_config_initialized = true; + + adev->ddev->mode_config.funcs = &amdgpu_mode_funcs; + + adev->ddev->mode_config.async_page_flip = true; + + adev->ddev->mode_config.max_width = 16384; + adev->ddev->mode_config.max_height = 16384; + + adev->ddev->mode_config.preferred_depth = 24; + adev->ddev->mode_config.prefer_shadow = 1; + + adev->ddev->mode_config.fb_base = adev->mc.aper_base; + + r = amdgpu_modeset_create_props(adev); + if (r) + return r; + + adev->ddev->mode_config.max_width = 16384; + adev->ddev->mode_config.max_height = 16384; + + /* allocate crtcs */ + for (i = 0; i < adev->mode_info.num_crtc; i++) { + r = dce_v6_0_crtc_init(adev, i); + if (r) + return r; + } + + ret = amdgpu_atombios_get_connector_info_from_object_table(adev); + if (ret) + amdgpu_print_display_setup(adev->ddev); + else + return -EINVAL; + + /* setup afmt */ + dce_v6_0_afmt_init(adev); + + r = dce_v6_0_audio_init(adev); + if (r) + return r; + + drm_kms_helper_poll_init(adev->ddev); + + return r; +} + +static int dce_v6_0_sw_fini(void *handle) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + kfree(adev->mode_info.bios_hardcoded_edid); + + drm_kms_helper_poll_fini(adev->ddev); + + dce_v6_0_audio_fini(adev); + + dce_v6_0_afmt_fini(adev); + + drm_mode_config_cleanup(adev->ddev); + adev->mode_info.mode_config_initialized = false; + + return 0; +} + +static int dce_v6_0_hw_init(void *handle) +{ + int i; + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + /* init dig PHYs, disp eng pll */ + amdgpu_atombios_encoder_init_dig(adev); + amdgpu_atombios_crtc_set_disp_eng_pll(adev, adev->clock.default_dispclk); + + /* initialize hpd */ + dce_v6_0_hpd_init(adev); + + for (i = 0; i < adev->mode_info.audio.num_pins; i++) { + dce_v6_0_audio_enable(adev, &adev->mode_info.audio.pin[i], false); + } + + dce_v6_0_pageflip_interrupt_init(adev); + + return 0; +} + +static int dce_v6_0_hw_fini(void *handle) +{ + int i; + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + dce_v6_0_hpd_fini(adev); + + for (i = 0; i < adev->mode_info.audio.num_pins; i++) { + dce_v6_0_audio_enable(adev, &adev->mode_info.audio.pin[i], false); + } + + dce_v6_0_pageflip_interrupt_fini(adev); + + return 0; +} + +static int dce_v6_0_suspend(void *handle) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + amdgpu_atombios_scratch_regs_save(adev); + + return dce_v6_0_hw_fini(handle); +} + +static int dce_v6_0_resume(void *handle) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + int ret; + + ret = dce_v6_0_hw_init(handle); + + amdgpu_atombios_scratch_regs_restore(adev); + + /* turn on the BL */ + if (adev->mode_info.bl_encoder) { + u8 bl_level = amdgpu_display_backlight_get_level(adev, + adev->mode_info.bl_encoder); + amdgpu_display_backlight_set_level(adev, adev->mode_info.bl_encoder, + bl_level); + } + + return ret; +} + +static bool dce_v6_0_is_idle(void *handle) +{ + return true; +} + +static int dce_v6_0_wait_for_idle(void *handle) +{ + return 0; +} + +static int dce_v6_0_soft_reset(void *handle) +{ + DRM_INFO("xxxx: dce_v6_0_soft_reset --- no impl!!\n"); + return 0; +} + +static void dce_v6_0_set_crtc_vblank_interrupt_state(struct amdgpu_device *adev, + int crtc, + enum amdgpu_interrupt_state state) +{ + u32 reg_block, interrupt_mask; + + if (crtc >= adev->mode_info.num_crtc) { + DRM_DEBUG("invalid crtc %d\n", crtc); + return; + } + + switch (crtc) { + case 0: + reg_block = SI_CRTC0_REGISTER_OFFSET; + break; + case 1: + reg_block = SI_CRTC1_REGISTER_OFFSET; + break; + case 2: + reg_block = SI_CRTC2_REGISTER_OFFSET; + break; + case 3: + reg_block = SI_CRTC3_REGISTER_OFFSET; + break; + case 4: + reg_block = SI_CRTC4_REGISTER_OFFSET; + break; + case 5: + reg_block = SI_CRTC5_REGISTER_OFFSET; + break; + default: + DRM_DEBUG("invalid crtc %d\n", crtc); + return; + } + + switch (state) { + case AMDGPU_IRQ_STATE_DISABLE: + interrupt_mask = RREG32(INT_MASK + reg_block); + interrupt_mask &= ~VBLANK_INT_MASK; + WREG32(INT_MASK + reg_block, interrupt_mask); + break; + case AMDGPU_IRQ_STATE_ENABLE: + interrupt_mask = RREG32(INT_MASK + reg_block); + interrupt_mask |= VBLANK_INT_MASK; + WREG32(INT_MASK + reg_block, interrupt_mask); + break; + default: + break; + } +} + +static void dce_v6_0_set_crtc_vline_interrupt_state(struct amdgpu_device *adev, + int crtc, + enum amdgpu_interrupt_state state) +{ + +} + +static int dce_v6_0_set_hpd_interrupt_state(struct amdgpu_device *adev, + struct amdgpu_irq_src *src, + unsigned type, + enum amdgpu_interrupt_state state) +{ + u32 dc_hpd_int_cntl_reg, dc_hpd_int_cntl; + + switch (type) { + case AMDGPU_HPD_1: + dc_hpd_int_cntl_reg = DC_HPD1_INT_CONTROL; + break; + case AMDGPU_HPD_2: + dc_hpd_int_cntl_reg = DC_HPD2_INT_CONTROL; + break; + case AMDGPU_HPD_3: + dc_hpd_int_cntl_reg = DC_HPD3_INT_CONTROL; + break; + case AMDGPU_HPD_4: + dc_hpd_int_cntl_reg = DC_HPD4_INT_CONTROL; + break; + case AMDGPU_HPD_5: + dc_hpd_int_cntl_reg = DC_HPD5_INT_CONTROL; + break; + case AMDGPU_HPD_6: + dc_hpd_int_cntl_reg = DC_HPD6_INT_CONTROL; + break; + default: + DRM_DEBUG("invalid hdp %d\n", type); + return 0; + } + + switch (state) { + case AMDGPU_IRQ_STATE_DISABLE: + dc_hpd_int_cntl = RREG32(dc_hpd_int_cntl_reg); + dc_hpd_int_cntl &= ~(DC_HPDx_INT_EN | DC_HPDx_RX_INT_EN); + WREG32(dc_hpd_int_cntl_reg, dc_hpd_int_cntl); + break; + case AMDGPU_IRQ_STATE_ENABLE: + dc_hpd_int_cntl = RREG32(dc_hpd_int_cntl_reg); + dc_hpd_int_cntl |= (DC_HPDx_INT_EN | DC_HPDx_RX_INT_EN); + WREG32(dc_hpd_int_cntl_reg, dc_hpd_int_cntl); + break; + default: + break; + } + + return 0; +} + +static int dce_v6_0_set_crtc_interrupt_state(struct amdgpu_device *adev, + struct amdgpu_irq_src *src, + unsigned type, + enum amdgpu_interrupt_state state) +{ + switch (type) { + case AMDGPU_CRTC_IRQ_VBLANK1: + dce_v6_0_set_crtc_vblank_interrupt_state(adev, 0, state); + break; + case AMDGPU_CRTC_IRQ_VBLANK2: + dce_v6_0_set_crtc_vblank_interrupt_state(adev, 1, state); + break; + case AMDGPU_CRTC_IRQ_VBLANK3: + dce_v6_0_set_crtc_vblank_interrupt_state(adev, 2, state); + break; + case AMDGPU_CRTC_IRQ_VBLANK4: + dce_v6_0_set_crtc_vblank_interrupt_state(adev, 3, state); + break; + case AMDGPU_CRTC_IRQ_VBLANK5: + dce_v6_0_set_crtc_vblank_interrupt_state(adev, 4, state); + break; + case AMDGPU_CRTC_IRQ_VBLANK6: + dce_v6_0_set_crtc_vblank_interrupt_state(adev, 5, state); + break; + case AMDGPU_CRTC_IRQ_VLINE1: + dce_v6_0_set_crtc_vline_interrupt_state(adev, 0, state); + break; + case AMDGPU_CRTC_IRQ_VLINE2: + dce_v6_0_set_crtc_vline_interrupt_state(adev, 1, state); + break; + case AMDGPU_CRTC_IRQ_VLINE3: + dce_v6_0_set_crtc_vline_interrupt_state(adev, 2, state); + break; + case AMDGPU_CRTC_IRQ_VLINE4: + dce_v6_0_set_crtc_vline_interrupt_state(adev, 3, state); + break; + case AMDGPU_CRTC_IRQ_VLINE5: + dce_v6_0_set_crtc_vline_interrupt_state(adev, 4, state); + break; + case AMDGPU_CRTC_IRQ_VLINE6: + dce_v6_0_set_crtc_vline_interrupt_state(adev, 5, state); + break; + default: + break; + } + return 0; +} + +static int dce_v6_0_crtc_irq(struct amdgpu_device *adev, + struct amdgpu_irq_src *source, + struct amdgpu_iv_entry *entry) +{ + unsigned crtc = entry->src_id - 1; + uint32_t disp_int = RREG32(interrupt_status_offsets[crtc].reg); + unsigned irq_type = amdgpu_crtc_idx_to_irq_type(adev, crtc); + + switch (entry->src_data) { + case 0: /* vblank */ + if (disp_int & interrupt_status_offsets[crtc].vblank) + WREG32(VBLANK_STATUS + crtc_offsets[crtc], VBLANK_ACK); + else + DRM_DEBUG("IH: IH event w/o asserted irq bit?\n"); + + if (amdgpu_irq_enabled(adev, source, irq_type)) { + drm_handle_vblank(adev->ddev, crtc); + } + DRM_DEBUG("IH: D%d vblank\n", crtc + 1); + break; + case 1: /* vline */ + if (disp_int & interrupt_status_offsets[crtc].vline) + WREG32(VLINE_STATUS + crtc_offsets[crtc], VLINE_ACK); + else + DRM_DEBUG("IH: IH event w/o asserted irq bit?\n"); + + DRM_DEBUG("IH: D%d vline\n", crtc + 1); + break; + default: + DRM_DEBUG("Unhandled interrupt: %d %d\n", entry->src_id, entry->src_data); + break; + } + + return 0; +} + +static int dce_v6_0_set_pageflip_interrupt_state(struct amdgpu_device *adev, + struct amdgpu_irq_src *src, + unsigned type, + enum amdgpu_interrupt_state state) +{ + u32 reg; + + if (type >= adev->mode_info.num_crtc) { + DRM_ERROR("invalid pageflip crtc %d\n", type); + return -EINVAL; + } + + reg = RREG32(GRPH_INT_CONTROL + crtc_offsets[type]); + if (state == AMDGPU_IRQ_STATE_DISABLE) + WREG32(GRPH_INT_CONTROL + crtc_offsets[type], + reg & ~GRPH_INTERRUPT_CONTROL__GRPH_PFLIP_INT_MASK_MASK); + else + WREG32(GRPH_INT_CONTROL + crtc_offsets[type], + reg | GRPH_INTERRUPT_CONTROL__GRPH_PFLIP_INT_MASK_MASK); + + return 0; +} + +static int dce_v6_0_pageflip_irq(struct amdgpu_device *adev, + struct amdgpu_irq_src *source, + struct amdgpu_iv_entry *entry) +{ + unsigned long flags; + unsigned crtc_id; + struct amdgpu_crtc *amdgpu_crtc; + struct amdgpu_flip_work *works; + + crtc_id = (entry->src_id - 8) >> 1; + amdgpu_crtc = adev->mode_info.crtcs[crtc_id]; + + if (crtc_id >= adev->mode_info.num_crtc) { + DRM_ERROR("invalid pageflip crtc %d\n", crtc_id); + return -EINVAL; + } + + if (RREG32(GRPH_INT_STATUS + crtc_offsets[crtc_id]) & + GRPH_INTERRUPT_STATUS__GRPH_PFLIP_INT_OCCURRED_MASK) + WREG32(GRPH_INT_STATUS + crtc_offsets[crtc_id], + GRPH_INTERRUPT_STATUS__GRPH_PFLIP_INT_CLEAR_MASK); + + /* IRQ could occur when in initial stage */ + if (amdgpu_crtc == NULL) + return 0; + + spin_lock_irqsave(&adev->ddev->event_lock, flags); + works = amdgpu_crtc->pflip_works; + if (amdgpu_crtc->pflip_status != AMDGPU_FLIP_SUBMITTED){ + DRM_DEBUG_DRIVER("amdgpu_crtc->pflip_status = %d != " + "AMDGPU_FLIP_SUBMITTED(%d)\n", + amdgpu_crtc->pflip_status, + AMDGPU_FLIP_SUBMITTED); + spin_unlock_irqrestore(&adev->ddev->event_lock, flags); + return 0; + } + + /* page flip completed. clean up */ + amdgpu_crtc->pflip_status = AMDGPU_FLIP_NONE; + amdgpu_crtc->pflip_works = NULL; + + /* wakeup usersapce */ + if (works->event) + drm_crtc_send_vblank_event(&amdgpu_crtc->base, works->event); + + spin_unlock_irqrestore(&adev->ddev->event_lock, flags); + + drm_crtc_vblank_put(&amdgpu_crtc->base); + schedule_work(&works->unpin_work); + + return 0; +} + +static int dce_v6_0_hpd_irq(struct amdgpu_device *adev, + struct amdgpu_irq_src *source, + struct amdgpu_iv_entry *entry) +{ + uint32_t disp_int, mask, int_control, tmp; + unsigned hpd; + + if (entry->src_data > 6) { + DRM_DEBUG("Unhandled interrupt: %d %d\n", entry->src_id, entry->src_data); + return 0; + } + + hpd = entry->src_data; + disp_int = RREG32(interrupt_status_offsets[hpd].reg); + mask = interrupt_status_offsets[hpd].hpd; + int_control = hpd_int_control_offsets[hpd]; + + if (disp_int & mask) { + tmp = RREG32(int_control); + tmp |= DC_HPD1_INT_CONTROL__DC_HPD1_INT_ACK_MASK; + WREG32(int_control, tmp); + schedule_work(&adev->hotplug_work); + DRM_INFO("IH: HPD%d\n", hpd + 1); + } + + return 0; + +} + +static int dce_v6_0_set_clockgating_state(void *handle, + enum amd_clockgating_state state) +{ + return 0; +} + +static int dce_v6_0_set_powergating_state(void *handle, + enum amd_powergating_state state) +{ + return 0; +} + +const struct amd_ip_funcs dce_v6_0_ip_funcs = { + .name = "dce_v6_0", + .early_init = dce_v6_0_early_init, + .late_init = NULL, + .sw_init = dce_v6_0_sw_init, + .sw_fini = dce_v6_0_sw_fini, + .hw_init = dce_v6_0_hw_init, + .hw_fini = dce_v6_0_hw_fini, + .suspend = dce_v6_0_suspend, + .resume = dce_v6_0_resume, + .is_idle = dce_v6_0_is_idle, + .wait_for_idle = dce_v6_0_wait_for_idle, + .soft_reset = dce_v6_0_soft_reset, + .set_clockgating_state = dce_v6_0_set_clockgating_state, + .set_powergating_state = dce_v6_0_set_powergating_state, +}; + +static void +dce_v6_0_encoder_mode_set(struct drm_encoder *encoder, + struct drm_display_mode *mode, + struct drm_display_mode *adjusted_mode) +{ + + struct amdgpu_encoder *amdgpu_encoder = to_amdgpu_encoder(encoder); + + amdgpu_encoder->pixel_clock = adjusted_mode->clock; + + /* need to call this here rather than in prepare() since we need some crtc info */ + amdgpu_atombios_encoder_dpms(encoder, DRM_MODE_DPMS_OFF); + + /* set scaler clears this on some chips */ + dce_v6_0_set_interleave(encoder->crtc, mode); + + if (amdgpu_atombios_encoder_get_encoder_mode(encoder) == ATOM_ENCODER_MODE_HDMI) { + dce_v6_0_afmt_enable(encoder, true); + dce_v6_0_afmt_setmode(encoder, adjusted_mode); + } +} + +static void dce_v6_0_encoder_prepare(struct drm_encoder *encoder) +{ + + struct amdgpu_device *adev = encoder->dev->dev_private; + struct amdgpu_encoder *amdgpu_encoder = to_amdgpu_encoder(encoder); + struct drm_connector *connector = amdgpu_get_connector_for_encoder(encoder); + + if ((amdgpu_encoder->active_device & + (ATOM_DEVICE_DFP_SUPPORT | ATOM_DEVICE_LCD_SUPPORT)) || + (amdgpu_encoder_get_dp_bridge_encoder_id(encoder) != + ENCODER_OBJECT_ID_NONE)) { + struct amdgpu_encoder_atom_dig *dig = amdgpu_encoder->enc_priv; + if (dig) { + dig->dig_encoder = dce_v6_0_pick_dig_encoder(encoder); + if (amdgpu_encoder->active_device & ATOM_DEVICE_DFP_SUPPORT) + dig->afmt = adev->mode_info.afmt[dig->dig_encoder]; + } + } + + amdgpu_atombios_scratch_regs_lock(adev, true); + + if (connector) { + struct amdgpu_connector *amdgpu_connector = to_amdgpu_connector(connector); + + /* select the clock/data port if it uses a router */ + if (amdgpu_connector->router.cd_valid) + amdgpu_i2c_router_select_cd_port(amdgpu_connector); + + /* turn eDP panel on for mode set */ + if (connector->connector_type == DRM_MODE_CONNECTOR_eDP) + amdgpu_atombios_encoder_set_edp_panel_power(connector, + ATOM_TRANSMITTER_ACTION_POWER_ON); + } + + /* this is needed for the pll/ss setup to work correctly in some cases */ + amdgpu_atombios_encoder_set_crtc_source(encoder); + /* set up the FMT blocks */ + dce_v6_0_program_fmt(encoder); +} + +static void dce_v6_0_encoder_commit(struct drm_encoder *encoder) +{ + + struct drm_device *dev = encoder->dev; + struct amdgpu_device *adev = dev->dev_private; + + /* need to call this here as we need the crtc set up */ + amdgpu_atombios_encoder_dpms(encoder, DRM_MODE_DPMS_ON); + amdgpu_atombios_scratch_regs_lock(adev, false); +} + +static void dce_v6_0_encoder_disable(struct drm_encoder *encoder) +{ + + struct amdgpu_encoder *amdgpu_encoder = to_amdgpu_encoder(encoder); + struct amdgpu_encoder_atom_dig *dig; + + amdgpu_atombios_encoder_dpms(encoder, DRM_MODE_DPMS_OFF); + + if (amdgpu_atombios_encoder_is_digital(encoder)) { + if (amdgpu_atombios_encoder_get_encoder_mode(encoder) == ATOM_ENCODER_MODE_HDMI) + dce_v6_0_afmt_enable(encoder, false); + dig = amdgpu_encoder->enc_priv; + dig->dig_encoder = -1; + } + amdgpu_encoder->active_device = 0; +} + +/* these are handled by the primary encoders */ +static void dce_v6_0_ext_prepare(struct drm_encoder *encoder) +{ + +} + +static void dce_v6_0_ext_commit(struct drm_encoder *encoder) +{ + +} + +static void +dce_v6_0_ext_mode_set(struct drm_encoder *encoder, + struct drm_display_mode *mode, + struct drm_display_mode *adjusted_mode) +{ + +} + +static void dce_v6_0_ext_disable(struct drm_encoder *encoder) +{ + +} + +static void +dce_v6_0_ext_dpms(struct drm_encoder *encoder, int mode) +{ + +} + +static bool dce_v6_0_ext_mode_fixup(struct drm_encoder *encoder, + const struct drm_display_mode *mode, + struct drm_display_mode *adjusted_mode) +{ + return true; +} + +static const struct drm_encoder_helper_funcs dce_v6_0_ext_helper_funcs = { + .dpms = dce_v6_0_ext_dpms, + .mode_fixup = dce_v6_0_ext_mode_fixup, + .prepare = dce_v6_0_ext_prepare, + .mode_set = dce_v6_0_ext_mode_set, + .commit = dce_v6_0_ext_commit, + .disable = dce_v6_0_ext_disable, + /* no detect for TMDS/LVDS yet */ +}; + +static const struct drm_encoder_helper_funcs dce_v6_0_dig_helper_funcs = { + .dpms = amdgpu_atombios_encoder_dpms, + .mode_fixup = amdgpu_atombios_encoder_mode_fixup, + .prepare = dce_v6_0_encoder_prepare, + .mode_set = dce_v6_0_encoder_mode_set, + .commit = dce_v6_0_encoder_commit, + .disable = dce_v6_0_encoder_disable, + .detect = amdgpu_atombios_encoder_dig_detect, +}; + +static const struct drm_encoder_helper_funcs dce_v6_0_dac_helper_funcs = { + .dpms = amdgpu_atombios_encoder_dpms, + .mode_fixup = amdgpu_atombios_encoder_mode_fixup, + .prepare = dce_v6_0_encoder_prepare, + .mode_set = dce_v6_0_encoder_mode_set, + .commit = dce_v6_0_encoder_commit, + .detect = amdgpu_atombios_encoder_dac_detect, +}; + +static void dce_v6_0_encoder_destroy(struct drm_encoder *encoder) +{ + struct amdgpu_encoder *amdgpu_encoder = to_amdgpu_encoder(encoder); + if (amdgpu_encoder->devices & (ATOM_DEVICE_LCD_SUPPORT)) + amdgpu_atombios_encoder_fini_backlight(amdgpu_encoder); + kfree(amdgpu_encoder->enc_priv); + drm_encoder_cleanup(encoder); + kfree(amdgpu_encoder); +} + +static const struct drm_encoder_funcs dce_v6_0_encoder_funcs = { + .destroy = dce_v6_0_encoder_destroy, +}; + +static void dce_v6_0_encoder_add(struct amdgpu_device *adev, + uint32_t encoder_enum, + uint32_t supported_device, + u16 caps) +{ + struct drm_device *dev = adev->ddev; + struct drm_encoder *encoder; + struct amdgpu_encoder *amdgpu_encoder; + + /* see if we already added it */ + list_for_each_entry(encoder, &dev->mode_config.encoder_list, head) { + amdgpu_encoder = to_amdgpu_encoder(encoder); + if (amdgpu_encoder->encoder_enum == encoder_enum) { + amdgpu_encoder->devices |= supported_device; + return; + } + + } + + /* add a new one */ + amdgpu_encoder = kzalloc(sizeof(struct amdgpu_encoder), GFP_KERNEL); + if (!amdgpu_encoder) + return; + + encoder = &amdgpu_encoder->base; + switch (adev->mode_info.num_crtc) { + case 1: + encoder->possible_crtcs = 0x1; + break; + case 2: + default: + encoder->possible_crtcs = 0x3; + break; + case 4: + encoder->possible_crtcs = 0xf; + break; + case 6: + encoder->possible_crtcs = 0x3f; + break; + } + + amdgpu_encoder->enc_priv = NULL; + + amdgpu_encoder->encoder_enum = encoder_enum; + amdgpu_encoder->encoder_id = (encoder_enum & OBJECT_ID_MASK) >> OBJECT_ID_SHIFT; + amdgpu_encoder->devices = supported_device; + amdgpu_encoder->rmx_type = RMX_OFF; + amdgpu_encoder->underscan_type = UNDERSCAN_OFF; + amdgpu_encoder->is_ext_encoder = false; + amdgpu_encoder->caps = caps; + + switch (amdgpu_encoder->encoder_id) { + case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_DAC1: + case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_DAC2: + drm_encoder_init(dev, encoder, &dce_v6_0_encoder_funcs, + DRM_MODE_ENCODER_DAC, NULL); + drm_encoder_helper_add(encoder, &dce_v6_0_dac_helper_funcs); + break; + case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_DVO1: + case ENCODER_OBJECT_ID_INTERNAL_UNIPHY: + case ENCODER_OBJECT_ID_INTERNAL_UNIPHY1: + case ENCODER_OBJECT_ID_INTERNAL_UNIPHY2: + case ENCODER_OBJECT_ID_INTERNAL_UNIPHY3: + if (amdgpu_encoder->devices & (ATOM_DEVICE_LCD_SUPPORT)) { + amdgpu_encoder->rmx_type = RMX_FULL; + drm_encoder_init(dev, encoder, &dce_v6_0_encoder_funcs, + DRM_MODE_ENCODER_LVDS, NULL); + amdgpu_encoder->enc_priv = amdgpu_atombios_encoder_get_lcd_info(amdgpu_encoder); + } else if (amdgpu_encoder->devices & (ATOM_DEVICE_CRT_SUPPORT)) { + drm_encoder_init(dev, encoder, &dce_v6_0_encoder_funcs, + DRM_MODE_ENCODER_DAC, NULL); + amdgpu_encoder->enc_priv = amdgpu_atombios_encoder_get_dig_info(amdgpu_encoder); + } else { + drm_encoder_init(dev, encoder, &dce_v6_0_encoder_funcs, + DRM_MODE_ENCODER_TMDS, NULL); + amdgpu_encoder->enc_priv = amdgpu_atombios_encoder_get_dig_info(amdgpu_encoder); + } + drm_encoder_helper_add(encoder, &dce_v6_0_dig_helper_funcs); + break; + case ENCODER_OBJECT_ID_SI170B: + case ENCODER_OBJECT_ID_CH7303: + case ENCODER_OBJECT_ID_EXTERNAL_SDVOA: + case ENCODER_OBJECT_ID_EXTERNAL_SDVOB: + case ENCODER_OBJECT_ID_TITFP513: + case ENCODER_OBJECT_ID_VT1623: + case ENCODER_OBJECT_ID_HDMI_SI1930: + case ENCODER_OBJECT_ID_TRAVIS: + case ENCODER_OBJECT_ID_NUTMEG: + /* these are handled by the primary encoders */ + amdgpu_encoder->is_ext_encoder = true; + if (amdgpu_encoder->devices & (ATOM_DEVICE_LCD_SUPPORT)) + drm_encoder_init(dev, encoder, &dce_v6_0_encoder_funcs, + DRM_MODE_ENCODER_LVDS, NULL); + else if (amdgpu_encoder->devices & (ATOM_DEVICE_CRT_SUPPORT)) + drm_encoder_init(dev, encoder, &dce_v6_0_encoder_funcs, + DRM_MODE_ENCODER_DAC, NULL); + else + drm_encoder_init(dev, encoder, &dce_v6_0_encoder_funcs, + DRM_MODE_ENCODER_TMDS, NULL); + drm_encoder_helper_add(encoder, &dce_v6_0_ext_helper_funcs); + break; + } +} + +static const struct amdgpu_display_funcs dce_v6_0_display_funcs = { + .set_vga_render_state = &dce_v6_0_set_vga_render_state, + .bandwidth_update = &dce_v6_0_bandwidth_update, + .vblank_get_counter = &dce_v6_0_vblank_get_counter, + .vblank_wait = &dce_v6_0_vblank_wait, + .is_display_hung = &dce_v6_0_is_display_hung, + .backlight_set_level = &amdgpu_atombios_encoder_set_backlight_level, + .backlight_get_level = &amdgpu_atombios_encoder_get_backlight_level, + .hpd_sense = &dce_v6_0_hpd_sense, + .hpd_set_polarity = &dce_v6_0_hpd_set_polarity, + .hpd_get_gpio_reg = &dce_v6_0_hpd_get_gpio_reg, + .page_flip = &dce_v6_0_page_flip, + .page_flip_get_scanoutpos = &dce_v6_0_crtc_get_scanoutpos, + .add_encoder = &dce_v6_0_encoder_add, + .add_connector = &amdgpu_connector_add, + .stop_mc_access = &dce_v6_0_stop_mc_access, + .resume_mc_access = &dce_v6_0_resume_mc_access, +}; + +static void dce_v6_0_set_display_funcs(struct amdgpu_device *adev) +{ + if (adev->mode_info.funcs == NULL) + adev->mode_info.funcs = &dce_v6_0_display_funcs; +} + +static const struct amdgpu_irq_src_funcs dce_v6_0_crtc_irq_funcs = { + .set = dce_v6_0_set_crtc_interrupt_state, + .process = dce_v6_0_crtc_irq, +}; + +static const struct amdgpu_irq_src_funcs dce_v6_0_pageflip_irq_funcs = { + .set = dce_v6_0_set_pageflip_interrupt_state, + .process = dce_v6_0_pageflip_irq, +}; + +static const struct amdgpu_irq_src_funcs dce_v6_0_hpd_irq_funcs = { + .set = dce_v6_0_set_hpd_interrupt_state, + .process = dce_v6_0_hpd_irq, +}; + +static void dce_v6_0_set_irq_funcs(struct amdgpu_device *adev) +{ + adev->crtc_irq.num_types = AMDGPU_CRTC_IRQ_LAST; + adev->crtc_irq.funcs = &dce_v6_0_crtc_irq_funcs; + + adev->pageflip_irq.num_types = AMDGPU_PAGEFLIP_IRQ_LAST; + adev->pageflip_irq.funcs = &dce_v6_0_pageflip_irq_funcs; + + adev->hpd_irq.num_types = AMDGPU_HPD_LAST; + adev->hpd_irq.funcs = &dce_v6_0_hpd_irq_funcs; +} diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v6_0.h b/drivers/gpu/drm/amd/amdgpu/dce_v6_0.h new file mode 100644 index 000000000000..6a5528105bb6 --- /dev/null +++ b/drivers/gpu/drm/amd/amdgpu/dce_v6_0.h @@ -0,0 +1,29 @@ +/* + * Copyright 2015 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + */ + +#ifndef __DCE_V6_0_H__ +#define __DCE_V6_0_H__ + +extern const struct amd_ip_funcs dce_v6_0_ip_funcs; + +#endif diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c new file mode 100644 index 000000000000..410b29c05671 --- /dev/null +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c @@ -0,0 +1,3233 @@ +/* + * Copyright 2015 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + */ +#include <linux/firmware.h> +#include "amdgpu.h" +#include "amdgpu_ih.h" +#include "amdgpu_gfx.h" +#include "amdgpu_ucode.h" +#include "si/clearstate_si.h" +#include "si/sid.h" + +#define GFX6_NUM_GFX_RINGS 1 +#define GFX6_NUM_COMPUTE_RINGS 2 +#define STATIC_PER_CU_PG_ENABLE (1 << 3) +#define DYN_PER_CU_PG_ENABLE (1 << 2) +#define RLC_SAVE_AND_RESTORE_STARTING_OFFSET 0x90 +#define RLC_CLEAR_STATE_DESCRIPTOR_OFFSET 0x3D + + +static void gfx_v6_0_set_ring_funcs(struct amdgpu_device *adev); +static void gfx_v6_0_set_irq_funcs(struct amdgpu_device *adev); +static void gfx_v6_0_get_cu_info(struct amdgpu_device *adev); + +MODULE_FIRMWARE("radeon/tahiti_pfp.bin"); +MODULE_FIRMWARE("radeon/tahiti_me.bin"); +MODULE_FIRMWARE("radeon/tahiti_ce.bin"); +MODULE_FIRMWARE("radeon/tahiti_rlc.bin"); + +MODULE_FIRMWARE("radeon/pitcairn_pfp.bin"); +MODULE_FIRMWARE("radeon/pitcairn_me.bin"); +MODULE_FIRMWARE("radeon/pitcairn_ce.bin"); +MODULE_FIRMWARE("radeon/pitcairn_rlc.bin"); + +MODULE_FIRMWARE("radeon/verde_pfp.bin"); +MODULE_FIRMWARE("radeon/verde_me.bin"); +MODULE_FIRMWARE("radeon/verde_ce.bin"); +MODULE_FIRMWARE("radeon/verde_rlc.bin"); + +MODULE_FIRMWARE("radeon/oland_pfp.bin"); +MODULE_FIRMWARE("radeon/oland_me.bin"); +MODULE_FIRMWARE("radeon/oland_ce.bin"); +MODULE_FIRMWARE("radeon/oland_rlc.bin"); + +MODULE_FIRMWARE("radeon/hainan_pfp.bin"); +MODULE_FIRMWARE("radeon/hainan_me.bin"); +MODULE_FIRMWARE("radeon/hainan_ce.bin"); +MODULE_FIRMWARE("radeon/hainan_rlc.bin"); + +static u32 gfx_v6_0_get_csb_size(struct amdgpu_device *adev); +static void gfx_v6_0_get_csb_buffer(struct amdgpu_device *adev, volatile u32 *buffer); +//static void gfx_v6_0_init_cp_pg_table(struct amdgpu_device *adev); +static void gfx_v6_0_init_pg(struct amdgpu_device *adev); + + +static const u32 verde_rlc_save_restore_register_list[] = +{ + (0x8000 << 16) | (0x98f4 >> 2), + 0x00000000, + (0x8040 << 16) | (0x98f4 >> 2), + 0x00000000, + (0x8000 << 16) | (0xe80 >> 2), + 0x00000000, + (0x8040 << 16) | (0xe80 >> 2), + 0x00000000, + (0x8000 << 16) | (0x89bc >> 2), + 0x00000000, + (0x8040 << 16) | (0x89bc >> 2), + 0x00000000, + (0x8000 << 16) | (0x8c1c >> 2), + 0x00000000, + (0x8040 << 16) | (0x8c1c >> 2), + 0x00000000, + (0x9c00 << 16) | (0x98f0 >> 2), + 0x00000000, + (0x9c00 << 16) | (0xe7c >> 2), + 0x00000000, + (0x8000 << 16) | (0x9148 >> 2), + 0x00000000, + (0x8040 << 16) | (0x9148 >> 2), + 0x00000000, + (0x9c00 << 16) | (0x9150 >> 2), + 0x00000000, + (0x9c00 << 16) | (0x897c >> 2), + 0x00000000, + (0x9c00 << 16) | (0x8d8c >> 2), + 0x00000000, + (0x9c00 << 16) | (0xac54 >> 2), + 0X00000000, + 0x3, + (0x9c00 << 16) | (0x98f8 >> 2), + 0x00000000, + (0x9c00 << 16) | (0x9910 >> 2), + 0x00000000, + (0x9c00 << 16) | (0x9914 >> 2), + 0x00000000, + (0x9c00 << 16) | (0x9918 >> 2), + 0x00000000, + (0x9c00 << 16) | (0x991c >> 2), + 0x00000000, + (0x9c00 << 16) | (0x9920 >> 2), + 0x00000000, + (0x9c00 << 16) | (0x9924 >> 2), + 0x00000000, + (0x9c00 << 16) | (0x9928 >> 2), + 0x00000000, + (0x9c00 << 16) | (0x992c >> 2), + 0x00000000, + (0x9c00 << 16) | (0x9930 >> 2), + 0x00000000, + (0x9c00 << 16) | (0x9934 >> 2), + 0x00000000, + (0x9c00 << 16) | (0x9938 >> 2), + 0x00000000, + (0x9c00 << 16) | (0x993c >> 2), + 0x00000000, + (0x9c00 << 16) | (0x9940 >> 2), + 0x00000000, + (0x9c00 << 16) | (0x9944 >> 2), + 0x00000000, + (0x9c00 << 16) | (0x9948 >> 2), + 0x00000000, + (0x9c00 << 16) | (0x994c >> 2), + 0x00000000, + (0x9c00 << 16) | (0x9950 >> 2), + 0x00000000, + (0x9c00 << 16) | (0x9954 >> 2), + 0x00000000, + (0x9c00 << 16) | (0x9958 >> 2), + 0x00000000, + (0x9c00 << 16) | (0x995c >> 2), + 0x00000000, + (0x9c00 << 16) | (0x9960 >> 2), + 0x00000000, + (0x9c00 << 16) | (0x9964 >> 2), + 0x00000000, + (0x9c00 << 16) | (0x9968 >> 2), + 0x00000000, + (0x9c00 << 16) | (0x996c >> 2), + 0x00000000, + (0x9c00 << 16) | (0x9970 >> 2), + 0x00000000, + (0x9c00 << 16) | (0x9974 >> 2), + 0x00000000, + (0x9c00 << 16) | (0x9978 >> 2), + 0x00000000, + (0x9c00 << 16) | (0x997c >> 2), + 0x00000000, + (0x9c00 << 16) | (0x9980 >> 2), + 0x00000000, + (0x9c00 << 16) | (0x9984 >> 2), + 0x00000000, + (0x9c00 << 16) | (0x9988 >> 2), + 0x00000000, + (0x9c00 << 16) | (0x998c >> 2), + 0x00000000, + (0x9c00 << 16) | (0x8c00 >> 2), + 0x00000000, + (0x9c00 << 16) | (0x8c14 >> 2), + 0x00000000, + (0x9c00 << 16) | (0x8c04 >> 2), + 0x00000000, + (0x9c00 << 16) | (0x8c08 >> 2), + 0x00000000, + (0x8000 << 16) | (0x9b7c >> 2), + 0x00000000, + (0x8040 << 16) | (0x9b7c >> 2), + 0x00000000, + (0x8000 << 16) | (0xe84 >> 2), + 0x00000000, + (0x8040 << 16) | (0xe84 >> 2), + 0x00000000, + (0x8000 << 16) | (0x89c0 >> 2), + 0x00000000, + (0x8040 << 16) | (0x89c0 >> 2), + 0x00000000, + (0x8000 << 16) | (0x914c >> 2), + 0x00000000, + (0x8040 << 16) | (0x914c >> 2), + 0x00000000, + (0x8000 << 16) | (0x8c20 >> 2), + 0x00000000, + (0x8040 << 16) | (0x8c20 >> 2), + 0x00000000, + (0x8000 << 16) | (0x9354 >> 2), + 0x00000000, + (0x8040 << 16) | (0x9354 >> 2), + 0x00000000, + (0x9c00 << 16) | (0x9060 >> 2), + 0x00000000, + (0x9c00 << 16) | (0x9364 >> 2), + 0x00000000, + (0x9c00 << 16) | (0x9100 >> 2), + 0x00000000, + (0x9c00 << 16) | (0x913c >> 2), + 0x00000000, + (0x8000 << 16) | (0x90e0 >> 2), + 0x00000000, + (0x8000 << 16) | (0x90e4 >> 2), + 0x00000000, + (0x8000 << 16) | (0x90e8 >> 2), + 0x00000000, + (0x8040 << 16) | (0x90e0 >> 2), + 0x00000000, + (0x8040 << 16) | (0x90e4 >> 2), + 0x00000000, + (0x8040 << 16) | (0x90e8 >> 2), + 0x00000000, + (0x9c00 << 16) | (0x8bcc >> 2), + 0x00000000, + (0x9c00 << 16) | (0x8b24 >> 2), + 0x00000000, + (0x9c00 << 16) | (0x88c4 >> 2), + 0x00000000, + (0x9c00 << 16) | (0x8e50 >> 2), + 0x00000000, + (0x9c00 << 16) | (0x8c0c >> 2), + 0x00000000, + (0x9c00 << 16) | (0x8e58 >> 2), + 0x00000000, + (0x9c00 << 16) | (0x8e5c >> 2), + 0x00000000, + (0x9c00 << 16) | (0x9508 >> 2), + 0x00000000, + (0x9c00 << 16) | (0x950c >> 2), + 0x00000000, + (0x9c00 << 16) | (0x9494 >> 2), + 0x00000000, + (0x9c00 << 16) | (0xac0c >> 2), + 0x00000000, + (0x9c00 << 16) | (0xac10 >> 2), + 0x00000000, + (0x9c00 << 16) | (0xac14 >> 2), + 0x00000000, + (0x9c00 << 16) | (0xae00 >> 2), + 0x00000000, + (0x9c00 << 16) | (0xac08 >> 2), + 0x00000000, + (0x9c00 << 16) | (0x88d4 >> 2), + 0x00000000, + (0x9c00 << 16) | (0x88c8 >> 2), + 0x00000000, + (0x9c00 << 16) | (0x88cc >> 2), + 0x00000000, + (0x9c00 << 16) | (0x89b0 >> 2), + 0x00000000, + (0x9c00 << 16) | (0x8b10 >> 2), + 0x00000000, + (0x9c00 << 16) | (0x8a14 >> 2), + 0x00000000, + (0x9c00 << 16) | (0x9830 >> 2), + 0x00000000, + (0x9c00 << 16) | (0x9834 >> 2), + 0x00000000, + (0x9c00 << 16) | (0x9838 >> 2), + 0x00000000, + (0x9c00 << 16) | (0x9a10 >> 2), + 0x00000000, + (0x8000 << 16) | (0x9870 >> 2), + 0x00000000, + (0x8000 << 16) | (0x9874 >> 2), + 0x00000000, + (0x8001 << 16) | (0x9870 >> 2), + 0x00000000, + (0x8001 << 16) | (0x9874 >> 2), + 0x00000000, + (0x8040 << 16) | (0x9870 >> 2), + 0x00000000, + (0x8040 << 16) | (0x9874 >> 2), + 0x00000000, + (0x8041 << 16) | (0x9870 >> 2), + 0x00000000, + (0x8041 << 16) | (0x9874 >> 2), + 0x00000000, + 0x00000000 +}; + +static int gfx_v6_0_init_microcode(struct amdgpu_device *adev) +{ + const char *chip_name; + char fw_name[30]; + int err; + const struct gfx_firmware_header_v1_0 *cp_hdr; + const struct rlc_firmware_header_v1_0 *rlc_hdr; + + DRM_DEBUG("\n"); + + switch (adev->asic_type) { + case CHIP_TAHITI: + chip_name = "tahiti"; + break; + case CHIP_PITCAIRN: + chip_name = "pitcairn"; + break; + case CHIP_VERDE: + chip_name = "verde"; + break; + case CHIP_OLAND: + chip_name = "oland"; + break; + case CHIP_HAINAN: + chip_name = "hainan"; + break; + default: BUG(); + } + + snprintf(fw_name, sizeof(fw_name), "radeon/%s_pfp.bin", chip_name); + err = request_firmware(&adev->gfx.pfp_fw, fw_name, adev->dev); + if (err) + goto out; + err = amdgpu_ucode_validate(adev->gfx.pfp_fw); + if (err) + goto out; + cp_hdr = (const struct gfx_firmware_header_v1_0 *)adev->gfx.pfp_fw->data; + adev->gfx.pfp_fw_version = le32_to_cpu(cp_hdr->header.ucode_version); + adev->gfx.pfp_feature_version = le32_to_cpu(cp_hdr->ucode_feature_version); + + snprintf(fw_name, sizeof(fw_name), "radeon/%s_me.bin", chip_name); + err = request_firmware(&adev->gfx.me_fw, fw_name, adev->dev); + if (err) + goto out; + err = amdgpu_ucode_validate(adev->gfx.me_fw); + if (err) + goto out; + cp_hdr = (const struct gfx_firmware_header_v1_0 *)adev->gfx.me_fw->data; + adev->gfx.me_fw_version = le32_to_cpu(cp_hdr->header.ucode_version); + adev->gfx.me_feature_version = le32_to_cpu(cp_hdr->ucode_feature_version); + + snprintf(fw_name, sizeof(fw_name), "radeon/%s_ce.bin", chip_name); + err = request_firmware(&adev->gfx.ce_fw, fw_name, adev->dev); + if (err) + goto out; + err = amdgpu_ucode_validate(adev->gfx.ce_fw); + if (err) + goto out; + cp_hdr = (const struct gfx_firmware_header_v1_0 *)adev->gfx.ce_fw->data; + adev->gfx.ce_fw_version = le32_to_cpu(cp_hdr->header.ucode_version); + adev->gfx.ce_feature_version = le32_to_cpu(cp_hdr->ucode_feature_version); + + snprintf(fw_name, sizeof(fw_name), "radeon/%s_rlc.bin", chip_name); + err = request_firmware(&adev->gfx.rlc_fw, fw_name, adev->dev); + if (err) + goto out; + err = amdgpu_ucode_validate(adev->gfx.rlc_fw); + rlc_hdr = (const struct rlc_firmware_header_v1_0 *)adev->gfx.rlc_fw->data; + adev->gfx.rlc_fw_version = le32_to_cpu(rlc_hdr->header.ucode_version); + adev->gfx.rlc_feature_version = le32_to_cpu(rlc_hdr->ucode_feature_version); + +out: + if (err) { + printk(KERN_ERR + "gfx6: Failed to load firmware \"%s\"\n", + fw_name); + release_firmware(adev->gfx.pfp_fw); + adev->gfx.pfp_fw = NULL; + release_firmware(adev->gfx.me_fw); + adev->gfx.me_fw = NULL; + release_firmware(adev->gfx.ce_fw); + adev->gfx.ce_fw = NULL; + release_firmware(adev->gfx.rlc_fw); + adev->gfx.rlc_fw = NULL; + } + return err; +} + +static void gfx_v6_0_tiling_mode_table_init(struct amdgpu_device *adev) +{ + const u32 num_tile_mode_states = 32; + u32 reg_offset, gb_tile_moden, split_equal_to_row_size; + + switch (adev->gfx.config.mem_row_size_in_kb) { + case 1: + split_equal_to_row_size = ADDR_SURF_TILE_SPLIT_1KB; + break; + case 2: + default: + split_equal_to_row_size = ADDR_SURF_TILE_SPLIT_2KB; + break; + case 4: + split_equal_to_row_size = ADDR_SURF_TILE_SPLIT_4KB; + break; + } + + if (adev->asic_type == CHIP_VERDE || + adev->asic_type == CHIP_OLAND || + adev->asic_type == CHIP_HAINAN) { + for (reg_offset = 0; reg_offset < num_tile_mode_states; reg_offset++) { + switch (reg_offset) { + case 0: + gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) | + MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) | + PIPE_CONFIG(ADDR_SURF_P4_8x16) | + TILE_SPLIT(ADDR_SURF_TILE_SPLIT_64B) | + NUM_BANKS(ADDR_SURF_16_BANK) | + BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4)); + break; + case 1: + gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) | + MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) | + PIPE_CONFIG(ADDR_SURF_P4_8x16) | + TILE_SPLIT(ADDR_SURF_TILE_SPLIT_128B) | + NUM_BANKS(ADDR_SURF_16_BANK) | + BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4)); + break; + case 2: + gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) | + MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) | + PIPE_CONFIG(ADDR_SURF_P4_8x16) | + TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) | + NUM_BANKS(ADDR_SURF_16_BANK) | + BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4)); + break; + case 3: + gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) | + MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) | + PIPE_CONFIG(ADDR_SURF_P4_8x16) | + TILE_SPLIT(ADDR_SURF_TILE_SPLIT_128B) | + NUM_BANKS(ADDR_SURF_16_BANK) | + BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4)); + break; + case 4: + gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) | + MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) | + PIPE_CONFIG(ADDR_SURF_P4_8x16) | + TILE_SPLIT(ADDR_SURF_TILE_SPLIT_64B) | + NUM_BANKS(ADDR_SURF_16_BANK) | + BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2)); + break; + case 5: + gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) | + MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) | + PIPE_CONFIG(ADDR_SURF_P4_8x16) | + TILE_SPLIT(split_equal_to_row_size) | + NUM_BANKS(ADDR_SURF_16_BANK) | + BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2)); + break; + case 6: + gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) | + MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) | + PIPE_CONFIG(ADDR_SURF_P4_8x16) | + TILE_SPLIT(split_equal_to_row_size) | + NUM_BANKS(ADDR_SURF_16_BANK) | + BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2)); + break; + case 7: + gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) | + MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) | + PIPE_CONFIG(ADDR_SURF_P4_8x16) | + TILE_SPLIT(split_equal_to_row_size) | + NUM_BANKS(ADDR_SURF_16_BANK) | + BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4)); + break; + case 8: + gb_tile_moden = (ARRAY_MODE(ARRAY_LINEAR_ALIGNED) | + MICRO_TILE_MODE(ADDR_SURF_DISPLAY_MICRO_TILING) | + PIPE_CONFIG(ADDR_SURF_P4_8x16) | + TILE_SPLIT(ADDR_SURF_TILE_SPLIT_64B) | + NUM_BANKS(ADDR_SURF_16_BANK) | + BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2)); + break; + case 9: + gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) | + MICRO_TILE_MODE(ADDR_SURF_DISPLAY_MICRO_TILING) | + PIPE_CONFIG(ADDR_SURF_P4_8x16) | + TILE_SPLIT(ADDR_SURF_TILE_SPLIT_64B) | + NUM_BANKS(ADDR_SURF_16_BANK) | + BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2)); + break; + case 10: + gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) | + MICRO_TILE_MODE(ADDR_SURF_DISPLAY_MICRO_TILING) | + PIPE_CONFIG(ADDR_SURF_P4_8x16) | + TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) | + NUM_BANKS(ADDR_SURF_16_BANK) | + BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4)); + break; + case 11: + gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) | + MICRO_TILE_MODE(ADDR_SURF_DISPLAY_MICRO_TILING) | + PIPE_CONFIG(ADDR_SURF_P4_8x16) | + TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) | + NUM_BANKS(ADDR_SURF_16_BANK) | + BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2)); + break; + case 12: + gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) | + MICRO_TILE_MODE(ADDR_SURF_DISPLAY_MICRO_TILING) | + PIPE_CONFIG(ADDR_SURF_P4_8x16) | + TILE_SPLIT(ADDR_SURF_TILE_SPLIT_512B) | + NUM_BANKS(ADDR_SURF_16_BANK) | + BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2)); + break; + case 13: + gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) | + MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) | + PIPE_CONFIG(ADDR_SURF_P4_8x16) | + TILE_SPLIT(ADDR_SURF_TILE_SPLIT_64B) | + NUM_BANKS(ADDR_SURF_16_BANK) | + BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2)); + break; + case 14: + gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) | + MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) | + PIPE_CONFIG(ADDR_SURF_P4_8x16) | + TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) | + NUM_BANKS(ADDR_SURF_16_BANK) | + BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2)); + break; + case 15: + gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) | + MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) | + PIPE_CONFIG(ADDR_SURF_P4_8x16) | + TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) | + NUM_BANKS(ADDR_SURF_16_BANK) | + BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2)); + break; + case 16: + gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) | + MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) | + PIPE_CONFIG(ADDR_SURF_P4_8x16) | + TILE_SPLIT(ADDR_SURF_TILE_SPLIT_512B) | + NUM_BANKS(ADDR_SURF_16_BANK) | + BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2)); + break; + case 17: + gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) | + MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) | + PIPE_CONFIG(ADDR_SURF_P4_8x16) | + TILE_SPLIT(split_equal_to_row_size) | + NUM_BANKS(ADDR_SURF_16_BANK) | + BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2)); + break; + case 21: + gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) | + MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) | + PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) | + TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) | + NUM_BANKS(ADDR_SURF_16_BANK) | + BANK_WIDTH(ADDR_SURF_BANK_WIDTH_2) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2)); + break; + case 22: + gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) | + MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) | + PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) | + TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) | + NUM_BANKS(ADDR_SURF_16_BANK) | + BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4)); + break; + case 23: + gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) | + MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) | + PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) | + TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) | + NUM_BANKS(ADDR_SURF_16_BANK) | + BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2)); + break; + case 24: + gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) | + MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) | + PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) | + TILE_SPLIT(ADDR_SURF_TILE_SPLIT_512B) | + NUM_BANKS(ADDR_SURF_16_BANK) | + BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2)); + break; + case 25: + gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) | + MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) | + PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) | + TILE_SPLIT(ADDR_SURF_TILE_SPLIT_1KB) | + NUM_BANKS(ADDR_SURF_8_BANK) | + BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1)); + break; + default: + gb_tile_moden = 0; + break; + } + adev->gfx.config.tile_mode_array[reg_offset] = gb_tile_moden; + WREG32(GB_TILE_MODE0 + reg_offset, gb_tile_moden); + } + } else if ((adev->asic_type == CHIP_TAHITI) || (adev->asic_type == CHIP_PITCAIRN)) { + for (reg_offset = 0; reg_offset < num_tile_mode_states; reg_offset++) { + switch (reg_offset) { + case 0: /* non-AA compressed depth or any compressed stencil */ + gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) | + MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) | + PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) | + TILE_SPLIT(ADDR_SURF_TILE_SPLIT_64B) | + NUM_BANKS(ADDR_SURF_16_BANK) | + BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2)); + break; + case 1: /* 2xAA/4xAA compressed depth only */ + gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) | + MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) | + PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) | + TILE_SPLIT(ADDR_SURF_TILE_SPLIT_128B) | + NUM_BANKS(ADDR_SURF_16_BANK) | + BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2)); + break; + case 2: /* 8xAA compressed depth only */ + gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) | + MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) | + PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) | + TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) | + NUM_BANKS(ADDR_SURF_16_BANK) | + BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2)); + break; + case 3: /* 2xAA/4xAA compressed depth with stencil (for depth buffer) */ + gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) | + MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) | + PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) | + TILE_SPLIT(ADDR_SURF_TILE_SPLIT_128B) | + NUM_BANKS(ADDR_SURF_16_BANK) | + BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2)); + break; + case 4: /* Maps w/ a dimension less than the 2D macro-tile dimensions (for mipmapped depth textures) */ + gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) | + MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) | + PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) | + TILE_SPLIT(ADDR_SURF_TILE_SPLIT_64B) | + NUM_BANKS(ADDR_SURF_16_BANK) | + BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2)); + break; + case 5: /* Uncompressed 16bpp depth - and stencil buffer allocated with it */ + gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) | + MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) | + PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) | + TILE_SPLIT(split_equal_to_row_size) | + NUM_BANKS(ADDR_SURF_16_BANK) | + BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2)); + break; + case 6: /* Uncompressed 32bpp depth - and stencil buffer allocated with it */ + gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) | + MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) | + PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) | + TILE_SPLIT(split_equal_to_row_size) | + NUM_BANKS(ADDR_SURF_16_BANK) | + BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1)); + break; + case 7: /* Uncompressed 8bpp stencil without depth (drivers typically do not use) */ + gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) | + MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) | + PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) | + TILE_SPLIT(split_equal_to_row_size) | + NUM_BANKS(ADDR_SURF_16_BANK) | + BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2)); + break; + case 8: /* 1D and 1D Array Surfaces */ + gb_tile_moden = (ARRAY_MODE(ARRAY_LINEAR_ALIGNED) | + MICRO_TILE_MODE(ADDR_SURF_DISPLAY_MICRO_TILING) | + PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) | + TILE_SPLIT(ADDR_SURF_TILE_SPLIT_64B) | + NUM_BANKS(ADDR_SURF_16_BANK) | + BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2)); + break; + case 9: /* Displayable maps. */ + gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) | + MICRO_TILE_MODE(ADDR_SURF_DISPLAY_MICRO_TILING) | + PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) | + TILE_SPLIT(ADDR_SURF_TILE_SPLIT_64B) | + NUM_BANKS(ADDR_SURF_16_BANK) | + BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2)); + break; + case 10: /* Display 8bpp. */ + gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) | + MICRO_TILE_MODE(ADDR_SURF_DISPLAY_MICRO_TILING) | + PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) | + TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) | + NUM_BANKS(ADDR_SURF_16_BANK) | + BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2)); + break; + case 11: /* Display 16bpp. */ + gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) | + MICRO_TILE_MODE(ADDR_SURF_DISPLAY_MICRO_TILING) | + PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) | + TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) | + NUM_BANKS(ADDR_SURF_16_BANK) | + BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2)); + break; + case 12: /* Display 32bpp. */ + gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) | + MICRO_TILE_MODE(ADDR_SURF_DISPLAY_MICRO_TILING) | + PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) | + TILE_SPLIT(ADDR_SURF_TILE_SPLIT_512B) | + NUM_BANKS(ADDR_SURF_16_BANK) | + BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1)); + break; + case 13: /* Thin. */ + gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) | + MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) | + PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) | + TILE_SPLIT(ADDR_SURF_TILE_SPLIT_64B) | + NUM_BANKS(ADDR_SURF_16_BANK) | + BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2)); + break; + case 14: /* Thin 8 bpp. */ + gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) | + MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) | + PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) | + TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) | + NUM_BANKS(ADDR_SURF_16_BANK) | + BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1)); + break; + case 15: /* Thin 16 bpp. */ + gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) | + MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) | + PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) | + TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) | + NUM_BANKS(ADDR_SURF_16_BANK) | + BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1)); + break; + case 16: /* Thin 32 bpp. */ + gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) | + MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) | + PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) | + TILE_SPLIT(ADDR_SURF_TILE_SPLIT_512B) | + NUM_BANKS(ADDR_SURF_16_BANK) | + BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1)); + break; + case 17: /* Thin 64 bpp. */ + gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) | + MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) | + PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) | + TILE_SPLIT(split_equal_to_row_size) | + NUM_BANKS(ADDR_SURF_16_BANK) | + BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1)); + break; + case 21: /* 8 bpp PRT. */ + gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) | + MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) | + PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) | + TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) | + NUM_BANKS(ADDR_SURF_16_BANK) | + BANK_WIDTH(ADDR_SURF_BANK_WIDTH_2) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2)); + break; + case 22: /* 16 bpp PRT */ + gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) | + MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) | + PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) | + TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) | + NUM_BANKS(ADDR_SURF_16_BANK) | + BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4)); + break; + case 23: /* 32 bpp PRT */ + gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) | + MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) | + PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) | + TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) | + NUM_BANKS(ADDR_SURF_16_BANK) | + BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2)); + break; + case 24: /* 64 bpp PRT */ + gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) | + MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) | + PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) | + TILE_SPLIT(ADDR_SURF_TILE_SPLIT_512B) | + NUM_BANKS(ADDR_SURF_16_BANK) | + BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2)); + break; + case 25: /* 128 bpp PRT */ + gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) | + MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) | + PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) | + TILE_SPLIT(ADDR_SURF_TILE_SPLIT_1KB) | + NUM_BANKS(ADDR_SURF_8_BANK) | + BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) | + BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) | + MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1)); + break; + default: + gb_tile_moden = 0; + break; + } + adev->gfx.config.tile_mode_array[reg_offset] = gb_tile_moden; + WREG32(GB_TILE_MODE0 + reg_offset, gb_tile_moden); + } + } else{ + + DRM_ERROR("unknown asic: 0x%x\n", adev->asic_type); + } + +} + +static void gfx_v6_0_select_se_sh(struct amdgpu_device *adev, u32 se_num, + u32 sh_num, u32 instance) +{ + u32 data; + + if (instance == 0xffffffff) + data = INSTANCE_BROADCAST_WRITES; + else + data = INSTANCE_INDEX(instance); + + if ((se_num == 0xffffffff) && (sh_num == 0xffffffff)) + data |= SH_BROADCAST_WRITES | SE_BROADCAST_WRITES; + else if (se_num == 0xffffffff) + data |= SE_BROADCAST_WRITES | SH_INDEX(sh_num); + else if (sh_num == 0xffffffff) + data |= SH_BROADCAST_WRITES | SE_INDEX(se_num); + else + data |= SH_INDEX(sh_num) | SE_INDEX(se_num); + WREG32(GRBM_GFX_INDEX, data); +} + +static u32 gfx_v6_0_create_bitmask(u32 bit_width) +{ + return (u32)(((u64)1 << bit_width) - 1); +} + +static u32 gfx_v6_0_get_rb_disabled(struct amdgpu_device *adev, + u32 max_rb_num_per_se, + u32 sh_per_se) +{ + u32 data, mask; + + data = RREG32(CC_RB_BACKEND_DISABLE); + data &= BACKEND_DISABLE_MASK; + data |= RREG32(GC_USER_RB_BACKEND_DISABLE); + + data >>= BACKEND_DISABLE_SHIFT; + + mask = gfx_v6_0_create_bitmask(max_rb_num_per_se / sh_per_se); + + return data & mask; +} + +static void gfx_v6_0_setup_rb(struct amdgpu_device *adev, + u32 se_num, u32 sh_per_se, + u32 max_rb_num_per_se) +{ + int i, j; + u32 data, mask; + u32 disabled_rbs = 0; + u32 enabled_rbs = 0; + + mutex_lock(&adev->grbm_idx_mutex); + for (i = 0; i < se_num; i++) { + for (j = 0; j < sh_per_se; j++) { + gfx_v6_0_select_se_sh(adev, i, j, 0xffffffff); + data = gfx_v6_0_get_rb_disabled(adev, max_rb_num_per_se, sh_per_se); + disabled_rbs |= data << ((i * sh_per_se + j) * TAHITI_RB_BITMAP_WIDTH_PER_SH); + } + } + gfx_v6_0_select_se_sh(adev, 0xffffffff, 0xffffffff, 0xffffffff); + mutex_unlock(&adev->grbm_idx_mutex); + + mask = 1; + for (i = 0; i < max_rb_num_per_se * se_num; i++) { + if (!(disabled_rbs & mask)) + enabled_rbs |= mask; + mask <<= 1; + } + + adev->gfx.config.backend_enable_mask = enabled_rbs; + adev->gfx.config.num_rbs = hweight32(enabled_rbs); + + mutex_lock(&adev->grbm_idx_mutex); + for (i = 0; i < se_num; i++) { + gfx_v6_0_select_se_sh(adev, i, 0xffffffff, 0xffffffff); + data = 0; + for (j = 0; j < sh_per_se; j++) { + switch (enabled_rbs & 3) { + case 1: + data |= (RASTER_CONFIG_RB_MAP_0 << (i * sh_per_se + j) * 2); + break; + case 2: + data |= (RASTER_CONFIG_RB_MAP_3 << (i * sh_per_se + j) * 2); + break; + case 3: + default: + data |= (RASTER_CONFIG_RB_MAP_2 << (i * sh_per_se + j) * 2); + break; + } + enabled_rbs >>= 2; + } + WREG32(PA_SC_RASTER_CONFIG, data); + } + gfx_v6_0_select_se_sh(adev, 0xffffffff, 0xffffffff, 0xffffffff); + mutex_unlock(&adev->grbm_idx_mutex); +} +/* +static void gmc_v6_0_init_compute_vmid(struct amdgpu_device *adev) +{ +} +*/ + +static u32 gfx_v6_0_get_cu_enabled(struct amdgpu_device *adev, u32 cu_per_sh) +{ + u32 data, mask; + + data = RREG32(CC_GC_SHADER_ARRAY_CONFIG); + data &= INACTIVE_CUS_MASK; + data |= RREG32(GC_USER_SHADER_ARRAY_CONFIG); + + data >>= INACTIVE_CUS_SHIFT; + + mask = gfx_v6_0_create_bitmask(cu_per_sh); + + return ~data & mask; +} + + +static void gfx_v6_0_setup_spi(struct amdgpu_device *adev, + u32 se_num, u32 sh_per_se, + u32 cu_per_sh) +{ + int i, j, k; + u32 data, mask; + u32 active_cu = 0; + + mutex_lock(&adev->grbm_idx_mutex); + for (i = 0; i < se_num; i++) { + for (j = 0; j < sh_per_se; j++) { + gfx_v6_0_select_se_sh(adev, i, j, 0xffffffff); + data = RREG32(SPI_STATIC_THREAD_MGMT_3); + active_cu = gfx_v6_0_get_cu_enabled(adev, cu_per_sh); + + mask = 1; + for (k = 0; k < 16; k++) { + mask <<= k; + if (active_cu & mask) { + data &= ~mask; + WREG32(SPI_STATIC_THREAD_MGMT_3, data); + break; + } + } + } + } + gfx_v6_0_select_se_sh(adev, 0xffffffff, 0xffffffff, 0xffffffff); + mutex_unlock(&adev->grbm_idx_mutex); +} + +static void gfx_v6_0_gpu_init(struct amdgpu_device *adev) +{ + u32 gb_addr_config = 0; + u32 mc_shared_chmap, mc_arb_ramcfg; + u32 sx_debug_1; + u32 hdp_host_path_cntl; + u32 tmp; + + switch (adev->asic_type) { + case CHIP_TAHITI: + adev->gfx.config.max_shader_engines = 2; + adev->gfx.config.max_tile_pipes = 12; + adev->gfx.config.max_cu_per_sh = 8; + adev->gfx.config.max_sh_per_se = 2; + adev->gfx.config.max_backends_per_se = 4; + adev->gfx.config.max_texture_channel_caches = 12; + adev->gfx.config.max_gprs = 256; + adev->gfx.config.max_gs_threads = 32; + adev->gfx.config.max_hw_contexts = 8; + + adev->gfx.config.sc_prim_fifo_size_frontend = 0x20; + adev->gfx.config.sc_prim_fifo_size_backend = 0x100; + adev->gfx.config.sc_hiz_tile_fifo_size = 0x30; + adev->gfx.config.sc_earlyz_tile_fifo_size = 0x130; + gb_addr_config = TAHITI_GB_ADDR_CONFIG_GOLDEN; + break; + case CHIP_PITCAIRN: + adev->gfx.config.max_shader_engines = 2; + adev->gfx.config.max_tile_pipes = 8; + adev->gfx.config.max_cu_per_sh = 5; + adev->gfx.config.max_sh_per_se = 2; + adev->gfx.config.max_backends_per_se = 4; + adev->gfx.config.max_texture_channel_caches = 8; + adev->gfx.config.max_gprs = 256; + adev->gfx.config.max_gs_threads = 32; + adev->gfx.config.max_hw_contexts = 8; + + adev->gfx.config.sc_prim_fifo_size_frontend = 0x20; + adev->gfx.config.sc_prim_fifo_size_backend = 0x100; + adev->gfx.config.sc_hiz_tile_fifo_size = 0x30; + adev->gfx.config.sc_earlyz_tile_fifo_size = 0x130; + gb_addr_config = TAHITI_GB_ADDR_CONFIG_GOLDEN; + break; + + case CHIP_VERDE: + adev->gfx.config.max_shader_engines = 1; + adev->gfx.config.max_tile_pipes = 4; + adev->gfx.config.max_cu_per_sh = 5; + adev->gfx.config.max_sh_per_se = 2; + adev->gfx.config.max_backends_per_se = 4; + adev->gfx.config.max_texture_channel_caches = 4; + adev->gfx.config.max_gprs = 256; + adev->gfx.config.max_gs_threads = 32; + adev->gfx.config.max_hw_contexts = 8; + + adev->gfx.config.sc_prim_fifo_size_frontend = 0x20; + adev->gfx.config.sc_prim_fifo_size_backend = 0x40; + adev->gfx.config.sc_hiz_tile_fifo_size = 0x30; + adev->gfx.config.sc_earlyz_tile_fifo_size = 0x130; + gb_addr_config = VERDE_GB_ADDR_CONFIG_GOLDEN; + break; + case CHIP_OLAND: + adev->gfx.config.max_shader_engines = 1; + adev->gfx.config.max_tile_pipes = 4; + adev->gfx.config.max_cu_per_sh = 6; + adev->gfx.config.max_sh_per_se = 1; + adev->gfx.config.max_backends_per_se = 2; + adev->gfx.config.max_texture_channel_caches = 4; + adev->gfx.config.max_gprs = 256; + adev->gfx.config.max_gs_threads = 16; + adev->gfx.config.max_hw_contexts = 8; + + adev->gfx.config.sc_prim_fifo_size_frontend = 0x20; + adev->gfx.config.sc_prim_fifo_size_backend = 0x40; + adev->gfx.config.sc_hiz_tile_fifo_size = 0x30; + adev->gfx.config.sc_earlyz_tile_fifo_size = 0x130; + gb_addr_config = VERDE_GB_ADDR_CONFIG_GOLDEN; + break; + case CHIP_HAINAN: + adev->gfx.config.max_shader_engines = 1; + adev->gfx.config.max_tile_pipes = 4; + adev->gfx.config.max_cu_per_sh = 5; + adev->gfx.config.max_sh_per_se = 1; + adev->gfx.config.max_backends_per_se = 1; + adev->gfx.config.max_texture_channel_caches = 2; + adev->gfx.config.max_gprs = 256; + adev->gfx.config.max_gs_threads = 16; + adev->gfx.config.max_hw_contexts = 8; + + adev->gfx.config.sc_prim_fifo_size_frontend = 0x20; + adev->gfx.config.sc_prim_fifo_size_backend = 0x40; + adev->gfx.config.sc_hiz_tile_fifo_size = 0x30; + adev->gfx.config.sc_earlyz_tile_fifo_size = 0x130; + gb_addr_config = HAINAN_GB_ADDR_CONFIG_GOLDEN; + break; + default: + BUG(); + break; + } + + WREG32(GRBM_CNTL, GRBM_READ_TIMEOUT(0xff)); + WREG32(SRBM_INT_CNTL, 1); + WREG32(SRBM_INT_ACK, 1); + + WREG32(BIF_FB_EN, FB_READ_EN | FB_WRITE_EN); + + mc_shared_chmap = RREG32(MC_SHARED_CHMAP); + mc_arb_ramcfg = RREG32(MC_ARB_RAMCFG); + + adev->gfx.config.num_tile_pipes = adev->gfx.config.max_tile_pipes; + adev->gfx.config.mem_max_burst_length_bytes = 256; + tmp = (mc_arb_ramcfg & NOOFCOLS_MASK) >> NOOFCOLS_SHIFT; + adev->gfx.config.mem_row_size_in_kb = (4 * (1 << (8 + tmp))) / 1024; + if (adev->gfx.config.mem_row_size_in_kb > 4) + adev->gfx.config.mem_row_size_in_kb = 4; + adev->gfx.config.shader_engine_tile_size = 32; + adev->gfx.config.num_gpus = 1; + adev->gfx.config.multi_gpu_tile_size = 64; + + gb_addr_config &= ~ROW_SIZE_MASK; + switch (adev->gfx.config.mem_row_size_in_kb) { + case 1: + default: + gb_addr_config |= ROW_SIZE(0); + break; + case 2: + gb_addr_config |= ROW_SIZE(1); + break; + case 4: + gb_addr_config |= ROW_SIZE(2); + break; + } + adev->gfx.config.gb_addr_config = gb_addr_config; + + WREG32(GB_ADDR_CONFIG, gb_addr_config); + WREG32(DMIF_ADDR_CONFIG, gb_addr_config); + WREG32(DMIF_ADDR_CALC, gb_addr_config); + WREG32(HDP_ADDR_CONFIG, gb_addr_config); + WREG32(DMA_TILING_CONFIG + DMA0_REGISTER_OFFSET, gb_addr_config); + WREG32(DMA_TILING_CONFIG + DMA1_REGISTER_OFFSET, gb_addr_config); +#if 0 + if (adev->has_uvd) { + WREG32(UVD_UDEC_ADDR_CONFIG, gb_addr_config); + WREG32(UVD_UDEC_DB_ADDR_CONFIG, gb_addr_config); + WREG32(UVD_UDEC_DBW_ADDR_CONFIG, gb_addr_config); + } +#endif + gfx_v6_0_tiling_mode_table_init(adev); + + gfx_v6_0_setup_rb(adev, adev->gfx.config.max_shader_engines, + adev->gfx.config.max_sh_per_se, + adev->gfx.config.max_backends_per_se); + + gfx_v6_0_setup_spi(adev, adev->gfx.config.max_shader_engines, + adev->gfx.config.max_sh_per_se, + adev->gfx.config.max_cu_per_sh); + + gfx_v6_0_get_cu_info(adev); + + WREG32(CP_QUEUE_THRESHOLDS, (ROQ_IB1_START(0x16) | + ROQ_IB2_START(0x2b))); + WREG32(CP_MEQ_THRESHOLDS, MEQ1_START(0x30) | MEQ2_START(0x60)); + + sx_debug_1 = RREG32(SX_DEBUG_1); + WREG32(SX_DEBUG_1, sx_debug_1); + + WREG32(SPI_CONFIG_CNTL_1, VTX_DONE_DELAY(4)); + + WREG32(PA_SC_FIFO_SIZE, (SC_FRONTEND_PRIM_FIFO_SIZE(adev->gfx.config.sc_prim_fifo_size_frontend) | + SC_BACKEND_PRIM_FIFO_SIZE(adev->gfx.config.sc_prim_fifo_size_backend) | + SC_HIZ_TILE_FIFO_SIZE(adev->gfx.config.sc_hiz_tile_fifo_size) | + SC_EARLYZ_TILE_FIFO_SIZE(adev->gfx.config.sc_earlyz_tile_fifo_size))); + + WREG32(VGT_NUM_INSTANCES, 1); + WREG32(CP_PERFMON_CNTL, 0); + WREG32(SQ_CONFIG, 0); + WREG32(PA_SC_FORCE_EOV_MAX_CNTS, (FORCE_EOV_MAX_CLK_CNT(4095) | + FORCE_EOV_MAX_REZ_CNT(255))); + + WREG32(VGT_CACHE_INVALIDATION, CACHE_INVALIDATION(VC_AND_TC) | + AUTO_INVLD_EN(ES_AND_GS_AUTO)); + + WREG32(VGT_GS_VERTEX_REUSE, 16); + WREG32(PA_SC_LINE_STIPPLE_STATE, 0); + + WREG32(CB_PERFCOUNTER0_SELECT0, 0); + WREG32(CB_PERFCOUNTER0_SELECT1, 0); + WREG32(CB_PERFCOUNTER1_SELECT0, 0); + WREG32(CB_PERFCOUNTER1_SELECT1, 0); + WREG32(CB_PERFCOUNTER2_SELECT0, 0); + WREG32(CB_PERFCOUNTER2_SELECT1, 0); + WREG32(CB_PERFCOUNTER3_SELECT0, 0); + WREG32(CB_PERFCOUNTER3_SELECT1, 0); + + hdp_host_path_cntl = RREG32(HDP_HOST_PATH_CNTL); + WREG32(HDP_HOST_PATH_CNTL, hdp_host_path_cntl); + + WREG32(PA_CL_ENHANCE, CLIP_VTX_REORDER_ENA | NUM_CLIP_SEQ(3)); + + udelay(50); +} + + +static void gfx_v6_0_scratch_init(struct amdgpu_device *adev) +{ + int i; + + adev->gfx.scratch.num_reg = 7; + adev->gfx.scratch.reg_base = SCRATCH_REG0; + for (i = 0; i < adev->gfx.scratch.num_reg; i++) { + adev->gfx.scratch.free[i] = true; + adev->gfx.scratch.reg[i] = adev->gfx.scratch.reg_base + i; + } +} + +static int gfx_v6_0_ring_test_ring(struct amdgpu_ring *ring) +{ + struct amdgpu_device *adev = ring->adev; + uint32_t scratch; + uint32_t tmp = 0; + unsigned i; + int r; + + r = amdgpu_gfx_scratch_get(adev, &scratch); + if (r) { + DRM_ERROR("amdgpu: cp failed to get scratch reg (%d).\n", r); + return r; + } + WREG32(scratch, 0xCAFEDEAD); + + r = amdgpu_ring_alloc(ring, 3); + if (r) { + DRM_ERROR("amdgpu: cp failed to lock ring %d (%d).\n", ring->idx, r); + amdgpu_gfx_scratch_free(adev, scratch); + return r; + } + amdgpu_ring_write(ring, PACKET3(PACKET3_SET_CONFIG_REG, 1)); + amdgpu_ring_write(ring, (scratch - PACKET3_SET_CONFIG_REG_START)); + amdgpu_ring_write(ring, 0xDEADBEEF); + amdgpu_ring_commit(ring); + + for (i = 0; i < adev->usec_timeout; i++) { + tmp = RREG32(scratch); + if (tmp == 0xDEADBEEF) + break; + DRM_UDELAY(1); + } + if (i < adev->usec_timeout) { + DRM_INFO("ring test on %d succeeded in %d usecs\n", ring->idx, i); + } else { + DRM_ERROR("amdgpu: ring %d test failed (scratch(0x%04X)=0x%08X)\n", + ring->idx, scratch, tmp); + r = -EINVAL; + } + amdgpu_gfx_scratch_free(adev, scratch); + return r; +} + +static void gfx_v6_0_ring_emit_hdp_flush(struct amdgpu_ring *ring) +{ + /* flush hdp cache */ + amdgpu_ring_write(ring, PACKET3(PACKET3_WRITE_DATA, 3)); + amdgpu_ring_write(ring, (WRITE_DATA_ENGINE_SEL(1) | + WRITE_DATA_DST_SEL(0))); + amdgpu_ring_write(ring, HDP_MEM_COHERENCY_FLUSH_CNTL); + amdgpu_ring_write(ring, 0); + amdgpu_ring_write(ring, 0x1); +} + +/** + * gfx_v6_0_ring_emit_hdp_invalidate - emit an hdp invalidate on the cp + * + * @adev: amdgpu_device pointer + * @ridx: amdgpu ring index + * + * Emits an hdp invalidate on the cp. + */ +static void gfx_v6_0_ring_emit_hdp_invalidate(struct amdgpu_ring *ring) +{ + amdgpu_ring_write(ring, PACKET3(PACKET3_WRITE_DATA, 3)); + amdgpu_ring_write(ring, (WRITE_DATA_ENGINE_SEL(1) | + WRITE_DATA_DST_SEL(0))); + amdgpu_ring_write(ring, HDP_DEBUG0); + amdgpu_ring_write(ring, 0); + amdgpu_ring_write(ring, 0x1); +} + +static void gfx_v6_0_ring_emit_fence(struct amdgpu_ring *ring, u64 addr, + u64 seq, unsigned flags) +{ + bool write64bit = flags & AMDGPU_FENCE_FLAG_64BIT; + bool int_sel = flags & AMDGPU_FENCE_FLAG_INT; + /* flush read cache over gart */ + amdgpu_ring_write(ring, PACKET3(PACKET3_SET_CONFIG_REG, 1)); + amdgpu_ring_write(ring, (CP_COHER_CNTL2 - PACKET3_SET_CONFIG_REG_START)); + amdgpu_ring_write(ring, 0); + amdgpu_ring_write(ring, PACKET3(PACKET3_SURFACE_SYNC, 3)); + amdgpu_ring_write(ring, PACKET3_TCL1_ACTION_ENA | + PACKET3_TC_ACTION_ENA | + PACKET3_SH_KCACHE_ACTION_ENA | + PACKET3_SH_ICACHE_ACTION_ENA); + amdgpu_ring_write(ring, 0xFFFFFFFF); + amdgpu_ring_write(ring, 0); + amdgpu_ring_write(ring, 10); /* poll interval */ + /* EVENT_WRITE_EOP - flush caches, send int */ + amdgpu_ring_write(ring, PACKET3(PACKET3_EVENT_WRITE_EOP, 4)); + amdgpu_ring_write(ring, EVENT_TYPE(CACHE_FLUSH_AND_INV_TS_EVENT) | EVENT_INDEX(5)); + amdgpu_ring_write(ring, addr & 0xfffffffc); + amdgpu_ring_write(ring, (upper_32_bits(addr) & 0xffff) | + DATA_SEL(write64bit ? 2 : 1) | INT_SEL(int_sel ? 2 : 0)); + amdgpu_ring_write(ring, lower_32_bits(seq)); + amdgpu_ring_write(ring, upper_32_bits(seq)); +} + +static void gfx_v6_0_ring_emit_ib(struct amdgpu_ring *ring, + struct amdgpu_ib *ib, + unsigned vm_id, bool ctx_switch) +{ + u32 header, control = 0; + + /* insert SWITCH_BUFFER packet before first IB in the ring frame */ + if (ctx_switch) { + amdgpu_ring_write(ring, PACKET3(PACKET3_SWITCH_BUFFER, 0)); + amdgpu_ring_write(ring, 0); + } + + if (ib->flags & AMDGPU_IB_FLAG_CE) + header = PACKET3(PACKET3_INDIRECT_BUFFER_CONST, 2); + else + header = PACKET3(PACKET3_INDIRECT_BUFFER, 2); + + control |= ib->length_dw | (vm_id << 24); + + amdgpu_ring_write(ring, header); + amdgpu_ring_write(ring, +#ifdef __BIG_ENDIAN + (2 << 0) | +#endif + (ib->gpu_addr & 0xFFFFFFFC)); + amdgpu_ring_write(ring, upper_32_bits(ib->gpu_addr) & 0xFFFF); + amdgpu_ring_write(ring, control); +} + +/** + * gfx_v6_0_ring_test_ib - basic ring IB test + * + * @ring: amdgpu_ring structure holding ring information + * + * Allocate an IB and execute it on the gfx ring (SI). + * Provides a basic gfx ring test to verify that IBs are working. + * Returns 0 on success, error on failure. + */ +static int gfx_v6_0_ring_test_ib(struct amdgpu_ring *ring, long timeout) +{ + struct amdgpu_device *adev = ring->adev; + struct amdgpu_ib ib; + struct fence *f = NULL; + uint32_t scratch; + uint32_t tmp = 0; + long r; + + r = amdgpu_gfx_scratch_get(adev, &scratch); + if (r) { + DRM_ERROR("amdgpu: failed to get scratch reg (%ld).\n", r); + return r; + } + WREG32(scratch, 0xCAFEDEAD); + memset(&ib, 0, sizeof(ib)); + r = amdgpu_ib_get(adev, NULL, 256, &ib); + if (r) { + DRM_ERROR("amdgpu: failed to get ib (%ld).\n", r); + goto err1; + } + ib.ptr[0] = PACKET3(PACKET3_SET_CONFIG_REG, 1); + ib.ptr[1] = ((scratch - PACKET3_SET_CONFIG_REG_START)); + ib.ptr[2] = 0xDEADBEEF; + ib.length_dw = 3; + + r = amdgpu_ib_schedule(ring, 1, &ib, NULL, NULL, &f); + if (r) + goto err2; + + r = fence_wait_timeout(f, false, timeout); + if (r == 0) { + DRM_ERROR("amdgpu: IB test timed out\n"); + r = -ETIMEDOUT; + goto err2; + } else if (r < 0) { + DRM_ERROR("amdgpu: fence wait failed (%ld).\n", r); + goto err2; + } + tmp = RREG32(scratch); + if (tmp == 0xDEADBEEF) { + DRM_INFO("ib test on ring %d succeeded\n", ring->idx); + r = 0; + } else { + DRM_ERROR("amdgpu: ib test failed (scratch(0x%04X)=0x%08X)\n", + scratch, tmp); + r = -EINVAL; + } + +err2: + amdgpu_ib_free(adev, &ib, NULL); + fence_put(f); +err1: + amdgpu_gfx_scratch_free(adev, scratch); + return r; +} + +static void gfx_v6_0_cp_gfx_enable(struct amdgpu_device *adev, bool enable) +{ + int i; + if (enable) + WREG32(CP_ME_CNTL, 0); + else { + WREG32(CP_ME_CNTL, (CP_ME_HALT | CP_PFP_HALT | CP_CE_HALT)); + WREG32(SCRATCH_UMSK, 0); + for (i = 0; i < adev->gfx.num_gfx_rings; i++) + adev->gfx.gfx_ring[i].ready = false; + for (i = 0; i < adev->gfx.num_compute_rings; i++) + adev->gfx.compute_ring[i].ready = false; + } + udelay(50); +} + +static int gfx_v6_0_cp_gfx_load_microcode(struct amdgpu_device *adev) +{ + unsigned i; + const struct gfx_firmware_header_v1_0 *pfp_hdr; + const struct gfx_firmware_header_v1_0 *ce_hdr; + const struct gfx_firmware_header_v1_0 *me_hdr; + const __le32 *fw_data; + u32 fw_size; + + if (!adev->gfx.me_fw || !adev->gfx.pfp_fw || !adev->gfx.ce_fw) + return -EINVAL; + + gfx_v6_0_cp_gfx_enable(adev, false); + pfp_hdr = (const struct gfx_firmware_header_v1_0 *)adev->gfx.pfp_fw->data; + ce_hdr = (const struct gfx_firmware_header_v1_0 *)adev->gfx.ce_fw->data; + me_hdr = (const struct gfx_firmware_header_v1_0 *)adev->gfx.me_fw->data; + + amdgpu_ucode_print_gfx_hdr(&pfp_hdr->header); + amdgpu_ucode_print_gfx_hdr(&ce_hdr->header); + amdgpu_ucode_print_gfx_hdr(&me_hdr->header); + + /* PFP */ + fw_data = (const __le32 *) + (adev->gfx.pfp_fw->data + le32_to_cpu(pfp_hdr->header.ucode_array_offset_bytes)); + fw_size = le32_to_cpu(pfp_hdr->header.ucode_size_bytes) / 4; + WREG32(CP_PFP_UCODE_ADDR, 0); + for (i = 0; i < fw_size; i++) + WREG32(CP_PFP_UCODE_DATA, le32_to_cpup(fw_data++)); + WREG32(CP_PFP_UCODE_ADDR, 0); + + /* CE */ + fw_data = (const __le32 *) + (adev->gfx.ce_fw->data + le32_to_cpu(ce_hdr->header.ucode_array_offset_bytes)); + fw_size = le32_to_cpu(ce_hdr->header.ucode_size_bytes) / 4; + WREG32(CP_CE_UCODE_ADDR, 0); + for (i = 0; i < fw_size; i++) + WREG32(CP_CE_UCODE_DATA, le32_to_cpup(fw_data++)); + WREG32(CP_CE_UCODE_ADDR, 0); + + /* ME */ + fw_data = (const __be32 *) + (adev->gfx.me_fw->data + le32_to_cpu(me_hdr->header.ucode_array_offset_bytes)); + fw_size = le32_to_cpu(me_hdr->header.ucode_size_bytes) / 4; + WREG32(CP_ME_RAM_WADDR, 0); + for (i = 0; i < fw_size; i++) + WREG32(CP_ME_RAM_DATA, le32_to_cpup(fw_data++)); + WREG32(CP_ME_RAM_WADDR, 0); + + + WREG32(CP_PFP_UCODE_ADDR, 0); + WREG32(CP_CE_UCODE_ADDR, 0); + WREG32(CP_ME_RAM_WADDR, 0); + WREG32(CP_ME_RAM_RADDR, 0); + return 0; +} + +static int gfx_v6_0_cp_gfx_start(struct amdgpu_device *adev) +{ + const struct cs_section_def *sect = NULL; + const struct cs_extent_def *ext = NULL; + struct amdgpu_ring *ring = &adev->gfx.gfx_ring[0]; + int r, i; + + r = amdgpu_ring_alloc(ring, 7 + 4); + if (r) { + DRM_ERROR("amdgpu: cp failed to lock ring (%d).\n", r); + return r; + } + amdgpu_ring_write(ring, PACKET3(PACKET3_ME_INITIALIZE, 5)); + amdgpu_ring_write(ring, 0x1); + amdgpu_ring_write(ring, 0x0); + amdgpu_ring_write(ring, adev->gfx.config.max_hw_contexts - 1); + amdgpu_ring_write(ring, PACKET3_ME_INITIALIZE_DEVICE_ID(1)); + amdgpu_ring_write(ring, 0); + amdgpu_ring_write(ring, 0); + + amdgpu_ring_write(ring, PACKET3(PACKET3_SET_BASE, 2)); + amdgpu_ring_write(ring, PACKET3_BASE_INDEX(CE_PARTITION_BASE)); + amdgpu_ring_write(ring, 0xc000); + amdgpu_ring_write(ring, 0xe000); + amdgpu_ring_commit(ring); + + gfx_v6_0_cp_gfx_enable(adev, true); + + r = amdgpu_ring_alloc(ring, gfx_v6_0_get_csb_size(adev) + 10); + if (r) { + DRM_ERROR("amdgpu: cp failed to lock ring (%d).\n", r); + return r; + } + + amdgpu_ring_write(ring, PACKET3(PACKET3_PREAMBLE_CNTL, 0)); + amdgpu_ring_write(ring, PACKET3_PREAMBLE_BEGIN_CLEAR_STATE); + + for (sect = adev->gfx.rlc.cs_data; sect->section != NULL; ++sect) { + for (ext = sect->section; ext->extent != NULL; ++ext) { + if (sect->id == SECT_CONTEXT) { + amdgpu_ring_write(ring, + PACKET3(PACKET3_SET_CONTEXT_REG, ext->reg_count)); + amdgpu_ring_write(ring, ext->reg_index - PACKET3_SET_CONTEXT_REG_START); + for (i = 0; i < ext->reg_count; i++) + amdgpu_ring_write(ring, ext->extent[i]); + } + } + } + + amdgpu_ring_write(ring, PACKET3(PACKET3_PREAMBLE_CNTL, 0)); + amdgpu_ring_write(ring, PACKET3_PREAMBLE_END_CLEAR_STATE); + + amdgpu_ring_write(ring, PACKET3(PACKET3_CLEAR_STATE, 0)); + amdgpu_ring_write(ring, 0); + + amdgpu_ring_write(ring, PACKET3(PACKET3_SET_CONTEXT_REG, 2)); + amdgpu_ring_write(ring, 0x00000316); + amdgpu_ring_write(ring, 0x0000000e); + amdgpu_ring_write(ring, 0x00000010); + + amdgpu_ring_commit(ring); + + return 0; +} + +static int gfx_v6_0_cp_gfx_resume(struct amdgpu_device *adev) +{ + struct amdgpu_ring *ring; + u32 tmp; + u32 rb_bufsz; + int r; + u64 rptr_addr; + + WREG32(CP_SEM_WAIT_TIMER, 0x0); + WREG32(CP_SEM_INCOMPLETE_TIMER_CNTL, 0x0); + + /* Set the write pointer delay */ + WREG32(CP_RB_WPTR_DELAY, 0); + + WREG32(CP_DEBUG, 0); + WREG32(SCRATCH_ADDR, 0); + + /* ring 0 - compute and gfx */ + /* Set ring buffer size */ + ring = &adev->gfx.gfx_ring[0]; + rb_bufsz = order_base_2(ring->ring_size / 8); + tmp = (order_base_2(AMDGPU_GPU_PAGE_SIZE/8) << 8) | rb_bufsz; + +#ifdef __BIG_ENDIAN + tmp |= BUF_SWAP_32BIT; +#endif + WREG32(CP_RB0_CNTL, tmp); + + /* Initialize the ring buffer's read and write pointers */ + WREG32(CP_RB0_CNTL, tmp | RB_RPTR_WR_ENA); + ring->wptr = 0; + WREG32(CP_RB0_WPTR, ring->wptr); + + /* set the wb address whether it's enabled or not */ + rptr_addr = adev->wb.gpu_addr + (ring->rptr_offs * 4); + WREG32(CP_RB0_RPTR_ADDR, lower_32_bits(rptr_addr)); + WREG32(CP_RB0_RPTR_ADDR_HI, upper_32_bits(rptr_addr) & 0xFF); + + WREG32(SCRATCH_UMSK, 0); + + mdelay(1); + WREG32(CP_RB0_CNTL, tmp); + + WREG32(CP_RB0_BASE, ring->gpu_addr >> 8); + + /* start the rings */ + gfx_v6_0_cp_gfx_start(adev); + ring->ready = true; + r = amdgpu_ring_test_ring(ring); + if (r) { + ring->ready = false; + return r; + } + + return 0; +} + +static u32 gfx_v6_0_ring_get_rptr(struct amdgpu_ring *ring) +{ + return ring->adev->wb.wb[ring->rptr_offs]; +} + +static u32 gfx_v6_0_ring_get_wptr(struct amdgpu_ring *ring) +{ + struct amdgpu_device *adev = ring->adev; + + if (ring == &adev->gfx.gfx_ring[0]) + return RREG32(CP_RB0_WPTR); + else if (ring == &adev->gfx.compute_ring[0]) + return RREG32(CP_RB1_WPTR); + else if (ring == &adev->gfx.compute_ring[1]) + return RREG32(CP_RB2_WPTR); + else + BUG(); +} + +static void gfx_v6_0_ring_set_wptr_gfx(struct amdgpu_ring *ring) +{ + struct amdgpu_device *adev = ring->adev; + + WREG32(CP_RB0_WPTR, ring->wptr); + (void)RREG32(CP_RB0_WPTR); +} + +static void gfx_v6_0_ring_set_wptr_compute(struct amdgpu_ring *ring) +{ + struct amdgpu_device *adev = ring->adev; + + if (ring == &adev->gfx.compute_ring[0]) { + WREG32(CP_RB1_WPTR, ring->wptr); + (void)RREG32(CP_RB1_WPTR); + } else if (ring == &adev->gfx.compute_ring[1]) { + WREG32(CP_RB2_WPTR, ring->wptr); + (void)RREG32(CP_RB2_WPTR); + } else { + BUG(); + } + +} + +static int gfx_v6_0_cp_compute_resume(struct amdgpu_device *adev) +{ + struct amdgpu_ring *ring; + u32 tmp; + u32 rb_bufsz; + int r; + u64 rptr_addr; + + /* ring1 - compute only */ + /* Set ring buffer size */ + + ring = &adev->gfx.compute_ring[0]; + rb_bufsz = order_base_2(ring->ring_size / 8); + tmp = (order_base_2(AMDGPU_GPU_PAGE_SIZE/8) << 8) | rb_bufsz; +#ifdef __BIG_ENDIAN + tmp |= BUF_SWAP_32BIT; +#endif + WREG32(CP_RB1_CNTL, tmp); + + WREG32(CP_RB1_CNTL, tmp | RB_RPTR_WR_ENA); + ring->wptr = 0; + WREG32(CP_RB1_WPTR, ring->wptr); + + rptr_addr = adev->wb.gpu_addr + (ring->rptr_offs * 4); + WREG32(CP_RB1_RPTR_ADDR, lower_32_bits(rptr_addr)); + WREG32(CP_RB1_RPTR_ADDR_HI, upper_32_bits(rptr_addr) & 0xFF); + + mdelay(1); + WREG32(CP_RB1_CNTL, tmp); + WREG32(CP_RB1_BASE, ring->gpu_addr >> 8); + + ring = &adev->gfx.compute_ring[1]; + rb_bufsz = order_base_2(ring->ring_size / 8); + tmp = (order_base_2(AMDGPU_GPU_PAGE_SIZE/8) << 8) | rb_bufsz; +#ifdef __BIG_ENDIAN + tmp |= BUF_SWAP_32BIT; +#endif + WREG32(CP_RB2_CNTL, tmp); + + WREG32(CP_RB2_CNTL, tmp | RB_RPTR_WR_ENA); + ring->wptr = 0; + WREG32(CP_RB2_WPTR, ring->wptr); + rptr_addr = adev->wb.gpu_addr + (ring->rptr_offs * 4); + WREG32(CP_RB2_RPTR_ADDR, lower_32_bits(rptr_addr)); + WREG32(CP_RB2_RPTR_ADDR_HI, upper_32_bits(rptr_addr) & 0xFF); + + mdelay(1); + WREG32(CP_RB2_CNTL, tmp); + WREG32(CP_RB2_BASE, ring->gpu_addr >> 8); + + adev->gfx.compute_ring[0].ready = true; + adev->gfx.compute_ring[1].ready = true; + + r = amdgpu_ring_test_ring(&adev->gfx.compute_ring[0]); + if (r) { + adev->gfx.compute_ring[0].ready = false; + return r; + } + + r = amdgpu_ring_test_ring(&adev->gfx.compute_ring[1]); + if (r) { + adev->gfx.compute_ring[1].ready = false; + return r; + } + + return 0; +} + +static void gfx_v6_0_cp_enable(struct amdgpu_device *adev, bool enable) +{ + gfx_v6_0_cp_gfx_enable(adev, enable); +} + +static int gfx_v6_0_cp_load_microcode(struct amdgpu_device *adev) +{ + return gfx_v6_0_cp_gfx_load_microcode(adev); +} + +static void gfx_v6_0_enable_gui_idle_interrupt(struct amdgpu_device *adev, + bool enable) +{ + u32 tmp = RREG32(CP_INT_CNTL_RING0); + u32 mask; + int i; + + if (enable) + tmp |= (CNTX_BUSY_INT_ENABLE | CNTX_EMPTY_INT_ENABLE); + else + tmp &= ~(CNTX_BUSY_INT_ENABLE | CNTX_EMPTY_INT_ENABLE); + WREG32(CP_INT_CNTL_RING0, tmp); + + if (!enable) { + /* read a gfx register */ + tmp = RREG32(DB_DEPTH_INFO); + + mask = RLC_BUSY_STATUS | GFX_POWER_STATUS | GFX_CLOCK_STATUS | GFX_LS_STATUS; + for (i = 0; i < adev->usec_timeout; i++) { + if ((RREG32(RLC_STAT) & mask) == (GFX_CLOCK_STATUS | GFX_POWER_STATUS)) + break; + udelay(1); + } + } +} + +static int gfx_v6_0_cp_resume(struct amdgpu_device *adev) +{ + int r; + + gfx_v6_0_enable_gui_idle_interrupt(adev, false); + + r = gfx_v6_0_cp_load_microcode(adev); + if (r) + return r; + + r = gfx_v6_0_cp_gfx_resume(adev); + if (r) + return r; + r = gfx_v6_0_cp_compute_resume(adev); + if (r) + return r; + + gfx_v6_0_enable_gui_idle_interrupt(adev, true); + + return 0; +} + +static void gfx_v6_0_ring_emit_pipeline_sync(struct amdgpu_ring *ring) +{ + int usepfp = (ring->type == AMDGPU_RING_TYPE_GFX); + uint32_t seq = ring->fence_drv.sync_seq; + uint64_t addr = ring->fence_drv.gpu_addr; + + amdgpu_ring_write(ring, PACKET3(PACKET3_WAIT_REG_MEM, 5)); + amdgpu_ring_write(ring, (WAIT_REG_MEM_MEM_SPACE(1) | /* memory */ + WAIT_REG_MEM_FUNCTION(3) | /* equal */ + WAIT_REG_MEM_ENGINE(usepfp))); /* pfp or me */ + amdgpu_ring_write(ring, addr & 0xfffffffc); + amdgpu_ring_write(ring, upper_32_bits(addr) & 0xffffffff); + amdgpu_ring_write(ring, seq); + amdgpu_ring_write(ring, 0xffffffff); + amdgpu_ring_write(ring, 4); /* poll interval */ + + if (usepfp) { + /* synce CE with ME to prevent CE fetch CEIB before context switch done */ + amdgpu_ring_write(ring, PACKET3(PACKET3_SWITCH_BUFFER, 0)); + amdgpu_ring_write(ring, 0); + amdgpu_ring_write(ring, PACKET3(PACKET3_SWITCH_BUFFER, 0)); + amdgpu_ring_write(ring, 0); + } +} + +static void gfx_v6_0_ring_emit_vm_flush(struct amdgpu_ring *ring, + unsigned vm_id, uint64_t pd_addr) +{ + int usepfp = (ring->type == AMDGPU_RING_TYPE_GFX); + + /* write new base address */ + amdgpu_ring_write(ring, PACKET3(PACKET3_WRITE_DATA, 3)); + amdgpu_ring_write(ring, (WRITE_DATA_ENGINE_SEL(1) | + WRITE_DATA_DST_SEL(0))); + if (vm_id < 8) { + amdgpu_ring_write(ring, (VM_CONTEXT0_PAGE_TABLE_BASE_ADDR + vm_id )); + } else { + amdgpu_ring_write(ring, (VM_CONTEXT8_PAGE_TABLE_BASE_ADDR + (vm_id - 8))); + } + amdgpu_ring_write(ring, 0); + amdgpu_ring_write(ring, pd_addr >> 12); + + /* bits 0-15 are the VM contexts0-15 */ + amdgpu_ring_write(ring, PACKET3(PACKET3_WRITE_DATA, 3)); + amdgpu_ring_write(ring, (WRITE_DATA_ENGINE_SEL(1) | + WRITE_DATA_DST_SEL(0))); + amdgpu_ring_write(ring, VM_INVALIDATE_REQUEST); + amdgpu_ring_write(ring, 0); + amdgpu_ring_write(ring, 1 << vm_id); + + /* wait for the invalidate to complete */ + amdgpu_ring_write(ring, PACKET3(PACKET3_WAIT_REG_MEM, 5)); + amdgpu_ring_write(ring, (WAIT_REG_MEM_FUNCTION(0) | /* always */ + WAIT_REG_MEM_ENGINE(0))); /* me */ + amdgpu_ring_write(ring, VM_INVALIDATE_REQUEST); + amdgpu_ring_write(ring, 0); + amdgpu_ring_write(ring, 0); /* ref */ + amdgpu_ring_write(ring, 0); /* mask */ + amdgpu_ring_write(ring, 0x20); /* poll interval */ + + if (usepfp) { + /* sync PFP to ME, otherwise we might get invalid PFP reads */ + amdgpu_ring_write(ring, PACKET3(PACKET3_PFP_SYNC_ME, 0)); + amdgpu_ring_write(ring, 0x0); + + /* synce CE with ME to prevent CE fetch CEIB before context switch done */ + amdgpu_ring_write(ring, PACKET3(PACKET3_SWITCH_BUFFER, 0)); + amdgpu_ring_write(ring, 0); + amdgpu_ring_write(ring, PACKET3(PACKET3_SWITCH_BUFFER, 0)); + amdgpu_ring_write(ring, 0); + } +} + + +static void gfx_v6_0_rlc_fini(struct amdgpu_device *adev) +{ + int r; + + if (adev->gfx.rlc.save_restore_obj) { + r = amdgpu_bo_reserve(adev->gfx.rlc.save_restore_obj, false); + if (unlikely(r != 0)) + dev_warn(adev->dev, "(%d) reserve RLC sr bo failed\n", r); + amdgpu_bo_unpin(adev->gfx.rlc.save_restore_obj); + amdgpu_bo_unreserve(adev->gfx.rlc.save_restore_obj); + + amdgpu_bo_unref(&adev->gfx.rlc.save_restore_obj); + adev->gfx.rlc.save_restore_obj = NULL; + } + + if (adev->gfx.rlc.clear_state_obj) { + r = amdgpu_bo_reserve(adev->gfx.rlc.clear_state_obj, false); + if (unlikely(r != 0)) + dev_warn(adev->dev, "(%d) reserve RLC c bo failed\n", r); + amdgpu_bo_unpin(adev->gfx.rlc.clear_state_obj); + amdgpu_bo_unreserve(adev->gfx.rlc.clear_state_obj); + + amdgpu_bo_unref(&adev->gfx.rlc.clear_state_obj); + adev->gfx.rlc.clear_state_obj = NULL; + } + + if (adev->gfx.rlc.cp_table_obj) { + r = amdgpu_bo_reserve(adev->gfx.rlc.cp_table_obj, false); + if (unlikely(r != 0)) + dev_warn(adev->dev, "(%d) reserve RLC cp table bo failed\n", r); + amdgpu_bo_unpin(adev->gfx.rlc.cp_table_obj); + amdgpu_bo_unreserve(adev->gfx.rlc.cp_table_obj); + + amdgpu_bo_unref(&adev->gfx.rlc.cp_table_obj); + adev->gfx.rlc.cp_table_obj = NULL; + } +} + +static int gfx_v6_0_rlc_init(struct amdgpu_device *adev) +{ + const u32 *src_ptr; + volatile u32 *dst_ptr; + u32 dws, i; + u64 reg_list_mc_addr; + const struct cs_section_def *cs_data; + int r; + + adev->gfx.rlc.reg_list = verde_rlc_save_restore_register_list; + adev->gfx.rlc.reg_list_size = + (u32)ARRAY_SIZE(verde_rlc_save_restore_register_list); + + adev->gfx.rlc.cs_data = si_cs_data; + src_ptr = adev->gfx.rlc.reg_list; + dws = adev->gfx.rlc.reg_list_size; + cs_data = adev->gfx.rlc.cs_data; + + if (src_ptr) { + /* save restore block */ + if (adev->gfx.rlc.save_restore_obj == NULL) { + + r = amdgpu_bo_create(adev, dws * 4, PAGE_SIZE, true, + AMDGPU_GEM_DOMAIN_VRAM, + AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED, + NULL, NULL, + &adev->gfx.rlc.save_restore_obj); + + if (r) { + dev_warn(adev->dev, "(%d) create RLC sr bo failed\n", r); + return r; + } + } + + r = amdgpu_bo_reserve(adev->gfx.rlc.save_restore_obj, false); + if (unlikely(r != 0)) { + gfx_v6_0_rlc_fini(adev); + return r; + } + r = amdgpu_bo_pin(adev->gfx.rlc.save_restore_obj, AMDGPU_GEM_DOMAIN_VRAM, + &adev->gfx.rlc.save_restore_gpu_addr); + if (r) { + amdgpu_bo_unreserve(adev->gfx.rlc.save_restore_obj); + dev_warn(adev->dev, "(%d) pin RLC sr bo failed\n", r); + gfx_v6_0_rlc_fini(adev); + return r; + } + + r = amdgpu_bo_kmap(adev->gfx.rlc.save_restore_obj, (void **)&adev->gfx.rlc.sr_ptr); + if (r) { + dev_warn(adev->dev, "(%d) map RLC sr bo failed\n", r); + gfx_v6_0_rlc_fini(adev); + return r; + } + /* write the sr buffer */ + dst_ptr = adev->gfx.rlc.sr_ptr; + for (i = 0; i < adev->gfx.rlc.reg_list_size; i++) + dst_ptr[i] = cpu_to_le32(src_ptr[i]); + amdgpu_bo_kunmap(adev->gfx.rlc.save_restore_obj); + amdgpu_bo_unreserve(adev->gfx.rlc.save_restore_obj); + } + + if (cs_data) { + /* clear state block */ + adev->gfx.rlc.clear_state_size = gfx_v6_0_get_csb_size(adev); + dws = adev->gfx.rlc.clear_state_size + (256 / 4); + + if (adev->gfx.rlc.clear_state_obj == NULL) { + r = amdgpu_bo_create(adev, dws * 4, PAGE_SIZE, true, + AMDGPU_GEM_DOMAIN_VRAM, + AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED, + NULL, NULL, + &adev->gfx.rlc.clear_state_obj); + + if (r) { + dev_warn(adev->dev, "(%d) create RLC c bo failed\n", r); + gfx_v6_0_rlc_fini(adev); + return r; + } + } + r = amdgpu_bo_reserve(adev->gfx.rlc.clear_state_obj, false); + if (unlikely(r != 0)) { + gfx_v6_0_rlc_fini(adev); + return r; + } + r = amdgpu_bo_pin(adev->gfx.rlc.clear_state_obj, AMDGPU_GEM_DOMAIN_VRAM, + &adev->gfx.rlc.clear_state_gpu_addr); + if (r) { + amdgpu_bo_unreserve(adev->gfx.rlc.clear_state_obj); + dev_warn(adev->dev, "(%d) pin RLC c bo failed\n", r); + gfx_v6_0_rlc_fini(adev); + return r; + } + + r = amdgpu_bo_kmap(adev->gfx.rlc.clear_state_obj, (void **)&adev->gfx.rlc.cs_ptr); + if (r) { + dev_warn(adev->dev, "(%d) map RLC c bo failed\n", r); + gfx_v6_0_rlc_fini(adev); + return r; + } + /* set up the cs buffer */ + dst_ptr = adev->gfx.rlc.cs_ptr; + reg_list_mc_addr = adev->gfx.rlc.clear_state_gpu_addr + 256; + dst_ptr[0] = cpu_to_le32(upper_32_bits(reg_list_mc_addr)); + dst_ptr[1] = cpu_to_le32(lower_32_bits(reg_list_mc_addr)); + dst_ptr[2] = cpu_to_le32(adev->gfx.rlc.clear_state_size); + gfx_v6_0_get_csb_buffer(adev, &dst_ptr[(256/4)]); + amdgpu_bo_kunmap(adev->gfx.rlc.clear_state_obj); + amdgpu_bo_unreserve(adev->gfx.rlc.clear_state_obj); + } + + return 0; +} + +static void gfx_v6_0_enable_lbpw(struct amdgpu_device *adev, bool enable) +{ + u32 tmp; + + tmp = RREG32(RLC_LB_CNTL); + if (enable) + tmp |= LOAD_BALANCE_ENABLE; + else + tmp &= ~LOAD_BALANCE_ENABLE; + WREG32(RLC_LB_CNTL, tmp); + + if (!enable) { + gfx_v6_0_select_se_sh(adev, 0xffffffff, 0xffffffff, 0xffffffff); + WREG32(SPI_LB_CU_MASK, 0x00ff); + } + +} + +static void gfx_v6_0_wait_for_rlc_serdes(struct amdgpu_device *adev) +{ + int i; + + for (i = 0; i < adev->usec_timeout; i++) { + if (RREG32(RLC_SERDES_MASTER_BUSY_0) == 0) + break; + udelay(1); + } + + for (i = 0; i < adev->usec_timeout; i++) { + if (RREG32(RLC_SERDES_MASTER_BUSY_1) == 0) + break; + udelay(1); + } +} + +static void gfx_v6_0_update_rlc(struct amdgpu_device *adev, u32 rlc) +{ + u32 tmp; + + tmp = RREG32(RLC_CNTL); + if (tmp != rlc) + WREG32(RLC_CNTL, rlc); +} + +static u32 gfx_v6_0_halt_rlc(struct amdgpu_device *adev) +{ + u32 data, orig; + + orig = data = RREG32(RLC_CNTL); + + if (data & RLC_ENABLE) { + data &= ~RLC_ENABLE; + WREG32(RLC_CNTL, data); + + gfx_v6_0_wait_for_rlc_serdes(adev); + } + + return orig; +} + +static void gfx_v6_0_rlc_stop(struct amdgpu_device *adev) +{ + WREG32(RLC_CNTL, 0); + + gfx_v6_0_enable_gui_idle_interrupt(adev, false); + gfx_v6_0_wait_for_rlc_serdes(adev); +} + +static void gfx_v6_0_rlc_start(struct amdgpu_device *adev) +{ + WREG32(RLC_CNTL, RLC_ENABLE); + + gfx_v6_0_enable_gui_idle_interrupt(adev, true); + + udelay(50); +} + +static void gfx_v6_0_rlc_reset(struct amdgpu_device *adev) +{ + u32 tmp = RREG32(GRBM_SOFT_RESET); + + tmp |= SOFT_RESET_RLC; + WREG32(GRBM_SOFT_RESET, tmp); + udelay(50); + tmp &= ~SOFT_RESET_RLC; + WREG32(GRBM_SOFT_RESET, tmp); + udelay(50); +} + +static bool gfx_v6_0_lbpw_supported(struct amdgpu_device *adev) +{ + u32 tmp; + + /* Enable LBPW only for DDR3 */ + tmp = RREG32(MC_SEQ_MISC0); + if ((tmp & 0xF0000000) == 0xB0000000) + return true; + return false; +} +static void gfx_v6_0_init_cg(struct amdgpu_device *adev) +{ +} + +static int gfx_v6_0_rlc_resume(struct amdgpu_device *adev) +{ + u32 i; + const struct rlc_firmware_header_v1_0 *hdr; + const __le32 *fw_data; + u32 fw_size; + + + if (!adev->gfx.rlc_fw) + return -EINVAL; + + gfx_v6_0_rlc_stop(adev); + gfx_v6_0_rlc_reset(adev); + gfx_v6_0_init_pg(adev); + gfx_v6_0_init_cg(adev); + + WREG32(RLC_RL_BASE, 0); + WREG32(RLC_RL_SIZE, 0); + WREG32(RLC_LB_CNTL, 0); + WREG32(RLC_LB_CNTR_MAX, 0xffffffff); + WREG32(RLC_LB_CNTR_INIT, 0); + WREG32(RLC_LB_INIT_CU_MASK, 0xffffffff); + + WREG32(RLC_MC_CNTL, 0); + WREG32(RLC_UCODE_CNTL, 0); + + hdr = (const struct rlc_firmware_header_v1_0 *)adev->gfx.rlc_fw->data; + fw_size = le32_to_cpu(hdr->header.ucode_size_bytes) / 4; + fw_data = (const __le32 *) + (adev->gfx.rlc_fw->data + le32_to_cpu(hdr->header.ucode_array_offset_bytes)); + + amdgpu_ucode_print_rlc_hdr(&hdr->header); + + for (i = 0; i < fw_size; i++) { + WREG32(RLC_UCODE_ADDR, i); + WREG32(RLC_UCODE_DATA, le32_to_cpup(fw_data++)); + } + WREG32(RLC_UCODE_ADDR, 0); + + gfx_v6_0_enable_lbpw(adev, gfx_v6_0_lbpw_supported(adev)); + gfx_v6_0_rlc_start(adev); + + return 0; +} + +static void gfx_v6_0_enable_cgcg(struct amdgpu_device *adev, bool enable) +{ + u32 data, orig, tmp; + + orig = data = RREG32(RLC_CGCG_CGLS_CTRL); + + if (enable && (adev->cg_flags & AMD_CG_SUPPORT_GFX_CGCG)) { + gfx_v6_0_enable_gui_idle_interrupt(adev, true); + + WREG32(RLC_GCPM_GENERAL_3, 0x00000080); + + tmp = gfx_v6_0_halt_rlc(adev); + + WREG32(RLC_SERDES_WR_MASTER_MASK_0, 0xffffffff); + WREG32(RLC_SERDES_WR_MASTER_MASK_1, 0xffffffff); + WREG32(RLC_SERDES_WR_CTRL, 0x00b000ff); + + gfx_v6_0_wait_for_rlc_serdes(adev); + gfx_v6_0_update_rlc(adev, tmp); + + WREG32(RLC_SERDES_WR_CTRL, 0x007000ff); + + data |= CGCG_EN | CGLS_EN; + } else { + gfx_v6_0_enable_gui_idle_interrupt(adev, false); + + RREG32(CB_CGTT_SCLK_CTRL); + RREG32(CB_CGTT_SCLK_CTRL); + RREG32(CB_CGTT_SCLK_CTRL); + RREG32(CB_CGTT_SCLK_CTRL); + + data &= ~(CGCG_EN | CGLS_EN); + } + + if (orig != data) + WREG32(RLC_CGCG_CGLS_CTRL, data); + +} + +static void gfx_v6_0_enable_mgcg(struct amdgpu_device *adev, bool enable) +{ + + u32 data, orig, tmp = 0; + + if (enable && (adev->cg_flags & AMD_CG_SUPPORT_GFX_MGCG)) { + orig = data = RREG32(CGTS_SM_CTRL_REG); + data = 0x96940200; + if (orig != data) + WREG32(CGTS_SM_CTRL_REG, data); + + if (adev->cg_flags & AMD_CG_SUPPORT_GFX_CP_LS) { + orig = data = RREG32(CP_MEM_SLP_CNTL); + data |= CP_MEM_LS_EN; + if (orig != data) + WREG32(CP_MEM_SLP_CNTL, data); + } + + orig = data = RREG32(RLC_CGTT_MGCG_OVERRIDE); + data &= 0xffffffc0; + if (orig != data) + WREG32(RLC_CGTT_MGCG_OVERRIDE, data); + + tmp = gfx_v6_0_halt_rlc(adev); + + WREG32(RLC_SERDES_WR_MASTER_MASK_0, 0xffffffff); + WREG32(RLC_SERDES_WR_MASTER_MASK_1, 0xffffffff); + WREG32(RLC_SERDES_WR_CTRL, 0x00d000ff); + + gfx_v6_0_update_rlc(adev, tmp); + } else { + orig = data = RREG32(RLC_CGTT_MGCG_OVERRIDE); + data |= 0x00000003; + if (orig != data) + WREG32(RLC_CGTT_MGCG_OVERRIDE, data); + + data = RREG32(CP_MEM_SLP_CNTL); + if (data & CP_MEM_LS_EN) { + data &= ~CP_MEM_LS_EN; + WREG32(CP_MEM_SLP_CNTL, data); + } + orig = data = RREG32(CGTS_SM_CTRL_REG); + data |= LS_OVERRIDE | OVERRIDE; + if (orig != data) + WREG32(CGTS_SM_CTRL_REG, data); + + tmp = gfx_v6_0_halt_rlc(adev); + + WREG32(RLC_SERDES_WR_MASTER_MASK_0, 0xffffffff); + WREG32(RLC_SERDES_WR_MASTER_MASK_1, 0xffffffff); + WREG32(RLC_SERDES_WR_CTRL, 0x00e000ff); + + gfx_v6_0_update_rlc(adev, tmp); + } +} +/* +static void gfx_v6_0_update_cg(struct amdgpu_device *adev, + bool enable) +{ + gfx_v6_0_enable_gui_idle_interrupt(adev, false); + if (enable) { + gfx_v6_0_enable_mgcg(adev, true); + gfx_v6_0_enable_cgcg(adev, true); + } else { + gfx_v6_0_enable_cgcg(adev, false); + gfx_v6_0_enable_mgcg(adev, false); + } + gfx_v6_0_enable_gui_idle_interrupt(adev, true); +} +*/ +static void gfx_v6_0_enable_sclk_slowdown_on_pu(struct amdgpu_device *adev, + bool enable) +{ +} + +static void gfx_v6_0_enable_sclk_slowdown_on_pd(struct amdgpu_device *adev, + bool enable) +{ +} + +static void gfx_v6_0_enable_cp_pg(struct amdgpu_device *adev, bool enable) +{ + u32 data, orig; + + orig = data = RREG32(RLC_PG_CNTL); + if (enable && (adev->pg_flags & AMD_PG_SUPPORT_CP)) + data &= ~0x8000; + else + data |= 0x8000; + if (orig != data) + WREG32(RLC_PG_CNTL, data); +} + +static void gfx_v6_0_enable_gds_pg(struct amdgpu_device *adev, bool enable) +{ +} +/* +static void gfx_v6_0_init_cp_pg_table(struct amdgpu_device *adev) +{ + const __le32 *fw_data; + volatile u32 *dst_ptr; + int me, i, max_me = 4; + u32 bo_offset = 0; + u32 table_offset, table_size; + + if (adev->asic_type == CHIP_KAVERI) + max_me = 5; + + if (adev->gfx.rlc.cp_table_ptr == NULL) + return; + + dst_ptr = adev->gfx.rlc.cp_table_ptr; + for (me = 0; me < max_me; me++) { + if (me == 0) { + const struct gfx_firmware_header_v1_0 *hdr = + (const struct gfx_firmware_header_v1_0 *)adev->gfx.ce_fw->data; + fw_data = (const __le32 *) + (adev->gfx.ce_fw->data + + le32_to_cpu(hdr->header.ucode_array_offset_bytes)); + table_offset = le32_to_cpu(hdr->jt_offset); + table_size = le32_to_cpu(hdr->jt_size); + } else if (me == 1) { + const struct gfx_firmware_header_v1_0 *hdr = + (const struct gfx_firmware_header_v1_0 *)adev->gfx.pfp_fw->data; + fw_data = (const __le32 *) + (adev->gfx.pfp_fw->data + + le32_to_cpu(hdr->header.ucode_array_offset_bytes)); + table_offset = le32_to_cpu(hdr->jt_offset); + table_size = le32_to_cpu(hdr->jt_size); + } else if (me == 2) { + const struct gfx_firmware_header_v1_0 *hdr = + (const struct gfx_firmware_header_v1_0 *)adev->gfx.me_fw->data; + fw_data = (const __le32 *) + (adev->gfx.me_fw->data + + le32_to_cpu(hdr->header.ucode_array_offset_bytes)); + table_offset = le32_to_cpu(hdr->jt_offset); + table_size = le32_to_cpu(hdr->jt_size); + } else if (me == 3) { + const struct gfx_firmware_header_v1_0 *hdr = + (const struct gfx_firmware_header_v1_0 *)adev->gfx.mec_fw->data; + fw_data = (const __le32 *) + (adev->gfx.mec_fw->data + + le32_to_cpu(hdr->header.ucode_array_offset_bytes)); + table_offset = le32_to_cpu(hdr->jt_offset); + table_size = le32_to_cpu(hdr->jt_size); + } else { + const struct gfx_firmware_header_v1_0 *hdr = + (const struct gfx_firmware_header_v1_0 *)adev->gfx.mec2_fw->data; + fw_data = (const __le32 *) + (adev->gfx.mec2_fw->data + + le32_to_cpu(hdr->header.ucode_array_offset_bytes)); + table_offset = le32_to_cpu(hdr->jt_offset); + table_size = le32_to_cpu(hdr->jt_size); + } + + for (i = 0; i < table_size; i ++) { + dst_ptr[bo_offset + i] = + cpu_to_le32(le32_to_cpu(fw_data[table_offset + i])); + } + + bo_offset += table_size; + } +} +*/ +static void gfx_v6_0_enable_gfx_cgpg(struct amdgpu_device *adev, + bool enable) +{ + + u32 tmp; + + if (enable && (adev->pg_flags & AMD_PG_SUPPORT_GFX_PG)) { + tmp = RLC_PUD(0x10) | RLC_PDD(0x10) | RLC_TTPD(0x10) | RLC_MSD(0x10); + WREG32(RLC_TTOP_D, tmp); + + tmp = RREG32(RLC_PG_CNTL); + tmp |= GFX_PG_ENABLE; + WREG32(RLC_PG_CNTL, tmp); + + tmp = RREG32(RLC_AUTO_PG_CTRL); + tmp |= AUTO_PG_EN; + WREG32(RLC_AUTO_PG_CTRL, tmp); + } else { + tmp = RREG32(RLC_AUTO_PG_CTRL); + tmp &= ~AUTO_PG_EN; + WREG32(RLC_AUTO_PG_CTRL, tmp); + + tmp = RREG32(DB_RENDER_CONTROL); + } +} + +static u32 gfx_v6_0_get_cu_active_bitmap(struct amdgpu_device *adev, + u32 se, u32 sh) +{ + + u32 mask = 0, tmp, tmp1; + int i; + + mutex_lock(&adev->grbm_idx_mutex); + gfx_v6_0_select_se_sh(adev, se, sh, 0xffffffff); + tmp = RREG32(CC_GC_SHADER_ARRAY_CONFIG); + tmp1 = RREG32(GC_USER_SHADER_ARRAY_CONFIG); + gfx_v6_0_select_se_sh(adev, 0xffffffff, 0xffffffff, 0xffffffff); + mutex_unlock(&adev->grbm_idx_mutex); + + tmp &= 0xffff0000; + + tmp |= tmp1; + tmp >>= 16; + + for (i = 0; i < adev->gfx.config.max_cu_per_sh; i ++) { + mask <<= 1; + mask |= 1; + } + + return (~tmp) & mask; +} + +static void gfx_v6_0_init_ao_cu_mask(struct amdgpu_device *adev) +{ + u32 i, j, k, active_cu_number = 0; + + u32 mask, counter, cu_bitmap; + u32 tmp = 0; + + for (i = 0; i < adev->gfx.config.max_shader_engines; i++) { + for (j = 0; j < adev->gfx.config.max_sh_per_se; j++) { + mask = 1; + cu_bitmap = 0; + counter = 0; + for (k = 0; k < adev->gfx.config.max_cu_per_sh; k++) { + if (gfx_v6_0_get_cu_active_bitmap(adev, i, j) & mask) { + if (counter < 2) + cu_bitmap |= mask; + counter++; + } + mask <<= 1; + } + + active_cu_number += counter; + tmp |= (cu_bitmap << (i * 16 + j * 8)); + } + } + + WREG32(RLC_PG_AO_CU_MASK, tmp); + + tmp = RREG32(RLC_MAX_PG_CU); + tmp &= ~MAX_PU_CU_MASK; + tmp |= MAX_PU_CU(active_cu_number); + WREG32(RLC_MAX_PG_CU, tmp); +} + +static void gfx_v6_0_enable_gfx_static_mgpg(struct amdgpu_device *adev, + bool enable) +{ + u32 data, orig; + + orig = data = RREG32(RLC_PG_CNTL); + if (enable && (adev->pg_flags & AMD_PG_SUPPORT_GFX_SMG)) + data |= STATIC_PER_CU_PG_ENABLE; + else + data &= ~STATIC_PER_CU_PG_ENABLE; + if (orig != data) + WREG32(RLC_PG_CNTL, data); +} + +static void gfx_v6_0_enable_gfx_dynamic_mgpg(struct amdgpu_device *adev, + bool enable) +{ + u32 data, orig; + + orig = data = RREG32(RLC_PG_CNTL); + if (enable && (adev->pg_flags & AMD_PG_SUPPORT_GFX_DMG)) + data |= DYN_PER_CU_PG_ENABLE; + else + data &= ~DYN_PER_CU_PG_ENABLE; + if (orig != data) + WREG32(RLC_PG_CNTL, data); +} + +static void gfx_v6_0_init_gfx_cgpg(struct amdgpu_device *adev) +{ + u32 tmp; + + WREG32(RLC_SAVE_AND_RESTORE_BASE, adev->gfx.rlc.save_restore_gpu_addr >> 8); + + tmp = RREG32(RLC_PG_CNTL); + tmp |= GFX_PG_SRC; + WREG32(RLC_PG_CNTL, tmp); + + WREG32(RLC_CLEAR_STATE_RESTORE_BASE, adev->gfx.rlc.clear_state_gpu_addr >> 8); + + tmp = RREG32(RLC_AUTO_PG_CTRL); + + tmp &= ~GRBM_REG_SGIT_MASK; + tmp |= GRBM_REG_SGIT(0x700); + tmp &= ~PG_AFTER_GRBM_REG_ST_MASK; + WREG32(RLC_AUTO_PG_CTRL, tmp); +} + +static void gfx_v6_0_update_gfx_pg(struct amdgpu_device *adev, bool enable) +{ + gfx_v6_0_enable_gfx_cgpg(adev, enable); + gfx_v6_0_enable_gfx_static_mgpg(adev, enable); + gfx_v6_0_enable_gfx_dynamic_mgpg(adev, enable); +} + +static u32 gfx_v6_0_get_csb_size(struct amdgpu_device *adev) +{ + u32 count = 0; + const struct cs_section_def *sect = NULL; + const struct cs_extent_def *ext = NULL; + + if (adev->gfx.rlc.cs_data == NULL) + return 0; + + /* begin clear state */ + count += 2; + /* context control state */ + count += 3; + + for (sect = adev->gfx.rlc.cs_data; sect->section != NULL; ++sect) { + for (ext = sect->section; ext->extent != NULL; ++ext) { + if (sect->id == SECT_CONTEXT) + count += 2 + ext->reg_count; + else + return 0; + } + } + /* pa_sc_raster_config */ + count += 3; + /* end clear state */ + count += 2; + /* clear state */ + count += 2; + + return count; +} + +static void gfx_v6_0_get_csb_buffer(struct amdgpu_device *adev, + volatile u32 *buffer) +{ + u32 count = 0, i; + const struct cs_section_def *sect = NULL; + const struct cs_extent_def *ext = NULL; + + if (adev->gfx.rlc.cs_data == NULL) + return; + if (buffer == NULL) + return; + + buffer[count++] = cpu_to_le32(PACKET3(PACKET3_PREAMBLE_CNTL, 0)); + buffer[count++] = cpu_to_le32(PACKET3_PREAMBLE_BEGIN_CLEAR_STATE); + + buffer[count++] = cpu_to_le32(PACKET3(PACKET3_CONTEXT_CONTROL, 1)); + buffer[count++] = cpu_to_le32(0x80000000); + buffer[count++] = cpu_to_le32(0x80000000); + + for (sect = adev->gfx.rlc.cs_data; sect->section != NULL; ++sect) { + for (ext = sect->section; ext->extent != NULL; ++ext) { + if (sect->id == SECT_CONTEXT) { + buffer[count++] = + cpu_to_le32(PACKET3(PACKET3_SET_CONTEXT_REG, ext->reg_count)); + buffer[count++] = cpu_to_le32(ext->reg_index - 0xa000); + for (i = 0; i < ext->reg_count; i++) + buffer[count++] = cpu_to_le32(ext->extent[i]); + } else { + return; + } + } + } + + buffer[count++] = cpu_to_le32(PACKET3(PACKET3_SET_CONTEXT_REG, 1)); + buffer[count++] = cpu_to_le32(PA_SC_RASTER_CONFIG - PACKET3_SET_CONTEXT_REG_START); + + switch (adev->asic_type) { + case CHIP_TAHITI: + case CHIP_PITCAIRN: + buffer[count++] = cpu_to_le32(0x2a00126a); + break; + case CHIP_VERDE: + buffer[count++] = cpu_to_le32(0x0000124a); + break; + case CHIP_OLAND: + buffer[count++] = cpu_to_le32(0x00000082); + break; + case CHIP_HAINAN: + buffer[count++] = cpu_to_le32(0x00000000); + break; + default: + buffer[count++] = cpu_to_le32(0x00000000); + break; + } + + buffer[count++] = cpu_to_le32(PACKET3(PACKET3_PREAMBLE_CNTL, 0)); + buffer[count++] = cpu_to_le32(PACKET3_PREAMBLE_END_CLEAR_STATE); + + buffer[count++] = cpu_to_le32(PACKET3(PACKET3_CLEAR_STATE, 0)); + buffer[count++] = cpu_to_le32(0); +} + +static void gfx_v6_0_init_pg(struct amdgpu_device *adev) +{ + if (adev->pg_flags & (AMD_PG_SUPPORT_GFX_PG | + AMD_PG_SUPPORT_GFX_SMG | + AMD_PG_SUPPORT_GFX_DMG | + AMD_PG_SUPPORT_CP | + AMD_PG_SUPPORT_GDS | + AMD_PG_SUPPORT_RLC_SMU_HS)) { + gfx_v6_0_enable_sclk_slowdown_on_pu(adev, true); + gfx_v6_0_enable_sclk_slowdown_on_pd(adev, true); + if (adev->pg_flags & AMD_PG_SUPPORT_GFX_PG) { + gfx_v6_0_init_gfx_cgpg(adev); + gfx_v6_0_enable_cp_pg(adev, true); + gfx_v6_0_enable_gds_pg(adev, true); + } else { + WREG32(RLC_SAVE_AND_RESTORE_BASE, adev->gfx.rlc.save_restore_gpu_addr >> 8); + WREG32(RLC_CLEAR_STATE_RESTORE_BASE, adev->gfx.rlc.clear_state_gpu_addr >> 8); + + } + gfx_v6_0_init_ao_cu_mask(adev); + gfx_v6_0_update_gfx_pg(adev, true); + } else { + + WREG32(RLC_SAVE_AND_RESTORE_BASE, adev->gfx.rlc.save_restore_gpu_addr >> 8); + WREG32(RLC_CLEAR_STATE_RESTORE_BASE, adev->gfx.rlc.clear_state_gpu_addr >> 8); + } +} + +static void gfx_v6_0_fini_pg(struct amdgpu_device *adev) +{ + if (adev->pg_flags & (AMD_PG_SUPPORT_GFX_PG | + AMD_PG_SUPPORT_GFX_SMG | + AMD_PG_SUPPORT_GFX_DMG | + AMD_PG_SUPPORT_CP | + AMD_PG_SUPPORT_GDS | + AMD_PG_SUPPORT_RLC_SMU_HS)) { + gfx_v6_0_update_gfx_pg(adev, false); + if (adev->pg_flags & AMD_PG_SUPPORT_GFX_PG) { + gfx_v6_0_enable_cp_pg(adev, false); + gfx_v6_0_enable_gds_pg(adev, false); + } + } +} + +static uint64_t gfx_v6_0_get_gpu_clock_counter(struct amdgpu_device *adev) +{ + uint64_t clock; + + mutex_lock(&adev->gfx.gpu_clock_mutex); + WREG32(RLC_CAPTURE_GPU_CLOCK_COUNT, 1); + clock = (uint64_t)RREG32(RLC_GPU_CLOCK_COUNT_LSB) | + ((uint64_t)RREG32(RLC_GPU_CLOCK_COUNT_MSB) << 32ULL); + mutex_unlock(&adev->gfx.gpu_clock_mutex); + return clock; +} + +static void gfx_v6_ring_emit_cntxcntl(struct amdgpu_ring *ring, uint32_t flags) +{ + amdgpu_ring_write(ring, PACKET3(PACKET3_CONTEXT_CONTROL, 1)); + amdgpu_ring_write(ring, 0x80000000); + amdgpu_ring_write(ring, 0); +} + +static unsigned gfx_v6_0_ring_get_emit_ib_size(struct amdgpu_ring *ring) +{ + return + 6; /* gfx_v6_0_ring_emit_ib */ +} + +static unsigned gfx_v6_0_ring_get_dma_frame_size_gfx(struct amdgpu_ring *ring) +{ + return + 5 + /* gfx_v6_0_ring_emit_hdp_flush */ + 5 + /* gfx_v6_0_ring_emit_hdp_invalidate */ + 14 + 14 + 14 + /* gfx_v6_0_ring_emit_fence x3 for user fence, vm fence */ + 7 + 4 + /* gfx_v6_0_ring_emit_pipeline_sync */ + 17 + 6 + /* gfx_v6_0_ring_emit_vm_flush */ + 3; /* gfx_v6_ring_emit_cntxcntl */ +} + +static unsigned gfx_v6_0_ring_get_dma_frame_size_compute(struct amdgpu_ring *ring) +{ + return + 5 + /* gfx_v6_0_ring_emit_hdp_flush */ + 5 + /* gfx_v6_0_ring_emit_hdp_invalidate */ + 7 + /* gfx_v6_0_ring_emit_pipeline_sync */ + 17 + /* gfx_v6_0_ring_emit_vm_flush */ + 14 + 14 + 14; /* gfx_v6_0_ring_emit_fence x3 for user fence, vm fence */ +} + +static const struct amdgpu_gfx_funcs gfx_v6_0_gfx_funcs = { + .get_gpu_clock_counter = &gfx_v6_0_get_gpu_clock_counter, + .select_se_sh = &gfx_v6_0_select_se_sh, +}; + +static int gfx_v6_0_early_init(void *handle) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + adev->gfx.num_gfx_rings = GFX6_NUM_GFX_RINGS; + adev->gfx.num_compute_rings = GFX6_NUM_COMPUTE_RINGS; + adev->gfx.funcs = &gfx_v6_0_gfx_funcs; + gfx_v6_0_set_ring_funcs(adev); + gfx_v6_0_set_irq_funcs(adev); + + return 0; +} + +static int gfx_v6_0_sw_init(void *handle) +{ + struct amdgpu_ring *ring; + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + int i, r; + + r = amdgpu_irq_add_id(adev, 181, &adev->gfx.eop_irq); + if (r) + return r; + + r = amdgpu_irq_add_id(adev, 184, &adev->gfx.priv_reg_irq); + if (r) + return r; + + r = amdgpu_irq_add_id(adev, 185, &adev->gfx.priv_inst_irq); + if (r) + return r; + + gfx_v6_0_scratch_init(adev); + + r = gfx_v6_0_init_microcode(adev); + if (r) { + DRM_ERROR("Failed to load gfx firmware!\n"); + return r; + } + + r = gfx_v6_0_rlc_init(adev); + if (r) { + DRM_ERROR("Failed to init rlc BOs!\n"); + return r; + } + + for (i = 0; i < adev->gfx.num_gfx_rings; i++) { + ring = &adev->gfx.gfx_ring[i]; + ring->ring_obj = NULL; + sprintf(ring->name, "gfx"); + r = amdgpu_ring_init(adev, ring, 1024, + 0x80000000, 0xf, + &adev->gfx.eop_irq, AMDGPU_CP_IRQ_GFX_EOP, + AMDGPU_RING_TYPE_GFX); + if (r) + return r; + } + + for (i = 0; i < adev->gfx.num_compute_rings; i++) { + unsigned irq_type; + + if ((i >= 32) || (i >= AMDGPU_MAX_COMPUTE_RINGS)) { + DRM_ERROR("Too many (%d) compute rings!\n", i); + break; + } + ring = &adev->gfx.compute_ring[i]; + ring->ring_obj = NULL; + ring->use_doorbell = false; + ring->doorbell_index = 0; + ring->me = 1; + ring->pipe = i; + ring->queue = i; + sprintf(ring->name, "comp %d.%d.%d", ring->me, ring->pipe, ring->queue); + irq_type = AMDGPU_CP_IRQ_COMPUTE_MEC1_PIPE0_EOP + ring->pipe; + r = amdgpu_ring_init(adev, ring, 1024, + 0x80000000, 0xf, + &adev->gfx.eop_irq, irq_type, + AMDGPU_RING_TYPE_COMPUTE); + if (r) + return r; + } + + return r; +} + +static int gfx_v6_0_sw_fini(void *handle) +{ + int i; + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + amdgpu_bo_unref(&adev->gds.oa_gfx_bo); + amdgpu_bo_unref(&adev->gds.gws_gfx_bo); + amdgpu_bo_unref(&adev->gds.gds_gfx_bo); + + for (i = 0; i < adev->gfx.num_gfx_rings; i++) + amdgpu_ring_fini(&adev->gfx.gfx_ring[i]); + for (i = 0; i < adev->gfx.num_compute_rings; i++) + amdgpu_ring_fini(&adev->gfx.compute_ring[i]); + + gfx_v6_0_rlc_fini(adev); + + return 0; +} + +static int gfx_v6_0_hw_init(void *handle) +{ + int r; + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + gfx_v6_0_gpu_init(adev); + + r = gfx_v6_0_rlc_resume(adev); + if (r) + return r; + + r = gfx_v6_0_cp_resume(adev); + if (r) + return r; + + adev->gfx.ce_ram_size = 0x8000; + + return r; +} + +static int gfx_v6_0_hw_fini(void *handle) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + gfx_v6_0_cp_enable(adev, false); + gfx_v6_0_rlc_stop(adev); + gfx_v6_0_fini_pg(adev); + + return 0; +} + +static int gfx_v6_0_suspend(void *handle) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + return gfx_v6_0_hw_fini(adev); +} + +static int gfx_v6_0_resume(void *handle) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + return gfx_v6_0_hw_init(adev); +} + +static bool gfx_v6_0_is_idle(void *handle) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + if (RREG32(GRBM_STATUS) & GRBM_STATUS__GUI_ACTIVE_MASK) + return false; + else + return true; +} + +static int gfx_v6_0_wait_for_idle(void *handle) +{ + unsigned i; + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + for (i = 0; i < adev->usec_timeout; i++) { + if (gfx_v6_0_is_idle(handle)) + return 0; + udelay(1); + } + return -ETIMEDOUT; +} + +static int gfx_v6_0_soft_reset(void *handle) +{ + return 0; +} + +static void gfx_v6_0_set_gfx_eop_interrupt_state(struct amdgpu_device *adev, + enum amdgpu_interrupt_state state) +{ + u32 cp_int_cntl; + + switch (state) { + case AMDGPU_IRQ_STATE_DISABLE: + cp_int_cntl = RREG32(CP_INT_CNTL_RING0); + cp_int_cntl &= ~CP_INT_CNTL_RING__TIME_STAMP_INT_ENABLE_MASK; + WREG32(CP_INT_CNTL_RING0, cp_int_cntl); + break; + case AMDGPU_IRQ_STATE_ENABLE: + cp_int_cntl = RREG32(CP_INT_CNTL_RING0); + cp_int_cntl |= CP_INT_CNTL_RING__TIME_STAMP_INT_ENABLE_MASK; + WREG32(CP_INT_CNTL_RING0, cp_int_cntl); + break; + default: + break; + } +} + +static void gfx_v6_0_set_compute_eop_interrupt_state(struct amdgpu_device *adev, + int ring, + enum amdgpu_interrupt_state state) +{ + u32 cp_int_cntl; + switch (state){ + case AMDGPU_IRQ_STATE_DISABLE: + if (ring == 0) { + cp_int_cntl = RREG32(CP_INT_CNTL_RING1); + cp_int_cntl &= ~CP_INT_CNTL_RING__TIME_STAMP_INT_ENABLE_MASK; + WREG32(CP_INT_CNTL_RING1, cp_int_cntl); + break; + } else { + cp_int_cntl = RREG32(CP_INT_CNTL_RING2); + cp_int_cntl &= ~CP_INT_CNTL_RING__TIME_STAMP_INT_ENABLE_MASK; + WREG32(CP_INT_CNTL_RING2, cp_int_cntl); + break; + + } + case AMDGPU_IRQ_STATE_ENABLE: + if (ring == 0) { + cp_int_cntl = RREG32(CP_INT_CNTL_RING1); + cp_int_cntl |= CP_INT_CNTL_RING__TIME_STAMP_INT_ENABLE_MASK; + WREG32(CP_INT_CNTL_RING1, cp_int_cntl); + break; + } else { + cp_int_cntl = RREG32(CP_INT_CNTL_RING2); + cp_int_cntl |= CP_INT_CNTL_RING__TIME_STAMP_INT_ENABLE_MASK; + WREG32(CP_INT_CNTL_RING2, cp_int_cntl); + break; + + } + + default: + BUG(); + break; + + } +} + +static int gfx_v6_0_set_priv_reg_fault_state(struct amdgpu_device *adev, + struct amdgpu_irq_src *src, + unsigned type, + enum amdgpu_interrupt_state state) +{ + u32 cp_int_cntl; + + switch (state) { + case AMDGPU_IRQ_STATE_DISABLE: + cp_int_cntl = RREG32(CP_INT_CNTL_RING0); + cp_int_cntl &= ~CP_INT_CNTL_RING0__PRIV_REG_INT_ENABLE_MASK; + WREG32(CP_INT_CNTL_RING0, cp_int_cntl); + break; + case AMDGPU_IRQ_STATE_ENABLE: + cp_int_cntl = RREG32(CP_INT_CNTL_RING0); + cp_int_cntl |= CP_INT_CNTL_RING0__PRIV_REG_INT_ENABLE_MASK; + WREG32(CP_INT_CNTL_RING0, cp_int_cntl); + break; + default: + break; + } + + return 0; +} + +static int gfx_v6_0_set_priv_inst_fault_state(struct amdgpu_device *adev, + struct amdgpu_irq_src *src, + unsigned type, + enum amdgpu_interrupt_state state) +{ + u32 cp_int_cntl; + + switch (state) { + case AMDGPU_IRQ_STATE_DISABLE: + cp_int_cntl = RREG32(CP_INT_CNTL_RING0); + cp_int_cntl &= ~CP_INT_CNTL_RING0__PRIV_INSTR_INT_ENABLE_MASK; + WREG32(CP_INT_CNTL_RING0, cp_int_cntl); + break; + case AMDGPU_IRQ_STATE_ENABLE: + cp_int_cntl = RREG32(CP_INT_CNTL_RING0); + cp_int_cntl |= CP_INT_CNTL_RING0__PRIV_INSTR_INT_ENABLE_MASK; + WREG32(CP_INT_CNTL_RING0, cp_int_cntl); + break; + default: + break; + } + + return 0; +} + +static int gfx_v6_0_set_eop_interrupt_state(struct amdgpu_device *adev, + struct amdgpu_irq_src *src, + unsigned type, + enum amdgpu_interrupt_state state) +{ + switch (type) { + case AMDGPU_CP_IRQ_GFX_EOP: + gfx_v6_0_set_gfx_eop_interrupt_state(adev, state); + break; + case AMDGPU_CP_IRQ_COMPUTE_MEC1_PIPE0_EOP: + gfx_v6_0_set_compute_eop_interrupt_state(adev, 0, state); + break; + case AMDGPU_CP_IRQ_COMPUTE_MEC1_PIPE1_EOP: + gfx_v6_0_set_compute_eop_interrupt_state(adev, 1, state); + break; + default: + break; + } + return 0; +} + +static int gfx_v6_0_eop_irq(struct amdgpu_device *adev, + struct amdgpu_irq_src *source, + struct amdgpu_iv_entry *entry) +{ + switch (entry->ring_id) { + case 0: + amdgpu_fence_process(&adev->gfx.gfx_ring[0]); + break; + case 1: + case 2: + amdgpu_fence_process(&adev->gfx.compute_ring[entry->ring_id -1]); + break; + default: + break; + } + return 0; +} + +static int gfx_v6_0_priv_reg_irq(struct amdgpu_device *adev, + struct amdgpu_irq_src *source, + struct amdgpu_iv_entry *entry) +{ + DRM_ERROR("Illegal register access in command stream\n"); + schedule_work(&adev->reset_work); + return 0; +} + +static int gfx_v6_0_priv_inst_irq(struct amdgpu_device *adev, + struct amdgpu_irq_src *source, + struct amdgpu_iv_entry *entry) +{ + DRM_ERROR("Illegal instruction in command stream\n"); + schedule_work(&adev->reset_work); + return 0; +} + +static int gfx_v6_0_set_clockgating_state(void *handle, + enum amd_clockgating_state state) +{ + bool gate = false; + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + if (state == AMD_CG_STATE_GATE) + gate = true; + + gfx_v6_0_enable_gui_idle_interrupt(adev, false); + if (gate) { + gfx_v6_0_enable_mgcg(adev, true); + gfx_v6_0_enable_cgcg(adev, true); + } else { + gfx_v6_0_enable_cgcg(adev, false); + gfx_v6_0_enable_mgcg(adev, false); + } + gfx_v6_0_enable_gui_idle_interrupt(adev, true); + + return 0; +} + +static int gfx_v6_0_set_powergating_state(void *handle, + enum amd_powergating_state state) +{ + bool gate = false; + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + if (state == AMD_PG_STATE_GATE) + gate = true; + + if (adev->pg_flags & (AMD_PG_SUPPORT_GFX_PG | + AMD_PG_SUPPORT_GFX_SMG | + AMD_PG_SUPPORT_GFX_DMG | + AMD_PG_SUPPORT_CP | + AMD_PG_SUPPORT_GDS | + AMD_PG_SUPPORT_RLC_SMU_HS)) { + gfx_v6_0_update_gfx_pg(adev, gate); + if (adev->pg_flags & AMD_PG_SUPPORT_GFX_PG) { + gfx_v6_0_enable_cp_pg(adev, gate); + gfx_v6_0_enable_gds_pg(adev, gate); + } + } + + return 0; +} + +const struct amd_ip_funcs gfx_v6_0_ip_funcs = { + .name = "gfx_v6_0", + .early_init = gfx_v6_0_early_init, + .late_init = NULL, + .sw_init = gfx_v6_0_sw_init, + .sw_fini = gfx_v6_0_sw_fini, + .hw_init = gfx_v6_0_hw_init, + .hw_fini = gfx_v6_0_hw_fini, + .suspend = gfx_v6_0_suspend, + .resume = gfx_v6_0_resume, + .is_idle = gfx_v6_0_is_idle, + .wait_for_idle = gfx_v6_0_wait_for_idle, + .soft_reset = gfx_v6_0_soft_reset, + .set_clockgating_state = gfx_v6_0_set_clockgating_state, + .set_powergating_state = gfx_v6_0_set_powergating_state, +}; + +static const struct amdgpu_ring_funcs gfx_v6_0_ring_funcs_gfx = { + .get_rptr = gfx_v6_0_ring_get_rptr, + .get_wptr = gfx_v6_0_ring_get_wptr, + .set_wptr = gfx_v6_0_ring_set_wptr_gfx, + .parse_cs = NULL, + .emit_ib = gfx_v6_0_ring_emit_ib, + .emit_fence = gfx_v6_0_ring_emit_fence, + .emit_pipeline_sync = gfx_v6_0_ring_emit_pipeline_sync, + .emit_vm_flush = gfx_v6_0_ring_emit_vm_flush, + .emit_hdp_flush = gfx_v6_0_ring_emit_hdp_flush, + .emit_hdp_invalidate = gfx_v6_0_ring_emit_hdp_invalidate, + .test_ring = gfx_v6_0_ring_test_ring, + .test_ib = gfx_v6_0_ring_test_ib, + .insert_nop = amdgpu_ring_insert_nop, + .emit_cntxcntl = gfx_v6_ring_emit_cntxcntl, + .get_emit_ib_size = gfx_v6_0_ring_get_emit_ib_size, + .get_dma_frame_size = gfx_v6_0_ring_get_dma_frame_size_gfx, +}; + +static const struct amdgpu_ring_funcs gfx_v6_0_ring_funcs_compute = { + .get_rptr = gfx_v6_0_ring_get_rptr, + .get_wptr = gfx_v6_0_ring_get_wptr, + .set_wptr = gfx_v6_0_ring_set_wptr_compute, + .parse_cs = NULL, + .emit_ib = gfx_v6_0_ring_emit_ib, + .emit_fence = gfx_v6_0_ring_emit_fence, + .emit_pipeline_sync = gfx_v6_0_ring_emit_pipeline_sync, + .emit_vm_flush = gfx_v6_0_ring_emit_vm_flush, + .emit_hdp_flush = gfx_v6_0_ring_emit_hdp_flush, + .emit_hdp_invalidate = gfx_v6_0_ring_emit_hdp_invalidate, + .test_ring = gfx_v6_0_ring_test_ring, + .test_ib = gfx_v6_0_ring_test_ib, + .insert_nop = amdgpu_ring_insert_nop, + .get_emit_ib_size = gfx_v6_0_ring_get_emit_ib_size, + .get_dma_frame_size = gfx_v6_0_ring_get_dma_frame_size_compute, +}; + +static void gfx_v6_0_set_ring_funcs(struct amdgpu_device *adev) +{ + int i; + + for (i = 0; i < adev->gfx.num_gfx_rings; i++) + adev->gfx.gfx_ring[i].funcs = &gfx_v6_0_ring_funcs_gfx; + for (i = 0; i < adev->gfx.num_compute_rings; i++) + adev->gfx.compute_ring[i].funcs = &gfx_v6_0_ring_funcs_compute; +} + +static const struct amdgpu_irq_src_funcs gfx_v6_0_eop_irq_funcs = { + .set = gfx_v6_0_set_eop_interrupt_state, + .process = gfx_v6_0_eop_irq, +}; + +static const struct amdgpu_irq_src_funcs gfx_v6_0_priv_reg_irq_funcs = { + .set = gfx_v6_0_set_priv_reg_fault_state, + .process = gfx_v6_0_priv_reg_irq, +}; + +static const struct amdgpu_irq_src_funcs gfx_v6_0_priv_inst_irq_funcs = { + .set = gfx_v6_0_set_priv_inst_fault_state, + .process = gfx_v6_0_priv_inst_irq, +}; + +static void gfx_v6_0_set_irq_funcs(struct amdgpu_device *adev) +{ + adev->gfx.eop_irq.num_types = AMDGPU_CP_IRQ_LAST; + adev->gfx.eop_irq.funcs = &gfx_v6_0_eop_irq_funcs; + + adev->gfx.priv_reg_irq.num_types = 1; + adev->gfx.priv_reg_irq.funcs = &gfx_v6_0_priv_reg_irq_funcs; + + adev->gfx.priv_inst_irq.num_types = 1; + adev->gfx.priv_inst_irq.funcs = &gfx_v6_0_priv_inst_irq_funcs; +} + +static void gfx_v6_0_get_cu_info(struct amdgpu_device *adev) +{ + int i, j, k, counter, active_cu_number = 0; + u32 mask, bitmap, ao_bitmap, ao_cu_mask = 0; + struct amdgpu_cu_info *cu_info = &adev->gfx.cu_info; + + memset(cu_info, 0, sizeof(*cu_info)); + + for (i = 0; i < adev->gfx.config.max_shader_engines; i++) { + for (j = 0; j < adev->gfx.config.max_sh_per_se; j++) { + mask = 1; + ao_bitmap = 0; + counter = 0; + bitmap = gfx_v6_0_get_cu_active_bitmap(adev, i, j); + cu_info->bitmap[i][j] = bitmap; + + for (k = 0; k < adev->gfx.config.max_cu_per_sh; k ++) { + if (bitmap & mask) { + if (counter < 2) + ao_bitmap |= mask; + counter ++; + } + mask <<= 1; + } + active_cu_number += counter; + ao_cu_mask |= (ao_bitmap << (i * 16 + j * 8)); + } + } + + cu_info->number = active_cu_number; + cu_info->ao_cu_mask = ao_cu_mask; +} diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.h b/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.h new file mode 100644 index 000000000000..b9657e72b248 --- /dev/null +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.h @@ -0,0 +1,29 @@ +/* + * Copyright 2015 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + */ + +#ifndef __GFX_V6_0_H__ +#define __GFX_V6_0_H__ + +extern const struct amd_ip_funcs gfx_v6_0_ip_funcs; + +#endif diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c index f4fbec3e224e..90102f123bb8 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c @@ -2096,6 +2096,25 @@ static void gfx_v7_0_ring_emit_ib_compute(struct amdgpu_ring *ring, amdgpu_ring_write(ring, control); } +static void gfx_v7_ring_emit_cntxcntl(struct amdgpu_ring *ring, uint32_t flags) +{ + uint32_t dw2 = 0; + + dw2 |= 0x80000000; /* set load_enable otherwise this package is just NOPs */ + if (flags & AMDGPU_HAVE_CTX_SWITCH) { + /* set load_global_config & load_global_uconfig */ + dw2 |= 0x8001; + /* set load_cs_sh_regs */ + dw2 |= 0x01000000; + /* set load_per_context_state & load_gfx_sh_regs */ + dw2 |= 0x10002; + } + + amdgpu_ring_write(ring, PACKET3(PACKET3_CONTEXT_CONTROL, 1)); + amdgpu_ring_write(ring, dw2); + amdgpu_ring_write(ring, 0); +} + /** * gfx_v7_0_ring_test_ib - basic ring IB test * @@ -2443,7 +2462,7 @@ static int gfx_v7_0_cp_gfx_resume(struct amdgpu_device *adev) return 0; } -static u32 gfx_v7_0_ring_get_rptr_gfx(struct amdgpu_ring *ring) +static u32 gfx_v7_0_ring_get_rptr(struct amdgpu_ring *ring) { return ring->adev->wb.wb[ring->rptr_offs]; } @@ -2463,11 +2482,6 @@ static void gfx_v7_0_ring_set_wptr_gfx(struct amdgpu_ring *ring) (void)RREG32(mmCP_RB0_WPTR); } -static u32 gfx_v7_0_ring_get_rptr_compute(struct amdgpu_ring *ring) -{ - return ring->adev->wb.wb[ring->rptr_offs]; -} - static u32 gfx_v7_0_ring_get_wptr_compute(struct amdgpu_ring *ring) { /* XXX check if swapping is necessary on BE */ @@ -4176,6 +4190,41 @@ static void gfx_v7_0_ring_emit_gds_switch(struct amdgpu_ring *ring, amdgpu_ring_write(ring, (1 << (oa_size + oa_base)) - (1 << oa_base)); } +static unsigned gfx_v7_0_ring_get_emit_ib_size_gfx(struct amdgpu_ring *ring) +{ + return + 4; /* gfx_v7_0_ring_emit_ib_gfx */ +} + +static unsigned gfx_v7_0_ring_get_dma_frame_size_gfx(struct amdgpu_ring *ring) +{ + return + 20 + /* gfx_v7_0_ring_emit_gds_switch */ + 7 + /* gfx_v7_0_ring_emit_hdp_flush */ + 5 + /* gfx_v7_0_ring_emit_hdp_invalidate */ + 12 + 12 + 12 + /* gfx_v7_0_ring_emit_fence_gfx x3 for user fence, vm fence */ + 7 + 4 + /* gfx_v7_0_ring_emit_pipeline_sync */ + 17 + 6 + /* gfx_v7_0_ring_emit_vm_flush */ + 3; /* gfx_v7_ring_emit_cntxcntl */ +} + +static unsigned gfx_v7_0_ring_get_emit_ib_size_compute(struct amdgpu_ring *ring) +{ + return + 4; /* gfx_v7_0_ring_emit_ib_compute */ +} + +static unsigned gfx_v7_0_ring_get_dma_frame_size_compute(struct amdgpu_ring *ring) +{ + return + 20 + /* gfx_v7_0_ring_emit_gds_switch */ + 7 + /* gfx_v7_0_ring_emit_hdp_flush */ + 5 + /* gfx_v7_0_ring_emit_hdp_invalidate */ + 7 + /* gfx_v7_0_ring_emit_pipeline_sync */ + 17 + /* gfx_v7_0_ring_emit_vm_flush */ + 7 + 7 + 7; /* gfx_v7_0_ring_emit_fence_compute x3 for user fence, vm fence */ +} + static const struct amdgpu_gfx_funcs gfx_v7_0_gfx_funcs = { .get_gpu_clock_counter = &gfx_v7_0_get_gpu_clock_counter, .select_se_sh = &gfx_v7_0_select_se_sh, @@ -4495,9 +4544,9 @@ static int gfx_v7_0_sw_fini(void *handle) int i; struct amdgpu_device *adev = (struct amdgpu_device *)handle; - amdgpu_bo_unref(&adev->gds.oa_gfx_bo); - amdgpu_bo_unref(&adev->gds.gws_gfx_bo); - amdgpu_bo_unref(&adev->gds.gds_gfx_bo); + amdgpu_bo_free_kernel(&adev->gds.oa_gfx_bo, NULL, NULL); + amdgpu_bo_free_kernel(&adev->gds.gws_gfx_bo, NULL, NULL); + amdgpu_bo_free_kernel(&adev->gds.gds_gfx_bo, NULL, NULL); for (i = 0; i < adev->gfx.num_gfx_rings; i++) amdgpu_ring_fini(&adev->gfx.gfx_ring[i]); @@ -4928,7 +4977,7 @@ const struct amd_ip_funcs gfx_v7_0_ip_funcs = { }; static const struct amdgpu_ring_funcs gfx_v7_0_ring_funcs_gfx = { - .get_rptr = gfx_v7_0_ring_get_rptr_gfx, + .get_rptr = gfx_v7_0_ring_get_rptr, .get_wptr = gfx_v7_0_ring_get_wptr_gfx, .set_wptr = gfx_v7_0_ring_set_wptr_gfx, .parse_cs = NULL, @@ -4943,10 +4992,13 @@ static const struct amdgpu_ring_funcs gfx_v7_0_ring_funcs_gfx = { .test_ib = gfx_v7_0_ring_test_ib, .insert_nop = amdgpu_ring_insert_nop, .pad_ib = amdgpu_ring_generic_pad_ib, + .emit_cntxcntl = gfx_v7_ring_emit_cntxcntl, + .get_emit_ib_size = gfx_v7_0_ring_get_emit_ib_size_gfx, + .get_dma_frame_size = gfx_v7_0_ring_get_dma_frame_size_gfx, }; static const struct amdgpu_ring_funcs gfx_v7_0_ring_funcs_compute = { - .get_rptr = gfx_v7_0_ring_get_rptr_compute, + .get_rptr = gfx_v7_0_ring_get_rptr, .get_wptr = gfx_v7_0_ring_get_wptr_compute, .set_wptr = gfx_v7_0_ring_set_wptr_compute, .parse_cs = NULL, @@ -4961,6 +5013,8 @@ static const struct amdgpu_ring_funcs gfx_v7_0_ring_funcs_compute = { .test_ib = gfx_v7_0_ring_test_ib, .insert_nop = amdgpu_ring_insert_nop, .pad_ib = amdgpu_ring_generic_pad_ib, + .get_emit_ib_size = gfx_v7_0_ring_get_emit_ib_size_compute, + .get_dma_frame_size = gfx_v7_0_ring_get_dma_frame_size_compute, }; static void gfx_v7_0_set_ring_funcs(struct amdgpu_device *adev) diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c index c6a63c2f91e3..47e270ad4fe3 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c @@ -2113,9 +2113,9 @@ static int gfx_v8_0_sw_fini(void *handle) int i; struct amdgpu_device *adev = (struct amdgpu_device *)handle; - amdgpu_bo_unref(&adev->gds.oa_gfx_bo); - amdgpu_bo_unref(&adev->gds.gws_gfx_bo); - amdgpu_bo_unref(&adev->gds.gds_gfx_bo); + amdgpu_bo_free_kernel(&adev->gds.oa_gfx_bo, NULL, NULL); + amdgpu_bo_free_kernel(&adev->gds.gws_gfx_bo, NULL, NULL); + amdgpu_bo_free_kernel(&adev->gds.gds_gfx_bo, NULL, NULL); for (i = 0; i < adev->gfx.num_gfx_rings; i++) amdgpu_ring_fini(&adev->gfx.gfx_ring[i]); @@ -3866,7 +3866,7 @@ static void gfx_v8_0_init_pg(struct amdgpu_device *adev) } } -void gfx_v8_0_rlc_stop(struct amdgpu_device *adev) +static void gfx_v8_0_rlc_stop(struct amdgpu_device *adev) { WREG32_FIELD(RLC_CNTL, RLC_ENABLE_F32, 0); @@ -5835,7 +5835,7 @@ static int gfx_v8_0_set_clockgating_state(void *handle, return 0; } -static u32 gfx_v8_0_ring_get_rptr_gfx(struct amdgpu_ring *ring) +static u32 gfx_v8_0_ring_get_rptr(struct amdgpu_ring *ring) { return ring->adev->wb.wb[ring->rptr_offs]; } @@ -5915,12 +5915,6 @@ static void gfx_v8_0_ring_emit_ib_gfx(struct amdgpu_ring *ring, { u32 header, control = 0; - /* insert SWITCH_BUFFER packet before first IB in the ring frame */ - if (ctx_switch) { - amdgpu_ring_write(ring, PACKET3(PACKET3_SWITCH_BUFFER, 0)); - amdgpu_ring_write(ring, 0); - } - if (ib->flags & AMDGPU_IB_FLAG_CE) header = PACKET3(PACKET3_INDIRECT_BUFFER_CONST, 2); else @@ -5990,14 +5984,6 @@ static void gfx_v8_0_ring_emit_pipeline_sync(struct amdgpu_ring *ring) amdgpu_ring_write(ring, seq); amdgpu_ring_write(ring, 0xffffffff); amdgpu_ring_write(ring, 4); /* poll interval */ - - if (usepfp) { - /* synce CE with ME to prevent CE fetch CEIB before context switch done */ - amdgpu_ring_write(ring, PACKET3(PACKET3_SWITCH_BUFFER, 0)); - amdgpu_ring_write(ring, 0); - amdgpu_ring_write(ring, PACKET3(PACKET3_SWITCH_BUFFER, 0)); - amdgpu_ring_write(ring, 0); - } } static void gfx_v8_0_ring_emit_vm_flush(struct amdgpu_ring *ring, @@ -6005,6 +5991,10 @@ static void gfx_v8_0_ring_emit_vm_flush(struct amdgpu_ring *ring, { int usepfp = (ring->type == AMDGPU_RING_TYPE_GFX); + /* GFX8 emits 128 dw nop to prevent DE do vm_flush before CE finish CEIB */ + if (usepfp) + amdgpu_ring_insert_nop(ring, 128); + amdgpu_ring_write(ring, PACKET3(PACKET3_WRITE_DATA, 3)); amdgpu_ring_write(ring, (WRITE_DATA_ENGINE_SEL(usepfp) | WRITE_DATA_DST_SEL(0)) | @@ -6044,18 +6034,11 @@ static void gfx_v8_0_ring_emit_vm_flush(struct amdgpu_ring *ring, /* sync PFP to ME, otherwise we might get invalid PFP reads */ amdgpu_ring_write(ring, PACKET3(PACKET3_PFP_SYNC_ME, 0)); amdgpu_ring_write(ring, 0x0); - amdgpu_ring_write(ring, PACKET3(PACKET3_SWITCH_BUFFER, 0)); - amdgpu_ring_write(ring, 0); - amdgpu_ring_write(ring, PACKET3(PACKET3_SWITCH_BUFFER, 0)); - amdgpu_ring_write(ring, 0); + /* GFX8 emits 128 dw nop to prevent CE access VM before vm_flush finish */ + amdgpu_ring_insert_nop(ring, 128); } } -static u32 gfx_v8_0_ring_get_rptr_compute(struct amdgpu_ring *ring) -{ - return ring->adev->wb.wb[ring->rptr_offs]; -} - static u32 gfx_v8_0_ring_get_wptr_compute(struct amdgpu_ring *ring) { return ring->adev->wb.wb[ring->wptr_offs]; @@ -6091,6 +6074,77 @@ static void gfx_v8_0_ring_emit_fence_compute(struct amdgpu_ring *ring, amdgpu_ring_write(ring, upper_32_bits(seq)); } +static void gfx_v8_ring_emit_sb(struct amdgpu_ring *ring) +{ + amdgpu_ring_write(ring, PACKET3(PACKET3_SWITCH_BUFFER, 0)); + amdgpu_ring_write(ring, 0); +} + +static void gfx_v8_ring_emit_cntxcntl(struct amdgpu_ring *ring, uint32_t flags) +{ + uint32_t dw2 = 0; + + dw2 |= 0x80000000; /* set load_enable otherwise this package is just NOPs */ + if (flags & AMDGPU_HAVE_CTX_SWITCH) { + /* set load_global_config & load_global_uconfig */ + dw2 |= 0x8001; + /* set load_cs_sh_regs */ + dw2 |= 0x01000000; + /* set load_per_context_state & load_gfx_sh_regs for GFX */ + dw2 |= 0x10002; + + /* set load_ce_ram if preamble presented */ + if (AMDGPU_PREAMBLE_IB_PRESENT & flags) + dw2 |= 0x10000000; + } else { + /* still load_ce_ram if this is the first time preamble presented + * although there is no context switch happens. + */ + if (AMDGPU_PREAMBLE_IB_PRESENT_FIRST & flags) + dw2 |= 0x10000000; + } + + amdgpu_ring_write(ring, PACKET3(PACKET3_CONTEXT_CONTROL, 1)); + amdgpu_ring_write(ring, dw2); + amdgpu_ring_write(ring, 0); +} + +static unsigned gfx_v8_0_ring_get_emit_ib_size_gfx(struct amdgpu_ring *ring) +{ + return + 4; /* gfx_v8_0_ring_emit_ib_gfx */ +} + +static unsigned gfx_v8_0_ring_get_dma_frame_size_gfx(struct amdgpu_ring *ring) +{ + return + 20 + /* gfx_v8_0_ring_emit_gds_switch */ + 7 + /* gfx_v8_0_ring_emit_hdp_flush */ + 5 + /* gfx_v8_0_ring_emit_hdp_invalidate */ + 6 + 6 + 6 +/* gfx_v8_0_ring_emit_fence_gfx x3 for user fence, vm fence */ + 7 + /* gfx_v8_0_ring_emit_pipeline_sync */ + 256 + 19 + /* gfx_v8_0_ring_emit_vm_flush */ + 2 + /* gfx_v8_ring_emit_sb */ + 3; /* gfx_v8_ring_emit_cntxcntl */ +} + +static unsigned gfx_v8_0_ring_get_emit_ib_size_compute(struct amdgpu_ring *ring) +{ + return + 4; /* gfx_v8_0_ring_emit_ib_compute */ +} + +static unsigned gfx_v8_0_ring_get_dma_frame_size_compute(struct amdgpu_ring *ring) +{ + return + 20 + /* gfx_v8_0_ring_emit_gds_switch */ + 7 + /* gfx_v8_0_ring_emit_hdp_flush */ + 5 + /* gfx_v8_0_ring_emit_hdp_invalidate */ + 7 + /* gfx_v8_0_ring_emit_pipeline_sync */ + 17 + /* gfx_v8_0_ring_emit_vm_flush */ + 7 + 7 + 7; /* gfx_v8_0_ring_emit_fence_compute x3 for user fence, vm fence */ +} + static void gfx_v8_0_set_gfx_eop_interrupt_state(struct amdgpu_device *adev, enum amdgpu_interrupt_state state) { @@ -6257,7 +6311,7 @@ const struct amd_ip_funcs gfx_v8_0_ip_funcs = { }; static const struct amdgpu_ring_funcs gfx_v8_0_ring_funcs_gfx = { - .get_rptr = gfx_v8_0_ring_get_rptr_gfx, + .get_rptr = gfx_v8_0_ring_get_rptr, .get_wptr = gfx_v8_0_ring_get_wptr_gfx, .set_wptr = gfx_v8_0_ring_set_wptr_gfx, .parse_cs = NULL, @@ -6272,10 +6326,14 @@ static const struct amdgpu_ring_funcs gfx_v8_0_ring_funcs_gfx = { .test_ib = gfx_v8_0_ring_test_ib, .insert_nop = amdgpu_ring_insert_nop, .pad_ib = amdgpu_ring_generic_pad_ib, + .emit_switch_buffer = gfx_v8_ring_emit_sb, + .emit_cntxcntl = gfx_v8_ring_emit_cntxcntl, + .get_emit_ib_size = gfx_v8_0_ring_get_emit_ib_size_gfx, + .get_dma_frame_size = gfx_v8_0_ring_get_dma_frame_size_gfx, }; static const struct amdgpu_ring_funcs gfx_v8_0_ring_funcs_compute = { - .get_rptr = gfx_v8_0_ring_get_rptr_compute, + .get_rptr = gfx_v8_0_ring_get_rptr, .get_wptr = gfx_v8_0_ring_get_wptr_compute, .set_wptr = gfx_v8_0_ring_set_wptr_compute, .parse_cs = NULL, @@ -6290,6 +6348,8 @@ static const struct amdgpu_ring_funcs gfx_v8_0_ring_funcs_compute = { .test_ib = gfx_v8_0_ring_test_ib, .insert_nop = amdgpu_ring_insert_nop, .pad_ib = amdgpu_ring_generic_pad_ib, + .get_emit_ib_size = gfx_v8_0_ring_get_emit_ib_size_compute, + .get_dma_frame_size = gfx_v8_0_ring_get_dma_frame_size_compute, }; static void gfx_v8_0_set_ring_funcs(struct amdgpu_device *adev) diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c new file mode 100644 index 000000000000..b13c8aaec078 --- /dev/null +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c @@ -0,0 +1,1071 @@ + +/* + * Copyright 2014 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + */ +#include <linux/firmware.h> +#include "drmP.h" +#include "amdgpu.h" +#include "gmc_v6_0.h" +#include "amdgpu_ucode.h" +#include "si/sid.h" + +static void gmc_v6_0_set_gart_funcs(struct amdgpu_device *adev); +static void gmc_v6_0_set_irq_funcs(struct amdgpu_device *adev); +static int gmc_v6_0_wait_for_idle(void *handle); + +MODULE_FIRMWARE("radeon/tahiti_mc.bin"); +MODULE_FIRMWARE("radeon/pitcairn_mc.bin"); +MODULE_FIRMWARE("radeon/verde_mc.bin"); +MODULE_FIRMWARE("radeon/oland_mc.bin"); + +static const u32 crtc_offsets[6] = +{ + SI_CRTC0_REGISTER_OFFSET, + SI_CRTC1_REGISTER_OFFSET, + SI_CRTC2_REGISTER_OFFSET, + SI_CRTC3_REGISTER_OFFSET, + SI_CRTC4_REGISTER_OFFSET, + SI_CRTC5_REGISTER_OFFSET +}; + +static void gmc_v6_0_mc_stop(struct amdgpu_device *adev, + struct amdgpu_mode_mc_save *save) +{ + u32 blackout; + + if (adev->mode_info.num_crtc) + amdgpu_display_stop_mc_access(adev, save); + + gmc_v6_0_wait_for_idle((void *)adev); + + blackout = RREG32(MC_SHARED_BLACKOUT_CNTL); + if (REG_GET_FIELD(blackout, mmMC_SHARED_BLACKOUT_CNTL, xxBLACKOUT_MODE) != 1) { + /* Block CPU access */ + WREG32(BIF_FB_EN, 0); + /* blackout the MC */ + blackout = REG_SET_FIELD(blackout, + mmMC_SHARED_BLACKOUT_CNTL, xxBLACKOUT_MODE, 0); + WREG32(MC_SHARED_BLACKOUT_CNTL, blackout | 1); + } + /* wait for the MC to settle */ + udelay(100); + +} + +static void gmc_v6_0_mc_resume(struct amdgpu_device *adev, + struct amdgpu_mode_mc_save *save) +{ + u32 tmp; + + /* unblackout the MC */ + tmp = RREG32(MC_SHARED_BLACKOUT_CNTL); + tmp = REG_SET_FIELD(tmp, mmMC_SHARED_BLACKOUT_CNTL, xxBLACKOUT_MODE, 0); + WREG32(MC_SHARED_BLACKOUT_CNTL, tmp); + /* allow CPU access */ + tmp = REG_SET_FIELD(0, mmBIF_FB_EN, xxFB_READ_EN, 1); + tmp = REG_SET_FIELD(tmp, mmBIF_FB_EN, xxFB_WRITE_EN, 1); + WREG32(BIF_FB_EN, tmp); + + if (adev->mode_info.num_crtc) + amdgpu_display_resume_mc_access(adev, save); + +} + +static int gmc_v6_0_init_microcode(struct amdgpu_device *adev) +{ + const char *chip_name; + char fw_name[30]; + int err; + + DRM_DEBUG("\n"); + + switch (adev->asic_type) { + case CHIP_TAHITI: + chip_name = "tahiti"; + break; + case CHIP_PITCAIRN: + chip_name = "pitcairn"; + break; + case CHIP_VERDE: + chip_name = "verde"; + break; + case CHIP_OLAND: + chip_name = "oland"; + break; + case CHIP_HAINAN: + chip_name = "hainan"; + break; + default: BUG(); + } + + snprintf(fw_name, sizeof(fw_name), "radeon/%s_mc.bin", chip_name); + err = request_firmware(&adev->mc.fw, fw_name, adev->dev); + if (err) + goto out; + + err = amdgpu_ucode_validate(adev->mc.fw); + +out: + if (err) { + dev_err(adev->dev, + "si_mc: Failed to load firmware \"%s\"\n", + fw_name); + release_firmware(adev->mc.fw); + adev->mc.fw = NULL; + } + return err; +} + +static int gmc_v6_0_mc_load_microcode(struct amdgpu_device *adev) +{ + const __le32 *new_fw_data = NULL; + u32 running; + const __le32 *new_io_mc_regs = NULL; + int i, regs_size, ucode_size; + const struct mc_firmware_header_v1_0 *hdr; + + if (!adev->mc.fw) + return -EINVAL; + + hdr = (const struct mc_firmware_header_v1_0 *)adev->mc.fw->data; + + amdgpu_ucode_print_mc_hdr(&hdr->header); + + adev->mc.fw_version = le32_to_cpu(hdr->header.ucode_version); + regs_size = le32_to_cpu(hdr->io_debug_size_bytes) / (4 * 2); + new_io_mc_regs = (const __le32 *) + (adev->mc.fw->data + le32_to_cpu(hdr->io_debug_array_offset_bytes)); + ucode_size = le32_to_cpu(hdr->header.ucode_size_bytes) / 4; + new_fw_data = (const __le32 *) + (adev->mc.fw->data + le32_to_cpu(hdr->header.ucode_array_offset_bytes)); + + running = RREG32(MC_SEQ_SUP_CNTL) & RUN_MASK; + + if (running == 0) { + + /* reset the engine and set to writable */ + WREG32(MC_SEQ_SUP_CNTL, 0x00000008); + WREG32(MC_SEQ_SUP_CNTL, 0x00000010); + + /* load mc io regs */ + for (i = 0; i < regs_size; i++) { + WREG32(MC_SEQ_IO_DEBUG_INDEX, le32_to_cpup(new_io_mc_regs++)); + WREG32(MC_SEQ_IO_DEBUG_DATA, le32_to_cpup(new_io_mc_regs++)); + } + /* load the MC ucode */ + for (i = 0; i < ucode_size; i++) { + WREG32(MC_SEQ_SUP_PGM, le32_to_cpup(new_fw_data++)); + } + + /* put the engine back into the active state */ + WREG32(MC_SEQ_SUP_CNTL, 0x00000008); + WREG32(MC_SEQ_SUP_CNTL, 0x00000004); + WREG32(MC_SEQ_SUP_CNTL, 0x00000001); + + /* wait for training to complete */ + for (i = 0; i < adev->usec_timeout; i++) { + if (RREG32(MC_SEQ_TRAIN_WAKEUP_CNTL) & TRAIN_DONE_D0) + break; + udelay(1); + } + for (i = 0; i < adev->usec_timeout; i++) { + if (RREG32(MC_SEQ_TRAIN_WAKEUP_CNTL) & TRAIN_DONE_D1) + break; + udelay(1); + } + + } + + return 0; +} + +static void gmc_v6_0_vram_gtt_location(struct amdgpu_device *adev, + struct amdgpu_mc *mc) +{ + if (mc->mc_vram_size > 0xFFC0000000ULL) { + dev_warn(adev->dev, "limiting VRAM\n"); + mc->real_vram_size = 0xFFC0000000ULL; + mc->mc_vram_size = 0xFFC0000000ULL; + } + amdgpu_vram_location(adev, &adev->mc, 0); + adev->mc.gtt_base_align = 0; + amdgpu_gtt_location(adev, mc); +} + +static void gmc_v6_0_mc_program(struct amdgpu_device *adev) +{ + struct amdgpu_mode_mc_save save; + u32 tmp; + int i, j; + + /* Initialize HDP */ + for (i = 0, j = 0; i < 32; i++, j += 0x6) { + WREG32((0xb05 + j), 0x00000000); + WREG32((0xb06 + j), 0x00000000); + WREG32((0xb07 + j), 0x00000000); + WREG32((0xb08 + j), 0x00000000); + WREG32((0xb09 + j), 0x00000000); + } + WREG32(HDP_REG_COHERENCY_FLUSH_CNTL, 0); + + gmc_v6_0_mc_stop(adev, &save); + + if (gmc_v6_0_wait_for_idle((void *)adev)) { + dev_warn(adev->dev, "Wait for MC idle timedout !\n"); + } + + WREG32(VGA_HDP_CONTROL, VGA_MEMORY_DISABLE); + /* Update configuration */ + WREG32(MC_VM_SYSTEM_APERTURE_LOW_ADDR, + adev->mc.vram_start >> 12); + WREG32(MC_VM_SYSTEM_APERTURE_HIGH_ADDR, + adev->mc.vram_end >> 12); + WREG32(MC_VM_SYSTEM_APERTURE_DEFAULT_ADDR, + adev->vram_scratch.gpu_addr >> 12); + tmp = ((adev->mc.vram_end >> 24) & 0xFFFF) << 16; + tmp |= ((adev->mc.vram_start >> 24) & 0xFFFF); + WREG32(MC_VM_FB_LOCATION, tmp); + /* XXX double check these! */ + WREG32(HDP_NONSURFACE_BASE, (adev->mc.vram_start >> 8)); + WREG32(HDP_NONSURFACE_INFO, (2 << 7) | (1 << 30)); + WREG32(HDP_NONSURFACE_SIZE, 0x3FFFFFFF); + WREG32(MC_VM_AGP_BASE, 0); + WREG32(MC_VM_AGP_TOP, 0x0FFFFFFF); + WREG32(MC_VM_AGP_BOT, 0x0FFFFFFF); + + if (gmc_v6_0_wait_for_idle((void *)adev)) { + dev_warn(adev->dev, "Wait for MC idle timedout !\n"); + } + gmc_v6_0_mc_resume(adev, &save); + amdgpu_display_set_vga_render_state(adev, false); +} + +static int gmc_v6_0_mc_init(struct amdgpu_device *adev) +{ + + u32 tmp; + int chansize, numchan; + + tmp = RREG32(MC_ARB_RAMCFG); + if (tmp & CHANSIZE_OVERRIDE) { + chansize = 16; + } else if (tmp & CHANSIZE_MASK) { + chansize = 64; + } else { + chansize = 32; + } + tmp = RREG32(MC_SHARED_CHMAP); + switch ((tmp & NOOFCHAN_MASK) >> NOOFCHAN_SHIFT) { + case 0: + default: + numchan = 1; + break; + case 1: + numchan = 2; + break; + case 2: + numchan = 4; + break; + case 3: + numchan = 8; + break; + case 4: + numchan = 3; + break; + case 5: + numchan = 6; + break; + case 6: + numchan = 10; + break; + case 7: + numchan = 12; + break; + case 8: + numchan = 16; + break; + } + adev->mc.vram_width = numchan * chansize; + /* Could aper size report 0 ? */ + adev->mc.aper_base = pci_resource_start(adev->pdev, 0); + adev->mc.aper_size = pci_resource_len(adev->pdev, 0); + /* size in MB on si */ + adev->mc.mc_vram_size = RREG32(CONFIG_MEMSIZE) * 1024ULL * 1024ULL; + adev->mc.real_vram_size = RREG32(CONFIG_MEMSIZE) * 1024ULL * 1024ULL; + adev->mc.visible_vram_size = adev->mc.aper_size; + + /* unless the user had overridden it, set the gart + * size equal to the 1024 or vram, whichever is larger. + */ + if (amdgpu_gart_size == -1) + adev->mc.gtt_size = amdgpu_ttm_get_gtt_mem_size(adev); + else + adev->mc.gtt_size = (uint64_t)amdgpu_gart_size << 20; + + gmc_v6_0_vram_gtt_location(adev, &adev->mc); + + return 0; +} + +static void gmc_v6_0_gart_flush_gpu_tlb(struct amdgpu_device *adev, + uint32_t vmid) +{ + WREG32(HDP_MEM_COHERENCY_FLUSH_CNTL, 0); + + WREG32(VM_INVALIDATE_REQUEST, 1 << vmid); +} + +static int gmc_v6_0_gart_set_pte_pde(struct amdgpu_device *adev, + void *cpu_pt_addr, + uint32_t gpu_page_idx, + uint64_t addr, + uint32_t flags) +{ + void __iomem *ptr = (void *)cpu_pt_addr; + uint64_t value; + + value = addr & 0xFFFFFFFFFFFFF000ULL; + value |= flags; + writeq(value, ptr + (gpu_page_idx * 8)); + + return 0; +} + +static void gmc_v6_0_set_fault_enable_default(struct amdgpu_device *adev, + bool value) +{ + u32 tmp; + + tmp = RREG32(VM_CONTEXT1_CNTL); + tmp = REG_SET_FIELD(tmp, mmVM_CONTEXT1_CNTL, + xxRANGE_PROTECTION_FAULT_ENABLE_DEFAULT, value); + tmp = REG_SET_FIELD(tmp, mmVM_CONTEXT1_CNTL, + xxDUMMY_PAGE_PROTECTION_FAULT_ENABLE_DEFAULT, value); + tmp = REG_SET_FIELD(tmp, mmVM_CONTEXT1_CNTL, + xxPDE0_PROTECTION_FAULT_ENABLE_DEFAULT, value); + tmp = REG_SET_FIELD(tmp, mmVM_CONTEXT1_CNTL, + xxVALID_PROTECTION_FAULT_ENABLE_DEFAULT, value); + tmp = REG_SET_FIELD(tmp, mmVM_CONTEXT1_CNTL, + xxREAD_PROTECTION_FAULT_ENABLE_DEFAULT, value); + tmp = REG_SET_FIELD(tmp, mmVM_CONTEXT1_CNTL, + xxWRITE_PROTECTION_FAULT_ENABLE_DEFAULT, value); + WREG32(VM_CONTEXT1_CNTL, tmp); +} + +static int gmc_v6_0_gart_enable(struct amdgpu_device *adev) +{ + int r, i; + + if (adev->gart.robj == NULL) { + dev_err(adev->dev, "No VRAM object for PCIE GART.\n"); + return -EINVAL; + } + r = amdgpu_gart_table_vram_pin(adev); + if (r) + return r; + /* Setup TLB control */ + WREG32(MC_VM_MX_L1_TLB_CNTL, + (0xA << 7) | + ENABLE_L1_TLB | + ENABLE_L1_FRAGMENT_PROCESSING | + SYSTEM_ACCESS_MODE_NOT_IN_SYS | + ENABLE_ADVANCED_DRIVER_MODEL | + SYSTEM_APERTURE_UNMAPPED_ACCESS_PASS_THRU); + /* Setup L2 cache */ + WREG32(VM_L2_CNTL, ENABLE_L2_CACHE | + ENABLE_L2_FRAGMENT_PROCESSING | + ENABLE_L2_PTE_CACHE_LRU_UPDATE_BY_WRITE | + ENABLE_L2_PDE0_CACHE_LRU_UPDATE_BY_WRITE | + EFFECTIVE_L2_QUEUE_SIZE(7) | + CONTEXT1_IDENTITY_ACCESS_MODE(1)); + WREG32(VM_L2_CNTL2, INVALIDATE_ALL_L1_TLBS | INVALIDATE_L2_CACHE); + WREG32(VM_L2_CNTL3, L2_CACHE_BIGK_ASSOCIATIVITY | + BANK_SELECT(4) | + L2_CACHE_BIGK_FRAGMENT_SIZE(4)); + /* setup context0 */ + WREG32(VM_CONTEXT0_PAGE_TABLE_START_ADDR, adev->mc.gtt_start >> 12); + WREG32(VM_CONTEXT0_PAGE_TABLE_END_ADDR, adev->mc.gtt_end >> 12); + WREG32(VM_CONTEXT0_PAGE_TABLE_BASE_ADDR, adev->gart.table_addr >> 12); + WREG32(VM_CONTEXT0_PROTECTION_FAULT_DEFAULT_ADDR, + (u32)(adev->dummy_page.addr >> 12)); + WREG32(VM_CONTEXT0_CNTL2, 0); + WREG32(VM_CONTEXT0_CNTL, (ENABLE_CONTEXT | PAGE_TABLE_DEPTH(0) | + RANGE_PROTECTION_FAULT_ENABLE_DEFAULT)); + + WREG32(0x575, 0); + WREG32(0x576, 0); + WREG32(0x577, 0); + + /* empty context1-15 */ + /* set vm size, must be a multiple of 4 */ + WREG32(VM_CONTEXT1_PAGE_TABLE_START_ADDR, 0); + WREG32(VM_CONTEXT1_PAGE_TABLE_END_ADDR, adev->vm_manager.max_pfn - 1); + /* Assign the pt base to something valid for now; the pts used for + * the VMs are determined by the application and setup and assigned + * on the fly in the vm part of radeon_gart.c + */ + for (i = 1; i < 16; i++) { + if (i < 8) + WREG32(VM_CONTEXT0_PAGE_TABLE_BASE_ADDR + i, + adev->gart.table_addr >> 12); + else + WREG32(VM_CONTEXT8_PAGE_TABLE_BASE_ADDR + i - 8, + adev->gart.table_addr >> 12); + } + + /* enable context1-15 */ + WREG32(VM_CONTEXT1_PROTECTION_FAULT_DEFAULT_ADDR, + (u32)(adev->dummy_page.addr >> 12)); + WREG32(VM_CONTEXT1_CNTL2, 4); + WREG32(VM_CONTEXT1_CNTL, ENABLE_CONTEXT | PAGE_TABLE_DEPTH(1) | + PAGE_TABLE_BLOCK_SIZE(amdgpu_vm_block_size - 9) | + RANGE_PROTECTION_FAULT_ENABLE_INTERRUPT | + RANGE_PROTECTION_FAULT_ENABLE_DEFAULT | + DUMMY_PAGE_PROTECTION_FAULT_ENABLE_INTERRUPT | + DUMMY_PAGE_PROTECTION_FAULT_ENABLE_DEFAULT | + PDE0_PROTECTION_FAULT_ENABLE_INTERRUPT | + PDE0_PROTECTION_FAULT_ENABLE_DEFAULT | + VALID_PROTECTION_FAULT_ENABLE_INTERRUPT | + VALID_PROTECTION_FAULT_ENABLE_DEFAULT | + READ_PROTECTION_FAULT_ENABLE_INTERRUPT | + READ_PROTECTION_FAULT_ENABLE_DEFAULT | + WRITE_PROTECTION_FAULT_ENABLE_INTERRUPT | + WRITE_PROTECTION_FAULT_ENABLE_DEFAULT); + + gmc_v6_0_gart_flush_gpu_tlb(adev, 0); + dev_info(adev->dev, "PCIE GART of %uM enabled (table at 0x%016llX).\n", + (unsigned)(adev->mc.gtt_size >> 20), + (unsigned long long)adev->gart.table_addr); + adev->gart.ready = true; + return 0; +} + +static int gmc_v6_0_gart_init(struct amdgpu_device *adev) +{ + int r; + + if (adev->gart.robj) { + dev_warn(adev->dev, "gmc_v6_0 PCIE GART already initialized\n"); + return 0; + } + r = amdgpu_gart_init(adev); + if (r) + return r; + adev->gart.table_size = adev->gart.num_gpu_pages * 8; + return amdgpu_gart_table_vram_alloc(adev); +} + +static void gmc_v6_0_gart_disable(struct amdgpu_device *adev) +{ + /*unsigned i; + + for (i = 1; i < 16; ++i) { + uint32_t reg; + if (i < 8) + reg = VM_CONTEXT0_PAGE_TABLE_BASE_ADDR + i ; + else + reg = VM_CONTEXT8_PAGE_TABLE_BASE_ADDR + (i - 8); + adev->vm_manager.saved_table_addr[i] = RREG32(reg); + }*/ + + /* Disable all tables */ + WREG32(VM_CONTEXT0_CNTL, 0); + WREG32(VM_CONTEXT1_CNTL, 0); + /* Setup TLB control */ + WREG32(MC_VM_MX_L1_TLB_CNTL, SYSTEM_ACCESS_MODE_NOT_IN_SYS | + SYSTEM_APERTURE_UNMAPPED_ACCESS_PASS_THRU); + /* Setup L2 cache */ + WREG32(VM_L2_CNTL, ENABLE_L2_PTE_CACHE_LRU_UPDATE_BY_WRITE | + ENABLE_L2_PDE0_CACHE_LRU_UPDATE_BY_WRITE | + EFFECTIVE_L2_QUEUE_SIZE(7) | + CONTEXT1_IDENTITY_ACCESS_MODE(1)); + WREG32(VM_L2_CNTL2, 0); + WREG32(VM_L2_CNTL3, L2_CACHE_BIGK_ASSOCIATIVITY | + L2_CACHE_BIGK_FRAGMENT_SIZE(0)); + amdgpu_gart_table_vram_unpin(adev); +} + +static void gmc_v6_0_gart_fini(struct amdgpu_device *adev) +{ + amdgpu_gart_table_vram_free(adev); + amdgpu_gart_fini(adev); +} + +static int gmc_v6_0_vm_init(struct amdgpu_device *adev) +{ + /* + * number of VMs + * VMID 0 is reserved for System + * amdgpu graphics/compute will use VMIDs 1-7 + * amdkfd will use VMIDs 8-15 + */ + adev->vm_manager.num_ids = AMDGPU_NUM_OF_VMIDS; + amdgpu_vm_manager_init(adev); + + /* base offset of vram pages */ + if (adev->flags & AMD_IS_APU) { + u64 tmp = RREG32(MC_VM_FB_OFFSET); + tmp <<= 22; + adev->vm_manager.vram_base_offset = tmp; + } else + adev->vm_manager.vram_base_offset = 0; + + return 0; +} + +static void gmc_v6_0_vm_fini(struct amdgpu_device *adev) +{ +} + +static void gmc_v6_0_vm_decode_fault(struct amdgpu_device *adev, + u32 status, u32 addr, u32 mc_client) +{ + u32 mc_id; + u32 vmid = REG_GET_FIELD(status, mmVM_CONTEXT1_PROTECTION_FAULT_STATUS, xxVMID); + u32 protections = REG_GET_FIELD(status, mmVM_CONTEXT1_PROTECTION_FAULT_STATUS, + xxPROTECTIONS); + char block[5] = { mc_client >> 24, (mc_client >> 16) & 0xff, + (mc_client >> 8) & 0xff, mc_client & 0xff, 0 }; + + mc_id = REG_GET_FIELD(status, mmVM_CONTEXT1_PROTECTION_FAULT_STATUS, + xxMEMORY_CLIENT_ID); + + dev_err(adev->dev, "VM fault (0x%02x, vmid %d) at page %u, %s from '%s' (0x%08x) (%d)\n", + protections, vmid, addr, + REG_GET_FIELD(status, mmVM_CONTEXT1_PROTECTION_FAULT_STATUS, + xxMEMORY_CLIENT_RW) ? + "write" : "read", block, mc_client, mc_id); +} + +/* +static const u32 mc_cg_registers[] = { + MC_HUB_MISC_HUB_CG, + MC_HUB_MISC_SIP_CG, + MC_HUB_MISC_VM_CG, + MC_XPB_CLK_GAT, + ATC_MISC_CG, + MC_CITF_MISC_WR_CG, + MC_CITF_MISC_RD_CG, + MC_CITF_MISC_VM_CG, + VM_L2_CG, +}; + +static const u32 mc_cg_ls_en[] = { + MC_HUB_MISC_HUB_CG__MEM_LS_ENABLE_MASK, + MC_HUB_MISC_SIP_CG__MEM_LS_ENABLE_MASK, + MC_HUB_MISC_VM_CG__MEM_LS_ENABLE_MASK, + MC_XPB_CLK_GAT__MEM_LS_ENABLE_MASK, + ATC_MISC_CG__MEM_LS_ENABLE_MASK, + MC_CITF_MISC_WR_CG__MEM_LS_ENABLE_MASK, + MC_CITF_MISC_RD_CG__MEM_LS_ENABLE_MASK, + MC_CITF_MISC_VM_CG__MEM_LS_ENABLE_MASK, + VM_L2_CG__MEM_LS_ENABLE_MASK, +}; + +static const u32 mc_cg_en[] = { + MC_HUB_MISC_HUB_CG__ENABLE_MASK, + MC_HUB_MISC_SIP_CG__ENABLE_MASK, + MC_HUB_MISC_VM_CG__ENABLE_MASK, + MC_XPB_CLK_GAT__ENABLE_MASK, + ATC_MISC_CG__ENABLE_MASK, + MC_CITF_MISC_WR_CG__ENABLE_MASK, + MC_CITF_MISC_RD_CG__ENABLE_MASK, + MC_CITF_MISC_VM_CG__ENABLE_MASK, + VM_L2_CG__ENABLE_MASK, +}; + +static void gmc_v6_0_enable_mc_ls(struct amdgpu_device *adev, + bool enable) +{ + int i; + u32 orig, data; + + for (i = 0; i < ARRAY_SIZE(mc_cg_registers); i++) { + orig = data = RREG32(mc_cg_registers[i]); + if (enable && (adev->cg_flags & AMDGPU_CG_SUPPORT_MC_LS)) + data |= mc_cg_ls_en[i]; + else + data &= ~mc_cg_ls_en[i]; + if (data != orig) + WREG32(mc_cg_registers[i], data); + } +} + +static void gmc_v6_0_enable_mc_mgcg(struct amdgpu_device *adev, + bool enable) +{ + int i; + u32 orig, data; + + for (i = 0; i < ARRAY_SIZE(mc_cg_registers); i++) { + orig = data = RREG32(mc_cg_registers[i]); + if (enable && (adev->cg_flags & AMDGPU_CG_SUPPORT_MC_MGCG)) + data |= mc_cg_en[i]; + else + data &= ~mc_cg_en[i]; + if (data != orig) + WREG32(mc_cg_registers[i], data); + } +} + +static void gmc_v6_0_enable_bif_mgls(struct amdgpu_device *adev, + bool enable) +{ + u32 orig, data; + + orig = data = RREG32_PCIE(ixPCIE_CNTL2); + + if (enable && (adev->cg_flags & AMDGPU_CG_SUPPORT_BIF_LS)) { + data = REG_SET_FIELD(data, PCIE_CNTL2, SLV_MEM_LS_EN, 1); + data = REG_SET_FIELD(data, PCIE_CNTL2, MST_MEM_LS_EN, 1); + data = REG_SET_FIELD(data, PCIE_CNTL2, REPLAY_MEM_LS_EN, 1); + data = REG_SET_FIELD(data, PCIE_CNTL2, SLV_MEM_AGGRESSIVE_LS_EN, 1); + } else { + data = REG_SET_FIELD(data, PCIE_CNTL2, SLV_MEM_LS_EN, 0); + data = REG_SET_FIELD(data, PCIE_CNTL2, MST_MEM_LS_EN, 0); + data = REG_SET_FIELD(data, PCIE_CNTL2, REPLAY_MEM_LS_EN, 0); + data = REG_SET_FIELD(data, PCIE_CNTL2, SLV_MEM_AGGRESSIVE_LS_EN, 0); + } + + if (orig != data) + WREG32_PCIE(ixPCIE_CNTL2, data); +} + +static void gmc_v6_0_enable_hdp_mgcg(struct amdgpu_device *adev, + bool enable) +{ + u32 orig, data; + + orig = data = RREG32(HDP_HOST_PATH_CNTL); + + if (enable && (adev->cg_flags & AMDGPU_CG_SUPPORT_HDP_MGCG)) + data = REG_SET_FIELD(data, HDP_HOST_PATH_CNTL, CLOCK_GATING_DIS, 0); + else + data = REG_SET_FIELD(data, HDP_HOST_PATH_CNTL, CLOCK_GATING_DIS, 1); + + if (orig != data) + WREG32(HDP_HOST_PATH_CNTL, data); +} + +static void gmc_v6_0_enable_hdp_ls(struct amdgpu_device *adev, + bool enable) +{ + u32 orig, data; + + orig = data = RREG32(HDP_MEM_POWER_LS); + + if (enable && (adev->cg_flags & AMDGPU_CG_SUPPORT_HDP_LS)) + data = REG_SET_FIELD(data, HDP_MEM_POWER_LS, LS_ENABLE, 1); + else + data = REG_SET_FIELD(data, HDP_MEM_POWER_LS, LS_ENABLE, 0); + + if (orig != data) + WREG32(HDP_MEM_POWER_LS, data); +} +*/ + +static int gmc_v6_0_convert_vram_type(int mc_seq_vram_type) +{ + switch (mc_seq_vram_type) { + case MC_SEQ_MISC0__MT__GDDR1: + return AMDGPU_VRAM_TYPE_GDDR1; + case MC_SEQ_MISC0__MT__DDR2: + return AMDGPU_VRAM_TYPE_DDR2; + case MC_SEQ_MISC0__MT__GDDR3: + return AMDGPU_VRAM_TYPE_GDDR3; + case MC_SEQ_MISC0__MT__GDDR4: + return AMDGPU_VRAM_TYPE_GDDR4; + case MC_SEQ_MISC0__MT__GDDR5: + return AMDGPU_VRAM_TYPE_GDDR5; + case MC_SEQ_MISC0__MT__DDR3: + return AMDGPU_VRAM_TYPE_DDR3; + default: + return AMDGPU_VRAM_TYPE_UNKNOWN; + } +} + +static int gmc_v6_0_early_init(void *handle) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + gmc_v6_0_set_gart_funcs(adev); + gmc_v6_0_set_irq_funcs(adev); + + if (adev->flags & AMD_IS_APU) { + adev->mc.vram_type = AMDGPU_VRAM_TYPE_UNKNOWN; + } else { + u32 tmp = RREG32(MC_SEQ_MISC0); + tmp &= MC_SEQ_MISC0__MT__MASK; + adev->mc.vram_type = gmc_v6_0_convert_vram_type(tmp); + } + + return 0; +} + +static int gmc_v6_0_late_init(void *handle) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + return amdgpu_irq_get(adev, &adev->mc.vm_fault, 0); +} + +static int gmc_v6_0_sw_init(void *handle) +{ + int r; + int dma_bits; + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + r = amdgpu_irq_add_id(adev, 146, &adev->mc.vm_fault); + if (r) + return r; + + r = amdgpu_irq_add_id(adev, 147, &adev->mc.vm_fault); + if (r) + return r; + + adev->vm_manager.max_pfn = amdgpu_vm_size << 18; + + adev->mc.mc_mask = 0xffffffffffULL; + + adev->need_dma32 = false; + dma_bits = adev->need_dma32 ? 32 : 40; + r = pci_set_dma_mask(adev->pdev, DMA_BIT_MASK(dma_bits)); + if (r) { + adev->need_dma32 = true; + dma_bits = 32; + dev_warn(adev->dev, "amdgpu: No suitable DMA available.\n"); + } + r = pci_set_consistent_dma_mask(adev->pdev, DMA_BIT_MASK(dma_bits)); + if (r) { + pci_set_consistent_dma_mask(adev->pdev, DMA_BIT_MASK(32)); + dev_warn(adev->dev, "amdgpu: No coherent DMA available.\n"); + } + + r = gmc_v6_0_init_microcode(adev); + if (r) { + dev_err(adev->dev, "Failed to load mc firmware!\n"); + return r; + } + + r = amdgpu_ttm_global_init(adev); + if (r) { + return r; + } + + r = gmc_v6_0_mc_init(adev); + if (r) + return r; + + r = amdgpu_bo_init(adev); + if (r) + return r; + + r = gmc_v6_0_gart_init(adev); + if (r) + return r; + + if (!adev->vm_manager.enabled) { + r = gmc_v6_0_vm_init(adev); + if (r) { + dev_err(adev->dev, "vm manager initialization failed (%d).\n", r); + return r; + } + adev->vm_manager.enabled = true; + } + + return r; +} + +static int gmc_v6_0_sw_fini(void *handle) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + if (adev->vm_manager.enabled) { + gmc_v6_0_vm_fini(adev); + adev->vm_manager.enabled = false; + } + gmc_v6_0_gart_fini(adev); + amdgpu_gem_force_release(adev); + amdgpu_bo_fini(adev); + + return 0; +} + +static int gmc_v6_0_hw_init(void *handle) +{ + int r; + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + gmc_v6_0_mc_program(adev); + + if (!(adev->flags & AMD_IS_APU)) { + r = gmc_v6_0_mc_load_microcode(adev); + if (r) { + dev_err(adev->dev, "Failed to load MC firmware!\n"); + return r; + } + } + + r = gmc_v6_0_gart_enable(adev); + if (r) + return r; + + return r; +} + +static int gmc_v6_0_hw_fini(void *handle) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + amdgpu_irq_put(adev, &adev->mc.vm_fault, 0); + gmc_v6_0_gart_disable(adev); + + return 0; +} + +static int gmc_v6_0_suspend(void *handle) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + if (adev->vm_manager.enabled) { + gmc_v6_0_vm_fini(adev); + adev->vm_manager.enabled = false; + } + gmc_v6_0_hw_fini(adev); + + return 0; +} + +static int gmc_v6_0_resume(void *handle) +{ + int r; + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + r = gmc_v6_0_hw_init(adev); + if (r) + return r; + + if (!adev->vm_manager.enabled) { + r = gmc_v6_0_vm_init(adev); + if (r) { + dev_err(adev->dev, "vm manager initialization failed (%d).\n", r); + return r; + } + adev->vm_manager.enabled = true; + } + + return r; +} + +static bool gmc_v6_0_is_idle(void *handle) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + u32 tmp = RREG32(SRBM_STATUS); + + if (tmp & (SRBM_STATUS__MCB_BUSY_MASK | SRBM_STATUS__MCB_NON_DISPLAY_BUSY_MASK | + SRBM_STATUS__MCC_BUSY_MASK | SRBM_STATUS__MCD_BUSY_MASK | SRBM_STATUS__VMC_BUSY_MASK)) + return false; + + return true; +} + +static int gmc_v6_0_wait_for_idle(void *handle) +{ + unsigned i; + u32 tmp; + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + for (i = 0; i < adev->usec_timeout; i++) { + tmp = RREG32(SRBM_STATUS) & (SRBM_STATUS__MCB_BUSY_MASK | + SRBM_STATUS__MCB_NON_DISPLAY_BUSY_MASK | + SRBM_STATUS__MCC_BUSY_MASK | + SRBM_STATUS__MCD_BUSY_MASK | + SRBM_STATUS__VMC_BUSY_MASK); + if (!tmp) + return 0; + udelay(1); + } + return -ETIMEDOUT; + +} + +static int gmc_v6_0_soft_reset(void *handle) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct amdgpu_mode_mc_save save; + u32 srbm_soft_reset = 0; + u32 tmp = RREG32(SRBM_STATUS); + + if (tmp & SRBM_STATUS__VMC_BUSY_MASK) + srbm_soft_reset = REG_SET_FIELD(srbm_soft_reset, + mmSRBM_SOFT_RESET, xxSOFT_RESET_VMC, 1); + + if (tmp & (SRBM_STATUS__MCB_BUSY_MASK | SRBM_STATUS__MCB_NON_DISPLAY_BUSY_MASK | + SRBM_STATUS__MCC_BUSY_MASK | SRBM_STATUS__MCD_BUSY_MASK)) { + if (!(adev->flags & AMD_IS_APU)) + srbm_soft_reset = REG_SET_FIELD(srbm_soft_reset, + mmSRBM_SOFT_RESET, xxSOFT_RESET_MC, 1); + } + + if (srbm_soft_reset) { + gmc_v6_0_mc_stop(adev, &save); + if (gmc_v6_0_wait_for_idle(adev)) { + dev_warn(adev->dev, "Wait for GMC idle timed out !\n"); + } + + + tmp = RREG32(SRBM_SOFT_RESET); + tmp |= srbm_soft_reset; + dev_info(adev->dev, "SRBM_SOFT_RESET=0x%08X\n", tmp); + WREG32(SRBM_SOFT_RESET, tmp); + tmp = RREG32(SRBM_SOFT_RESET); + + udelay(50); + + tmp &= ~srbm_soft_reset; + WREG32(SRBM_SOFT_RESET, tmp); + tmp = RREG32(SRBM_SOFT_RESET); + + udelay(50); + + gmc_v6_0_mc_resume(adev, &save); + udelay(50); + } + + return 0; +} + +static int gmc_v6_0_vm_fault_interrupt_state(struct amdgpu_device *adev, + struct amdgpu_irq_src *src, + unsigned type, + enum amdgpu_interrupt_state state) +{ + u32 tmp; + u32 bits = (VM_CONTEXT1_CNTL__RANGE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK | + VM_CONTEXT1_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK | + VM_CONTEXT1_CNTL__PDE0_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK | + VM_CONTEXT1_CNTL__VALID_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK | + VM_CONTEXT1_CNTL__READ_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK | + VM_CONTEXT1_CNTL__WRITE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK); + + switch (state) { + case AMDGPU_IRQ_STATE_DISABLE: + tmp = RREG32(VM_CONTEXT0_CNTL); + tmp &= ~bits; + WREG32(VM_CONTEXT0_CNTL, tmp); + tmp = RREG32(VM_CONTEXT1_CNTL); + tmp &= ~bits; + WREG32(VM_CONTEXT1_CNTL, tmp); + break; + case AMDGPU_IRQ_STATE_ENABLE: + tmp = RREG32(VM_CONTEXT0_CNTL); + tmp |= bits; + WREG32(VM_CONTEXT0_CNTL, tmp); + tmp = RREG32(VM_CONTEXT1_CNTL); + tmp |= bits; + WREG32(VM_CONTEXT1_CNTL, tmp); + break; + default: + break; + } + + return 0; +} + +static int gmc_v6_0_process_interrupt(struct amdgpu_device *adev, + struct amdgpu_irq_src *source, + struct amdgpu_iv_entry *entry) +{ + u32 addr, status; + + addr = RREG32(VM_CONTEXT1_PROTECTION_FAULT_ADDR); + status = RREG32(VM_CONTEXT1_PROTECTION_FAULT_STATUS); + WREG32_P(VM_CONTEXT1_CNTL2, 1, ~1); + + if (!addr && !status) + return 0; + + if (amdgpu_vm_fault_stop == AMDGPU_VM_FAULT_STOP_FIRST) + gmc_v6_0_set_fault_enable_default(adev, false); + + dev_err(adev->dev, "GPU fault detected: %d 0x%08x\n", + entry->src_id, entry->src_data); + dev_err(adev->dev, " VM_CONTEXT1_PROTECTION_FAULT_ADDR 0x%08X\n", + addr); + dev_err(adev->dev, " VM_CONTEXT1_PROTECTION_FAULT_STATUS 0x%08X\n", + status); + gmc_v6_0_vm_decode_fault(adev, status, addr, 0); + + return 0; +} + +static int gmc_v6_0_set_clockgating_state(void *handle, + enum amd_clockgating_state state) +{ + return 0; +} + +static int gmc_v6_0_set_powergating_state(void *handle, + enum amd_powergating_state state) +{ + return 0; +} + +const struct amd_ip_funcs gmc_v6_0_ip_funcs = { + .name = "gmc_v6_0", + .early_init = gmc_v6_0_early_init, + .late_init = gmc_v6_0_late_init, + .sw_init = gmc_v6_0_sw_init, + .sw_fini = gmc_v6_0_sw_fini, + .hw_init = gmc_v6_0_hw_init, + .hw_fini = gmc_v6_0_hw_fini, + .suspend = gmc_v6_0_suspend, + .resume = gmc_v6_0_resume, + .is_idle = gmc_v6_0_is_idle, + .wait_for_idle = gmc_v6_0_wait_for_idle, + .soft_reset = gmc_v6_0_soft_reset, + .set_clockgating_state = gmc_v6_0_set_clockgating_state, + .set_powergating_state = gmc_v6_0_set_powergating_state, +}; + +static const struct amdgpu_gart_funcs gmc_v6_0_gart_funcs = { + .flush_gpu_tlb = gmc_v6_0_gart_flush_gpu_tlb, + .set_pte_pde = gmc_v6_0_gart_set_pte_pde, +}; + +static const struct amdgpu_irq_src_funcs gmc_v6_0_irq_funcs = { + .set = gmc_v6_0_vm_fault_interrupt_state, + .process = gmc_v6_0_process_interrupt, +}; + +static void gmc_v6_0_set_gart_funcs(struct amdgpu_device *adev) +{ + if (adev->gart.gart_funcs == NULL) + adev->gart.gart_funcs = &gmc_v6_0_gart_funcs; +} + +static void gmc_v6_0_set_irq_funcs(struct amdgpu_device *adev) +{ + adev->mc.vm_fault.num_types = 1; + adev->mc.vm_fault.funcs = &gmc_v6_0_irq_funcs; +} + diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.h b/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.h new file mode 100644 index 000000000000..42c4fc676cd4 --- /dev/null +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.h @@ -0,0 +1,29 @@ +/* + * Copyright 2015 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + */ + +#ifndef __GMC_V6_0_H__ +#define __GMC_V6_0_H__ + +extern const struct amd_ip_funcs gmc_v6_0_ip_funcs; + +#endif diff --git a/drivers/gpu/drm/amd/amdgpu/iceland_smc.c b/drivers/gpu/drm/amd/amdgpu/iceland_smc.c index 211839913728..ef7c27d7356a 100644 --- a/drivers/gpu/drm/amd/amdgpu/iceland_smc.c +++ b/drivers/gpu/drm/amd/amdgpu/iceland_smc.c @@ -121,7 +121,7 @@ out: return result; } -void iceland_start_smc(struct amdgpu_device *adev) +static void iceland_start_smc(struct amdgpu_device *adev) { uint32_t val = RREG32_SMC(ixSMC_SYSCON_RESET_CNTL); @@ -129,7 +129,7 @@ void iceland_start_smc(struct amdgpu_device *adev) WREG32_SMC(ixSMC_SYSCON_RESET_CNTL, val); } -void iceland_reset_smc(struct amdgpu_device *adev) +static void iceland_reset_smc(struct amdgpu_device *adev) { uint32_t val = RREG32_SMC(ixSMC_SYSCON_RESET_CNTL); @@ -145,7 +145,7 @@ static int iceland_program_jump_on_start(struct amdgpu_device *adev) return 0; } -void iceland_stop_smc_clock(struct amdgpu_device *adev) +static void iceland_stop_smc_clock(struct amdgpu_device *adev) { uint32_t val = RREG32_SMC(ixSMC_SYSCON_CLOCK_CNTL_0); @@ -153,7 +153,7 @@ void iceland_stop_smc_clock(struct amdgpu_device *adev) WREG32_SMC(ixSMC_SYSCON_CLOCK_CNTL_0, val); } -void iceland_start_smc_clock(struct amdgpu_device *adev) +static void iceland_start_smc_clock(struct amdgpu_device *adev) { uint32_t val = RREG32_SMC(ixSMC_SYSCON_CLOCK_CNTL_0); diff --git a/drivers/gpu/drm/amd/amdgpu/r600_dpm.h b/drivers/gpu/drm/amd/amdgpu/r600_dpm.h new file mode 100644 index 000000000000..055321f61ca7 --- /dev/null +++ b/drivers/gpu/drm/amd/amdgpu/r600_dpm.h @@ -0,0 +1,127 @@ +/* + * Copyright 2011 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + */ +#ifndef __R600_DPM_H__ +#define __R600_DPM_H__ + +#define R600_ASI_DFLT 10000 +#define R600_BSP_DFLT 0x41EB +#define R600_BSU_DFLT 0x2 +#define R600_AH_DFLT 5 +#define R600_RLP_DFLT 25 +#define R600_RMP_DFLT 65 +#define R600_LHP_DFLT 40 +#define R600_LMP_DFLT 15 +#define R600_TD_DFLT 0 +#define R600_UTC_DFLT_00 0x24 +#define R600_UTC_DFLT_01 0x22 +#define R600_UTC_DFLT_02 0x22 +#define R600_UTC_DFLT_03 0x22 +#define R600_UTC_DFLT_04 0x22 +#define R600_UTC_DFLT_05 0x22 +#define R600_UTC_DFLT_06 0x22 +#define R600_UTC_DFLT_07 0x22 +#define R600_UTC_DFLT_08 0x22 +#define R600_UTC_DFLT_09 0x22 +#define R600_UTC_DFLT_10 0x22 +#define R600_UTC_DFLT_11 0x22 +#define R600_UTC_DFLT_12 0x22 +#define R600_UTC_DFLT_13 0x22 +#define R600_UTC_DFLT_14 0x22 +#define R600_DTC_DFLT_00 0x24 +#define R600_DTC_DFLT_01 0x22 +#define R600_DTC_DFLT_02 0x22 +#define R600_DTC_DFLT_03 0x22 +#define R600_DTC_DFLT_04 0x22 +#define R600_DTC_DFLT_05 0x22 +#define R600_DTC_DFLT_06 0x22 +#define R600_DTC_DFLT_07 0x22 +#define R600_DTC_DFLT_08 0x22 +#define R600_DTC_DFLT_09 0x22 +#define R600_DTC_DFLT_10 0x22 +#define R600_DTC_DFLT_11 0x22 +#define R600_DTC_DFLT_12 0x22 +#define R600_DTC_DFLT_13 0x22 +#define R600_DTC_DFLT_14 0x22 +#define R600_VRC_DFLT 0x0000C003 +#define R600_VOLTAGERESPONSETIME_DFLT 1000 +#define R600_BACKBIASRESPONSETIME_DFLT 1000 +#define R600_VRU_DFLT 0x3 +#define R600_SPLLSTEPTIME_DFLT 0x1000 +#define R600_SPLLSTEPUNIT_DFLT 0x3 +#define R600_TPU_DFLT 0 +#define R600_TPC_DFLT 0x200 +#define R600_SSTU_DFLT 0 +#define R600_SST_DFLT 0x00C8 +#define R600_GICST_DFLT 0x200 +#define R600_FCT_DFLT 0x0400 +#define R600_FCTU_DFLT 0 +#define R600_CTXCGTT3DRPHC_DFLT 0x20 +#define R600_CTXCGTT3DRSDC_DFLT 0x40 +#define R600_VDDC3DOORPHC_DFLT 0x100 +#define R600_VDDC3DOORSDC_DFLT 0x7 +#define R600_VDDC3DOORSU_DFLT 0 +#define R600_MPLLLOCKTIME_DFLT 100 +#define R600_MPLLRESETTIME_DFLT 150 +#define R600_VCOSTEPPCT_DFLT 20 +#define R600_ENDINGVCOSTEPPCT_DFLT 5 +#define R600_REFERENCEDIVIDER_DFLT 4 + +#define R600_PM_NUMBER_OF_TC 15 +#define R600_PM_NUMBER_OF_SCLKS 20 +#define R600_PM_NUMBER_OF_MCLKS 4 +#define R600_PM_NUMBER_OF_VOLTAGE_LEVELS 4 +#define R600_PM_NUMBER_OF_ACTIVITY_LEVELS 3 + +/* XXX are these ok? */ +#define R600_TEMP_RANGE_MIN (90 * 1000) +#define R600_TEMP_RANGE_MAX (120 * 1000) + +#define FDO_PWM_MODE_STATIC 1 +#define FDO_PWM_MODE_STATIC_RPM 5 + +enum r600_power_level { + R600_POWER_LEVEL_LOW = 0, + R600_POWER_LEVEL_MEDIUM = 1, + R600_POWER_LEVEL_HIGH = 2, + R600_POWER_LEVEL_CTXSW = 3, +}; + +enum r600_td { + R600_TD_AUTO, + R600_TD_UP, + R600_TD_DOWN, +}; + +enum r600_display_watermark { + R600_DISPLAY_WATERMARK_LOW = 0, + R600_DISPLAY_WATERMARK_HIGH = 1, +}; + +enum r600_display_gap +{ + R600_PM_DISPLAY_GAP_VBLANK_OR_WM = 0, + R600_PM_DISPLAY_GAP_VBLANK = 1, + R600_PM_DISPLAY_GAP_WATERMARK = 2, + R600_PM_DISPLAY_GAP_IGNORE = 3, +}; +#endif diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v2_4.c b/drivers/gpu/drm/amd/amdgpu/sdma_v2_4.c index 9d02ba27bb87..9ae307505190 100644 --- a/drivers/gpu/drm/amd/amdgpu/sdma_v2_4.c +++ b/drivers/gpu/drm/amd/amdgpu/sdma_v2_4.c @@ -902,6 +902,22 @@ static void sdma_v2_4_ring_emit_vm_flush(struct amdgpu_ring *ring, SDMA_PKT_POLL_REGMEM_DW5_INTERVAL(10)); /* retry count, poll interval */ } +static unsigned sdma_v2_4_ring_get_emit_ib_size(struct amdgpu_ring *ring) +{ + return + 7 + 6; /* sdma_v2_4_ring_emit_ib */ +} + +static unsigned sdma_v2_4_ring_get_dma_frame_size(struct amdgpu_ring *ring) +{ + return + 6 + /* sdma_v2_4_ring_emit_hdp_flush */ + 3 + /* sdma_v2_4_ring_emit_hdp_invalidate */ + 6 + /* sdma_v2_4_ring_emit_pipeline_sync */ + 12 + /* sdma_v2_4_ring_emit_vm_flush */ + 10 + 10 + 10; /* sdma_v2_4_ring_emit_fence x3 for user fence, vm fence */ +} + static int sdma_v2_4_early_init(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; @@ -1220,6 +1236,8 @@ static const struct amdgpu_ring_funcs sdma_v2_4_ring_funcs = { .test_ib = sdma_v2_4_ring_test_ib, .insert_nop = sdma_v2_4_ring_insert_nop, .pad_ib = sdma_v2_4_ring_pad_ib, + .get_emit_ib_size = sdma_v2_4_ring_get_emit_ib_size, + .get_dma_frame_size = sdma_v2_4_ring_get_dma_frame_size, }; static void sdma_v2_4_set_ring_funcs(struct amdgpu_device *adev) diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c b/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c index ddb8081f78ac..f325fd86430b 100644 --- a/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c +++ b/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c @@ -495,31 +495,6 @@ static void sdma_v3_0_ring_emit_fence(struct amdgpu_ring *ring, u64 addr, u64 se amdgpu_ring_write(ring, SDMA_PKT_TRAP_INT_CONTEXT_INT_CONTEXT(0)); } -unsigned init_cond_exec(struct amdgpu_ring *ring) -{ - unsigned ret; - amdgpu_ring_write(ring, SDMA_PKT_HEADER_OP(SDMA_OP_COND_EXE)); - amdgpu_ring_write(ring, lower_32_bits(ring->cond_exe_gpu_addr)); - amdgpu_ring_write(ring, upper_32_bits(ring->cond_exe_gpu_addr)); - amdgpu_ring_write(ring, 1); - ret = ring->wptr;/* this is the offset we need patch later */ - amdgpu_ring_write(ring, 0x55aa55aa);/* insert dummy here and patch it later */ - return ret; -} - -void patch_cond_exec(struct amdgpu_ring *ring, unsigned offset) -{ - unsigned cur; - BUG_ON(ring->ring[offset] != 0x55aa55aa); - - cur = ring->wptr - 1; - if (likely(cur > offset)) - ring->ring[offset] = cur - offset; - else - ring->ring[offset] = (ring->ring_size>>2) - offset + cur; -} - - /** * sdma_v3_0_gfx_stop - stop the gfx async dma engines * @@ -1129,6 +1104,22 @@ static void sdma_v3_0_ring_emit_vm_flush(struct amdgpu_ring *ring, SDMA_PKT_POLL_REGMEM_DW5_INTERVAL(10)); /* retry count, poll interval */ } +static unsigned sdma_v3_0_ring_get_emit_ib_size(struct amdgpu_ring *ring) +{ + return + 7 + 6; /* sdma_v3_0_ring_emit_ib */ +} + +static unsigned sdma_v3_0_ring_get_dma_frame_size(struct amdgpu_ring *ring) +{ + return + 6 + /* sdma_v3_0_ring_emit_hdp_flush */ + 3 + /* sdma_v3_0_ring_emit_hdp_invalidate */ + 6 + /* sdma_v3_0_ring_emit_pipeline_sync */ + 12 + /* sdma_v3_0_ring_emit_vm_flush */ + 10 + 10 + 10; /* sdma_v3_0_ring_emit_fence x3 for user fence, vm fence */ +} + static int sdma_v3_0_early_init(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; @@ -1590,6 +1581,8 @@ static const struct amdgpu_ring_funcs sdma_v3_0_ring_funcs = { .test_ib = sdma_v3_0_ring_test_ib, .insert_nop = sdma_v3_0_ring_insert_nop, .pad_ib = sdma_v3_0_ring_pad_ib, + .get_emit_ib_size = sdma_v3_0_ring_get_emit_ib_size, + .get_dma_frame_size = sdma_v3_0_ring_get_dma_frame_size, }; static void sdma_v3_0_set_ring_funcs(struct amdgpu_device *adev) diff --git a/drivers/gpu/drm/amd/amdgpu/si.c b/drivers/gpu/drm/amd/amdgpu/si.c new file mode 100644 index 000000000000..fee76b8a536f --- /dev/null +++ b/drivers/gpu/drm/amd/amdgpu/si.c @@ -0,0 +1,1965 @@ +/* + * Copyright 2015 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + */ + +#include <linux/firmware.h> +#include <linux/slab.h> +#include <linux/module.h> +#include "drmP.h" +#include "amdgpu.h" +#include "amdgpu_atombios.h" +#include "amdgpu_ih.h" +#include "amdgpu_uvd.h" +#include "amdgpu_vce.h" +#include "atom.h" +#include "amdgpu_powerplay.h" +#include "si/sid.h" +#include "si_ih.h" +#include "gfx_v6_0.h" +#include "gmc_v6_0.h" +#include "si_dma.h" +#include "dce_v6_0.h" +#include "si.h" + +static const u32 tahiti_golden_registers[] = +{ + 0x2684, 0x00010000, 0x00018208, + 0x260c, 0xffffffff, 0x00000000, + 0x260d, 0xf00fffff, 0x00000400, + 0x260e, 0x0002021c, 0x00020200, + 0x031e, 0x00000080, 0x00000000, + 0x340c, 0x000300c0, 0x00800040, + 0x360c, 0x000300c0, 0x00800040, + 0x16ec, 0x000000f0, 0x00000070, + 0x16f0, 0x00200000, 0x50100000, + 0x1c0c, 0x31000311, 0x00000011, + 0x09df, 0x00000003, 0x000007ff, + 0x0903, 0x000007ff, 0x00000000, + 0x2285, 0xf000001f, 0x00000007, + 0x22c9, 0xffffffff, 0x00ffffff, + 0x22c4, 0x0000ff0f, 0x00000000, + 0xa293, 0x07ffffff, 0x4e000000, + 0xa0d4, 0x3f3f3fff, 0x2a00126a, + 0x000c, 0x000000ff, 0x0040, + 0x000d, 0x00000040, 0x00004040, + 0x2440, 0x07ffffff, 0x03000000, + 0x23a2, 0x01ff1f3f, 0x00000000, + 0x23a1, 0x01ff1f3f, 0x00000000, + 0x2418, 0x0000007f, 0x00000020, + 0x2542, 0x00010000, 0x00010000, + 0x2b05, 0x00000200, 0x000002fb, + 0x2b04, 0xffffffff, 0x0000543b, + 0x2b03, 0xffffffff, 0xa9210876, + 0x2234, 0xffffffff, 0x000fff40, + 0x2235, 0x0000001f, 0x00000010, + 0x0504, 0x20000000, 0x20fffed8, + 0x0570, 0x000c0fc0, 0x000c0400 +}; + +static const u32 tahiti_golden_registers2[] = +{ + 0x0319, 0x00000001, 0x00000001 +}; + +static const u32 tahiti_golden_rlc_registers[] = +{ + 0x3109, 0xffffffff, 0x00601005, + 0x311f, 0xffffffff, 0x10104040, + 0x3122, 0xffffffff, 0x0100000a, + 0x30c5, 0xffffffff, 0x00000800, + 0x30c3, 0xffffffff, 0x800000f4, + 0x3d2a, 0xffffffff, 0x00000000 +}; + +static const u32 pitcairn_golden_registers[] = +{ + 0x2684, 0x00010000, 0x00018208, + 0x260c, 0xffffffff, 0x00000000, + 0x260d, 0xf00fffff, 0x00000400, + 0x260e, 0x0002021c, 0x00020200, + 0x031e, 0x00000080, 0x00000000, + 0x340c, 0x000300c0, 0x00800040, + 0x360c, 0x000300c0, 0x00800040, + 0x16ec, 0x000000f0, 0x00000070, + 0x16f0, 0x00200000, 0x50100000, + 0x1c0c, 0x31000311, 0x00000011, + 0x0ab9, 0x00073ffe, 0x000022a2, + 0x0903, 0x000007ff, 0x00000000, + 0x2285, 0xf000001f, 0x00000007, + 0x22c9, 0xffffffff, 0x00ffffff, + 0x22c4, 0x0000ff0f, 0x00000000, + 0xa293, 0x07ffffff, 0x4e000000, + 0xa0d4, 0x3f3f3fff, 0x2a00126a, + 0x000c, 0x000000ff, 0x0040, + 0x000d, 0x00000040, 0x00004040, + 0x2440, 0x07ffffff, 0x03000000, + 0x2418, 0x0000007f, 0x00000020, + 0x2542, 0x00010000, 0x00010000, + 0x2b05, 0x000003ff, 0x000000f7, + 0x2b04, 0xffffffff, 0x00000000, + 0x2b03, 0xffffffff, 0x32761054, + 0x2235, 0x0000001f, 0x00000010, + 0x0570, 0x000c0fc0, 0x000c0400 +}; + +static const u32 pitcairn_golden_rlc_registers[] = +{ + 0x3109, 0xffffffff, 0x00601004, + 0x311f, 0xffffffff, 0x10102020, + 0x3122, 0xffffffff, 0x01000020, + 0x30c5, 0xffffffff, 0x00000800, + 0x30c3, 0xffffffff, 0x800000a4 +}; + +static const u32 verde_pg_init[] = +{ + 0xd4f, 0xffffffff, 0x40000, + 0xd4e, 0xffffffff, 0x200010ff, + 0xd4f, 0xffffffff, 0x0, + 0xd4f, 0xffffffff, 0x0, + 0xd4f, 0xffffffff, 0x0, + 0xd4f, 0xffffffff, 0x0, + 0xd4f, 0xffffffff, 0x0, + 0xd4f, 0xffffffff, 0x7007, + 0xd4e, 0xffffffff, 0x300010ff, + 0xd4f, 0xffffffff, 0x0, + 0xd4f, 0xffffffff, 0x0, + 0xd4f, 0xffffffff, 0x0, + 0xd4f, 0xffffffff, 0x0, + 0xd4f, 0xffffffff, 0x0, + 0xd4f, 0xffffffff, 0x400000, + 0xd4e, 0xffffffff, 0x100010ff, + 0xd4f, 0xffffffff, 0x0, + 0xd4f, 0xffffffff, 0x0, + 0xd4f, 0xffffffff, 0x0, + 0xd4f, 0xffffffff, 0x0, + 0xd4f, 0xffffffff, 0x0, + 0xd4f, 0xffffffff, 0x120200, + 0xd4e, 0xffffffff, 0x500010ff, + 0xd4f, 0xffffffff, 0x0, + 0xd4f, 0xffffffff, 0x0, + 0xd4f, 0xffffffff, 0x0, + 0xd4f, 0xffffffff, 0x0, + 0xd4f, 0xffffffff, 0x0, + 0xd4f, 0xffffffff, 0x1e1e16, + 0xd4e, 0xffffffff, 0x600010ff, + 0xd4f, 0xffffffff, 0x0, + 0xd4f, 0xffffffff, 0x0, + 0xd4f, 0xffffffff, 0x0, + 0xd4f, 0xffffffff, 0x0, + 0xd4f, 0xffffffff, 0x0, + 0xd4f, 0xffffffff, 0x171f1e, + 0xd4e, 0xffffffff, 0x700010ff, + 0xd4f, 0xffffffff, 0x0, + 0xd4f, 0xffffffff, 0x0, + 0xd4f, 0xffffffff, 0x0, + 0xd4f, 0xffffffff, 0x0, + 0xd4f, 0xffffffff, 0x0, + 0xd4f, 0xffffffff, 0x0, + 0xd4e, 0xffffffff, 0x9ff, + 0xd40, 0xffffffff, 0x0, + 0xd41, 0xffffffff, 0x10000800, + 0xd41, 0xffffffff, 0xf, + 0xd41, 0xffffffff, 0xf, + 0xd40, 0xffffffff, 0x4, + 0xd41, 0xffffffff, 0x1000051e, + 0xd41, 0xffffffff, 0xffff, + 0xd41, 0xffffffff, 0xffff, + 0xd40, 0xffffffff, 0x8, + 0xd41, 0xffffffff, 0x80500, + 0xd40, 0xffffffff, 0x12, + 0xd41, 0xffffffff, 0x9050c, + 0xd40, 0xffffffff, 0x1d, + 0xd41, 0xffffffff, 0xb052c, + 0xd40, 0xffffffff, 0x2a, + 0xd41, 0xffffffff, 0x1053e, + 0xd40, 0xffffffff, 0x2d, + 0xd41, 0xffffffff, 0x10546, + 0xd40, 0xffffffff, 0x30, + 0xd41, 0xffffffff, 0xa054e, + 0xd40, 0xffffffff, 0x3c, + 0xd41, 0xffffffff, 0x1055f, + 0xd40, 0xffffffff, 0x3f, + 0xd41, 0xffffffff, 0x10567, + 0xd40, 0xffffffff, 0x42, + 0xd41, 0xffffffff, 0x1056f, + 0xd40, 0xffffffff, 0x45, + 0xd41, 0xffffffff, 0x10572, + 0xd40, 0xffffffff, 0x48, + 0xd41, 0xffffffff, 0x20575, + 0xd40, 0xffffffff, 0x4c, + 0xd41, 0xffffffff, 0x190801, + 0xd40, 0xffffffff, 0x67, + 0xd41, 0xffffffff, 0x1082a, + 0xd40, 0xffffffff, 0x6a, + 0xd41, 0xffffffff, 0x1b082d, + 0xd40, 0xffffffff, 0x87, + 0xd41, 0xffffffff, 0x310851, + 0xd40, 0xffffffff, 0xba, + 0xd41, 0xffffffff, 0x891, + 0xd40, 0xffffffff, 0xbc, + 0xd41, 0xffffffff, 0x893, + 0xd40, 0xffffffff, 0xbe, + 0xd41, 0xffffffff, 0x20895, + 0xd40, 0xffffffff, 0xc2, + 0xd41, 0xffffffff, 0x20899, + 0xd40, 0xffffffff, 0xc6, + 0xd41, 0xffffffff, 0x2089d, + 0xd40, 0xffffffff, 0xca, + 0xd41, 0xffffffff, 0x8a1, + 0xd40, 0xffffffff, 0xcc, + 0xd41, 0xffffffff, 0x8a3, + 0xd40, 0xffffffff, 0xce, + 0xd41, 0xffffffff, 0x308a5, + 0xd40, 0xffffffff, 0xd3, + 0xd41, 0xffffffff, 0x6d08cd, + 0xd40, 0xffffffff, 0x142, + 0xd41, 0xffffffff, 0x2000095a, + 0xd41, 0xffffffff, 0x1, + 0xd40, 0xffffffff, 0x144, + 0xd41, 0xffffffff, 0x301f095b, + 0xd40, 0xffffffff, 0x165, + 0xd41, 0xffffffff, 0xc094d, + 0xd40, 0xffffffff, 0x173, + 0xd41, 0xffffffff, 0xf096d, + 0xd40, 0xffffffff, 0x184, + 0xd41, 0xffffffff, 0x15097f, + 0xd40, 0xffffffff, 0x19b, + 0xd41, 0xffffffff, 0xc0998, + 0xd40, 0xffffffff, 0x1a9, + 0xd41, 0xffffffff, 0x409a7, + 0xd40, 0xffffffff, 0x1af, + 0xd41, 0xffffffff, 0xcdc, + 0xd40, 0xffffffff, 0x1b1, + 0xd41, 0xffffffff, 0x800, + 0xd42, 0xffffffff, 0x6c9b2000, + 0xd44, 0xfc00, 0x2000, + 0xd51, 0xffffffff, 0xfc0, + 0xa35, 0x00000100, 0x100 +}; + +static const u32 verde_golden_rlc_registers[] = +{ + 0x3109, 0xffffffff, 0x033f1005, + 0x311f, 0xffffffff, 0x10808020, + 0x3122, 0xffffffff, 0x00800008, + 0x30c5, 0xffffffff, 0x00001000, + 0x30c3, 0xffffffff, 0x80010014 +}; + +static const u32 verde_golden_registers[] = +{ + 0x2684, 0x00010000, 0x00018208, + 0x260c, 0xffffffff, 0x00000000, + 0x260d, 0xf00fffff, 0x00000400, + 0x260e, 0x0002021c, 0x00020200, + 0x031e, 0x00000080, 0x00000000, + 0x340c, 0x000300c0, 0x00800040, + 0x340c, 0x000300c0, 0x00800040, + 0x360c, 0x000300c0, 0x00800040, + 0x360c, 0x000300c0, 0x00800040, + 0x16ec, 0x000000f0, 0x00000070, + 0x16f0, 0x00200000, 0x50100000, + + 0x1c0c, 0x31000311, 0x00000011, + 0x0ab9, 0x00073ffe, 0x000022a2, + 0x0ab9, 0x00073ffe, 0x000022a2, + 0x0ab9, 0x00073ffe, 0x000022a2, + 0x0903, 0x000007ff, 0x00000000, + 0x0903, 0x000007ff, 0x00000000, + 0x0903, 0x000007ff, 0x00000000, + 0x2285, 0xf000001f, 0x00000007, + 0x2285, 0xf000001f, 0x00000007, + 0x2285, 0xf000001f, 0x00000007, + 0x2285, 0xffffffff, 0x00ffffff, + 0x22c4, 0x0000ff0f, 0x00000000, + + 0xa293, 0x07ffffff, 0x4e000000, + 0xa0d4, 0x3f3f3fff, 0x0000124a, + 0xa0d4, 0x3f3f3fff, 0x0000124a, + 0xa0d4, 0x3f3f3fff, 0x0000124a, + 0x000c, 0x000000ff, 0x0040, + 0x000d, 0x00000040, 0x00004040, + 0x2440, 0x07ffffff, 0x03000000, + 0x2440, 0x07ffffff, 0x03000000, + 0x23a2, 0x01ff1f3f, 0x00000000, + 0x23a3, 0x01ff1f3f, 0x00000000, + 0x23a2, 0x01ff1f3f, 0x00000000, + 0x23a1, 0x01ff1f3f, 0x00000000, + 0x23a1, 0x01ff1f3f, 0x00000000, + + 0x23a1, 0x01ff1f3f, 0x00000000, + 0x2418, 0x0000007f, 0x00000020, + 0x2542, 0x00010000, 0x00010000, + 0x2b01, 0x000003ff, 0x00000003, + 0x2b05, 0x000003ff, 0x00000003, + 0x2b05, 0x000003ff, 0x00000003, + 0x2b04, 0xffffffff, 0x00000000, + 0x2b04, 0xffffffff, 0x00000000, + 0x2b04, 0xffffffff, 0x00000000, + 0x2b03, 0xffffffff, 0x00001032, + 0x2b03, 0xffffffff, 0x00001032, + 0x2b03, 0xffffffff, 0x00001032, + 0x2235, 0x0000001f, 0x00000010, + 0x2235, 0x0000001f, 0x00000010, + 0x2235, 0x0000001f, 0x00000010, + 0x0570, 0x000c0fc0, 0x000c0400 +}; + +static const u32 oland_golden_registers[] = +{ + 0x2684, 0x00010000, 0x00018208, + 0x260c, 0xffffffff, 0x00000000, + 0x260d, 0xf00fffff, 0x00000400, + 0x260e, 0x0002021c, 0x00020200, + 0x031e, 0x00000080, 0x00000000, + 0x340c, 0x000300c0, 0x00800040, + 0x360c, 0x000300c0, 0x00800040, + 0x16ec, 0x000000f0, 0x00000070, + 0x16f9, 0x00200000, 0x50100000, + 0x1c0c, 0x31000311, 0x00000011, + 0x0ab9, 0x00073ffe, 0x000022a2, + 0x0903, 0x000007ff, 0x00000000, + 0x2285, 0xf000001f, 0x00000007, + 0x22c9, 0xffffffff, 0x00ffffff, + 0x22c4, 0x0000ff0f, 0x00000000, + 0xa293, 0x07ffffff, 0x4e000000, + 0xa0d4, 0x3f3f3fff, 0x00000082, + 0x000c, 0x000000ff, 0x0040, + 0x000d, 0x00000040, 0x00004040, + 0x2440, 0x07ffffff, 0x03000000, + 0x2418, 0x0000007f, 0x00000020, + 0x2542, 0x00010000, 0x00010000, + 0x2b05, 0x000003ff, 0x000000f3, + 0x2b04, 0xffffffff, 0x00000000, + 0x2b03, 0xffffffff, 0x00003210, + 0x2235, 0x0000001f, 0x00000010, + 0x0570, 0x000c0fc0, 0x000c0400 +}; + +static const u32 oland_golden_rlc_registers[] = +{ + 0x3109, 0xffffffff, 0x00601005, + 0x311f, 0xffffffff, 0x10104040, + 0x3122, 0xffffffff, 0x0100000a, + 0x30c5, 0xffffffff, 0x00000800, + 0x30c3, 0xffffffff, 0x800000f4 +}; + +static const u32 hainan_golden_registers[] = +{ + 0x2684, 0x00010000, 0x00018208, + 0x260c, 0xffffffff, 0x00000000, + 0x260d, 0xf00fffff, 0x00000400, + 0x260e, 0x0002021c, 0x00020200, + 0x4595, 0xff000fff, 0x00000100, + 0x340c, 0x000300c0, 0x00800040, + 0x3630, 0xff000fff, 0x00000100, + 0x360c, 0x000300c0, 0x00800040, + 0x0ab9, 0x00073ffe, 0x000022a2, + 0x0903, 0x000007ff, 0x00000000, + 0x2285, 0xf000001f, 0x00000007, + 0x22c9, 0xffffffff, 0x00ffffff, + 0x22c4, 0x0000ff0f, 0x00000000, + 0xa393, 0x07ffffff, 0x4e000000, + 0xa0d4, 0x3f3f3fff, 0x00000000, + 0x000c, 0x000000ff, 0x0040, + 0x000d, 0x00000040, 0x00004040, + 0x2440, 0x03e00000, 0x03600000, + 0x2418, 0x0000007f, 0x00000020, + 0x2542, 0x00010000, 0x00010000, + 0x2b05, 0x000003ff, 0x000000f1, + 0x2b04, 0xffffffff, 0x00000000, + 0x2b03, 0xffffffff, 0x00003210, + 0x2235, 0x0000001f, 0x00000010, + 0x0570, 0x000c0fc0, 0x000c0400 +}; + +static const u32 hainan_golden_registers2[] = +{ + 0x263e, 0xffffffff, 0x02010001 +}; + +static const u32 tahiti_mgcg_cgcg_init[] = +{ + 0x3100, 0xffffffff, 0xfffffffc, + 0x200b, 0xffffffff, 0xe0000000, + 0x2698, 0xffffffff, 0x00000100, + 0x24a9, 0xffffffff, 0x00000100, + 0x3059, 0xffffffff, 0x00000100, + 0x25dd, 0xffffffff, 0x00000100, + 0x2261, 0xffffffff, 0x06000100, + 0x2286, 0xffffffff, 0x00000100, + 0x24a8, 0xffffffff, 0x00000100, + 0x30e0, 0xffffffff, 0x00000100, + 0x22ca, 0xffffffff, 0x00000100, + 0x2451, 0xffffffff, 0x00000100, + 0x2362, 0xffffffff, 0x00000100, + 0x2363, 0xffffffff, 0x00000100, + 0x240c, 0xffffffff, 0x00000100, + 0x240d, 0xffffffff, 0x00000100, + 0x240e, 0xffffffff, 0x00000100, + 0x240f, 0xffffffff, 0x00000100, + 0x2b60, 0xffffffff, 0x00000100, + 0x2b15, 0xffffffff, 0x00000100, + 0x225f, 0xffffffff, 0x06000100, + 0x261a, 0xffffffff, 0x00000100, + 0x2544, 0xffffffff, 0x00000100, + 0x2bc1, 0xffffffff, 0x00000100, + 0x2b81, 0xffffffff, 0x00000100, + 0x2527, 0xffffffff, 0x00000100, + 0x200b, 0xffffffff, 0xe0000000, + 0x2458, 0xffffffff, 0x00010000, + 0x2459, 0xffffffff, 0x00030002, + 0x245a, 0xffffffff, 0x00040007, + 0x245b, 0xffffffff, 0x00060005, + 0x245c, 0xffffffff, 0x00090008, + 0x245d, 0xffffffff, 0x00020001, + 0x245e, 0xffffffff, 0x00040003, + 0x245f, 0xffffffff, 0x00000007, + 0x2460, 0xffffffff, 0x00060005, + 0x2461, 0xffffffff, 0x00090008, + 0x2462, 0xffffffff, 0x00030002, + 0x2463, 0xffffffff, 0x00050004, + 0x2464, 0xffffffff, 0x00000008, + 0x2465, 0xffffffff, 0x00070006, + 0x2466, 0xffffffff, 0x000a0009, + 0x2467, 0xffffffff, 0x00040003, + 0x2468, 0xffffffff, 0x00060005, + 0x2469, 0xffffffff, 0x00000009, + 0x246a, 0xffffffff, 0x00080007, + 0x246b, 0xffffffff, 0x000b000a, + 0x246c, 0xffffffff, 0x00050004, + 0x246d, 0xffffffff, 0x00070006, + 0x246e, 0xffffffff, 0x0008000b, + 0x246f, 0xffffffff, 0x000a0009, + 0x2470, 0xffffffff, 0x000d000c, + 0x2471, 0xffffffff, 0x00060005, + 0x2472, 0xffffffff, 0x00080007, + 0x2473, 0xffffffff, 0x0000000b, + 0x2474, 0xffffffff, 0x000a0009, + 0x2475, 0xffffffff, 0x000d000c, + 0x2476, 0xffffffff, 0x00070006, + 0x2477, 0xffffffff, 0x00090008, + 0x2478, 0xffffffff, 0x0000000c, + 0x2479, 0xffffffff, 0x000b000a, + 0x247a, 0xffffffff, 0x000e000d, + 0x247b, 0xffffffff, 0x00080007, + 0x247c, 0xffffffff, 0x000a0009, + 0x247d, 0xffffffff, 0x0000000d, + 0x247e, 0xffffffff, 0x000c000b, + 0x247f, 0xffffffff, 0x000f000e, + 0x2480, 0xffffffff, 0x00090008, + 0x2481, 0xffffffff, 0x000b000a, + 0x2482, 0xffffffff, 0x000c000f, + 0x2483, 0xffffffff, 0x000e000d, + 0x2484, 0xffffffff, 0x00110010, + 0x2485, 0xffffffff, 0x000a0009, + 0x2486, 0xffffffff, 0x000c000b, + 0x2487, 0xffffffff, 0x0000000f, + 0x2488, 0xffffffff, 0x000e000d, + 0x2489, 0xffffffff, 0x00110010, + 0x248a, 0xffffffff, 0x000b000a, + 0x248b, 0xffffffff, 0x000d000c, + 0x248c, 0xffffffff, 0x00000010, + 0x248d, 0xffffffff, 0x000f000e, + 0x248e, 0xffffffff, 0x00120011, + 0x248f, 0xffffffff, 0x000c000b, + 0x2490, 0xffffffff, 0x000e000d, + 0x2491, 0xffffffff, 0x00000011, + 0x2492, 0xffffffff, 0x0010000f, + 0x2493, 0xffffffff, 0x00130012, + 0x2494, 0xffffffff, 0x000d000c, + 0x2495, 0xffffffff, 0x000f000e, + 0x2496, 0xffffffff, 0x00100013, + 0x2497, 0xffffffff, 0x00120011, + 0x2498, 0xffffffff, 0x00150014, + 0x2499, 0xffffffff, 0x000e000d, + 0x249a, 0xffffffff, 0x0010000f, + 0x249b, 0xffffffff, 0x00000013, + 0x249c, 0xffffffff, 0x00120011, + 0x249d, 0xffffffff, 0x00150014, + 0x249e, 0xffffffff, 0x000f000e, + 0x249f, 0xffffffff, 0x00110010, + 0x24a0, 0xffffffff, 0x00000014, + 0x24a1, 0xffffffff, 0x00130012, + 0x24a2, 0xffffffff, 0x00160015, + 0x24a3, 0xffffffff, 0x0010000f, + 0x24a4, 0xffffffff, 0x00120011, + 0x24a5, 0xffffffff, 0x00000015, + 0x24a6, 0xffffffff, 0x00140013, + 0x24a7, 0xffffffff, 0x00170016, + 0x2454, 0xffffffff, 0x96940200, + 0x21c2, 0xffffffff, 0x00900100, + 0x311e, 0xffffffff, 0x00000080, + 0x3101, 0xffffffff, 0x0020003f, + 0xc, 0xffffffff, 0x0000001c, + 0xd, 0x000f0000, 0x000f0000, + 0x583, 0xffffffff, 0x00000100, + 0x409, 0xffffffff, 0x00000100, + 0x40b, 0x00000101, 0x00000000, + 0x82a, 0xffffffff, 0x00000104, + 0x993, 0x000c0000, 0x000c0000, + 0x992, 0x000c0000, 0x000c0000, + 0x1579, 0xff000fff, 0x00000100, + 0x157a, 0x00000001, 0x00000001, + 0xbd4, 0x00000001, 0x00000001, + 0xc33, 0xc0000fff, 0x00000104, + 0x3079, 0x00000001, 0x00000001, + 0x3430, 0xfffffff0, 0x00000100, + 0x3630, 0xfffffff0, 0x00000100 +}; +static const u32 pitcairn_mgcg_cgcg_init[] = +{ + 0x3100, 0xffffffff, 0xfffffffc, + 0x200b, 0xffffffff, 0xe0000000, + 0x2698, 0xffffffff, 0x00000100, + 0x24a9, 0xffffffff, 0x00000100, + 0x3059, 0xffffffff, 0x00000100, + 0x25dd, 0xffffffff, 0x00000100, + 0x2261, 0xffffffff, 0x06000100, + 0x2286, 0xffffffff, 0x00000100, + 0x24a8, 0xffffffff, 0x00000100, + 0x30e0, 0xffffffff, 0x00000100, + 0x22ca, 0xffffffff, 0x00000100, + 0x2451, 0xffffffff, 0x00000100, + 0x2362, 0xffffffff, 0x00000100, + 0x2363, 0xffffffff, 0x00000100, + 0x240c, 0xffffffff, 0x00000100, + 0x240d, 0xffffffff, 0x00000100, + 0x240e, 0xffffffff, 0x00000100, + 0x240f, 0xffffffff, 0x00000100, + 0x2b60, 0xffffffff, 0x00000100, + 0x2b15, 0xffffffff, 0x00000100, + 0x225f, 0xffffffff, 0x06000100, + 0x261a, 0xffffffff, 0x00000100, + 0x2544, 0xffffffff, 0x00000100, + 0x2bc1, 0xffffffff, 0x00000100, + 0x2b81, 0xffffffff, 0x00000100, + 0x2527, 0xffffffff, 0x00000100, + 0x200b, 0xffffffff, 0xe0000000, + 0x2458, 0xffffffff, 0x00010000, + 0x2459, 0xffffffff, 0x00030002, + 0x245a, 0xffffffff, 0x00040007, + 0x245b, 0xffffffff, 0x00060005, + 0x245c, 0xffffffff, 0x00090008, + 0x245d, 0xffffffff, 0x00020001, + 0x245e, 0xffffffff, 0x00040003, + 0x245f, 0xffffffff, 0x00000007, + 0x2460, 0xffffffff, 0x00060005, + 0x2461, 0xffffffff, 0x00090008, + 0x2462, 0xffffffff, 0x00030002, + 0x2463, 0xffffffff, 0x00050004, + 0x2464, 0xffffffff, 0x00000008, + 0x2465, 0xffffffff, 0x00070006, + 0x2466, 0xffffffff, 0x000a0009, + 0x2467, 0xffffffff, 0x00040003, + 0x2468, 0xffffffff, 0x00060005, + 0x2469, 0xffffffff, 0x00000009, + 0x246a, 0xffffffff, 0x00080007, + 0x246b, 0xffffffff, 0x000b000a, + 0x246c, 0xffffffff, 0x00050004, + 0x246d, 0xffffffff, 0x00070006, + 0x246e, 0xffffffff, 0x0008000b, + 0x246f, 0xffffffff, 0x000a0009, + 0x2470, 0xffffffff, 0x000d000c, + 0x2480, 0xffffffff, 0x00090008, + 0x2481, 0xffffffff, 0x000b000a, + 0x2482, 0xffffffff, 0x000c000f, + 0x2483, 0xffffffff, 0x000e000d, + 0x2484, 0xffffffff, 0x00110010, + 0x2485, 0xffffffff, 0x000a0009, + 0x2486, 0xffffffff, 0x000c000b, + 0x2487, 0xffffffff, 0x0000000f, + 0x2488, 0xffffffff, 0x000e000d, + 0x2489, 0xffffffff, 0x00110010, + 0x248a, 0xffffffff, 0x000b000a, + 0x248b, 0xffffffff, 0x000d000c, + 0x248c, 0xffffffff, 0x00000010, + 0x248d, 0xffffffff, 0x000f000e, + 0x248e, 0xffffffff, 0x00120011, + 0x248f, 0xffffffff, 0x000c000b, + 0x2490, 0xffffffff, 0x000e000d, + 0x2491, 0xffffffff, 0x00000011, + 0x2492, 0xffffffff, 0x0010000f, + 0x2493, 0xffffffff, 0x00130012, + 0x2494, 0xffffffff, 0x000d000c, + 0x2495, 0xffffffff, 0x000f000e, + 0x2496, 0xffffffff, 0x00100013, + 0x2497, 0xffffffff, 0x00120011, + 0x2498, 0xffffffff, 0x00150014, + 0x2454, 0xffffffff, 0x96940200, + 0x21c2, 0xffffffff, 0x00900100, + 0x311e, 0xffffffff, 0x00000080, + 0x3101, 0xffffffff, 0x0020003f, + 0xc, 0xffffffff, 0x0000001c, + 0xd, 0x000f0000, 0x000f0000, + 0x583, 0xffffffff, 0x00000100, + 0x409, 0xffffffff, 0x00000100, + 0x40b, 0x00000101, 0x00000000, + 0x82a, 0xffffffff, 0x00000104, + 0x1579, 0xff000fff, 0x00000100, + 0x157a, 0x00000001, 0x00000001, + 0xbd4, 0x00000001, 0x00000001, + 0xc33, 0xc0000fff, 0x00000104, + 0x3079, 0x00000001, 0x00000001, + 0x3430, 0xfffffff0, 0x00000100, + 0x3630, 0xfffffff0, 0x00000100 +}; +static const u32 verde_mgcg_cgcg_init[] = +{ + 0x3100, 0xffffffff, 0xfffffffc, + 0x200b, 0xffffffff, 0xe0000000, + 0x2698, 0xffffffff, 0x00000100, + 0x24a9, 0xffffffff, 0x00000100, + 0x3059, 0xffffffff, 0x00000100, + 0x25dd, 0xffffffff, 0x00000100, + 0x2261, 0xffffffff, 0x06000100, + 0x2286, 0xffffffff, 0x00000100, + 0x24a8, 0xffffffff, 0x00000100, + 0x30e0, 0xffffffff, 0x00000100, + 0x22ca, 0xffffffff, 0x00000100, + 0x2451, 0xffffffff, 0x00000100, + 0x2362, 0xffffffff, 0x00000100, + 0x2363, 0xffffffff, 0x00000100, + 0x240c, 0xffffffff, 0x00000100, + 0x240d, 0xffffffff, 0x00000100, + 0x240e, 0xffffffff, 0x00000100, + 0x240f, 0xffffffff, 0x00000100, + 0x2b60, 0xffffffff, 0x00000100, + 0x2b15, 0xffffffff, 0x00000100, + 0x225f, 0xffffffff, 0x06000100, + 0x261a, 0xffffffff, 0x00000100, + 0x2544, 0xffffffff, 0x00000100, + 0x2bc1, 0xffffffff, 0x00000100, + 0x2b81, 0xffffffff, 0x00000100, + 0x2527, 0xffffffff, 0x00000100, + 0x200b, 0xffffffff, 0xe0000000, + 0x2458, 0xffffffff, 0x00010000, + 0x2459, 0xffffffff, 0x00030002, + 0x245a, 0xffffffff, 0x00040007, + 0x245b, 0xffffffff, 0x00060005, + 0x245c, 0xffffffff, 0x00090008, + 0x245d, 0xffffffff, 0x00020001, + 0x245e, 0xffffffff, 0x00040003, + 0x245f, 0xffffffff, 0x00000007, + 0x2460, 0xffffffff, 0x00060005, + 0x2461, 0xffffffff, 0x00090008, + 0x2462, 0xffffffff, 0x00030002, + 0x2463, 0xffffffff, 0x00050004, + 0x2464, 0xffffffff, 0x00000008, + 0x2465, 0xffffffff, 0x00070006, + 0x2466, 0xffffffff, 0x000a0009, + 0x2467, 0xffffffff, 0x00040003, + 0x2468, 0xffffffff, 0x00060005, + 0x2469, 0xffffffff, 0x00000009, + 0x246a, 0xffffffff, 0x00080007, + 0x246b, 0xffffffff, 0x000b000a, + 0x246c, 0xffffffff, 0x00050004, + 0x246d, 0xffffffff, 0x00070006, + 0x246e, 0xffffffff, 0x0008000b, + 0x246f, 0xffffffff, 0x000a0009, + 0x2470, 0xffffffff, 0x000d000c, + 0x2480, 0xffffffff, 0x00090008, + 0x2481, 0xffffffff, 0x000b000a, + 0x2482, 0xffffffff, 0x000c000f, + 0x2483, 0xffffffff, 0x000e000d, + 0x2484, 0xffffffff, 0x00110010, + 0x2485, 0xffffffff, 0x000a0009, + 0x2486, 0xffffffff, 0x000c000b, + 0x2487, 0xffffffff, 0x0000000f, + 0x2488, 0xffffffff, 0x000e000d, + 0x2489, 0xffffffff, 0x00110010, + 0x248a, 0xffffffff, 0x000b000a, + 0x248b, 0xffffffff, 0x000d000c, + 0x248c, 0xffffffff, 0x00000010, + 0x248d, 0xffffffff, 0x000f000e, + 0x248e, 0xffffffff, 0x00120011, + 0x248f, 0xffffffff, 0x000c000b, + 0x2490, 0xffffffff, 0x000e000d, + 0x2491, 0xffffffff, 0x00000011, + 0x2492, 0xffffffff, 0x0010000f, + 0x2493, 0xffffffff, 0x00130012, + 0x2494, 0xffffffff, 0x000d000c, + 0x2495, 0xffffffff, 0x000f000e, + 0x2496, 0xffffffff, 0x00100013, + 0x2497, 0xffffffff, 0x00120011, + 0x2498, 0xffffffff, 0x00150014, + 0x2454, 0xffffffff, 0x96940200, + 0x21c2, 0xffffffff, 0x00900100, + 0x311e, 0xffffffff, 0x00000080, + 0x3101, 0xffffffff, 0x0020003f, + 0xc, 0xffffffff, 0x0000001c, + 0xd, 0x000f0000, 0x000f0000, + 0x583, 0xffffffff, 0x00000100, + 0x409, 0xffffffff, 0x00000100, + 0x40b, 0x00000101, 0x00000000, + 0x82a, 0xffffffff, 0x00000104, + 0x993, 0x000c0000, 0x000c0000, + 0x992, 0x000c0000, 0x000c0000, + 0x1579, 0xff000fff, 0x00000100, + 0x157a, 0x00000001, 0x00000001, + 0xbd4, 0x00000001, 0x00000001, + 0xc33, 0xc0000fff, 0x00000104, + 0x3079, 0x00000001, 0x00000001, + 0x3430, 0xfffffff0, 0x00000100, + 0x3630, 0xfffffff0, 0x00000100 +}; +static const u32 oland_mgcg_cgcg_init[] = +{ + 0x3100, 0xffffffff, 0xfffffffc, + 0x200b, 0xffffffff, 0xe0000000, + 0x2698, 0xffffffff, 0x00000100, + 0x24a9, 0xffffffff, 0x00000100, + 0x3059, 0xffffffff, 0x00000100, + 0x25dd, 0xffffffff, 0x00000100, + 0x2261, 0xffffffff, 0x06000100, + 0x2286, 0xffffffff, 0x00000100, + 0x24a8, 0xffffffff, 0x00000100, + 0x30e0, 0xffffffff, 0x00000100, + 0x22ca, 0xffffffff, 0x00000100, + 0x2451, 0xffffffff, 0x00000100, + 0x2362, 0xffffffff, 0x00000100, + 0x2363, 0xffffffff, 0x00000100, + 0x240c, 0xffffffff, 0x00000100, + 0x240d, 0xffffffff, 0x00000100, + 0x240e, 0xffffffff, 0x00000100, + 0x240f, 0xffffffff, 0x00000100, + 0x2b60, 0xffffffff, 0x00000100, + 0x2b15, 0xffffffff, 0x00000100, + 0x225f, 0xffffffff, 0x06000100, + 0x261a, 0xffffffff, 0x00000100, + 0x2544, 0xffffffff, 0x00000100, + 0x2bc1, 0xffffffff, 0x00000100, + 0x2b81, 0xffffffff, 0x00000100, + 0x2527, 0xffffffff, 0x00000100, + 0x200b, 0xffffffff, 0xe0000000, + 0x2458, 0xffffffff, 0x00010000, + 0x2459, 0xffffffff, 0x00030002, + 0x245a, 0xffffffff, 0x00040007, + 0x245b, 0xffffffff, 0x00060005, + 0x245c, 0xffffffff, 0x00090008, + 0x245d, 0xffffffff, 0x00020001, + 0x245e, 0xffffffff, 0x00040003, + 0x245f, 0xffffffff, 0x00000007, + 0x2460, 0xffffffff, 0x00060005, + 0x2461, 0xffffffff, 0x00090008, + 0x2462, 0xffffffff, 0x00030002, + 0x2463, 0xffffffff, 0x00050004, + 0x2464, 0xffffffff, 0x00000008, + 0x2465, 0xffffffff, 0x00070006, + 0x2466, 0xffffffff, 0x000a0009, + 0x2467, 0xffffffff, 0x00040003, + 0x2468, 0xffffffff, 0x00060005, + 0x2469, 0xffffffff, 0x00000009, + 0x246a, 0xffffffff, 0x00080007, + 0x246b, 0xffffffff, 0x000b000a, + 0x246c, 0xffffffff, 0x00050004, + 0x246d, 0xffffffff, 0x00070006, + 0x246e, 0xffffffff, 0x0008000b, + 0x246f, 0xffffffff, 0x000a0009, + 0x2470, 0xffffffff, 0x000d000c, + 0x2471, 0xffffffff, 0x00060005, + 0x2472, 0xffffffff, 0x00080007, + 0x2473, 0xffffffff, 0x0000000b, + 0x2474, 0xffffffff, 0x000a0009, + 0x2475, 0xffffffff, 0x000d000c, + 0x2454, 0xffffffff, 0x96940200, + 0x21c2, 0xffffffff, 0x00900100, + 0x311e, 0xffffffff, 0x00000080, + 0x3101, 0xffffffff, 0x0020003f, + 0xc, 0xffffffff, 0x0000001c, + 0xd, 0x000f0000, 0x000f0000, + 0x583, 0xffffffff, 0x00000100, + 0x409, 0xffffffff, 0x00000100, + 0x40b, 0x00000101, 0x00000000, + 0x82a, 0xffffffff, 0x00000104, + 0x993, 0x000c0000, 0x000c0000, + 0x992, 0x000c0000, 0x000c0000, + 0x1579, 0xff000fff, 0x00000100, + 0x157a, 0x00000001, 0x00000001, + 0xbd4, 0x00000001, 0x00000001, + 0xc33, 0xc0000fff, 0x00000104, + 0x3079, 0x00000001, 0x00000001, + 0x3430, 0xfffffff0, 0x00000100, + 0x3630, 0xfffffff0, 0x00000100 +}; +static const u32 hainan_mgcg_cgcg_init[] = +{ + 0x3100, 0xffffffff, 0xfffffffc, + 0x200b, 0xffffffff, 0xe0000000, + 0x2698, 0xffffffff, 0x00000100, + 0x24a9, 0xffffffff, 0x00000100, + 0x3059, 0xffffffff, 0x00000100, + 0x25dd, 0xffffffff, 0x00000100, + 0x2261, 0xffffffff, 0x06000100, + 0x2286, 0xffffffff, 0x00000100, + 0x24a8, 0xffffffff, 0x00000100, + 0x30e0, 0xffffffff, 0x00000100, + 0x22ca, 0xffffffff, 0x00000100, + 0x2451, 0xffffffff, 0x00000100, + 0x2362, 0xffffffff, 0x00000100, + 0x2363, 0xffffffff, 0x00000100, + 0x240c, 0xffffffff, 0x00000100, + 0x240d, 0xffffffff, 0x00000100, + 0x240e, 0xffffffff, 0x00000100, + 0x240f, 0xffffffff, 0x00000100, + 0x2b60, 0xffffffff, 0x00000100, + 0x2b15, 0xffffffff, 0x00000100, + 0x225f, 0xffffffff, 0x06000100, + 0x261a, 0xffffffff, 0x00000100, + 0x2544, 0xffffffff, 0x00000100, + 0x2bc1, 0xffffffff, 0x00000100, + 0x2b81, 0xffffffff, 0x00000100, + 0x2527, 0xffffffff, 0x00000100, + 0x200b, 0xffffffff, 0xe0000000, + 0x2458, 0xffffffff, 0x00010000, + 0x2459, 0xffffffff, 0x00030002, + 0x245a, 0xffffffff, 0x00040007, + 0x245b, 0xffffffff, 0x00060005, + 0x245c, 0xffffffff, 0x00090008, + 0x245d, 0xffffffff, 0x00020001, + 0x245e, 0xffffffff, 0x00040003, + 0x245f, 0xffffffff, 0x00000007, + 0x2460, 0xffffffff, 0x00060005, + 0x2461, 0xffffffff, 0x00090008, + 0x2462, 0xffffffff, 0x00030002, + 0x2463, 0xffffffff, 0x00050004, + 0x2464, 0xffffffff, 0x00000008, + 0x2465, 0xffffffff, 0x00070006, + 0x2466, 0xffffffff, 0x000a0009, + 0x2467, 0xffffffff, 0x00040003, + 0x2468, 0xffffffff, 0x00060005, + 0x2469, 0xffffffff, 0x00000009, + 0x246a, 0xffffffff, 0x00080007, + 0x246b, 0xffffffff, 0x000b000a, + 0x246c, 0xffffffff, 0x00050004, + 0x246d, 0xffffffff, 0x00070006, + 0x246e, 0xffffffff, 0x0008000b, + 0x246f, 0xffffffff, 0x000a0009, + 0x2470, 0xffffffff, 0x000d000c, + 0x2471, 0xffffffff, 0x00060005, + 0x2472, 0xffffffff, 0x00080007, + 0x2473, 0xffffffff, 0x0000000b, + 0x2474, 0xffffffff, 0x000a0009, + 0x2475, 0xffffffff, 0x000d000c, + 0x2454, 0xffffffff, 0x96940200, + 0x21c2, 0xffffffff, 0x00900100, + 0x311e, 0xffffffff, 0x00000080, + 0x3101, 0xffffffff, 0x0020003f, + 0xc, 0xffffffff, 0x0000001c, + 0xd, 0x000f0000, 0x000f0000, + 0x583, 0xffffffff, 0x00000100, + 0x409, 0xffffffff, 0x00000100, + 0x82a, 0xffffffff, 0x00000104, + 0x993, 0x000c0000, 0x000c0000, + 0x992, 0x000c0000, 0x000c0000, + 0xbd4, 0x00000001, 0x00000001, + 0xc33, 0xc0000fff, 0x00000104, + 0x3079, 0x00000001, 0x00000001, + 0x3430, 0xfffffff0, 0x00000100, + 0x3630, 0xfffffff0, 0x00000100 +}; + +static u32 si_pcie_rreg(struct amdgpu_device *adev, u32 reg) +{ + unsigned long flags; + u32 r; + + spin_lock_irqsave(&adev->pcie_idx_lock, flags); + WREG32(AMDGPU_PCIE_INDEX, reg); + (void)RREG32(AMDGPU_PCIE_INDEX); + r = RREG32(AMDGPU_PCIE_DATA); + spin_unlock_irqrestore(&adev->pcie_idx_lock, flags); + return r; +} + +static void si_pcie_wreg(struct amdgpu_device *adev, u32 reg, u32 v) +{ + unsigned long flags; + + spin_lock_irqsave(&adev->pcie_idx_lock, flags); + WREG32(AMDGPU_PCIE_INDEX, reg); + (void)RREG32(AMDGPU_PCIE_INDEX); + WREG32(AMDGPU_PCIE_DATA, v); + (void)RREG32(AMDGPU_PCIE_DATA); + spin_unlock_irqrestore(&adev->pcie_idx_lock, flags); +} + +u32 si_pciep_rreg(struct amdgpu_device *adev, u32 reg) +{ + unsigned long flags; + u32 r; + + spin_lock_irqsave(&adev->pcie_idx_lock, flags); + WREG32(PCIE_PORT_INDEX, ((reg) & 0xff)); + (void)RREG32(PCIE_PORT_INDEX); + r = RREG32(PCIE_PORT_DATA); + spin_unlock_irqrestore(&adev->pcie_idx_lock, flags); + return r; +} + +void si_pciep_wreg(struct amdgpu_device *adev, u32 reg, u32 v) +{ + unsigned long flags; + + spin_lock_irqsave(&adev->pcie_idx_lock, flags); + WREG32(PCIE_PORT_INDEX, ((reg) & 0xff)); + (void)RREG32(PCIE_PORT_INDEX); + WREG32(PCIE_PORT_DATA, (v)); + (void)RREG32(PCIE_PORT_DATA); + spin_unlock_irqrestore(&adev->pcie_idx_lock, flags); +} + +static u32 si_smc_rreg(struct amdgpu_device *adev, u32 reg) +{ + unsigned long flags; + u32 r; + + spin_lock_irqsave(&adev->smc_idx_lock, flags); + WREG32(SMC_IND_INDEX_0, (reg)); + r = RREG32(SMC_IND_DATA_0); + spin_unlock_irqrestore(&adev->smc_idx_lock, flags); + return r; +} + +static void si_smc_wreg(struct amdgpu_device *adev, u32 reg, u32 v) +{ + unsigned long flags; + + spin_lock_irqsave(&adev->smc_idx_lock, flags); + WREG32(SMC_IND_INDEX_0, (reg)); + WREG32(SMC_IND_DATA_0, (v)); + spin_unlock_irqrestore(&adev->smc_idx_lock, flags); +} + +static u32 si_get_virtual_caps(struct amdgpu_device *adev) +{ + /* SI does not support SR-IOV */ + return 0; +} + +static struct amdgpu_allowed_register_entry si_allowed_read_registers[] = { + {GRBM_STATUS, false}, + {GB_ADDR_CONFIG, false}, + {MC_ARB_RAMCFG, false}, + {GB_TILE_MODE0, false}, + {GB_TILE_MODE1, false}, + {GB_TILE_MODE2, false}, + {GB_TILE_MODE3, false}, + {GB_TILE_MODE4, false}, + {GB_TILE_MODE5, false}, + {GB_TILE_MODE6, false}, + {GB_TILE_MODE7, false}, + {GB_TILE_MODE8, false}, + {GB_TILE_MODE9, false}, + {GB_TILE_MODE10, false}, + {GB_TILE_MODE11, false}, + {GB_TILE_MODE12, false}, + {GB_TILE_MODE13, false}, + {GB_TILE_MODE14, false}, + {GB_TILE_MODE15, false}, + {GB_TILE_MODE16, false}, + {GB_TILE_MODE17, false}, + {GB_TILE_MODE18, false}, + {GB_TILE_MODE19, false}, + {GB_TILE_MODE20, false}, + {GB_TILE_MODE21, false}, + {GB_TILE_MODE22, false}, + {GB_TILE_MODE23, false}, + {GB_TILE_MODE24, false}, + {GB_TILE_MODE25, false}, + {GB_TILE_MODE26, false}, + {GB_TILE_MODE27, false}, + {GB_TILE_MODE28, false}, + {GB_TILE_MODE29, false}, + {GB_TILE_MODE30, false}, + {GB_TILE_MODE31, false}, + {CC_RB_BACKEND_DISABLE, false, true}, + {GC_USER_RB_BACKEND_DISABLE, false, true}, + {PA_SC_RASTER_CONFIG, false, true}, +}; + +static uint32_t si_read_indexed_register(struct amdgpu_device *adev, + u32 se_num, u32 sh_num, + u32 reg_offset) +{ + uint32_t val; + + mutex_lock(&adev->grbm_idx_mutex); + if (se_num != 0xffffffff || sh_num != 0xffffffff) + amdgpu_gfx_select_se_sh(adev, se_num, sh_num, 0xffffffff); + + val = RREG32(reg_offset); + + if (se_num != 0xffffffff || sh_num != 0xffffffff) + amdgpu_gfx_select_se_sh(adev, 0xffffffff, 0xffffffff, 0xffffffff); + mutex_unlock(&adev->grbm_idx_mutex); + return val; +} + +static int si_read_register(struct amdgpu_device *adev, u32 se_num, + u32 sh_num, u32 reg_offset, u32 *value) +{ + uint32_t i; + + *value = 0; + for (i = 0; i < ARRAY_SIZE(si_allowed_read_registers); i++) { + if (reg_offset != si_allowed_read_registers[i].reg_offset) + continue; + + if (!si_allowed_read_registers[i].untouched) + *value = si_allowed_read_registers[i].grbm_indexed ? + si_read_indexed_register(adev, se_num, + sh_num, reg_offset) : + RREG32(reg_offset); + return 0; + } + return -EINVAL; +} + +static bool si_read_disabled_bios(struct amdgpu_device *adev) +{ + u32 bus_cntl; + u32 d1vga_control = 0; + u32 d2vga_control = 0; + u32 vga_render_control = 0; + u32 rom_cntl; + bool r; + + bus_cntl = RREG32(R600_BUS_CNTL); + if (adev->mode_info.num_crtc) { + d1vga_control = RREG32(AVIVO_D1VGA_CONTROL); + d2vga_control = RREG32(AVIVO_D2VGA_CONTROL); + vga_render_control = RREG32(VGA_RENDER_CONTROL); + } + rom_cntl = RREG32(R600_ROM_CNTL); + + /* enable the rom */ + WREG32(R600_BUS_CNTL, (bus_cntl & ~R600_BIOS_ROM_DIS)); + if (adev->mode_info.num_crtc) { + /* Disable VGA mode */ + WREG32(AVIVO_D1VGA_CONTROL, + (d1vga_control & ~(AVIVO_DVGA_CONTROL_MODE_ENABLE | + AVIVO_DVGA_CONTROL_TIMING_SELECT))); + WREG32(AVIVO_D2VGA_CONTROL, + (d2vga_control & ~(AVIVO_DVGA_CONTROL_MODE_ENABLE | + AVIVO_DVGA_CONTROL_TIMING_SELECT))); + WREG32(VGA_RENDER_CONTROL, + (vga_render_control & C_000300_VGA_VSTATUS_CNTL)); + } + WREG32(R600_ROM_CNTL, rom_cntl | R600_SCK_OVERWRITE); + + r = amdgpu_read_bios(adev); + + /* restore regs */ + WREG32(R600_BUS_CNTL, bus_cntl); + if (adev->mode_info.num_crtc) { + WREG32(AVIVO_D1VGA_CONTROL, d1vga_control); + WREG32(AVIVO_D2VGA_CONTROL, d2vga_control); + WREG32(VGA_RENDER_CONTROL, vga_render_control); + } + WREG32(R600_ROM_CNTL, rom_cntl); + return r; +} + +//xxx: not implemented +static int si_asic_reset(struct amdgpu_device *adev) +{ + return 0; +} + +static void si_vga_set_state(struct amdgpu_device *adev, bool state) +{ + uint32_t temp; + + temp = RREG32(CONFIG_CNTL); + if (state == false) { + temp &= ~(1<<0); + temp |= (1<<1); + } else { + temp &= ~(1<<1); + } + WREG32(CONFIG_CNTL, temp); +} + +static u32 si_get_xclk(struct amdgpu_device *adev) +{ + u32 reference_clock = adev->clock.spll.reference_freq; + u32 tmp; + + tmp = RREG32(CG_CLKPIN_CNTL_2); + if (tmp & MUX_TCLK_TO_XCLK) + return TCLK; + + tmp = RREG32(CG_CLKPIN_CNTL); + if (tmp & XTALIN_DIVIDE) + return reference_clock / 4; + + return reference_clock; +} + +//xxx:not implemented +static int si_set_uvd_clocks(struct amdgpu_device *adev, u32 vclk, u32 dclk) +{ + return 0; +} + +static const struct amdgpu_asic_funcs si_asic_funcs = +{ + .read_disabled_bios = &si_read_disabled_bios, + .read_register = &si_read_register, + .reset = &si_asic_reset, + .set_vga_state = &si_vga_set_state, + .get_xclk = &si_get_xclk, + .set_uvd_clocks = &si_set_uvd_clocks, + .set_vce_clocks = NULL, + .get_virtual_caps = &si_get_virtual_caps, +}; + +static uint32_t si_get_rev_id(struct amdgpu_device *adev) +{ + return (RREG32(CC_DRM_ID_STRAPS) & CC_DRM_ID_STRAPS__ATI_REV_ID_MASK) + >> CC_DRM_ID_STRAPS__ATI_REV_ID__SHIFT; +} + +static int si_common_early_init(void *handle) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + adev->smc_rreg = &si_smc_rreg; + adev->smc_wreg = &si_smc_wreg; + adev->pcie_rreg = &si_pcie_rreg; + adev->pcie_wreg = &si_pcie_wreg; + adev->pciep_rreg = &si_pciep_rreg; + adev->pciep_wreg = &si_pciep_wreg; + adev->uvd_ctx_rreg = NULL; + adev->uvd_ctx_wreg = NULL; + adev->didt_rreg = NULL; + adev->didt_wreg = NULL; + + adev->asic_funcs = &si_asic_funcs; + + adev->rev_id = si_get_rev_id(adev); + adev->external_rev_id = 0xFF; + switch (adev->asic_type) { + case CHIP_TAHITI: + adev->cg_flags = + AMD_CG_SUPPORT_GFX_MGCG | + AMD_CG_SUPPORT_GFX_MGLS | + /*AMD_CG_SUPPORT_GFX_CGCG |*/ + AMD_CG_SUPPORT_GFX_CGLS | + AMD_CG_SUPPORT_GFX_CGTS | + AMD_CG_SUPPORT_GFX_CP_LS | + AMD_CG_SUPPORT_MC_MGCG | + AMD_CG_SUPPORT_SDMA_MGCG | + AMD_CG_SUPPORT_BIF_LS | + AMD_CG_SUPPORT_VCE_MGCG | + AMD_CG_SUPPORT_UVD_MGCG | + AMD_CG_SUPPORT_HDP_LS | + AMD_CG_SUPPORT_HDP_MGCG; + adev->pg_flags = 0; + break; + case CHIP_PITCAIRN: + adev->cg_flags = + AMD_CG_SUPPORT_GFX_MGCG | + AMD_CG_SUPPORT_GFX_MGLS | + /*AMD_CG_SUPPORT_GFX_CGCG |*/ + AMD_CG_SUPPORT_GFX_CGLS | + AMD_CG_SUPPORT_GFX_CGTS | + AMD_CG_SUPPORT_GFX_CP_LS | + AMD_CG_SUPPORT_GFX_RLC_LS | + AMD_CG_SUPPORT_MC_LS | + AMD_CG_SUPPORT_MC_MGCG | + AMD_CG_SUPPORT_SDMA_MGCG | + AMD_CG_SUPPORT_BIF_LS | + AMD_CG_SUPPORT_VCE_MGCG | + AMD_CG_SUPPORT_UVD_MGCG | + AMD_CG_SUPPORT_HDP_LS | + AMD_CG_SUPPORT_HDP_MGCG; + adev->pg_flags = 0; + break; + + case CHIP_VERDE: + adev->cg_flags = + AMD_CG_SUPPORT_GFX_MGCG | + AMD_CG_SUPPORT_GFX_MGLS | + AMD_CG_SUPPORT_GFX_CGLS | + AMD_CG_SUPPORT_GFX_CGTS | + AMD_CG_SUPPORT_GFX_CGTS_LS | + AMD_CG_SUPPORT_GFX_CP_LS | + AMD_CG_SUPPORT_MC_LS | + AMD_CG_SUPPORT_MC_MGCG | + AMD_CG_SUPPORT_SDMA_MGCG | + AMD_CG_SUPPORT_SDMA_LS | + AMD_CG_SUPPORT_BIF_LS | + AMD_CG_SUPPORT_VCE_MGCG | + AMD_CG_SUPPORT_UVD_MGCG | + AMD_CG_SUPPORT_HDP_LS | + AMD_CG_SUPPORT_HDP_MGCG; + adev->pg_flags = 0; + //??? + adev->external_rev_id = adev->rev_id + 0x14; + break; + case CHIP_OLAND: + adev->cg_flags = + AMD_CG_SUPPORT_GFX_MGCG | + AMD_CG_SUPPORT_GFX_MGLS | + /*AMD_CG_SUPPORT_GFX_CGCG |*/ + AMD_CG_SUPPORT_GFX_CGLS | + AMD_CG_SUPPORT_GFX_CGTS | + AMD_CG_SUPPORT_GFX_CP_LS | + AMD_CG_SUPPORT_GFX_RLC_LS | + AMD_CG_SUPPORT_MC_LS | + AMD_CG_SUPPORT_MC_MGCG | + AMD_CG_SUPPORT_SDMA_MGCG | + AMD_CG_SUPPORT_BIF_LS | + AMD_CG_SUPPORT_UVD_MGCG | + AMD_CG_SUPPORT_HDP_LS | + AMD_CG_SUPPORT_HDP_MGCG; + adev->pg_flags = 0; + break; + case CHIP_HAINAN: + adev->cg_flags = + AMD_CG_SUPPORT_GFX_MGCG | + AMD_CG_SUPPORT_GFX_MGLS | + /*AMD_CG_SUPPORT_GFX_CGCG |*/ + AMD_CG_SUPPORT_GFX_CGLS | + AMD_CG_SUPPORT_GFX_CGTS | + AMD_CG_SUPPORT_GFX_CP_LS | + AMD_CG_SUPPORT_GFX_RLC_LS | + AMD_CG_SUPPORT_MC_LS | + AMD_CG_SUPPORT_MC_MGCG | + AMD_CG_SUPPORT_SDMA_MGCG | + AMD_CG_SUPPORT_BIF_LS | + AMD_CG_SUPPORT_HDP_LS | + AMD_CG_SUPPORT_HDP_MGCG; + adev->pg_flags = 0; + break; + + default: + return -EINVAL; + } + + return 0; +} + +static int si_common_sw_init(void *handle) +{ + return 0; +} + +static int si_common_sw_fini(void *handle) +{ + return 0; +} + + +static void si_init_golden_registers(struct amdgpu_device *adev) +{ + switch (adev->asic_type) { + case CHIP_TAHITI: + amdgpu_program_register_sequence(adev, + tahiti_golden_registers, + (const u32)ARRAY_SIZE(tahiti_golden_registers)); + amdgpu_program_register_sequence(adev, + tahiti_golden_rlc_registers, + (const u32)ARRAY_SIZE(tahiti_golden_rlc_registers)); + amdgpu_program_register_sequence(adev, + tahiti_mgcg_cgcg_init, + (const u32)ARRAY_SIZE(tahiti_mgcg_cgcg_init)); + amdgpu_program_register_sequence(adev, + tahiti_golden_registers2, + (const u32)ARRAY_SIZE(tahiti_golden_registers2)); + break; + case CHIP_PITCAIRN: + amdgpu_program_register_sequence(adev, + pitcairn_golden_registers, + (const u32)ARRAY_SIZE(pitcairn_golden_registers)); + amdgpu_program_register_sequence(adev, + pitcairn_golden_rlc_registers, + (const u32)ARRAY_SIZE(pitcairn_golden_rlc_registers)); + amdgpu_program_register_sequence(adev, + pitcairn_mgcg_cgcg_init, + (const u32)ARRAY_SIZE(pitcairn_mgcg_cgcg_init)); + case CHIP_VERDE: + amdgpu_program_register_sequence(adev, + verde_golden_registers, + (const u32)ARRAY_SIZE(verde_golden_registers)); + amdgpu_program_register_sequence(adev, + verde_golden_rlc_registers, + (const u32)ARRAY_SIZE(verde_golden_rlc_registers)); + amdgpu_program_register_sequence(adev, + verde_mgcg_cgcg_init, + (const u32)ARRAY_SIZE(verde_mgcg_cgcg_init)); + amdgpu_program_register_sequence(adev, + verde_pg_init, + (const u32)ARRAY_SIZE(verde_pg_init)); + break; + case CHIP_OLAND: + amdgpu_program_register_sequence(adev, + oland_golden_registers, + (const u32)ARRAY_SIZE(oland_golden_registers)); + amdgpu_program_register_sequence(adev, + oland_golden_rlc_registers, + (const u32)ARRAY_SIZE(oland_golden_rlc_registers)); + amdgpu_program_register_sequence(adev, + oland_mgcg_cgcg_init, + (const u32)ARRAY_SIZE(oland_mgcg_cgcg_init)); + case CHIP_HAINAN: + amdgpu_program_register_sequence(adev, + hainan_golden_registers, + (const u32)ARRAY_SIZE(hainan_golden_registers)); + amdgpu_program_register_sequence(adev, + hainan_golden_registers2, + (const u32)ARRAY_SIZE(hainan_golden_registers2)); + amdgpu_program_register_sequence(adev, + hainan_mgcg_cgcg_init, + (const u32)ARRAY_SIZE(hainan_mgcg_cgcg_init)); + break; + + + default: + BUG(); + } +} + +static void si_pcie_gen3_enable(struct amdgpu_device *adev) +{ + struct pci_dev *root = adev->pdev->bus->self; + int bridge_pos, gpu_pos; + u32 speed_cntl, mask, current_data_rate; + int ret, i; + u16 tmp16; + + if (pci_is_root_bus(adev->pdev->bus)) + return; + + if (amdgpu_pcie_gen2 == 0) + return; + + if (adev->flags & AMD_IS_APU) + return; + + ret = drm_pcie_get_speed_cap_mask(adev->ddev, &mask); + if (ret != 0) + return; + + if (!(mask & (DRM_PCIE_SPEED_50 | DRM_PCIE_SPEED_80))) + return; + + speed_cntl = RREG32_PCIE_PORT(PCIE_LC_SPEED_CNTL); + current_data_rate = (speed_cntl & LC_CURRENT_DATA_RATE_MASK) >> + LC_CURRENT_DATA_RATE_SHIFT; + if (mask & DRM_PCIE_SPEED_80) { + if (current_data_rate == 2) { + DRM_INFO("PCIE gen 3 link speeds already enabled\n"); + return; + } + DRM_INFO("enabling PCIE gen 3 link speeds, disable with amdgpu.pcie_gen2=0\n"); + } else if (mask & DRM_PCIE_SPEED_50) { + if (current_data_rate == 1) { + DRM_INFO("PCIE gen 2 link speeds already enabled\n"); + return; + } + DRM_INFO("enabling PCIE gen 2 link speeds, disable with amdgpu.pcie_gen2=0\n"); + } + + bridge_pos = pci_pcie_cap(root); + if (!bridge_pos) + return; + + gpu_pos = pci_pcie_cap(adev->pdev); + if (!gpu_pos) + return; + + if (mask & DRM_PCIE_SPEED_80) { + if (current_data_rate != 2) { + u16 bridge_cfg, gpu_cfg; + u16 bridge_cfg2, gpu_cfg2; + u32 max_lw, current_lw, tmp; + + pci_read_config_word(root, bridge_pos + PCI_EXP_LNKCTL, &bridge_cfg); + pci_read_config_word(adev->pdev, gpu_pos + PCI_EXP_LNKCTL, &gpu_cfg); + + tmp16 = bridge_cfg | PCI_EXP_LNKCTL_HAWD; + pci_write_config_word(root, bridge_pos + PCI_EXP_LNKCTL, tmp16); + + tmp16 = gpu_cfg | PCI_EXP_LNKCTL_HAWD; + pci_write_config_word(adev->pdev, gpu_pos + PCI_EXP_LNKCTL, tmp16); + + tmp = RREG32_PCIE(PCIE_LC_STATUS1); + max_lw = (tmp & LC_DETECTED_LINK_WIDTH_MASK) >> LC_DETECTED_LINK_WIDTH_SHIFT; + current_lw = (tmp & LC_OPERATING_LINK_WIDTH_MASK) >> LC_OPERATING_LINK_WIDTH_SHIFT; + + if (current_lw < max_lw) { + tmp = RREG32_PCIE_PORT(PCIE_LC_LINK_WIDTH_CNTL); + if (tmp & LC_RENEGOTIATION_SUPPORT) { + tmp &= ~(LC_LINK_WIDTH_MASK | LC_UPCONFIGURE_DIS); + tmp |= (max_lw << LC_LINK_WIDTH_SHIFT); + tmp |= LC_UPCONFIGURE_SUPPORT | LC_RENEGOTIATE_EN | LC_RECONFIG_NOW; + WREG32_PCIE_PORT(PCIE_LC_LINK_WIDTH_CNTL, tmp); + } + } + + for (i = 0; i < 10; i++) { + pci_read_config_word(adev->pdev, gpu_pos + PCI_EXP_DEVSTA, &tmp16); + if (tmp16 & PCI_EXP_DEVSTA_TRPND) + break; + + pci_read_config_word(root, bridge_pos + PCI_EXP_LNKCTL, &bridge_cfg); + pci_read_config_word(adev->pdev, gpu_pos + PCI_EXP_LNKCTL, &gpu_cfg); + + pci_read_config_word(root, bridge_pos + PCI_EXP_LNKCTL2, &bridge_cfg2); + pci_read_config_word(adev->pdev, gpu_pos + PCI_EXP_LNKCTL2, &gpu_cfg2); + + tmp = RREG32_PCIE_PORT(PCIE_LC_CNTL4); + tmp |= LC_SET_QUIESCE; + WREG32_PCIE_PORT(PCIE_LC_CNTL4, tmp); + + tmp = RREG32_PCIE_PORT(PCIE_LC_CNTL4); + tmp |= LC_REDO_EQ; + WREG32_PCIE_PORT(PCIE_LC_CNTL4, tmp); + + mdelay(100); + + pci_read_config_word(root, bridge_pos + PCI_EXP_LNKCTL, &tmp16); + tmp16 &= ~PCI_EXP_LNKCTL_HAWD; + tmp16 |= (bridge_cfg & PCI_EXP_LNKCTL_HAWD); + pci_write_config_word(root, bridge_pos + PCI_EXP_LNKCTL, tmp16); + + pci_read_config_word(adev->pdev, gpu_pos + PCI_EXP_LNKCTL, &tmp16); + tmp16 &= ~PCI_EXP_LNKCTL_HAWD; + tmp16 |= (gpu_cfg & PCI_EXP_LNKCTL_HAWD); + pci_write_config_word(adev->pdev, gpu_pos + PCI_EXP_LNKCTL, tmp16); + + pci_read_config_word(root, bridge_pos + PCI_EXP_LNKCTL2, &tmp16); + tmp16 &= ~((1 << 4) | (7 << 9)); + tmp16 |= (bridge_cfg2 & ((1 << 4) | (7 << 9))); + pci_write_config_word(root, bridge_pos + PCI_EXP_LNKCTL2, tmp16); + + pci_read_config_word(adev->pdev, gpu_pos + PCI_EXP_LNKCTL2, &tmp16); + tmp16 &= ~((1 << 4) | (7 << 9)); + tmp16 |= (gpu_cfg2 & ((1 << 4) | (7 << 9))); + pci_write_config_word(adev->pdev, gpu_pos + PCI_EXP_LNKCTL2, tmp16); + + tmp = RREG32_PCIE_PORT(PCIE_LC_CNTL4); + tmp &= ~LC_SET_QUIESCE; + WREG32_PCIE_PORT(PCIE_LC_CNTL4, tmp); + } + } + } + + speed_cntl |= LC_FORCE_EN_SW_SPEED_CHANGE | LC_FORCE_DIS_HW_SPEED_CHANGE; + speed_cntl &= ~LC_FORCE_DIS_SW_SPEED_CHANGE; + WREG32_PCIE_PORT(PCIE_LC_SPEED_CNTL, speed_cntl); + + pci_read_config_word(adev->pdev, gpu_pos + PCI_EXP_LNKCTL2, &tmp16); + tmp16 &= ~0xf; + if (mask & DRM_PCIE_SPEED_80) + tmp16 |= 3; + else if (mask & DRM_PCIE_SPEED_50) + tmp16 |= 2; + else + tmp16 |= 1; + pci_write_config_word(adev->pdev, gpu_pos + PCI_EXP_LNKCTL2, tmp16); + + speed_cntl = RREG32_PCIE_PORT(PCIE_LC_SPEED_CNTL); + speed_cntl |= LC_INITIATE_LINK_SPEED_CHANGE; + WREG32_PCIE_PORT(PCIE_LC_SPEED_CNTL, speed_cntl); + + for (i = 0; i < adev->usec_timeout; i++) { + speed_cntl = RREG32_PCIE_PORT(PCIE_LC_SPEED_CNTL); + if ((speed_cntl & LC_INITIATE_LINK_SPEED_CHANGE) == 0) + break; + udelay(1); + } +} + +static inline u32 si_pif_phy0_rreg(struct amdgpu_device *adev, u32 reg) +{ + unsigned long flags; + u32 r; + + spin_lock_irqsave(&adev->pcie_idx_lock, flags); + WREG32(EVERGREEN_PIF_PHY0_INDEX, ((reg) & 0xffff)); + r = RREG32(EVERGREEN_PIF_PHY0_DATA); + spin_unlock_irqrestore(&adev->pcie_idx_lock, flags); + return r; +} + +static inline void si_pif_phy0_wreg(struct amdgpu_device *adev, u32 reg, u32 v) +{ + unsigned long flags; + + spin_lock_irqsave(&adev->pcie_idx_lock, flags); + WREG32(EVERGREEN_PIF_PHY0_INDEX, ((reg) & 0xffff)); + WREG32(EVERGREEN_PIF_PHY0_DATA, (v)); + spin_unlock_irqrestore(&adev->pcie_idx_lock, flags); +} + +static inline u32 si_pif_phy1_rreg(struct amdgpu_device *adev, u32 reg) +{ + unsigned long flags; + u32 r; + + spin_lock_irqsave(&adev->pcie_idx_lock, flags); + WREG32(EVERGREEN_PIF_PHY1_INDEX, ((reg) & 0xffff)); + r = RREG32(EVERGREEN_PIF_PHY1_DATA); + spin_unlock_irqrestore(&adev->pcie_idx_lock, flags); + return r; +} + +static inline void si_pif_phy1_wreg(struct amdgpu_device *adev, u32 reg, u32 v) +{ + unsigned long flags; + + spin_lock_irqsave(&adev->pcie_idx_lock, flags); + WREG32(EVERGREEN_PIF_PHY1_INDEX, ((reg) & 0xffff)); + WREG32(EVERGREEN_PIF_PHY1_DATA, (v)); + spin_unlock_irqrestore(&adev->pcie_idx_lock, flags); +} +static void si_program_aspm(struct amdgpu_device *adev) +{ + u32 data, orig; + bool disable_l0s = false, disable_l1 = false, disable_plloff_in_l1 = false; + bool disable_clkreq = false; + + if (amdgpu_aspm == 0) + return; + + if (adev->flags & AMD_IS_APU) + return; + orig = data = RREG32_PCIE_PORT(PCIE_LC_N_FTS_CNTL); + data &= ~LC_XMIT_N_FTS_MASK; + data |= LC_XMIT_N_FTS(0x24) | LC_XMIT_N_FTS_OVERRIDE_EN; + if (orig != data) + WREG32_PCIE_PORT(PCIE_LC_N_FTS_CNTL, data); + + orig = data = RREG32_PCIE_PORT(PCIE_LC_CNTL3); + data |= LC_GO_TO_RECOVERY; + if (orig != data) + WREG32_PCIE_PORT(PCIE_LC_CNTL3, data); + + orig = data = RREG32_PCIE(PCIE_P_CNTL); + data |= P_IGNORE_EDB_ERR; + if (orig != data) + WREG32_PCIE(PCIE_P_CNTL, data); + + orig = data = RREG32_PCIE_PORT(PCIE_LC_CNTL); + data &= ~(LC_L0S_INACTIVITY_MASK | LC_L1_INACTIVITY_MASK); + data |= LC_PMI_TO_L1_DIS; + if (!disable_l0s) + data |= LC_L0S_INACTIVITY(7); + + if (!disable_l1) { + data |= LC_L1_INACTIVITY(7); + data &= ~LC_PMI_TO_L1_DIS; + if (orig != data) + WREG32_PCIE_PORT(PCIE_LC_CNTL, data); + + if (!disable_plloff_in_l1) { + bool clk_req_support; + + orig = data = si_pif_phy0_rreg(adev,PB0_PIF_PWRDOWN_0); + data &= ~(PLL_POWER_STATE_IN_OFF_0_MASK | PLL_POWER_STATE_IN_TXS2_0_MASK); + data |= PLL_POWER_STATE_IN_OFF_0(7) | PLL_POWER_STATE_IN_TXS2_0(7); + if (orig != data) + si_pif_phy0_wreg(adev,PB0_PIF_PWRDOWN_0, data); + + orig = data = si_pif_phy0_rreg(adev,PB0_PIF_PWRDOWN_1); + data &= ~(PLL_POWER_STATE_IN_OFF_1_MASK | PLL_POWER_STATE_IN_TXS2_1_MASK); + data |= PLL_POWER_STATE_IN_OFF_1(7) | PLL_POWER_STATE_IN_TXS2_1(7); + if (orig != data) + si_pif_phy0_wreg(adev,PB0_PIF_PWRDOWN_1, data); + + orig = data = si_pif_phy1_rreg(adev,PB1_PIF_PWRDOWN_0); + data &= ~(PLL_POWER_STATE_IN_OFF_0_MASK | PLL_POWER_STATE_IN_TXS2_0_MASK); + data |= PLL_POWER_STATE_IN_OFF_0(7) | PLL_POWER_STATE_IN_TXS2_0(7); + if (orig != data) + si_pif_phy1_wreg(adev,PB1_PIF_PWRDOWN_0, data); + + orig = data = si_pif_phy1_rreg(adev,PB1_PIF_PWRDOWN_1); + data &= ~(PLL_POWER_STATE_IN_OFF_1_MASK | PLL_POWER_STATE_IN_TXS2_1_MASK); + data |= PLL_POWER_STATE_IN_OFF_1(7) | PLL_POWER_STATE_IN_TXS2_1(7); + if (orig != data) + si_pif_phy1_wreg(adev,PB1_PIF_PWRDOWN_1, data); + + if ((adev->family != CHIP_OLAND) && (adev->family != CHIP_HAINAN)) { + orig = data = si_pif_phy0_rreg(adev,PB0_PIF_PWRDOWN_0); + data &= ~PLL_RAMP_UP_TIME_0_MASK; + if (orig != data) + si_pif_phy0_wreg(adev,PB0_PIF_PWRDOWN_0, data); + + orig = data = si_pif_phy0_rreg(adev,PB0_PIF_PWRDOWN_1); + data &= ~PLL_RAMP_UP_TIME_1_MASK; + if (orig != data) + si_pif_phy0_wreg(adev,PB0_PIF_PWRDOWN_1, data); + + orig = data = si_pif_phy0_rreg(adev,PB0_PIF_PWRDOWN_2); + data &= ~PLL_RAMP_UP_TIME_2_MASK; + if (orig != data) + si_pif_phy0_wreg(adev,PB0_PIF_PWRDOWN_2, data); + + orig = data = si_pif_phy0_rreg(adev,PB0_PIF_PWRDOWN_3); + data &= ~PLL_RAMP_UP_TIME_3_MASK; + if (orig != data) + si_pif_phy0_wreg(adev,PB0_PIF_PWRDOWN_3, data); + + orig = data = si_pif_phy1_rreg(adev,PB1_PIF_PWRDOWN_0); + data &= ~PLL_RAMP_UP_TIME_0_MASK; + if (orig != data) + si_pif_phy1_wreg(adev,PB1_PIF_PWRDOWN_0, data); + + orig = data = si_pif_phy1_rreg(adev,PB1_PIF_PWRDOWN_1); + data &= ~PLL_RAMP_UP_TIME_1_MASK; + if (orig != data) + si_pif_phy1_wreg(adev,PB1_PIF_PWRDOWN_1, data); + + orig = data = si_pif_phy1_rreg(adev,PB1_PIF_PWRDOWN_2); + data &= ~PLL_RAMP_UP_TIME_2_MASK; + if (orig != data) + si_pif_phy1_wreg(adev,PB1_PIF_PWRDOWN_2, data); + + orig = data = si_pif_phy1_rreg(adev,PB1_PIF_PWRDOWN_3); + data &= ~PLL_RAMP_UP_TIME_3_MASK; + if (orig != data) + si_pif_phy1_wreg(adev,PB1_PIF_PWRDOWN_3, data); + } + orig = data = RREG32_PCIE_PORT(PCIE_LC_LINK_WIDTH_CNTL); + data &= ~LC_DYN_LANES_PWR_STATE_MASK; + data |= LC_DYN_LANES_PWR_STATE(3); + if (orig != data) + WREG32_PCIE_PORT(PCIE_LC_LINK_WIDTH_CNTL, data); + + orig = data = si_pif_phy0_rreg(adev,PB0_PIF_CNTL); + data &= ~LS2_EXIT_TIME_MASK; + if ((adev->family == CHIP_OLAND) || (adev->family == CHIP_HAINAN)) + data |= LS2_EXIT_TIME(5); + if (orig != data) + si_pif_phy0_wreg(adev,PB0_PIF_CNTL, data); + + orig = data = si_pif_phy1_rreg(adev,PB1_PIF_CNTL); + data &= ~LS2_EXIT_TIME_MASK; + if ((adev->family == CHIP_OLAND) || (adev->family == CHIP_HAINAN)) + data |= LS2_EXIT_TIME(5); + if (orig != data) + si_pif_phy1_wreg(adev,PB1_PIF_CNTL, data); + + if (!disable_clkreq && + !pci_is_root_bus(adev->pdev->bus)) { + struct pci_dev *root = adev->pdev->bus->self; + u32 lnkcap; + + clk_req_support = false; + pcie_capability_read_dword(root, PCI_EXP_LNKCAP, &lnkcap); + if (lnkcap & PCI_EXP_LNKCAP_CLKPM) + clk_req_support = true; + } else { + clk_req_support = false; + } + + if (clk_req_support) { + orig = data = RREG32_PCIE_PORT(PCIE_LC_CNTL2); + data |= LC_ALLOW_PDWN_IN_L1 | LC_ALLOW_PDWN_IN_L23; + if (orig != data) + WREG32_PCIE_PORT(PCIE_LC_CNTL2, data); + + orig = data = RREG32(THM_CLK_CNTL); + data &= ~(CMON_CLK_SEL_MASK | TMON_CLK_SEL_MASK); + data |= CMON_CLK_SEL(1) | TMON_CLK_SEL(1); + if (orig != data) + WREG32(THM_CLK_CNTL, data); + + orig = data = RREG32(MISC_CLK_CNTL); + data &= ~(DEEP_SLEEP_CLK_SEL_MASK | ZCLK_SEL_MASK); + data |= DEEP_SLEEP_CLK_SEL(1) | ZCLK_SEL(1); + if (orig != data) + WREG32(MISC_CLK_CNTL, data); + + orig = data = RREG32(CG_CLKPIN_CNTL); + data &= ~BCLK_AS_XCLK; + if (orig != data) + WREG32(CG_CLKPIN_CNTL, data); + + orig = data = RREG32(CG_CLKPIN_CNTL_2); + data &= ~FORCE_BIF_REFCLK_EN; + if (orig != data) + WREG32(CG_CLKPIN_CNTL_2, data); + + orig = data = RREG32(MPLL_BYPASSCLK_SEL); + data &= ~MPLL_CLKOUT_SEL_MASK; + data |= MPLL_CLKOUT_SEL(4); + if (orig != data) + WREG32(MPLL_BYPASSCLK_SEL, data); + + orig = data = RREG32(SPLL_CNTL_MODE); + data &= ~SPLL_REFCLK_SEL_MASK; + if (orig != data) + WREG32(SPLL_CNTL_MODE, data); + } + } + } else { + if (orig != data) + WREG32_PCIE_PORT(PCIE_LC_CNTL, data); + } + + orig = data = RREG32_PCIE(PCIE_CNTL2); + data |= SLV_MEM_LS_EN | MST_MEM_LS_EN | REPLAY_MEM_LS_EN; + if (orig != data) + WREG32_PCIE(PCIE_CNTL2, data); + + if (!disable_l0s) { + data = RREG32_PCIE_PORT(PCIE_LC_N_FTS_CNTL); + if((data & LC_N_FTS_MASK) == LC_N_FTS_MASK) { + data = RREG32_PCIE(PCIE_LC_STATUS1); + if ((data & LC_REVERSE_XMIT) && (data & LC_REVERSE_RCVR)) { + orig = data = RREG32_PCIE_PORT(PCIE_LC_CNTL); + data &= ~LC_L0S_INACTIVITY_MASK; + if (orig != data) + WREG32_PCIE_PORT(PCIE_LC_CNTL, data); + } + } + } +} + +static void si_fix_pci_max_read_req_size(struct amdgpu_device *adev) +{ + int readrq; + u16 v; + + readrq = pcie_get_readrq(adev->pdev); + v = ffs(readrq) - 8; + if ((v == 0) || (v == 6) || (v == 7)) + pcie_set_readrq(adev->pdev, 512); +} + +static int si_common_hw_init(void *handle) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + si_fix_pci_max_read_req_size(adev); + si_init_golden_registers(adev); + si_pcie_gen3_enable(adev); + si_program_aspm(adev); + + return 0; +} + +static int si_common_hw_fini(void *handle) +{ + return 0; +} + +static int si_common_suspend(void *handle) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + return si_common_hw_fini(adev); +} + +static int si_common_resume(void *handle) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + return si_common_hw_init(adev); +} + +static bool si_common_is_idle(void *handle) +{ + return true; +} + +static int si_common_wait_for_idle(void *handle) +{ + return 0; +} + +static int si_common_soft_reset(void *handle) +{ + return 0; +} + +static int si_common_set_clockgating_state(void *handle, + enum amd_clockgating_state state) +{ + return 0; +} + +static int si_common_set_powergating_state(void *handle, + enum amd_powergating_state state) +{ + return 0; +} + +const struct amd_ip_funcs si_common_ip_funcs = { + .name = "si_common", + .early_init = si_common_early_init, + .late_init = NULL, + .sw_init = si_common_sw_init, + .sw_fini = si_common_sw_fini, + .hw_init = si_common_hw_init, + .hw_fini = si_common_hw_fini, + .suspend = si_common_suspend, + .resume = si_common_resume, + .is_idle = si_common_is_idle, + .wait_for_idle = si_common_wait_for_idle, + .soft_reset = si_common_soft_reset, + .set_clockgating_state = si_common_set_clockgating_state, + .set_powergating_state = si_common_set_powergating_state, +}; + +static const struct amdgpu_ip_block_version verde_ip_blocks[] = +{ + { + .type = AMD_IP_BLOCK_TYPE_COMMON, + .major = 1, + .minor = 0, + .rev = 0, + .funcs = &si_common_ip_funcs, + }, + { + .type = AMD_IP_BLOCK_TYPE_GMC, + .major = 6, + .minor = 0, + .rev = 0, + .funcs = &gmc_v6_0_ip_funcs, + }, + { + .type = AMD_IP_BLOCK_TYPE_IH, + .major = 1, + .minor = 0, + .rev = 0, + .funcs = &si_ih_ip_funcs, + }, + { + .type = AMD_IP_BLOCK_TYPE_SMC, + .major = 6, + .minor = 0, + .rev = 0, + .funcs = &amdgpu_pp_ip_funcs, + }, + { + .type = AMD_IP_BLOCK_TYPE_DCE, + .major = 6, + .minor = 0, + .rev = 0, + .funcs = &dce_v6_0_ip_funcs, + }, + { + .type = AMD_IP_BLOCK_TYPE_GFX, + .major = 6, + .minor = 0, + .rev = 0, + .funcs = &gfx_v6_0_ip_funcs, + }, + { + .type = AMD_IP_BLOCK_TYPE_SDMA, + .major = 1, + .minor = 0, + .rev = 0, + .funcs = &si_dma_ip_funcs, + }, +/* { + .type = AMD_IP_BLOCK_TYPE_UVD, + .major = 3, + .minor = 1, + .rev = 0, + .funcs = &si_null_ip_funcs, + }, + { + .type = AMD_IP_BLOCK_TYPE_VCE, + .major = 1, + .minor = 0, + .rev = 0, + .funcs = &si_null_ip_funcs, + }, + */ +}; + + +static const struct amdgpu_ip_block_version hainan_ip_blocks[] = +{ + { + .type = AMD_IP_BLOCK_TYPE_COMMON, + .major = 1, + .minor = 0, + .rev = 0, + .funcs = &si_common_ip_funcs, + }, + { + .type = AMD_IP_BLOCK_TYPE_GMC, + .major = 6, + .minor = 0, + .rev = 0, + .funcs = &gmc_v6_0_ip_funcs, + }, + { + .type = AMD_IP_BLOCK_TYPE_IH, + .major = 1, + .minor = 0, + .rev = 0, + .funcs = &si_ih_ip_funcs, + }, + { + .type = AMD_IP_BLOCK_TYPE_SMC, + .major = 6, + .minor = 0, + .rev = 0, + .funcs = &amdgpu_pp_ip_funcs, + }, + { + .type = AMD_IP_BLOCK_TYPE_GFX, + .major = 6, + .minor = 0, + .rev = 0, + .funcs = &gfx_v6_0_ip_funcs, + }, + { + .type = AMD_IP_BLOCK_TYPE_SDMA, + .major = 1, + .minor = 0, + .rev = 0, + .funcs = &si_dma_ip_funcs, + }, +}; + +int si_set_ip_blocks(struct amdgpu_device *adev) +{ + switch (adev->asic_type) { + case CHIP_VERDE: + case CHIP_TAHITI: + case CHIP_PITCAIRN: + case CHIP_OLAND: + adev->ip_blocks = verde_ip_blocks; + adev->num_ip_blocks = ARRAY_SIZE(verde_ip_blocks); + break; + case CHIP_HAINAN: + adev->ip_blocks = hainan_ip_blocks; + adev->num_ip_blocks = ARRAY_SIZE(hainan_ip_blocks); + break; + default: + BUG(); + } + return 0; +} + diff --git a/drivers/gpu/drm/amd/amdgpu/si.h b/drivers/gpu/drm/amd/amdgpu/si.h new file mode 100644 index 000000000000..959d7b63e0e5 --- /dev/null +++ b/drivers/gpu/drm/amd/amdgpu/si.h @@ -0,0 +1,33 @@ +/* + * Copyright 2015 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + */ + +#ifndef __SI_H__ +#define __SI_H__ + +extern const struct amd_ip_funcs si_common_ip_funcs; + +void si_srbm_select(struct amdgpu_device *adev, + u32 me, u32 pipe, u32 queue, u32 vmid); +int si_set_ip_blocks(struct amdgpu_device *adev); + +#endif diff --git a/drivers/gpu/drm/amd/amdgpu/si_dma.c b/drivers/gpu/drm/amd/amdgpu/si_dma.c new file mode 100644 index 000000000000..de358193a8f9 --- /dev/null +++ b/drivers/gpu/drm/amd/amdgpu/si_dma.c @@ -0,0 +1,915 @@ +/* + * Copyright 2015 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + * Authors: Alex Deucher + */ +#include <drm/drmP.h> +#include "amdgpu.h" +#include "amdgpu_trace.h" +#include "si/sid.h" + +const u32 sdma_offsets[SDMA_MAX_INSTANCE] = +{ + DMA0_REGISTER_OFFSET, + DMA1_REGISTER_OFFSET +}; + +static void si_dma_set_ring_funcs(struct amdgpu_device *adev); +static void si_dma_set_buffer_funcs(struct amdgpu_device *adev); +static void si_dma_set_vm_pte_funcs(struct amdgpu_device *adev); +static void si_dma_set_irq_funcs(struct amdgpu_device *adev); + +static uint32_t si_dma_ring_get_rptr(struct amdgpu_ring *ring) +{ + return ring->adev->wb.wb[ring->rptr_offs>>2]; +} + +static uint32_t si_dma_ring_get_wptr(struct amdgpu_ring *ring) +{ + struct amdgpu_device *adev = ring->adev; + u32 me = (ring == &adev->sdma.instance[0].ring) ? 0 : 1; + + return (RREG32(DMA_RB_WPTR + sdma_offsets[me]) & 0x3fffc) >> 2; +} + +static void si_dma_ring_set_wptr(struct amdgpu_ring *ring) +{ + struct amdgpu_device *adev = ring->adev; + u32 me = (ring == &adev->sdma.instance[0].ring) ? 0 : 1; + + WREG32(DMA_RB_WPTR + sdma_offsets[me], (ring->wptr << 2) & 0x3fffc); +} + +static void si_dma_ring_emit_ib(struct amdgpu_ring *ring, + struct amdgpu_ib *ib, + unsigned vm_id, bool ctx_switch) +{ + /* The indirect buffer packet must end on an 8 DW boundary in the DMA ring. + * Pad as necessary with NOPs. + */ + while ((ring->wptr & 7) != 5) + amdgpu_ring_write(ring, DMA_PACKET(DMA_PACKET_NOP, 0, 0, 0, 0)); + amdgpu_ring_write(ring, DMA_IB_PACKET(DMA_PACKET_INDIRECT_BUFFER, vm_id, 0)); + amdgpu_ring_write(ring, (ib->gpu_addr & 0xFFFFFFE0)); + amdgpu_ring_write(ring, (ib->length_dw << 12) | (upper_32_bits(ib->gpu_addr) & 0xFF)); + +} + +static void si_dma_ring_emit_hdp_flush(struct amdgpu_ring *ring) +{ + amdgpu_ring_write(ring, DMA_PACKET(DMA_PACKET_SRBM_WRITE, 0, 0, 0, 0)); + amdgpu_ring_write(ring, (0xf << 16) | (HDP_MEM_COHERENCY_FLUSH_CNTL)); + amdgpu_ring_write(ring, 1); +} + +static void si_dma_ring_emit_hdp_invalidate(struct amdgpu_ring *ring) +{ + amdgpu_ring_write(ring, DMA_PACKET(DMA_PACKET_SRBM_WRITE, 0, 0, 0, 0)); + amdgpu_ring_write(ring, (0xf << 16) | (HDP_DEBUG0)); + amdgpu_ring_write(ring, 1); +} + +/** + * si_dma_ring_emit_fence - emit a fence on the DMA ring + * + * @ring: amdgpu ring pointer + * @fence: amdgpu fence object + * + * Add a DMA fence packet to the ring to write + * the fence seq number and DMA trap packet to generate + * an interrupt if needed (VI). + */ +static void si_dma_ring_emit_fence(struct amdgpu_ring *ring, u64 addr, u64 seq, + unsigned flags) +{ + + bool write64bit = flags & AMDGPU_FENCE_FLAG_64BIT; + /* write the fence */ + amdgpu_ring_write(ring, DMA_PACKET(DMA_PACKET_FENCE, 0, 0, 0, 0)); + amdgpu_ring_write(ring, addr & 0xfffffffc); + amdgpu_ring_write(ring, (upper_32_bits(addr) & 0xff)); + amdgpu_ring_write(ring, seq); + /* optionally write high bits as well */ + if (write64bit) { + addr += 4; + amdgpu_ring_write(ring, DMA_PACKET(DMA_PACKET_FENCE, 0, 0, 0, 0)); + amdgpu_ring_write(ring, addr & 0xfffffffc); + amdgpu_ring_write(ring, (upper_32_bits(addr) & 0xff)); + amdgpu_ring_write(ring, upper_32_bits(seq)); + } + /* generate an interrupt */ + amdgpu_ring_write(ring, DMA_PACKET(DMA_PACKET_TRAP, 0, 0, 0, 0)); +} + +static void si_dma_stop(struct amdgpu_device *adev) +{ + struct amdgpu_ring *ring; + u32 rb_cntl; + unsigned i; + + for (i = 0; i < adev->sdma.num_instances; i++) { + ring = &adev->sdma.instance[i].ring; + /* dma0 */ + rb_cntl = RREG32(DMA_RB_CNTL + sdma_offsets[i]); + rb_cntl &= ~DMA_RB_ENABLE; + WREG32(DMA_RB_CNTL + sdma_offsets[i], rb_cntl); + + if (adev->mman.buffer_funcs_ring == ring) + amdgpu_ttm_set_active_vram_size(adev, adev->mc.visible_vram_size); + ring->ready = false; + } +} + +static int si_dma_start(struct amdgpu_device *adev) +{ + struct amdgpu_ring *ring; + u32 rb_cntl, dma_cntl, ib_cntl, rb_bufsz; + int i, r; + uint64_t rptr_addr; + + for (i = 0; i < adev->sdma.num_instances; i++) { + ring = &adev->sdma.instance[i].ring; + + WREG32(DMA_SEM_INCOMPLETE_TIMER_CNTL + sdma_offsets[i], 0); + WREG32(DMA_SEM_WAIT_FAIL_TIMER_CNTL + sdma_offsets[i], 0); + + /* Set ring buffer size in dwords */ + rb_bufsz = order_base_2(ring->ring_size / 4); + rb_cntl = rb_bufsz << 1; +#ifdef __BIG_ENDIAN + rb_cntl |= DMA_RB_SWAP_ENABLE | DMA_RPTR_WRITEBACK_SWAP_ENABLE; +#endif + WREG32(DMA_RB_CNTL + sdma_offsets[i], rb_cntl); + + /* Initialize the ring buffer's read and write pointers */ + WREG32(DMA_RB_RPTR + sdma_offsets[i], 0); + WREG32(DMA_RB_WPTR + sdma_offsets[i], 0); + + rptr_addr = adev->wb.gpu_addr + (ring->rptr_offs * 4); + + WREG32(DMA_RB_RPTR_ADDR_LO + sdma_offsets[i], lower_32_bits(rptr_addr)); + WREG32(DMA_RB_RPTR_ADDR_HI + sdma_offsets[i], upper_32_bits(rptr_addr) & 0xFF); + + rb_cntl |= DMA_RPTR_WRITEBACK_ENABLE; + + WREG32(DMA_RB_BASE + sdma_offsets[i], ring->gpu_addr >> 8); + + /* enable DMA IBs */ + ib_cntl = DMA_IB_ENABLE | CMD_VMID_FORCE; +#ifdef __BIG_ENDIAN + ib_cntl |= DMA_IB_SWAP_ENABLE; +#endif + WREG32(DMA_IB_CNTL + sdma_offsets[i], ib_cntl); + + dma_cntl = RREG32(DMA_CNTL + sdma_offsets[i]); + dma_cntl &= ~CTXEMPTY_INT_ENABLE; + WREG32(DMA_CNTL + sdma_offsets[i], dma_cntl); + + ring->wptr = 0; + WREG32(DMA_RB_WPTR + sdma_offsets[i], ring->wptr << 2); + WREG32(DMA_RB_CNTL + sdma_offsets[i], rb_cntl | DMA_RB_ENABLE); + + ring->ready = true; + + r = amdgpu_ring_test_ring(ring); + if (r) { + ring->ready = false; + return r; + } + + if (adev->mman.buffer_funcs_ring == ring) + amdgpu_ttm_set_active_vram_size(adev, adev->mc.real_vram_size); + } + + return 0; +} + +/** + * si_dma_ring_test_ring - simple async dma engine test + * + * @ring: amdgpu_ring structure holding ring information + * + * Test the DMA engine by writing using it to write an + * value to memory. (VI). + * Returns 0 for success, error for failure. + */ +static int si_dma_ring_test_ring(struct amdgpu_ring *ring) +{ + struct amdgpu_device *adev = ring->adev; + unsigned i; + unsigned index; + int r; + u32 tmp; + u64 gpu_addr; + + r = amdgpu_wb_get(adev, &index); + if (r) { + dev_err(adev->dev, "(%d) failed to allocate wb slot\n", r); + return r; + } + + gpu_addr = adev->wb.gpu_addr + (index * 4); + tmp = 0xCAFEDEAD; + adev->wb.wb[index] = cpu_to_le32(tmp); + + r = amdgpu_ring_alloc(ring, 4); + if (r) { + DRM_ERROR("amdgpu: dma failed to lock ring %d (%d).\n", ring->idx, r); + amdgpu_wb_free(adev, index); + return r; + } + + amdgpu_ring_write(ring, DMA_PACKET(DMA_PACKET_WRITE, 0, 0, 0, 1)); + amdgpu_ring_write(ring, lower_32_bits(gpu_addr)); + amdgpu_ring_write(ring, upper_32_bits(gpu_addr) & 0xff); + amdgpu_ring_write(ring, 0xDEADBEEF); + amdgpu_ring_commit(ring); + + for (i = 0; i < adev->usec_timeout; i++) { + tmp = le32_to_cpu(adev->wb.wb[index]); + if (tmp == 0xDEADBEEF) + break; + DRM_UDELAY(1); + } + + if (i < adev->usec_timeout) { + DRM_INFO("ring test on %d succeeded in %d usecs\n", ring->idx, i); + } else { + DRM_ERROR("amdgpu: ring %d test failed (0x%08X)\n", + ring->idx, tmp); + r = -EINVAL; + } + amdgpu_wb_free(adev, index); + + return r; +} + +/** + * si_dma_ring_test_ib - test an IB on the DMA engine + * + * @ring: amdgpu_ring structure holding ring information + * + * Test a simple IB in the DMA ring (VI). + * Returns 0 on success, error on failure. + */ +static int si_dma_ring_test_ib(struct amdgpu_ring *ring, long timeout) +{ + struct amdgpu_device *adev = ring->adev; + struct amdgpu_ib ib; + struct fence *f = NULL; + unsigned index; + u32 tmp = 0; + u64 gpu_addr; + long r; + + r = amdgpu_wb_get(adev, &index); + if (r) { + dev_err(adev->dev, "(%ld) failed to allocate wb slot\n", r); + return r; + } + + gpu_addr = adev->wb.gpu_addr + (index * 4); + tmp = 0xCAFEDEAD; + adev->wb.wb[index] = cpu_to_le32(tmp); + memset(&ib, 0, sizeof(ib)); + r = amdgpu_ib_get(adev, NULL, 256, &ib); + if (r) { + DRM_ERROR("amdgpu: failed to get ib (%ld).\n", r); + goto err0; + } + + ib.ptr[0] = DMA_PACKET(DMA_PACKET_WRITE, 0, 0, 0, 1); + ib.ptr[1] = lower_32_bits(gpu_addr); + ib.ptr[2] = upper_32_bits(gpu_addr) & 0xff; + ib.ptr[3] = 0xDEADBEEF; + ib.length_dw = 4; + r = amdgpu_ib_schedule(ring, 1, &ib, NULL, NULL, &f); + if (r) + goto err1; + + r = fence_wait_timeout(f, false, timeout); + if (r == 0) { + DRM_ERROR("amdgpu: IB test timed out\n"); + r = -ETIMEDOUT; + goto err1; + } else if (r < 0) { + DRM_ERROR("amdgpu: fence wait failed (%ld).\n", r); + goto err1; + } + tmp = le32_to_cpu(adev->wb.wb[index]); + if (tmp == 0xDEADBEEF) { + DRM_INFO("ib test on ring %d succeeded\n", ring->idx); + r = 0; + } else { + DRM_ERROR("amdgpu: ib test failed (0x%08X)\n", tmp); + r = -EINVAL; + } + +err1: + amdgpu_ib_free(adev, &ib, NULL); + fence_put(f); +err0: + amdgpu_wb_free(adev, index); + return r; +} + +/** + * cik_dma_vm_copy_pte - update PTEs by copying them from the GART + * + * @ib: indirect buffer to fill with commands + * @pe: addr of the page entry + * @src: src addr to copy from + * @count: number of page entries to update + * + * Update PTEs by copying them from the GART using DMA (SI). + */ +static void si_dma_vm_copy_pte(struct amdgpu_ib *ib, + uint64_t pe, uint64_t src, + unsigned count) +{ + unsigned bytes = count * 8; + + ib->ptr[ib->length_dw++] = DMA_PACKET(DMA_PACKET_COPY, + 1, 0, 0, bytes); + ib->ptr[ib->length_dw++] = lower_32_bits(pe); + ib->ptr[ib->length_dw++] = lower_32_bits(src); + ib->ptr[ib->length_dw++] = upper_32_bits(pe) & 0xff; + ib->ptr[ib->length_dw++] = upper_32_bits(src) & 0xff; +} + +/** + * si_dma_vm_write_pte - update PTEs by writing them manually + * + * @ib: indirect buffer to fill with commands + * @pe: addr of the page entry + * @value: dst addr to write into pe + * @count: number of page entries to update + * @incr: increase next addr by incr bytes + * + * Update PTEs by writing them manually using DMA (SI). + */ +static void si_dma_vm_write_pte(struct amdgpu_ib *ib, uint64_t pe, + uint64_t value, unsigned count, + uint32_t incr) +{ + unsigned ndw = count * 2; + + ib->ptr[ib->length_dw++] = DMA_PACKET(DMA_PACKET_WRITE, 0, 0, 0, ndw); + ib->ptr[ib->length_dw++] = lower_32_bits(pe); + ib->ptr[ib->length_dw++] = upper_32_bits(pe); + for (; ndw > 0; ndw -= 2) { + ib->ptr[ib->length_dw++] = lower_32_bits(value); + ib->ptr[ib->length_dw++] = upper_32_bits(value); + value += incr; + } +} + +/** + * si_dma_vm_set_pte_pde - update the page tables using sDMA + * + * @ib: indirect buffer to fill with commands + * @pe: addr of the page entry + * @addr: dst addr to write into pe + * @count: number of page entries to update + * @incr: increase next addr by incr bytes + * @flags: access flags + * + * Update the page tables using sDMA (CIK). + */ +static void si_dma_vm_set_pte_pde(struct amdgpu_ib *ib, + uint64_t pe, + uint64_t addr, unsigned count, + uint32_t incr, uint32_t flags) +{ + uint64_t value; + unsigned ndw; + + while (count) { + ndw = count * 2; + if (ndw > 0xFFFFE) + ndw = 0xFFFFE; + + if (flags & AMDGPU_PTE_VALID) + value = addr; + else + value = 0; + + /* for physically contiguous pages (vram) */ + ib->ptr[ib->length_dw++] = DMA_PTE_PDE_PACKET(ndw); + ib->ptr[ib->length_dw++] = pe; /* dst addr */ + ib->ptr[ib->length_dw++] = upper_32_bits(pe) & 0xff; + ib->ptr[ib->length_dw++] = flags; /* mask */ + ib->ptr[ib->length_dw++] = 0; + ib->ptr[ib->length_dw++] = value; /* value */ + ib->ptr[ib->length_dw++] = upper_32_bits(value); + ib->ptr[ib->length_dw++] = incr; /* increment size */ + ib->ptr[ib->length_dw++] = 0; + pe += ndw * 4; + addr += (ndw / 2) * incr; + count -= ndw / 2; + } +} + +/** + * si_dma_pad_ib - pad the IB to the required number of dw + * + * @ib: indirect buffer to fill with padding + * + */ +static void si_dma_ring_pad_ib(struct amdgpu_ring *ring, struct amdgpu_ib *ib) +{ + while (ib->length_dw & 0x7) + ib->ptr[ib->length_dw++] = DMA_PACKET(DMA_PACKET_NOP, 0, 0, 0, 0); +} + +/** + * cik_sdma_ring_emit_pipeline_sync - sync the pipeline + * + * @ring: amdgpu_ring pointer + * + * Make sure all previous operations are completed (CIK). + */ +static void si_dma_ring_emit_pipeline_sync(struct amdgpu_ring *ring) +{ + uint32_t seq = ring->fence_drv.sync_seq; + uint64_t addr = ring->fence_drv.gpu_addr; + + /* wait for idle */ + amdgpu_ring_write(ring, DMA_PACKET(DMA_PACKET_POLL_REG_MEM, 0, 0, 0, 0) | + (1 << 27)); /* Poll memory */ + amdgpu_ring_write(ring, lower_32_bits(addr)); + amdgpu_ring_write(ring, (0xff << 16) | upper_32_bits(addr)); /* retry, addr_hi */ + amdgpu_ring_write(ring, 0xffffffff); /* mask */ + amdgpu_ring_write(ring, seq); /* value */ + amdgpu_ring_write(ring, (3 << 28) | 0x20); /* func(equal) | poll interval */ +} + +/** + * si_dma_ring_emit_vm_flush - cik vm flush using sDMA + * + * @ring: amdgpu_ring pointer + * @vm: amdgpu_vm pointer + * + * Update the page table base and flush the VM TLB + * using sDMA (VI). + */ +static void si_dma_ring_emit_vm_flush(struct amdgpu_ring *ring, + unsigned vm_id, uint64_t pd_addr) +{ + amdgpu_ring_write(ring, DMA_PACKET(DMA_PACKET_SRBM_WRITE, 0, 0, 0, 0)); + if (vm_id < 8) + amdgpu_ring_write(ring, (0xf << 16) | (VM_CONTEXT0_PAGE_TABLE_BASE_ADDR + vm_id)); + else + amdgpu_ring_write(ring, (0xf << 16) | (VM_CONTEXT8_PAGE_TABLE_BASE_ADDR + (vm_id - 8))); + amdgpu_ring_write(ring, pd_addr >> 12); + + /* bits 0-7 are the VM contexts0-7 */ + amdgpu_ring_write(ring, DMA_PACKET(DMA_PACKET_SRBM_WRITE, 0, 0, 0, 0)); + amdgpu_ring_write(ring, (0xf << 16) | (VM_INVALIDATE_REQUEST)); + amdgpu_ring_write(ring, 1 << vm_id); + + /* wait for invalidate to complete */ + amdgpu_ring_write(ring, DMA_PACKET(DMA_PACKET_POLL_REG_MEM, 0, 0, 0, 0)); + amdgpu_ring_write(ring, VM_INVALIDATE_REQUEST); + amdgpu_ring_write(ring, 0xff << 16); /* retry */ + amdgpu_ring_write(ring, 1 << vm_id); /* mask */ + amdgpu_ring_write(ring, 0); /* value */ + amdgpu_ring_write(ring, (0 << 28) | 0x20); /* func(always) | poll interval */ +} + +static unsigned si_dma_ring_get_emit_ib_size(struct amdgpu_ring *ring) +{ + return + 7 + 3; /* si_dma_ring_emit_ib */ +} + +static unsigned si_dma_ring_get_dma_frame_size(struct amdgpu_ring *ring) +{ + return + 3 + /* si_dma_ring_emit_hdp_flush */ + 3 + /* si_dma_ring_emit_hdp_invalidate */ + 6 + /* si_dma_ring_emit_pipeline_sync */ + 12 + /* si_dma_ring_emit_vm_flush */ + 9 + 9 + 9; /* si_dma_ring_emit_fence x3 for user fence, vm fence */ +} + +static int si_dma_early_init(void *handle) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + adev->sdma.num_instances = 2; + + si_dma_set_ring_funcs(adev); + si_dma_set_buffer_funcs(adev); + si_dma_set_vm_pte_funcs(adev); + si_dma_set_irq_funcs(adev); + + return 0; +} + +static int si_dma_sw_init(void *handle) +{ + struct amdgpu_ring *ring; + int r, i; + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + /* DMA0 trap event */ + r = amdgpu_irq_add_id(adev, 224, &adev->sdma.trap_irq); + if (r) + return r; + + /* DMA1 trap event */ + r = amdgpu_irq_add_id(adev, 244, &adev->sdma.trap_irq_1); + if (r) + return r; + + for (i = 0; i < adev->sdma.num_instances; i++) { + ring = &adev->sdma.instance[i].ring; + ring->ring_obj = NULL; + ring->use_doorbell = false; + sprintf(ring->name, "sdma%d", i); + r = amdgpu_ring_init(adev, ring, 1024, + DMA_PACKET(DMA_PACKET_NOP, 0, 0, 0, 0), 0xf, + &adev->sdma.trap_irq, + (i == 0) ? + AMDGPU_SDMA_IRQ_TRAP0 : AMDGPU_SDMA_IRQ_TRAP1, + AMDGPU_RING_TYPE_SDMA); + if (r) + return r; + } + + return r; +} + +static int si_dma_sw_fini(void *handle) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + int i; + + for (i = 0; i < adev->sdma.num_instances; i++) + amdgpu_ring_fini(&adev->sdma.instance[i].ring); + + return 0; +} + +static int si_dma_hw_init(void *handle) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + return si_dma_start(adev); +} + +static int si_dma_hw_fini(void *handle) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + si_dma_stop(adev); + + return 0; +} + +static int si_dma_suspend(void *handle) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + return si_dma_hw_fini(adev); +} + +static int si_dma_resume(void *handle) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + return si_dma_hw_init(adev); +} + +static bool si_dma_is_idle(void *handle) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + u32 tmp = RREG32(SRBM_STATUS2); + + if (tmp & (DMA_BUSY_MASK | DMA1_BUSY_MASK)) + return false; + + return true; +} + +static int si_dma_wait_for_idle(void *handle) +{ + unsigned i; + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + for (i = 0; i < adev->usec_timeout; i++) { + if (si_dma_is_idle(handle)) + return 0; + udelay(1); + } + return -ETIMEDOUT; +} + +static int si_dma_soft_reset(void *handle) +{ + DRM_INFO("si_dma_soft_reset --- not implemented !!!!!!!\n"); + return 0; +} + +static int si_dma_set_trap_irq_state(struct amdgpu_device *adev, + struct amdgpu_irq_src *src, + unsigned type, + enum amdgpu_interrupt_state state) +{ + u32 sdma_cntl; + + switch (type) { + case AMDGPU_SDMA_IRQ_TRAP0: + switch (state) { + case AMDGPU_IRQ_STATE_DISABLE: + sdma_cntl = RREG32(DMA_CNTL + DMA0_REGISTER_OFFSET); + sdma_cntl &= ~TRAP_ENABLE; + WREG32(DMA_CNTL + DMA0_REGISTER_OFFSET, sdma_cntl); + break; + case AMDGPU_IRQ_STATE_ENABLE: + sdma_cntl = RREG32(DMA_CNTL + DMA0_REGISTER_OFFSET); + sdma_cntl |= TRAP_ENABLE; + WREG32(DMA_CNTL + DMA0_REGISTER_OFFSET, sdma_cntl); + break; + default: + break; + } + break; + case AMDGPU_SDMA_IRQ_TRAP1: + switch (state) { + case AMDGPU_IRQ_STATE_DISABLE: + sdma_cntl = RREG32(DMA_CNTL + DMA1_REGISTER_OFFSET); + sdma_cntl &= ~TRAP_ENABLE; + WREG32(DMA_CNTL + DMA1_REGISTER_OFFSET, sdma_cntl); + break; + case AMDGPU_IRQ_STATE_ENABLE: + sdma_cntl = RREG32(DMA_CNTL + DMA1_REGISTER_OFFSET); + sdma_cntl |= TRAP_ENABLE; + WREG32(DMA_CNTL + DMA1_REGISTER_OFFSET, sdma_cntl); + break; + default: + break; + } + break; + default: + break; + } + return 0; +} + +static int si_dma_process_trap_irq(struct amdgpu_device *adev, + struct amdgpu_irq_src *source, + struct amdgpu_iv_entry *entry) +{ + amdgpu_fence_process(&adev->sdma.instance[0].ring); + + return 0; +} + +static int si_dma_process_trap_irq_1(struct amdgpu_device *adev, + struct amdgpu_irq_src *source, + struct amdgpu_iv_entry *entry) +{ + amdgpu_fence_process(&adev->sdma.instance[1].ring); + + return 0; +} + +static int si_dma_process_illegal_inst_irq(struct amdgpu_device *adev, + struct amdgpu_irq_src *source, + struct amdgpu_iv_entry *entry) +{ + DRM_ERROR("Illegal instruction in SDMA command stream\n"); + schedule_work(&adev->reset_work); + return 0; +} + +static int si_dma_set_clockgating_state(void *handle, + enum amd_clockgating_state state) +{ + u32 orig, data, offset; + int i; + bool enable; + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + enable = (state == AMD_CG_STATE_GATE) ? true : false; + + if (enable && (adev->cg_flags & AMD_CG_SUPPORT_SDMA_MGCG)) { + for (i = 0; i < adev->sdma.num_instances; i++) { + if (i == 0) + offset = DMA0_REGISTER_OFFSET; + else + offset = DMA1_REGISTER_OFFSET; + orig = data = RREG32(DMA_POWER_CNTL + offset); + data &= ~MEM_POWER_OVERRIDE; + if (data != orig) + WREG32(DMA_POWER_CNTL + offset, data); + WREG32(DMA_CLK_CTRL + offset, 0x00000100); + } + } else { + for (i = 0; i < adev->sdma.num_instances; i++) { + if (i == 0) + offset = DMA0_REGISTER_OFFSET; + else + offset = DMA1_REGISTER_OFFSET; + orig = data = RREG32(DMA_POWER_CNTL + offset); + data |= MEM_POWER_OVERRIDE; + if (data != orig) + WREG32(DMA_POWER_CNTL + offset, data); + + orig = data = RREG32(DMA_CLK_CTRL + offset); + data = 0xff000000; + if (data != orig) + WREG32(DMA_CLK_CTRL + offset, data); + } + } + + return 0; +} + +static int si_dma_set_powergating_state(void *handle, + enum amd_powergating_state state) +{ + u32 tmp; + + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + WREG32(DMA_PGFSM_WRITE, 0x00002000); + WREG32(DMA_PGFSM_CONFIG, 0x100010ff); + + for (tmp = 0; tmp < 5; tmp++) + WREG32(DMA_PGFSM_WRITE, 0); + + return 0; +} + +const struct amd_ip_funcs si_dma_ip_funcs = { + .name = "si_dma", + .early_init = si_dma_early_init, + .late_init = NULL, + .sw_init = si_dma_sw_init, + .sw_fini = si_dma_sw_fini, + .hw_init = si_dma_hw_init, + .hw_fini = si_dma_hw_fini, + .suspend = si_dma_suspend, + .resume = si_dma_resume, + .is_idle = si_dma_is_idle, + .wait_for_idle = si_dma_wait_for_idle, + .soft_reset = si_dma_soft_reset, + .set_clockgating_state = si_dma_set_clockgating_state, + .set_powergating_state = si_dma_set_powergating_state, +}; + +static const struct amdgpu_ring_funcs si_dma_ring_funcs = { + .get_rptr = si_dma_ring_get_rptr, + .get_wptr = si_dma_ring_get_wptr, + .set_wptr = si_dma_ring_set_wptr, + .parse_cs = NULL, + .emit_ib = si_dma_ring_emit_ib, + .emit_fence = si_dma_ring_emit_fence, + .emit_pipeline_sync = si_dma_ring_emit_pipeline_sync, + .emit_vm_flush = si_dma_ring_emit_vm_flush, + .emit_hdp_flush = si_dma_ring_emit_hdp_flush, + .emit_hdp_invalidate = si_dma_ring_emit_hdp_invalidate, + .test_ring = si_dma_ring_test_ring, + .test_ib = si_dma_ring_test_ib, + .insert_nop = amdgpu_ring_insert_nop, + .pad_ib = si_dma_ring_pad_ib, + .get_emit_ib_size = si_dma_ring_get_emit_ib_size, + .get_dma_frame_size = si_dma_ring_get_dma_frame_size, +}; + +static void si_dma_set_ring_funcs(struct amdgpu_device *adev) +{ + int i; + + for (i = 0; i < adev->sdma.num_instances; i++) + adev->sdma.instance[i].ring.funcs = &si_dma_ring_funcs; +} + +static const struct amdgpu_irq_src_funcs si_dma_trap_irq_funcs = { + .set = si_dma_set_trap_irq_state, + .process = si_dma_process_trap_irq, +}; + +static const struct amdgpu_irq_src_funcs si_dma_trap_irq_funcs_1 = { + .set = si_dma_set_trap_irq_state, + .process = si_dma_process_trap_irq_1, +}; + +static const struct amdgpu_irq_src_funcs si_dma_illegal_inst_irq_funcs = { + .process = si_dma_process_illegal_inst_irq, +}; + +static void si_dma_set_irq_funcs(struct amdgpu_device *adev) +{ + adev->sdma.trap_irq.num_types = AMDGPU_SDMA_IRQ_LAST; + adev->sdma.trap_irq.funcs = &si_dma_trap_irq_funcs; + adev->sdma.trap_irq_1.funcs = &si_dma_trap_irq_funcs_1; + adev->sdma.illegal_inst_irq.funcs = &si_dma_illegal_inst_irq_funcs; +} + +/** + * si_dma_emit_copy_buffer - copy buffer using the sDMA engine + * + * @ring: amdgpu_ring structure holding ring information + * @src_offset: src GPU address + * @dst_offset: dst GPU address + * @byte_count: number of bytes to xfer + * + * Copy GPU buffers using the DMA engine (VI). + * Used by the amdgpu ttm implementation to move pages if + * registered as the asic copy callback. + */ +static void si_dma_emit_copy_buffer(struct amdgpu_ib *ib, + uint64_t src_offset, + uint64_t dst_offset, + uint32_t byte_count) +{ + ib->ptr[ib->length_dw++] = DMA_PACKET(DMA_PACKET_COPY, + 1, 0, 0, byte_count); + ib->ptr[ib->length_dw++] = lower_32_bits(dst_offset); + ib->ptr[ib->length_dw++] = lower_32_bits(src_offset); + ib->ptr[ib->length_dw++] = upper_32_bits(dst_offset) & 0xff; + ib->ptr[ib->length_dw++] = upper_32_bits(src_offset) & 0xff; +} + +/** + * si_dma_emit_fill_buffer - fill buffer using the sDMA engine + * + * @ring: amdgpu_ring structure holding ring information + * @src_data: value to write to buffer + * @dst_offset: dst GPU address + * @byte_count: number of bytes to xfer + * + * Fill GPU buffers using the DMA engine (VI). + */ +static void si_dma_emit_fill_buffer(struct amdgpu_ib *ib, + uint32_t src_data, + uint64_t dst_offset, + uint32_t byte_count) +{ + ib->ptr[ib->length_dw++] = DMA_PACKET(DMA_PACKET_CONSTANT_FILL, + 0, 0, 0, byte_count / 4); + ib->ptr[ib->length_dw++] = lower_32_bits(dst_offset); + ib->ptr[ib->length_dw++] = src_data; + ib->ptr[ib->length_dw++] = upper_32_bits(dst_offset) << 16; +} + + +static const struct amdgpu_buffer_funcs si_dma_buffer_funcs = { + .copy_max_bytes = 0xffff8, + .copy_num_dw = 5, + .emit_copy_buffer = si_dma_emit_copy_buffer, + + .fill_max_bytes = 0xffff8, + .fill_num_dw = 4, + .emit_fill_buffer = si_dma_emit_fill_buffer, +}; + +static void si_dma_set_buffer_funcs(struct amdgpu_device *adev) +{ + if (adev->mman.buffer_funcs == NULL) { + adev->mman.buffer_funcs = &si_dma_buffer_funcs; + adev->mman.buffer_funcs_ring = &adev->sdma.instance[0].ring; + } +} + +static const struct amdgpu_vm_pte_funcs si_dma_vm_pte_funcs = { + .copy_pte = si_dma_vm_copy_pte, + .write_pte = si_dma_vm_write_pte, + .set_pte_pde = si_dma_vm_set_pte_pde, +}; + +static void si_dma_set_vm_pte_funcs(struct amdgpu_device *adev) +{ + unsigned i; + + if (adev->vm_manager.vm_pte_funcs == NULL) { + adev->vm_manager.vm_pte_funcs = &si_dma_vm_pte_funcs; + for (i = 0; i < adev->sdma.num_instances; i++) + adev->vm_manager.vm_pte_rings[i] = + &adev->sdma.instance[i].ring; + + adev->vm_manager.vm_pte_num_rings = adev->sdma.num_instances; + } +} diff --git a/drivers/gpu/drm/amd/amdgpu/si_dma.h b/drivers/gpu/drm/amd/amdgpu/si_dma.h new file mode 100644 index 000000000000..3a3e0c78a54b --- /dev/null +++ b/drivers/gpu/drm/amd/amdgpu/si_dma.h @@ -0,0 +1,29 @@ +/* + * Copyright 2015 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + */ + +#ifndef __SI_DMA_H__ +#define __SI_DMA_H__ + +extern const struct amd_ip_funcs si_dma_ip_funcs; + +#endif diff --git a/drivers/gpu/drm/amd/amdgpu/si_dpm.c b/drivers/gpu/drm/amd/amdgpu/si_dpm.c new file mode 100644 index 000000000000..e2db4a734676 --- /dev/null +++ b/drivers/gpu/drm/amd/amdgpu/si_dpm.c @@ -0,0 +1,7993 @@ +/* + * Copyright 2013 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + */ + +#include "drmP.h" +#include "amdgpu.h" +#include "amdgpu_pm.h" +#include "amdgpu_dpm.h" +#include "amdgpu_atombios.h" +#include "si/sid.h" +#include "r600_dpm.h" +#include "si_dpm.h" +#include "atom.h" +#include "../include/pptable.h" +#include <linux/math64.h> +#include <linux/seq_file.h> +#include <linux/firmware.h> + +#define MC_CG_ARB_FREQ_F0 0x0a +#define MC_CG_ARB_FREQ_F1 0x0b +#define MC_CG_ARB_FREQ_F2 0x0c +#define MC_CG_ARB_FREQ_F3 0x0d + +#define SMC_RAM_END 0x20000 + +#define SCLK_MIN_DEEPSLEEP_FREQ 1350 + + +/* sizeof(ATOM_PPLIB_EXTENDEDHEADER) */ +#define SIZE_OF_ATOM_PPLIB_EXTENDEDHEADER_V2 12 +#define SIZE_OF_ATOM_PPLIB_EXTENDEDHEADER_V3 14 +#define SIZE_OF_ATOM_PPLIB_EXTENDEDHEADER_V4 16 +#define SIZE_OF_ATOM_PPLIB_EXTENDEDHEADER_V5 18 +#define SIZE_OF_ATOM_PPLIB_EXTENDEDHEADER_V6 20 +#define SIZE_OF_ATOM_PPLIB_EXTENDEDHEADER_V7 22 + +#define BIOS_SCRATCH_4 0x5cd + +MODULE_FIRMWARE("radeon/tahiti_smc.bin"); +MODULE_FIRMWARE("radeon/tahiti_k_smc.bin"); +MODULE_FIRMWARE("radeon/pitcairn_smc.bin"); +MODULE_FIRMWARE("radeon/pitcairn_k_smc.bin"); +MODULE_FIRMWARE("radeon/verde_smc.bin"); +MODULE_FIRMWARE("radeon/verde_k_smc.bin"); +MODULE_FIRMWARE("radeon/oland_smc.bin"); +MODULE_FIRMWARE("radeon/oland_k_smc.bin"); +MODULE_FIRMWARE("radeon/hainan_smc.bin"); +MODULE_FIRMWARE("radeon/hainan_k_smc.bin"); + +union power_info { + struct _ATOM_POWERPLAY_INFO info; + struct _ATOM_POWERPLAY_INFO_V2 info_2; + struct _ATOM_POWERPLAY_INFO_V3 info_3; + struct _ATOM_PPLIB_POWERPLAYTABLE pplib; + struct _ATOM_PPLIB_POWERPLAYTABLE2 pplib2; + struct _ATOM_PPLIB_POWERPLAYTABLE3 pplib3; + struct _ATOM_PPLIB_POWERPLAYTABLE4 pplib4; + struct _ATOM_PPLIB_POWERPLAYTABLE5 pplib5; +}; + +union fan_info { + struct _ATOM_PPLIB_FANTABLE fan; + struct _ATOM_PPLIB_FANTABLE2 fan2; + struct _ATOM_PPLIB_FANTABLE3 fan3; +}; + +union pplib_clock_info { + struct _ATOM_PPLIB_R600_CLOCK_INFO r600; + struct _ATOM_PPLIB_RS780_CLOCK_INFO rs780; + struct _ATOM_PPLIB_EVERGREEN_CLOCK_INFO evergreen; + struct _ATOM_PPLIB_SUMO_CLOCK_INFO sumo; + struct _ATOM_PPLIB_SI_CLOCK_INFO si; +}; + +static const u32 r600_utc[R600_PM_NUMBER_OF_TC] = +{ + R600_UTC_DFLT_00, + R600_UTC_DFLT_01, + R600_UTC_DFLT_02, + R600_UTC_DFLT_03, + R600_UTC_DFLT_04, + R600_UTC_DFLT_05, + R600_UTC_DFLT_06, + R600_UTC_DFLT_07, + R600_UTC_DFLT_08, + R600_UTC_DFLT_09, + R600_UTC_DFLT_10, + R600_UTC_DFLT_11, + R600_UTC_DFLT_12, + R600_UTC_DFLT_13, + R600_UTC_DFLT_14, +}; + +static const u32 r600_dtc[R600_PM_NUMBER_OF_TC] = +{ + R600_DTC_DFLT_00, + R600_DTC_DFLT_01, + R600_DTC_DFLT_02, + R600_DTC_DFLT_03, + R600_DTC_DFLT_04, + R600_DTC_DFLT_05, + R600_DTC_DFLT_06, + R600_DTC_DFLT_07, + R600_DTC_DFLT_08, + R600_DTC_DFLT_09, + R600_DTC_DFLT_10, + R600_DTC_DFLT_11, + R600_DTC_DFLT_12, + R600_DTC_DFLT_13, + R600_DTC_DFLT_14, +}; + +static const struct si_cac_config_reg cac_weights_tahiti[] = +{ + { 0x0, 0x0000ffff, 0, 0xc, SISLANDS_CACCONFIG_CGIND }, + { 0x0, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x1, 0x0000ffff, 0, 0x101, SISLANDS_CACCONFIG_CGIND }, + { 0x1, 0xffff0000, 16, 0xc, SISLANDS_CACCONFIG_CGIND }, + { 0x2, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x3, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x3, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x4, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x4, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x5, 0x0000ffff, 0, 0x8fc, SISLANDS_CACCONFIG_CGIND }, + { 0x5, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x6, 0x0000ffff, 0, 0x95, SISLANDS_CACCONFIG_CGIND }, + { 0x6, 0xffff0000, 16, 0x34e, SISLANDS_CACCONFIG_CGIND }, + { 0x18f, 0x0000ffff, 0, 0x1a1, SISLANDS_CACCONFIG_CGIND }, + { 0x7, 0x0000ffff, 0, 0xda, SISLANDS_CACCONFIG_CGIND }, + { 0x7, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x8, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x8, 0xffff0000, 16, 0x46, SISLANDS_CACCONFIG_CGIND }, + { 0x9, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0xa, 0x0000ffff, 0, 0x208, SISLANDS_CACCONFIG_CGIND }, + { 0xb, 0x0000ffff, 0, 0xe7, SISLANDS_CACCONFIG_CGIND }, + { 0xb, 0xffff0000, 16, 0x948, SISLANDS_CACCONFIG_CGIND }, + { 0xc, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0xd, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0xd, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0xe, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0xf, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0xf, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x10, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x10, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x11, 0x0000ffff, 0, 0x167, SISLANDS_CACCONFIG_CGIND }, + { 0x11, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x12, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x13, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x13, 0xffff0000, 16, 0x35, SISLANDS_CACCONFIG_CGIND }, + { 0x14, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x15, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x15, 0xffff0000, 16, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x4e, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x16, 0x0000ffff, 0, 0x31, SISLANDS_CACCONFIG_CGIND }, + { 0x16, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x17, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x18, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x18, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x19, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x19, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x1a, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x1a, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x1b, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x1b, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x1c, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x1c, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x1d, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x1d, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x1e, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x1e, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x1f, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x1f, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x20, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x6d, 0x0000ffff, 0, 0x18e, SISLANDS_CACCONFIG_CGIND }, + { 0xFFFFFFFF } +}; + +static const struct si_cac_config_reg lcac_tahiti[] = +{ + { 0x143, 0x0001fffe, 1, 0x3, SISLANDS_CACCONFIG_CGIND }, + { 0x143, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x146, 0x0001fffe, 1, 0x3, SISLANDS_CACCONFIG_CGIND }, + { 0x146, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x149, 0x0001fffe, 1, 0x3, SISLANDS_CACCONFIG_CGIND }, + { 0x149, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x14c, 0x0001fffe, 1, 0x3, SISLANDS_CACCONFIG_CGIND }, + { 0x14c, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x98, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x98, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x9b, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x9b, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x9e, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x9e, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x101, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x101, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x104, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x104, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x107, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x107, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x10a, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x10a, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x10d, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x10d, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x8c, 0x0001fffe, 1, 0x8, SISLANDS_CACCONFIG_CGIND }, + { 0x8c, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x8f, 0x0001fffe, 1, 0x8, SISLANDS_CACCONFIG_CGIND }, + { 0x8f, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x92, 0x0001fffe, 1, 0x8, SISLANDS_CACCONFIG_CGIND }, + { 0x92, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x95, 0x0001fffe, 1, 0x8, SISLANDS_CACCONFIG_CGIND }, + { 0x95, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x14f, 0x0001fffe, 1, 0x8, SISLANDS_CACCONFIG_CGIND }, + { 0x14f, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x152, 0x0001fffe, 1, 0x8, SISLANDS_CACCONFIG_CGIND }, + { 0x152, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x155, 0x0001fffe, 1, 0x8, SISLANDS_CACCONFIG_CGIND }, + { 0x155, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x158, 0x0001fffe, 1, 0x8, SISLANDS_CACCONFIG_CGIND }, + { 0x158, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x110, 0x0001fffe, 1, 0x8, SISLANDS_CACCONFIG_CGIND }, + { 0x110, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x113, 0x0001fffe, 1, 0x8, SISLANDS_CACCONFIG_CGIND }, + { 0x113, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x116, 0x0001fffe, 1, 0x8, SISLANDS_CACCONFIG_CGIND }, + { 0x116, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x119, 0x0001fffe, 1, 0x8, SISLANDS_CACCONFIG_CGIND }, + { 0x119, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x11c, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x11c, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x11f, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x11f, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x122, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x122, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x125, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x125, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x128, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x128, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x12b, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x12b, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x15b, 0x0001fffe, 1, 0x4, SISLANDS_CACCONFIG_CGIND }, + { 0x15b, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x15e, 0x0001fffe, 1, 0x4, SISLANDS_CACCONFIG_CGIND }, + { 0x15e, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x161, 0x0001fffe, 1, 0x4, SISLANDS_CACCONFIG_CGIND }, + { 0x161, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x164, 0x0001fffe, 1, 0x4, SISLANDS_CACCONFIG_CGIND }, + { 0x164, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x167, 0x0001fffe, 1, 0x4, SISLANDS_CACCONFIG_CGIND }, + { 0x167, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x16a, 0x0001fffe, 1, 0x4, SISLANDS_CACCONFIG_CGIND }, + { 0x16a, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x16d, 0x0001fffe, 1, 0x6, SISLANDS_CACCONFIG_CGIND }, + { 0x16d, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x170, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x170, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x173, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x173, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x176, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x176, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x179, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x179, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x17c, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x17c, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x17f, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x17f, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0xFFFFFFFF } + +}; + +static const struct si_cac_config_reg cac_override_tahiti[] = +{ + { 0xFFFFFFFF } +}; + +static const struct si_powertune_data powertune_data_tahiti = +{ + ((1 << 16) | 27027), + 6, + 0, + 4, + 95, + { + 0UL, + 0UL, + 4521550UL, + 309631529UL, + -1270850L, + 4513710L, + 40 + }, + 595000000UL, + 12, + { + 0, + 0, + 0, + 0, + 0, + 0, + 0, + 0 + }, + true +}; + +static const struct si_dte_data dte_data_tahiti = +{ + { 1159409, 0, 0, 0, 0 }, + { 777, 0, 0, 0, 0 }, + 2, + 54000, + 127000, + 25, + 2, + 10, + 13, + { 27, 31, 35, 39, 43, 47, 54, 61, 67, 74, 81, 88, 95, 0, 0, 0 }, + { 240888759, 221057860, 235370597, 162287531, 158510299, 131423027, 116673180, 103067515, 87941937, 76209048, 68209175, 64090048, 58301890, 0, 0, 0 }, + { 12024, 11189, 11451, 8411, 7939, 6666, 5681, 4905, 4241, 3720, 3354, 3122, 2890, 0, 0, 0 }, + 85, + false +}; + +#if 0 +static const struct si_dte_data dte_data_tahiti_le = +{ + { 0x1E8480, 0x7A1200, 0x2160EC0, 0x3938700, 0 }, + { 0x7D, 0x7D, 0x4E4, 0xB00, 0 }, + 0x5, + 0xAFC8, + 0x64, + 0x32, + 1, + 0, + 0x10, + { 0x78, 0x7C, 0x82, 0x88, 0x8E, 0x94, 0x9A, 0xA0, 0xA6, 0xAC, 0xB0, 0xB4, 0xB8, 0xBC, 0xC0, 0xC4 }, + { 0x3938700, 0x3938700, 0x3938700, 0x3938700, 0x3938700, 0x3938700, 0x3938700, 0x3938700, 0x3938700, 0x3938700, 0x3938700, 0x3938700, 0x3938700, 0x3938700, 0x3938700, 0x3938700 }, + { 0x2AF8, 0x2AF8, 0x29BB, 0x27F9, 0x2637, 0x2475, 0x22B3, 0x20F1, 0x1F2F, 0x1D6D, 0x1734, 0x1414, 0x10F4, 0xDD4, 0xAB4, 0x794 }, + 85, + true +}; +#endif + +static const struct si_dte_data dte_data_tahiti_pro = +{ + { 0x1E8480, 0x3D0900, 0x989680, 0x2625A00, 0x0 }, + { 0x0, 0x0, 0x0, 0x0, 0x0 }, + 5, + 45000, + 100, + 0xA, + 1, + 0, + 0x10, + { 0x96, 0xB4, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF }, + { 0x895440, 0x3D0900, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680 }, + { 0x7D0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0 }, + 90, + true +}; + +static const struct si_dte_data dte_data_new_zealand = +{ + { 0x1E8480, 0x3D0900, 0x989680, 0x2625A00, 0 }, + { 0x29B, 0x3E9, 0x537, 0x7D2, 0 }, + 0x5, + 0xAFC8, + 0x69, + 0x32, + 1, + 0, + 0x10, + { 0x82, 0xA0, 0xB4, 0xFE, 0xFE, 0xFE, 0xFE, 0xFE, 0xFE, 0xFE, 0xFE, 0xFE, 0xFE, 0xFE, 0xFE, 0xFE }, + { 0x895440, 0x3D0900, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680 }, + { 0xDAC, 0x1388, 0x685, 0x685, 0x685, 0x685, 0x685, 0x685, 0x685, 0x685, 0x685, 0x685, 0x685, 0x685, 0x685, 0x685 }, + 85, + true +}; + +static const struct si_dte_data dte_data_aruba_pro = +{ + { 0x1E8480, 0x3D0900, 0x989680, 0x2625A00, 0x0 }, + { 0x0, 0x0, 0x0, 0x0, 0x0 }, + 5, + 45000, + 100, + 0xA, + 1, + 0, + 0x10, + { 0x96, 0xB4, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF }, + { 0x895440, 0x3D0900, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680 }, + { 0x1000, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0 }, + 90, + true +}; + +static const struct si_dte_data dte_data_malta = +{ + { 0x1E8480, 0x3D0900, 0x989680, 0x2625A00, 0x0 }, + { 0x0, 0x0, 0x0, 0x0, 0x0 }, + 5, + 45000, + 100, + 0xA, + 1, + 0, + 0x10, + { 0x96, 0xB4, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF }, + { 0x895440, 0x3D0900, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680 }, + { 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0 }, + 90, + true +}; + +static const struct si_cac_config_reg cac_weights_pitcairn[] = +{ + { 0x0, 0x0000ffff, 0, 0x8a, SISLANDS_CACCONFIG_CGIND }, + { 0x0, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x1, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x1, 0xffff0000, 16, 0x24d, SISLANDS_CACCONFIG_CGIND }, + { 0x2, 0x0000ffff, 0, 0x19, SISLANDS_CACCONFIG_CGIND }, + { 0x3, 0x0000ffff, 0, 0x118, SISLANDS_CACCONFIG_CGIND }, + { 0x3, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x4, 0x0000ffff, 0, 0x76, SISLANDS_CACCONFIG_CGIND }, + { 0x4, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x5, 0x0000ffff, 0, 0xc11, SISLANDS_CACCONFIG_CGIND }, + { 0x5, 0xffff0000, 16, 0x7f3, SISLANDS_CACCONFIG_CGIND }, + { 0x6, 0x0000ffff, 0, 0x403, SISLANDS_CACCONFIG_CGIND }, + { 0x6, 0xffff0000, 16, 0x367, SISLANDS_CACCONFIG_CGIND }, + { 0x18f, 0x0000ffff, 0, 0x4c9, SISLANDS_CACCONFIG_CGIND }, + { 0x7, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x7, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x8, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x8, 0xffff0000, 16, 0x45d, SISLANDS_CACCONFIG_CGIND }, + { 0x9, 0x0000ffff, 0, 0x36d, SISLANDS_CACCONFIG_CGIND }, + { 0xa, 0x0000ffff, 0, 0x534, SISLANDS_CACCONFIG_CGIND }, + { 0xb, 0x0000ffff, 0, 0x5da, SISLANDS_CACCONFIG_CGIND }, + { 0xb, 0xffff0000, 16, 0x880, SISLANDS_CACCONFIG_CGIND }, + { 0xc, 0x0000ffff, 0, 0x201, SISLANDS_CACCONFIG_CGIND }, + { 0xd, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0xd, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0xe, 0x0000ffff, 0, 0x9f, SISLANDS_CACCONFIG_CGIND }, + { 0xf, 0x0000ffff, 0, 0x1f, SISLANDS_CACCONFIG_CGIND }, + { 0xf, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x10, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x10, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x11, 0x0000ffff, 0, 0x5de, SISLANDS_CACCONFIG_CGIND }, + { 0x11, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x12, 0x0000ffff, 0, 0x7b, SISLANDS_CACCONFIG_CGIND }, + { 0x13, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x13, 0xffff0000, 16, 0x13, SISLANDS_CACCONFIG_CGIND }, + { 0x14, 0x0000ffff, 0, 0xf9, SISLANDS_CACCONFIG_CGIND }, + { 0x15, 0x0000ffff, 0, 0x66, SISLANDS_CACCONFIG_CGIND }, + { 0x15, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x4e, 0x0000ffff, 0, 0x13, SISLANDS_CACCONFIG_CGIND }, + { 0x16, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x16, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x17, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x18, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x18, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x19, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x19, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x1a, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x1a, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x1b, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x1b, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x1c, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x1c, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x1d, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x1d, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x1e, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x1e, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x1f, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x1f, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x20, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x6d, 0x0000ffff, 0, 0x186, SISLANDS_CACCONFIG_CGIND }, + { 0xFFFFFFFF } +}; + +static const struct si_cac_config_reg lcac_pitcairn[] = +{ + { 0x98, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x98, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x104, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x104, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x110, 0x0001fffe, 1, 0x5, SISLANDS_CACCONFIG_CGIND }, + { 0x110, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x14f, 0x0001fffe, 1, 0x5, SISLANDS_CACCONFIG_CGIND }, + { 0x14f, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x8c, 0x0001fffe, 1, 0x5, SISLANDS_CACCONFIG_CGIND }, + { 0x8c, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x143, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x143, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x9b, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x9b, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x107, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x107, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x113, 0x0001fffe, 1, 0x5, SISLANDS_CACCONFIG_CGIND }, + { 0x113, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x152, 0x0001fffe, 1, 0x5, SISLANDS_CACCONFIG_CGIND }, + { 0x152, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x8f, 0x0001fffe, 1, 0x5, SISLANDS_CACCONFIG_CGIND }, + { 0x8f, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x146, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x146, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x9e, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x9e, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x10a, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x10a, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x116, 0x0001fffe, 1, 0x5, SISLANDS_CACCONFIG_CGIND }, + { 0x116, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x155, 0x0001fffe, 1, 0x5, SISLANDS_CACCONFIG_CGIND }, + { 0x155, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x92, 0x0001fffe, 1, 0x5, SISLANDS_CACCONFIG_CGIND }, + { 0x92, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x149, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x149, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x101, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x101, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x10d, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x10d, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x119, 0x0001fffe, 1, 0x5, SISLANDS_CACCONFIG_CGIND }, + { 0x119, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x158, 0x0001fffe, 1, 0x5, SISLANDS_CACCONFIG_CGIND }, + { 0x158, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x95, 0x0001fffe, 1, 0x5, SISLANDS_CACCONFIG_CGIND }, + { 0x95, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x14c, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x14c, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x11c, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x11c, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x11f, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x11f, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x122, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x122, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x125, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x125, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x128, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x128, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x12b, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x12b, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x164, 0x0001fffe, 1, 0x4, SISLANDS_CACCONFIG_CGIND }, + { 0x164, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x167, 0x0001fffe, 1, 0x4, SISLANDS_CACCONFIG_CGIND }, + { 0x167, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x16a, 0x0001fffe, 1, 0x4, SISLANDS_CACCONFIG_CGIND }, + { 0x16a, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x15e, 0x0001fffe, 1, 0x4, SISLANDS_CACCONFIG_CGIND }, + { 0x15e, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x161, 0x0001fffe, 1, 0x4, SISLANDS_CACCONFIG_CGIND }, + { 0x161, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x15b, 0x0001fffe, 1, 0x4, SISLANDS_CACCONFIG_CGIND }, + { 0x15b, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x16d, 0x0001fffe, 1, 0x4, SISLANDS_CACCONFIG_CGIND }, + { 0x16d, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x170, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x170, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x173, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x173, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x176, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x176, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x179, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x179, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x17c, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x17c, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x17f, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x17f, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0xFFFFFFFF } +}; + +static const struct si_cac_config_reg cac_override_pitcairn[] = +{ + { 0xFFFFFFFF } +}; + +static const struct si_powertune_data powertune_data_pitcairn = +{ + ((1 << 16) | 27027), + 5, + 0, + 6, + 100, + { + 51600000UL, + 1800000UL, + 7194395UL, + 309631529UL, + -1270850L, + 4513710L, + 100 + }, + 117830498UL, + 12, + { + 0, + 0, + 0, + 0, + 0, + 0, + 0, + 0 + }, + true +}; + +static const struct si_dte_data dte_data_pitcairn = +{ + { 0, 0, 0, 0, 0 }, + { 0, 0, 0, 0, 0 }, + 0, + 0, + 0, + 0, + 0, + 0, + 0, + { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 }, + { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 }, + { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 }, + 0, + false +}; + +static const struct si_dte_data dte_data_curacao_xt = +{ + { 0x1E8480, 0x3D0900, 0x989680, 0x2625A00, 0x0 }, + { 0x0, 0x0, 0x0, 0x0, 0x0 }, + 5, + 45000, + 100, + 0xA, + 1, + 0, + 0x10, + { 0x96, 0xB4, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF }, + { 0x895440, 0x3D0900, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680 }, + { 0x1D17, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0 }, + 90, + true +}; + +static const struct si_dte_data dte_data_curacao_pro = +{ + { 0x1E8480, 0x3D0900, 0x989680, 0x2625A00, 0x0 }, + { 0x0, 0x0, 0x0, 0x0, 0x0 }, + 5, + 45000, + 100, + 0xA, + 1, + 0, + 0x10, + { 0x96, 0xB4, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF }, + { 0x895440, 0x3D0900, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680 }, + { 0x1D17, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0 }, + 90, + true +}; + +static const struct si_dte_data dte_data_neptune_xt = +{ + { 0x1E8480, 0x3D0900, 0x989680, 0x2625A00, 0x0 }, + { 0x0, 0x0, 0x0, 0x0, 0x0 }, + 5, + 45000, + 100, + 0xA, + 1, + 0, + 0x10, + { 0x96, 0xB4, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF }, + { 0x895440, 0x3D0900, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680 }, + { 0x3A2F, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0 }, + 90, + true +}; + +static const struct si_cac_config_reg cac_weights_chelsea_pro[] = +{ + { 0x0, 0x0000ffff, 0, 0x82, SISLANDS_CACCONFIG_CGIND }, + { 0x0, 0xffff0000, 16, 0x4F, SISLANDS_CACCONFIG_CGIND }, + { 0x1, 0x0000ffff, 0, 0x153, SISLANDS_CACCONFIG_CGIND }, + { 0x1, 0xffff0000, 16, 0x52, SISLANDS_CACCONFIG_CGIND }, + { 0x2, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x3, 0x0000ffff, 0, 0x135, SISLANDS_CACCONFIG_CGIND }, + { 0x3, 0xffff0000, 16, 0x4F, SISLANDS_CACCONFIG_CGIND }, + { 0x4, 0x0000ffff, 0, 0x135, SISLANDS_CACCONFIG_CGIND }, + { 0x4, 0xffff0000, 16, 0xAC, SISLANDS_CACCONFIG_CGIND }, + { 0x5, 0x0000ffff, 0, 0x118, SISLANDS_CACCONFIG_CGIND }, + { 0x5, 0xffff0000, 16, 0xBE, SISLANDS_CACCONFIG_CGIND }, + { 0x6, 0x0000ffff, 0, 0x110, SISLANDS_CACCONFIG_CGIND }, + { 0x6, 0xffff0000, 16, 0x4CD, SISLANDS_CACCONFIG_CGIND }, + { 0x18f, 0x0000ffff, 0, 0x30, SISLANDS_CACCONFIG_CGIND }, + { 0x7, 0x0000ffff, 0, 0x37, SISLANDS_CACCONFIG_CGIND }, + { 0x7, 0xffff0000, 16, 0x27, SISLANDS_CACCONFIG_CGIND }, + { 0x8, 0x0000ffff, 0, 0xC3, SISLANDS_CACCONFIG_CGIND }, + { 0x8, 0xffff0000, 16, 0x35, SISLANDS_CACCONFIG_CGIND }, + { 0x9, 0x0000ffff, 0, 0x28, SISLANDS_CACCONFIG_CGIND }, + { 0xa, 0x0000ffff, 0, 0x26C, SISLANDS_CACCONFIG_CGIND }, + { 0xb, 0x0000ffff, 0, 0x3B2, SISLANDS_CACCONFIG_CGIND }, + { 0xb, 0xffff0000, 16, 0x99D, SISLANDS_CACCONFIG_CGIND }, + { 0xc, 0x0000ffff, 0, 0xA3F, SISLANDS_CACCONFIG_CGIND }, + { 0xd, 0x0000ffff, 0, 0xA, SISLANDS_CACCONFIG_CGIND }, + { 0xd, 0xffff0000, 16, 0xA, SISLANDS_CACCONFIG_CGIND }, + { 0xe, 0x0000ffff, 0, 0x5, SISLANDS_CACCONFIG_CGIND }, + { 0xf, 0x0000ffff, 0, 0x3, SISLANDS_CACCONFIG_CGIND }, + { 0xf, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x10, 0x0000ffff, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x10, 0xffff0000, 16, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x11, 0x0000ffff, 0, 0x5, SISLANDS_CACCONFIG_CGIND }, + { 0x11, 0xffff0000, 16, 0x15, SISLANDS_CACCONFIG_CGIND }, + { 0x12, 0x0000ffff, 0, 0x34, SISLANDS_CACCONFIG_CGIND }, + { 0x13, 0x0000ffff, 0, 0x4, SISLANDS_CACCONFIG_CGIND }, + { 0x13, 0xffff0000, 16, 0x4, SISLANDS_CACCONFIG_CGIND }, + { 0x14, 0x0000ffff, 0, 0x2BD, SISLANDS_CACCONFIG_CGIND }, + { 0x15, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x15, 0xffff0000, 16, 0x6, SISLANDS_CACCONFIG_CGIND }, + { 0x4e, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x16, 0x0000ffff, 0, 0x30, SISLANDS_CACCONFIG_CGIND }, + { 0x16, 0xffff0000, 16, 0x7A, SISLANDS_CACCONFIG_CGIND }, + { 0x17, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x18, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x18, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x19, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x19, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x1a, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1a, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1b, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1b, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1c, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1c, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1d, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1d, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1e, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1e, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1f, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1f, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x20, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x6d, 0x0000ffff, 0, 0x100, SISLANDS_CACCONFIG_CGIND }, + { 0xFFFFFFFF } +}; + +static const struct si_cac_config_reg cac_weights_chelsea_xt[] = +{ + { 0x0, 0x0000ffff, 0, 0x82, SISLANDS_CACCONFIG_CGIND }, + { 0x0, 0xffff0000, 16, 0x4F, SISLANDS_CACCONFIG_CGIND }, + { 0x1, 0x0000ffff, 0, 0x153, SISLANDS_CACCONFIG_CGIND }, + { 0x1, 0xffff0000, 16, 0x52, SISLANDS_CACCONFIG_CGIND }, + { 0x2, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x3, 0x0000ffff, 0, 0x135, SISLANDS_CACCONFIG_CGIND }, + { 0x3, 0xffff0000, 16, 0x4F, SISLANDS_CACCONFIG_CGIND }, + { 0x4, 0x0000ffff, 0, 0x135, SISLANDS_CACCONFIG_CGIND }, + { 0x4, 0xffff0000, 16, 0xAC, SISLANDS_CACCONFIG_CGIND }, + { 0x5, 0x0000ffff, 0, 0x118, SISLANDS_CACCONFIG_CGIND }, + { 0x5, 0xffff0000, 16, 0xBE, SISLANDS_CACCONFIG_CGIND }, + { 0x6, 0x0000ffff, 0, 0x110, SISLANDS_CACCONFIG_CGIND }, + { 0x6, 0xffff0000, 16, 0x4CD, SISLANDS_CACCONFIG_CGIND }, + { 0x18f, 0x0000ffff, 0, 0x30, SISLANDS_CACCONFIG_CGIND }, + { 0x7, 0x0000ffff, 0, 0x37, SISLANDS_CACCONFIG_CGIND }, + { 0x7, 0xffff0000, 16, 0x27, SISLANDS_CACCONFIG_CGIND }, + { 0x8, 0x0000ffff, 0, 0xC3, SISLANDS_CACCONFIG_CGIND }, + { 0x8, 0xffff0000, 16, 0x35, SISLANDS_CACCONFIG_CGIND }, + { 0x9, 0x0000ffff, 0, 0x28, SISLANDS_CACCONFIG_CGIND }, + { 0xa, 0x0000ffff, 0, 0x26C, SISLANDS_CACCONFIG_CGIND }, + { 0xb, 0x0000ffff, 0, 0x3B2, SISLANDS_CACCONFIG_CGIND }, + { 0xb, 0xffff0000, 16, 0x99D, SISLANDS_CACCONFIG_CGIND }, + { 0xc, 0x0000ffff, 0, 0xA3F, SISLANDS_CACCONFIG_CGIND }, + { 0xd, 0x0000ffff, 0, 0xA, SISLANDS_CACCONFIG_CGIND }, + { 0xd, 0xffff0000, 16, 0xA, SISLANDS_CACCONFIG_CGIND }, + { 0xe, 0x0000ffff, 0, 0x5, SISLANDS_CACCONFIG_CGIND }, + { 0xf, 0x0000ffff, 0, 0x3, SISLANDS_CACCONFIG_CGIND }, + { 0xf, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x10, 0x0000ffff, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x10, 0xffff0000, 16, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x11, 0x0000ffff, 0, 0x5, SISLANDS_CACCONFIG_CGIND }, + { 0x11, 0xffff0000, 16, 0x15, SISLANDS_CACCONFIG_CGIND }, + { 0x12, 0x0000ffff, 0, 0x34, SISLANDS_CACCONFIG_CGIND }, + { 0x13, 0x0000ffff, 0, 0x4, SISLANDS_CACCONFIG_CGIND }, + { 0x13, 0xffff0000, 16, 0x4, SISLANDS_CACCONFIG_CGIND }, + { 0x14, 0x0000ffff, 0, 0x30A, SISLANDS_CACCONFIG_CGIND }, + { 0x15, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x15, 0xffff0000, 16, 0x6, SISLANDS_CACCONFIG_CGIND }, + { 0x4e, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x16, 0x0000ffff, 0, 0x30, SISLANDS_CACCONFIG_CGIND }, + { 0x16, 0xffff0000, 16, 0x7A, SISLANDS_CACCONFIG_CGIND }, + { 0x17, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x18, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x18, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x19, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x19, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x1a, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1a, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1b, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1b, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1c, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1c, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1d, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1d, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1e, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1e, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1f, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1f, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x20, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x6d, 0x0000ffff, 0, 0x100, SISLANDS_CACCONFIG_CGIND }, + { 0xFFFFFFFF } +}; + +static const struct si_cac_config_reg cac_weights_heathrow[] = +{ + { 0x0, 0x0000ffff, 0, 0x82, SISLANDS_CACCONFIG_CGIND }, + { 0x0, 0xffff0000, 16, 0x4F, SISLANDS_CACCONFIG_CGIND }, + { 0x1, 0x0000ffff, 0, 0x153, SISLANDS_CACCONFIG_CGIND }, + { 0x1, 0xffff0000, 16, 0x52, SISLANDS_CACCONFIG_CGIND }, + { 0x2, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x3, 0x0000ffff, 0, 0x135, SISLANDS_CACCONFIG_CGIND }, + { 0x3, 0xffff0000, 16, 0x4F, SISLANDS_CACCONFIG_CGIND }, + { 0x4, 0x0000ffff, 0, 0x135, SISLANDS_CACCONFIG_CGIND }, + { 0x4, 0xffff0000, 16, 0xAC, SISLANDS_CACCONFIG_CGIND }, + { 0x5, 0x0000ffff, 0, 0x118, SISLANDS_CACCONFIG_CGIND }, + { 0x5, 0xffff0000, 16, 0xBE, SISLANDS_CACCONFIG_CGIND }, + { 0x6, 0x0000ffff, 0, 0x110, SISLANDS_CACCONFIG_CGIND }, + { 0x6, 0xffff0000, 16, 0x4CD, SISLANDS_CACCONFIG_CGIND }, + { 0x18f, 0x0000ffff, 0, 0x30, SISLANDS_CACCONFIG_CGIND }, + { 0x7, 0x0000ffff, 0, 0x37, SISLANDS_CACCONFIG_CGIND }, + { 0x7, 0xffff0000, 16, 0x27, SISLANDS_CACCONFIG_CGIND }, + { 0x8, 0x0000ffff, 0, 0xC3, SISLANDS_CACCONFIG_CGIND }, + { 0x8, 0xffff0000, 16, 0x35, SISLANDS_CACCONFIG_CGIND }, + { 0x9, 0x0000ffff, 0, 0x28, SISLANDS_CACCONFIG_CGIND }, + { 0xa, 0x0000ffff, 0, 0x26C, SISLANDS_CACCONFIG_CGIND }, + { 0xb, 0x0000ffff, 0, 0x3B2, SISLANDS_CACCONFIG_CGIND }, + { 0xb, 0xffff0000, 16, 0x99D, SISLANDS_CACCONFIG_CGIND }, + { 0xc, 0x0000ffff, 0, 0xA3F, SISLANDS_CACCONFIG_CGIND }, + { 0xd, 0x0000ffff, 0, 0xA, SISLANDS_CACCONFIG_CGIND }, + { 0xd, 0xffff0000, 16, 0xA, SISLANDS_CACCONFIG_CGIND }, + { 0xe, 0x0000ffff, 0, 0x5, SISLANDS_CACCONFIG_CGIND }, + { 0xf, 0x0000ffff, 0, 0x3, SISLANDS_CACCONFIG_CGIND }, + { 0xf, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x10, 0x0000ffff, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x10, 0xffff0000, 16, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x11, 0x0000ffff, 0, 0x5, SISLANDS_CACCONFIG_CGIND }, + { 0x11, 0xffff0000, 16, 0x15, SISLANDS_CACCONFIG_CGIND }, + { 0x12, 0x0000ffff, 0, 0x34, SISLANDS_CACCONFIG_CGIND }, + { 0x13, 0x0000ffff, 0, 0x4, SISLANDS_CACCONFIG_CGIND }, + { 0x13, 0xffff0000, 16, 0x4, SISLANDS_CACCONFIG_CGIND }, + { 0x14, 0x0000ffff, 0, 0x362, SISLANDS_CACCONFIG_CGIND }, + { 0x15, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x15, 0xffff0000, 16, 0x6, SISLANDS_CACCONFIG_CGIND }, + { 0x4e, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x16, 0x0000ffff, 0, 0x30, SISLANDS_CACCONFIG_CGIND }, + { 0x16, 0xffff0000, 16, 0x7A, SISLANDS_CACCONFIG_CGIND }, + { 0x17, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x18, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x18, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x19, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x19, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x1a, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1a, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1b, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1b, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1c, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1c, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1d, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1d, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1e, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1e, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1f, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1f, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x20, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x6d, 0x0000ffff, 0, 0x100, SISLANDS_CACCONFIG_CGIND }, + { 0xFFFFFFFF } +}; + +static const struct si_cac_config_reg cac_weights_cape_verde_pro[] = +{ + { 0x0, 0x0000ffff, 0, 0x82, SISLANDS_CACCONFIG_CGIND }, + { 0x0, 0xffff0000, 16, 0x4F, SISLANDS_CACCONFIG_CGIND }, + { 0x1, 0x0000ffff, 0, 0x153, SISLANDS_CACCONFIG_CGIND }, + { 0x1, 0xffff0000, 16, 0x52, SISLANDS_CACCONFIG_CGIND }, + { 0x2, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x3, 0x0000ffff, 0, 0x135, SISLANDS_CACCONFIG_CGIND }, + { 0x3, 0xffff0000, 16, 0x4F, SISLANDS_CACCONFIG_CGIND }, + { 0x4, 0x0000ffff, 0, 0x135, SISLANDS_CACCONFIG_CGIND }, + { 0x4, 0xffff0000, 16, 0xAC, SISLANDS_CACCONFIG_CGIND }, + { 0x5, 0x0000ffff, 0, 0x118, SISLANDS_CACCONFIG_CGIND }, + { 0x5, 0xffff0000, 16, 0xBE, SISLANDS_CACCONFIG_CGIND }, + { 0x6, 0x0000ffff, 0, 0x110, SISLANDS_CACCONFIG_CGIND }, + { 0x6, 0xffff0000, 16, 0x4CD, SISLANDS_CACCONFIG_CGIND }, + { 0x18f, 0x0000ffff, 0, 0x30, SISLANDS_CACCONFIG_CGIND }, + { 0x7, 0x0000ffff, 0, 0x37, SISLANDS_CACCONFIG_CGIND }, + { 0x7, 0xffff0000, 16, 0x27, SISLANDS_CACCONFIG_CGIND }, + { 0x8, 0x0000ffff, 0, 0xC3, SISLANDS_CACCONFIG_CGIND }, + { 0x8, 0xffff0000, 16, 0x35, SISLANDS_CACCONFIG_CGIND }, + { 0x9, 0x0000ffff, 0, 0x28, SISLANDS_CACCONFIG_CGIND }, + { 0xa, 0x0000ffff, 0, 0x26C, SISLANDS_CACCONFIG_CGIND }, + { 0xb, 0x0000ffff, 0, 0x3B2, SISLANDS_CACCONFIG_CGIND }, + { 0xb, 0xffff0000, 16, 0x99D, SISLANDS_CACCONFIG_CGIND }, + { 0xc, 0x0000ffff, 0, 0xA3F, SISLANDS_CACCONFIG_CGIND }, + { 0xd, 0x0000ffff, 0, 0xA, SISLANDS_CACCONFIG_CGIND }, + { 0xd, 0xffff0000, 16, 0xA, SISLANDS_CACCONFIG_CGIND }, + { 0xe, 0x0000ffff, 0, 0x5, SISLANDS_CACCONFIG_CGIND }, + { 0xf, 0x0000ffff, 0, 0x3, SISLANDS_CACCONFIG_CGIND }, + { 0xf, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x10, 0x0000ffff, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x10, 0xffff0000, 16, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x11, 0x0000ffff, 0, 0x5, SISLANDS_CACCONFIG_CGIND }, + { 0x11, 0xffff0000, 16, 0x15, SISLANDS_CACCONFIG_CGIND }, + { 0x12, 0x0000ffff, 0, 0x34, SISLANDS_CACCONFIG_CGIND }, + { 0x13, 0x0000ffff, 0, 0x4, SISLANDS_CACCONFIG_CGIND }, + { 0x13, 0xffff0000, 16, 0x4, SISLANDS_CACCONFIG_CGIND }, + { 0x14, 0x0000ffff, 0, 0x315, SISLANDS_CACCONFIG_CGIND }, + { 0x15, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x15, 0xffff0000, 16, 0x6, SISLANDS_CACCONFIG_CGIND }, + { 0x4e, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x16, 0x0000ffff, 0, 0x30, SISLANDS_CACCONFIG_CGIND }, + { 0x16, 0xffff0000, 16, 0x7A, SISLANDS_CACCONFIG_CGIND }, + { 0x17, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x18, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x18, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x19, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x19, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x1a, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1a, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1b, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1b, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1c, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1c, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1d, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1d, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1e, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1e, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1f, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1f, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x20, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x6d, 0x0000ffff, 0, 0x100, SISLANDS_CACCONFIG_CGIND }, + { 0xFFFFFFFF } +}; + +static const struct si_cac_config_reg cac_weights_cape_verde[] = +{ + { 0x0, 0x0000ffff, 0, 0x82, SISLANDS_CACCONFIG_CGIND }, + { 0x0, 0xffff0000, 16, 0x4F, SISLANDS_CACCONFIG_CGIND }, + { 0x1, 0x0000ffff, 0, 0x153, SISLANDS_CACCONFIG_CGIND }, + { 0x1, 0xffff0000, 16, 0x52, SISLANDS_CACCONFIG_CGIND }, + { 0x2, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x3, 0x0000ffff, 0, 0x135, SISLANDS_CACCONFIG_CGIND }, + { 0x3, 0xffff0000, 16, 0x4F, SISLANDS_CACCONFIG_CGIND }, + { 0x4, 0x0000ffff, 0, 0x135, SISLANDS_CACCONFIG_CGIND }, + { 0x4, 0xffff0000, 16, 0xAC, SISLANDS_CACCONFIG_CGIND }, + { 0x5, 0x0000ffff, 0, 0x118, SISLANDS_CACCONFIG_CGIND }, + { 0x5, 0xffff0000, 16, 0xBE, SISLANDS_CACCONFIG_CGIND }, + { 0x6, 0x0000ffff, 0, 0x110, SISLANDS_CACCONFIG_CGIND }, + { 0x6, 0xffff0000, 16, 0x4CD, SISLANDS_CACCONFIG_CGIND }, + { 0x18f, 0x0000ffff, 0, 0x30, SISLANDS_CACCONFIG_CGIND }, + { 0x7, 0x0000ffff, 0, 0x37, SISLANDS_CACCONFIG_CGIND }, + { 0x7, 0xffff0000, 16, 0x27, SISLANDS_CACCONFIG_CGIND }, + { 0x8, 0x0000ffff, 0, 0xC3, SISLANDS_CACCONFIG_CGIND }, + { 0x8, 0xffff0000, 16, 0x35, SISLANDS_CACCONFIG_CGIND }, + { 0x9, 0x0000ffff, 0, 0x28, SISLANDS_CACCONFIG_CGIND }, + { 0xa, 0x0000ffff, 0, 0x26C, SISLANDS_CACCONFIG_CGIND }, + { 0xb, 0x0000ffff, 0, 0x3B2, SISLANDS_CACCONFIG_CGIND }, + { 0xb, 0xffff0000, 16, 0x99D, SISLANDS_CACCONFIG_CGIND }, + { 0xc, 0x0000ffff, 0, 0xA3F, SISLANDS_CACCONFIG_CGIND }, + { 0xd, 0x0000ffff, 0, 0xA, SISLANDS_CACCONFIG_CGIND }, + { 0xd, 0xffff0000, 16, 0xA, SISLANDS_CACCONFIG_CGIND }, + { 0xe, 0x0000ffff, 0, 0x5, SISLANDS_CACCONFIG_CGIND }, + { 0xf, 0x0000ffff, 0, 0x3, SISLANDS_CACCONFIG_CGIND }, + { 0xf, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x10, 0x0000ffff, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x10, 0xffff0000, 16, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x11, 0x0000ffff, 0, 0x5, SISLANDS_CACCONFIG_CGIND }, + { 0x11, 0xffff0000, 16, 0x15, SISLANDS_CACCONFIG_CGIND }, + { 0x12, 0x0000ffff, 0, 0x34, SISLANDS_CACCONFIG_CGIND }, + { 0x13, 0x0000ffff, 0, 0x4, SISLANDS_CACCONFIG_CGIND }, + { 0x13, 0xffff0000, 16, 0x4, SISLANDS_CACCONFIG_CGIND }, + { 0x14, 0x0000ffff, 0, 0x3BA, SISLANDS_CACCONFIG_CGIND }, + { 0x15, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x15, 0xffff0000, 16, 0x6, SISLANDS_CACCONFIG_CGIND }, + { 0x4e, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x16, 0x0000ffff, 0, 0x30, SISLANDS_CACCONFIG_CGIND }, + { 0x16, 0xffff0000, 16, 0x7A, SISLANDS_CACCONFIG_CGIND }, + { 0x17, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x18, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x18, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x19, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x19, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x1a, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1a, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1b, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1b, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1c, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1c, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1d, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1d, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1e, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1e, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1f, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1f, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x20, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x6d, 0x0000ffff, 0, 0x100, SISLANDS_CACCONFIG_CGIND }, + { 0xFFFFFFFF } +}; + +static const struct si_cac_config_reg lcac_cape_verde[] = +{ + { 0x98, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x98, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x104, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x104, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x110, 0x0001fffe, 1, 0x5, SISLANDS_CACCONFIG_CGIND }, + { 0x110, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x14f, 0x0001fffe, 1, 0x5, SISLANDS_CACCONFIG_CGIND }, + { 0x14f, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x8c, 0x0001fffe, 1, 0x5, SISLANDS_CACCONFIG_CGIND }, + { 0x8c, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x143, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x143, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x9b, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x9b, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x107, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x107, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x113, 0x0001fffe, 1, 0x5, SISLANDS_CACCONFIG_CGIND }, + { 0x113, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x152, 0x0001fffe, 1, 0x5, SISLANDS_CACCONFIG_CGIND }, + { 0x152, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x8f, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x8f, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x146, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x146, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x11c, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x11c, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x11f, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x11f, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x164, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x164, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x167, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x167, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x16a, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x16a, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x15e, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x15e, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x161, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x161, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x15b, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x15b, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x16d, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x16d, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x170, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x170, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x173, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x173, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x176, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x176, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x179, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x179, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x17c, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x17c, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x17f, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x17f, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0xFFFFFFFF } +}; + +static const struct si_cac_config_reg cac_override_cape_verde[] = +{ + { 0xFFFFFFFF } +}; + +static const struct si_powertune_data powertune_data_cape_verde = +{ + ((1 << 16) | 0x6993), + 5, + 0, + 7, + 105, + { + 0UL, + 0UL, + 7194395UL, + 309631529UL, + -1270850L, + 4513710L, + 100 + }, + 117830498UL, + 12, + { + 0, + 0, + 0, + 0, + 0, + 0, + 0, + 0 + }, + true +}; + +static const struct si_dte_data dte_data_cape_verde = +{ + { 0, 0, 0, 0, 0 }, + { 0, 0, 0, 0, 0 }, + 0, + 0, + 0, + 0, + 0, + 0, + 0, + { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 }, + { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 }, + { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 }, + 0, + false +}; + +static const struct si_dte_data dte_data_venus_xtx = +{ + { 0x1E8480, 0x3D0900, 0x989680, 0x2625A00, 0x0 }, + { 0x71C, 0xAAB, 0xE39, 0x11C7, 0x0 }, + 5, + 55000, + 0x69, + 0xA, + 1, + 0, + 0x3, + { 0x96, 0xB4, 0xFF, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0 }, + { 0x895440, 0x3D0900, 0x989680, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0 }, + { 0xD6D8, 0x88B8, 0x1555, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0 }, + 90, + true +}; + +static const struct si_dte_data dte_data_venus_xt = +{ + { 0x1E8480, 0x3D0900, 0x989680, 0x2625A00, 0x0 }, + { 0xBDA, 0x11C7, 0x17B4, 0x1DA1, 0x0 }, + 5, + 55000, + 0x69, + 0xA, + 1, + 0, + 0x3, + { 0x96, 0xB4, 0xFF, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0 }, + { 0x895440, 0x3D0900, 0x989680, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0 }, + { 0xAFC8, 0x88B8, 0x238E, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0 }, + 90, + true +}; + +static const struct si_dte_data dte_data_venus_pro = +{ + { 0x1E8480, 0x3D0900, 0x989680, 0x2625A00, 0x0 }, + { 0x11C7, 0x1AAB, 0x238E, 0x2C72, 0x0 }, + 5, + 55000, + 0x69, + 0xA, + 1, + 0, + 0x3, + { 0x96, 0xB4, 0xFF, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0 }, + { 0x895440, 0x3D0900, 0x989680, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0 }, + { 0x88B8, 0x88B8, 0x3555, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0 }, + 90, + true +}; + +static const struct si_cac_config_reg cac_weights_oland[] = +{ + { 0x0, 0x0000ffff, 0, 0x82, SISLANDS_CACCONFIG_CGIND }, + { 0x0, 0xffff0000, 16, 0x4F, SISLANDS_CACCONFIG_CGIND }, + { 0x1, 0x0000ffff, 0, 0x153, SISLANDS_CACCONFIG_CGIND }, + { 0x1, 0xffff0000, 16, 0x52, SISLANDS_CACCONFIG_CGIND }, + { 0x2, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x3, 0x0000ffff, 0, 0x135, SISLANDS_CACCONFIG_CGIND }, + { 0x3, 0xffff0000, 16, 0x4F, SISLANDS_CACCONFIG_CGIND }, + { 0x4, 0x0000ffff, 0, 0x135, SISLANDS_CACCONFIG_CGIND }, + { 0x4, 0xffff0000, 16, 0xAC, SISLANDS_CACCONFIG_CGIND }, + { 0x5, 0x0000ffff, 0, 0x118, SISLANDS_CACCONFIG_CGIND }, + { 0x5, 0xffff0000, 16, 0xBE, SISLANDS_CACCONFIG_CGIND }, + { 0x6, 0x0000ffff, 0, 0x110, SISLANDS_CACCONFIG_CGIND }, + { 0x6, 0xffff0000, 16, 0x4CD, SISLANDS_CACCONFIG_CGIND }, + { 0x18f, 0x0000ffff, 0, 0x30, SISLANDS_CACCONFIG_CGIND }, + { 0x7, 0x0000ffff, 0, 0x37, SISLANDS_CACCONFIG_CGIND }, + { 0x7, 0xffff0000, 16, 0x27, SISLANDS_CACCONFIG_CGIND }, + { 0x8, 0x0000ffff, 0, 0xC3, SISLANDS_CACCONFIG_CGIND }, + { 0x8, 0xffff0000, 16, 0x35, SISLANDS_CACCONFIG_CGIND }, + { 0x9, 0x0000ffff, 0, 0x28, SISLANDS_CACCONFIG_CGIND }, + { 0xa, 0x0000ffff, 0, 0x26C, SISLANDS_CACCONFIG_CGIND }, + { 0xb, 0x0000ffff, 0, 0x3B2, SISLANDS_CACCONFIG_CGIND }, + { 0xb, 0xffff0000, 16, 0x99D, SISLANDS_CACCONFIG_CGIND }, + { 0xc, 0x0000ffff, 0, 0xA3F, SISLANDS_CACCONFIG_CGIND }, + { 0xd, 0x0000ffff, 0, 0xA, SISLANDS_CACCONFIG_CGIND }, + { 0xd, 0xffff0000, 16, 0xA, SISLANDS_CACCONFIG_CGIND }, + { 0xe, 0x0000ffff, 0, 0x5, SISLANDS_CACCONFIG_CGIND }, + { 0xf, 0x0000ffff, 0, 0x3, SISLANDS_CACCONFIG_CGIND }, + { 0xf, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x10, 0x0000ffff, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x10, 0xffff0000, 16, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x11, 0x0000ffff, 0, 0x5, SISLANDS_CACCONFIG_CGIND }, + { 0x11, 0xffff0000, 16, 0x15, SISLANDS_CACCONFIG_CGIND }, + { 0x12, 0x0000ffff, 0, 0x34, SISLANDS_CACCONFIG_CGIND }, + { 0x13, 0x0000ffff, 0, 0x4, SISLANDS_CACCONFIG_CGIND }, + { 0x13, 0xffff0000, 16, 0x4, SISLANDS_CACCONFIG_CGIND }, + { 0x14, 0x0000ffff, 0, 0x3BA, SISLANDS_CACCONFIG_CGIND }, + { 0x15, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x15, 0xffff0000, 16, 0x6, SISLANDS_CACCONFIG_CGIND }, + { 0x4e, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x16, 0x0000ffff, 0, 0x30, SISLANDS_CACCONFIG_CGIND }, + { 0x16, 0xffff0000, 16, 0x7A, SISLANDS_CACCONFIG_CGIND }, + { 0x17, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x18, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x18, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x19, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x19, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x1a, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1a, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1b, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1b, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1c, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1c, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1d, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1d, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1e, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1e, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1f, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1f, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x20, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x6d, 0x0000ffff, 0, 0x100, SISLANDS_CACCONFIG_CGIND }, + { 0xFFFFFFFF } +}; + +static const struct si_cac_config_reg cac_weights_mars_pro[] = +{ + { 0x0, 0x0000ffff, 0, 0x43, SISLANDS_CACCONFIG_CGIND }, + { 0x0, 0xffff0000, 16, 0x29, SISLANDS_CACCONFIG_CGIND }, + { 0x1, 0x0000ffff, 0, 0xAF, SISLANDS_CACCONFIG_CGIND }, + { 0x1, 0xffff0000, 16, 0x2A, SISLANDS_CACCONFIG_CGIND }, + { 0x2, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x3, 0x0000ffff, 0, 0xA0, SISLANDS_CACCONFIG_CGIND }, + { 0x3, 0xffff0000, 16, 0x29, SISLANDS_CACCONFIG_CGIND }, + { 0x4, 0x0000ffff, 0, 0xA0, SISLANDS_CACCONFIG_CGIND }, + { 0x4, 0xffff0000, 16, 0x59, SISLANDS_CACCONFIG_CGIND }, + { 0x5, 0x0000ffff, 0, 0x1A5, SISLANDS_CACCONFIG_CGIND }, + { 0x5, 0xffff0000, 16, 0x1D6, SISLANDS_CACCONFIG_CGIND }, + { 0x6, 0x0000ffff, 0, 0x2A3, SISLANDS_CACCONFIG_CGIND }, + { 0x6, 0xffff0000, 16, 0x8FD, SISLANDS_CACCONFIG_CGIND }, + { 0x18f, 0x0000ffff, 0, 0x76, SISLANDS_CACCONFIG_CGIND }, + { 0x7, 0x0000ffff, 0, 0x8A, SISLANDS_CACCONFIG_CGIND }, + { 0x7, 0xffff0000, 16, 0xA3, SISLANDS_CACCONFIG_CGIND }, + { 0x8, 0x0000ffff, 0, 0x71, SISLANDS_CACCONFIG_CGIND }, + { 0x8, 0xffff0000, 16, 0x36, SISLANDS_CACCONFIG_CGIND }, + { 0x9, 0x0000ffff, 0, 0xA6, SISLANDS_CACCONFIG_CGIND }, + { 0xa, 0x0000ffff, 0, 0x81, SISLANDS_CACCONFIG_CGIND }, + { 0xb, 0x0000ffff, 0, 0x3D2, SISLANDS_CACCONFIG_CGIND }, + { 0xb, 0xffff0000, 16, 0x27C, SISLANDS_CACCONFIG_CGIND }, + { 0xc, 0x0000ffff, 0, 0xA96, SISLANDS_CACCONFIG_CGIND }, + { 0xd, 0x0000ffff, 0, 0x5, SISLANDS_CACCONFIG_CGIND }, + { 0xd, 0xffff0000, 16, 0x5, SISLANDS_CACCONFIG_CGIND }, + { 0xe, 0x0000ffff, 0, 0xB, SISLANDS_CACCONFIG_CGIND }, + { 0xf, 0x0000ffff, 0, 0x3, SISLANDS_CACCONFIG_CGIND }, + { 0xf, 0xffff0000, 16, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x10, 0x0000ffff, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x10, 0xffff0000, 16, 0x4, SISLANDS_CACCONFIG_CGIND }, + { 0x11, 0x0000ffff, 0, 0x15, SISLANDS_CACCONFIG_CGIND }, + { 0x11, 0xffff0000, 16, 0x7, SISLANDS_CACCONFIG_CGIND }, + { 0x12, 0x0000ffff, 0, 0x36, SISLANDS_CACCONFIG_CGIND }, + { 0x13, 0x0000ffff, 0, 0x10, SISLANDS_CACCONFIG_CGIND }, + { 0x13, 0xffff0000, 16, 0x10, SISLANDS_CACCONFIG_CGIND }, + { 0x14, 0x0000ffff, 0, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x15, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x15, 0xffff0000, 16, 0x6, SISLANDS_CACCONFIG_CGIND }, + { 0x4e, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x16, 0x0000ffff, 0, 0x32, SISLANDS_CACCONFIG_CGIND }, + { 0x16, 0xffff0000, 16, 0x7E, SISLANDS_CACCONFIG_CGIND }, + { 0x17, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x18, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x18, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x19, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x19, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x1a, 0x0000ffff, 0, 0x280, SISLANDS_CACCONFIG_CGIND }, + { 0x1a, 0xffff0000, 16, 0x7, SISLANDS_CACCONFIG_CGIND }, + { 0x1b, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x1b, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x1c, 0x0000ffff, 0, 0x3C, SISLANDS_CACCONFIG_CGIND }, + { 0x1c, 0xffff0000, 16, 0x203, SISLANDS_CACCONFIG_CGIND }, + { 0x1d, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x1d, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x1e, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1e, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1f, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1f, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x20, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x6d, 0x0000ffff, 0, 0xB4, SISLANDS_CACCONFIG_CGIND }, + { 0xFFFFFFFF } +}; + +static const struct si_cac_config_reg cac_weights_mars_xt[] = +{ + { 0x0, 0x0000ffff, 0, 0x43, SISLANDS_CACCONFIG_CGIND }, + { 0x0, 0xffff0000, 16, 0x29, SISLANDS_CACCONFIG_CGIND }, + { 0x1, 0x0000ffff, 0, 0xAF, SISLANDS_CACCONFIG_CGIND }, + { 0x1, 0xffff0000, 16, 0x2A, SISLANDS_CACCONFIG_CGIND }, + { 0x2, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x3, 0x0000ffff, 0, 0xA0, SISLANDS_CACCONFIG_CGIND }, + { 0x3, 0xffff0000, 16, 0x29, SISLANDS_CACCONFIG_CGIND }, + { 0x4, 0x0000ffff, 0, 0xA0, SISLANDS_CACCONFIG_CGIND }, + { 0x4, 0xffff0000, 16, 0x59, SISLANDS_CACCONFIG_CGIND }, + { 0x5, 0x0000ffff, 0, 0x1A5, SISLANDS_CACCONFIG_CGIND }, + { 0x5, 0xffff0000, 16, 0x1D6, SISLANDS_CACCONFIG_CGIND }, + { 0x6, 0x0000ffff, 0, 0x2A3, SISLANDS_CACCONFIG_CGIND }, + { 0x6, 0xffff0000, 16, 0x8FD, SISLANDS_CACCONFIG_CGIND }, + { 0x18f, 0x0000ffff, 0, 0x76, SISLANDS_CACCONFIG_CGIND }, + { 0x7, 0x0000ffff, 0, 0x8A, SISLANDS_CACCONFIG_CGIND }, + { 0x7, 0xffff0000, 16, 0xA3, SISLANDS_CACCONFIG_CGIND }, + { 0x8, 0x0000ffff, 0, 0x71, SISLANDS_CACCONFIG_CGIND }, + { 0x8, 0xffff0000, 16, 0x36, SISLANDS_CACCONFIG_CGIND }, + { 0x9, 0x0000ffff, 0, 0xA6, SISLANDS_CACCONFIG_CGIND }, + { 0xa, 0x0000ffff, 0, 0x81, SISLANDS_CACCONFIG_CGIND }, + { 0xb, 0x0000ffff, 0, 0x3D2, SISLANDS_CACCONFIG_CGIND }, + { 0xb, 0xffff0000, 16, 0x27C, SISLANDS_CACCONFIG_CGIND }, + { 0xc, 0x0000ffff, 0, 0xA96, SISLANDS_CACCONFIG_CGIND }, + { 0xd, 0x0000ffff, 0, 0x5, SISLANDS_CACCONFIG_CGIND }, + { 0xd, 0xffff0000, 16, 0x5, SISLANDS_CACCONFIG_CGIND }, + { 0xe, 0x0000ffff, 0, 0xB, SISLANDS_CACCONFIG_CGIND }, + { 0xf, 0x0000ffff, 0, 0x3, SISLANDS_CACCONFIG_CGIND }, + { 0xf, 0xffff0000, 16, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x10, 0x0000ffff, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x10, 0xffff0000, 16, 0x4, SISLANDS_CACCONFIG_CGIND }, + { 0x11, 0x0000ffff, 0, 0x15, SISLANDS_CACCONFIG_CGIND }, + { 0x11, 0xffff0000, 16, 0x7, SISLANDS_CACCONFIG_CGIND }, + { 0x12, 0x0000ffff, 0, 0x36, SISLANDS_CACCONFIG_CGIND }, + { 0x13, 0x0000ffff, 0, 0x10, SISLANDS_CACCONFIG_CGIND }, + { 0x13, 0xffff0000, 16, 0x10, SISLANDS_CACCONFIG_CGIND }, + { 0x14, 0x0000ffff, 0, 0x60, SISLANDS_CACCONFIG_CGIND }, + { 0x15, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x15, 0xffff0000, 16, 0x6, SISLANDS_CACCONFIG_CGIND }, + { 0x4e, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x16, 0x0000ffff, 0, 0x32, SISLANDS_CACCONFIG_CGIND }, + { 0x16, 0xffff0000, 16, 0x7E, SISLANDS_CACCONFIG_CGIND }, + { 0x17, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x18, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x18, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x19, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x19, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x1a, 0x0000ffff, 0, 0x280, SISLANDS_CACCONFIG_CGIND }, + { 0x1a, 0xffff0000, 16, 0x7, SISLANDS_CACCONFIG_CGIND }, + { 0x1b, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x1b, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x1c, 0x0000ffff, 0, 0x3C, SISLANDS_CACCONFIG_CGIND }, + { 0x1c, 0xffff0000, 16, 0x203, SISLANDS_CACCONFIG_CGIND }, + { 0x1d, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x1d, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x1e, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1e, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1f, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1f, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x20, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x6d, 0x0000ffff, 0, 0xB4, SISLANDS_CACCONFIG_CGIND }, + { 0xFFFFFFFF } +}; + +static const struct si_cac_config_reg cac_weights_oland_pro[] = +{ + { 0x0, 0x0000ffff, 0, 0x43, SISLANDS_CACCONFIG_CGIND }, + { 0x0, 0xffff0000, 16, 0x29, SISLANDS_CACCONFIG_CGIND }, + { 0x1, 0x0000ffff, 0, 0xAF, SISLANDS_CACCONFIG_CGIND }, + { 0x1, 0xffff0000, 16, 0x2A, SISLANDS_CACCONFIG_CGIND }, + { 0x2, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x3, 0x0000ffff, 0, 0xA0, SISLANDS_CACCONFIG_CGIND }, + { 0x3, 0xffff0000, 16, 0x29, SISLANDS_CACCONFIG_CGIND }, + { 0x4, 0x0000ffff, 0, 0xA0, SISLANDS_CACCONFIG_CGIND }, + { 0x4, 0xffff0000, 16, 0x59, SISLANDS_CACCONFIG_CGIND }, + { 0x5, 0x0000ffff, 0, 0x1A5, SISLANDS_CACCONFIG_CGIND }, + { 0x5, 0xffff0000, 16, 0x1D6, SISLANDS_CACCONFIG_CGIND }, + { 0x6, 0x0000ffff, 0, 0x2A3, SISLANDS_CACCONFIG_CGIND }, + { 0x6, 0xffff0000, 16, 0x8FD, SISLANDS_CACCONFIG_CGIND }, + { 0x18f, 0x0000ffff, 0, 0x76, SISLANDS_CACCONFIG_CGIND }, + { 0x7, 0x0000ffff, 0, 0x8A, SISLANDS_CACCONFIG_CGIND }, + { 0x7, 0xffff0000, 16, 0xA3, SISLANDS_CACCONFIG_CGIND }, + { 0x8, 0x0000ffff, 0, 0x71, SISLANDS_CACCONFIG_CGIND }, + { 0x8, 0xffff0000, 16, 0x36, SISLANDS_CACCONFIG_CGIND }, + { 0x9, 0x0000ffff, 0, 0xA6, SISLANDS_CACCONFIG_CGIND }, + { 0xa, 0x0000ffff, 0, 0x81, SISLANDS_CACCONFIG_CGIND }, + { 0xb, 0x0000ffff, 0, 0x3D2, SISLANDS_CACCONFIG_CGIND }, + { 0xb, 0xffff0000, 16, 0x27C, SISLANDS_CACCONFIG_CGIND }, + { 0xc, 0x0000ffff, 0, 0xA96, SISLANDS_CACCONFIG_CGIND }, + { 0xd, 0x0000ffff, 0, 0x5, SISLANDS_CACCONFIG_CGIND }, + { 0xd, 0xffff0000, 16, 0x5, SISLANDS_CACCONFIG_CGIND }, + { 0xe, 0x0000ffff, 0, 0xB, SISLANDS_CACCONFIG_CGIND }, + { 0xf, 0x0000ffff, 0, 0x3, SISLANDS_CACCONFIG_CGIND }, + { 0xf, 0xffff0000, 16, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x10, 0x0000ffff, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x10, 0xffff0000, 16, 0x4, SISLANDS_CACCONFIG_CGIND }, + { 0x11, 0x0000ffff, 0, 0x15, SISLANDS_CACCONFIG_CGIND }, + { 0x11, 0xffff0000, 16, 0x7, SISLANDS_CACCONFIG_CGIND }, + { 0x12, 0x0000ffff, 0, 0x36, SISLANDS_CACCONFIG_CGIND }, + { 0x13, 0x0000ffff, 0, 0x10, SISLANDS_CACCONFIG_CGIND }, + { 0x13, 0xffff0000, 16, 0x10, SISLANDS_CACCONFIG_CGIND }, + { 0x14, 0x0000ffff, 0, 0x90, SISLANDS_CACCONFIG_CGIND }, + { 0x15, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x15, 0xffff0000, 16, 0x6, SISLANDS_CACCONFIG_CGIND }, + { 0x4e, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x16, 0x0000ffff, 0, 0x32, SISLANDS_CACCONFIG_CGIND }, + { 0x16, 0xffff0000, 16, 0x7E, SISLANDS_CACCONFIG_CGIND }, + { 0x17, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x18, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x18, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x19, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x19, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x1a, 0x0000ffff, 0, 0x280, SISLANDS_CACCONFIG_CGIND }, + { 0x1a, 0xffff0000, 16, 0x7, SISLANDS_CACCONFIG_CGIND }, + { 0x1b, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x1b, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x1c, 0x0000ffff, 0, 0x3C, SISLANDS_CACCONFIG_CGIND }, + { 0x1c, 0xffff0000, 16, 0x203, SISLANDS_CACCONFIG_CGIND }, + { 0x1d, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x1d, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x1e, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1e, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1f, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1f, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x20, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x6d, 0x0000ffff, 0, 0xB4, SISLANDS_CACCONFIG_CGIND }, + { 0xFFFFFFFF } +}; + +static const struct si_cac_config_reg cac_weights_oland_xt[] = +{ + { 0x0, 0x0000ffff, 0, 0x43, SISLANDS_CACCONFIG_CGIND }, + { 0x0, 0xffff0000, 16, 0x29, SISLANDS_CACCONFIG_CGIND }, + { 0x1, 0x0000ffff, 0, 0xAF, SISLANDS_CACCONFIG_CGIND }, + { 0x1, 0xffff0000, 16, 0x2A, SISLANDS_CACCONFIG_CGIND }, + { 0x2, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x3, 0x0000ffff, 0, 0xA0, SISLANDS_CACCONFIG_CGIND }, + { 0x3, 0xffff0000, 16, 0x29, SISLANDS_CACCONFIG_CGIND }, + { 0x4, 0x0000ffff, 0, 0xA0, SISLANDS_CACCONFIG_CGIND }, + { 0x4, 0xffff0000, 16, 0x59, SISLANDS_CACCONFIG_CGIND }, + { 0x5, 0x0000ffff, 0, 0x1A5, SISLANDS_CACCONFIG_CGIND }, + { 0x5, 0xffff0000, 16, 0x1D6, SISLANDS_CACCONFIG_CGIND }, + { 0x6, 0x0000ffff, 0, 0x2A3, SISLANDS_CACCONFIG_CGIND }, + { 0x6, 0xffff0000, 16, 0x8FD, SISLANDS_CACCONFIG_CGIND }, + { 0x18f, 0x0000ffff, 0, 0x76, SISLANDS_CACCONFIG_CGIND }, + { 0x7, 0x0000ffff, 0, 0x8A, SISLANDS_CACCONFIG_CGIND }, + { 0x7, 0xffff0000, 16, 0xA3, SISLANDS_CACCONFIG_CGIND }, + { 0x8, 0x0000ffff, 0, 0x71, SISLANDS_CACCONFIG_CGIND }, + { 0x8, 0xffff0000, 16, 0x36, SISLANDS_CACCONFIG_CGIND }, + { 0x9, 0x0000ffff, 0, 0xA6, SISLANDS_CACCONFIG_CGIND }, + { 0xa, 0x0000ffff, 0, 0x81, SISLANDS_CACCONFIG_CGIND }, + { 0xb, 0x0000ffff, 0, 0x3D2, SISLANDS_CACCONFIG_CGIND }, + { 0xb, 0xffff0000, 16, 0x27C, SISLANDS_CACCONFIG_CGIND }, + { 0xc, 0x0000ffff, 0, 0xA96, SISLANDS_CACCONFIG_CGIND }, + { 0xd, 0x0000ffff, 0, 0x5, SISLANDS_CACCONFIG_CGIND }, + { 0xd, 0xffff0000, 16, 0x5, SISLANDS_CACCONFIG_CGIND }, + { 0xe, 0x0000ffff, 0, 0xB, SISLANDS_CACCONFIG_CGIND }, + { 0xf, 0x0000ffff, 0, 0x3, SISLANDS_CACCONFIG_CGIND }, + { 0xf, 0xffff0000, 16, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x10, 0x0000ffff, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x10, 0xffff0000, 16, 0x4, SISLANDS_CACCONFIG_CGIND }, + { 0x11, 0x0000ffff, 0, 0x15, SISLANDS_CACCONFIG_CGIND }, + { 0x11, 0xffff0000, 16, 0x7, SISLANDS_CACCONFIG_CGIND }, + { 0x12, 0x0000ffff, 0, 0x36, SISLANDS_CACCONFIG_CGIND }, + { 0x13, 0x0000ffff, 0, 0x10, SISLANDS_CACCONFIG_CGIND }, + { 0x13, 0xffff0000, 16, 0x10, SISLANDS_CACCONFIG_CGIND }, + { 0x14, 0x0000ffff, 0, 0x120, SISLANDS_CACCONFIG_CGIND }, + { 0x15, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x15, 0xffff0000, 16, 0x6, SISLANDS_CACCONFIG_CGIND }, + { 0x4e, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x16, 0x0000ffff, 0, 0x32, SISLANDS_CACCONFIG_CGIND }, + { 0x16, 0xffff0000, 16, 0x7E, SISLANDS_CACCONFIG_CGIND }, + { 0x17, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x18, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x18, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x19, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x19, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x1a, 0x0000ffff, 0, 0x280, SISLANDS_CACCONFIG_CGIND }, + { 0x1a, 0xffff0000, 16, 0x7, SISLANDS_CACCONFIG_CGIND }, + { 0x1b, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x1b, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x1c, 0x0000ffff, 0, 0x3C, SISLANDS_CACCONFIG_CGIND }, + { 0x1c, 0xffff0000, 16, 0x203, SISLANDS_CACCONFIG_CGIND }, + { 0x1d, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x1d, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x1e, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1e, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1f, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1f, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x20, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x6d, 0x0000ffff, 0, 0xB4, SISLANDS_CACCONFIG_CGIND }, + { 0xFFFFFFFF } +}; + +static const struct si_cac_config_reg lcac_oland[] = +{ + { 0x98, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x98, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x104, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x104, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x110, 0x0001fffe, 1, 0x6, SISLANDS_CACCONFIG_CGIND }, + { 0x110, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x14f, 0x0001fffe, 1, 0x6, SISLANDS_CACCONFIG_CGIND }, + { 0x14f, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x8c, 0x0001fffe, 1, 0x6, SISLANDS_CACCONFIG_CGIND }, + { 0x8c, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x143, 0x0001fffe, 1, 0x4, SISLANDS_CACCONFIG_CGIND }, + { 0x143, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x11c, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x11c, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x11f, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x11f, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x164, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x164, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x167, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x167, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x16a, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x16a, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x15e, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x15e, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x161, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x161, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x15b, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x15b, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x16d, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x16d, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x170, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x170, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x173, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x173, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x176, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x176, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x179, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x179, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x17c, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x17c, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x17f, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x17f, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0xFFFFFFFF } +}; + +static const struct si_cac_config_reg lcac_mars_pro[] = +{ + { 0x98, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x98, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x104, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x104, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x110, 0x0001fffe, 1, 0x6, SISLANDS_CACCONFIG_CGIND }, + { 0x110, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x14f, 0x0001fffe, 1, 0x6, SISLANDS_CACCONFIG_CGIND }, + { 0x14f, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x8c, 0x0001fffe, 1, 0x6, SISLANDS_CACCONFIG_CGIND }, + { 0x8c, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x143, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x143, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x11c, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x11c, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x11f, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x11f, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x164, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x164, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x167, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x167, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x16a, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x16a, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x15e, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x15e, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x161, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x161, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x15b, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x15b, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x16d, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND }, + { 0x16d, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x170, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x170, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x173, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x173, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x176, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x176, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x179, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x179, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x17c, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x17c, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x17f, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0x17f, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND }, + { 0xFFFFFFFF } +}; + +static const struct si_cac_config_reg cac_override_oland[] = +{ + { 0xFFFFFFFF } +}; + +static const struct si_powertune_data powertune_data_oland = +{ + ((1 << 16) | 0x6993), + 5, + 0, + 7, + 105, + { + 0UL, + 0UL, + 7194395UL, + 309631529UL, + -1270850L, + 4513710L, + 100 + }, + 117830498UL, + 12, + { + 0, + 0, + 0, + 0, + 0, + 0, + 0, + 0 + }, + true +}; + +static const struct si_powertune_data powertune_data_mars_pro = +{ + ((1 << 16) | 0x6993), + 5, + 0, + 7, + 105, + { + 0UL, + 0UL, + 7194395UL, + 309631529UL, + -1270850L, + 4513710L, + 100 + }, + 117830498UL, + 12, + { + 0, + 0, + 0, + 0, + 0, + 0, + 0, + 0 + }, + true +}; + +static const struct si_dte_data dte_data_oland = +{ + { 0, 0, 0, 0, 0 }, + { 0, 0, 0, 0, 0 }, + 0, + 0, + 0, + 0, + 0, + 0, + 0, + { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 }, + { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 }, + { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 }, + 0, + false +}; + +static const struct si_dte_data dte_data_mars_pro = +{ + { 0x1E8480, 0x3D0900, 0x989680, 0x2625A00, 0x0 }, + { 0x0, 0x0, 0x0, 0x0, 0x0 }, + 5, + 55000, + 105, + 0xA, + 1, + 0, + 0x10, + { 0x96, 0xB4, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF }, + { 0x895440, 0x3D0900, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680 }, + { 0xF627, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0 }, + 90, + true +}; + +static const struct si_dte_data dte_data_sun_xt = +{ + { 0x1E8480, 0x3D0900, 0x989680, 0x2625A00, 0x0 }, + { 0x0, 0x0, 0x0, 0x0, 0x0 }, + 5, + 55000, + 105, + 0xA, + 1, + 0, + 0x10, + { 0x96, 0xB4, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF }, + { 0x895440, 0x3D0900, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680 }, + { 0xD555, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0 }, + 90, + true +}; + + +static const struct si_cac_config_reg cac_weights_hainan[] = +{ + { 0x0, 0x0000ffff, 0, 0x2d9, SISLANDS_CACCONFIG_CGIND }, + { 0x0, 0xffff0000, 16, 0x22b, SISLANDS_CACCONFIG_CGIND }, + { 0x1, 0x0000ffff, 0, 0x21c, SISLANDS_CACCONFIG_CGIND }, + { 0x1, 0xffff0000, 16, 0x1dc, SISLANDS_CACCONFIG_CGIND }, + { 0x2, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x3, 0x0000ffff, 0, 0x24e, SISLANDS_CACCONFIG_CGIND }, + { 0x3, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x4, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x4, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x5, 0x0000ffff, 0, 0x35e, SISLANDS_CACCONFIG_CGIND }, + { 0x5, 0xffff0000, 16, 0x1143, SISLANDS_CACCONFIG_CGIND }, + { 0x6, 0x0000ffff, 0, 0xe17, SISLANDS_CACCONFIG_CGIND }, + { 0x6, 0xffff0000, 16, 0x441, SISLANDS_CACCONFIG_CGIND }, + { 0x18f, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x7, 0x0000ffff, 0, 0x28b, SISLANDS_CACCONFIG_CGIND }, + { 0x7, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x8, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x8, 0xffff0000, 16, 0xabe, SISLANDS_CACCONFIG_CGIND }, + { 0x9, 0x0000ffff, 0, 0xf11, SISLANDS_CACCONFIG_CGIND }, + { 0xa, 0x0000ffff, 0, 0x907, SISLANDS_CACCONFIG_CGIND }, + { 0xb, 0x0000ffff, 0, 0xb45, SISLANDS_CACCONFIG_CGIND }, + { 0xb, 0xffff0000, 16, 0xd1e, SISLANDS_CACCONFIG_CGIND }, + { 0xc, 0x0000ffff, 0, 0xa2c, SISLANDS_CACCONFIG_CGIND }, + { 0xd, 0x0000ffff, 0, 0x62, SISLANDS_CACCONFIG_CGIND }, + { 0xd, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0xe, 0x0000ffff, 0, 0x1f3, SISLANDS_CACCONFIG_CGIND }, + { 0xf, 0x0000ffff, 0, 0x42, SISLANDS_CACCONFIG_CGIND }, + { 0xf, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x10, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x10, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x11, 0x0000ffff, 0, 0x709, SISLANDS_CACCONFIG_CGIND }, + { 0x11, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x12, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x13, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x13, 0xffff0000, 16, 0x3a, SISLANDS_CACCONFIG_CGIND }, + { 0x14, 0x0000ffff, 0, 0x357, SISLANDS_CACCONFIG_CGIND }, + { 0x15, 0x0000ffff, 0, 0x9f, SISLANDS_CACCONFIG_CGIND }, + { 0x15, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x4e, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x16, 0x0000ffff, 0, 0x314, SISLANDS_CACCONFIG_CGIND }, + { 0x16, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x17, 0x0000ffff, 0, 0x6d, SISLANDS_CACCONFIG_CGIND }, + { 0x18, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x18, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x19, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x19, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND }, + { 0x1a, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1a, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1b, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1b, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1c, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1c, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1d, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1d, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1e, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1e, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1f, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x1f, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x20, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND }, + { 0x6d, 0x0000ffff, 0, 0x1b9, SISLANDS_CACCONFIG_CGIND }, + { 0xFFFFFFFF } +}; + +static const struct si_powertune_data powertune_data_hainan = +{ + ((1 << 16) | 0x6993), + 5, + 0, + 9, + 105, + { + 0UL, + 0UL, + 7194395UL, + 309631529UL, + -1270850L, + 4513710L, + 100 + }, + 117830498UL, + 12, + { + 0, + 0, + 0, + 0, + 0, + 0, + 0, + 0 + }, + true +}; + +static struct rv7xx_power_info *rv770_get_pi(struct amdgpu_device *adev); +static struct evergreen_power_info *evergreen_get_pi(struct amdgpu_device *adev); +static struct ni_power_info *ni_get_pi(struct amdgpu_device *adev); +static struct si_ps *si_get_ps(struct amdgpu_ps *rps); + +static int si_populate_voltage_value(struct amdgpu_device *adev, + const struct atom_voltage_table *table, + u16 value, SISLANDS_SMC_VOLTAGE_VALUE *voltage); +static int si_get_std_voltage_value(struct amdgpu_device *adev, + SISLANDS_SMC_VOLTAGE_VALUE *voltage, + u16 *std_voltage); +static int si_write_smc_soft_register(struct amdgpu_device *adev, + u16 reg_offset, u32 value); +static int si_convert_power_level_to_smc(struct amdgpu_device *adev, + struct rv7xx_pl *pl, + SISLANDS_SMC_HW_PERFORMANCE_LEVEL *level); +static int si_calculate_sclk_params(struct amdgpu_device *adev, + u32 engine_clock, + SISLANDS_SMC_SCLK_VALUE *sclk); + +static void si_thermal_start_smc_fan_control(struct amdgpu_device *adev); +static void si_fan_ctrl_set_default_mode(struct amdgpu_device *adev); +static void si_dpm_set_dpm_funcs(struct amdgpu_device *adev); +static void si_dpm_set_irq_funcs(struct amdgpu_device *adev); + +static struct si_power_info *si_get_pi(struct amdgpu_device *adev) +{ + struct si_power_info *pi = adev->pm.dpm.priv; + return pi; +} + +static void si_calculate_leakage_for_v_and_t_formula(const struct ni_leakage_coeffients *coeff, + u16 v, s32 t, u32 ileakage, u32 *leakage) +{ + s64 kt, kv, leakage_w, i_leakage, vddc; + s64 temperature, t_slope, t_intercept, av, bv, t_ref; + s64 tmp; + + i_leakage = div64_s64(drm_int2fixp(ileakage), 100); + vddc = div64_s64(drm_int2fixp(v), 1000); + temperature = div64_s64(drm_int2fixp(t), 1000); + + t_slope = div64_s64(drm_int2fixp(coeff->t_slope), 100000000); + t_intercept = div64_s64(drm_int2fixp(coeff->t_intercept), 100000000); + av = div64_s64(drm_int2fixp(coeff->av), 100000000); + bv = div64_s64(drm_int2fixp(coeff->bv), 100000000); + t_ref = drm_int2fixp(coeff->t_ref); + + tmp = drm_fixp_mul(t_slope, vddc) + t_intercept; + kt = drm_fixp_exp(drm_fixp_mul(tmp, temperature)); + kt = drm_fixp_div(kt, drm_fixp_exp(drm_fixp_mul(tmp, t_ref))); + kv = drm_fixp_mul(av, drm_fixp_exp(drm_fixp_mul(bv, vddc))); + + leakage_w = drm_fixp_mul(drm_fixp_mul(drm_fixp_mul(i_leakage, kt), kv), vddc); + + *leakage = drm_fixp2int(leakage_w * 1000); +} + +static void si_calculate_leakage_for_v_and_t(struct amdgpu_device *adev, + const struct ni_leakage_coeffients *coeff, + u16 v, + s32 t, + u32 i_leakage, + u32 *leakage) +{ + si_calculate_leakage_for_v_and_t_formula(coeff, v, t, i_leakage, leakage); +} + +static void si_calculate_leakage_for_v_formula(const struct ni_leakage_coeffients *coeff, + const u32 fixed_kt, u16 v, + u32 ileakage, u32 *leakage) +{ + s64 kt, kv, leakage_w, i_leakage, vddc; + + i_leakage = div64_s64(drm_int2fixp(ileakage), 100); + vddc = div64_s64(drm_int2fixp(v), 1000); + + kt = div64_s64(drm_int2fixp(fixed_kt), 100000000); + kv = drm_fixp_mul(div64_s64(drm_int2fixp(coeff->av), 100000000), + drm_fixp_exp(drm_fixp_mul(div64_s64(drm_int2fixp(coeff->bv), 100000000), vddc))); + + leakage_w = drm_fixp_mul(drm_fixp_mul(drm_fixp_mul(i_leakage, kt), kv), vddc); + + *leakage = drm_fixp2int(leakage_w * 1000); +} + +static void si_calculate_leakage_for_v(struct amdgpu_device *adev, + const struct ni_leakage_coeffients *coeff, + const u32 fixed_kt, + u16 v, + u32 i_leakage, + u32 *leakage) +{ + si_calculate_leakage_for_v_formula(coeff, fixed_kt, v, i_leakage, leakage); +} + + +static void si_update_dte_from_pl2(struct amdgpu_device *adev, + struct si_dte_data *dte_data) +{ + u32 p_limit1 = adev->pm.dpm.tdp_limit; + u32 p_limit2 = adev->pm.dpm.near_tdp_limit; + u32 k = dte_data->k; + u32 t_max = dte_data->max_t; + u32 t_split[5] = { 10, 15, 20, 25, 30 }; + u32 t_0 = dte_data->t0; + u32 i; + + if (p_limit2 != 0 && p_limit2 <= p_limit1) { + dte_data->tdep_count = 3; + + for (i = 0; i < k; i++) { + dte_data->r[i] = + (t_split[i] * (t_max - t_0/(u32)1000) * (1 << 14)) / + (p_limit2 * (u32)100); + } + + dte_data->tdep_r[1] = dte_data->r[4] * 2; + + for (i = 2; i < SMC_SISLANDS_DTE_MAX_TEMPERATURE_DEPENDENT_ARRAY_SIZE; i++) { + dte_data->tdep_r[i] = dte_data->r[4]; + } + } else { + DRM_ERROR("Invalid PL2! DTE will not be updated.\n"); + } +} + +static struct rv7xx_power_info *rv770_get_pi(struct amdgpu_device *adev) +{ + struct rv7xx_power_info *pi = adev->pm.dpm.priv; + + return pi; +} + +static struct ni_power_info *ni_get_pi(struct amdgpu_device *adev) +{ + struct ni_power_info *pi = adev->pm.dpm.priv; + + return pi; +} + +static struct si_ps *si_get_ps(struct amdgpu_ps *aps) +{ + struct si_ps *ps = aps->ps_priv; + + return ps; +} + +static void si_initialize_powertune_defaults(struct amdgpu_device *adev) +{ + struct ni_power_info *ni_pi = ni_get_pi(adev); + struct si_power_info *si_pi = si_get_pi(adev); + bool update_dte_from_pl2 = false; + + if (adev->asic_type == CHIP_TAHITI) { + si_pi->cac_weights = cac_weights_tahiti; + si_pi->lcac_config = lcac_tahiti; + si_pi->cac_override = cac_override_tahiti; + si_pi->powertune_data = &powertune_data_tahiti; + si_pi->dte_data = dte_data_tahiti; + + switch (adev->pdev->device) { + case 0x6798: + si_pi->dte_data.enable_dte_by_default = true; + break; + case 0x6799: + si_pi->dte_data = dte_data_new_zealand; + break; + case 0x6790: + case 0x6791: + case 0x6792: + case 0x679E: + si_pi->dte_data = dte_data_aruba_pro; + update_dte_from_pl2 = true; + break; + case 0x679B: + si_pi->dte_data = dte_data_malta; + update_dte_from_pl2 = true; + break; + case 0x679A: + si_pi->dte_data = dte_data_tahiti_pro; + update_dte_from_pl2 = true; + break; + default: + if (si_pi->dte_data.enable_dte_by_default == true) + DRM_ERROR("DTE is not enabled!\n"); + break; + } + } else if (adev->asic_type == CHIP_PITCAIRN) { + si_pi->cac_weights = cac_weights_pitcairn; + si_pi->lcac_config = lcac_pitcairn; + si_pi->cac_override = cac_override_pitcairn; + si_pi->powertune_data = &powertune_data_pitcairn; + + switch (adev->pdev->device) { + case 0x6810: + case 0x6818: + si_pi->dte_data = dte_data_curacao_xt; + update_dte_from_pl2 = true; + break; + case 0x6819: + case 0x6811: + si_pi->dte_data = dte_data_curacao_pro; + update_dte_from_pl2 = true; + break; + case 0x6800: + case 0x6806: + si_pi->dte_data = dte_data_neptune_xt; + update_dte_from_pl2 = true; + break; + default: + si_pi->dte_data = dte_data_pitcairn; + break; + } + } else if (adev->asic_type == CHIP_VERDE) { + si_pi->lcac_config = lcac_cape_verde; + si_pi->cac_override = cac_override_cape_verde; + si_pi->powertune_data = &powertune_data_cape_verde; + + switch (adev->pdev->device) { + case 0x683B: + case 0x683F: + case 0x6829: + case 0x6835: + si_pi->cac_weights = cac_weights_cape_verde_pro; + si_pi->dte_data = dte_data_cape_verde; + break; + case 0x682C: + si_pi->cac_weights = cac_weights_cape_verde_pro; + si_pi->dte_data = dte_data_sun_xt; + break; + case 0x6825: + case 0x6827: + si_pi->cac_weights = cac_weights_heathrow; + si_pi->dte_data = dte_data_cape_verde; + break; + case 0x6824: + case 0x682D: + si_pi->cac_weights = cac_weights_chelsea_xt; + si_pi->dte_data = dte_data_cape_verde; + break; + case 0x682F: + si_pi->cac_weights = cac_weights_chelsea_pro; + si_pi->dte_data = dte_data_cape_verde; + break; + case 0x6820: + si_pi->cac_weights = cac_weights_heathrow; + si_pi->dte_data = dte_data_venus_xtx; + break; + case 0x6821: + si_pi->cac_weights = cac_weights_heathrow; + si_pi->dte_data = dte_data_venus_xt; + break; + case 0x6823: + case 0x682B: + case 0x6822: + case 0x682A: + si_pi->cac_weights = cac_weights_chelsea_pro; + si_pi->dte_data = dte_data_venus_pro; + break; + default: + si_pi->cac_weights = cac_weights_cape_verde; + si_pi->dte_data = dte_data_cape_verde; + break; + } + } else if (adev->asic_type == CHIP_OLAND) { + si_pi->lcac_config = lcac_mars_pro; + si_pi->cac_override = cac_override_oland; + si_pi->powertune_data = &powertune_data_mars_pro; + si_pi->dte_data = dte_data_mars_pro; + + switch (adev->pdev->device) { + case 0x6601: + case 0x6621: + case 0x6603: + case 0x6605: + si_pi->cac_weights = cac_weights_mars_pro; + update_dte_from_pl2 = true; + break; + case 0x6600: + case 0x6606: + case 0x6620: + case 0x6604: + si_pi->cac_weights = cac_weights_mars_xt; + update_dte_from_pl2 = true; + break; + case 0x6611: + case 0x6613: + case 0x6608: + si_pi->cac_weights = cac_weights_oland_pro; + update_dte_from_pl2 = true; + break; + case 0x6610: + si_pi->cac_weights = cac_weights_oland_xt; + update_dte_from_pl2 = true; + break; + default: + si_pi->cac_weights = cac_weights_oland; + si_pi->lcac_config = lcac_oland; + si_pi->cac_override = cac_override_oland; + si_pi->powertune_data = &powertune_data_oland; + si_pi->dte_data = dte_data_oland; + break; + } + } else if (adev->asic_type == CHIP_HAINAN) { + si_pi->cac_weights = cac_weights_hainan; + si_pi->lcac_config = lcac_oland; + si_pi->cac_override = cac_override_oland; + si_pi->powertune_data = &powertune_data_hainan; + si_pi->dte_data = dte_data_sun_xt; + update_dte_from_pl2 = true; + } else { + DRM_ERROR("Unknown SI asic revision, failed to initialize PowerTune!\n"); + return; + } + + ni_pi->enable_power_containment = false; + ni_pi->enable_cac = false; + ni_pi->enable_sq_ramping = false; + si_pi->enable_dte = false; + + if (si_pi->powertune_data->enable_powertune_by_default) { + ni_pi->enable_power_containment = true; + ni_pi->enable_cac = true; + if (si_pi->dte_data.enable_dte_by_default) { + si_pi->enable_dte = true; + if (update_dte_from_pl2) + si_update_dte_from_pl2(adev, &si_pi->dte_data); + + } + ni_pi->enable_sq_ramping = true; + } + + ni_pi->driver_calculate_cac_leakage = true; + ni_pi->cac_configuration_required = true; + + if (ni_pi->cac_configuration_required) { + ni_pi->support_cac_long_term_average = true; + si_pi->dyn_powertune_data.l2_lta_window_size = + si_pi->powertune_data->l2_lta_window_size_default; + si_pi->dyn_powertune_data.lts_truncate = + si_pi->powertune_data->lts_truncate_default; + } else { + ni_pi->support_cac_long_term_average = false; + si_pi->dyn_powertune_data.l2_lta_window_size = 0; + si_pi->dyn_powertune_data.lts_truncate = 0; + } + + si_pi->dyn_powertune_data.disable_uvd_powertune = false; +} + +static u32 si_get_smc_power_scaling_factor(struct amdgpu_device *adev) +{ + return 1; +} + +static u32 si_calculate_cac_wintime(struct amdgpu_device *adev) +{ + u32 xclk; + u32 wintime; + u32 cac_window; + u32 cac_window_size; + + xclk = amdgpu_asic_get_xclk(adev); + + if (xclk == 0) + return 0; + + cac_window = RREG32(CG_CAC_CTRL) & CAC_WINDOW_MASK; + cac_window_size = ((cac_window & 0xFFFF0000) >> 16) * (cac_window & 0x0000FFFF); + + wintime = (cac_window_size * 100) / xclk; + + return wintime; +} + +static u32 si_scale_power_for_smc(u32 power_in_watts, u32 scaling_factor) +{ + return power_in_watts; +} + +static int si_calculate_adjusted_tdp_limits(struct amdgpu_device *adev, + bool adjust_polarity, + u32 tdp_adjustment, + u32 *tdp_limit, + u32 *near_tdp_limit) +{ + u32 adjustment_delta, max_tdp_limit; + + if (tdp_adjustment > (u32)adev->pm.dpm.tdp_od_limit) + return -EINVAL; + + max_tdp_limit = ((100 + 100) * adev->pm.dpm.tdp_limit) / 100; + + if (adjust_polarity) { + *tdp_limit = ((100 + tdp_adjustment) * adev->pm.dpm.tdp_limit) / 100; + *near_tdp_limit = adev->pm.dpm.near_tdp_limit_adjusted + (*tdp_limit - adev->pm.dpm.tdp_limit); + } else { + *tdp_limit = ((100 - tdp_adjustment) * adev->pm.dpm.tdp_limit) / 100; + adjustment_delta = adev->pm.dpm.tdp_limit - *tdp_limit; + if (adjustment_delta < adev->pm.dpm.near_tdp_limit_adjusted) + *near_tdp_limit = adev->pm.dpm.near_tdp_limit_adjusted - adjustment_delta; + else + *near_tdp_limit = 0; + } + + if ((*tdp_limit <= 0) || (*tdp_limit > max_tdp_limit)) + return -EINVAL; + if ((*near_tdp_limit <= 0) || (*near_tdp_limit > *tdp_limit)) + return -EINVAL; + + return 0; +} + +static int si_populate_smc_tdp_limits(struct amdgpu_device *adev, + struct amdgpu_ps *amdgpu_state) +{ + struct ni_power_info *ni_pi = ni_get_pi(adev); + struct si_power_info *si_pi = si_get_pi(adev); + + if (ni_pi->enable_power_containment) { + SISLANDS_SMC_STATETABLE *smc_table = &si_pi->smc_statetable; + PP_SIslands_PAPMParameters *papm_parm; + struct amdgpu_ppm_table *ppm = adev->pm.dpm.dyn_state.ppm_table; + u32 scaling_factor = si_get_smc_power_scaling_factor(adev); + u32 tdp_limit; + u32 near_tdp_limit; + int ret; + + if (scaling_factor == 0) + return -EINVAL; + + memset(smc_table, 0, sizeof(SISLANDS_SMC_STATETABLE)); + + ret = si_calculate_adjusted_tdp_limits(adev, + false, /* ??? */ + adev->pm.dpm.tdp_adjustment, + &tdp_limit, + &near_tdp_limit); + if (ret) + return ret; + + smc_table->dpm2Params.TDPLimit = + cpu_to_be32(si_scale_power_for_smc(tdp_limit, scaling_factor) * 1000); + smc_table->dpm2Params.NearTDPLimit = + cpu_to_be32(si_scale_power_for_smc(near_tdp_limit, scaling_factor) * 1000); + smc_table->dpm2Params.SafePowerLimit = + cpu_to_be32(si_scale_power_for_smc((near_tdp_limit * SISLANDS_DPM2_TDP_SAFE_LIMIT_PERCENT) / 100, scaling_factor) * 1000); + + ret = amdgpu_si_copy_bytes_to_smc(adev, + (si_pi->state_table_start + offsetof(SISLANDS_SMC_STATETABLE, dpm2Params) + + offsetof(PP_SIslands_DPM2Parameters, TDPLimit)), + (u8 *)(&(smc_table->dpm2Params.TDPLimit)), + sizeof(u32) * 3, + si_pi->sram_end); + if (ret) + return ret; + + if (si_pi->enable_ppm) { + papm_parm = &si_pi->papm_parm; + memset(papm_parm, 0, sizeof(PP_SIslands_PAPMParameters)); + papm_parm->NearTDPLimitTherm = cpu_to_be32(ppm->dgpu_tdp); + papm_parm->dGPU_T_Limit = cpu_to_be32(ppm->tj_max); + papm_parm->dGPU_T_Warning = cpu_to_be32(95); + papm_parm->dGPU_T_Hysteresis = cpu_to_be32(5); + papm_parm->PlatformPowerLimit = 0xffffffff; + papm_parm->NearTDPLimitPAPM = 0xffffffff; + + ret = amdgpu_si_copy_bytes_to_smc(adev, si_pi->papm_cfg_table_start, + (u8 *)papm_parm, + sizeof(PP_SIslands_PAPMParameters), + si_pi->sram_end); + if (ret) + return ret; + } + } + return 0; +} + +static int si_populate_smc_tdp_limits_2(struct amdgpu_device *adev, + struct amdgpu_ps *amdgpu_state) +{ + struct ni_power_info *ni_pi = ni_get_pi(adev); + struct si_power_info *si_pi = si_get_pi(adev); + + if (ni_pi->enable_power_containment) { + SISLANDS_SMC_STATETABLE *smc_table = &si_pi->smc_statetable; + u32 scaling_factor = si_get_smc_power_scaling_factor(adev); + int ret; + + memset(smc_table, 0, sizeof(SISLANDS_SMC_STATETABLE)); + + smc_table->dpm2Params.NearTDPLimit = + cpu_to_be32(si_scale_power_for_smc(adev->pm.dpm.near_tdp_limit_adjusted, scaling_factor) * 1000); + smc_table->dpm2Params.SafePowerLimit = + cpu_to_be32(si_scale_power_for_smc((adev->pm.dpm.near_tdp_limit_adjusted * SISLANDS_DPM2_TDP_SAFE_LIMIT_PERCENT) / 100, scaling_factor) * 1000); + + ret = amdgpu_si_copy_bytes_to_smc(adev, + (si_pi->state_table_start + + offsetof(SISLANDS_SMC_STATETABLE, dpm2Params) + + offsetof(PP_SIslands_DPM2Parameters, NearTDPLimit)), + (u8 *)(&(smc_table->dpm2Params.NearTDPLimit)), + sizeof(u32) * 2, + si_pi->sram_end); + if (ret) + return ret; + } + + return 0; +} + +static u16 si_calculate_power_efficiency_ratio(struct amdgpu_device *adev, + const u16 prev_std_vddc, + const u16 curr_std_vddc) +{ + u64 margin = (u64)SISLANDS_DPM2_PWREFFICIENCYRATIO_MARGIN; + u64 prev_vddc = (u64)prev_std_vddc; + u64 curr_vddc = (u64)curr_std_vddc; + u64 pwr_efficiency_ratio, n, d; + + if ((prev_vddc == 0) || (curr_vddc == 0)) + return 0; + + n = div64_u64((u64)1024 * curr_vddc * curr_vddc * ((u64)1000 + margin), (u64)1000); + d = prev_vddc * prev_vddc; + pwr_efficiency_ratio = div64_u64(n, d); + + if (pwr_efficiency_ratio > (u64)0xFFFF) + return 0; + + return (u16)pwr_efficiency_ratio; +} + +static bool si_should_disable_uvd_powertune(struct amdgpu_device *adev, + struct amdgpu_ps *amdgpu_state) +{ + struct si_power_info *si_pi = si_get_pi(adev); + + if (si_pi->dyn_powertune_data.disable_uvd_powertune && + amdgpu_state->vclk && amdgpu_state->dclk) + return true; + + return false; +} + +struct evergreen_power_info *evergreen_get_pi(struct amdgpu_device *adev) +{ + struct evergreen_power_info *pi = adev->pm.dpm.priv; + + return pi; +} + +static int si_populate_power_containment_values(struct amdgpu_device *adev, + struct amdgpu_ps *amdgpu_state, + SISLANDS_SMC_SWSTATE *smc_state) +{ + struct evergreen_power_info *eg_pi = evergreen_get_pi(adev); + struct ni_power_info *ni_pi = ni_get_pi(adev); + struct si_ps *state = si_get_ps(amdgpu_state); + SISLANDS_SMC_VOLTAGE_VALUE vddc; + u32 prev_sclk; + u32 max_sclk; + u32 min_sclk; + u16 prev_std_vddc; + u16 curr_std_vddc; + int i; + u16 pwr_efficiency_ratio; + u8 max_ps_percent; + bool disable_uvd_power_tune; + int ret; + + if (ni_pi->enable_power_containment == false) + return 0; + + if (state->performance_level_count == 0) + return -EINVAL; + + if (smc_state->levelCount != state->performance_level_count) + return -EINVAL; + + disable_uvd_power_tune = si_should_disable_uvd_powertune(adev, amdgpu_state); + + smc_state->levels[0].dpm2.MaxPS = 0; + smc_state->levels[0].dpm2.NearTDPDec = 0; + smc_state->levels[0].dpm2.AboveSafeInc = 0; + smc_state->levels[0].dpm2.BelowSafeInc = 0; + smc_state->levels[0].dpm2.PwrEfficiencyRatio = 0; + + for (i = 1; i < state->performance_level_count; i++) { + prev_sclk = state->performance_levels[i-1].sclk; + max_sclk = state->performance_levels[i].sclk; + if (i == 1) + max_ps_percent = SISLANDS_DPM2_MAXPS_PERCENT_M; + else + max_ps_percent = SISLANDS_DPM2_MAXPS_PERCENT_H; + + if (prev_sclk > max_sclk) + return -EINVAL; + + if ((max_ps_percent == 0) || + (prev_sclk == max_sclk) || + disable_uvd_power_tune) + min_sclk = max_sclk; + else if (i == 1) + min_sclk = prev_sclk; + else + min_sclk = (prev_sclk * (u32)max_ps_percent) / 100; + + if (min_sclk < state->performance_levels[0].sclk) + min_sclk = state->performance_levels[0].sclk; + + if (min_sclk == 0) + return -EINVAL; + + ret = si_populate_voltage_value(adev, &eg_pi->vddc_voltage_table, + state->performance_levels[i-1].vddc, &vddc); + if (ret) + return ret; + + ret = si_get_std_voltage_value(adev, &vddc, &prev_std_vddc); + if (ret) + return ret; + + ret = si_populate_voltage_value(adev, &eg_pi->vddc_voltage_table, + state->performance_levels[i].vddc, &vddc); + if (ret) + return ret; + + ret = si_get_std_voltage_value(adev, &vddc, &curr_std_vddc); + if (ret) + return ret; + + pwr_efficiency_ratio = si_calculate_power_efficiency_ratio(adev, + prev_std_vddc, curr_std_vddc); + + smc_state->levels[i].dpm2.MaxPS = (u8)((SISLANDS_DPM2_MAX_PULSE_SKIP * (max_sclk - min_sclk)) / max_sclk); + smc_state->levels[i].dpm2.NearTDPDec = SISLANDS_DPM2_NEAR_TDP_DEC; + smc_state->levels[i].dpm2.AboveSafeInc = SISLANDS_DPM2_ABOVE_SAFE_INC; + smc_state->levels[i].dpm2.BelowSafeInc = SISLANDS_DPM2_BELOW_SAFE_INC; + smc_state->levels[i].dpm2.PwrEfficiencyRatio = cpu_to_be16(pwr_efficiency_ratio); + } + + return 0; +} + +static int si_populate_sq_ramping_values(struct amdgpu_device *adev, + struct amdgpu_ps *amdgpu_state, + SISLANDS_SMC_SWSTATE *smc_state) +{ + struct ni_power_info *ni_pi = ni_get_pi(adev); + struct si_ps *state = si_get_ps(amdgpu_state); + u32 sq_power_throttle, sq_power_throttle2; + bool enable_sq_ramping = ni_pi->enable_sq_ramping; + int i; + + if (state->performance_level_count == 0) + return -EINVAL; + + if (smc_state->levelCount != state->performance_level_count) + return -EINVAL; + + if (adev->pm.dpm.sq_ramping_threshold == 0) + return -EINVAL; + + if (SISLANDS_DPM2_SQ_RAMP_MAX_POWER > (MAX_POWER_MASK >> MAX_POWER_SHIFT)) + enable_sq_ramping = false; + + if (SISLANDS_DPM2_SQ_RAMP_MIN_POWER > (MIN_POWER_MASK >> MIN_POWER_SHIFT)) + enable_sq_ramping = false; + + if (SISLANDS_DPM2_SQ_RAMP_MAX_POWER_DELTA > (MAX_POWER_DELTA_MASK >> MAX_POWER_DELTA_SHIFT)) + enable_sq_ramping = false; + + if (SISLANDS_DPM2_SQ_RAMP_STI_SIZE > (STI_SIZE_MASK >> STI_SIZE_SHIFT)) + enable_sq_ramping = false; + + if (SISLANDS_DPM2_SQ_RAMP_LTI_RATIO > (LTI_RATIO_MASK >> LTI_RATIO_SHIFT)) + enable_sq_ramping = false; + + for (i = 0; i < state->performance_level_count; i++) { + sq_power_throttle = 0; + sq_power_throttle2 = 0; + + if ((state->performance_levels[i].sclk >= adev->pm.dpm.sq_ramping_threshold) && + enable_sq_ramping) { + sq_power_throttle |= MAX_POWER(SISLANDS_DPM2_SQ_RAMP_MAX_POWER); + sq_power_throttle |= MIN_POWER(SISLANDS_DPM2_SQ_RAMP_MIN_POWER); + sq_power_throttle2 |= MAX_POWER_DELTA(SISLANDS_DPM2_SQ_RAMP_MAX_POWER_DELTA); + sq_power_throttle2 |= STI_SIZE(SISLANDS_DPM2_SQ_RAMP_STI_SIZE); + sq_power_throttle2 |= LTI_RATIO(SISLANDS_DPM2_SQ_RAMP_LTI_RATIO); + } else { + sq_power_throttle |= MAX_POWER_MASK | MIN_POWER_MASK; + sq_power_throttle2 |= MAX_POWER_DELTA_MASK | STI_SIZE_MASK | LTI_RATIO_MASK; + } + + smc_state->levels[i].SQPowerThrottle = cpu_to_be32(sq_power_throttle); + smc_state->levels[i].SQPowerThrottle_2 = cpu_to_be32(sq_power_throttle2); + } + + return 0; +} + +static int si_enable_power_containment(struct amdgpu_device *adev, + struct amdgpu_ps *amdgpu_new_state, + bool enable) +{ + struct ni_power_info *ni_pi = ni_get_pi(adev); + PPSMC_Result smc_result; + int ret = 0; + + if (ni_pi->enable_power_containment) { + if (enable) { + if (!si_should_disable_uvd_powertune(adev, amdgpu_new_state)) { + smc_result = amdgpu_si_send_msg_to_smc(adev, PPSMC_TDPClampingActive); + if (smc_result != PPSMC_Result_OK) { + ret = -EINVAL; + ni_pi->pc_enabled = false; + } else { + ni_pi->pc_enabled = true; + } + } + } else { + smc_result = amdgpu_si_send_msg_to_smc(adev, PPSMC_TDPClampingInactive); + if (smc_result != PPSMC_Result_OK) + ret = -EINVAL; + ni_pi->pc_enabled = false; + } + } + + return ret; +} + +static int si_initialize_smc_dte_tables(struct amdgpu_device *adev) +{ + struct si_power_info *si_pi = si_get_pi(adev); + int ret = 0; + struct si_dte_data *dte_data = &si_pi->dte_data; + Smc_SIslands_DTE_Configuration *dte_tables = NULL; + u32 table_size; + u8 tdep_count; + u32 i; + + if (dte_data == NULL) + si_pi->enable_dte = false; + + if (si_pi->enable_dte == false) + return 0; + + if (dte_data->k <= 0) + return -EINVAL; + + dte_tables = kzalloc(sizeof(Smc_SIslands_DTE_Configuration), GFP_KERNEL); + if (dte_tables == NULL) { + si_pi->enable_dte = false; + return -ENOMEM; + } + + table_size = dte_data->k; + + if (table_size > SMC_SISLANDS_DTE_MAX_FILTER_STAGES) + table_size = SMC_SISLANDS_DTE_MAX_FILTER_STAGES; + + tdep_count = dte_data->tdep_count; + if (tdep_count > SMC_SISLANDS_DTE_MAX_TEMPERATURE_DEPENDENT_ARRAY_SIZE) + tdep_count = SMC_SISLANDS_DTE_MAX_TEMPERATURE_DEPENDENT_ARRAY_SIZE; + + dte_tables->K = cpu_to_be32(table_size); + dte_tables->T0 = cpu_to_be32(dte_data->t0); + dte_tables->MaxT = cpu_to_be32(dte_data->max_t); + dte_tables->WindowSize = dte_data->window_size; + dte_tables->temp_select = dte_data->temp_select; + dte_tables->DTE_mode = dte_data->dte_mode; + dte_tables->Tthreshold = cpu_to_be32(dte_data->t_threshold); + + if (tdep_count > 0) + table_size--; + + for (i = 0; i < table_size; i++) { + dte_tables->tau[i] = cpu_to_be32(dte_data->tau[i]); + dte_tables->R[i] = cpu_to_be32(dte_data->r[i]); + } + + dte_tables->Tdep_count = tdep_count; + + for (i = 0; i < (u32)tdep_count; i++) { + dte_tables->T_limits[i] = dte_data->t_limits[i]; + dte_tables->Tdep_tau[i] = cpu_to_be32(dte_data->tdep_tau[i]); + dte_tables->Tdep_R[i] = cpu_to_be32(dte_data->tdep_r[i]); + } + + ret = amdgpu_si_copy_bytes_to_smc(adev, si_pi->dte_table_start, + (u8 *)dte_tables, + sizeof(Smc_SIslands_DTE_Configuration), + si_pi->sram_end); + kfree(dte_tables); + + return ret; +} + +static int si_get_cac_std_voltage_max_min(struct amdgpu_device *adev, + u16 *max, u16 *min) +{ + struct si_power_info *si_pi = si_get_pi(adev); + struct amdgpu_cac_leakage_table *table = + &adev->pm.dpm.dyn_state.cac_leakage_table; + u32 i; + u32 v0_loadline; + + if (table == NULL) + return -EINVAL; + + *max = 0; + *min = 0xFFFF; + + for (i = 0; i < table->count; i++) { + if (table->entries[i].vddc > *max) + *max = table->entries[i].vddc; + if (table->entries[i].vddc < *min) + *min = table->entries[i].vddc; + } + + if (si_pi->powertune_data->lkge_lut_v0_percent > 100) + return -EINVAL; + + v0_loadline = (*min) * (100 - si_pi->powertune_data->lkge_lut_v0_percent) / 100; + + if (v0_loadline > 0xFFFFUL) + return -EINVAL; + + *min = (u16)v0_loadline; + + if ((*min > *max) || (*max == 0) || (*min == 0)) + return -EINVAL; + + return 0; +} + +static u16 si_get_cac_std_voltage_step(u16 max, u16 min) +{ + return ((max - min) + (SMC_SISLANDS_LKGE_LUT_NUM_OF_VOLT_ENTRIES - 1)) / + SMC_SISLANDS_LKGE_LUT_NUM_OF_VOLT_ENTRIES; +} + +static int si_init_dte_leakage_table(struct amdgpu_device *adev, + PP_SIslands_CacConfig *cac_tables, + u16 vddc_max, u16 vddc_min, u16 vddc_step, + u16 t0, u16 t_step) +{ + struct si_power_info *si_pi = si_get_pi(adev); + u32 leakage; + unsigned int i, j; + s32 t; + u32 smc_leakage; + u32 scaling_factor; + u16 voltage; + + scaling_factor = si_get_smc_power_scaling_factor(adev); + + for (i = 0; i < SMC_SISLANDS_LKGE_LUT_NUM_OF_TEMP_ENTRIES ; i++) { + t = (1000 * (i * t_step + t0)); + + for (j = 0; j < SMC_SISLANDS_LKGE_LUT_NUM_OF_VOLT_ENTRIES; j++) { + voltage = vddc_max - (vddc_step * j); + + si_calculate_leakage_for_v_and_t(adev, + &si_pi->powertune_data->leakage_coefficients, + voltage, + t, + si_pi->dyn_powertune_data.cac_leakage, + &leakage); + + smc_leakage = si_scale_power_for_smc(leakage, scaling_factor) / 4; + + if (smc_leakage > 0xFFFF) + smc_leakage = 0xFFFF; + + cac_tables->cac_lkge_lut[i][SMC_SISLANDS_LKGE_LUT_NUM_OF_VOLT_ENTRIES-1-j] = + cpu_to_be16((u16)smc_leakage); + } + } + return 0; +} + +static int si_init_simplified_leakage_table(struct amdgpu_device *adev, + PP_SIslands_CacConfig *cac_tables, + u16 vddc_max, u16 vddc_min, u16 vddc_step) +{ + struct si_power_info *si_pi = si_get_pi(adev); + u32 leakage; + unsigned int i, j; + u32 smc_leakage; + u32 scaling_factor; + u16 voltage; + + scaling_factor = si_get_smc_power_scaling_factor(adev); + + for (j = 0; j < SMC_SISLANDS_LKGE_LUT_NUM_OF_VOLT_ENTRIES; j++) { + voltage = vddc_max - (vddc_step * j); + + si_calculate_leakage_for_v(adev, + &si_pi->powertune_data->leakage_coefficients, + si_pi->powertune_data->fixed_kt, + voltage, + si_pi->dyn_powertune_data.cac_leakage, + &leakage); + + smc_leakage = si_scale_power_for_smc(leakage, scaling_factor) / 4; + + if (smc_leakage > 0xFFFF) + smc_leakage = 0xFFFF; + + for (i = 0; i < SMC_SISLANDS_LKGE_LUT_NUM_OF_TEMP_ENTRIES ; i++) + cac_tables->cac_lkge_lut[i][SMC_SISLANDS_LKGE_LUT_NUM_OF_VOLT_ENTRIES-1-j] = + cpu_to_be16((u16)smc_leakage); + } + return 0; +} + +static int si_initialize_smc_cac_tables(struct amdgpu_device *adev) +{ + struct ni_power_info *ni_pi = ni_get_pi(adev); + struct si_power_info *si_pi = si_get_pi(adev); + PP_SIslands_CacConfig *cac_tables = NULL; + u16 vddc_max, vddc_min, vddc_step; + u16 t0, t_step; + u32 load_line_slope, reg; + int ret = 0; + u32 ticks_per_us = amdgpu_asic_get_xclk(adev) / 100; + + if (ni_pi->enable_cac == false) + return 0; + + cac_tables = kzalloc(sizeof(PP_SIslands_CacConfig), GFP_KERNEL); + if (!cac_tables) + return -ENOMEM; + + reg = RREG32(CG_CAC_CTRL) & ~CAC_WINDOW_MASK; + reg |= CAC_WINDOW(si_pi->powertune_data->cac_window); + WREG32(CG_CAC_CTRL, reg); + + si_pi->dyn_powertune_data.cac_leakage = adev->pm.dpm.cac_leakage; + si_pi->dyn_powertune_data.dc_pwr_value = + si_pi->powertune_data->dc_cac[NISLANDS_DCCAC_LEVEL_0]; + si_pi->dyn_powertune_data.wintime = si_calculate_cac_wintime(adev); + si_pi->dyn_powertune_data.shift_n = si_pi->powertune_data->shift_n_default; + + si_pi->dyn_powertune_data.leakage_minimum_temperature = 80 * 1000; + + ret = si_get_cac_std_voltage_max_min(adev, &vddc_max, &vddc_min); + if (ret) + goto done_free; + + vddc_step = si_get_cac_std_voltage_step(vddc_max, vddc_min); + vddc_min = vddc_max - (vddc_step * (SMC_SISLANDS_LKGE_LUT_NUM_OF_VOLT_ENTRIES - 1)); + t_step = 4; + t0 = 60; + + if (si_pi->enable_dte || ni_pi->driver_calculate_cac_leakage) + ret = si_init_dte_leakage_table(adev, cac_tables, + vddc_max, vddc_min, vddc_step, + t0, t_step); + else + ret = si_init_simplified_leakage_table(adev, cac_tables, + vddc_max, vddc_min, vddc_step); + if (ret) + goto done_free; + + load_line_slope = ((u32)adev->pm.dpm.load_line_slope << SMC_SISLANDS_SCALE_R) / 100; + + cac_tables->l2numWin_TDP = cpu_to_be32(si_pi->dyn_powertune_data.l2_lta_window_size); + cac_tables->lts_truncate_n = si_pi->dyn_powertune_data.lts_truncate; + cac_tables->SHIFT_N = si_pi->dyn_powertune_data.shift_n; + cac_tables->lkge_lut_V0 = cpu_to_be32((u32)vddc_min); + cac_tables->lkge_lut_Vstep = cpu_to_be32((u32)vddc_step); + cac_tables->R_LL = cpu_to_be32(load_line_slope); + cac_tables->WinTime = cpu_to_be32(si_pi->dyn_powertune_data.wintime); + cac_tables->calculation_repeats = cpu_to_be32(2); + cac_tables->dc_cac = cpu_to_be32(0); + cac_tables->log2_PG_LKG_SCALE = 12; + cac_tables->cac_temp = si_pi->powertune_data->operating_temp; + cac_tables->lkge_lut_T0 = cpu_to_be32((u32)t0); + cac_tables->lkge_lut_Tstep = cpu_to_be32((u32)t_step); + + ret = amdgpu_si_copy_bytes_to_smc(adev, si_pi->cac_table_start, + (u8 *)cac_tables, + sizeof(PP_SIslands_CacConfig), + si_pi->sram_end); + + if (ret) + goto done_free; + + ret = si_write_smc_soft_register(adev, SI_SMC_SOFT_REGISTER_ticks_per_us, ticks_per_us); + +done_free: + if (ret) { + ni_pi->enable_cac = false; + ni_pi->enable_power_containment = false; + } + + kfree(cac_tables); + + return ret; +} + +static int si_program_cac_config_registers(struct amdgpu_device *adev, + const struct si_cac_config_reg *cac_config_regs) +{ + const struct si_cac_config_reg *config_regs = cac_config_regs; + u32 data = 0, offset; + + if (!config_regs) + return -EINVAL; + + while (config_regs->offset != 0xFFFFFFFF) { + switch (config_regs->type) { + case SISLANDS_CACCONFIG_CGIND: + offset = SMC_CG_IND_START + config_regs->offset; + if (offset < SMC_CG_IND_END) + data = RREG32_SMC(offset); + break; + default: + data = RREG32(config_regs->offset); + break; + } + + data &= ~config_regs->mask; + data |= ((config_regs->value << config_regs->shift) & config_regs->mask); + + switch (config_regs->type) { + case SISLANDS_CACCONFIG_CGIND: + offset = SMC_CG_IND_START + config_regs->offset; + if (offset < SMC_CG_IND_END) + WREG32_SMC(offset, data); + break; + default: + WREG32(config_regs->offset, data); + break; + } + config_regs++; + } + return 0; +} + +static int si_initialize_hardware_cac_manager(struct amdgpu_device *adev) +{ + struct ni_power_info *ni_pi = ni_get_pi(adev); + struct si_power_info *si_pi = si_get_pi(adev); + int ret; + + if ((ni_pi->enable_cac == false) || + (ni_pi->cac_configuration_required == false)) + return 0; + + ret = si_program_cac_config_registers(adev, si_pi->lcac_config); + if (ret) + return ret; + ret = si_program_cac_config_registers(adev, si_pi->cac_override); + if (ret) + return ret; + ret = si_program_cac_config_registers(adev, si_pi->cac_weights); + if (ret) + return ret; + + return 0; +} + +static int si_enable_smc_cac(struct amdgpu_device *adev, + struct amdgpu_ps *amdgpu_new_state, + bool enable) +{ + struct ni_power_info *ni_pi = ni_get_pi(adev); + struct si_power_info *si_pi = si_get_pi(adev); + PPSMC_Result smc_result; + int ret = 0; + + if (ni_pi->enable_cac) { + if (enable) { + if (!si_should_disable_uvd_powertune(adev, amdgpu_new_state)) { + if (ni_pi->support_cac_long_term_average) { + smc_result = amdgpu_si_send_msg_to_smc(adev, PPSMC_CACLongTermAvgEnable); + if (smc_result != PPSMC_Result_OK) + ni_pi->support_cac_long_term_average = false; + } + + smc_result = amdgpu_si_send_msg_to_smc(adev, PPSMC_MSG_EnableCac); + if (smc_result != PPSMC_Result_OK) { + ret = -EINVAL; + ni_pi->cac_enabled = false; + } else { + ni_pi->cac_enabled = true; + } + + if (si_pi->enable_dte) { + smc_result = amdgpu_si_send_msg_to_smc(adev, PPSMC_MSG_EnableDTE); + if (smc_result != PPSMC_Result_OK) + ret = -EINVAL; + } + } + } else if (ni_pi->cac_enabled) { + if (si_pi->enable_dte) + smc_result = amdgpu_si_send_msg_to_smc(adev, PPSMC_MSG_DisableDTE); + + smc_result = amdgpu_si_send_msg_to_smc(adev, PPSMC_MSG_DisableCac); + + ni_pi->cac_enabled = false; + + if (ni_pi->support_cac_long_term_average) + smc_result = amdgpu_si_send_msg_to_smc(adev, PPSMC_CACLongTermAvgDisable); + } + } + return ret; +} + +static int si_init_smc_spll_table(struct amdgpu_device *adev) +{ + struct ni_power_info *ni_pi = ni_get_pi(adev); + struct si_power_info *si_pi = si_get_pi(adev); + SMC_SISLANDS_SPLL_DIV_TABLE *spll_table; + SISLANDS_SMC_SCLK_VALUE sclk_params; + u32 fb_div, p_div; + u32 clk_s, clk_v; + u32 sclk = 0; + int ret = 0; + u32 tmp; + int i; + + if (si_pi->spll_table_start == 0) + return -EINVAL; + + spll_table = kzalloc(sizeof(SMC_SISLANDS_SPLL_DIV_TABLE), GFP_KERNEL); + if (spll_table == NULL) + return -ENOMEM; + + for (i = 0; i < 256; i++) { + ret = si_calculate_sclk_params(adev, sclk, &sclk_params); + if (ret) + break; + p_div = (sclk_params.vCG_SPLL_FUNC_CNTL & SPLL_PDIV_A_MASK) >> SPLL_PDIV_A_SHIFT; + fb_div = (sclk_params.vCG_SPLL_FUNC_CNTL_3 & SPLL_FB_DIV_MASK) >> SPLL_FB_DIV_SHIFT; + clk_s = (sclk_params.vCG_SPLL_SPREAD_SPECTRUM & CLK_S_MASK) >> CLK_S_SHIFT; + clk_v = (sclk_params.vCG_SPLL_SPREAD_SPECTRUM_2 & CLK_V_MASK) >> CLK_V_SHIFT; + + fb_div &= ~0x00001FFF; + fb_div >>= 1; + clk_v >>= 6; + + if (p_div & ~(SMC_SISLANDS_SPLL_DIV_TABLE_PDIV_MASK >> SMC_SISLANDS_SPLL_DIV_TABLE_PDIV_SHIFT)) + ret = -EINVAL; + if (fb_div & ~(SMC_SISLANDS_SPLL_DIV_TABLE_FBDIV_MASK >> SMC_SISLANDS_SPLL_DIV_TABLE_FBDIV_SHIFT)) + ret = -EINVAL; + if (clk_s & ~(SMC_SISLANDS_SPLL_DIV_TABLE_CLKS_MASK >> SMC_SISLANDS_SPLL_DIV_TABLE_CLKS_SHIFT)) + ret = -EINVAL; + if (clk_v & ~(SMC_SISLANDS_SPLL_DIV_TABLE_CLKV_MASK >> SMC_SISLANDS_SPLL_DIV_TABLE_CLKV_SHIFT)) + ret = -EINVAL; + + if (ret) + break; + + tmp = ((fb_div << SMC_SISLANDS_SPLL_DIV_TABLE_FBDIV_SHIFT) & SMC_SISLANDS_SPLL_DIV_TABLE_FBDIV_MASK) | + ((p_div << SMC_SISLANDS_SPLL_DIV_TABLE_PDIV_SHIFT) & SMC_SISLANDS_SPLL_DIV_TABLE_PDIV_MASK); + spll_table->freq[i] = cpu_to_be32(tmp); + + tmp = ((clk_v << SMC_SISLANDS_SPLL_DIV_TABLE_CLKV_SHIFT) & SMC_SISLANDS_SPLL_DIV_TABLE_CLKV_MASK) | + ((clk_s << SMC_SISLANDS_SPLL_DIV_TABLE_CLKS_SHIFT) & SMC_SISLANDS_SPLL_DIV_TABLE_CLKS_MASK); + spll_table->ss[i] = cpu_to_be32(tmp); + + sclk += 512; + } + + + if (!ret) + ret = amdgpu_si_copy_bytes_to_smc(adev, si_pi->spll_table_start, + (u8 *)spll_table, + sizeof(SMC_SISLANDS_SPLL_DIV_TABLE), + si_pi->sram_end); + + if (ret) + ni_pi->enable_power_containment = false; + + kfree(spll_table); + + return ret; +} + +struct si_dpm_quirk { + u32 chip_vendor; + u32 chip_device; + u32 subsys_vendor; + u32 subsys_device; + u32 max_sclk; + u32 max_mclk; +}; + +/* cards with dpm stability problems */ +static struct si_dpm_quirk si_dpm_quirk_list[] = { + /* PITCAIRN - https://bugs.freedesktop.org/show_bug.cgi?id=76490 */ + { PCI_VENDOR_ID_ATI, 0x6810, 0x1462, 0x3036, 0, 120000 }, + { PCI_VENDOR_ID_ATI, 0x6811, 0x174b, 0xe271, 0, 120000 }, + { PCI_VENDOR_ID_ATI, 0x6810, 0x174b, 0xe271, 85000, 90000 }, + { PCI_VENDOR_ID_ATI, 0x6811, 0x1462, 0x2015, 0, 120000 }, + { PCI_VENDOR_ID_ATI, 0x6811, 0x1043, 0x2015, 0, 120000 }, + { 0, 0, 0, 0 }, +}; + +static u16 si_get_lower_of_leakage_and_vce_voltage(struct amdgpu_device *adev, + u16 vce_voltage) +{ + u16 highest_leakage = 0; + struct si_power_info *si_pi = si_get_pi(adev); + int i; + + for (i = 0; i < si_pi->leakage_voltage.count; i++){ + if (highest_leakage < si_pi->leakage_voltage.entries[i].voltage) + highest_leakage = si_pi->leakage_voltage.entries[i].voltage; + } + + if (si_pi->leakage_voltage.count && (highest_leakage < vce_voltage)) + return highest_leakage; + + return vce_voltage; +} + +static int si_get_vce_clock_voltage(struct amdgpu_device *adev, + u32 evclk, u32 ecclk, u16 *voltage) +{ + u32 i; + int ret = -EINVAL; + struct amdgpu_vce_clock_voltage_dependency_table *table = + &adev->pm.dpm.dyn_state.vce_clock_voltage_dependency_table; + + if (((evclk == 0) && (ecclk == 0)) || + (table && (table->count == 0))) { + *voltage = 0; + return 0; + } + + for (i = 0; i < table->count; i++) { + if ((evclk <= table->entries[i].evclk) && + (ecclk <= table->entries[i].ecclk)) { + *voltage = table->entries[i].v; + ret = 0; + break; + } + } + + /* if no match return the highest voltage */ + if (ret) + *voltage = table->entries[table->count - 1].v; + + *voltage = si_get_lower_of_leakage_and_vce_voltage(adev, *voltage); + + return ret; +} + +static bool si_dpm_vblank_too_short(struct amdgpu_device *adev) +{ + + u32 vblank_time = amdgpu_dpm_get_vblank_time(adev); + /* we never hit the non-gddr5 limit so disable it */ + u32 switch_limit = adev->mc.vram_type == AMDGPU_VRAM_TYPE_GDDR5 ? 450 : 0; + + if (vblank_time < switch_limit) + return true; + else + return false; + +} + +static int ni_copy_and_switch_arb_sets(struct amdgpu_device *adev, + u32 arb_freq_src, u32 arb_freq_dest) +{ + u32 mc_arb_dram_timing; + u32 mc_arb_dram_timing2; + u32 burst_time; + u32 mc_cg_config; + + switch (arb_freq_src) { + case MC_CG_ARB_FREQ_F0: + mc_arb_dram_timing = RREG32(MC_ARB_DRAM_TIMING); + mc_arb_dram_timing2 = RREG32(MC_ARB_DRAM_TIMING2); + burst_time = (RREG32(MC_ARB_BURST_TIME) & STATE0_MASK) >> STATE0_SHIFT; + break; + case MC_CG_ARB_FREQ_F1: + mc_arb_dram_timing = RREG32(MC_ARB_DRAM_TIMING_1); + mc_arb_dram_timing2 = RREG32(MC_ARB_DRAM_TIMING2_1); + burst_time = (RREG32(MC_ARB_BURST_TIME) & STATE1_MASK) >> STATE1_SHIFT; + break; + case MC_CG_ARB_FREQ_F2: + mc_arb_dram_timing = RREG32(MC_ARB_DRAM_TIMING_2); + mc_arb_dram_timing2 = RREG32(MC_ARB_DRAM_TIMING2_2); + burst_time = (RREG32(MC_ARB_BURST_TIME) & STATE2_MASK) >> STATE2_SHIFT; + break; + case MC_CG_ARB_FREQ_F3: + mc_arb_dram_timing = RREG32(MC_ARB_DRAM_TIMING_3); + mc_arb_dram_timing2 = RREG32(MC_ARB_DRAM_TIMING2_3); + burst_time = (RREG32(MC_ARB_BURST_TIME) & STATE3_MASK) >> STATE3_SHIFT; + break; + default: + return -EINVAL; + } + + switch (arb_freq_dest) { + case MC_CG_ARB_FREQ_F0: + WREG32(MC_ARB_DRAM_TIMING, mc_arb_dram_timing); + WREG32(MC_ARB_DRAM_TIMING2, mc_arb_dram_timing2); + WREG32_P(MC_ARB_BURST_TIME, STATE0(burst_time), ~STATE0_MASK); + break; + case MC_CG_ARB_FREQ_F1: + WREG32(MC_ARB_DRAM_TIMING_1, mc_arb_dram_timing); + WREG32(MC_ARB_DRAM_TIMING2_1, mc_arb_dram_timing2); + WREG32_P(MC_ARB_BURST_TIME, STATE1(burst_time), ~STATE1_MASK); + break; + case MC_CG_ARB_FREQ_F2: + WREG32(MC_ARB_DRAM_TIMING_2, mc_arb_dram_timing); + WREG32(MC_ARB_DRAM_TIMING2_2, mc_arb_dram_timing2); + WREG32_P(MC_ARB_BURST_TIME, STATE2(burst_time), ~STATE2_MASK); + break; + case MC_CG_ARB_FREQ_F3: + WREG32(MC_ARB_DRAM_TIMING_3, mc_arb_dram_timing); + WREG32(MC_ARB_DRAM_TIMING2_3, mc_arb_dram_timing2); + WREG32_P(MC_ARB_BURST_TIME, STATE3(burst_time), ~STATE3_MASK); + break; + default: + return -EINVAL; + } + + mc_cg_config = RREG32(MC_CG_CONFIG) | 0x0000000F; + WREG32(MC_CG_CONFIG, mc_cg_config); + WREG32_P(MC_ARB_CG, CG_ARB_REQ(arb_freq_dest), ~CG_ARB_REQ_MASK); + + return 0; +} + +static void ni_update_current_ps(struct amdgpu_device *adev, + struct amdgpu_ps *rps) +{ + struct si_ps *new_ps = si_get_ps(rps); + struct evergreen_power_info *eg_pi = evergreen_get_pi(adev); + struct ni_power_info *ni_pi = ni_get_pi(adev); + + eg_pi->current_rps = *rps; + ni_pi->current_ps = *new_ps; + eg_pi->current_rps.ps_priv = &ni_pi->current_ps; +} + +static void ni_update_requested_ps(struct amdgpu_device *adev, + struct amdgpu_ps *rps) +{ + struct si_ps *new_ps = si_get_ps(rps); + struct evergreen_power_info *eg_pi = evergreen_get_pi(adev); + struct ni_power_info *ni_pi = ni_get_pi(adev); + + eg_pi->requested_rps = *rps; + ni_pi->requested_ps = *new_ps; + eg_pi->requested_rps.ps_priv = &ni_pi->requested_ps; +} + +static void ni_set_uvd_clock_before_set_eng_clock(struct amdgpu_device *adev, + struct amdgpu_ps *new_ps, + struct amdgpu_ps *old_ps) +{ + struct si_ps *new_state = si_get_ps(new_ps); + struct si_ps *current_state = si_get_ps(old_ps); + + if ((new_ps->vclk == old_ps->vclk) && + (new_ps->dclk == old_ps->dclk)) + return; + + if (new_state->performance_levels[new_state->performance_level_count - 1].sclk >= + current_state->performance_levels[current_state->performance_level_count - 1].sclk) + return; + + amdgpu_asic_set_uvd_clocks(adev, new_ps->vclk, new_ps->dclk); +} + +static void ni_set_uvd_clock_after_set_eng_clock(struct amdgpu_device *adev, + struct amdgpu_ps *new_ps, + struct amdgpu_ps *old_ps) +{ + struct si_ps *new_state = si_get_ps(new_ps); + struct si_ps *current_state = si_get_ps(old_ps); + + if ((new_ps->vclk == old_ps->vclk) && + (new_ps->dclk == old_ps->dclk)) + return; + + if (new_state->performance_levels[new_state->performance_level_count - 1].sclk < + current_state->performance_levels[current_state->performance_level_count - 1].sclk) + return; + + amdgpu_asic_set_uvd_clocks(adev, new_ps->vclk, new_ps->dclk); +} + +static u16 btc_find_voltage(struct atom_voltage_table *table, u16 voltage) +{ + unsigned int i; + + for (i = 0; i < table->count; i++) + if (voltage <= table->entries[i].value) + return table->entries[i].value; + + return table->entries[table->count - 1].value; +} + +static u32 btc_find_valid_clock(struct amdgpu_clock_array *clocks, + u32 max_clock, u32 requested_clock) +{ + unsigned int i; + + if ((clocks == NULL) || (clocks->count == 0)) + return (requested_clock < max_clock) ? requested_clock : max_clock; + + for (i = 0; i < clocks->count; i++) { + if (clocks->values[i] >= requested_clock) + return (clocks->values[i] < max_clock) ? clocks->values[i] : max_clock; + } + + return (clocks->values[clocks->count - 1] < max_clock) ? + clocks->values[clocks->count - 1] : max_clock; +} + +static u32 btc_get_valid_mclk(struct amdgpu_device *adev, + u32 max_mclk, u32 requested_mclk) +{ + return btc_find_valid_clock(&adev->pm.dpm.dyn_state.valid_mclk_values, + max_mclk, requested_mclk); +} + +static u32 btc_get_valid_sclk(struct amdgpu_device *adev, + u32 max_sclk, u32 requested_sclk) +{ + return btc_find_valid_clock(&adev->pm.dpm.dyn_state.valid_sclk_values, + max_sclk, requested_sclk); +} + +static void btc_get_max_clock_from_voltage_dependency_table(struct amdgpu_clock_voltage_dependency_table *table, + u32 *max_clock) +{ + u32 i, clock = 0; + + if ((table == NULL) || (table->count == 0)) { + *max_clock = clock; + return; + } + + for (i = 0; i < table->count; i++) { + if (clock < table->entries[i].clk) + clock = table->entries[i].clk; + } + *max_clock = clock; +} + +static void btc_apply_voltage_dependency_rules(struct amdgpu_clock_voltage_dependency_table *table, + u32 clock, u16 max_voltage, u16 *voltage) +{ + u32 i; + + if ((table == NULL) || (table->count == 0)) + return; + + for (i= 0; i < table->count; i++) { + if (clock <= table->entries[i].clk) { + if (*voltage < table->entries[i].v) + *voltage = (u16)((table->entries[i].v < max_voltage) ? + table->entries[i].v : max_voltage); + return; + } + } + + *voltage = (*voltage > max_voltage) ? *voltage : max_voltage; +} + +static void btc_adjust_clock_combinations(struct amdgpu_device *adev, + const struct amdgpu_clock_and_voltage_limits *max_limits, + struct rv7xx_pl *pl) +{ + + if ((pl->mclk == 0) || (pl->sclk == 0)) + return; + + if (pl->mclk == pl->sclk) + return; + + if (pl->mclk > pl->sclk) { + if (((pl->mclk + (pl->sclk - 1)) / pl->sclk) > adev->pm.dpm.dyn_state.mclk_sclk_ratio) + pl->sclk = btc_get_valid_sclk(adev, + max_limits->sclk, + (pl->mclk + + (adev->pm.dpm.dyn_state.mclk_sclk_ratio - 1)) / + adev->pm.dpm.dyn_state.mclk_sclk_ratio); + } else { + if ((pl->sclk - pl->mclk) > adev->pm.dpm.dyn_state.sclk_mclk_delta) + pl->mclk = btc_get_valid_mclk(adev, + max_limits->mclk, + pl->sclk - + adev->pm.dpm.dyn_state.sclk_mclk_delta); + } +} + +static void btc_apply_voltage_delta_rules(struct amdgpu_device *adev, + u16 max_vddc, u16 max_vddci, + u16 *vddc, u16 *vddci) +{ + struct evergreen_power_info *eg_pi = evergreen_get_pi(adev); + u16 new_voltage; + + if ((0 == *vddc) || (0 == *vddci)) + return; + + if (*vddc > *vddci) { + if ((*vddc - *vddci) > adev->pm.dpm.dyn_state.vddc_vddci_delta) { + new_voltage = btc_find_voltage(&eg_pi->vddci_voltage_table, + (*vddc - adev->pm.dpm.dyn_state.vddc_vddci_delta)); + *vddci = (new_voltage < max_vddci) ? new_voltage : max_vddci; + } + } else { + if ((*vddci - *vddc) > adev->pm.dpm.dyn_state.vddc_vddci_delta) { + new_voltage = btc_find_voltage(&eg_pi->vddc_voltage_table, + (*vddci - adev->pm.dpm.dyn_state.vddc_vddci_delta)); + *vddc = (new_voltage < max_vddc) ? new_voltage : max_vddc; + } + } +} + +static enum amdgpu_pcie_gen r600_get_pcie_gen_support(struct amdgpu_device *adev, + u32 sys_mask, + enum amdgpu_pcie_gen asic_gen, + enum amdgpu_pcie_gen default_gen) +{ + switch (asic_gen) { + case AMDGPU_PCIE_GEN1: + return AMDGPU_PCIE_GEN1; + case AMDGPU_PCIE_GEN2: + return AMDGPU_PCIE_GEN2; + case AMDGPU_PCIE_GEN3: + return AMDGPU_PCIE_GEN3; + default: + if ((sys_mask & DRM_PCIE_SPEED_80) && (default_gen == AMDGPU_PCIE_GEN3)) + return AMDGPU_PCIE_GEN3; + else if ((sys_mask & DRM_PCIE_SPEED_50) && (default_gen == AMDGPU_PCIE_GEN2)) + return AMDGPU_PCIE_GEN2; + else + return AMDGPU_PCIE_GEN1; + } + return AMDGPU_PCIE_GEN1; +} + +static void r600_calculate_u_and_p(u32 i, u32 r_c, u32 p_b, + u32 *p, u32 *u) +{ + u32 b_c = 0; + u32 i_c; + u32 tmp; + + i_c = (i * r_c) / 100; + tmp = i_c >> p_b; + + while (tmp) { + b_c++; + tmp >>= 1; + } + + *u = (b_c + 1) / 2; + *p = i_c / (1 << (2 * (*u))); +} + +static int r600_calculate_at(u32 t, u32 h, u32 fh, u32 fl, u32 *tl, u32 *th) +{ + u32 k, a, ah, al; + u32 t1; + + if ((fl == 0) || (fh == 0) || (fl > fh)) + return -EINVAL; + + k = (100 * fh) / fl; + t1 = (t * (k - 100)); + a = (1000 * (100 * h + t1)) / (10000 + (t1 / 100)); + a = (a + 5) / 10; + ah = ((a * t) + 5000) / 10000; + al = a - ah; + + *th = t - ah; + *tl = t + al; + + return 0; +} + +static bool r600_is_uvd_state(u32 class, u32 class2) +{ + if (class & ATOM_PPLIB_CLASSIFICATION_UVDSTATE) + return true; + if (class & ATOM_PPLIB_CLASSIFICATION_HD2STATE) + return true; + if (class & ATOM_PPLIB_CLASSIFICATION_HDSTATE) + return true; + if (class & ATOM_PPLIB_CLASSIFICATION_SDSTATE) + return true; + if (class2 & ATOM_PPLIB_CLASSIFICATION2_MVC) + return true; + return false; +} + +static u8 rv770_get_memory_module_index(struct amdgpu_device *adev) +{ + return (u8) ((RREG32(BIOS_SCRATCH_4) >> 16) & 0xff); +} + +static void rv770_get_max_vddc(struct amdgpu_device *adev) +{ + struct rv7xx_power_info *pi = rv770_get_pi(adev); + u16 vddc; + + if (amdgpu_atombios_get_max_vddc(adev, 0, 0, &vddc)) + pi->max_vddc = 0; + else + pi->max_vddc = vddc; +} + +static void rv770_get_engine_memory_ss(struct amdgpu_device *adev) +{ + struct rv7xx_power_info *pi = rv770_get_pi(adev); + struct amdgpu_atom_ss ss; + + pi->sclk_ss = amdgpu_atombios_get_asic_ss_info(adev, &ss, + ASIC_INTERNAL_ENGINE_SS, 0); + pi->mclk_ss = amdgpu_atombios_get_asic_ss_info(adev, &ss, + ASIC_INTERNAL_MEMORY_SS, 0); + + if (pi->sclk_ss || pi->mclk_ss) + pi->dynamic_ss = true; + else + pi->dynamic_ss = false; +} + + +static void si_apply_state_adjust_rules(struct amdgpu_device *adev, + struct amdgpu_ps *rps) +{ + struct si_ps *ps = si_get_ps(rps); + struct amdgpu_clock_and_voltage_limits *max_limits; + bool disable_mclk_switching = false; + bool disable_sclk_switching = false; + u32 mclk, sclk; + u16 vddc, vddci, min_vce_voltage = 0; + u32 max_sclk_vddc, max_mclk_vddci, max_mclk_vddc; + u32 max_sclk = 0, max_mclk = 0; + int i; + struct si_dpm_quirk *p = si_dpm_quirk_list; + + /* Apply dpm quirks */ + while (p && p->chip_device != 0) { + if (adev->pdev->vendor == p->chip_vendor && + adev->pdev->device == p->chip_device && + adev->pdev->subsystem_vendor == p->subsys_vendor && + adev->pdev->subsystem_device == p->subsys_device) { + max_sclk = p->max_sclk; + max_mclk = p->max_mclk; + break; + } + ++p; + } + + if (rps->vce_active) { + rps->evclk = adev->pm.dpm.vce_states[adev->pm.dpm.vce_level].evclk; + rps->ecclk = adev->pm.dpm.vce_states[adev->pm.dpm.vce_level].ecclk; + si_get_vce_clock_voltage(adev, rps->evclk, rps->ecclk, + &min_vce_voltage); + } else { + rps->evclk = 0; + rps->ecclk = 0; + } + + if ((adev->pm.dpm.new_active_crtc_count > 1) || + si_dpm_vblank_too_short(adev)) + disable_mclk_switching = true; + + if (rps->vclk || rps->dclk) { + disable_mclk_switching = true; + disable_sclk_switching = true; + } + + if (adev->pm.dpm.ac_power) + max_limits = &adev->pm.dpm.dyn_state.max_clock_voltage_on_ac; + else + max_limits = &adev->pm.dpm.dyn_state.max_clock_voltage_on_dc; + + for (i = ps->performance_level_count - 2; i >= 0; i--) { + if (ps->performance_levels[i].vddc > ps->performance_levels[i+1].vddc) + ps->performance_levels[i].vddc = ps->performance_levels[i+1].vddc; + } + if (adev->pm.dpm.ac_power == false) { + for (i = 0; i < ps->performance_level_count; i++) { + if (ps->performance_levels[i].mclk > max_limits->mclk) + ps->performance_levels[i].mclk = max_limits->mclk; + if (ps->performance_levels[i].sclk > max_limits->sclk) + ps->performance_levels[i].sclk = max_limits->sclk; + if (ps->performance_levels[i].vddc > max_limits->vddc) + ps->performance_levels[i].vddc = max_limits->vddc; + if (ps->performance_levels[i].vddci > max_limits->vddci) + ps->performance_levels[i].vddci = max_limits->vddci; + } + } + + /* limit clocks to max supported clocks based on voltage dependency tables */ + btc_get_max_clock_from_voltage_dependency_table(&adev->pm.dpm.dyn_state.vddc_dependency_on_sclk, + &max_sclk_vddc); + btc_get_max_clock_from_voltage_dependency_table(&adev->pm.dpm.dyn_state.vddci_dependency_on_mclk, + &max_mclk_vddci); + btc_get_max_clock_from_voltage_dependency_table(&adev->pm.dpm.dyn_state.vddc_dependency_on_mclk, + &max_mclk_vddc); + + for (i = 0; i < ps->performance_level_count; i++) { + if (max_sclk_vddc) { + if (ps->performance_levels[i].sclk > max_sclk_vddc) + ps->performance_levels[i].sclk = max_sclk_vddc; + } + if (max_mclk_vddci) { + if (ps->performance_levels[i].mclk > max_mclk_vddci) + ps->performance_levels[i].mclk = max_mclk_vddci; + } + if (max_mclk_vddc) { + if (ps->performance_levels[i].mclk > max_mclk_vddc) + ps->performance_levels[i].mclk = max_mclk_vddc; + } + if (max_mclk) { + if (ps->performance_levels[i].mclk > max_mclk) + ps->performance_levels[i].mclk = max_mclk; + } + if (max_sclk) { + if (ps->performance_levels[i].sclk > max_sclk) + ps->performance_levels[i].sclk = max_sclk; + } + } + + /* XXX validate the min clocks required for display */ + + if (disable_mclk_switching) { + mclk = ps->performance_levels[ps->performance_level_count - 1].mclk; + vddci = ps->performance_levels[ps->performance_level_count - 1].vddci; + } else { + mclk = ps->performance_levels[0].mclk; + vddci = ps->performance_levels[0].vddci; + } + + if (disable_sclk_switching) { + sclk = ps->performance_levels[ps->performance_level_count - 1].sclk; + vddc = ps->performance_levels[ps->performance_level_count - 1].vddc; + } else { + sclk = ps->performance_levels[0].sclk; + vddc = ps->performance_levels[0].vddc; + } + + if (rps->vce_active) { + if (sclk < adev->pm.dpm.vce_states[adev->pm.dpm.vce_level].sclk) + sclk = adev->pm.dpm.vce_states[adev->pm.dpm.vce_level].sclk; + if (mclk < adev->pm.dpm.vce_states[adev->pm.dpm.vce_level].mclk) + mclk = adev->pm.dpm.vce_states[adev->pm.dpm.vce_level].mclk; + } + + /* adjusted low state */ + ps->performance_levels[0].sclk = sclk; + ps->performance_levels[0].mclk = mclk; + ps->performance_levels[0].vddc = vddc; + ps->performance_levels[0].vddci = vddci; + + if (disable_sclk_switching) { + sclk = ps->performance_levels[0].sclk; + for (i = 1; i < ps->performance_level_count; i++) { + if (sclk < ps->performance_levels[i].sclk) + sclk = ps->performance_levels[i].sclk; + } + for (i = 0; i < ps->performance_level_count; i++) { + ps->performance_levels[i].sclk = sclk; + ps->performance_levels[i].vddc = vddc; + } + } else { + for (i = 1; i < ps->performance_level_count; i++) { + if (ps->performance_levels[i].sclk < ps->performance_levels[i - 1].sclk) + ps->performance_levels[i].sclk = ps->performance_levels[i - 1].sclk; + if (ps->performance_levels[i].vddc < ps->performance_levels[i - 1].vddc) + ps->performance_levels[i].vddc = ps->performance_levels[i - 1].vddc; + } + } + + if (disable_mclk_switching) { + mclk = ps->performance_levels[0].mclk; + for (i = 1; i < ps->performance_level_count; i++) { + if (mclk < ps->performance_levels[i].mclk) + mclk = ps->performance_levels[i].mclk; + } + for (i = 0; i < ps->performance_level_count; i++) { + ps->performance_levels[i].mclk = mclk; + ps->performance_levels[i].vddci = vddci; + } + } else { + for (i = 1; i < ps->performance_level_count; i++) { + if (ps->performance_levels[i].mclk < ps->performance_levels[i - 1].mclk) + ps->performance_levels[i].mclk = ps->performance_levels[i - 1].mclk; + if (ps->performance_levels[i].vddci < ps->performance_levels[i - 1].vddci) + ps->performance_levels[i].vddci = ps->performance_levels[i - 1].vddci; + } + } + + for (i = 0; i < ps->performance_level_count; i++) + btc_adjust_clock_combinations(adev, max_limits, + &ps->performance_levels[i]); + + for (i = 0; i < ps->performance_level_count; i++) { + if (ps->performance_levels[i].vddc < min_vce_voltage) + ps->performance_levels[i].vddc = min_vce_voltage; + btc_apply_voltage_dependency_rules(&adev->pm.dpm.dyn_state.vddc_dependency_on_sclk, + ps->performance_levels[i].sclk, + max_limits->vddc, &ps->performance_levels[i].vddc); + btc_apply_voltage_dependency_rules(&adev->pm.dpm.dyn_state.vddci_dependency_on_mclk, + ps->performance_levels[i].mclk, + max_limits->vddci, &ps->performance_levels[i].vddci); + btc_apply_voltage_dependency_rules(&adev->pm.dpm.dyn_state.vddc_dependency_on_mclk, + ps->performance_levels[i].mclk, + max_limits->vddc, &ps->performance_levels[i].vddc); + btc_apply_voltage_dependency_rules(&adev->pm.dpm.dyn_state.vddc_dependency_on_dispclk, + adev->clock.current_dispclk, + max_limits->vddc, &ps->performance_levels[i].vddc); + } + + for (i = 0; i < ps->performance_level_count; i++) { + btc_apply_voltage_delta_rules(adev, + max_limits->vddc, max_limits->vddci, + &ps->performance_levels[i].vddc, + &ps->performance_levels[i].vddci); + } + + ps->dc_compatible = true; + for (i = 0; i < ps->performance_level_count; i++) { + if (ps->performance_levels[i].vddc > adev->pm.dpm.dyn_state.max_clock_voltage_on_dc.vddc) + ps->dc_compatible = false; + } +} + +#if 0 +static int si_read_smc_soft_register(struct amdgpu_device *adev, + u16 reg_offset, u32 *value) +{ + struct si_power_info *si_pi = si_get_pi(adev); + + return amdgpu_si_read_smc_sram_dword(adev, + si_pi->soft_regs_start + reg_offset, value, + si_pi->sram_end); +} +#endif + +static int si_write_smc_soft_register(struct amdgpu_device *adev, + u16 reg_offset, u32 value) +{ + struct si_power_info *si_pi = si_get_pi(adev); + + return amdgpu_si_write_smc_sram_dword(adev, + si_pi->soft_regs_start + reg_offset, + value, si_pi->sram_end); +} + +static bool si_is_special_1gb_platform(struct amdgpu_device *adev) +{ + bool ret = false; + u32 tmp, width, row, column, bank, density; + bool is_memory_gddr5, is_special; + + tmp = RREG32(MC_SEQ_MISC0); + is_memory_gddr5 = (MC_SEQ_MISC0_GDDR5_VALUE == ((tmp & MC_SEQ_MISC0_GDDR5_MASK) >> MC_SEQ_MISC0_GDDR5_SHIFT)); + is_special = (MC_SEQ_MISC0_REV_ID_VALUE == ((tmp & MC_SEQ_MISC0_REV_ID_MASK) >> MC_SEQ_MISC0_REV_ID_SHIFT)) + & (MC_SEQ_MISC0_VEN_ID_VALUE == ((tmp & MC_SEQ_MISC0_VEN_ID_MASK) >> MC_SEQ_MISC0_VEN_ID_SHIFT)); + + WREG32(MC_SEQ_IO_DEBUG_INDEX, 0xb); + width = ((RREG32(MC_SEQ_IO_DEBUG_DATA) >> 1) & 1) ? 16 : 32; + + tmp = RREG32(MC_ARB_RAMCFG); + row = ((tmp & NOOFROWS_MASK) >> NOOFROWS_SHIFT) + 10; + column = ((tmp & NOOFCOLS_MASK) >> NOOFCOLS_SHIFT) + 8; + bank = ((tmp & NOOFBANK_MASK) >> NOOFBANK_SHIFT) + 2; + + density = (1 << (row + column - 20 + bank)) * width; + + if ((adev->pdev->device == 0x6819) && + is_memory_gddr5 && is_special && (density == 0x400)) + ret = true; + + return ret; +} + +static void si_get_leakage_vddc(struct amdgpu_device *adev) +{ + struct si_power_info *si_pi = si_get_pi(adev); + u16 vddc, count = 0; + int i, ret; + + for (i = 0; i < SISLANDS_MAX_LEAKAGE_COUNT; i++) { + ret = amdgpu_atombios_get_leakage_vddc_based_on_leakage_idx(adev, &vddc, SISLANDS_LEAKAGE_INDEX0 + i); + + if (!ret && (vddc > 0) && (vddc != (SISLANDS_LEAKAGE_INDEX0 + i))) { + si_pi->leakage_voltage.entries[count].voltage = vddc; + si_pi->leakage_voltage.entries[count].leakage_index = + SISLANDS_LEAKAGE_INDEX0 + i; + count++; + } + } + si_pi->leakage_voltage.count = count; +} + +static int si_get_leakage_voltage_from_leakage_index(struct amdgpu_device *adev, + u32 index, u16 *leakage_voltage) +{ + struct si_power_info *si_pi = si_get_pi(adev); + int i; + + if (leakage_voltage == NULL) + return -EINVAL; + + if ((index & 0xff00) != 0xff00) + return -EINVAL; + + if ((index & 0xff) > SISLANDS_MAX_LEAKAGE_COUNT + 1) + return -EINVAL; + + if (index < SISLANDS_LEAKAGE_INDEX0) + return -EINVAL; + + for (i = 0; i < si_pi->leakage_voltage.count; i++) { + if (si_pi->leakage_voltage.entries[i].leakage_index == index) { + *leakage_voltage = si_pi->leakage_voltage.entries[i].voltage; + return 0; + } + } + return -EAGAIN; +} + +static void si_set_dpm_event_sources(struct amdgpu_device *adev, u32 sources) +{ + struct rv7xx_power_info *pi = rv770_get_pi(adev); + bool want_thermal_protection; + enum amdgpu_dpm_event_src dpm_event_src; + + switch (sources) { + case 0: + default: + want_thermal_protection = false; + break; + case (1 << AMDGPU_DPM_AUTO_THROTTLE_SRC_THERMAL): + want_thermal_protection = true; + dpm_event_src = AMDGPU_DPM_EVENT_SRC_DIGITAL; + break; + case (1 << AMDGPU_DPM_AUTO_THROTTLE_SRC_EXTERNAL): + want_thermal_protection = true; + dpm_event_src = AMDGPU_DPM_EVENT_SRC_EXTERNAL; + break; + case ((1 << AMDGPU_DPM_AUTO_THROTTLE_SRC_EXTERNAL) | + (1 << AMDGPU_DPM_AUTO_THROTTLE_SRC_THERMAL)): + want_thermal_protection = true; + dpm_event_src = AMDGPU_DPM_EVENT_SRC_DIGIAL_OR_EXTERNAL; + break; + } + + if (want_thermal_protection) { + WREG32_P(CG_THERMAL_CTRL, DPM_EVENT_SRC(dpm_event_src), ~DPM_EVENT_SRC_MASK); + if (pi->thermal_protection) + WREG32_P(GENERAL_PWRMGT, 0, ~THERMAL_PROTECTION_DIS); + } else { + WREG32_P(GENERAL_PWRMGT, THERMAL_PROTECTION_DIS, ~THERMAL_PROTECTION_DIS); + } +} + +static void si_enable_auto_throttle_source(struct amdgpu_device *adev, + enum amdgpu_dpm_auto_throttle_src source, + bool enable) +{ + struct rv7xx_power_info *pi = rv770_get_pi(adev); + + if (enable) { + if (!(pi->active_auto_throttle_sources & (1 << source))) { + pi->active_auto_throttle_sources |= 1 << source; + si_set_dpm_event_sources(adev, pi->active_auto_throttle_sources); + } + } else { + if (pi->active_auto_throttle_sources & (1 << source)) { + pi->active_auto_throttle_sources &= ~(1 << source); + si_set_dpm_event_sources(adev, pi->active_auto_throttle_sources); + } + } +} + +static void si_start_dpm(struct amdgpu_device *adev) +{ + WREG32_P(GENERAL_PWRMGT, GLOBAL_PWRMGT_EN, ~GLOBAL_PWRMGT_EN); +} + +static void si_stop_dpm(struct amdgpu_device *adev) +{ + WREG32_P(GENERAL_PWRMGT, 0, ~GLOBAL_PWRMGT_EN); +} + +static void si_enable_sclk_control(struct amdgpu_device *adev, bool enable) +{ + if (enable) + WREG32_P(SCLK_PWRMGT_CNTL, 0, ~SCLK_PWRMGT_OFF); + else + WREG32_P(SCLK_PWRMGT_CNTL, SCLK_PWRMGT_OFF, ~SCLK_PWRMGT_OFF); + +} + +#if 0 +static int si_notify_hardware_of_thermal_state(struct amdgpu_device *adev, + u32 thermal_level) +{ + PPSMC_Result ret; + + if (thermal_level == 0) { + ret = amdgpu_si_send_msg_to_smc(adev, PPSMC_MSG_EnableThermalInterrupt); + if (ret == PPSMC_Result_OK) + return 0; + else + return -EINVAL; + } + return 0; +} + +static void si_notify_hardware_vpu_recovery_event(struct amdgpu_device *adev) +{ + si_write_smc_soft_register(adev, SI_SMC_SOFT_REGISTER_tdr_is_about_to_happen, true); +} +#endif + +#if 0 +static int si_notify_hw_of_powersource(struct amdgpu_device *adev, bool ac_power) +{ + if (ac_power) + return (amdgpu_si_send_msg_to_smc(adev, PPSMC_MSG_RunningOnAC) == PPSMC_Result_OK) ? + 0 : -EINVAL; + + return 0; +} +#endif + +static PPSMC_Result si_send_msg_to_smc_with_parameter(struct amdgpu_device *adev, + PPSMC_Msg msg, u32 parameter) +{ + WREG32(SMC_SCRATCH0, parameter); + return amdgpu_si_send_msg_to_smc(adev, msg); +} + +static int si_restrict_performance_levels_before_switch(struct amdgpu_device *adev) +{ + if (amdgpu_si_send_msg_to_smc(adev, PPSMC_MSG_NoForcedLevel) != PPSMC_Result_OK) + return -EINVAL; + + return (si_send_msg_to_smc_with_parameter(adev, PPSMC_MSG_SetEnabledLevels, 1) == PPSMC_Result_OK) ? + 0 : -EINVAL; +} + +static int si_dpm_force_performance_level(struct amdgpu_device *adev, + enum amdgpu_dpm_forced_level level) +{ + struct amdgpu_ps *rps = adev->pm.dpm.current_ps; + struct si_ps *ps = si_get_ps(rps); + u32 levels = ps->performance_level_count; + + if (level == AMDGPU_DPM_FORCED_LEVEL_HIGH) { + if (si_send_msg_to_smc_with_parameter(adev, PPSMC_MSG_SetEnabledLevels, levels) != PPSMC_Result_OK) + return -EINVAL; + + if (si_send_msg_to_smc_with_parameter(adev, PPSMC_MSG_SetForcedLevels, 1) != PPSMC_Result_OK) + return -EINVAL; + } else if (level == AMDGPU_DPM_FORCED_LEVEL_LOW) { + if (si_send_msg_to_smc_with_parameter(adev, PPSMC_MSG_SetForcedLevels, 0) != PPSMC_Result_OK) + return -EINVAL; + + if (si_send_msg_to_smc_with_parameter(adev, PPSMC_MSG_SetEnabledLevels, 1) != PPSMC_Result_OK) + return -EINVAL; + } else if (level == AMDGPU_DPM_FORCED_LEVEL_AUTO) { + if (si_send_msg_to_smc_with_parameter(adev, PPSMC_MSG_SetForcedLevels, 0) != PPSMC_Result_OK) + return -EINVAL; + + if (si_send_msg_to_smc_with_parameter(adev, PPSMC_MSG_SetEnabledLevels, levels) != PPSMC_Result_OK) + return -EINVAL; + } + + adev->pm.dpm.forced_level = level; + + return 0; +} + +#if 0 +static int si_set_boot_state(struct amdgpu_device *adev) +{ + return (amdgpu_si_send_msg_to_smc(adev, PPSMC_MSG_SwitchToInitialState) == PPSMC_Result_OK) ? + 0 : -EINVAL; +} +#endif + +static int si_set_sw_state(struct amdgpu_device *adev) +{ + return (amdgpu_si_send_msg_to_smc(adev, PPSMC_MSG_SwitchToSwState) == PPSMC_Result_OK) ? + 0 : -EINVAL; +} + +static int si_halt_smc(struct amdgpu_device *adev) +{ + if (amdgpu_si_send_msg_to_smc(adev, PPSMC_MSG_Halt) != PPSMC_Result_OK) + return -EINVAL; + + return (amdgpu_si_wait_for_smc_inactive(adev) == PPSMC_Result_OK) ? + 0 : -EINVAL; +} + +static int si_resume_smc(struct amdgpu_device *adev) +{ + if (amdgpu_si_send_msg_to_smc(adev, PPSMC_FlushDataCache) != PPSMC_Result_OK) + return -EINVAL; + + return (amdgpu_si_send_msg_to_smc(adev, PPSMC_MSG_Resume) == PPSMC_Result_OK) ? + 0 : -EINVAL; +} + +static void si_dpm_start_smc(struct amdgpu_device *adev) +{ + amdgpu_si_program_jump_on_start(adev); + amdgpu_si_start_smc(adev); + amdgpu_si_smc_clock(adev, true); +} + +static void si_dpm_stop_smc(struct amdgpu_device *adev) +{ + amdgpu_si_reset_smc(adev); + amdgpu_si_smc_clock(adev, false); +} + +static int si_process_firmware_header(struct amdgpu_device *adev) +{ + struct si_power_info *si_pi = si_get_pi(adev); + u32 tmp; + int ret; + + ret = amdgpu_si_read_smc_sram_dword(adev, + SISLANDS_SMC_FIRMWARE_HEADER_LOCATION + + SISLANDS_SMC_FIRMWARE_HEADER_stateTable, + &tmp, si_pi->sram_end); + if (ret) + return ret; + + si_pi->state_table_start = tmp; + + ret = amdgpu_si_read_smc_sram_dword(adev, + SISLANDS_SMC_FIRMWARE_HEADER_LOCATION + + SISLANDS_SMC_FIRMWARE_HEADER_softRegisters, + &tmp, si_pi->sram_end); + if (ret) + return ret; + + si_pi->soft_regs_start = tmp; + + ret = amdgpu_si_read_smc_sram_dword(adev, + SISLANDS_SMC_FIRMWARE_HEADER_LOCATION + + SISLANDS_SMC_FIRMWARE_HEADER_mcRegisterTable, + &tmp, si_pi->sram_end); + if (ret) + return ret; + + si_pi->mc_reg_table_start = tmp; + + ret = amdgpu_si_read_smc_sram_dword(adev, + SISLANDS_SMC_FIRMWARE_HEADER_LOCATION + + SISLANDS_SMC_FIRMWARE_HEADER_fanTable, + &tmp, si_pi->sram_end); + if (ret) + return ret; + + si_pi->fan_table_start = tmp; + + ret = amdgpu_si_read_smc_sram_dword(adev, + SISLANDS_SMC_FIRMWARE_HEADER_LOCATION + + SISLANDS_SMC_FIRMWARE_HEADER_mcArbDramAutoRefreshTable, + &tmp, si_pi->sram_end); + if (ret) + return ret; + + si_pi->arb_table_start = tmp; + + ret = amdgpu_si_read_smc_sram_dword(adev, + SISLANDS_SMC_FIRMWARE_HEADER_LOCATION + + SISLANDS_SMC_FIRMWARE_HEADER_CacConfigTable, + &tmp, si_pi->sram_end); + if (ret) + return ret; + + si_pi->cac_table_start = tmp; + + ret = amdgpu_si_read_smc_sram_dword(adev, + SISLANDS_SMC_FIRMWARE_HEADER_LOCATION + + SISLANDS_SMC_FIRMWARE_HEADER_DteConfiguration, + &tmp, si_pi->sram_end); + if (ret) + return ret; + + si_pi->dte_table_start = tmp; + + ret = amdgpu_si_read_smc_sram_dword(adev, + SISLANDS_SMC_FIRMWARE_HEADER_LOCATION + + SISLANDS_SMC_FIRMWARE_HEADER_spllTable, + &tmp, si_pi->sram_end); + if (ret) + return ret; + + si_pi->spll_table_start = tmp; + + ret = amdgpu_si_read_smc_sram_dword(adev, + SISLANDS_SMC_FIRMWARE_HEADER_LOCATION + + SISLANDS_SMC_FIRMWARE_HEADER_PAPMParameters, + &tmp, si_pi->sram_end); + if (ret) + return ret; + + si_pi->papm_cfg_table_start = tmp; + + return ret; +} + +static void si_read_clock_registers(struct amdgpu_device *adev) +{ + struct si_power_info *si_pi = si_get_pi(adev); + + si_pi->clock_registers.cg_spll_func_cntl = RREG32(CG_SPLL_FUNC_CNTL); + si_pi->clock_registers.cg_spll_func_cntl_2 = RREG32(CG_SPLL_FUNC_CNTL_2); + si_pi->clock_registers.cg_spll_func_cntl_3 = RREG32(CG_SPLL_FUNC_CNTL_3); + si_pi->clock_registers.cg_spll_func_cntl_4 = RREG32(CG_SPLL_FUNC_CNTL_4); + si_pi->clock_registers.cg_spll_spread_spectrum = RREG32(CG_SPLL_SPREAD_SPECTRUM); + si_pi->clock_registers.cg_spll_spread_spectrum_2 = RREG32(CG_SPLL_SPREAD_SPECTRUM_2); + si_pi->clock_registers.dll_cntl = RREG32(DLL_CNTL); + si_pi->clock_registers.mclk_pwrmgt_cntl = RREG32(MCLK_PWRMGT_CNTL); + si_pi->clock_registers.mpll_ad_func_cntl = RREG32(MPLL_AD_FUNC_CNTL); + si_pi->clock_registers.mpll_dq_func_cntl = RREG32(MPLL_DQ_FUNC_CNTL); + si_pi->clock_registers.mpll_func_cntl = RREG32(MPLL_FUNC_CNTL); + si_pi->clock_registers.mpll_func_cntl_1 = RREG32(MPLL_FUNC_CNTL_1); + si_pi->clock_registers.mpll_func_cntl_2 = RREG32(MPLL_FUNC_CNTL_2); + si_pi->clock_registers.mpll_ss1 = RREG32(MPLL_SS1); + si_pi->clock_registers.mpll_ss2 = RREG32(MPLL_SS2); +} + +static void si_enable_thermal_protection(struct amdgpu_device *adev, + bool enable) +{ + if (enable) + WREG32_P(GENERAL_PWRMGT, 0, ~THERMAL_PROTECTION_DIS); + else + WREG32_P(GENERAL_PWRMGT, THERMAL_PROTECTION_DIS, ~THERMAL_PROTECTION_DIS); +} + +static void si_enable_acpi_power_management(struct amdgpu_device *adev) +{ + WREG32_P(GENERAL_PWRMGT, STATIC_PM_EN, ~STATIC_PM_EN); +} + +#if 0 +static int si_enter_ulp_state(struct amdgpu_device *adev) +{ + WREG32(SMC_MESSAGE_0, PPSMC_MSG_SwitchToMinimumPower); + + udelay(25000); + + return 0; +} + +static int si_exit_ulp_state(struct amdgpu_device *adev) +{ + int i; + + WREG32(SMC_MESSAGE_0, PPSMC_MSG_ResumeFromMinimumPower); + + udelay(7000); + + for (i = 0; i < adev->usec_timeout; i++) { + if (RREG32(SMC_RESP_0) == 1) + break; + udelay(1000); + } + + return 0; +} +#endif + +static int si_notify_smc_display_change(struct amdgpu_device *adev, + bool has_display) +{ + PPSMC_Msg msg = has_display ? + PPSMC_MSG_HasDisplay : PPSMC_MSG_NoDisplay; + + return (amdgpu_si_send_msg_to_smc(adev, msg) == PPSMC_Result_OK) ? + 0 : -EINVAL; +} + +static void si_program_response_times(struct amdgpu_device *adev) +{ + u32 voltage_response_time, backbias_response_time, acpi_delay_time, vbi_time_out; + u32 vddc_dly, acpi_dly, vbi_dly; + u32 reference_clock; + + si_write_smc_soft_register(adev, SI_SMC_SOFT_REGISTER_mvdd_chg_time, 1); + + voltage_response_time = (u32)adev->pm.dpm.voltage_response_time; + backbias_response_time = (u32)adev->pm.dpm.backbias_response_time; + + if (voltage_response_time == 0) + voltage_response_time = 1000; + + acpi_delay_time = 15000; + vbi_time_out = 100000; + + reference_clock = amdgpu_asic_get_xclk(adev); + + vddc_dly = (voltage_response_time * reference_clock) / 100; + acpi_dly = (acpi_delay_time * reference_clock) / 100; + vbi_dly = (vbi_time_out * reference_clock) / 100; + + si_write_smc_soft_register(adev, SI_SMC_SOFT_REGISTER_delay_vreg, vddc_dly); + si_write_smc_soft_register(adev, SI_SMC_SOFT_REGISTER_delay_acpi, acpi_dly); + si_write_smc_soft_register(adev, SI_SMC_SOFT_REGISTER_mclk_chg_timeout, vbi_dly); + si_write_smc_soft_register(adev, SI_SMC_SOFT_REGISTER_mc_block_delay, 0xAA); +} + +static void si_program_ds_registers(struct amdgpu_device *adev) +{ + struct evergreen_power_info *eg_pi = evergreen_get_pi(adev); + u32 tmp; + + /* DEEP_SLEEP_CLK_SEL field should be 0x10 on tahiti A0 */ + if (adev->asic_type == CHIP_TAHITI && adev->rev_id == 0x0) + tmp = 0x10; + else + tmp = 0x1; + + if (eg_pi->sclk_deep_sleep) { + WREG32_P(MISC_CLK_CNTL, DEEP_SLEEP_CLK_SEL(tmp), ~DEEP_SLEEP_CLK_SEL_MASK); + WREG32_P(CG_SPLL_AUTOSCALE_CNTL, AUTOSCALE_ON_SS_CLEAR, + ~AUTOSCALE_ON_SS_CLEAR); + } +} + +static void si_program_display_gap(struct amdgpu_device *adev) +{ + u32 tmp, pipe; + int i; + + tmp = RREG32(CG_DISPLAY_GAP_CNTL) & ~(DISP1_GAP_MASK | DISP2_GAP_MASK); + if (adev->pm.dpm.new_active_crtc_count > 0) + tmp |= DISP1_GAP(R600_PM_DISPLAY_GAP_VBLANK_OR_WM); + else + tmp |= DISP1_GAP(R600_PM_DISPLAY_GAP_IGNORE); + + if (adev->pm.dpm.new_active_crtc_count > 1) + tmp |= DISP2_GAP(R600_PM_DISPLAY_GAP_VBLANK_OR_WM); + else + tmp |= DISP2_GAP(R600_PM_DISPLAY_GAP_IGNORE); + + WREG32(CG_DISPLAY_GAP_CNTL, tmp); + + tmp = RREG32(DCCG_DISP_SLOW_SELECT_REG); + pipe = (tmp & DCCG_DISP1_SLOW_SELECT_MASK) >> DCCG_DISP1_SLOW_SELECT_SHIFT; + + if ((adev->pm.dpm.new_active_crtc_count > 0) && + (!(adev->pm.dpm.new_active_crtcs & (1 << pipe)))) { + /* find the first active crtc */ + for (i = 0; i < adev->mode_info.num_crtc; i++) { + if (adev->pm.dpm.new_active_crtcs & (1 << i)) + break; + } + if (i == adev->mode_info.num_crtc) + pipe = 0; + else + pipe = i; + + tmp &= ~DCCG_DISP1_SLOW_SELECT_MASK; + tmp |= DCCG_DISP1_SLOW_SELECT(pipe); + WREG32(DCCG_DISP_SLOW_SELECT_REG, tmp); + } + + /* Setting this to false forces the performance state to low if the crtcs are disabled. + * This can be a problem on PowerXpress systems or if you want to use the card + * for offscreen rendering or compute if there are no crtcs enabled. + */ + si_notify_smc_display_change(adev, adev->pm.dpm.new_active_crtc_count > 0); +} + +static void si_enable_spread_spectrum(struct amdgpu_device *adev, bool enable) +{ + struct rv7xx_power_info *pi = rv770_get_pi(adev); + + if (enable) { + if (pi->sclk_ss) + WREG32_P(GENERAL_PWRMGT, DYN_SPREAD_SPECTRUM_EN, ~DYN_SPREAD_SPECTRUM_EN); + } else { + WREG32_P(CG_SPLL_SPREAD_SPECTRUM, 0, ~SSEN); + WREG32_P(GENERAL_PWRMGT, 0, ~DYN_SPREAD_SPECTRUM_EN); + } +} + +static void si_setup_bsp(struct amdgpu_device *adev) +{ + struct rv7xx_power_info *pi = rv770_get_pi(adev); + u32 xclk = amdgpu_asic_get_xclk(adev); + + r600_calculate_u_and_p(pi->asi, + xclk, + 16, + &pi->bsp, + &pi->bsu); + + r600_calculate_u_and_p(pi->pasi, + xclk, + 16, + &pi->pbsp, + &pi->pbsu); + + + pi->dsp = BSP(pi->bsp) | BSU(pi->bsu); + pi->psp = BSP(pi->pbsp) | BSU(pi->pbsu); + + WREG32(CG_BSP, pi->dsp); +} + +static void si_program_git(struct amdgpu_device *adev) +{ + WREG32_P(CG_GIT, CG_GICST(R600_GICST_DFLT), ~CG_GICST_MASK); +} + +static void si_program_tp(struct amdgpu_device *adev) +{ + int i; + enum r600_td td = R600_TD_DFLT; + + for (i = 0; i < R600_PM_NUMBER_OF_TC; i++) + WREG32(CG_FFCT_0 + i, (UTC_0(r600_utc[i]) | DTC_0(r600_dtc[i]))); + + if (td == R600_TD_AUTO) + WREG32_P(SCLK_PWRMGT_CNTL, 0, ~FIR_FORCE_TREND_SEL); + else + WREG32_P(SCLK_PWRMGT_CNTL, FIR_FORCE_TREND_SEL, ~FIR_FORCE_TREND_SEL); + + if (td == R600_TD_UP) + WREG32_P(SCLK_PWRMGT_CNTL, 0, ~FIR_TREND_MODE); + + if (td == R600_TD_DOWN) + WREG32_P(SCLK_PWRMGT_CNTL, FIR_TREND_MODE, ~FIR_TREND_MODE); +} + +static void si_program_tpp(struct amdgpu_device *adev) +{ + WREG32(CG_TPC, R600_TPC_DFLT); +} + +static void si_program_sstp(struct amdgpu_device *adev) +{ + WREG32(CG_SSP, (SSTU(R600_SSTU_DFLT) | SST(R600_SST_DFLT))); +} + +static void si_enable_display_gap(struct amdgpu_device *adev) +{ + u32 tmp = RREG32(CG_DISPLAY_GAP_CNTL); + + tmp &= ~(DISP1_GAP_MASK | DISP2_GAP_MASK); + tmp |= (DISP1_GAP(R600_PM_DISPLAY_GAP_IGNORE) | + DISP2_GAP(R600_PM_DISPLAY_GAP_IGNORE)); + + tmp &= ~(DISP1_GAP_MCHG_MASK | DISP2_GAP_MCHG_MASK); + tmp |= (DISP1_GAP_MCHG(R600_PM_DISPLAY_GAP_VBLANK) | + DISP2_GAP_MCHG(R600_PM_DISPLAY_GAP_IGNORE)); + WREG32(CG_DISPLAY_GAP_CNTL, tmp); +} + +static void si_program_vc(struct amdgpu_device *adev) +{ + struct rv7xx_power_info *pi = rv770_get_pi(adev); + + WREG32(CG_FTV, pi->vrc); +} + +static void si_clear_vc(struct amdgpu_device *adev) +{ + WREG32(CG_FTV, 0); +} + +static u8 si_get_ddr3_mclk_frequency_ratio(u32 memory_clock) +{ + u8 mc_para_index; + + if (memory_clock < 10000) + mc_para_index = 0; + else if (memory_clock >= 80000) + mc_para_index = 0x0f; + else + mc_para_index = (u8)((memory_clock - 10000) / 5000 + 1); + return mc_para_index; +} + +static u8 si_get_mclk_frequency_ratio(u32 memory_clock, bool strobe_mode) +{ + u8 mc_para_index; + + if (strobe_mode) { + if (memory_clock < 12500) + mc_para_index = 0x00; + else if (memory_clock > 47500) + mc_para_index = 0x0f; + else + mc_para_index = (u8)((memory_clock - 10000) / 2500); + } else { + if (memory_clock < 65000) + mc_para_index = 0x00; + else if (memory_clock > 135000) + mc_para_index = 0x0f; + else + mc_para_index = (u8)((memory_clock - 60000) / 5000); + } + return mc_para_index; +} + +static u8 si_get_strobe_mode_settings(struct amdgpu_device *adev, u32 mclk) +{ + struct rv7xx_power_info *pi = rv770_get_pi(adev); + bool strobe_mode = false; + u8 result = 0; + + if (mclk <= pi->mclk_strobe_mode_threshold) + strobe_mode = true; + + if (adev->mc.vram_type == AMDGPU_VRAM_TYPE_GDDR5) + result = si_get_mclk_frequency_ratio(mclk, strobe_mode); + else + result = si_get_ddr3_mclk_frequency_ratio(mclk); + + if (strobe_mode) + result |= SISLANDS_SMC_STROBE_ENABLE; + + return result; +} + +static int si_upload_firmware(struct amdgpu_device *adev) +{ + struct si_power_info *si_pi = si_get_pi(adev); + + amdgpu_si_reset_smc(adev); + amdgpu_si_smc_clock(adev, false); + + return amdgpu_si_load_smc_ucode(adev, si_pi->sram_end); +} + +static bool si_validate_phase_shedding_tables(struct amdgpu_device *adev, + const struct atom_voltage_table *table, + const struct amdgpu_phase_shedding_limits_table *limits) +{ + u32 data, num_bits, num_levels; + + if ((table == NULL) || (limits == NULL)) + return false; + + data = table->mask_low; + + num_bits = hweight32(data); + + if (num_bits == 0) + return false; + + num_levels = (1 << num_bits); + + if (table->count != num_levels) + return false; + + if (limits->count != (num_levels - 1)) + return false; + + return true; +} + +static void si_trim_voltage_table_to_fit_state_table(struct amdgpu_device *adev, + u32 max_voltage_steps, + struct atom_voltage_table *voltage_table) +{ + unsigned int i, diff; + + if (voltage_table->count <= max_voltage_steps) + return; + + diff = voltage_table->count - max_voltage_steps; + + for (i= 0; i < max_voltage_steps; i++) + voltage_table->entries[i] = voltage_table->entries[i + diff]; + + voltage_table->count = max_voltage_steps; +} + +static int si_get_svi2_voltage_table(struct amdgpu_device *adev, + struct amdgpu_clock_voltage_dependency_table *voltage_dependency_table, + struct atom_voltage_table *voltage_table) +{ + u32 i; + + if (voltage_dependency_table == NULL) + return -EINVAL; + + voltage_table->mask_low = 0; + voltage_table->phase_delay = 0; + + voltage_table->count = voltage_dependency_table->count; + for (i = 0; i < voltage_table->count; i++) { + voltage_table->entries[i].value = voltage_dependency_table->entries[i].v; + voltage_table->entries[i].smio_low = 0; + } + + return 0; +} + +static int si_construct_voltage_tables(struct amdgpu_device *adev) +{ + struct rv7xx_power_info *pi = rv770_get_pi(adev); + struct evergreen_power_info *eg_pi = evergreen_get_pi(adev); + struct si_power_info *si_pi = si_get_pi(adev); + int ret; + + if (pi->voltage_control) { + ret = amdgpu_atombios_get_voltage_table(adev, VOLTAGE_TYPE_VDDC, + VOLTAGE_OBJ_GPIO_LUT, &eg_pi->vddc_voltage_table); + if (ret) + return ret; + + if (eg_pi->vddc_voltage_table.count > SISLANDS_MAX_NO_VREG_STEPS) + si_trim_voltage_table_to_fit_state_table(adev, + SISLANDS_MAX_NO_VREG_STEPS, + &eg_pi->vddc_voltage_table); + } else if (si_pi->voltage_control_svi2) { + ret = si_get_svi2_voltage_table(adev, + &adev->pm.dpm.dyn_state.vddc_dependency_on_mclk, + &eg_pi->vddc_voltage_table); + if (ret) + return ret; + } else { + return -EINVAL; + } + + if (eg_pi->vddci_control) { + ret = amdgpu_atombios_get_voltage_table(adev, VOLTAGE_TYPE_VDDCI, + VOLTAGE_OBJ_GPIO_LUT, &eg_pi->vddci_voltage_table); + if (ret) + return ret; + + if (eg_pi->vddci_voltage_table.count > SISLANDS_MAX_NO_VREG_STEPS) + si_trim_voltage_table_to_fit_state_table(adev, + SISLANDS_MAX_NO_VREG_STEPS, + &eg_pi->vddci_voltage_table); + } + if (si_pi->vddci_control_svi2) { + ret = si_get_svi2_voltage_table(adev, + &adev->pm.dpm.dyn_state.vddci_dependency_on_mclk, + &eg_pi->vddci_voltage_table); + if (ret) + return ret; + } + + if (pi->mvdd_control) { + ret = amdgpu_atombios_get_voltage_table(adev, VOLTAGE_TYPE_MVDDC, + VOLTAGE_OBJ_GPIO_LUT, &si_pi->mvdd_voltage_table); + + if (ret) { + pi->mvdd_control = false; + return ret; + } + + if (si_pi->mvdd_voltage_table.count == 0) { + pi->mvdd_control = false; + return -EINVAL; + } + + if (si_pi->mvdd_voltage_table.count > SISLANDS_MAX_NO_VREG_STEPS) + si_trim_voltage_table_to_fit_state_table(adev, + SISLANDS_MAX_NO_VREG_STEPS, + &si_pi->mvdd_voltage_table); + } + + if (si_pi->vddc_phase_shed_control) { + ret = amdgpu_atombios_get_voltage_table(adev, VOLTAGE_TYPE_VDDC, + VOLTAGE_OBJ_PHASE_LUT, &si_pi->vddc_phase_shed_table); + if (ret) + si_pi->vddc_phase_shed_control = false; + + if ((si_pi->vddc_phase_shed_table.count == 0) || + (si_pi->vddc_phase_shed_table.count > SISLANDS_MAX_NO_VREG_STEPS)) + si_pi->vddc_phase_shed_control = false; + } + + return 0; +} + +static void si_populate_smc_voltage_table(struct amdgpu_device *adev, + const struct atom_voltage_table *voltage_table, + SISLANDS_SMC_STATETABLE *table) +{ + unsigned int i; + + for (i = 0; i < voltage_table->count; i++) + table->lowSMIO[i] |= cpu_to_be32(voltage_table->entries[i].smio_low); +} + +static int si_populate_smc_voltage_tables(struct amdgpu_device *adev, + SISLANDS_SMC_STATETABLE *table) +{ + struct rv7xx_power_info *pi = rv770_get_pi(adev); + struct evergreen_power_info *eg_pi = evergreen_get_pi(adev); + struct si_power_info *si_pi = si_get_pi(adev); + u8 i; + + if (si_pi->voltage_control_svi2) { + si_write_smc_soft_register(adev, SI_SMC_SOFT_REGISTER_svi_rework_gpio_id_svc, + si_pi->svc_gpio_id); + si_write_smc_soft_register(adev, SI_SMC_SOFT_REGISTER_svi_rework_gpio_id_svd, + si_pi->svd_gpio_id); + si_write_smc_soft_register(adev, SI_SMC_SOFT_REGISTER_svi_rework_plat_type, + 2); + } else { + if (eg_pi->vddc_voltage_table.count) { + si_populate_smc_voltage_table(adev, &eg_pi->vddc_voltage_table, table); + table->voltageMaskTable.lowMask[SISLANDS_SMC_VOLTAGEMASK_VDDC] = + cpu_to_be32(eg_pi->vddc_voltage_table.mask_low); + + for (i = 0; i < eg_pi->vddc_voltage_table.count; i++) { + if (pi->max_vddc_in_table <= eg_pi->vddc_voltage_table.entries[i].value) { + table->maxVDDCIndexInPPTable = i; + break; + } + } + } + + if (eg_pi->vddci_voltage_table.count) { + si_populate_smc_voltage_table(adev, &eg_pi->vddci_voltage_table, table); + + table->voltageMaskTable.lowMask[SISLANDS_SMC_VOLTAGEMASK_VDDCI] = + cpu_to_be32(eg_pi->vddci_voltage_table.mask_low); + } + + + if (si_pi->mvdd_voltage_table.count) { + si_populate_smc_voltage_table(adev, &si_pi->mvdd_voltage_table, table); + + table->voltageMaskTable.lowMask[SISLANDS_SMC_VOLTAGEMASK_MVDD] = + cpu_to_be32(si_pi->mvdd_voltage_table.mask_low); + } + + if (si_pi->vddc_phase_shed_control) { + if (si_validate_phase_shedding_tables(adev, &si_pi->vddc_phase_shed_table, + &adev->pm.dpm.dyn_state.phase_shedding_limits_table)) { + si_populate_smc_voltage_table(adev, &si_pi->vddc_phase_shed_table, table); + + table->phaseMaskTable.lowMask[SISLANDS_SMC_VOLTAGEMASK_VDDC] = + cpu_to_be32(si_pi->vddc_phase_shed_table.mask_low); + + si_write_smc_soft_register(adev, SI_SMC_SOFT_REGISTER_phase_shedding_delay, + (u32)si_pi->vddc_phase_shed_table.phase_delay); + } else { + si_pi->vddc_phase_shed_control = false; + } + } + } + + return 0; +} + +static int si_populate_voltage_value(struct amdgpu_device *adev, + const struct atom_voltage_table *table, + u16 value, SISLANDS_SMC_VOLTAGE_VALUE *voltage) +{ + unsigned int i; + + for (i = 0; i < table->count; i++) { + if (value <= table->entries[i].value) { + voltage->index = (u8)i; + voltage->value = cpu_to_be16(table->entries[i].value); + break; + } + } + + if (i >= table->count) + return -EINVAL; + + return 0; +} + +static int si_populate_mvdd_value(struct amdgpu_device *adev, u32 mclk, + SISLANDS_SMC_VOLTAGE_VALUE *voltage) +{ + struct rv7xx_power_info *pi = rv770_get_pi(adev); + struct si_power_info *si_pi = si_get_pi(adev); + + if (pi->mvdd_control) { + if (mclk <= pi->mvdd_split_frequency) + voltage->index = 0; + else + voltage->index = (u8)(si_pi->mvdd_voltage_table.count) - 1; + + voltage->value = cpu_to_be16(si_pi->mvdd_voltage_table.entries[voltage->index].value); + } + return 0; +} + +static int si_get_std_voltage_value(struct amdgpu_device *adev, + SISLANDS_SMC_VOLTAGE_VALUE *voltage, + u16 *std_voltage) +{ + u16 v_index; + bool voltage_found = false; + *std_voltage = be16_to_cpu(voltage->value); + + if (adev->pm.dpm.dyn_state.cac_leakage_table.entries) { + if (adev->pm.dpm.platform_caps & ATOM_PP_PLATFORM_CAP_NEW_CAC_VOLTAGE) { + if (adev->pm.dpm.dyn_state.vddc_dependency_on_sclk.entries == NULL) + return -EINVAL; + + for (v_index = 0; (u32)v_index < adev->pm.dpm.dyn_state.vddc_dependency_on_sclk.count; v_index++) { + if (be16_to_cpu(voltage->value) == + (u16)adev->pm.dpm.dyn_state.vddc_dependency_on_sclk.entries[v_index].v) { + voltage_found = true; + if ((u32)v_index < adev->pm.dpm.dyn_state.cac_leakage_table.count) + *std_voltage = + adev->pm.dpm.dyn_state.cac_leakage_table.entries[v_index].vddc; + else + *std_voltage = + adev->pm.dpm.dyn_state.cac_leakage_table.entries[adev->pm.dpm.dyn_state.cac_leakage_table.count-1].vddc; + break; + } + } + + if (!voltage_found) { + for (v_index = 0; (u32)v_index < adev->pm.dpm.dyn_state.vddc_dependency_on_sclk.count; v_index++) { + if (be16_to_cpu(voltage->value) <= + (u16)adev->pm.dpm.dyn_state.vddc_dependency_on_sclk.entries[v_index].v) { + voltage_found = true; + if ((u32)v_index < adev->pm.dpm.dyn_state.cac_leakage_table.count) + *std_voltage = + adev->pm.dpm.dyn_state.cac_leakage_table.entries[v_index].vddc; + else + *std_voltage = + adev->pm.dpm.dyn_state.cac_leakage_table.entries[adev->pm.dpm.dyn_state.cac_leakage_table.count-1].vddc; + break; + } + } + } + } else { + if ((u32)voltage->index < adev->pm.dpm.dyn_state.cac_leakage_table.count) + *std_voltage = adev->pm.dpm.dyn_state.cac_leakage_table.entries[voltage->index].vddc; + } + } + + return 0; +} + +static int si_populate_std_voltage_value(struct amdgpu_device *adev, + u16 value, u8 index, + SISLANDS_SMC_VOLTAGE_VALUE *voltage) +{ + voltage->index = index; + voltage->value = cpu_to_be16(value); + + return 0; +} + +static int si_populate_phase_shedding_value(struct amdgpu_device *adev, + const struct amdgpu_phase_shedding_limits_table *limits, + u16 voltage, u32 sclk, u32 mclk, + SISLANDS_SMC_VOLTAGE_VALUE *smc_voltage) +{ + unsigned int i; + + for (i = 0; i < limits->count; i++) { + if ((voltage <= limits->entries[i].voltage) && + (sclk <= limits->entries[i].sclk) && + (mclk <= limits->entries[i].mclk)) + break; + } + + smc_voltage->phase_settings = (u8)i; + + return 0; +} + +static int si_init_arb_table_index(struct amdgpu_device *adev) +{ + struct si_power_info *si_pi = si_get_pi(adev); + u32 tmp; + int ret; + + ret = amdgpu_si_read_smc_sram_dword(adev, si_pi->arb_table_start, + &tmp, si_pi->sram_end); + if (ret) + return ret; + + tmp &= 0x00FFFFFF; + tmp |= MC_CG_ARB_FREQ_F1 << 24; + + return amdgpu_si_write_smc_sram_dword(adev, si_pi->arb_table_start, + tmp, si_pi->sram_end); +} + +static int si_initial_switch_from_arb_f0_to_f1(struct amdgpu_device *adev) +{ + return ni_copy_and_switch_arb_sets(adev, MC_CG_ARB_FREQ_F0, MC_CG_ARB_FREQ_F1); +} + +static int si_reset_to_default(struct amdgpu_device *adev) +{ + return (amdgpu_si_send_msg_to_smc(adev, PPSMC_MSG_ResetToDefaults) == PPSMC_Result_OK) ? + 0 : -EINVAL; +} + +static int si_force_switch_to_arb_f0(struct amdgpu_device *adev) +{ + struct si_power_info *si_pi = si_get_pi(adev); + u32 tmp; + int ret; + + ret = amdgpu_si_read_smc_sram_dword(adev, si_pi->arb_table_start, + &tmp, si_pi->sram_end); + if (ret) + return ret; + + tmp = (tmp >> 24) & 0xff; + + if (tmp == MC_CG_ARB_FREQ_F0) + return 0; + + return ni_copy_and_switch_arb_sets(adev, tmp, MC_CG_ARB_FREQ_F0); +} + +static u32 si_calculate_memory_refresh_rate(struct amdgpu_device *adev, + u32 engine_clock) +{ + u32 dram_rows; + u32 dram_refresh_rate; + u32 mc_arb_rfsh_rate; + u32 tmp = (RREG32(MC_ARB_RAMCFG) & NOOFROWS_MASK) >> NOOFROWS_SHIFT; + + if (tmp >= 4) + dram_rows = 16384; + else + dram_rows = 1 << (tmp + 10); + + dram_refresh_rate = 1 << ((RREG32(MC_SEQ_MISC0) & 0x3) + 3); + mc_arb_rfsh_rate = ((engine_clock * 10) * dram_refresh_rate / dram_rows - 32) / 64; + + return mc_arb_rfsh_rate; +} + +static int si_populate_memory_timing_parameters(struct amdgpu_device *adev, + struct rv7xx_pl *pl, + SMC_SIslands_MCArbDramTimingRegisterSet *arb_regs) +{ + u32 dram_timing; + u32 dram_timing2; + u32 burst_time; + + arb_regs->mc_arb_rfsh_rate = + (u8)si_calculate_memory_refresh_rate(adev, pl->sclk); + + amdgpu_atombios_set_engine_dram_timings(adev, + pl->sclk, + pl->mclk); + + dram_timing = RREG32(MC_ARB_DRAM_TIMING); + dram_timing2 = RREG32(MC_ARB_DRAM_TIMING2); + burst_time = RREG32(MC_ARB_BURST_TIME) & STATE0_MASK; + + arb_regs->mc_arb_dram_timing = cpu_to_be32(dram_timing); + arb_regs->mc_arb_dram_timing2 = cpu_to_be32(dram_timing2); + arb_regs->mc_arb_burst_time = (u8)burst_time; + + return 0; +} + +static int si_do_program_memory_timing_parameters(struct amdgpu_device *adev, + struct amdgpu_ps *amdgpu_state, + unsigned int first_arb_set) +{ + struct si_power_info *si_pi = si_get_pi(adev); + struct si_ps *state = si_get_ps(amdgpu_state); + SMC_SIslands_MCArbDramTimingRegisterSet arb_regs = { 0 }; + int i, ret = 0; + + for (i = 0; i < state->performance_level_count; i++) { + ret = si_populate_memory_timing_parameters(adev, &state->performance_levels[i], &arb_regs); + if (ret) + break; + ret = amdgpu_si_copy_bytes_to_smc(adev, + si_pi->arb_table_start + + offsetof(SMC_SIslands_MCArbDramTimingRegisters, data) + + sizeof(SMC_SIslands_MCArbDramTimingRegisterSet) * (first_arb_set + i), + (u8 *)&arb_regs, + sizeof(SMC_SIslands_MCArbDramTimingRegisterSet), + si_pi->sram_end); + if (ret) + break; + } + + return ret; +} + +static int si_program_memory_timing_parameters(struct amdgpu_device *adev, + struct amdgpu_ps *amdgpu_new_state) +{ + return si_do_program_memory_timing_parameters(adev, amdgpu_new_state, + SISLANDS_DRIVER_STATE_ARB_INDEX); +} + +static int si_populate_initial_mvdd_value(struct amdgpu_device *adev, + struct SISLANDS_SMC_VOLTAGE_VALUE *voltage) +{ + struct rv7xx_power_info *pi = rv770_get_pi(adev); + struct si_power_info *si_pi = si_get_pi(adev); + + if (pi->mvdd_control) + return si_populate_voltage_value(adev, &si_pi->mvdd_voltage_table, + si_pi->mvdd_bootup_value, voltage); + + return 0; +} + +static int si_populate_smc_initial_state(struct amdgpu_device *adev, + struct amdgpu_ps *amdgpu_initial_state, + SISLANDS_SMC_STATETABLE *table) +{ + struct si_ps *initial_state = si_get_ps(amdgpu_initial_state); + struct rv7xx_power_info *pi = rv770_get_pi(adev); + struct evergreen_power_info *eg_pi = evergreen_get_pi(adev); + struct si_power_info *si_pi = si_get_pi(adev); + u32 reg; + int ret; + + table->initialState.levels[0].mclk.vDLL_CNTL = + cpu_to_be32(si_pi->clock_registers.dll_cntl); + table->initialState.levels[0].mclk.vMCLK_PWRMGT_CNTL = + cpu_to_be32(si_pi->clock_registers.mclk_pwrmgt_cntl); + table->initialState.levels[0].mclk.vMPLL_AD_FUNC_CNTL = + cpu_to_be32(si_pi->clock_registers.mpll_ad_func_cntl); + table->initialState.levels[0].mclk.vMPLL_DQ_FUNC_CNTL = + cpu_to_be32(si_pi->clock_registers.mpll_dq_func_cntl); + table->initialState.levels[0].mclk.vMPLL_FUNC_CNTL = + cpu_to_be32(si_pi->clock_registers.mpll_func_cntl); + table->initialState.levels[0].mclk.vMPLL_FUNC_CNTL_1 = + cpu_to_be32(si_pi->clock_registers.mpll_func_cntl_1); + table->initialState.levels[0].mclk.vMPLL_FUNC_CNTL_2 = + cpu_to_be32(si_pi->clock_registers.mpll_func_cntl_2); + table->initialState.levels[0].mclk.vMPLL_SS = + cpu_to_be32(si_pi->clock_registers.mpll_ss1); + table->initialState.levels[0].mclk.vMPLL_SS2 = + cpu_to_be32(si_pi->clock_registers.mpll_ss2); + + table->initialState.levels[0].mclk.mclk_value = + cpu_to_be32(initial_state->performance_levels[0].mclk); + + table->initialState.levels[0].sclk.vCG_SPLL_FUNC_CNTL = + cpu_to_be32(si_pi->clock_registers.cg_spll_func_cntl); + table->initialState.levels[0].sclk.vCG_SPLL_FUNC_CNTL_2 = + cpu_to_be32(si_pi->clock_registers.cg_spll_func_cntl_2); + table->initialState.levels[0].sclk.vCG_SPLL_FUNC_CNTL_3 = + cpu_to_be32(si_pi->clock_registers.cg_spll_func_cntl_3); + table->initialState.levels[0].sclk.vCG_SPLL_FUNC_CNTL_4 = + cpu_to_be32(si_pi->clock_registers.cg_spll_func_cntl_4); + table->initialState.levels[0].sclk.vCG_SPLL_SPREAD_SPECTRUM = + cpu_to_be32(si_pi->clock_registers.cg_spll_spread_spectrum); + table->initialState.levels[0].sclk.vCG_SPLL_SPREAD_SPECTRUM_2 = + cpu_to_be32(si_pi->clock_registers.cg_spll_spread_spectrum_2); + + table->initialState.levels[0].sclk.sclk_value = + cpu_to_be32(initial_state->performance_levels[0].sclk); + + table->initialState.levels[0].arbRefreshState = + SISLANDS_INITIAL_STATE_ARB_INDEX; + + table->initialState.levels[0].ACIndex = 0; + + ret = si_populate_voltage_value(adev, &eg_pi->vddc_voltage_table, + initial_state->performance_levels[0].vddc, + &table->initialState.levels[0].vddc); + + if (!ret) { + u16 std_vddc; + + ret = si_get_std_voltage_value(adev, + &table->initialState.levels[0].vddc, + &std_vddc); + if (!ret) + si_populate_std_voltage_value(adev, std_vddc, + table->initialState.levels[0].vddc.index, + &table->initialState.levels[0].std_vddc); + } + + if (eg_pi->vddci_control) + si_populate_voltage_value(adev, + &eg_pi->vddci_voltage_table, + initial_state->performance_levels[0].vddci, + &table->initialState.levels[0].vddci); + + if (si_pi->vddc_phase_shed_control) + si_populate_phase_shedding_value(adev, + &adev->pm.dpm.dyn_state.phase_shedding_limits_table, + initial_state->performance_levels[0].vddc, + initial_state->performance_levels[0].sclk, + initial_state->performance_levels[0].mclk, + &table->initialState.levels[0].vddc); + + si_populate_initial_mvdd_value(adev, &table->initialState.levels[0].mvdd); + + reg = CG_R(0xffff) | CG_L(0); + table->initialState.levels[0].aT = cpu_to_be32(reg); + table->initialState.levels[0].bSP = cpu_to_be32(pi->dsp); + table->initialState.levels[0].gen2PCIE = (u8)si_pi->boot_pcie_gen; + + if (adev->mc.vram_type == AMDGPU_VRAM_TYPE_GDDR5) { + table->initialState.levels[0].strobeMode = + si_get_strobe_mode_settings(adev, + initial_state->performance_levels[0].mclk); + + if (initial_state->performance_levels[0].mclk > pi->mclk_edc_enable_threshold) + table->initialState.levels[0].mcFlags = SISLANDS_SMC_MC_EDC_RD_FLAG | SISLANDS_SMC_MC_EDC_WR_FLAG; + else + table->initialState.levels[0].mcFlags = 0; + } + + table->initialState.levelCount = 1; + + table->initialState.flags |= PPSMC_SWSTATE_FLAG_DC; + + table->initialState.levels[0].dpm2.MaxPS = 0; + table->initialState.levels[0].dpm2.NearTDPDec = 0; + table->initialState.levels[0].dpm2.AboveSafeInc = 0; + table->initialState.levels[0].dpm2.BelowSafeInc = 0; + table->initialState.levels[0].dpm2.PwrEfficiencyRatio = 0; + + reg = MIN_POWER_MASK | MAX_POWER_MASK; + table->initialState.levels[0].SQPowerThrottle = cpu_to_be32(reg); + + reg = MAX_POWER_DELTA_MASK | STI_SIZE_MASK | LTI_RATIO_MASK; + table->initialState.levels[0].SQPowerThrottle_2 = cpu_to_be32(reg); + + return 0; +} + +static int si_populate_smc_acpi_state(struct amdgpu_device *adev, + SISLANDS_SMC_STATETABLE *table) +{ + struct rv7xx_power_info *pi = rv770_get_pi(adev); + struct evergreen_power_info *eg_pi = evergreen_get_pi(adev); + struct si_power_info *si_pi = si_get_pi(adev); + u32 spll_func_cntl = si_pi->clock_registers.cg_spll_func_cntl; + u32 spll_func_cntl_2 = si_pi->clock_registers.cg_spll_func_cntl_2; + u32 spll_func_cntl_3 = si_pi->clock_registers.cg_spll_func_cntl_3; + u32 spll_func_cntl_4 = si_pi->clock_registers.cg_spll_func_cntl_4; + u32 dll_cntl = si_pi->clock_registers.dll_cntl; + u32 mclk_pwrmgt_cntl = si_pi->clock_registers.mclk_pwrmgt_cntl; + u32 mpll_ad_func_cntl = si_pi->clock_registers.mpll_ad_func_cntl; + u32 mpll_dq_func_cntl = si_pi->clock_registers.mpll_dq_func_cntl; + u32 mpll_func_cntl = si_pi->clock_registers.mpll_func_cntl; + u32 mpll_func_cntl_1 = si_pi->clock_registers.mpll_func_cntl_1; + u32 mpll_func_cntl_2 = si_pi->clock_registers.mpll_func_cntl_2; + u32 reg; + int ret; + + table->ACPIState = table->initialState; + + table->ACPIState.flags &= ~PPSMC_SWSTATE_FLAG_DC; + + if (pi->acpi_vddc) { + ret = si_populate_voltage_value(adev, &eg_pi->vddc_voltage_table, + pi->acpi_vddc, &table->ACPIState.levels[0].vddc); + if (!ret) { + u16 std_vddc; + + ret = si_get_std_voltage_value(adev, + &table->ACPIState.levels[0].vddc, &std_vddc); + if (!ret) + si_populate_std_voltage_value(adev, std_vddc, + table->ACPIState.levels[0].vddc.index, + &table->ACPIState.levels[0].std_vddc); + } + table->ACPIState.levels[0].gen2PCIE = si_pi->acpi_pcie_gen; + + if (si_pi->vddc_phase_shed_control) { + si_populate_phase_shedding_value(adev, + &adev->pm.dpm.dyn_state.phase_shedding_limits_table, + pi->acpi_vddc, + 0, + 0, + &table->ACPIState.levels[0].vddc); + } + } else { + ret = si_populate_voltage_value(adev, &eg_pi->vddc_voltage_table, + pi->min_vddc_in_table, &table->ACPIState.levels[0].vddc); + if (!ret) { + u16 std_vddc; + + ret = si_get_std_voltage_value(adev, + &table->ACPIState.levels[0].vddc, &std_vddc); + + if (!ret) + si_populate_std_voltage_value(adev, std_vddc, + table->ACPIState.levels[0].vddc.index, + &table->ACPIState.levels[0].std_vddc); + } + table->ACPIState.levels[0].gen2PCIE = (u8)r600_get_pcie_gen_support(adev, + si_pi->sys_pcie_mask, + si_pi->boot_pcie_gen, + AMDGPU_PCIE_GEN1); + + if (si_pi->vddc_phase_shed_control) + si_populate_phase_shedding_value(adev, + &adev->pm.dpm.dyn_state.phase_shedding_limits_table, + pi->min_vddc_in_table, + 0, + 0, + &table->ACPIState.levels[0].vddc); + } + + if (pi->acpi_vddc) { + if (eg_pi->acpi_vddci) + si_populate_voltage_value(adev, &eg_pi->vddci_voltage_table, + eg_pi->acpi_vddci, + &table->ACPIState.levels[0].vddci); + } + + mclk_pwrmgt_cntl |= MRDCK0_RESET | MRDCK1_RESET; + mclk_pwrmgt_cntl &= ~(MRDCK0_PDNB | MRDCK1_PDNB); + + dll_cntl &= ~(MRDCK0_BYPASS | MRDCK1_BYPASS); + + spll_func_cntl_2 &= ~SCLK_MUX_SEL_MASK; + spll_func_cntl_2 |= SCLK_MUX_SEL(4); + + table->ACPIState.levels[0].mclk.vDLL_CNTL = + cpu_to_be32(dll_cntl); + table->ACPIState.levels[0].mclk.vMCLK_PWRMGT_CNTL = + cpu_to_be32(mclk_pwrmgt_cntl); + table->ACPIState.levels[0].mclk.vMPLL_AD_FUNC_CNTL = + cpu_to_be32(mpll_ad_func_cntl); + table->ACPIState.levels[0].mclk.vMPLL_DQ_FUNC_CNTL = + cpu_to_be32(mpll_dq_func_cntl); + table->ACPIState.levels[0].mclk.vMPLL_FUNC_CNTL = + cpu_to_be32(mpll_func_cntl); + table->ACPIState.levels[0].mclk.vMPLL_FUNC_CNTL_1 = + cpu_to_be32(mpll_func_cntl_1); + table->ACPIState.levels[0].mclk.vMPLL_FUNC_CNTL_2 = + cpu_to_be32(mpll_func_cntl_2); + table->ACPIState.levels[0].mclk.vMPLL_SS = + cpu_to_be32(si_pi->clock_registers.mpll_ss1); + table->ACPIState.levels[0].mclk.vMPLL_SS2 = + cpu_to_be32(si_pi->clock_registers.mpll_ss2); + + table->ACPIState.levels[0].sclk.vCG_SPLL_FUNC_CNTL = + cpu_to_be32(spll_func_cntl); + table->ACPIState.levels[0].sclk.vCG_SPLL_FUNC_CNTL_2 = + cpu_to_be32(spll_func_cntl_2); + table->ACPIState.levels[0].sclk.vCG_SPLL_FUNC_CNTL_3 = + cpu_to_be32(spll_func_cntl_3); + table->ACPIState.levels[0].sclk.vCG_SPLL_FUNC_CNTL_4 = + cpu_to_be32(spll_func_cntl_4); + + table->ACPIState.levels[0].mclk.mclk_value = 0; + table->ACPIState.levels[0].sclk.sclk_value = 0; + + si_populate_mvdd_value(adev, 0, &table->ACPIState.levels[0].mvdd); + + if (eg_pi->dynamic_ac_timing) + table->ACPIState.levels[0].ACIndex = 0; + + table->ACPIState.levels[0].dpm2.MaxPS = 0; + table->ACPIState.levels[0].dpm2.NearTDPDec = 0; + table->ACPIState.levels[0].dpm2.AboveSafeInc = 0; + table->ACPIState.levels[0].dpm2.BelowSafeInc = 0; + table->ACPIState.levels[0].dpm2.PwrEfficiencyRatio = 0; + + reg = MIN_POWER_MASK | MAX_POWER_MASK; + table->ACPIState.levels[0].SQPowerThrottle = cpu_to_be32(reg); + + reg = MAX_POWER_DELTA_MASK | STI_SIZE_MASK | LTI_RATIO_MASK; + table->ACPIState.levels[0].SQPowerThrottle_2 = cpu_to_be32(reg); + + return 0; +} + +static int si_populate_ulv_state(struct amdgpu_device *adev, + SISLANDS_SMC_SWSTATE *state) +{ + struct evergreen_power_info *eg_pi = evergreen_get_pi(adev); + struct si_power_info *si_pi = si_get_pi(adev); + struct si_ulv_param *ulv = &si_pi->ulv; + u32 sclk_in_sr = 1350; /* ??? */ + int ret; + + ret = si_convert_power_level_to_smc(adev, &ulv->pl, + &state->levels[0]); + if (!ret) { + if (eg_pi->sclk_deep_sleep) { + if (sclk_in_sr <= SCLK_MIN_DEEPSLEEP_FREQ) + state->levels[0].stateFlags |= PPSMC_STATEFLAG_DEEPSLEEP_BYPASS; + else + state->levels[0].stateFlags |= PPSMC_STATEFLAG_DEEPSLEEP_THROTTLE; + } + if (ulv->one_pcie_lane_in_ulv) + state->flags |= PPSMC_SWSTATE_FLAG_PCIE_X1; + state->levels[0].arbRefreshState = (u8)(SISLANDS_ULV_STATE_ARB_INDEX); + state->levels[0].ACIndex = 1; + state->levels[0].std_vddc = state->levels[0].vddc; + state->levelCount = 1; + + state->flags |= PPSMC_SWSTATE_FLAG_DC; + } + + return ret; +} + +static int si_program_ulv_memory_timing_parameters(struct amdgpu_device *adev) +{ + struct si_power_info *si_pi = si_get_pi(adev); + struct si_ulv_param *ulv = &si_pi->ulv; + SMC_SIslands_MCArbDramTimingRegisterSet arb_regs = { 0 }; + int ret; + + ret = si_populate_memory_timing_parameters(adev, &ulv->pl, + &arb_regs); + if (ret) + return ret; + + si_write_smc_soft_register(adev, SI_SMC_SOFT_REGISTER_ulv_volt_change_delay, + ulv->volt_change_delay); + + ret = amdgpu_si_copy_bytes_to_smc(adev, + si_pi->arb_table_start + + offsetof(SMC_SIslands_MCArbDramTimingRegisters, data) + + sizeof(SMC_SIslands_MCArbDramTimingRegisterSet) * SISLANDS_ULV_STATE_ARB_INDEX, + (u8 *)&arb_regs, + sizeof(SMC_SIslands_MCArbDramTimingRegisterSet), + si_pi->sram_end); + + return ret; +} + +static void si_get_mvdd_configuration(struct amdgpu_device *adev) +{ + struct rv7xx_power_info *pi = rv770_get_pi(adev); + + pi->mvdd_split_frequency = 30000; +} + +static int si_init_smc_table(struct amdgpu_device *adev) +{ + struct si_power_info *si_pi = si_get_pi(adev); + struct amdgpu_ps *amdgpu_boot_state = adev->pm.dpm.boot_ps; + const struct si_ulv_param *ulv = &si_pi->ulv; + SISLANDS_SMC_STATETABLE *table = &si_pi->smc_statetable; + int ret; + u32 lane_width; + u32 vr_hot_gpio; + + si_populate_smc_voltage_tables(adev, table); + + switch (adev->pm.int_thermal_type) { + case THERMAL_TYPE_SI: + case THERMAL_TYPE_EMC2103_WITH_INTERNAL: + table->thermalProtectType = PPSMC_THERMAL_PROTECT_TYPE_INTERNAL; + break; + case THERMAL_TYPE_NONE: + table->thermalProtectType = PPSMC_THERMAL_PROTECT_TYPE_NONE; + break; + default: + table->thermalProtectType = PPSMC_THERMAL_PROTECT_TYPE_EXTERNAL; + break; + } + + if (adev->pm.dpm.platform_caps & ATOM_PP_PLATFORM_CAP_HARDWAREDC) + table->systemFlags |= PPSMC_SYSTEMFLAG_GPIO_DC; + + if (adev->pm.dpm.platform_caps & ATOM_PP_PLATFORM_CAP_REGULATOR_HOT) { + if ((adev->pdev->device != 0x6818) && (adev->pdev->device != 0x6819)) + table->systemFlags |= PPSMC_SYSTEMFLAG_REGULATOR_HOT; + } + + if (adev->pm.dpm.platform_caps & ATOM_PP_PLATFORM_CAP_STEPVDDC) + table->systemFlags |= PPSMC_SYSTEMFLAG_STEPVDDC; + + if (adev->mc.vram_type == AMDGPU_VRAM_TYPE_GDDR5) + table->systemFlags |= PPSMC_SYSTEMFLAG_GDDR5; + + if (adev->pm.dpm.platform_caps & ATOM_PP_PLATFORM_CAP_REVERT_GPIO5_POLARITY) + table->extraFlags |= PPSMC_EXTRAFLAGS_AC2DC_GPIO5_POLARITY_HIGH; + + if (adev->pm.dpm.platform_caps & ATOM_PP_PLATFORM_CAP_VRHOT_GPIO_CONFIGURABLE) { + table->systemFlags |= PPSMC_SYSTEMFLAG_REGULATOR_HOT_PROG_GPIO; + vr_hot_gpio = adev->pm.dpm.backbias_response_time; + si_write_smc_soft_register(adev, SI_SMC_SOFT_REGISTER_vr_hot_gpio, + vr_hot_gpio); + } + + ret = si_populate_smc_initial_state(adev, amdgpu_boot_state, table); + if (ret) + return ret; + + ret = si_populate_smc_acpi_state(adev, table); + if (ret) + return ret; + + table->driverState = table->initialState; + + ret = si_do_program_memory_timing_parameters(adev, amdgpu_boot_state, + SISLANDS_INITIAL_STATE_ARB_INDEX); + if (ret) + return ret; + + if (ulv->supported && ulv->pl.vddc) { + ret = si_populate_ulv_state(adev, &table->ULVState); + if (ret) + return ret; + + ret = si_program_ulv_memory_timing_parameters(adev); + if (ret) + return ret; + + WREG32(CG_ULV_CONTROL, ulv->cg_ulv_control); + WREG32(CG_ULV_PARAMETER, ulv->cg_ulv_parameter); + + lane_width = amdgpu_get_pcie_lanes(adev); + si_write_smc_soft_register(adev, SI_SMC_SOFT_REGISTER_non_ulv_pcie_link_width, lane_width); + } else { + table->ULVState = table->initialState; + } + + return amdgpu_si_copy_bytes_to_smc(adev, si_pi->state_table_start, + (u8 *)table, sizeof(SISLANDS_SMC_STATETABLE), + si_pi->sram_end); +} + +static int si_calculate_sclk_params(struct amdgpu_device *adev, + u32 engine_clock, + SISLANDS_SMC_SCLK_VALUE *sclk) +{ + struct rv7xx_power_info *pi = rv770_get_pi(adev); + struct si_power_info *si_pi = si_get_pi(adev); + struct atom_clock_dividers dividers; + u32 spll_func_cntl = si_pi->clock_registers.cg_spll_func_cntl; + u32 spll_func_cntl_2 = si_pi->clock_registers.cg_spll_func_cntl_2; + u32 spll_func_cntl_3 = si_pi->clock_registers.cg_spll_func_cntl_3; + u32 spll_func_cntl_4 = si_pi->clock_registers.cg_spll_func_cntl_4; + u32 cg_spll_spread_spectrum = si_pi->clock_registers.cg_spll_spread_spectrum; + u32 cg_spll_spread_spectrum_2 = si_pi->clock_registers.cg_spll_spread_spectrum_2; + u64 tmp; + u32 reference_clock = adev->clock.spll.reference_freq; + u32 reference_divider; + u32 fbdiv; + int ret; + + ret = amdgpu_atombios_get_clock_dividers(adev, COMPUTE_ENGINE_PLL_PARAM, + engine_clock, false, ÷rs); + if (ret) + return ret; + + reference_divider = 1 + dividers.ref_div; + + tmp = (u64) engine_clock * reference_divider * dividers.post_div * 16384; + do_div(tmp, reference_clock); + fbdiv = (u32) tmp; + + spll_func_cntl &= ~(SPLL_PDIV_A_MASK | SPLL_REF_DIV_MASK); + spll_func_cntl |= SPLL_REF_DIV(dividers.ref_div); + spll_func_cntl |= SPLL_PDIV_A(dividers.post_div); + + spll_func_cntl_2 &= ~SCLK_MUX_SEL_MASK; + spll_func_cntl_2 |= SCLK_MUX_SEL(2); + + spll_func_cntl_3 &= ~SPLL_FB_DIV_MASK; + spll_func_cntl_3 |= SPLL_FB_DIV(fbdiv); + spll_func_cntl_3 |= SPLL_DITHEN; + + if (pi->sclk_ss) { + struct amdgpu_atom_ss ss; + u32 vco_freq = engine_clock * dividers.post_div; + + if (amdgpu_atombios_get_asic_ss_info(adev, &ss, + ASIC_INTERNAL_ENGINE_SS, vco_freq)) { + u32 clk_s = reference_clock * 5 / (reference_divider * ss.rate); + u32 clk_v = 4 * ss.percentage * fbdiv / (clk_s * 10000); + + cg_spll_spread_spectrum &= ~CLK_S_MASK; + cg_spll_spread_spectrum |= CLK_S(clk_s); + cg_spll_spread_spectrum |= SSEN; + + cg_spll_spread_spectrum_2 &= ~CLK_V_MASK; + cg_spll_spread_spectrum_2 |= CLK_V(clk_v); + } + } + + sclk->sclk_value = engine_clock; + sclk->vCG_SPLL_FUNC_CNTL = spll_func_cntl; + sclk->vCG_SPLL_FUNC_CNTL_2 = spll_func_cntl_2; + sclk->vCG_SPLL_FUNC_CNTL_3 = spll_func_cntl_3; + sclk->vCG_SPLL_FUNC_CNTL_4 = spll_func_cntl_4; + sclk->vCG_SPLL_SPREAD_SPECTRUM = cg_spll_spread_spectrum; + sclk->vCG_SPLL_SPREAD_SPECTRUM_2 = cg_spll_spread_spectrum_2; + + return 0; +} + +static int si_populate_sclk_value(struct amdgpu_device *adev, + u32 engine_clock, + SISLANDS_SMC_SCLK_VALUE *sclk) +{ + SISLANDS_SMC_SCLK_VALUE sclk_tmp; + int ret; + + ret = si_calculate_sclk_params(adev, engine_clock, &sclk_tmp); + if (!ret) { + sclk->sclk_value = cpu_to_be32(sclk_tmp.sclk_value); + sclk->vCG_SPLL_FUNC_CNTL = cpu_to_be32(sclk_tmp.vCG_SPLL_FUNC_CNTL); + sclk->vCG_SPLL_FUNC_CNTL_2 = cpu_to_be32(sclk_tmp.vCG_SPLL_FUNC_CNTL_2); + sclk->vCG_SPLL_FUNC_CNTL_3 = cpu_to_be32(sclk_tmp.vCG_SPLL_FUNC_CNTL_3); + sclk->vCG_SPLL_FUNC_CNTL_4 = cpu_to_be32(sclk_tmp.vCG_SPLL_FUNC_CNTL_4); + sclk->vCG_SPLL_SPREAD_SPECTRUM = cpu_to_be32(sclk_tmp.vCG_SPLL_SPREAD_SPECTRUM); + sclk->vCG_SPLL_SPREAD_SPECTRUM_2 = cpu_to_be32(sclk_tmp.vCG_SPLL_SPREAD_SPECTRUM_2); + } + + return ret; +} + +static int si_populate_mclk_value(struct amdgpu_device *adev, + u32 engine_clock, + u32 memory_clock, + SISLANDS_SMC_MCLK_VALUE *mclk, + bool strobe_mode, + bool dll_state_on) +{ + struct rv7xx_power_info *pi = rv770_get_pi(adev); + struct si_power_info *si_pi = si_get_pi(adev); + u32 dll_cntl = si_pi->clock_registers.dll_cntl; + u32 mclk_pwrmgt_cntl = si_pi->clock_registers.mclk_pwrmgt_cntl; + u32 mpll_ad_func_cntl = si_pi->clock_registers.mpll_ad_func_cntl; + u32 mpll_dq_func_cntl = si_pi->clock_registers.mpll_dq_func_cntl; + u32 mpll_func_cntl = si_pi->clock_registers.mpll_func_cntl; + u32 mpll_func_cntl_1 = si_pi->clock_registers.mpll_func_cntl_1; + u32 mpll_func_cntl_2 = si_pi->clock_registers.mpll_func_cntl_2; + u32 mpll_ss1 = si_pi->clock_registers.mpll_ss1; + u32 mpll_ss2 = si_pi->clock_registers.mpll_ss2; + struct atom_mpll_param mpll_param; + int ret; + + ret = amdgpu_atombios_get_memory_pll_dividers(adev, memory_clock, strobe_mode, &mpll_param); + if (ret) + return ret; + + mpll_func_cntl &= ~BWCTRL_MASK; + mpll_func_cntl |= BWCTRL(mpll_param.bwcntl); + + mpll_func_cntl_1 &= ~(CLKF_MASK | CLKFRAC_MASK | VCO_MODE_MASK); + mpll_func_cntl_1 |= CLKF(mpll_param.clkf) | + CLKFRAC(mpll_param.clkfrac) | VCO_MODE(mpll_param.vco_mode); + + mpll_ad_func_cntl &= ~YCLK_POST_DIV_MASK; + mpll_ad_func_cntl |= YCLK_POST_DIV(mpll_param.post_div); + + if (adev->mc.vram_type == AMDGPU_VRAM_TYPE_GDDR5) { + mpll_dq_func_cntl &= ~(YCLK_SEL_MASK | YCLK_POST_DIV_MASK); + mpll_dq_func_cntl |= YCLK_SEL(mpll_param.yclk_sel) | + YCLK_POST_DIV(mpll_param.post_div); + } + + if (pi->mclk_ss) { + struct amdgpu_atom_ss ss; + u32 freq_nom; + u32 tmp; + u32 reference_clock = adev->clock.mpll.reference_freq; + + if (adev->mc.vram_type == AMDGPU_VRAM_TYPE_GDDR5) + freq_nom = memory_clock * 4; + else + freq_nom = memory_clock * 2; + + tmp = freq_nom / reference_clock; + tmp = tmp * tmp; + if (amdgpu_atombios_get_asic_ss_info(adev, &ss, + ASIC_INTERNAL_MEMORY_SS, freq_nom)) { + u32 clks = reference_clock * 5 / ss.rate; + u32 clkv = (u32)((((131 * ss.percentage * ss.rate) / 100) * tmp) / freq_nom); + + mpll_ss1 &= ~CLKV_MASK; + mpll_ss1 |= CLKV(clkv); + + mpll_ss2 &= ~CLKS_MASK; + mpll_ss2 |= CLKS(clks); + } + } + + mclk_pwrmgt_cntl &= ~DLL_SPEED_MASK; + mclk_pwrmgt_cntl |= DLL_SPEED(mpll_param.dll_speed); + + if (dll_state_on) + mclk_pwrmgt_cntl |= MRDCK0_PDNB | MRDCK1_PDNB; + else + mclk_pwrmgt_cntl &= ~(MRDCK0_PDNB | MRDCK1_PDNB); + + mclk->mclk_value = cpu_to_be32(memory_clock); + mclk->vMPLL_FUNC_CNTL = cpu_to_be32(mpll_func_cntl); + mclk->vMPLL_FUNC_CNTL_1 = cpu_to_be32(mpll_func_cntl_1); + mclk->vMPLL_FUNC_CNTL_2 = cpu_to_be32(mpll_func_cntl_2); + mclk->vMPLL_AD_FUNC_CNTL = cpu_to_be32(mpll_ad_func_cntl); + mclk->vMPLL_DQ_FUNC_CNTL = cpu_to_be32(mpll_dq_func_cntl); + mclk->vMCLK_PWRMGT_CNTL = cpu_to_be32(mclk_pwrmgt_cntl); + mclk->vDLL_CNTL = cpu_to_be32(dll_cntl); + mclk->vMPLL_SS = cpu_to_be32(mpll_ss1); + mclk->vMPLL_SS2 = cpu_to_be32(mpll_ss2); + + return 0; +} + +static void si_populate_smc_sp(struct amdgpu_device *adev, + struct amdgpu_ps *amdgpu_state, + SISLANDS_SMC_SWSTATE *smc_state) +{ + struct si_ps *ps = si_get_ps(amdgpu_state); + struct rv7xx_power_info *pi = rv770_get_pi(adev); + int i; + + for (i = 0; i < ps->performance_level_count - 1; i++) + smc_state->levels[i].bSP = cpu_to_be32(pi->dsp); + + smc_state->levels[ps->performance_level_count - 1].bSP = + cpu_to_be32(pi->psp); +} + +static int si_convert_power_level_to_smc(struct amdgpu_device *adev, + struct rv7xx_pl *pl, + SISLANDS_SMC_HW_PERFORMANCE_LEVEL *level) +{ + struct rv7xx_power_info *pi = rv770_get_pi(adev); + struct evergreen_power_info *eg_pi = evergreen_get_pi(adev); + struct si_power_info *si_pi = si_get_pi(adev); + int ret; + bool dll_state_on; + u16 std_vddc; + bool gmc_pg = false; + + if (eg_pi->pcie_performance_request && + (si_pi->force_pcie_gen != AMDGPU_PCIE_GEN_INVALID)) + level->gen2PCIE = (u8)si_pi->force_pcie_gen; + else + level->gen2PCIE = (u8)pl->pcie_gen; + + ret = si_populate_sclk_value(adev, pl->sclk, &level->sclk); + if (ret) + return ret; + + level->mcFlags = 0; + + if (pi->mclk_stutter_mode_threshold && + (pl->mclk <= pi->mclk_stutter_mode_threshold) && + !eg_pi->uvd_enabled && + (RREG32(DPG_PIPE_STUTTER_CONTROL) & STUTTER_ENABLE) && + (adev->pm.dpm.new_active_crtc_count <= 2)) { + level->mcFlags |= SISLANDS_SMC_MC_STUTTER_EN; + + if (gmc_pg) + level->mcFlags |= SISLANDS_SMC_MC_PG_EN; + } + + if (adev->mc.vram_type == AMDGPU_VRAM_TYPE_GDDR5) { + if (pl->mclk > pi->mclk_edc_enable_threshold) + level->mcFlags |= SISLANDS_SMC_MC_EDC_RD_FLAG; + + if (pl->mclk > eg_pi->mclk_edc_wr_enable_threshold) + level->mcFlags |= SISLANDS_SMC_MC_EDC_WR_FLAG; + + level->strobeMode = si_get_strobe_mode_settings(adev, pl->mclk); + + if (level->strobeMode & SISLANDS_SMC_STROBE_ENABLE) { + if (si_get_mclk_frequency_ratio(pl->mclk, true) >= + ((RREG32(MC_SEQ_MISC7) >> 16) & 0xf)) + dll_state_on = ((RREG32(MC_SEQ_MISC5) >> 1) & 0x1) ? true : false; + else + dll_state_on = ((RREG32(MC_SEQ_MISC6) >> 1) & 0x1) ? true : false; + } else { + dll_state_on = false; + } + } else { + level->strobeMode = si_get_strobe_mode_settings(adev, + pl->mclk); + + dll_state_on = ((RREG32(MC_SEQ_MISC5) >> 1) & 0x1) ? true : false; + } + + ret = si_populate_mclk_value(adev, + pl->sclk, + pl->mclk, + &level->mclk, + (level->strobeMode & SISLANDS_SMC_STROBE_ENABLE) != 0, dll_state_on); + if (ret) + return ret; + + ret = si_populate_voltage_value(adev, + &eg_pi->vddc_voltage_table, + pl->vddc, &level->vddc); + if (ret) + return ret; + + + ret = si_get_std_voltage_value(adev, &level->vddc, &std_vddc); + if (ret) + return ret; + + ret = si_populate_std_voltage_value(adev, std_vddc, + level->vddc.index, &level->std_vddc); + if (ret) + return ret; + + if (eg_pi->vddci_control) { + ret = si_populate_voltage_value(adev, &eg_pi->vddci_voltage_table, + pl->vddci, &level->vddci); + if (ret) + return ret; + } + + if (si_pi->vddc_phase_shed_control) { + ret = si_populate_phase_shedding_value(adev, + &adev->pm.dpm.dyn_state.phase_shedding_limits_table, + pl->vddc, + pl->sclk, + pl->mclk, + &level->vddc); + if (ret) + return ret; + } + + level->MaxPoweredUpCU = si_pi->max_cu; + + ret = si_populate_mvdd_value(adev, pl->mclk, &level->mvdd); + + return ret; +} + +static int si_populate_smc_t(struct amdgpu_device *adev, + struct amdgpu_ps *amdgpu_state, + SISLANDS_SMC_SWSTATE *smc_state) +{ + struct rv7xx_power_info *pi = rv770_get_pi(adev); + struct si_ps *state = si_get_ps(amdgpu_state); + u32 a_t; + u32 t_l, t_h; + u32 high_bsp; + int i, ret; + + if (state->performance_level_count >= 9) + return -EINVAL; + + if (state->performance_level_count < 2) { + a_t = CG_R(0xffff) | CG_L(0); + smc_state->levels[0].aT = cpu_to_be32(a_t); + return 0; + } + + smc_state->levels[0].aT = cpu_to_be32(0); + + for (i = 0; i <= state->performance_level_count - 2; i++) { + ret = r600_calculate_at( + (50 / SISLANDS_MAX_HARDWARE_POWERLEVELS) * 100 * (i + 1), + 100 * R600_AH_DFLT, + state->performance_levels[i + 1].sclk, + state->performance_levels[i].sclk, + &t_l, + &t_h); + + if (ret) { + t_h = (i + 1) * 1000 - 50 * R600_AH_DFLT; + t_l = (i + 1) * 1000 + 50 * R600_AH_DFLT; + } + + a_t = be32_to_cpu(smc_state->levels[i].aT) & ~CG_R_MASK; + a_t |= CG_R(t_l * pi->bsp / 20000); + smc_state->levels[i].aT = cpu_to_be32(a_t); + + high_bsp = (i == state->performance_level_count - 2) ? + pi->pbsp : pi->bsp; + a_t = CG_R(0xffff) | CG_L(t_h * high_bsp / 20000); + smc_state->levels[i + 1].aT = cpu_to_be32(a_t); + } + + return 0; +} + +static int si_disable_ulv(struct amdgpu_device *adev) +{ + struct si_power_info *si_pi = si_get_pi(adev); + struct si_ulv_param *ulv = &si_pi->ulv; + + if (ulv->supported) + return (amdgpu_si_send_msg_to_smc(adev, PPSMC_MSG_DisableULV) == PPSMC_Result_OK) ? + 0 : -EINVAL; + + return 0; +} + +static bool si_is_state_ulv_compatible(struct amdgpu_device *adev, + struct amdgpu_ps *amdgpu_state) +{ + const struct si_power_info *si_pi = si_get_pi(adev); + const struct si_ulv_param *ulv = &si_pi->ulv; + const struct si_ps *state = si_get_ps(amdgpu_state); + int i; + + if (state->performance_levels[0].mclk != ulv->pl.mclk) + return false; + + /* XXX validate against display requirements! */ + + for (i = 0; i < adev->pm.dpm.dyn_state.vddc_dependency_on_dispclk.count; i++) { + if (adev->clock.current_dispclk <= + adev->pm.dpm.dyn_state.vddc_dependency_on_dispclk.entries[i].clk) { + if (ulv->pl.vddc < + adev->pm.dpm.dyn_state.vddc_dependency_on_dispclk.entries[i].v) + return false; + } + } + + if ((amdgpu_state->vclk != 0) || (amdgpu_state->dclk != 0)) + return false; + + return true; +} + +static int si_set_power_state_conditionally_enable_ulv(struct amdgpu_device *adev, + struct amdgpu_ps *amdgpu_new_state) +{ + const struct si_power_info *si_pi = si_get_pi(adev); + const struct si_ulv_param *ulv = &si_pi->ulv; + + if (ulv->supported) { + if (si_is_state_ulv_compatible(adev, amdgpu_new_state)) + return (amdgpu_si_send_msg_to_smc(adev, PPSMC_MSG_EnableULV) == PPSMC_Result_OK) ? + 0 : -EINVAL; + } + return 0; +} + +static int si_convert_power_state_to_smc(struct amdgpu_device *adev, + struct amdgpu_ps *amdgpu_state, + SISLANDS_SMC_SWSTATE *smc_state) +{ + struct evergreen_power_info *eg_pi = evergreen_get_pi(adev); + struct ni_power_info *ni_pi = ni_get_pi(adev); + struct si_power_info *si_pi = si_get_pi(adev); + struct si_ps *state = si_get_ps(amdgpu_state); + int i, ret; + u32 threshold; + u32 sclk_in_sr = 1350; /* ??? */ + + if (state->performance_level_count > SISLANDS_MAX_HARDWARE_POWERLEVELS) + return -EINVAL; + + threshold = state->performance_levels[state->performance_level_count-1].sclk * 100 / 100; + + if (amdgpu_state->vclk && amdgpu_state->dclk) { + eg_pi->uvd_enabled = true; + if (eg_pi->smu_uvd_hs) + smc_state->flags |= PPSMC_SWSTATE_FLAG_UVD; + } else { + eg_pi->uvd_enabled = false; + } + + if (state->dc_compatible) + smc_state->flags |= PPSMC_SWSTATE_FLAG_DC; + + smc_state->levelCount = 0; + for (i = 0; i < state->performance_level_count; i++) { + if (eg_pi->sclk_deep_sleep) { + if ((i == 0) || si_pi->sclk_deep_sleep_above_low) { + if (sclk_in_sr <= SCLK_MIN_DEEPSLEEP_FREQ) + smc_state->levels[i].stateFlags |= PPSMC_STATEFLAG_DEEPSLEEP_BYPASS; + else + smc_state->levels[i].stateFlags |= PPSMC_STATEFLAG_DEEPSLEEP_THROTTLE; + } + } + + ret = si_convert_power_level_to_smc(adev, &state->performance_levels[i], + &smc_state->levels[i]); + smc_state->levels[i].arbRefreshState = + (u8)(SISLANDS_DRIVER_STATE_ARB_INDEX + i); + + if (ret) + return ret; + + if (ni_pi->enable_power_containment) + smc_state->levels[i].displayWatermark = + (state->performance_levels[i].sclk < threshold) ? + PPSMC_DISPLAY_WATERMARK_LOW : PPSMC_DISPLAY_WATERMARK_HIGH; + else + smc_state->levels[i].displayWatermark = (i < 2) ? + PPSMC_DISPLAY_WATERMARK_LOW : PPSMC_DISPLAY_WATERMARK_HIGH; + + if (eg_pi->dynamic_ac_timing) + smc_state->levels[i].ACIndex = SISLANDS_MCREGISTERTABLE_FIRST_DRIVERSTATE_SLOT + i; + else + smc_state->levels[i].ACIndex = 0; + + smc_state->levelCount++; + } + + si_write_smc_soft_register(adev, + SI_SMC_SOFT_REGISTER_watermark_threshold, + threshold / 512); + + si_populate_smc_sp(adev, amdgpu_state, smc_state); + + ret = si_populate_power_containment_values(adev, amdgpu_state, smc_state); + if (ret) + ni_pi->enable_power_containment = false; + + ret = si_populate_sq_ramping_values(adev, amdgpu_state, smc_state); + if (ret) + ni_pi->enable_sq_ramping = false; + + return si_populate_smc_t(adev, amdgpu_state, smc_state); +} + +static int si_upload_sw_state(struct amdgpu_device *adev, + struct amdgpu_ps *amdgpu_new_state) +{ + struct si_power_info *si_pi = si_get_pi(adev); + struct si_ps *new_state = si_get_ps(amdgpu_new_state); + int ret; + u32 address = si_pi->state_table_start + + offsetof(SISLANDS_SMC_STATETABLE, driverState); + u32 state_size = sizeof(SISLANDS_SMC_SWSTATE) + + ((new_state->performance_level_count - 1) * + sizeof(SISLANDS_SMC_HW_PERFORMANCE_LEVEL)); + SISLANDS_SMC_SWSTATE *smc_state = &si_pi->smc_statetable.driverState; + + memset(smc_state, 0, state_size); + + ret = si_convert_power_state_to_smc(adev, amdgpu_new_state, smc_state); + if (ret) + return ret; + + return amdgpu_si_copy_bytes_to_smc(adev, address, (u8 *)smc_state, + state_size, si_pi->sram_end); +} + +static int si_upload_ulv_state(struct amdgpu_device *adev) +{ + struct si_power_info *si_pi = si_get_pi(adev); + struct si_ulv_param *ulv = &si_pi->ulv; + int ret = 0; + + if (ulv->supported && ulv->pl.vddc) { + u32 address = si_pi->state_table_start + + offsetof(SISLANDS_SMC_STATETABLE, ULVState); + SISLANDS_SMC_SWSTATE *smc_state = &si_pi->smc_statetable.ULVState; + u32 state_size = sizeof(SISLANDS_SMC_SWSTATE); + + memset(smc_state, 0, state_size); + + ret = si_populate_ulv_state(adev, smc_state); + if (!ret) + ret = amdgpu_si_copy_bytes_to_smc(adev, address, (u8 *)smc_state, + state_size, si_pi->sram_end); + } + + return ret; +} + +static int si_upload_smc_data(struct amdgpu_device *adev) +{ + struct amdgpu_crtc *amdgpu_crtc = NULL; + int i; + + if (adev->pm.dpm.new_active_crtc_count == 0) + return 0; + + for (i = 0; i < adev->mode_info.num_crtc; i++) { + if (adev->pm.dpm.new_active_crtcs & (1 << i)) { + amdgpu_crtc = adev->mode_info.crtcs[i]; + break; + } + } + + if (amdgpu_crtc == NULL) + return 0; + + if (amdgpu_crtc->line_time <= 0) + return 0; + + if (si_write_smc_soft_register(adev, + SI_SMC_SOFT_REGISTER_crtc_index, + amdgpu_crtc->crtc_id) != PPSMC_Result_OK) + return 0; + + if (si_write_smc_soft_register(adev, + SI_SMC_SOFT_REGISTER_mclk_change_block_cp_min, + amdgpu_crtc->wm_high / amdgpu_crtc->line_time) != PPSMC_Result_OK) + return 0; + + if (si_write_smc_soft_register(adev, + SI_SMC_SOFT_REGISTER_mclk_change_block_cp_max, + amdgpu_crtc->wm_low / amdgpu_crtc->line_time) != PPSMC_Result_OK) + return 0; + + return 0; +} + +static int si_set_mc_special_registers(struct amdgpu_device *adev, + struct si_mc_reg_table *table) +{ + u8 i, j, k; + u32 temp_reg; + + for (i = 0, j = table->last; i < table->last; i++) { + if (j >= SMC_SISLANDS_MC_REGISTER_ARRAY_SIZE) + return -EINVAL; + switch (table->mc_reg_address[i].s1) { + case MC_SEQ_MISC1: + temp_reg = RREG32(MC_PMG_CMD_EMRS); + table->mc_reg_address[j].s1 = MC_PMG_CMD_EMRS; + table->mc_reg_address[j].s0 = MC_SEQ_PMG_CMD_EMRS_LP; + for (k = 0; k < table->num_entries; k++) + table->mc_reg_table_entry[k].mc_data[j] = + ((temp_reg & 0xffff0000)) | + ((table->mc_reg_table_entry[k].mc_data[i] & 0xffff0000) >> 16); + j++; + if (j >= SMC_SISLANDS_MC_REGISTER_ARRAY_SIZE) + return -EINVAL; + + temp_reg = RREG32(MC_PMG_CMD_MRS); + table->mc_reg_address[j].s1 = MC_PMG_CMD_MRS; + table->mc_reg_address[j].s0 = MC_SEQ_PMG_CMD_MRS_LP; + for (k = 0; k < table->num_entries; k++) { + table->mc_reg_table_entry[k].mc_data[j] = + (temp_reg & 0xffff0000) | + (table->mc_reg_table_entry[k].mc_data[i] & 0x0000ffff); + if (adev->mc.vram_type != AMDGPU_VRAM_TYPE_GDDR5) + table->mc_reg_table_entry[k].mc_data[j] |= 0x100; + } + j++; + if (j >= SMC_SISLANDS_MC_REGISTER_ARRAY_SIZE) + return -EINVAL; + + if (adev->mc.vram_type != AMDGPU_VRAM_TYPE_GDDR5) { + table->mc_reg_address[j].s1 = MC_PMG_AUTO_CMD; + table->mc_reg_address[j].s0 = MC_PMG_AUTO_CMD; + for (k = 0; k < table->num_entries; k++) + table->mc_reg_table_entry[k].mc_data[j] = + (table->mc_reg_table_entry[k].mc_data[i] & 0xffff0000) >> 16; + j++; + if (j >= SMC_SISLANDS_MC_REGISTER_ARRAY_SIZE) + return -EINVAL; + } + break; + case MC_SEQ_RESERVE_M: + temp_reg = RREG32(MC_PMG_CMD_MRS1); + table->mc_reg_address[j].s1 = MC_PMG_CMD_MRS1; + table->mc_reg_address[j].s0 = MC_SEQ_PMG_CMD_MRS1_LP; + for(k = 0; k < table->num_entries; k++) + table->mc_reg_table_entry[k].mc_data[j] = + (temp_reg & 0xffff0000) | + (table->mc_reg_table_entry[k].mc_data[i] & 0x0000ffff); + j++; + if (j >= SMC_SISLANDS_MC_REGISTER_ARRAY_SIZE) + return -EINVAL; + break; + default: + break; + } + } + + table->last = j; + + return 0; +} + +static bool si_check_s0_mc_reg_index(u16 in_reg, u16 *out_reg) +{ + bool result = true; + switch (in_reg) { + case MC_SEQ_RAS_TIMING: + *out_reg = MC_SEQ_RAS_TIMING_LP; + break; + case MC_SEQ_CAS_TIMING: + *out_reg = MC_SEQ_CAS_TIMING_LP; + break; + case MC_SEQ_MISC_TIMING: + *out_reg = MC_SEQ_MISC_TIMING_LP; + break; + case MC_SEQ_MISC_TIMING2: + *out_reg = MC_SEQ_MISC_TIMING2_LP; + break; + case MC_SEQ_RD_CTL_D0: + *out_reg = MC_SEQ_RD_CTL_D0_LP; + break; + case MC_SEQ_RD_CTL_D1: + *out_reg = MC_SEQ_RD_CTL_D1_LP; + break; + case MC_SEQ_WR_CTL_D0: + *out_reg = MC_SEQ_WR_CTL_D0_LP; + break; + case MC_SEQ_WR_CTL_D1: + *out_reg = MC_SEQ_WR_CTL_D1_LP; + break; + case MC_PMG_CMD_EMRS: + *out_reg = MC_SEQ_PMG_CMD_EMRS_LP; + break; + case MC_PMG_CMD_MRS: + *out_reg = MC_SEQ_PMG_CMD_MRS_LP; + break; + case MC_PMG_CMD_MRS1: + *out_reg = MC_SEQ_PMG_CMD_MRS1_LP; + break; + case MC_SEQ_PMG_TIMING: + *out_reg = MC_SEQ_PMG_TIMING_LP; + break; + case MC_PMG_CMD_MRS2: + *out_reg = MC_SEQ_PMG_CMD_MRS2_LP; + break; + case MC_SEQ_WR_CTL_2: + *out_reg = MC_SEQ_WR_CTL_2_LP; + break; + default: + result = false; + break; + } + + return result; +} + +static void si_set_valid_flag(struct si_mc_reg_table *table) +{ + u8 i, j; + + for (i = 0; i < table->last; i++) { + for (j = 1; j < table->num_entries; j++) { + if (table->mc_reg_table_entry[j-1].mc_data[i] != table->mc_reg_table_entry[j].mc_data[i]) { + table->valid_flag |= 1 << i; + break; + } + } + } +} + +static void si_set_s0_mc_reg_index(struct si_mc_reg_table *table) +{ + u32 i; + u16 address; + + for (i = 0; i < table->last; i++) + table->mc_reg_address[i].s0 = si_check_s0_mc_reg_index(table->mc_reg_address[i].s1, &address) ? + address : table->mc_reg_address[i].s1; + +} + +static int si_copy_vbios_mc_reg_table(struct atom_mc_reg_table *table, + struct si_mc_reg_table *si_table) +{ + u8 i, j; + + if (table->last > SMC_SISLANDS_MC_REGISTER_ARRAY_SIZE) + return -EINVAL; + if (table->num_entries > MAX_AC_TIMING_ENTRIES) + return -EINVAL; + + for (i = 0; i < table->last; i++) + si_table->mc_reg_address[i].s1 = table->mc_reg_address[i].s1; + si_table->last = table->last; + + for (i = 0; i < table->num_entries; i++) { + si_table->mc_reg_table_entry[i].mclk_max = + table->mc_reg_table_entry[i].mclk_max; + for (j = 0; j < table->last; j++) { + si_table->mc_reg_table_entry[i].mc_data[j] = + table->mc_reg_table_entry[i].mc_data[j]; + } + } + si_table->num_entries = table->num_entries; + + return 0; +} + +static int si_initialize_mc_reg_table(struct amdgpu_device *adev) +{ + struct si_power_info *si_pi = si_get_pi(adev); + struct atom_mc_reg_table *table; + struct si_mc_reg_table *si_table = &si_pi->mc_reg_table; + u8 module_index = rv770_get_memory_module_index(adev); + int ret; + + table = kzalloc(sizeof(struct atom_mc_reg_table), GFP_KERNEL); + if (!table) + return -ENOMEM; + + WREG32(MC_SEQ_RAS_TIMING_LP, RREG32(MC_SEQ_RAS_TIMING)); + WREG32(MC_SEQ_CAS_TIMING_LP, RREG32(MC_SEQ_CAS_TIMING)); + WREG32(MC_SEQ_MISC_TIMING_LP, RREG32(MC_SEQ_MISC_TIMING)); + WREG32(MC_SEQ_MISC_TIMING2_LP, RREG32(MC_SEQ_MISC_TIMING2)); + WREG32(MC_SEQ_PMG_CMD_EMRS_LP, RREG32(MC_PMG_CMD_EMRS)); + WREG32(MC_SEQ_PMG_CMD_MRS_LP, RREG32(MC_PMG_CMD_MRS)); + WREG32(MC_SEQ_PMG_CMD_MRS1_LP, RREG32(MC_PMG_CMD_MRS1)); + WREG32(MC_SEQ_WR_CTL_D0_LP, RREG32(MC_SEQ_WR_CTL_D0)); + WREG32(MC_SEQ_WR_CTL_D1_LP, RREG32(MC_SEQ_WR_CTL_D1)); + WREG32(MC_SEQ_RD_CTL_D0_LP, RREG32(MC_SEQ_RD_CTL_D0)); + WREG32(MC_SEQ_RD_CTL_D1_LP, RREG32(MC_SEQ_RD_CTL_D1)); + WREG32(MC_SEQ_PMG_TIMING_LP, RREG32(MC_SEQ_PMG_TIMING)); + WREG32(MC_SEQ_PMG_CMD_MRS2_LP, RREG32(MC_PMG_CMD_MRS2)); + WREG32(MC_SEQ_WR_CTL_2_LP, RREG32(MC_SEQ_WR_CTL_2)); + + ret = amdgpu_atombios_init_mc_reg_table(adev, module_index, table); + if (ret) + goto init_mc_done; + + ret = si_copy_vbios_mc_reg_table(table, si_table); + if (ret) + goto init_mc_done; + + si_set_s0_mc_reg_index(si_table); + + ret = si_set_mc_special_registers(adev, si_table); + if (ret) + goto init_mc_done; + + si_set_valid_flag(si_table); + +init_mc_done: + kfree(table); + + return ret; + +} + +static void si_populate_mc_reg_addresses(struct amdgpu_device *adev, + SMC_SIslands_MCRegisters *mc_reg_table) +{ + struct si_power_info *si_pi = si_get_pi(adev); + u32 i, j; + + for (i = 0, j = 0; j < si_pi->mc_reg_table.last; j++) { + if (si_pi->mc_reg_table.valid_flag & (1 << j)) { + if (i >= SMC_SISLANDS_MC_REGISTER_ARRAY_SIZE) + break; + mc_reg_table->address[i].s0 = + cpu_to_be16(si_pi->mc_reg_table.mc_reg_address[j].s0); + mc_reg_table->address[i].s1 = + cpu_to_be16(si_pi->mc_reg_table.mc_reg_address[j].s1); + i++; + } + } + mc_reg_table->last = (u8)i; +} + +static void si_convert_mc_registers(const struct si_mc_reg_entry *entry, + SMC_SIslands_MCRegisterSet *data, + u32 num_entries, u32 valid_flag) +{ + u32 i, j; + + for(i = 0, j = 0; j < num_entries; j++) { + if (valid_flag & (1 << j)) { + data->value[i] = cpu_to_be32(entry->mc_data[j]); + i++; + } + } +} + +static void si_convert_mc_reg_table_entry_to_smc(struct amdgpu_device *adev, + struct rv7xx_pl *pl, + SMC_SIslands_MCRegisterSet *mc_reg_table_data) +{ + struct si_power_info *si_pi = si_get_pi(adev); + u32 i = 0; + + for (i = 0; i < si_pi->mc_reg_table.num_entries; i++) { + if (pl->mclk <= si_pi->mc_reg_table.mc_reg_table_entry[i].mclk_max) + break; + } + + if ((i == si_pi->mc_reg_table.num_entries) && (i > 0)) + --i; + + si_convert_mc_registers(&si_pi->mc_reg_table.mc_reg_table_entry[i], + mc_reg_table_data, si_pi->mc_reg_table.last, + si_pi->mc_reg_table.valid_flag); +} + +static void si_convert_mc_reg_table_to_smc(struct amdgpu_device *adev, + struct amdgpu_ps *amdgpu_state, + SMC_SIslands_MCRegisters *mc_reg_table) +{ + struct si_ps *state = si_get_ps(amdgpu_state); + int i; + + for (i = 0; i < state->performance_level_count; i++) { + si_convert_mc_reg_table_entry_to_smc(adev, + &state->performance_levels[i], + &mc_reg_table->data[SISLANDS_MCREGISTERTABLE_FIRST_DRIVERSTATE_SLOT + i]); + } +} + +static int si_populate_mc_reg_table(struct amdgpu_device *adev, + struct amdgpu_ps *amdgpu_boot_state) +{ + struct si_ps *boot_state = si_get_ps(amdgpu_boot_state); + struct si_power_info *si_pi = si_get_pi(adev); + struct si_ulv_param *ulv = &si_pi->ulv; + SMC_SIslands_MCRegisters *smc_mc_reg_table = &si_pi->smc_mc_reg_table; + + memset(smc_mc_reg_table, 0, sizeof(SMC_SIslands_MCRegisters)); + + si_write_smc_soft_register(adev, SI_SMC_SOFT_REGISTER_seq_index, 1); + + si_populate_mc_reg_addresses(adev, smc_mc_reg_table); + + si_convert_mc_reg_table_entry_to_smc(adev, &boot_state->performance_levels[0], + &smc_mc_reg_table->data[SISLANDS_MCREGISTERTABLE_INITIAL_SLOT]); + + si_convert_mc_registers(&si_pi->mc_reg_table.mc_reg_table_entry[0], + &smc_mc_reg_table->data[SISLANDS_MCREGISTERTABLE_ACPI_SLOT], + si_pi->mc_reg_table.last, + si_pi->mc_reg_table.valid_flag); + + if (ulv->supported && ulv->pl.vddc != 0) + si_convert_mc_reg_table_entry_to_smc(adev, &ulv->pl, + &smc_mc_reg_table->data[SISLANDS_MCREGISTERTABLE_ULV_SLOT]); + else + si_convert_mc_registers(&si_pi->mc_reg_table.mc_reg_table_entry[0], + &smc_mc_reg_table->data[SISLANDS_MCREGISTERTABLE_ULV_SLOT], + si_pi->mc_reg_table.last, + si_pi->mc_reg_table.valid_flag); + + si_convert_mc_reg_table_to_smc(adev, amdgpu_boot_state, smc_mc_reg_table); + + return amdgpu_si_copy_bytes_to_smc(adev, si_pi->mc_reg_table_start, + (u8 *)smc_mc_reg_table, + sizeof(SMC_SIslands_MCRegisters), si_pi->sram_end); +} + +static int si_upload_mc_reg_table(struct amdgpu_device *adev, + struct amdgpu_ps *amdgpu_new_state) +{ + struct si_ps *new_state = si_get_ps(amdgpu_new_state); + struct si_power_info *si_pi = si_get_pi(adev); + u32 address = si_pi->mc_reg_table_start + + offsetof(SMC_SIslands_MCRegisters, + data[SISLANDS_MCREGISTERTABLE_FIRST_DRIVERSTATE_SLOT]); + SMC_SIslands_MCRegisters *smc_mc_reg_table = &si_pi->smc_mc_reg_table; + + memset(smc_mc_reg_table, 0, sizeof(SMC_SIslands_MCRegisters)); + + si_convert_mc_reg_table_to_smc(adev, amdgpu_new_state, smc_mc_reg_table); + + return amdgpu_si_copy_bytes_to_smc(adev, address, + (u8 *)&smc_mc_reg_table->data[SISLANDS_MCREGISTERTABLE_FIRST_DRIVERSTATE_SLOT], + sizeof(SMC_SIslands_MCRegisterSet) * new_state->performance_level_count, + si_pi->sram_end); +} + +static void si_enable_voltage_control(struct amdgpu_device *adev, bool enable) +{ + if (enable) + WREG32_P(GENERAL_PWRMGT, VOLT_PWRMGT_EN, ~VOLT_PWRMGT_EN); + else + WREG32_P(GENERAL_PWRMGT, 0, ~VOLT_PWRMGT_EN); +} + +static enum amdgpu_pcie_gen si_get_maximum_link_speed(struct amdgpu_device *adev, + struct amdgpu_ps *amdgpu_state) +{ + struct si_ps *state = si_get_ps(amdgpu_state); + int i; + u16 pcie_speed, max_speed = 0; + + for (i = 0; i < state->performance_level_count; i++) { + pcie_speed = state->performance_levels[i].pcie_gen; + if (max_speed < pcie_speed) + max_speed = pcie_speed; + } + return max_speed; +} + +static u16 si_get_current_pcie_speed(struct amdgpu_device *adev) +{ + u32 speed_cntl; + + speed_cntl = RREG32_PCIE_PORT(PCIE_LC_SPEED_CNTL) & LC_CURRENT_DATA_RATE_MASK; + speed_cntl >>= LC_CURRENT_DATA_RATE_SHIFT; + + return (u16)speed_cntl; +} + +static void si_request_link_speed_change_before_state_change(struct amdgpu_device *adev, + struct amdgpu_ps *amdgpu_new_state, + struct amdgpu_ps *amdgpu_current_state) +{ + struct si_power_info *si_pi = si_get_pi(adev); + enum amdgpu_pcie_gen target_link_speed = si_get_maximum_link_speed(adev, amdgpu_new_state); + enum amdgpu_pcie_gen current_link_speed; + + if (si_pi->force_pcie_gen == AMDGPU_PCIE_GEN_INVALID) + current_link_speed = si_get_maximum_link_speed(adev, amdgpu_current_state); + else + current_link_speed = si_pi->force_pcie_gen; + + si_pi->force_pcie_gen = AMDGPU_PCIE_GEN_INVALID; + si_pi->pspp_notify_required = false; + if (target_link_speed > current_link_speed) { + switch (target_link_speed) { +#if defined(CONFIG_ACPI) + case AMDGPU_PCIE_GEN3: + if (amdgpu_acpi_pcie_performance_request(adev, PCIE_PERF_REQ_PECI_GEN3, false) == 0) + break; + si_pi->force_pcie_gen = AMDGPU_PCIE_GEN2; + if (current_link_speed == AMDGPU_PCIE_GEN2) + break; + case AMDGPU_PCIE_GEN2: + if (amdgpu_acpi_pcie_performance_request(adev, PCIE_PERF_REQ_PECI_GEN2, false) == 0) + break; +#endif + default: + si_pi->force_pcie_gen = si_get_current_pcie_speed(adev); + break; + } + } else { + if (target_link_speed < current_link_speed) + si_pi->pspp_notify_required = true; + } +} + +static void si_notify_link_speed_change_after_state_change(struct amdgpu_device *adev, + struct amdgpu_ps *amdgpu_new_state, + struct amdgpu_ps *amdgpu_current_state) +{ + struct si_power_info *si_pi = si_get_pi(adev); + enum amdgpu_pcie_gen target_link_speed = si_get_maximum_link_speed(adev, amdgpu_new_state); + u8 request; + + if (si_pi->pspp_notify_required) { + if (target_link_speed == AMDGPU_PCIE_GEN3) + request = PCIE_PERF_REQ_PECI_GEN3; + else if (target_link_speed == AMDGPU_PCIE_GEN2) + request = PCIE_PERF_REQ_PECI_GEN2; + else + request = PCIE_PERF_REQ_PECI_GEN1; + + if ((request == PCIE_PERF_REQ_PECI_GEN1) && + (si_get_current_pcie_speed(adev) > 0)) + return; + +#if defined(CONFIG_ACPI) + amdgpu_acpi_pcie_performance_request(adev, request, false); +#endif + } +} + +#if 0 +static int si_ds_request(struct amdgpu_device *adev, + bool ds_status_on, u32 count_write) +{ + struct evergreen_power_info *eg_pi = evergreen_get_pi(adev); + + if (eg_pi->sclk_deep_sleep) { + if (ds_status_on) + return (amdgpu_si_send_msg_to_smc(adev, PPSMC_MSG_CancelThrottleOVRDSCLKDS) == + PPSMC_Result_OK) ? + 0 : -EINVAL; + else + return (amdgpu_si_send_msg_to_smc(adev, PPSMC_MSG_ThrottleOVRDSCLKDS) == + PPSMC_Result_OK) ? 0 : -EINVAL; + } + return 0; +} +#endif + +static void si_set_max_cu_value(struct amdgpu_device *adev) +{ + struct si_power_info *si_pi = si_get_pi(adev); + + if (adev->asic_type == CHIP_VERDE) { + switch (adev->pdev->device) { + case 0x6820: + case 0x6825: + case 0x6821: + case 0x6823: + case 0x6827: + si_pi->max_cu = 10; + break; + case 0x682D: + case 0x6824: + case 0x682F: + case 0x6826: + si_pi->max_cu = 8; + break; + case 0x6828: + case 0x6830: + case 0x6831: + case 0x6838: + case 0x6839: + case 0x683D: + si_pi->max_cu = 10; + break; + case 0x683B: + case 0x683F: + case 0x6829: + si_pi->max_cu = 8; + break; + default: + si_pi->max_cu = 0; + break; + } + } else { + si_pi->max_cu = 0; + } +} + +static int si_patch_single_dependency_table_based_on_leakage(struct amdgpu_device *adev, + struct amdgpu_clock_voltage_dependency_table *table) +{ + u32 i; + int j; + u16 leakage_voltage; + + if (table) { + for (i = 0; i < table->count; i++) { + switch (si_get_leakage_voltage_from_leakage_index(adev, + table->entries[i].v, + &leakage_voltage)) { + case 0: + table->entries[i].v = leakage_voltage; + break; + case -EAGAIN: + return -EINVAL; + case -EINVAL: + default: + break; + } + } + + for (j = (table->count - 2); j >= 0; j--) { + table->entries[j].v = (table->entries[j].v <= table->entries[j + 1].v) ? + table->entries[j].v : table->entries[j + 1].v; + } + } + return 0; +} + +static int si_patch_dependency_tables_based_on_leakage(struct amdgpu_device *adev) +{ + int ret = 0; + + ret = si_patch_single_dependency_table_based_on_leakage(adev, + &adev->pm.dpm.dyn_state.vddc_dependency_on_sclk); + if (ret) + DRM_ERROR("Could not patch vddc_on_sclk leakage table\n"); + ret = si_patch_single_dependency_table_based_on_leakage(adev, + &adev->pm.dpm.dyn_state.vddc_dependency_on_mclk); + if (ret) + DRM_ERROR("Could not patch vddc_on_mclk leakage table\n"); + ret = si_patch_single_dependency_table_based_on_leakage(adev, + &adev->pm.dpm.dyn_state.vddci_dependency_on_mclk); + if (ret) + DRM_ERROR("Could not patch vddci_on_mclk leakage table\n"); + return ret; +} + +static void si_set_pcie_lane_width_in_smc(struct amdgpu_device *adev, + struct amdgpu_ps *amdgpu_new_state, + struct amdgpu_ps *amdgpu_current_state) +{ + u32 lane_width; + u32 new_lane_width = + (amdgpu_new_state->caps & ATOM_PPLIB_PCIE_LINK_WIDTH_MASK) >> ATOM_PPLIB_PCIE_LINK_WIDTH_SHIFT; + u32 current_lane_width = + (amdgpu_current_state->caps & ATOM_PPLIB_PCIE_LINK_WIDTH_MASK) >> ATOM_PPLIB_PCIE_LINK_WIDTH_SHIFT; + + if (new_lane_width != current_lane_width) { + amdgpu_set_pcie_lanes(adev, new_lane_width); + lane_width = amdgpu_get_pcie_lanes(adev); + si_write_smc_soft_register(adev, SI_SMC_SOFT_REGISTER_non_ulv_pcie_link_width, lane_width); + } +} + +static void si_dpm_setup_asic(struct amdgpu_device *adev) +{ + si_read_clock_registers(adev); + si_enable_acpi_power_management(adev); +} + +static int si_thermal_enable_alert(struct amdgpu_device *adev, + bool enable) +{ + u32 thermal_int = RREG32(CG_THERMAL_INT); + + if (enable) { + PPSMC_Result result; + + thermal_int &= ~(THERM_INT_MASK_HIGH | THERM_INT_MASK_LOW); + WREG32(CG_THERMAL_INT, thermal_int); + result = amdgpu_si_send_msg_to_smc(adev, PPSMC_MSG_EnableThermalInterrupt); + if (result != PPSMC_Result_OK) { + DRM_DEBUG_KMS("Could not enable thermal interrupts.\n"); + return -EINVAL; + } + } else { + thermal_int |= THERM_INT_MASK_HIGH | THERM_INT_MASK_LOW; + WREG32(CG_THERMAL_INT, thermal_int); + } + + return 0; +} + +static int si_thermal_set_temperature_range(struct amdgpu_device *adev, + int min_temp, int max_temp) +{ + int low_temp = 0 * 1000; + int high_temp = 255 * 1000; + + if (low_temp < min_temp) + low_temp = min_temp; + if (high_temp > max_temp) + high_temp = max_temp; + if (high_temp < low_temp) { + DRM_ERROR("invalid thermal range: %d - %d\n", low_temp, high_temp); + return -EINVAL; + } + + WREG32_P(CG_THERMAL_INT, DIG_THERM_INTH(high_temp / 1000), ~DIG_THERM_INTH_MASK); + WREG32_P(CG_THERMAL_INT, DIG_THERM_INTL(low_temp / 1000), ~DIG_THERM_INTL_MASK); + WREG32_P(CG_THERMAL_CTRL, DIG_THERM_DPM(high_temp / 1000), ~DIG_THERM_DPM_MASK); + + adev->pm.dpm.thermal.min_temp = low_temp; + adev->pm.dpm.thermal.max_temp = high_temp; + + return 0; +} + +static void si_fan_ctrl_set_static_mode(struct amdgpu_device *adev, u32 mode) +{ + struct si_power_info *si_pi = si_get_pi(adev); + u32 tmp; + + if (si_pi->fan_ctrl_is_in_default_mode) { + tmp = (RREG32(CG_FDO_CTRL2) & FDO_PWM_MODE_MASK) >> FDO_PWM_MODE_SHIFT; + si_pi->fan_ctrl_default_mode = tmp; + tmp = (RREG32(CG_FDO_CTRL2) & TMIN_MASK) >> TMIN_SHIFT; + si_pi->t_min = tmp; + si_pi->fan_ctrl_is_in_default_mode = false; + } + + tmp = RREG32(CG_FDO_CTRL2) & ~TMIN_MASK; + tmp |= TMIN(0); + WREG32(CG_FDO_CTRL2, tmp); + + tmp = RREG32(CG_FDO_CTRL2) & ~FDO_PWM_MODE_MASK; + tmp |= FDO_PWM_MODE(mode); + WREG32(CG_FDO_CTRL2, tmp); +} + +static int si_thermal_setup_fan_table(struct amdgpu_device *adev) +{ + struct si_power_info *si_pi = si_get_pi(adev); + PP_SIslands_FanTable fan_table = { FDO_MODE_HARDWARE }; + u32 duty100; + u32 t_diff1, t_diff2, pwm_diff1, pwm_diff2; + u16 fdo_min, slope1, slope2; + u32 reference_clock, tmp; + int ret; + u64 tmp64; + + if (!si_pi->fan_table_start) { + adev->pm.dpm.fan.ucode_fan_control = false; + return 0; + } + + duty100 = (RREG32(CG_FDO_CTRL1) & FMAX_DUTY100_MASK) >> FMAX_DUTY100_SHIFT; + + if (duty100 == 0) { + adev->pm.dpm.fan.ucode_fan_control = false; + return 0; + } + + tmp64 = (u64)adev->pm.dpm.fan.pwm_min * duty100; + do_div(tmp64, 10000); + fdo_min = (u16)tmp64; + + t_diff1 = adev->pm.dpm.fan.t_med - adev->pm.dpm.fan.t_min; + t_diff2 = adev->pm.dpm.fan.t_high - adev->pm.dpm.fan.t_med; + + pwm_diff1 = adev->pm.dpm.fan.pwm_med - adev->pm.dpm.fan.pwm_min; + pwm_diff2 = adev->pm.dpm.fan.pwm_high - adev->pm.dpm.fan.pwm_med; + + slope1 = (u16)((50 + ((16 * duty100 * pwm_diff1) / t_diff1)) / 100); + slope2 = (u16)((50 + ((16 * duty100 * pwm_diff2) / t_diff2)) / 100); + + fan_table.temp_min = cpu_to_be16((50 + adev->pm.dpm.fan.t_min) / 100); + fan_table.temp_med = cpu_to_be16((50 + adev->pm.dpm.fan.t_med) / 100); + fan_table.temp_max = cpu_to_be16((50 + adev->pm.dpm.fan.t_max) / 100); + fan_table.slope1 = cpu_to_be16(slope1); + fan_table.slope2 = cpu_to_be16(slope2); + fan_table.fdo_min = cpu_to_be16(fdo_min); + fan_table.hys_down = cpu_to_be16(adev->pm.dpm.fan.t_hyst); + fan_table.hys_up = cpu_to_be16(1); + fan_table.hys_slope = cpu_to_be16(1); + fan_table.temp_resp_lim = cpu_to_be16(5); + reference_clock = amdgpu_asic_get_xclk(adev); + + fan_table.refresh_period = cpu_to_be32((adev->pm.dpm.fan.cycle_delay * + reference_clock) / 1600); + fan_table.fdo_max = cpu_to_be16((u16)duty100); + + tmp = (RREG32(CG_MULT_THERMAL_CTRL) & TEMP_SEL_MASK) >> TEMP_SEL_SHIFT; + fan_table.temp_src = (uint8_t)tmp; + + ret = amdgpu_si_copy_bytes_to_smc(adev, + si_pi->fan_table_start, + (u8 *)(&fan_table), + sizeof(fan_table), + si_pi->sram_end); + + if (ret) { + DRM_ERROR("Failed to load fan table to the SMC."); + adev->pm.dpm.fan.ucode_fan_control = false; + } + + return ret; +} + +static int si_fan_ctrl_start_smc_fan_control(struct amdgpu_device *adev) +{ + struct si_power_info *si_pi = si_get_pi(adev); + PPSMC_Result ret; + + ret = amdgpu_si_send_msg_to_smc(adev, PPSMC_StartFanControl); + if (ret == PPSMC_Result_OK) { + si_pi->fan_is_controlled_by_smc = true; + return 0; + } else { + return -EINVAL; + } +} + +static int si_fan_ctrl_stop_smc_fan_control(struct amdgpu_device *adev) +{ + struct si_power_info *si_pi = si_get_pi(adev); + PPSMC_Result ret; + + ret = amdgpu_si_send_msg_to_smc(adev, PPSMC_StopFanControl); + + if (ret == PPSMC_Result_OK) { + si_pi->fan_is_controlled_by_smc = false; + return 0; + } else { + return -EINVAL; + } +} + +static int si_dpm_get_fan_speed_percent(struct amdgpu_device *adev, + u32 *speed) +{ + u32 duty, duty100; + u64 tmp64; + + if (adev->pm.no_fan) + return -ENOENT; + + duty100 = (RREG32(CG_FDO_CTRL1) & FMAX_DUTY100_MASK) >> FMAX_DUTY100_SHIFT; + duty = (RREG32(CG_THERMAL_STATUS) & FDO_PWM_DUTY_MASK) >> FDO_PWM_DUTY_SHIFT; + + if (duty100 == 0) + return -EINVAL; + + tmp64 = (u64)duty * 100; + do_div(tmp64, duty100); + *speed = (u32)tmp64; + + if (*speed > 100) + *speed = 100; + + return 0; +} + +static int si_dpm_set_fan_speed_percent(struct amdgpu_device *adev, + u32 speed) +{ + struct si_power_info *si_pi = si_get_pi(adev); + u32 tmp; + u32 duty, duty100; + u64 tmp64; + + if (adev->pm.no_fan) + return -ENOENT; + + if (si_pi->fan_is_controlled_by_smc) + return -EINVAL; + + if (speed > 100) + return -EINVAL; + + duty100 = (RREG32(CG_FDO_CTRL1) & FMAX_DUTY100_MASK) >> FMAX_DUTY100_SHIFT; + + if (duty100 == 0) + return -EINVAL; + + tmp64 = (u64)speed * duty100; + do_div(tmp64, 100); + duty = (u32)tmp64; + + tmp = RREG32(CG_FDO_CTRL0) & ~FDO_STATIC_DUTY_MASK; + tmp |= FDO_STATIC_DUTY(duty); + WREG32(CG_FDO_CTRL0, tmp); + + return 0; +} + +static void si_dpm_set_fan_control_mode(struct amdgpu_device *adev, u32 mode) +{ + if (mode) { + /* stop auto-manage */ + if (adev->pm.dpm.fan.ucode_fan_control) + si_fan_ctrl_stop_smc_fan_control(adev); + si_fan_ctrl_set_static_mode(adev, mode); + } else { + /* restart auto-manage */ + if (adev->pm.dpm.fan.ucode_fan_control) + si_thermal_start_smc_fan_control(adev); + else + si_fan_ctrl_set_default_mode(adev); + } +} + +static u32 si_dpm_get_fan_control_mode(struct amdgpu_device *adev) +{ + struct si_power_info *si_pi = si_get_pi(adev); + u32 tmp; + + if (si_pi->fan_is_controlled_by_smc) + return 0; + + tmp = RREG32(CG_FDO_CTRL2) & FDO_PWM_MODE_MASK; + return (tmp >> FDO_PWM_MODE_SHIFT); +} + +#if 0 +static int si_fan_ctrl_get_fan_speed_rpm(struct amdgpu_device *adev, + u32 *speed) +{ + u32 tach_period; + u32 xclk = amdgpu_asic_get_xclk(adev); + + if (adev->pm.no_fan) + return -ENOENT; + + if (adev->pm.fan_pulses_per_revolution == 0) + return -ENOENT; + + tach_period = (RREG32(CG_TACH_STATUS) & TACH_PERIOD_MASK) >> TACH_PERIOD_SHIFT; + if (tach_period == 0) + return -ENOENT; + + *speed = 60 * xclk * 10000 / tach_period; + + return 0; +} + +static int si_fan_ctrl_set_fan_speed_rpm(struct amdgpu_device *adev, + u32 speed) +{ + u32 tach_period, tmp; + u32 xclk = amdgpu_asic_get_xclk(adev); + + if (adev->pm.no_fan) + return -ENOENT; + + if (adev->pm.fan_pulses_per_revolution == 0) + return -ENOENT; + + if ((speed < adev->pm.fan_min_rpm) || + (speed > adev->pm.fan_max_rpm)) + return -EINVAL; + + if (adev->pm.dpm.fan.ucode_fan_control) + si_fan_ctrl_stop_smc_fan_control(adev); + + tach_period = 60 * xclk * 10000 / (8 * speed); + tmp = RREG32(CG_TACH_CTRL) & ~TARGET_PERIOD_MASK; + tmp |= TARGET_PERIOD(tach_period); + WREG32(CG_TACH_CTRL, tmp); + + si_fan_ctrl_set_static_mode(adev, FDO_PWM_MODE_STATIC_RPM); + + return 0; +} +#endif + +static void si_fan_ctrl_set_default_mode(struct amdgpu_device *adev) +{ + struct si_power_info *si_pi = si_get_pi(adev); + u32 tmp; + + if (!si_pi->fan_ctrl_is_in_default_mode) { + tmp = RREG32(CG_FDO_CTRL2) & ~FDO_PWM_MODE_MASK; + tmp |= FDO_PWM_MODE(si_pi->fan_ctrl_default_mode); + WREG32(CG_FDO_CTRL2, tmp); + + tmp = RREG32(CG_FDO_CTRL2) & ~TMIN_MASK; + tmp |= TMIN(si_pi->t_min); + WREG32(CG_FDO_CTRL2, tmp); + si_pi->fan_ctrl_is_in_default_mode = true; + } +} + +static void si_thermal_start_smc_fan_control(struct amdgpu_device *adev) +{ + if (adev->pm.dpm.fan.ucode_fan_control) { + si_fan_ctrl_start_smc_fan_control(adev); + si_fan_ctrl_set_static_mode(adev, FDO_PWM_MODE_STATIC); + } +} + +static void si_thermal_initialize(struct amdgpu_device *adev) +{ + u32 tmp; + + if (adev->pm.fan_pulses_per_revolution) { + tmp = RREG32(CG_TACH_CTRL) & ~EDGE_PER_REV_MASK; + tmp |= EDGE_PER_REV(adev->pm.fan_pulses_per_revolution -1); + WREG32(CG_TACH_CTRL, tmp); + } + + tmp = RREG32(CG_FDO_CTRL2) & ~TACH_PWM_RESP_RATE_MASK; + tmp |= TACH_PWM_RESP_RATE(0x28); + WREG32(CG_FDO_CTRL2, tmp); +} + +static int si_thermal_start_thermal_controller(struct amdgpu_device *adev) +{ + int ret; + + si_thermal_initialize(adev); + ret = si_thermal_set_temperature_range(adev, R600_TEMP_RANGE_MIN, R600_TEMP_RANGE_MAX); + if (ret) + return ret; + ret = si_thermal_enable_alert(adev, true); + if (ret) + return ret; + if (adev->pm.dpm.fan.ucode_fan_control) { + ret = si_halt_smc(adev); + if (ret) + return ret; + ret = si_thermal_setup_fan_table(adev); + if (ret) + return ret; + ret = si_resume_smc(adev); + if (ret) + return ret; + si_thermal_start_smc_fan_control(adev); + } + + return 0; +} + +static void si_thermal_stop_thermal_controller(struct amdgpu_device *adev) +{ + if (!adev->pm.no_fan) { + si_fan_ctrl_set_default_mode(adev); + si_fan_ctrl_stop_smc_fan_control(adev); + } +} + +static int si_dpm_enable(struct amdgpu_device *adev) +{ + struct rv7xx_power_info *pi = rv770_get_pi(adev); + struct evergreen_power_info *eg_pi = evergreen_get_pi(adev); + struct si_power_info *si_pi = si_get_pi(adev); + struct amdgpu_ps *boot_ps = adev->pm.dpm.boot_ps; + int ret; + + if (amdgpu_si_is_smc_running(adev)) + return -EINVAL; + if (pi->voltage_control || si_pi->voltage_control_svi2) + si_enable_voltage_control(adev, true); + if (pi->mvdd_control) + si_get_mvdd_configuration(adev); + if (pi->voltage_control || si_pi->voltage_control_svi2) { + ret = si_construct_voltage_tables(adev); + if (ret) { + DRM_ERROR("si_construct_voltage_tables failed\n"); + return ret; + } + } + if (eg_pi->dynamic_ac_timing) { + ret = si_initialize_mc_reg_table(adev); + if (ret) + eg_pi->dynamic_ac_timing = false; + } + if (pi->dynamic_ss) + si_enable_spread_spectrum(adev, true); + if (pi->thermal_protection) + si_enable_thermal_protection(adev, true); + si_setup_bsp(adev); + si_program_git(adev); + si_program_tp(adev); + si_program_tpp(adev); + si_program_sstp(adev); + si_enable_display_gap(adev); + si_program_vc(adev); + ret = si_upload_firmware(adev); + if (ret) { + DRM_ERROR("si_upload_firmware failed\n"); + return ret; + } + ret = si_process_firmware_header(adev); + if (ret) { + DRM_ERROR("si_process_firmware_header failed\n"); + return ret; + } + ret = si_initial_switch_from_arb_f0_to_f1(adev); + if (ret) { + DRM_ERROR("si_initial_switch_from_arb_f0_to_f1 failed\n"); + return ret; + } + ret = si_init_smc_table(adev); + if (ret) { + DRM_ERROR("si_init_smc_table failed\n"); + return ret; + } + ret = si_init_smc_spll_table(adev); + if (ret) { + DRM_ERROR("si_init_smc_spll_table failed\n"); + return ret; + } + ret = si_init_arb_table_index(adev); + if (ret) { + DRM_ERROR("si_init_arb_table_index failed\n"); + return ret; + } + if (eg_pi->dynamic_ac_timing) { + ret = si_populate_mc_reg_table(adev, boot_ps); + if (ret) { + DRM_ERROR("si_populate_mc_reg_table failed\n"); + return ret; + } + } + ret = si_initialize_smc_cac_tables(adev); + if (ret) { + DRM_ERROR("si_initialize_smc_cac_tables failed\n"); + return ret; + } + ret = si_initialize_hardware_cac_manager(adev); + if (ret) { + DRM_ERROR("si_initialize_hardware_cac_manager failed\n"); + return ret; + } + ret = si_initialize_smc_dte_tables(adev); + if (ret) { + DRM_ERROR("si_initialize_smc_dte_tables failed\n"); + return ret; + } + ret = si_populate_smc_tdp_limits(adev, boot_ps); + if (ret) { + DRM_ERROR("si_populate_smc_tdp_limits failed\n"); + return ret; + } + ret = si_populate_smc_tdp_limits_2(adev, boot_ps); + if (ret) { + DRM_ERROR("si_populate_smc_tdp_limits_2 failed\n"); + return ret; + } + si_program_response_times(adev); + si_program_ds_registers(adev); + si_dpm_start_smc(adev); + ret = si_notify_smc_display_change(adev, false); + if (ret) { + DRM_ERROR("si_notify_smc_display_change failed\n"); + return ret; + } + si_enable_sclk_control(adev, true); + si_start_dpm(adev); + + si_enable_auto_throttle_source(adev, AMDGPU_DPM_AUTO_THROTTLE_SRC_THERMAL, true); + si_thermal_start_thermal_controller(adev); + ni_update_current_ps(adev, boot_ps); + + return 0; +} + +static int si_set_temperature_range(struct amdgpu_device *adev) +{ + int ret; + + ret = si_thermal_enable_alert(adev, false); + if (ret) + return ret; + ret = si_thermal_set_temperature_range(adev, R600_TEMP_RANGE_MIN, R600_TEMP_RANGE_MAX); + if (ret) + return ret; + ret = si_thermal_enable_alert(adev, true); + if (ret) + return ret; + + return ret; +} + +static void si_dpm_disable(struct amdgpu_device *adev) +{ + struct rv7xx_power_info *pi = rv770_get_pi(adev); + struct amdgpu_ps *boot_ps = adev->pm.dpm.boot_ps; + + if (!amdgpu_si_is_smc_running(adev)) + return; + si_thermal_stop_thermal_controller(adev); + si_disable_ulv(adev); + si_clear_vc(adev); + if (pi->thermal_protection) + si_enable_thermal_protection(adev, false); + si_enable_power_containment(adev, boot_ps, false); + si_enable_smc_cac(adev, boot_ps, false); + si_enable_spread_spectrum(adev, false); + si_enable_auto_throttle_source(adev, AMDGPU_DPM_AUTO_THROTTLE_SRC_THERMAL, false); + si_stop_dpm(adev); + si_reset_to_default(adev); + si_dpm_stop_smc(adev); + si_force_switch_to_arb_f0(adev); + + ni_update_current_ps(adev, boot_ps); +} + +static int si_dpm_pre_set_power_state(struct amdgpu_device *adev) +{ + struct evergreen_power_info *eg_pi = evergreen_get_pi(adev); + struct amdgpu_ps requested_ps = *adev->pm.dpm.requested_ps; + struct amdgpu_ps *new_ps = &requested_ps; + + ni_update_requested_ps(adev, new_ps); + si_apply_state_adjust_rules(adev, &eg_pi->requested_rps); + + return 0; +} + +static int si_power_control_set_level(struct amdgpu_device *adev) +{ + struct amdgpu_ps *new_ps = adev->pm.dpm.requested_ps; + int ret; + + ret = si_restrict_performance_levels_before_switch(adev); + if (ret) + return ret; + ret = si_halt_smc(adev); + if (ret) + return ret; + ret = si_populate_smc_tdp_limits(adev, new_ps); + if (ret) + return ret; + ret = si_populate_smc_tdp_limits_2(adev, new_ps); + if (ret) + return ret; + ret = si_resume_smc(adev); + if (ret) + return ret; + ret = si_set_sw_state(adev); + if (ret) + return ret; + return 0; +} + +static int si_dpm_set_power_state(struct amdgpu_device *adev) +{ + struct evergreen_power_info *eg_pi = evergreen_get_pi(adev); + struct amdgpu_ps *new_ps = &eg_pi->requested_rps; + struct amdgpu_ps *old_ps = &eg_pi->current_rps; + int ret; + + ret = si_disable_ulv(adev); + if (ret) { + DRM_ERROR("si_disable_ulv failed\n"); + return ret; + } + ret = si_restrict_performance_levels_before_switch(adev); + if (ret) { + DRM_ERROR("si_restrict_performance_levels_before_switch failed\n"); + return ret; + } + if (eg_pi->pcie_performance_request) + si_request_link_speed_change_before_state_change(adev, new_ps, old_ps); + ni_set_uvd_clock_before_set_eng_clock(adev, new_ps, old_ps); + ret = si_enable_power_containment(adev, new_ps, false); + if (ret) { + DRM_ERROR("si_enable_power_containment failed\n"); + return ret; + } + ret = si_enable_smc_cac(adev, new_ps, false); + if (ret) { + DRM_ERROR("si_enable_smc_cac failed\n"); + return ret; + } + ret = si_halt_smc(adev); + if (ret) { + DRM_ERROR("si_halt_smc failed\n"); + return ret; + } + ret = si_upload_sw_state(adev, new_ps); + if (ret) { + DRM_ERROR("si_upload_sw_state failed\n"); + return ret; + } + ret = si_upload_smc_data(adev); + if (ret) { + DRM_ERROR("si_upload_smc_data failed\n"); + return ret; + } + ret = si_upload_ulv_state(adev); + if (ret) { + DRM_ERROR("si_upload_ulv_state failed\n"); + return ret; + } + if (eg_pi->dynamic_ac_timing) { + ret = si_upload_mc_reg_table(adev, new_ps); + if (ret) { + DRM_ERROR("si_upload_mc_reg_table failed\n"); + return ret; + } + } + ret = si_program_memory_timing_parameters(adev, new_ps); + if (ret) { + DRM_ERROR("si_program_memory_timing_parameters failed\n"); + return ret; + } + si_set_pcie_lane_width_in_smc(adev, new_ps, old_ps); + + ret = si_resume_smc(adev); + if (ret) { + DRM_ERROR("si_resume_smc failed\n"); + return ret; + } + ret = si_set_sw_state(adev); + if (ret) { + DRM_ERROR("si_set_sw_state failed\n"); + return ret; + } + ni_set_uvd_clock_after_set_eng_clock(adev, new_ps, old_ps); + if (eg_pi->pcie_performance_request) + si_notify_link_speed_change_after_state_change(adev, new_ps, old_ps); + ret = si_set_power_state_conditionally_enable_ulv(adev, new_ps); + if (ret) { + DRM_ERROR("si_set_power_state_conditionally_enable_ulv failed\n"); + return ret; + } + ret = si_enable_smc_cac(adev, new_ps, true); + if (ret) { + DRM_ERROR("si_enable_smc_cac failed\n"); + return ret; + } + ret = si_enable_power_containment(adev, new_ps, true); + if (ret) { + DRM_ERROR("si_enable_power_containment failed\n"); + return ret; + } + + ret = si_power_control_set_level(adev); + if (ret) { + DRM_ERROR("si_power_control_set_level failed\n"); + return ret; + } + + return 0; +} + +static void si_dpm_post_set_power_state(struct amdgpu_device *adev) +{ + struct evergreen_power_info *eg_pi = evergreen_get_pi(adev); + struct amdgpu_ps *new_ps = &eg_pi->requested_rps; + + ni_update_current_ps(adev, new_ps); +} + +#if 0 +void si_dpm_reset_asic(struct amdgpu_device *adev) +{ + si_restrict_performance_levels_before_switch(adev); + si_disable_ulv(adev); + si_set_boot_state(adev); +} +#endif + +static void si_dpm_display_configuration_changed(struct amdgpu_device *adev) +{ + si_program_display_gap(adev); +} + + +static void si_parse_pplib_non_clock_info(struct amdgpu_device *adev, + struct amdgpu_ps *rps, + struct _ATOM_PPLIB_NONCLOCK_INFO *non_clock_info, + u8 table_rev) +{ + rps->caps = le32_to_cpu(non_clock_info->ulCapsAndSettings); + rps->class = le16_to_cpu(non_clock_info->usClassification); + rps->class2 = le16_to_cpu(non_clock_info->usClassification2); + + if (ATOM_PPLIB_NONCLOCKINFO_VER1 < table_rev) { + rps->vclk = le32_to_cpu(non_clock_info->ulVCLK); + rps->dclk = le32_to_cpu(non_clock_info->ulDCLK); + } else if (r600_is_uvd_state(rps->class, rps->class2)) { + rps->vclk = RV770_DEFAULT_VCLK_FREQ; + rps->dclk = RV770_DEFAULT_DCLK_FREQ; + } else { + rps->vclk = 0; + rps->dclk = 0; + } + + if (rps->class & ATOM_PPLIB_CLASSIFICATION_BOOT) + adev->pm.dpm.boot_ps = rps; + if (rps->class & ATOM_PPLIB_CLASSIFICATION_UVDSTATE) + adev->pm.dpm.uvd_ps = rps; +} + +static void si_parse_pplib_clock_info(struct amdgpu_device *adev, + struct amdgpu_ps *rps, int index, + union pplib_clock_info *clock_info) +{ + struct rv7xx_power_info *pi = rv770_get_pi(adev); + struct evergreen_power_info *eg_pi = evergreen_get_pi(adev); + struct si_power_info *si_pi = si_get_pi(adev); + struct si_ps *ps = si_get_ps(rps); + u16 leakage_voltage; + struct rv7xx_pl *pl = &ps->performance_levels[index]; + int ret; + + ps->performance_level_count = index + 1; + + pl->sclk = le16_to_cpu(clock_info->si.usEngineClockLow); + pl->sclk |= clock_info->si.ucEngineClockHigh << 16; + pl->mclk = le16_to_cpu(clock_info->si.usMemoryClockLow); + pl->mclk |= clock_info->si.ucMemoryClockHigh << 16; + + pl->vddc = le16_to_cpu(clock_info->si.usVDDC); + pl->vddci = le16_to_cpu(clock_info->si.usVDDCI); + pl->flags = le32_to_cpu(clock_info->si.ulFlags); + pl->pcie_gen = r600_get_pcie_gen_support(adev, + si_pi->sys_pcie_mask, + si_pi->boot_pcie_gen, + clock_info->si.ucPCIEGen); + + /* patch up vddc if necessary */ + ret = si_get_leakage_voltage_from_leakage_index(adev, pl->vddc, + &leakage_voltage); + if (ret == 0) + pl->vddc = leakage_voltage; + + if (rps->class & ATOM_PPLIB_CLASSIFICATION_ACPI) { + pi->acpi_vddc = pl->vddc; + eg_pi->acpi_vddci = pl->vddci; + si_pi->acpi_pcie_gen = pl->pcie_gen; + } + + if ((rps->class2 & ATOM_PPLIB_CLASSIFICATION2_ULV) && + index == 0) { + /* XXX disable for A0 tahiti */ + si_pi->ulv.supported = false; + si_pi->ulv.pl = *pl; + si_pi->ulv.one_pcie_lane_in_ulv = false; + si_pi->ulv.volt_change_delay = SISLANDS_ULVVOLTAGECHANGEDELAY_DFLT; + si_pi->ulv.cg_ulv_parameter = SISLANDS_CGULVPARAMETER_DFLT; + si_pi->ulv.cg_ulv_control = SISLANDS_CGULVCONTROL_DFLT; + } + + if (pi->min_vddc_in_table > pl->vddc) + pi->min_vddc_in_table = pl->vddc; + + if (pi->max_vddc_in_table < pl->vddc) + pi->max_vddc_in_table = pl->vddc; + + /* patch up boot state */ + if (rps->class & ATOM_PPLIB_CLASSIFICATION_BOOT) { + u16 vddc, vddci, mvdd; + amdgpu_atombios_get_default_voltages(adev, &vddc, &vddci, &mvdd); + pl->mclk = adev->clock.default_mclk; + pl->sclk = adev->clock.default_sclk; + pl->vddc = vddc; + pl->vddci = vddci; + si_pi->mvdd_bootup_value = mvdd; + } + + if ((rps->class & ATOM_PPLIB_CLASSIFICATION_UI_MASK) == + ATOM_PPLIB_CLASSIFICATION_UI_PERFORMANCE) { + adev->pm.dpm.dyn_state.max_clock_voltage_on_ac.sclk = pl->sclk; + adev->pm.dpm.dyn_state.max_clock_voltage_on_ac.mclk = pl->mclk; + adev->pm.dpm.dyn_state.max_clock_voltage_on_ac.vddc = pl->vddc; + adev->pm.dpm.dyn_state.max_clock_voltage_on_ac.vddci = pl->vddci; + } +} + +union pplib_power_state { + struct _ATOM_PPLIB_STATE v1; + struct _ATOM_PPLIB_STATE_V2 v2; +}; + +static int si_parse_power_table(struct amdgpu_device *adev) +{ + struct amdgpu_mode_info *mode_info = &adev->mode_info; + struct _ATOM_PPLIB_NONCLOCK_INFO *non_clock_info; + union pplib_power_state *power_state; + int i, j, k, non_clock_array_index, clock_array_index; + union pplib_clock_info *clock_info; + struct _StateArray *state_array; + struct _ClockInfoArray *clock_info_array; + struct _NonClockInfoArray *non_clock_info_array; + union power_info *power_info; + int index = GetIndexIntoMasterTable(DATA, PowerPlayInfo); + u16 data_offset; + u8 frev, crev; + u8 *power_state_offset; + struct si_ps *ps; + + if (!amdgpu_atom_parse_data_header(mode_info->atom_context, index, NULL, + &frev, &crev, &data_offset)) + return -EINVAL; + power_info = (union power_info *)(mode_info->atom_context->bios + data_offset); + + amdgpu_add_thermal_controller(adev); + + state_array = (struct _StateArray *) + (mode_info->atom_context->bios + data_offset + + le16_to_cpu(power_info->pplib.usStateArrayOffset)); + clock_info_array = (struct _ClockInfoArray *) + (mode_info->atom_context->bios + data_offset + + le16_to_cpu(power_info->pplib.usClockInfoArrayOffset)); + non_clock_info_array = (struct _NonClockInfoArray *) + (mode_info->atom_context->bios + data_offset + + le16_to_cpu(power_info->pplib.usNonClockInfoArrayOffset)); + + adev->pm.dpm.ps = kzalloc(sizeof(struct amdgpu_ps) * + state_array->ucNumEntries, GFP_KERNEL); + if (!adev->pm.dpm.ps) + return -ENOMEM; + power_state_offset = (u8 *)state_array->states; + for (i = 0; i < state_array->ucNumEntries; i++) { + u8 *idx; + power_state = (union pplib_power_state *)power_state_offset; + non_clock_array_index = power_state->v2.nonClockInfoIndex; + non_clock_info = (struct _ATOM_PPLIB_NONCLOCK_INFO *) + &non_clock_info_array->nonClockInfo[non_clock_array_index]; + ps = kzalloc(sizeof(struct si_ps), GFP_KERNEL); + if (ps == NULL) { + kfree(adev->pm.dpm.ps); + return -ENOMEM; + } + adev->pm.dpm.ps[i].ps_priv = ps; + si_parse_pplib_non_clock_info(adev, &adev->pm.dpm.ps[i], + non_clock_info, + non_clock_info_array->ucEntrySize); + k = 0; + idx = (u8 *)&power_state->v2.clockInfoIndex[0]; + for (j = 0; j < power_state->v2.ucNumDPMLevels; j++) { + clock_array_index = idx[j]; + if (clock_array_index >= clock_info_array->ucNumEntries) + continue; + if (k >= SISLANDS_MAX_HARDWARE_POWERLEVELS) + break; + clock_info = (union pplib_clock_info *) + ((u8 *)&clock_info_array->clockInfo[0] + + (clock_array_index * clock_info_array->ucEntrySize)); + si_parse_pplib_clock_info(adev, + &adev->pm.dpm.ps[i], k, + clock_info); + k++; + } + power_state_offset += 2 + power_state->v2.ucNumDPMLevels; + } + adev->pm.dpm.num_ps = state_array->ucNumEntries; + + /* fill in the vce power states */ + for (i = 0; i < AMDGPU_MAX_VCE_LEVELS; i++) { + u32 sclk, mclk; + clock_array_index = adev->pm.dpm.vce_states[i].clk_idx; + clock_info = (union pplib_clock_info *) + &clock_info_array->clockInfo[clock_array_index * clock_info_array->ucEntrySize]; + sclk = le16_to_cpu(clock_info->si.usEngineClockLow); + sclk |= clock_info->si.ucEngineClockHigh << 16; + mclk = le16_to_cpu(clock_info->si.usMemoryClockLow); + mclk |= clock_info->si.ucMemoryClockHigh << 16; + adev->pm.dpm.vce_states[i].sclk = sclk; + adev->pm.dpm.vce_states[i].mclk = mclk; + } + + return 0; +} + +static int si_dpm_init(struct amdgpu_device *adev) +{ + struct rv7xx_power_info *pi; + struct evergreen_power_info *eg_pi; + struct ni_power_info *ni_pi; + struct si_power_info *si_pi; + struct atom_clock_dividers dividers; + int ret; + u32 mask; + + si_pi = kzalloc(sizeof(struct si_power_info), GFP_KERNEL); + if (si_pi == NULL) + return -ENOMEM; + adev->pm.dpm.priv = si_pi; + ni_pi = &si_pi->ni; + eg_pi = &ni_pi->eg; + pi = &eg_pi->rv7xx; + + ret = drm_pcie_get_speed_cap_mask(adev->ddev, &mask); + if (ret) + si_pi->sys_pcie_mask = 0; + else + si_pi->sys_pcie_mask = mask; + si_pi->force_pcie_gen = AMDGPU_PCIE_GEN_INVALID; + si_pi->boot_pcie_gen = si_get_current_pcie_speed(adev); + + si_set_max_cu_value(adev); + + rv770_get_max_vddc(adev); + si_get_leakage_vddc(adev); + si_patch_dependency_tables_based_on_leakage(adev); + + pi->acpi_vddc = 0; + eg_pi->acpi_vddci = 0; + pi->min_vddc_in_table = 0; + pi->max_vddc_in_table = 0; + + ret = amdgpu_get_platform_caps(adev); + if (ret) + return ret; + + ret = amdgpu_parse_extended_power_table(adev); + if (ret) + return ret; + + ret = si_parse_power_table(adev); + if (ret) + return ret; + + adev->pm.dpm.dyn_state.vddc_dependency_on_dispclk.entries = + kzalloc(4 * sizeof(struct amdgpu_clock_voltage_dependency_entry), GFP_KERNEL); + if (!adev->pm.dpm.dyn_state.vddc_dependency_on_dispclk.entries) { + amdgpu_free_extended_power_table(adev); + return -ENOMEM; + } + adev->pm.dpm.dyn_state.vddc_dependency_on_dispclk.count = 4; + adev->pm.dpm.dyn_state.vddc_dependency_on_dispclk.entries[0].clk = 0; + adev->pm.dpm.dyn_state.vddc_dependency_on_dispclk.entries[0].v = 0; + adev->pm.dpm.dyn_state.vddc_dependency_on_dispclk.entries[1].clk = 36000; + adev->pm.dpm.dyn_state.vddc_dependency_on_dispclk.entries[1].v = 720; + adev->pm.dpm.dyn_state.vddc_dependency_on_dispclk.entries[2].clk = 54000; + adev->pm.dpm.dyn_state.vddc_dependency_on_dispclk.entries[2].v = 810; + adev->pm.dpm.dyn_state.vddc_dependency_on_dispclk.entries[3].clk = 72000; + adev->pm.dpm.dyn_state.vddc_dependency_on_dispclk.entries[3].v = 900; + + if (adev->pm.dpm.voltage_response_time == 0) + adev->pm.dpm.voltage_response_time = R600_VOLTAGERESPONSETIME_DFLT; + if (adev->pm.dpm.backbias_response_time == 0) + adev->pm.dpm.backbias_response_time = R600_BACKBIASRESPONSETIME_DFLT; + + ret = amdgpu_atombios_get_clock_dividers(adev, COMPUTE_ENGINE_PLL_PARAM, + 0, false, ÷rs); + if (ret) + pi->ref_div = dividers.ref_div + 1; + else + pi->ref_div = R600_REFERENCEDIVIDER_DFLT; + + eg_pi->smu_uvd_hs = false; + + pi->mclk_strobe_mode_threshold = 40000; + if (si_is_special_1gb_platform(adev)) + pi->mclk_stutter_mode_threshold = 0; + else + pi->mclk_stutter_mode_threshold = pi->mclk_strobe_mode_threshold; + pi->mclk_edc_enable_threshold = 40000; + eg_pi->mclk_edc_wr_enable_threshold = 40000; + + ni_pi->mclk_rtt_mode_threshold = eg_pi->mclk_edc_wr_enable_threshold; + + pi->voltage_control = + amdgpu_atombios_is_voltage_gpio(adev, SET_VOLTAGE_TYPE_ASIC_VDDC, + VOLTAGE_OBJ_GPIO_LUT); + if (!pi->voltage_control) { + si_pi->voltage_control_svi2 = + amdgpu_atombios_is_voltage_gpio(adev, SET_VOLTAGE_TYPE_ASIC_VDDC, + VOLTAGE_OBJ_SVID2); + if (si_pi->voltage_control_svi2) + amdgpu_atombios_get_svi2_info(adev, SET_VOLTAGE_TYPE_ASIC_VDDC, + &si_pi->svd_gpio_id, &si_pi->svc_gpio_id); + } + + pi->mvdd_control = + amdgpu_atombios_is_voltage_gpio(adev, SET_VOLTAGE_TYPE_ASIC_MVDDC, + VOLTAGE_OBJ_GPIO_LUT); + + eg_pi->vddci_control = + amdgpu_atombios_is_voltage_gpio(adev, SET_VOLTAGE_TYPE_ASIC_VDDCI, + VOLTAGE_OBJ_GPIO_LUT); + if (!eg_pi->vddci_control) + si_pi->vddci_control_svi2 = + amdgpu_atombios_is_voltage_gpio(adev, SET_VOLTAGE_TYPE_ASIC_VDDCI, + VOLTAGE_OBJ_SVID2); + + si_pi->vddc_phase_shed_control = + amdgpu_atombios_is_voltage_gpio(adev, SET_VOLTAGE_TYPE_ASIC_VDDC, + VOLTAGE_OBJ_PHASE_LUT); + + rv770_get_engine_memory_ss(adev); + + pi->asi = RV770_ASI_DFLT; + pi->pasi = CYPRESS_HASI_DFLT; + pi->vrc = SISLANDS_VRC_DFLT; + + pi->gfx_clock_gating = true; + + eg_pi->sclk_deep_sleep = true; + si_pi->sclk_deep_sleep_above_low = false; + + if (adev->pm.int_thermal_type != THERMAL_TYPE_NONE) + pi->thermal_protection = true; + else + pi->thermal_protection = false; + + eg_pi->dynamic_ac_timing = true; + + eg_pi->light_sleep = true; +#if defined(CONFIG_ACPI) + eg_pi->pcie_performance_request = + amdgpu_acpi_is_pcie_performance_request_supported(adev); +#else + eg_pi->pcie_performance_request = false; +#endif + + si_pi->sram_end = SMC_RAM_END; + + adev->pm.dpm.dyn_state.mclk_sclk_ratio = 4; + adev->pm.dpm.dyn_state.sclk_mclk_delta = 15000; + adev->pm.dpm.dyn_state.vddc_vddci_delta = 200; + adev->pm.dpm.dyn_state.valid_sclk_values.count = 0; + adev->pm.dpm.dyn_state.valid_sclk_values.values = NULL; + adev->pm.dpm.dyn_state.valid_mclk_values.count = 0; + adev->pm.dpm.dyn_state.valid_mclk_values.values = NULL; + + si_initialize_powertune_defaults(adev); + + /* make sure dc limits are valid */ + if ((adev->pm.dpm.dyn_state.max_clock_voltage_on_dc.sclk == 0) || + (adev->pm.dpm.dyn_state.max_clock_voltage_on_dc.mclk == 0)) + adev->pm.dpm.dyn_state.max_clock_voltage_on_dc = + adev->pm.dpm.dyn_state.max_clock_voltage_on_ac; + + si_pi->fan_ctrl_is_in_default_mode = true; + + return 0; +} + +static void si_dpm_fini(struct amdgpu_device *adev) +{ + int i; + + if (adev->pm.dpm.ps) + for (i = 0; i < adev->pm.dpm.num_ps; i++) + kfree(adev->pm.dpm.ps[i].ps_priv); + kfree(adev->pm.dpm.ps); + kfree(adev->pm.dpm.priv); + kfree(adev->pm.dpm.dyn_state.vddc_dependency_on_dispclk.entries); + amdgpu_free_extended_power_table(adev); +} + +static void si_dpm_debugfs_print_current_performance_level(struct amdgpu_device *adev, + struct seq_file *m) +{ + struct evergreen_power_info *eg_pi = evergreen_get_pi(adev); + struct amdgpu_ps *rps = &eg_pi->current_rps; + struct si_ps *ps = si_get_ps(rps); + struct rv7xx_pl *pl; + u32 current_index = + (RREG32(TARGET_AND_CURRENT_PROFILE_INDEX) & CURRENT_STATE_INDEX_MASK) >> + CURRENT_STATE_INDEX_SHIFT; + + if (current_index >= ps->performance_level_count) { + seq_printf(m, "invalid dpm profile %d\n", current_index); + } else { + pl = &ps->performance_levels[current_index]; + seq_printf(m, "uvd vclk: %d dclk: %d\n", rps->vclk, rps->dclk); + seq_printf(m, "power level %d sclk: %u mclk: %u vddc: %u vddci: %u pcie gen: %u\n", + current_index, pl->sclk, pl->mclk, pl->vddc, pl->vddci, pl->pcie_gen + 1); + } +} + +static int si_dpm_set_interrupt_state(struct amdgpu_device *adev, + struct amdgpu_irq_src *source, + unsigned type, + enum amdgpu_interrupt_state state) +{ + u32 cg_thermal_int; + + switch (type) { + case AMDGPU_THERMAL_IRQ_LOW_TO_HIGH: + switch (state) { + case AMDGPU_IRQ_STATE_DISABLE: + cg_thermal_int = RREG32_SMC(CG_THERMAL_INT); + cg_thermal_int |= THERM_INT_MASK_HIGH; + WREG32_SMC(CG_THERMAL_INT, cg_thermal_int); + break; + case AMDGPU_IRQ_STATE_ENABLE: + cg_thermal_int = RREG32_SMC(CG_THERMAL_INT); + cg_thermal_int &= ~THERM_INT_MASK_HIGH; + WREG32_SMC(CG_THERMAL_INT, cg_thermal_int); + break; + default: + break; + } + break; + + case AMDGPU_THERMAL_IRQ_HIGH_TO_LOW: + switch (state) { + case AMDGPU_IRQ_STATE_DISABLE: + cg_thermal_int = RREG32_SMC(CG_THERMAL_INT); + cg_thermal_int |= THERM_INT_MASK_LOW; + WREG32_SMC(CG_THERMAL_INT, cg_thermal_int); + break; + case AMDGPU_IRQ_STATE_ENABLE: + cg_thermal_int = RREG32_SMC(CG_THERMAL_INT); + cg_thermal_int &= ~THERM_INT_MASK_LOW; + WREG32_SMC(CG_THERMAL_INT, cg_thermal_int); + break; + default: + break; + } + break; + + default: + break; + } + return 0; +} + +static int si_dpm_process_interrupt(struct amdgpu_device *adev, + struct amdgpu_irq_src *source, + struct amdgpu_iv_entry *entry) +{ + bool queue_thermal = false; + + if (entry == NULL) + return -EINVAL; + + switch (entry->src_id) { + case 230: /* thermal low to high */ + DRM_DEBUG("IH: thermal low to high\n"); + adev->pm.dpm.thermal.high_to_low = false; + queue_thermal = true; + break; + case 231: /* thermal high to low */ + DRM_DEBUG("IH: thermal high to low\n"); + adev->pm.dpm.thermal.high_to_low = true; + queue_thermal = true; + break; + default: + break; + } + + if (queue_thermal) + schedule_work(&adev->pm.dpm.thermal.work); + + return 0; +} + +static int si_dpm_late_init(void *handle) +{ + int ret; + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + if (!amdgpu_dpm) + return 0; + + /* init the sysfs and debugfs files late */ + ret = amdgpu_pm_sysfs_init(adev); + if (ret) + return ret; + + ret = si_set_temperature_range(adev); + if (ret) + return ret; +#if 0 //TODO ? + si_dpm_powergate_uvd(adev, true); +#endif + return 0; +} + +/** + * si_dpm_init_microcode - load ucode images from disk + * + * @adev: amdgpu_device pointer + * + * Use the firmware interface to load the ucode images into + * the driver (not loaded into hw). + * Returns 0 on success, error on failure. + */ +static int si_dpm_init_microcode(struct amdgpu_device *adev) +{ + const char *chip_name; + char fw_name[30]; + int err; + + DRM_DEBUG("\n"); + switch (adev->asic_type) { + case CHIP_TAHITI: + chip_name = "tahiti"; + break; + case CHIP_PITCAIRN: + if ((adev->pdev->revision == 0x81) || + (adev->pdev->device == 0x6810) || + (adev->pdev->device == 0x6811) || + (adev->pdev->device == 0x6816) || + (adev->pdev->device == 0x6817) || + (adev->pdev->device == 0x6806)) + chip_name = "pitcairn_k"; + else + chip_name = "pitcairn"; + break; + case CHIP_VERDE: + if ((adev->pdev->revision == 0x81) || + (adev->pdev->revision == 0x83) || + (adev->pdev->revision == 0x87) || + (adev->pdev->device == 0x6820) || + (adev->pdev->device == 0x6821) || + (adev->pdev->device == 0x6822) || + (adev->pdev->device == 0x6823) || + (adev->pdev->device == 0x682A) || + (adev->pdev->device == 0x682B)) + chip_name = "verde_k"; + else + chip_name = "verde"; + break; + case CHIP_OLAND: + if ((adev->pdev->revision == 0xC7) || + (adev->pdev->revision == 0x80) || + (adev->pdev->revision == 0x81) || + (adev->pdev->revision == 0x83) || + (adev->pdev->device == 0x6604) || + (adev->pdev->device == 0x6605)) + chip_name = "oland_k"; + else + chip_name = "oland"; + break; + case CHIP_HAINAN: + if ((adev->pdev->revision == 0x81) || + (adev->pdev->revision == 0x83) || + (adev->pdev->revision == 0xC3) || + (adev->pdev->device == 0x6664) || + (adev->pdev->device == 0x6665) || + (adev->pdev->device == 0x6667)) + chip_name = "hainan_k"; + else + chip_name = "hainan"; + break; + default: BUG(); + } + + snprintf(fw_name, sizeof(fw_name), "radeon/%s_smc.bin", chip_name); + err = request_firmware(&adev->pm.fw, fw_name, adev->dev); + if (err) + goto out; + err = amdgpu_ucode_validate(adev->pm.fw); + +out: + if (err) { + DRM_ERROR("si_smc: Failed to load firmware. err = %d\"%s\"\n", + err, fw_name); + release_firmware(adev->pm.fw); + adev->pm.fw = NULL; + } + return err; + +} + +static int si_dpm_sw_init(void *handle) +{ + int ret; + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + ret = amdgpu_irq_add_id(adev, 230, &adev->pm.dpm.thermal.irq); + if (ret) + return ret; + + ret = amdgpu_irq_add_id(adev, 231, &adev->pm.dpm.thermal.irq); + if (ret) + return ret; + + /* default to balanced state */ + adev->pm.dpm.state = POWER_STATE_TYPE_BALANCED; + adev->pm.dpm.user_state = POWER_STATE_TYPE_BALANCED; + adev->pm.dpm.forced_level = AMDGPU_DPM_FORCED_LEVEL_AUTO; + adev->pm.default_sclk = adev->clock.default_sclk; + adev->pm.default_mclk = adev->clock.default_mclk; + adev->pm.current_sclk = adev->clock.default_sclk; + adev->pm.current_mclk = adev->clock.default_mclk; + adev->pm.int_thermal_type = THERMAL_TYPE_NONE; + + if (amdgpu_dpm == 0) + return 0; + + ret = si_dpm_init_microcode(adev); + if (ret) + return ret; + + INIT_WORK(&adev->pm.dpm.thermal.work, amdgpu_dpm_thermal_work_handler); + mutex_lock(&adev->pm.mutex); + ret = si_dpm_init(adev); + if (ret) + goto dpm_failed; + adev->pm.dpm.current_ps = adev->pm.dpm.requested_ps = adev->pm.dpm.boot_ps; + if (amdgpu_dpm == 1) + amdgpu_pm_print_power_states(adev); + mutex_unlock(&adev->pm.mutex); + DRM_INFO("amdgpu: dpm initialized\n"); + + return 0; + +dpm_failed: + si_dpm_fini(adev); + mutex_unlock(&adev->pm.mutex); + DRM_ERROR("amdgpu: dpm initialization failed\n"); + return ret; +} + +static int si_dpm_sw_fini(void *handle) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + mutex_lock(&adev->pm.mutex); + amdgpu_pm_sysfs_fini(adev); + si_dpm_fini(adev); + mutex_unlock(&adev->pm.mutex); + + return 0; +} + +static int si_dpm_hw_init(void *handle) +{ + int ret; + + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + if (!amdgpu_dpm) + return 0; + + mutex_lock(&adev->pm.mutex); + si_dpm_setup_asic(adev); + ret = si_dpm_enable(adev); + if (ret) + adev->pm.dpm_enabled = false; + else + adev->pm.dpm_enabled = true; + mutex_unlock(&adev->pm.mutex); + + return ret; +} + +static int si_dpm_hw_fini(void *handle) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + if (adev->pm.dpm_enabled) { + mutex_lock(&adev->pm.mutex); + si_dpm_disable(adev); + mutex_unlock(&adev->pm.mutex); + } + + return 0; +} + +static int si_dpm_suspend(void *handle) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + if (adev->pm.dpm_enabled) { + mutex_lock(&adev->pm.mutex); + /* disable dpm */ + si_dpm_disable(adev); + /* reset the power state */ + adev->pm.dpm.current_ps = adev->pm.dpm.requested_ps = adev->pm.dpm.boot_ps; + mutex_unlock(&adev->pm.mutex); + } + return 0; +} + +static int si_dpm_resume(void *handle) +{ + int ret; + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + if (adev->pm.dpm_enabled) { + /* asic init will reset to the boot state */ + mutex_lock(&adev->pm.mutex); + si_dpm_setup_asic(adev); + ret = si_dpm_enable(adev); + if (ret) + adev->pm.dpm_enabled = false; + else + adev->pm.dpm_enabled = true; + mutex_unlock(&adev->pm.mutex); + if (adev->pm.dpm_enabled) + amdgpu_pm_compute_clocks(adev); + } + return 0; +} + +static bool si_dpm_is_idle(void *handle) +{ + /* XXX */ + return true; +} + +static int si_dpm_wait_for_idle(void *handle) +{ + /* XXX */ + return 0; +} + +static int si_dpm_soft_reset(void *handle) +{ + return 0; +} + +static int si_dpm_set_clockgating_state(void *handle, + enum amd_clockgating_state state) +{ + return 0; +} + +static int si_dpm_set_powergating_state(void *handle, + enum amd_powergating_state state) +{ + return 0; +} + +/* get temperature in millidegrees */ +static int si_dpm_get_temp(struct amdgpu_device *adev) +{ + u32 temp; + int actual_temp = 0; + + temp = (RREG32(CG_MULT_THERMAL_STATUS) & CTF_TEMP_MASK) >> + CTF_TEMP_SHIFT; + + if (temp & 0x200) + actual_temp = 255; + else + actual_temp = temp & 0x1ff; + + actual_temp = (actual_temp * 1000); + + return actual_temp; +} + +static u32 si_dpm_get_sclk(struct amdgpu_device *adev, bool low) +{ + struct evergreen_power_info *eg_pi = evergreen_get_pi(adev); + struct si_ps *requested_state = si_get_ps(&eg_pi->requested_rps); + + if (low) + return requested_state->performance_levels[0].sclk; + else + return requested_state->performance_levels[requested_state->performance_level_count - 1].sclk; +} + +static u32 si_dpm_get_mclk(struct amdgpu_device *adev, bool low) +{ + struct evergreen_power_info *eg_pi = evergreen_get_pi(adev); + struct si_ps *requested_state = si_get_ps(&eg_pi->requested_rps); + + if (low) + return requested_state->performance_levels[0].mclk; + else + return requested_state->performance_levels[requested_state->performance_level_count - 1].mclk; +} + +static void si_dpm_print_power_state(struct amdgpu_device *adev, + struct amdgpu_ps *rps) +{ + struct si_ps *ps = si_get_ps(rps); + struct rv7xx_pl *pl; + int i; + + amdgpu_dpm_print_class_info(rps->class, rps->class2); + amdgpu_dpm_print_cap_info(rps->caps); + DRM_INFO("\tuvd vclk: %d dclk: %d\n", rps->vclk, rps->dclk); + for (i = 0; i < ps->performance_level_count; i++) { + pl = &ps->performance_levels[i]; + if (adev->asic_type >= CHIP_TAHITI) + DRM_INFO("\t\tpower level %d sclk: %u mclk: %u vddc: %u vddci: %u pcie gen: %u\n", + i, pl->sclk, pl->mclk, pl->vddc, pl->vddci, pl->pcie_gen + 1); + else + DRM_INFO("\t\tpower level %d sclk: %u mclk: %u vddc: %u vddci: %u\n", + i, pl->sclk, pl->mclk, pl->vddc, pl->vddci); + } + amdgpu_dpm_print_ps_status(adev, rps); +} + +static int si_dpm_early_init(void *handle) +{ + + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + si_dpm_set_dpm_funcs(adev); + si_dpm_set_irq_funcs(adev); + return 0; +} + + +const struct amd_ip_funcs si_dpm_ip_funcs = { + .name = "si_dpm", + .early_init = si_dpm_early_init, + .late_init = si_dpm_late_init, + .sw_init = si_dpm_sw_init, + .sw_fini = si_dpm_sw_fini, + .hw_init = si_dpm_hw_init, + .hw_fini = si_dpm_hw_fini, + .suspend = si_dpm_suspend, + .resume = si_dpm_resume, + .is_idle = si_dpm_is_idle, + .wait_for_idle = si_dpm_wait_for_idle, + .soft_reset = si_dpm_soft_reset, + .set_clockgating_state = si_dpm_set_clockgating_state, + .set_powergating_state = si_dpm_set_powergating_state, +}; + +static const struct amdgpu_dpm_funcs si_dpm_funcs = { + .get_temperature = &si_dpm_get_temp, + .pre_set_power_state = &si_dpm_pre_set_power_state, + .set_power_state = &si_dpm_set_power_state, + .post_set_power_state = &si_dpm_post_set_power_state, + .display_configuration_changed = &si_dpm_display_configuration_changed, + .get_sclk = &si_dpm_get_sclk, + .get_mclk = &si_dpm_get_mclk, + .print_power_state = &si_dpm_print_power_state, + .debugfs_print_current_performance_level = &si_dpm_debugfs_print_current_performance_level, + .force_performance_level = &si_dpm_force_performance_level, + .vblank_too_short = &si_dpm_vblank_too_short, + .set_fan_control_mode = &si_dpm_set_fan_control_mode, + .get_fan_control_mode = &si_dpm_get_fan_control_mode, + .set_fan_speed_percent = &si_dpm_set_fan_speed_percent, + .get_fan_speed_percent = &si_dpm_get_fan_speed_percent, +}; + +static void si_dpm_set_dpm_funcs(struct amdgpu_device *adev) +{ + if (adev->pm.funcs == NULL) + adev->pm.funcs = &si_dpm_funcs; +} + +static const struct amdgpu_irq_src_funcs si_dpm_irq_funcs = { + .set = si_dpm_set_interrupt_state, + .process = si_dpm_process_interrupt, +}; + +static void si_dpm_set_irq_funcs(struct amdgpu_device *adev) +{ + adev->pm.dpm.thermal.irq.num_types = AMDGPU_THERMAL_IRQ_LAST; + adev->pm.dpm.thermal.irq.funcs = &si_dpm_irq_funcs; +} + diff --git a/drivers/gpu/drm/amd/amdgpu/si_dpm.h b/drivers/gpu/drm/amd/amdgpu/si_dpm.h new file mode 100644 index 000000000000..51ce21c5f4fb --- /dev/null +++ b/drivers/gpu/drm/amd/amdgpu/si_dpm.h @@ -0,0 +1,1015 @@ +/* + * Copyright 2012 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + */ +#ifndef __SI_DPM_H__ +#define __SI_DPM_H__ + +#include "amdgpu_atombios.h" +#include "sislands_smc.h" + +#define MC_CG_CONFIG 0x96f +#define MC_ARB_CG 0x9fa +#define CG_ARB_REQ(x) ((x) << 0) +#define CG_ARB_REQ_MASK (0xff << 0) + +#define MC_ARB_DRAM_TIMING_1 0x9fc +#define MC_ARB_DRAM_TIMING_2 0x9fd +#define MC_ARB_DRAM_TIMING_3 0x9fe +#define MC_ARB_DRAM_TIMING2_1 0x9ff +#define MC_ARB_DRAM_TIMING2_2 0xa00 +#define MC_ARB_DRAM_TIMING2_3 0xa01 + +#define MAX_NO_OF_MVDD_VALUES 2 +#define MAX_NO_VREG_STEPS 32 +#define NISLANDS_MAX_SMC_PERFORMANCE_LEVELS_PER_SWSTATE 16 +#define SMC_NISLANDS_MC_REGISTER_ARRAY_SIZE 32 +#define SMC_NISLANDS_MC_REGISTER_ARRAY_SET_COUNT 20 +#define RV770_ASI_DFLT 1000 +#define CYPRESS_HASI_DFLT 400000 +#define PCIE_PERF_REQ_PECI_GEN1 2 +#define PCIE_PERF_REQ_PECI_GEN2 3 +#define PCIE_PERF_REQ_PECI_GEN3 4 +#define RV770_DEFAULT_VCLK_FREQ 53300 /* 10 khz */ +#define RV770_DEFAULT_DCLK_FREQ 40000 /* 10 khz */ + +#define SMC_EVERGREEN_MC_REGISTER_ARRAY_SIZE 16 + +#define RV770_SMC_TABLE_ADDRESS 0xB000 +#define RV770_SMC_PERFORMANCE_LEVELS_PER_SWSTATE 3 + +#define SMC_STROBE_RATIO 0x0F +#define SMC_STROBE_ENABLE 0x10 + +#define SMC_MC_EDC_RD_FLAG 0x01 +#define SMC_MC_EDC_WR_FLAG 0x02 +#define SMC_MC_RTT_ENABLE 0x04 +#define SMC_MC_STUTTER_EN 0x08 + +#define RV770_SMC_VOLTAGEMASK_VDDC 0 +#define RV770_SMC_VOLTAGEMASK_MVDD 1 +#define RV770_SMC_VOLTAGEMASK_VDDCI 2 +#define RV770_SMC_VOLTAGEMASK_MAX 4 + +#define NISLANDS_MAX_SMC_PERFORMANCE_LEVELS_PER_SWSTATE 16 +#define NISLANDS_SMC_STROBE_RATIO 0x0F +#define NISLANDS_SMC_STROBE_ENABLE 0x10 + +#define NISLANDS_SMC_MC_EDC_RD_FLAG 0x01 +#define NISLANDS_SMC_MC_EDC_WR_FLAG 0x02 +#define NISLANDS_SMC_MC_RTT_ENABLE 0x04 +#define NISLANDS_SMC_MC_STUTTER_EN 0x08 + +#define MAX_NO_VREG_STEPS 32 + +#define NISLANDS_SMC_VOLTAGEMASK_VDDC 0 +#define NISLANDS_SMC_VOLTAGEMASK_MVDD 1 +#define NISLANDS_SMC_VOLTAGEMASK_VDDCI 2 +#define NISLANDS_SMC_VOLTAGEMASK_MAX 4 + +#define SISLANDS_MCREGISTERTABLE_INITIAL_SLOT 0 +#define SISLANDS_MCREGISTERTABLE_ACPI_SLOT 1 +#define SISLANDS_MCREGISTERTABLE_ULV_SLOT 2 +#define SISLANDS_MCREGISTERTABLE_FIRST_DRIVERSTATE_SLOT 3 + +#define SISLANDS_LEAKAGE_INDEX0 0xff01 +#define SISLANDS_MAX_LEAKAGE_COUNT 4 + +#define SISLANDS_MAX_HARDWARE_POWERLEVELS 5 +#define SISLANDS_INITIAL_STATE_ARB_INDEX 0 +#define SISLANDS_ACPI_STATE_ARB_INDEX 1 +#define SISLANDS_ULV_STATE_ARB_INDEX 2 +#define SISLANDS_DRIVER_STATE_ARB_INDEX 3 + +#define SISLANDS_DPM2_MAX_PULSE_SKIP 256 + +#define SISLANDS_DPM2_NEAR_TDP_DEC 10 +#define SISLANDS_DPM2_ABOVE_SAFE_INC 5 +#define SISLANDS_DPM2_BELOW_SAFE_INC 20 + +#define SISLANDS_DPM2_TDP_SAFE_LIMIT_PERCENT 80 + +#define SISLANDS_DPM2_MAXPS_PERCENT_H 99 +#define SISLANDS_DPM2_MAXPS_PERCENT_M 99 + +#define SISLANDS_DPM2_SQ_RAMP_MAX_POWER 0x3FFF +#define SISLANDS_DPM2_SQ_RAMP_MIN_POWER 0x12 +#define SISLANDS_DPM2_SQ_RAMP_MAX_POWER_DELTA 0x15 +#define SISLANDS_DPM2_SQ_RAMP_STI_SIZE 0x1E +#define SISLANDS_DPM2_SQ_RAMP_LTI_RATIO 0xF + +#define SISLANDS_DPM2_PWREFFICIENCYRATIO_MARGIN 10 + +#define SISLANDS_VRC_DFLT 0xC000B3 +#define SISLANDS_ULVVOLTAGECHANGEDELAY_DFLT 1687 +#define SISLANDS_CGULVPARAMETER_DFLT 0x00040035 +#define SISLANDS_CGULVCONTROL_DFLT 0x1f007550 + +#define SI_ASI_DFLT 10000 +#define SI_BSP_DFLT 0x41EB +#define SI_BSU_DFLT 0x2 +#define SI_AH_DFLT 5 +#define SI_RLP_DFLT 25 +#define SI_RMP_DFLT 65 +#define SI_LHP_DFLT 40 +#define SI_LMP_DFLT 15 +#define SI_TD_DFLT 0 +#define SI_UTC_DFLT_00 0x24 +#define SI_UTC_DFLT_01 0x22 +#define SI_UTC_DFLT_02 0x22 +#define SI_UTC_DFLT_03 0x22 +#define SI_UTC_DFLT_04 0x22 +#define SI_UTC_DFLT_05 0x22 +#define SI_UTC_DFLT_06 0x22 +#define SI_UTC_DFLT_07 0x22 +#define SI_UTC_DFLT_08 0x22 +#define SI_UTC_DFLT_09 0x22 +#define SI_UTC_DFLT_10 0x22 +#define SI_UTC_DFLT_11 0x22 +#define SI_UTC_DFLT_12 0x22 +#define SI_UTC_DFLT_13 0x22 +#define SI_UTC_DFLT_14 0x22 +#define SI_DTC_DFLT_00 0x24 +#define SI_DTC_DFLT_01 0x22 +#define SI_DTC_DFLT_02 0x22 +#define SI_DTC_DFLT_03 0x22 +#define SI_DTC_DFLT_04 0x22 +#define SI_DTC_DFLT_05 0x22 +#define SI_DTC_DFLT_06 0x22 +#define SI_DTC_DFLT_07 0x22 +#define SI_DTC_DFLT_08 0x22 +#define SI_DTC_DFLT_09 0x22 +#define SI_DTC_DFLT_10 0x22 +#define SI_DTC_DFLT_11 0x22 +#define SI_DTC_DFLT_12 0x22 +#define SI_DTC_DFLT_13 0x22 +#define SI_DTC_DFLT_14 0x22 +#define SI_VRC_DFLT 0x0000C003 +#define SI_VOLTAGERESPONSETIME_DFLT 1000 +#define SI_BACKBIASRESPONSETIME_DFLT 1000 +#define SI_VRU_DFLT 0x3 +#define SI_SPLLSTEPTIME_DFLT 0x1000 +#define SI_SPLLSTEPUNIT_DFLT 0x3 +#define SI_TPU_DFLT 0 +#define SI_TPC_DFLT 0x200 +#define SI_SSTU_DFLT 0 +#define SI_SST_DFLT 0x00C8 +#define SI_GICST_DFLT 0x200 +#define SI_FCT_DFLT 0x0400 +#define SI_FCTU_DFLT 0 +#define SI_CTXCGTT3DRPHC_DFLT 0x20 +#define SI_CTXCGTT3DRSDC_DFLT 0x40 +#define SI_VDDC3DOORPHC_DFLT 0x100 +#define SI_VDDC3DOORSDC_DFLT 0x7 +#define SI_VDDC3DOORSU_DFLT 0 +#define SI_MPLLLOCKTIME_DFLT 100 +#define SI_MPLLRESETTIME_DFLT 150 +#define SI_VCOSTEPPCT_DFLT 20 +#define SI_ENDINGVCOSTEPPCT_DFLT 5 +#define SI_REFERENCEDIVIDER_DFLT 4 + +#define SI_PM_NUMBER_OF_TC 15 +#define SI_PM_NUMBER_OF_SCLKS 20 +#define SI_PM_NUMBER_OF_MCLKS 4 +#define SI_PM_NUMBER_OF_VOLTAGE_LEVELS 4 +#define SI_PM_NUMBER_OF_ACTIVITY_LEVELS 3 + +/* XXX are these ok? */ +#define SI_TEMP_RANGE_MIN (90 * 1000) +#define SI_TEMP_RANGE_MAX (120 * 1000) + +#define FDO_PWM_MODE_STATIC 1 +#define FDO_PWM_MODE_STATIC_RPM 5 + +enum ni_dc_cac_level +{ + NISLANDS_DCCAC_LEVEL_0 = 0, + NISLANDS_DCCAC_LEVEL_1, + NISLANDS_DCCAC_LEVEL_2, + NISLANDS_DCCAC_LEVEL_3, + NISLANDS_DCCAC_LEVEL_4, + NISLANDS_DCCAC_LEVEL_5, + NISLANDS_DCCAC_LEVEL_6, + NISLANDS_DCCAC_LEVEL_7, + NISLANDS_DCCAC_MAX_LEVELS +}; + +enum si_cac_config_reg_type +{ + SISLANDS_CACCONFIG_MMR = 0, + SISLANDS_CACCONFIG_CGIND, + SISLANDS_CACCONFIG_MAX +}; + +enum si_power_level { + SI_POWER_LEVEL_LOW = 0, + SI_POWER_LEVEL_MEDIUM = 1, + SI_POWER_LEVEL_HIGH = 2, + SI_POWER_LEVEL_CTXSW = 3, +}; + +enum si_td { + SI_TD_AUTO, + SI_TD_UP, + SI_TD_DOWN, +}; + +enum si_display_watermark { + SI_DISPLAY_WATERMARK_LOW = 0, + SI_DISPLAY_WATERMARK_HIGH = 1, +}; + +enum si_display_gap +{ + SI_PM_DISPLAY_GAP_VBLANK_OR_WM = 0, + SI_PM_DISPLAY_GAP_VBLANK = 1, + SI_PM_DISPLAY_GAP_WATERMARK = 2, + SI_PM_DISPLAY_GAP_IGNORE = 3, +}; + +extern const struct amd_ip_funcs si_dpm_ip_funcs; + +struct ni_leakage_coeffients +{ + u32 at; + u32 bt; + u32 av; + u32 bv; + s32 t_slope; + s32 t_intercept; + u32 t_ref; +}; + +struct SMC_Evergreen_MCRegisterAddress +{ + uint16_t s0; + uint16_t s1; +}; + +typedef struct SMC_Evergreen_MCRegisterAddress SMC_Evergreen_MCRegisterAddress; + +struct evergreen_mc_reg_entry { + u32 mclk_max; + u32 mc_data[SMC_EVERGREEN_MC_REGISTER_ARRAY_SIZE]; +}; + +struct evergreen_mc_reg_table { + u8 last; + u8 num_entries; + u16 valid_flag; + struct evergreen_mc_reg_entry mc_reg_table_entry[MAX_AC_TIMING_ENTRIES]; + SMC_Evergreen_MCRegisterAddress mc_reg_address[SMC_EVERGREEN_MC_REGISTER_ARRAY_SIZE]; +}; + +struct SMC_Evergreen_MCRegisterSet +{ + uint32_t value[SMC_EVERGREEN_MC_REGISTER_ARRAY_SIZE]; +}; + +typedef struct SMC_Evergreen_MCRegisterSet SMC_Evergreen_MCRegisterSet; + +struct SMC_Evergreen_MCRegisters +{ + uint8_t last; + uint8_t reserved[3]; + SMC_Evergreen_MCRegisterAddress address[SMC_EVERGREEN_MC_REGISTER_ARRAY_SIZE]; + SMC_Evergreen_MCRegisterSet data[5]; +}; + +typedef struct SMC_Evergreen_MCRegisters SMC_Evergreen_MCRegisters; + +struct SMC_NIslands_MCRegisterSet +{ + uint32_t value[SMC_NISLANDS_MC_REGISTER_ARRAY_SIZE]; +}; + +typedef struct SMC_NIslands_MCRegisterSet SMC_NIslands_MCRegisterSet; + +struct ni_mc_reg_entry { + u32 mclk_max; + u32 mc_data[SMC_NISLANDS_MC_REGISTER_ARRAY_SIZE]; +}; + +struct SMC_NIslands_MCRegisterAddress +{ + uint16_t s0; + uint16_t s1; +}; + +typedef struct SMC_NIslands_MCRegisterAddress SMC_NIslands_MCRegisterAddress; + +struct SMC_NIslands_MCRegisters +{ + uint8_t last; + uint8_t reserved[3]; + SMC_NIslands_MCRegisterAddress address[SMC_NISLANDS_MC_REGISTER_ARRAY_SIZE]; + SMC_NIslands_MCRegisterSet data[SMC_NISLANDS_MC_REGISTER_ARRAY_SET_COUNT]; +}; + +typedef struct SMC_NIslands_MCRegisters SMC_NIslands_MCRegisters; + +struct evergreen_ulv_param { + bool supported; + struct rv7xx_pl *pl; +}; + +struct evergreen_arb_registers { + u32 mc_arb_dram_timing; + u32 mc_arb_dram_timing2; + u32 mc_arb_rfsh_rate; + u32 mc_arb_burst_time; +}; + +struct at { + u32 rlp; + u32 rmp; + u32 lhp; + u32 lmp; +}; + +struct ni_clock_registers { + u32 cg_spll_func_cntl; + u32 cg_spll_func_cntl_2; + u32 cg_spll_func_cntl_3; + u32 cg_spll_func_cntl_4; + u32 cg_spll_spread_spectrum; + u32 cg_spll_spread_spectrum_2; + u32 mclk_pwrmgt_cntl; + u32 dll_cntl; + u32 mpll_ad_func_cntl; + u32 mpll_ad_func_cntl_2; + u32 mpll_dq_func_cntl; + u32 mpll_dq_func_cntl_2; + u32 mpll_ss1; + u32 mpll_ss2; +}; + +struct RV770_SMC_SCLK_VALUE +{ + uint32_t vCG_SPLL_FUNC_CNTL; + uint32_t vCG_SPLL_FUNC_CNTL_2; + uint32_t vCG_SPLL_FUNC_CNTL_3; + uint32_t vCG_SPLL_SPREAD_SPECTRUM; + uint32_t vCG_SPLL_SPREAD_SPECTRUM_2; + uint32_t sclk_value; +}; + +typedef struct RV770_SMC_SCLK_VALUE RV770_SMC_SCLK_VALUE; + +struct RV770_SMC_MCLK_VALUE +{ + uint32_t vMPLL_AD_FUNC_CNTL; + uint32_t vMPLL_AD_FUNC_CNTL_2; + uint32_t vMPLL_DQ_FUNC_CNTL; + uint32_t vMPLL_DQ_FUNC_CNTL_2; + uint32_t vMCLK_PWRMGT_CNTL; + uint32_t vDLL_CNTL; + uint32_t vMPLL_SS; + uint32_t vMPLL_SS2; + uint32_t mclk_value; +}; + +typedef struct RV770_SMC_MCLK_VALUE RV770_SMC_MCLK_VALUE; + + +struct RV730_SMC_MCLK_VALUE +{ + uint32_t vMCLK_PWRMGT_CNTL; + uint32_t vDLL_CNTL; + uint32_t vMPLL_FUNC_CNTL; + uint32_t vMPLL_FUNC_CNTL2; + uint32_t vMPLL_FUNC_CNTL3; + uint32_t vMPLL_SS; + uint32_t vMPLL_SS2; + uint32_t mclk_value; +}; + +typedef struct RV730_SMC_MCLK_VALUE RV730_SMC_MCLK_VALUE; + +struct RV770_SMC_VOLTAGE_VALUE +{ + uint16_t value; + uint8_t index; + uint8_t padding; +}; + +typedef struct RV770_SMC_VOLTAGE_VALUE RV770_SMC_VOLTAGE_VALUE; + +union RV7XX_SMC_MCLK_VALUE +{ + RV770_SMC_MCLK_VALUE mclk770; + RV730_SMC_MCLK_VALUE mclk730; +}; + +typedef union RV7XX_SMC_MCLK_VALUE RV7XX_SMC_MCLK_VALUE, *LPRV7XX_SMC_MCLK_VALUE; + +struct RV770_SMC_HW_PERFORMANCE_LEVEL +{ + uint8_t arbValue; + union{ + uint8_t seqValue; + uint8_t ACIndex; + }; + uint8_t displayWatermark; + uint8_t gen2PCIE; + uint8_t gen2XSP; + uint8_t backbias; + uint8_t strobeMode; + uint8_t mcFlags; + uint32_t aT; + uint32_t bSP; + RV770_SMC_SCLK_VALUE sclk; + RV7XX_SMC_MCLK_VALUE mclk; + RV770_SMC_VOLTAGE_VALUE vddc; + RV770_SMC_VOLTAGE_VALUE mvdd; + RV770_SMC_VOLTAGE_VALUE vddci; + uint8_t reserved1; + uint8_t reserved2; + uint8_t stateFlags; + uint8_t padding; +}; + +typedef struct RV770_SMC_HW_PERFORMANCE_LEVEL RV770_SMC_HW_PERFORMANCE_LEVEL; + +struct RV770_SMC_SWSTATE +{ + uint8_t flags; + uint8_t padding1; + uint8_t padding2; + uint8_t padding3; + RV770_SMC_HW_PERFORMANCE_LEVEL levels[RV770_SMC_PERFORMANCE_LEVELS_PER_SWSTATE]; +}; + +typedef struct RV770_SMC_SWSTATE RV770_SMC_SWSTATE; + +struct RV770_SMC_VOLTAGEMASKTABLE +{ + uint8_t highMask[RV770_SMC_VOLTAGEMASK_MAX]; + uint32_t lowMask[RV770_SMC_VOLTAGEMASK_MAX]; +}; + +typedef struct RV770_SMC_VOLTAGEMASKTABLE RV770_SMC_VOLTAGEMASKTABLE; + +struct RV770_SMC_STATETABLE +{ + uint8_t thermalProtectType; + uint8_t systemFlags; + uint8_t maxVDDCIndexInPPTable; + uint8_t extraFlags; + uint8_t highSMIO[MAX_NO_VREG_STEPS]; + uint32_t lowSMIO[MAX_NO_VREG_STEPS]; + RV770_SMC_VOLTAGEMASKTABLE voltageMaskTable; + RV770_SMC_SWSTATE initialState; + RV770_SMC_SWSTATE ACPIState; + RV770_SMC_SWSTATE driverState; + RV770_SMC_SWSTATE ULVState; +}; + +typedef struct RV770_SMC_STATETABLE RV770_SMC_STATETABLE; + +struct vddc_table_entry { + u16 vddc; + u8 vddc_index; + u8 high_smio; + u32 low_smio; +}; + +struct rv770_clock_registers { + u32 cg_spll_func_cntl; + u32 cg_spll_func_cntl_2; + u32 cg_spll_func_cntl_3; + u32 cg_spll_spread_spectrum; + u32 cg_spll_spread_spectrum_2; + u32 mpll_ad_func_cntl; + u32 mpll_ad_func_cntl_2; + u32 mpll_dq_func_cntl; + u32 mpll_dq_func_cntl_2; + u32 mclk_pwrmgt_cntl; + u32 dll_cntl; + u32 mpll_ss1; + u32 mpll_ss2; +}; + +struct rv730_clock_registers { + u32 cg_spll_func_cntl; + u32 cg_spll_func_cntl_2; + u32 cg_spll_func_cntl_3; + u32 cg_spll_spread_spectrum; + u32 cg_spll_spread_spectrum_2; + u32 mclk_pwrmgt_cntl; + u32 dll_cntl; + u32 mpll_func_cntl; + u32 mpll_func_cntl2; + u32 mpll_func_cntl3; + u32 mpll_ss; + u32 mpll_ss2; +}; + +union r7xx_clock_registers { + struct rv770_clock_registers rv770; + struct rv730_clock_registers rv730; +}; + +struct rv7xx_power_info { + /* flags */ + bool mem_gddr5; + bool pcie_gen2; + bool dynamic_pcie_gen2; + bool acpi_pcie_gen2; + bool boot_in_gen2; + bool voltage_control; /* vddc */ + bool mvdd_control; + bool sclk_ss; + bool mclk_ss; + bool dynamic_ss; + bool gfx_clock_gating; + bool mg_clock_gating; + bool mgcgtssm; + bool power_gating; + bool thermal_protection; + bool display_gap; + bool dcodt; + bool ulps; + /* registers */ + union r7xx_clock_registers clk_regs; + u32 s0_vid_lower_smio_cntl; + /* voltage */ + u32 vddc_mask_low; + u32 mvdd_mask_low; + u32 mvdd_split_frequency; + u32 mvdd_low_smio[MAX_NO_OF_MVDD_VALUES]; + u16 max_vddc; + u16 max_vddc_in_table; + u16 min_vddc_in_table; + struct vddc_table_entry vddc_table[MAX_NO_VREG_STEPS]; + u8 valid_vddc_entries; + /* dc odt */ + u32 mclk_odt_threshold; + u8 odt_value_0[2]; + u8 odt_value_1[2]; + /* stored values */ + u32 boot_sclk; + u16 acpi_vddc; + u32 ref_div; + u32 active_auto_throttle_sources; + u32 mclk_stutter_mode_threshold; + u32 mclk_strobe_mode_threshold; + u32 mclk_edc_enable_threshold; + u32 bsp; + u32 bsu; + u32 pbsp; + u32 pbsu; + u32 dsp; + u32 psp; + u32 asi; + u32 pasi; + u32 vrc; + u32 restricted_levels; + u32 rlp; + u32 rmp; + u32 lhp; + u32 lmp; + /* smc offsets */ + u16 state_table_start; + u16 soft_regs_start; + u16 sram_end; + /* scratch structs */ + RV770_SMC_STATETABLE smc_statetable; +}; + +struct rv7xx_pl { + u32 sclk; + u32 mclk; + u16 vddc; + u16 vddci; /* eg+ only */ + u32 flags; + enum amdgpu_pcie_gen pcie_gen; /* si+ only */ +}; + +struct rv7xx_ps { + struct rv7xx_pl high; + struct rv7xx_pl medium; + struct rv7xx_pl low; + bool dc_compatible; +}; + +struct si_ps { + u16 performance_level_count; + bool dc_compatible; + struct rv7xx_pl performance_levels[NISLANDS_MAX_SMC_PERFORMANCE_LEVELS_PER_SWSTATE]; +}; + +struct ni_mc_reg_table { + u8 last; + u8 num_entries; + u16 valid_flag; + struct ni_mc_reg_entry mc_reg_table_entry[MAX_AC_TIMING_ENTRIES]; + SMC_NIslands_MCRegisterAddress mc_reg_address[SMC_NISLANDS_MC_REGISTER_ARRAY_SIZE]; +}; + +struct ni_cac_data +{ + struct ni_leakage_coeffients leakage_coefficients; + u32 i_leakage; + s32 leakage_minimum_temperature; + u32 pwr_const; + u32 dc_cac_value; + u32 bif_cac_value; + u32 lkge_pwr; + u8 mc_wr_weight; + u8 mc_rd_weight; + u8 allow_ovrflw; + u8 num_win_tdp; + u8 l2num_win_tdp; + u8 lts_truncate_n; +}; + +struct evergreen_power_info { + /* must be first! */ + struct rv7xx_power_info rv7xx; + /* flags */ + bool vddci_control; + bool dynamic_ac_timing; + bool abm; + bool mcls; + bool light_sleep; + bool memory_transition; + bool pcie_performance_request; + bool pcie_performance_request_registered; + bool sclk_deep_sleep; + bool dll_default_on; + bool ls_clock_gating; + bool smu_uvd_hs; + bool uvd_enabled; + /* stored values */ + u16 acpi_vddci; + u8 mvdd_high_index; + u8 mvdd_low_index; + u32 mclk_edc_wr_enable_threshold; + struct evergreen_mc_reg_table mc_reg_table; + struct atom_voltage_table vddc_voltage_table; + struct atom_voltage_table vddci_voltage_table; + struct evergreen_arb_registers bootup_arb_registers; + struct evergreen_ulv_param ulv; + struct at ats[2]; + /* smc offsets */ + u16 mc_reg_table_start; + struct amdgpu_ps current_rps; + struct rv7xx_ps current_ps; + struct amdgpu_ps requested_rps; + struct rv7xx_ps requested_ps; +}; + +struct PP_NIslands_Dpm2PerfLevel +{ + uint8_t MaxPS; + uint8_t TgtAct; + uint8_t MaxPS_StepInc; + uint8_t MaxPS_StepDec; + uint8_t PSST; + uint8_t NearTDPDec; + uint8_t AboveSafeInc; + uint8_t BelowSafeInc; + uint8_t PSDeltaLimit; + uint8_t PSDeltaWin; + uint8_t Reserved[6]; +}; + +typedef struct PP_NIslands_Dpm2PerfLevel PP_NIslands_Dpm2PerfLevel; + +struct PP_NIslands_DPM2Parameters +{ + uint32_t TDPLimit; + uint32_t NearTDPLimit; + uint32_t SafePowerLimit; + uint32_t PowerBoostLimit; +}; +typedef struct PP_NIslands_DPM2Parameters PP_NIslands_DPM2Parameters; + +struct NISLANDS_SMC_SCLK_VALUE +{ + uint32_t vCG_SPLL_FUNC_CNTL; + uint32_t vCG_SPLL_FUNC_CNTL_2; + uint32_t vCG_SPLL_FUNC_CNTL_3; + uint32_t vCG_SPLL_FUNC_CNTL_4; + uint32_t vCG_SPLL_SPREAD_SPECTRUM; + uint32_t vCG_SPLL_SPREAD_SPECTRUM_2; + uint32_t sclk_value; +}; + +typedef struct NISLANDS_SMC_SCLK_VALUE NISLANDS_SMC_SCLK_VALUE; + +struct NISLANDS_SMC_MCLK_VALUE +{ + uint32_t vMPLL_FUNC_CNTL; + uint32_t vMPLL_FUNC_CNTL_1; + uint32_t vMPLL_FUNC_CNTL_2; + uint32_t vMPLL_AD_FUNC_CNTL; + uint32_t vMPLL_AD_FUNC_CNTL_2; + uint32_t vMPLL_DQ_FUNC_CNTL; + uint32_t vMPLL_DQ_FUNC_CNTL_2; + uint32_t vMCLK_PWRMGT_CNTL; + uint32_t vDLL_CNTL; + uint32_t vMPLL_SS; + uint32_t vMPLL_SS2; + uint32_t mclk_value; +}; + +typedef struct NISLANDS_SMC_MCLK_VALUE NISLANDS_SMC_MCLK_VALUE; + +struct NISLANDS_SMC_VOLTAGE_VALUE +{ + uint16_t value; + uint8_t index; + uint8_t padding; +}; + +typedef struct NISLANDS_SMC_VOLTAGE_VALUE NISLANDS_SMC_VOLTAGE_VALUE; + +struct NISLANDS_SMC_HW_PERFORMANCE_LEVEL +{ + uint8_t arbValue; + uint8_t ACIndex; + uint8_t displayWatermark; + uint8_t gen2PCIE; + uint8_t reserved1; + uint8_t reserved2; + uint8_t strobeMode; + uint8_t mcFlags; + uint32_t aT; + uint32_t bSP; + NISLANDS_SMC_SCLK_VALUE sclk; + NISLANDS_SMC_MCLK_VALUE mclk; + NISLANDS_SMC_VOLTAGE_VALUE vddc; + NISLANDS_SMC_VOLTAGE_VALUE mvdd; + NISLANDS_SMC_VOLTAGE_VALUE vddci; + NISLANDS_SMC_VOLTAGE_VALUE std_vddc; + uint32_t powergate_en; + uint8_t hUp; + uint8_t hDown; + uint8_t stateFlags; + uint8_t arbRefreshState; + uint32_t SQPowerThrottle; + uint32_t SQPowerThrottle_2; + uint32_t reserved[2]; + PP_NIslands_Dpm2PerfLevel dpm2; +}; + +typedef struct NISLANDS_SMC_HW_PERFORMANCE_LEVEL NISLANDS_SMC_HW_PERFORMANCE_LEVEL; + +struct NISLANDS_SMC_SWSTATE +{ + uint8_t flags; + uint8_t levelCount; + uint8_t padding2; + uint8_t padding3; + NISLANDS_SMC_HW_PERFORMANCE_LEVEL levels[1]; +}; + +typedef struct NISLANDS_SMC_SWSTATE NISLANDS_SMC_SWSTATE; + +struct NISLANDS_SMC_VOLTAGEMASKTABLE +{ + uint8_t highMask[NISLANDS_SMC_VOLTAGEMASK_MAX]; + uint32_t lowMask[NISLANDS_SMC_VOLTAGEMASK_MAX]; +}; + +typedef struct NISLANDS_SMC_VOLTAGEMASKTABLE NISLANDS_SMC_VOLTAGEMASKTABLE; + +#define NISLANDS_MAX_NO_VREG_STEPS 32 + +struct NISLANDS_SMC_STATETABLE +{ + uint8_t thermalProtectType; + uint8_t systemFlags; + uint8_t maxVDDCIndexInPPTable; + uint8_t extraFlags; + uint8_t highSMIO[NISLANDS_MAX_NO_VREG_STEPS]; + uint32_t lowSMIO[NISLANDS_MAX_NO_VREG_STEPS]; + NISLANDS_SMC_VOLTAGEMASKTABLE voltageMaskTable; + PP_NIslands_DPM2Parameters dpm2Params; + NISLANDS_SMC_SWSTATE initialState; + NISLANDS_SMC_SWSTATE ACPIState; + NISLANDS_SMC_SWSTATE ULVState; + NISLANDS_SMC_SWSTATE driverState; + NISLANDS_SMC_HW_PERFORMANCE_LEVEL dpmLevels[NISLANDS_MAX_SMC_PERFORMANCE_LEVELS_PER_SWSTATE - 1]; +}; + +typedef struct NISLANDS_SMC_STATETABLE NISLANDS_SMC_STATETABLE; + +struct ni_power_info { + /* must be first! */ + struct evergreen_power_info eg; + struct ni_clock_registers clock_registers; + struct ni_mc_reg_table mc_reg_table; + u32 mclk_rtt_mode_threshold; + /* flags */ + bool use_power_boost_limit; + bool support_cac_long_term_average; + bool cac_enabled; + bool cac_configuration_required; + bool driver_calculate_cac_leakage; + bool pc_enabled; + bool enable_power_containment; + bool enable_cac; + bool enable_sq_ramping; + /* smc offsets */ + u16 arb_table_start; + u16 fan_table_start; + u16 cac_table_start; + u16 spll_table_start; + /* CAC stuff */ + struct ni_cac_data cac_data; + u32 dc_cac_table[NISLANDS_DCCAC_MAX_LEVELS]; + const struct ni_cac_weights *cac_weights; + u8 lta_window_size; + u8 lts_truncate; + struct si_ps current_ps; + struct si_ps requested_ps; + /* scratch structs */ + SMC_NIslands_MCRegisters smc_mc_reg_table; + NISLANDS_SMC_STATETABLE smc_statetable; +}; + +struct si_cac_config_reg +{ + u32 offset; + u32 mask; + u32 shift; + u32 value; + enum si_cac_config_reg_type type; +}; + +struct si_powertune_data +{ + u32 cac_window; + u32 l2_lta_window_size_default; + u8 lts_truncate_default; + u8 shift_n_default; + u8 operating_temp; + struct ni_leakage_coeffients leakage_coefficients; + u32 fixed_kt; + u32 lkge_lut_v0_percent; + u8 dc_cac[NISLANDS_DCCAC_MAX_LEVELS]; + bool enable_powertune_by_default; +}; + +struct si_dyn_powertune_data +{ + u32 cac_leakage; + s32 leakage_minimum_temperature; + u32 wintime; + u32 l2_lta_window_size; + u8 lts_truncate; + u8 shift_n; + u8 dc_pwr_value; + bool disable_uvd_powertune; +}; + +struct si_dte_data +{ + u32 tau[SMC_SISLANDS_DTE_MAX_FILTER_STAGES]; + u32 r[SMC_SISLANDS_DTE_MAX_FILTER_STAGES]; + u32 k; + u32 t0; + u32 max_t; + u8 window_size; + u8 temp_select; + u8 dte_mode; + u8 tdep_count; + u8 t_limits[SMC_SISLANDS_DTE_MAX_TEMPERATURE_DEPENDENT_ARRAY_SIZE]; + u32 tdep_tau[SMC_SISLANDS_DTE_MAX_TEMPERATURE_DEPENDENT_ARRAY_SIZE]; + u32 tdep_r[SMC_SISLANDS_DTE_MAX_TEMPERATURE_DEPENDENT_ARRAY_SIZE]; + u32 t_threshold; + bool enable_dte_by_default; +}; + +struct si_clock_registers { + u32 cg_spll_func_cntl; + u32 cg_spll_func_cntl_2; + u32 cg_spll_func_cntl_3; + u32 cg_spll_func_cntl_4; + u32 cg_spll_spread_spectrum; + u32 cg_spll_spread_spectrum_2; + u32 dll_cntl; + u32 mclk_pwrmgt_cntl; + u32 mpll_ad_func_cntl; + u32 mpll_dq_func_cntl; + u32 mpll_func_cntl; + u32 mpll_func_cntl_1; + u32 mpll_func_cntl_2; + u32 mpll_ss1; + u32 mpll_ss2; +}; + +struct si_mc_reg_entry { + u32 mclk_max; + u32 mc_data[SMC_SISLANDS_MC_REGISTER_ARRAY_SIZE]; +}; + +struct si_mc_reg_table { + u8 last; + u8 num_entries; + u16 valid_flag; + struct si_mc_reg_entry mc_reg_table_entry[MAX_AC_TIMING_ENTRIES]; + SMC_NIslands_MCRegisterAddress mc_reg_address[SMC_SISLANDS_MC_REGISTER_ARRAY_SIZE]; +}; + +struct si_leakage_voltage_entry +{ + u16 voltage; + u16 leakage_index; +}; + +struct si_leakage_voltage +{ + u16 count; + struct si_leakage_voltage_entry entries[SISLANDS_MAX_LEAKAGE_COUNT]; +}; + + +struct si_ulv_param { + bool supported; + u32 cg_ulv_control; + u32 cg_ulv_parameter; + u32 volt_change_delay; + struct rv7xx_pl pl; + bool one_pcie_lane_in_ulv; +}; + +struct si_power_info { + /* must be first! */ + struct ni_power_info ni; + struct si_clock_registers clock_registers; + struct si_mc_reg_table mc_reg_table; + struct atom_voltage_table mvdd_voltage_table; + struct atom_voltage_table vddc_phase_shed_table; + struct si_leakage_voltage leakage_voltage; + u16 mvdd_bootup_value; + struct si_ulv_param ulv; + u32 max_cu; + /* pcie gen */ + enum amdgpu_pcie_gen force_pcie_gen; + enum amdgpu_pcie_gen boot_pcie_gen; + enum amdgpu_pcie_gen acpi_pcie_gen; + u32 sys_pcie_mask; + /* flags */ + bool enable_dte; + bool enable_ppm; + bool vddc_phase_shed_control; + bool pspp_notify_required; + bool sclk_deep_sleep_above_low; + bool voltage_control_svi2; + bool vddci_control_svi2; + /* smc offsets */ + u32 sram_end; + u32 state_table_start; + u32 soft_regs_start; + u32 mc_reg_table_start; + u32 arb_table_start; + u32 cac_table_start; + u32 dte_table_start; + u32 spll_table_start; + u32 papm_cfg_table_start; + u32 fan_table_start; + /* CAC stuff */ + const struct si_cac_config_reg *cac_weights; + const struct si_cac_config_reg *lcac_config; + const struct si_cac_config_reg *cac_override; + const struct si_powertune_data *powertune_data; + struct si_dyn_powertune_data dyn_powertune_data; + /* DTE stuff */ + struct si_dte_data dte_data; + /* scratch structs */ + SMC_SIslands_MCRegisters smc_mc_reg_table; + SISLANDS_SMC_STATETABLE smc_statetable; + PP_SIslands_PAPMParameters papm_parm; + /* SVI2 */ + u8 svd_gpio_id; + u8 svc_gpio_id; + /* fan control */ + bool fan_ctrl_is_in_default_mode; + u32 t_min; + u32 fan_ctrl_default_mode; + bool fan_is_controlled_by_smc; +}; + +#endif diff --git a/drivers/gpu/drm/amd/amdgpu/si_ih.c b/drivers/gpu/drm/amd/amdgpu/si_ih.c new file mode 100644 index 000000000000..8fae3d4a2360 --- /dev/null +++ b/drivers/gpu/drm/amd/amdgpu/si_ih.c @@ -0,0 +1,299 @@ +/* + * Copyright 2015 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + */ +#include "drmP.h" +#include "amdgpu.h" +#include "amdgpu_ih.h" +#include "si/sid.h" +#include "si_ih.h" + +static void si_ih_set_interrupt_funcs(struct amdgpu_device *adev); + +static void si_ih_enable_interrupts(struct amdgpu_device *adev) +{ + u32 ih_cntl = RREG32(IH_CNTL); + u32 ih_rb_cntl = RREG32(IH_RB_CNTL); + + ih_cntl |= ENABLE_INTR; + ih_rb_cntl |= IH_RB_ENABLE; + WREG32(IH_CNTL, ih_cntl); + WREG32(IH_RB_CNTL, ih_rb_cntl); + adev->irq.ih.enabled = true; +} + +static void si_ih_disable_interrupts(struct amdgpu_device *adev) +{ + u32 ih_rb_cntl = RREG32(IH_RB_CNTL); + u32 ih_cntl = RREG32(IH_CNTL); + + ih_rb_cntl &= ~IH_RB_ENABLE; + ih_cntl &= ~ENABLE_INTR; + WREG32(IH_RB_CNTL, ih_rb_cntl); + WREG32(IH_CNTL, ih_cntl); + WREG32(IH_RB_RPTR, 0); + WREG32(IH_RB_WPTR, 0); + adev->irq.ih.enabled = false; + adev->irq.ih.rptr = 0; +} + +static int si_ih_irq_init(struct amdgpu_device *adev) +{ + int rb_bufsz; + u32 interrupt_cntl, ih_cntl, ih_rb_cntl; + u64 wptr_off; + + si_ih_disable_interrupts(adev); + WREG32(INTERRUPT_CNTL2, adev->irq.ih.gpu_addr >> 8); + interrupt_cntl = RREG32(INTERRUPT_CNTL); + interrupt_cntl &= ~IH_DUMMY_RD_OVERRIDE; + interrupt_cntl &= ~IH_REQ_NONSNOOP_EN; + WREG32(INTERRUPT_CNTL, interrupt_cntl); + + WREG32(IH_RB_BASE, adev->irq.ih.gpu_addr >> 8); + rb_bufsz = order_base_2(adev->irq.ih.ring_size / 4); + + ih_rb_cntl = IH_WPTR_OVERFLOW_ENABLE | + IH_WPTR_OVERFLOW_CLEAR | + (rb_bufsz << 1) | + IH_WPTR_WRITEBACK_ENABLE; + + wptr_off = adev->wb.gpu_addr + (adev->irq.ih.wptr_offs * 4); + WREG32(IH_RB_WPTR_ADDR_LO, lower_32_bits(wptr_off)); + WREG32(IH_RB_WPTR_ADDR_HI, upper_32_bits(wptr_off) & 0xFF); + WREG32(IH_RB_CNTL, ih_rb_cntl); + WREG32(IH_RB_RPTR, 0); + WREG32(IH_RB_WPTR, 0); + + ih_cntl = MC_WRREQ_CREDIT(0x10) | MC_WR_CLEAN_CNT(0x10) | MC_VMID(0); + if (adev->irq.msi_enabled) + ih_cntl |= RPTR_REARM; + WREG32(IH_CNTL, ih_cntl); + + pci_set_master(adev->pdev); + si_ih_enable_interrupts(adev); + + return 0; +} + +static void si_ih_irq_disable(struct amdgpu_device *adev) +{ + si_ih_disable_interrupts(adev); + mdelay(1); +} + +static u32 si_ih_get_wptr(struct amdgpu_device *adev) +{ + u32 wptr, tmp; + + wptr = le32_to_cpu(adev->wb.wb[adev->irq.ih.wptr_offs]); + + if (wptr & IH_RB_WPTR__RB_OVERFLOW_MASK) { + wptr &= ~IH_RB_WPTR__RB_OVERFLOW_MASK; + dev_warn(adev->dev, "IH ring buffer overflow (0x%08X, 0x%08X, 0x%08X)\n", + wptr, adev->irq.ih.rptr, (wptr + 16) & adev->irq.ih.ptr_mask); + adev->irq.ih.rptr = (wptr + 16) & adev->irq.ih.ptr_mask; + tmp = RREG32(IH_RB_CNTL); + tmp |= IH_RB_CNTL__WPTR_OVERFLOW_CLEAR_MASK; + WREG32(IH_RB_CNTL, tmp); + } + return (wptr & adev->irq.ih.ptr_mask); +} + +static void si_ih_decode_iv(struct amdgpu_device *adev, + struct amdgpu_iv_entry *entry) +{ + u32 ring_index = adev->irq.ih.rptr >> 2; + uint32_t dw[4]; + + dw[0] = le32_to_cpu(adev->irq.ih.ring[ring_index + 0]); + dw[1] = le32_to_cpu(adev->irq.ih.ring[ring_index + 1]); + dw[2] = le32_to_cpu(adev->irq.ih.ring[ring_index + 2]); + dw[3] = le32_to_cpu(adev->irq.ih.ring[ring_index + 3]); + + entry->src_id = dw[0] & 0xff; + entry->src_data = dw[1] & 0xfffffff; + entry->ring_id = dw[2] & 0xff; + entry->vm_id = (dw[2] >> 8) & 0xff; + + adev->irq.ih.rptr += 16; +} + +static void si_ih_set_rptr(struct amdgpu_device *adev) +{ + WREG32(IH_RB_RPTR, adev->irq.ih.rptr); +} + +static int si_ih_early_init(void *handle) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + si_ih_set_interrupt_funcs(adev); + + return 0; +} + +static int si_ih_sw_init(void *handle) +{ + int r; + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + r = amdgpu_ih_ring_init(adev, 64 * 1024, false); + if (r) + return r; + + return amdgpu_irq_init(adev); +} + +static int si_ih_sw_fini(void *handle) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + amdgpu_irq_fini(adev); + amdgpu_ih_ring_fini(adev); + + return 0; +} + +static int si_ih_hw_init(void *handle) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + return si_ih_irq_init(adev); +} + +static int si_ih_hw_fini(void *handle) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + si_ih_irq_disable(adev); + + return 0; +} + +static int si_ih_suspend(void *handle) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + return si_ih_hw_fini(adev); +} + +static int si_ih_resume(void *handle) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + return si_ih_hw_init(adev); +} + +static bool si_ih_is_idle(void *handle) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + u32 tmp = RREG32(SRBM_STATUS); + + if (tmp & SRBM_STATUS__IH_BUSY_MASK) + return false; + + return true; +} + +static int si_ih_wait_for_idle(void *handle) +{ + unsigned i; + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + for (i = 0; i < adev->usec_timeout; i++) { + if (si_ih_is_idle(handle)) + return 0; + udelay(1); + } + return -ETIMEDOUT; +} + +static int si_ih_soft_reset(void *handle) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + u32 srbm_soft_reset = 0; + u32 tmp = RREG32(SRBM_STATUS); + + if (tmp & SRBM_STATUS__IH_BUSY_MASK) + srbm_soft_reset |= SRBM_SOFT_RESET__SOFT_RESET_IH_MASK; + + if (srbm_soft_reset) { + tmp = RREG32(SRBM_SOFT_RESET); + tmp |= srbm_soft_reset; + dev_info(adev->dev, "SRBM_SOFT_RESET=0x%08X\n", tmp); + WREG32(SRBM_SOFT_RESET, tmp); + tmp = RREG32(SRBM_SOFT_RESET); + + udelay(50); + + tmp &= ~srbm_soft_reset; + WREG32(SRBM_SOFT_RESET, tmp); + tmp = RREG32(SRBM_SOFT_RESET); + + udelay(50); + } + + return 0; +} + +static int si_ih_set_clockgating_state(void *handle, + enum amd_clockgating_state state) +{ + return 0; +} + +static int si_ih_set_powergating_state(void *handle, + enum amd_powergating_state state) +{ + return 0; +} + +const struct amd_ip_funcs si_ih_ip_funcs = { + .name = "si_ih", + .early_init = si_ih_early_init, + .late_init = NULL, + .sw_init = si_ih_sw_init, + .sw_fini = si_ih_sw_fini, + .hw_init = si_ih_hw_init, + .hw_fini = si_ih_hw_fini, + .suspend = si_ih_suspend, + .resume = si_ih_resume, + .is_idle = si_ih_is_idle, + .wait_for_idle = si_ih_wait_for_idle, + .soft_reset = si_ih_soft_reset, + .set_clockgating_state = si_ih_set_clockgating_state, + .set_powergating_state = si_ih_set_powergating_state, +}; + +static const struct amdgpu_ih_funcs si_ih_funcs = { + .get_wptr = si_ih_get_wptr, + .decode_iv = si_ih_decode_iv, + .set_rptr = si_ih_set_rptr +}; + +static void si_ih_set_interrupt_funcs(struct amdgpu_device *adev) +{ + if (adev->irq.ih_funcs == NULL) + adev->irq.ih_funcs = &si_ih_funcs; +} + diff --git a/drivers/gpu/drm/amd/amdgpu/si_ih.h b/drivers/gpu/drm/amd/amdgpu/si_ih.h new file mode 100644 index 000000000000..f3e3a954369c --- /dev/null +++ b/drivers/gpu/drm/amd/amdgpu/si_ih.h @@ -0,0 +1,29 @@ +/* + * Copyright 2015 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + */ + +#ifndef __SI_IH_H__ +#define __SI_IH_H__ + +extern const struct amd_ip_funcs si_ih_ip_funcs; + +#endif diff --git a/drivers/gpu/drm/amd/amdgpu/si_smc.c b/drivers/gpu/drm/amd/amdgpu/si_smc.c new file mode 100644 index 000000000000..668ba99d6c05 --- /dev/null +++ b/drivers/gpu/drm/amd/amdgpu/si_smc.c @@ -0,0 +1,273 @@ +/* + * Copyright 2011 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + * Authors: Alex Deucher + */ + +#include <linux/firmware.h> +#include "drmP.h" +#include "amdgpu.h" +#include "si/sid.h" +#include "ppsmc.h" +#include "amdgpu_ucode.h" +#include "sislands_smc.h" + +static int si_set_smc_sram_address(struct amdgpu_device *adev, + u32 smc_address, u32 limit) +{ + if (smc_address & 3) + return -EINVAL; + if ((smc_address + 3) > limit) + return -EINVAL; + + WREG32(SMC_IND_INDEX_0, smc_address); + WREG32_P(SMC_IND_ACCESS_CNTL, 0, ~AUTO_INCREMENT_IND_0); + + return 0; +} + +int amdgpu_si_copy_bytes_to_smc(struct amdgpu_device *adev, + u32 smc_start_address, + const u8 *src, u32 byte_count, u32 limit) +{ + unsigned long flags; + int ret = 0; + u32 data, original_data, addr, extra_shift; + + if (smc_start_address & 3) + return -EINVAL; + if ((smc_start_address + byte_count) > limit) + return -EINVAL; + + addr = smc_start_address; + + spin_lock_irqsave(&adev->smc_idx_lock, flags); + while (byte_count >= 4) { + /* SMC address space is BE */ + data = (src[0] << 24) | (src[1] << 16) | (src[2] << 8) | src[3]; + + ret = si_set_smc_sram_address(adev, addr, limit); + if (ret) + goto done; + + WREG32(SMC_IND_DATA_0, data); + + src += 4; + byte_count -= 4; + addr += 4; + } + + /* RMW for the final bytes */ + if (byte_count > 0) { + data = 0; + + ret = si_set_smc_sram_address(adev, addr, limit); + if (ret) + goto done; + + original_data = RREG32(SMC_IND_DATA_0); + extra_shift = 8 * (4 - byte_count); + + while (byte_count > 0) { + /* SMC address space is BE */ + data = (data << 8) + *src++; + byte_count--; + } + + data <<= extra_shift; + data |= (original_data & ~((~0UL) << extra_shift)); + + ret = si_set_smc_sram_address(adev, addr, limit); + if (ret) + goto done; + + WREG32(SMC_IND_DATA_0, data); + } + +done: + spin_unlock_irqrestore(&adev->smc_idx_lock, flags); + + return ret; +} + +void amdgpu_si_start_smc(struct amdgpu_device *adev) +{ + u32 tmp = RREG32_SMC(SMC_SYSCON_RESET_CNTL); + + tmp &= ~RST_REG; + + WREG32_SMC(SMC_SYSCON_RESET_CNTL, tmp); +} + +void amdgpu_si_reset_smc(struct amdgpu_device *adev) +{ + u32 tmp; + + RREG32(CB_CGTT_SCLK_CTRL); + RREG32(CB_CGTT_SCLK_CTRL); + RREG32(CB_CGTT_SCLK_CTRL); + RREG32(CB_CGTT_SCLK_CTRL); + + tmp = RREG32_SMC(SMC_SYSCON_RESET_CNTL) | + RST_REG; + WREG32_SMC(SMC_SYSCON_RESET_CNTL, tmp); +} + +int amdgpu_si_program_jump_on_start(struct amdgpu_device *adev) +{ + static const u8 data[] = { 0x0E, 0x00, 0x40, 0x40 }; + + return amdgpu_si_copy_bytes_to_smc(adev, 0x0, data, 4, sizeof(data)+1); +} + +void amdgpu_si_smc_clock(struct amdgpu_device *adev, bool enable) +{ + u32 tmp = RREG32_SMC(SMC_SYSCON_CLOCK_CNTL_0); + + if (enable) + tmp &= ~CK_DISABLE; + else + tmp |= CK_DISABLE; + + WREG32_SMC(SMC_SYSCON_CLOCK_CNTL_0, tmp); +} + +bool amdgpu_si_is_smc_running(struct amdgpu_device *adev) +{ + u32 rst = RREG32_SMC(SMC_SYSCON_RESET_CNTL); + u32 clk = RREG32_SMC(SMC_SYSCON_CLOCK_CNTL_0); + + if (!(rst & RST_REG) && !(clk & CK_DISABLE)) + return true; + + return false; +} + +PPSMC_Result amdgpu_si_send_msg_to_smc(struct amdgpu_device *adev, + PPSMC_Msg msg) +{ + u32 tmp; + int i; + + if (!amdgpu_si_is_smc_running(adev)) + return PPSMC_Result_Failed; + + WREG32(SMC_MESSAGE_0, msg); + + for (i = 0; i < adev->usec_timeout; i++) { + tmp = RREG32(SMC_RESP_0); + if (tmp != 0) + break; + udelay(1); + } + + return (PPSMC_Result)RREG32(SMC_RESP_0); +} + +PPSMC_Result amdgpu_si_wait_for_smc_inactive(struct amdgpu_device *adev) +{ + u32 tmp; + int i; + + if (!amdgpu_si_is_smc_running(adev)) + return PPSMC_Result_OK; + + for (i = 0; i < adev->usec_timeout; i++) { + tmp = RREG32_SMC(SMC_SYSCON_CLOCK_CNTL_0); + if ((tmp & CKEN) == 0) + break; + udelay(1); + } + + return PPSMC_Result_OK; +} + +int amdgpu_si_load_smc_ucode(struct amdgpu_device *adev, u32 limit) +{ + const struct smc_firmware_header_v1_0 *hdr; + unsigned long flags; + u32 ucode_start_address; + u32 ucode_size; + const u8 *src; + u32 data; + + if (!adev->pm.fw) + return -EINVAL; + + hdr = (const struct smc_firmware_header_v1_0 *)adev->pm.fw->data; + + amdgpu_ucode_print_smc_hdr(&hdr->header); + + adev->pm.fw_version = le32_to_cpu(hdr->header.ucode_version); + ucode_start_address = le32_to_cpu(hdr->ucode_start_addr); + ucode_size = le32_to_cpu(hdr->header.ucode_size_bytes); + src = (const u8 *) + (adev->pm.fw->data + le32_to_cpu(hdr->header.ucode_array_offset_bytes)); + if (ucode_size & 3) + return -EINVAL; + + spin_lock_irqsave(&adev->smc_idx_lock, flags); + WREG32(SMC_IND_INDEX_0, ucode_start_address); + WREG32_P(SMC_IND_ACCESS_CNTL, AUTO_INCREMENT_IND_0, ~AUTO_INCREMENT_IND_0); + while (ucode_size >= 4) { + /* SMC address space is BE */ + data = (src[0] << 24) | (src[1] << 16) | (src[2] << 8) | src[3]; + + WREG32(SMC_IND_DATA_0, data); + + src += 4; + ucode_size -= 4; + } + WREG32_P(SMC_IND_ACCESS_CNTL, 0, ~AUTO_INCREMENT_IND_0); + spin_unlock_irqrestore(&adev->smc_idx_lock, flags); + + return 0; +} + +int amdgpu_si_read_smc_sram_dword(struct amdgpu_device *adev, u32 smc_address, + u32 *value, u32 limit) +{ + unsigned long flags; + int ret; + + spin_lock_irqsave(&adev->smc_idx_lock, flags); + ret = si_set_smc_sram_address(adev, smc_address, limit); + if (ret == 0) + *value = RREG32(SMC_IND_DATA_0); + spin_unlock_irqrestore(&adev->smc_idx_lock, flags); + + return ret; +} + +int amdgpu_si_write_smc_sram_dword(struct amdgpu_device *adev, u32 smc_address, + u32 value, u32 limit) +{ + unsigned long flags; + int ret; + + spin_lock_irqsave(&adev->smc_idx_lock, flags); + ret = si_set_smc_sram_address(adev, smc_address, limit); + if (ret == 0) + WREG32(SMC_IND_DATA_0, value); + spin_unlock_irqrestore(&adev->smc_idx_lock, flags); + + return ret; +} diff --git a/drivers/gpu/drm/amd/amdgpu/sislands_smc.h b/drivers/gpu/drm/amd/amdgpu/sislands_smc.h new file mode 100644 index 000000000000..ee4b846e58fa --- /dev/null +++ b/drivers/gpu/drm/amd/amdgpu/sislands_smc.h @@ -0,0 +1,422 @@ +/* + * Copyright 2013 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + */ +#ifndef PP_SISLANDS_SMC_H +#define PP_SISLANDS_SMC_H + +#include "ppsmc.h" + +#pragma pack(push, 1) + +#define SISLANDS_MAX_SMC_PERFORMANCE_LEVELS_PER_SWSTATE 16 + +struct PP_SIslands_Dpm2PerfLevel +{ + uint8_t MaxPS; + uint8_t TgtAct; + uint8_t MaxPS_StepInc; + uint8_t MaxPS_StepDec; + uint8_t PSSamplingTime; + uint8_t NearTDPDec; + uint8_t AboveSafeInc; + uint8_t BelowSafeInc; + uint8_t PSDeltaLimit; + uint8_t PSDeltaWin; + uint16_t PwrEfficiencyRatio; + uint8_t Reserved[4]; +}; + +typedef struct PP_SIslands_Dpm2PerfLevel PP_SIslands_Dpm2PerfLevel; + +struct PP_SIslands_DPM2Status +{ + uint32_t dpm2Flags; + uint8_t CurrPSkip; + uint8_t CurrPSkipPowerShift; + uint8_t CurrPSkipTDP; + uint8_t CurrPSkipOCP; + uint8_t MaxSPLLIndex; + uint8_t MinSPLLIndex; + uint8_t CurrSPLLIndex; + uint8_t InfSweepMode; + uint8_t InfSweepDir; + uint8_t TDPexceeded; + uint8_t reserved; + uint8_t SwitchDownThreshold; + uint32_t SwitchDownCounter; + uint32_t SysScalingFactor; +}; + +typedef struct PP_SIslands_DPM2Status PP_SIslands_DPM2Status; + +struct PP_SIslands_DPM2Parameters +{ + uint32_t TDPLimit; + uint32_t NearTDPLimit; + uint32_t SafePowerLimit; + uint32_t PowerBoostLimit; + uint32_t MinLimitDelta; +}; +typedef struct PP_SIslands_DPM2Parameters PP_SIslands_DPM2Parameters; + +struct PP_SIslands_PAPMStatus +{ + uint32_t EstimatedDGPU_T; + uint32_t EstimatedDGPU_P; + uint32_t EstimatedAPU_T; + uint32_t EstimatedAPU_P; + uint8_t dGPU_T_Limit_Exceeded; + uint8_t reserved[3]; +}; +typedef struct PP_SIslands_PAPMStatus PP_SIslands_PAPMStatus; + +struct PP_SIslands_PAPMParameters +{ + uint32_t NearTDPLimitTherm; + uint32_t NearTDPLimitPAPM; + uint32_t PlatformPowerLimit; + uint32_t dGPU_T_Limit; + uint32_t dGPU_T_Warning; + uint32_t dGPU_T_Hysteresis; +}; +typedef struct PP_SIslands_PAPMParameters PP_SIslands_PAPMParameters; + +struct SISLANDS_SMC_SCLK_VALUE +{ + uint32_t vCG_SPLL_FUNC_CNTL; + uint32_t vCG_SPLL_FUNC_CNTL_2; + uint32_t vCG_SPLL_FUNC_CNTL_3; + uint32_t vCG_SPLL_FUNC_CNTL_4; + uint32_t vCG_SPLL_SPREAD_SPECTRUM; + uint32_t vCG_SPLL_SPREAD_SPECTRUM_2; + uint32_t sclk_value; +}; + +typedef struct SISLANDS_SMC_SCLK_VALUE SISLANDS_SMC_SCLK_VALUE; + +struct SISLANDS_SMC_MCLK_VALUE +{ + uint32_t vMPLL_FUNC_CNTL; + uint32_t vMPLL_FUNC_CNTL_1; + uint32_t vMPLL_FUNC_CNTL_2; + uint32_t vMPLL_AD_FUNC_CNTL; + uint32_t vMPLL_DQ_FUNC_CNTL; + uint32_t vMCLK_PWRMGT_CNTL; + uint32_t vDLL_CNTL; + uint32_t vMPLL_SS; + uint32_t vMPLL_SS2; + uint32_t mclk_value; +}; + +typedef struct SISLANDS_SMC_MCLK_VALUE SISLANDS_SMC_MCLK_VALUE; + +struct SISLANDS_SMC_VOLTAGE_VALUE +{ + uint16_t value; + uint8_t index; + uint8_t phase_settings; +}; + +typedef struct SISLANDS_SMC_VOLTAGE_VALUE SISLANDS_SMC_VOLTAGE_VALUE; + +struct SISLANDS_SMC_HW_PERFORMANCE_LEVEL +{ + uint8_t ACIndex; + uint8_t displayWatermark; + uint8_t gen2PCIE; + uint8_t UVDWatermark; + uint8_t VCEWatermark; + uint8_t strobeMode; + uint8_t mcFlags; + uint8_t padding; + uint32_t aT; + uint32_t bSP; + SISLANDS_SMC_SCLK_VALUE sclk; + SISLANDS_SMC_MCLK_VALUE mclk; + SISLANDS_SMC_VOLTAGE_VALUE vddc; + SISLANDS_SMC_VOLTAGE_VALUE mvdd; + SISLANDS_SMC_VOLTAGE_VALUE vddci; + SISLANDS_SMC_VOLTAGE_VALUE std_vddc; + uint8_t hysteresisUp; + uint8_t hysteresisDown; + uint8_t stateFlags; + uint8_t arbRefreshState; + uint32_t SQPowerThrottle; + uint32_t SQPowerThrottle_2; + uint32_t MaxPoweredUpCU; + SISLANDS_SMC_VOLTAGE_VALUE high_temp_vddc; + SISLANDS_SMC_VOLTAGE_VALUE low_temp_vddc; + uint32_t reserved[2]; + PP_SIslands_Dpm2PerfLevel dpm2; +}; + +#define SISLANDS_SMC_STROBE_RATIO 0x0F +#define SISLANDS_SMC_STROBE_ENABLE 0x10 + +#define SISLANDS_SMC_MC_EDC_RD_FLAG 0x01 +#define SISLANDS_SMC_MC_EDC_WR_FLAG 0x02 +#define SISLANDS_SMC_MC_RTT_ENABLE 0x04 +#define SISLANDS_SMC_MC_STUTTER_EN 0x08 +#define SISLANDS_SMC_MC_PG_EN 0x10 + +typedef struct SISLANDS_SMC_HW_PERFORMANCE_LEVEL SISLANDS_SMC_HW_PERFORMANCE_LEVEL; + +struct SISLANDS_SMC_SWSTATE +{ + uint8_t flags; + uint8_t levelCount; + uint8_t padding2; + uint8_t padding3; + SISLANDS_SMC_HW_PERFORMANCE_LEVEL levels[1]; +}; + +typedef struct SISLANDS_SMC_SWSTATE SISLANDS_SMC_SWSTATE; + +#define SISLANDS_SMC_VOLTAGEMASK_VDDC 0 +#define SISLANDS_SMC_VOLTAGEMASK_MVDD 1 +#define SISLANDS_SMC_VOLTAGEMASK_VDDCI 2 +#define SISLANDS_SMC_VOLTAGEMASK_MAX 4 + +struct SISLANDS_SMC_VOLTAGEMASKTABLE +{ + uint32_t lowMask[SISLANDS_SMC_VOLTAGEMASK_MAX]; +}; + +typedef struct SISLANDS_SMC_VOLTAGEMASKTABLE SISLANDS_SMC_VOLTAGEMASKTABLE; + +#define SISLANDS_MAX_NO_VREG_STEPS 32 + +struct SISLANDS_SMC_STATETABLE +{ + uint8_t thermalProtectType; + uint8_t systemFlags; + uint8_t maxVDDCIndexInPPTable; + uint8_t extraFlags; + uint32_t lowSMIO[SISLANDS_MAX_NO_VREG_STEPS]; + SISLANDS_SMC_VOLTAGEMASKTABLE voltageMaskTable; + SISLANDS_SMC_VOLTAGEMASKTABLE phaseMaskTable; + PP_SIslands_DPM2Parameters dpm2Params; + SISLANDS_SMC_SWSTATE initialState; + SISLANDS_SMC_SWSTATE ACPIState; + SISLANDS_SMC_SWSTATE ULVState; + SISLANDS_SMC_SWSTATE driverState; + SISLANDS_SMC_HW_PERFORMANCE_LEVEL dpmLevels[SISLANDS_MAX_SMC_PERFORMANCE_LEVELS_PER_SWSTATE - 1]; +}; + +typedef struct SISLANDS_SMC_STATETABLE SISLANDS_SMC_STATETABLE; + +#define SI_SMC_SOFT_REGISTER_mclk_chg_timeout 0x0 +#define SI_SMC_SOFT_REGISTER_delay_vreg 0xC +#define SI_SMC_SOFT_REGISTER_delay_acpi 0x28 +#define SI_SMC_SOFT_REGISTER_seq_index 0x5C +#define SI_SMC_SOFT_REGISTER_mvdd_chg_time 0x60 +#define SI_SMC_SOFT_REGISTER_mclk_switch_lim 0x70 +#define SI_SMC_SOFT_REGISTER_watermark_threshold 0x78 +#define SI_SMC_SOFT_REGISTER_phase_shedding_delay 0x88 +#define SI_SMC_SOFT_REGISTER_ulv_volt_change_delay 0x8C +#define SI_SMC_SOFT_REGISTER_mc_block_delay 0x98 +#define SI_SMC_SOFT_REGISTER_ticks_per_us 0xA8 +#define SI_SMC_SOFT_REGISTER_crtc_index 0xC4 +#define SI_SMC_SOFT_REGISTER_mclk_change_block_cp_min 0xC8 +#define SI_SMC_SOFT_REGISTER_mclk_change_block_cp_max 0xCC +#define SI_SMC_SOFT_REGISTER_non_ulv_pcie_link_width 0xF4 +#define SI_SMC_SOFT_REGISTER_tdr_is_about_to_happen 0xFC +#define SI_SMC_SOFT_REGISTER_vr_hot_gpio 0x100 +#define SI_SMC_SOFT_REGISTER_svi_rework_plat_type 0x118 +#define SI_SMC_SOFT_REGISTER_svi_rework_gpio_id_svd 0x11c +#define SI_SMC_SOFT_REGISTER_svi_rework_gpio_id_svc 0x120 + +struct PP_SIslands_FanTable +{ + uint8_t fdo_mode; + uint8_t padding; + int16_t temp_min; + int16_t temp_med; + int16_t temp_max; + int16_t slope1; + int16_t slope2; + int16_t fdo_min; + int16_t hys_up; + int16_t hys_down; + int16_t hys_slope; + int16_t temp_resp_lim; + int16_t temp_curr; + int16_t slope_curr; + int16_t pwm_curr; + uint32_t refresh_period; + int16_t fdo_max; + uint8_t temp_src; + int8_t padding2; +}; + +typedef struct PP_SIslands_FanTable PP_SIslands_FanTable; + +#define SMC_SISLANDS_LKGE_LUT_NUM_OF_TEMP_ENTRIES 16 +#define SMC_SISLANDS_LKGE_LUT_NUM_OF_VOLT_ENTRIES 32 + +#define SMC_SISLANDS_SCALE_I 7 +#define SMC_SISLANDS_SCALE_R 12 + +struct PP_SIslands_CacConfig +{ + uint16_t cac_lkge_lut[SMC_SISLANDS_LKGE_LUT_NUM_OF_TEMP_ENTRIES][SMC_SISLANDS_LKGE_LUT_NUM_OF_VOLT_ENTRIES]; + uint32_t lkge_lut_V0; + uint32_t lkge_lut_Vstep; + uint32_t WinTime; + uint32_t R_LL; + uint32_t calculation_repeats; + uint32_t l2numWin_TDP; + uint32_t dc_cac; + uint8_t lts_truncate_n; + uint8_t SHIFT_N; + uint8_t log2_PG_LKG_SCALE; + uint8_t cac_temp; + uint32_t lkge_lut_T0; + uint32_t lkge_lut_Tstep; +}; + +typedef struct PP_SIslands_CacConfig PP_SIslands_CacConfig; + +#define SMC_SISLANDS_MC_REGISTER_ARRAY_SIZE 16 +#define SMC_SISLANDS_MC_REGISTER_ARRAY_SET_COUNT 20 + +struct SMC_SIslands_MCRegisterAddress +{ + uint16_t s0; + uint16_t s1; +}; + +typedef struct SMC_SIslands_MCRegisterAddress SMC_SIslands_MCRegisterAddress; + +struct SMC_SIslands_MCRegisterSet +{ + uint32_t value[SMC_SISLANDS_MC_REGISTER_ARRAY_SIZE]; +}; + +typedef struct SMC_SIslands_MCRegisterSet SMC_SIslands_MCRegisterSet; + +struct SMC_SIslands_MCRegisters +{ + uint8_t last; + uint8_t reserved[3]; + SMC_SIslands_MCRegisterAddress address[SMC_SISLANDS_MC_REGISTER_ARRAY_SIZE]; + SMC_SIslands_MCRegisterSet data[SMC_SISLANDS_MC_REGISTER_ARRAY_SET_COUNT]; +}; + +typedef struct SMC_SIslands_MCRegisters SMC_SIslands_MCRegisters; + +struct SMC_SIslands_MCArbDramTimingRegisterSet +{ + uint32_t mc_arb_dram_timing; + uint32_t mc_arb_dram_timing2; + uint8_t mc_arb_rfsh_rate; + uint8_t mc_arb_burst_time; + uint8_t padding[2]; +}; + +typedef struct SMC_SIslands_MCArbDramTimingRegisterSet SMC_SIslands_MCArbDramTimingRegisterSet; + +struct SMC_SIslands_MCArbDramTimingRegisters +{ + uint8_t arb_current; + uint8_t reserved[3]; + SMC_SIslands_MCArbDramTimingRegisterSet data[16]; +}; + +typedef struct SMC_SIslands_MCArbDramTimingRegisters SMC_SIslands_MCArbDramTimingRegisters; + +struct SMC_SISLANDS_SPLL_DIV_TABLE +{ + uint32_t freq[256]; + uint32_t ss[256]; +}; + +#define SMC_SISLANDS_SPLL_DIV_TABLE_FBDIV_MASK 0x01ffffff +#define SMC_SISLANDS_SPLL_DIV_TABLE_FBDIV_SHIFT 0 +#define SMC_SISLANDS_SPLL_DIV_TABLE_PDIV_MASK 0xfe000000 +#define SMC_SISLANDS_SPLL_DIV_TABLE_PDIV_SHIFT 25 +#define SMC_SISLANDS_SPLL_DIV_TABLE_CLKV_MASK 0x000fffff +#define SMC_SISLANDS_SPLL_DIV_TABLE_CLKV_SHIFT 0 +#define SMC_SISLANDS_SPLL_DIV_TABLE_CLKS_MASK 0xfff00000 +#define SMC_SISLANDS_SPLL_DIV_TABLE_CLKS_SHIFT 20 + +typedef struct SMC_SISLANDS_SPLL_DIV_TABLE SMC_SISLANDS_SPLL_DIV_TABLE; + +#define SMC_SISLANDS_DTE_MAX_FILTER_STAGES 5 + +#define SMC_SISLANDS_DTE_MAX_TEMPERATURE_DEPENDENT_ARRAY_SIZE 16 + +struct Smc_SIslands_DTE_Configuration +{ + uint32_t tau[SMC_SISLANDS_DTE_MAX_FILTER_STAGES]; + uint32_t R[SMC_SISLANDS_DTE_MAX_FILTER_STAGES]; + uint32_t K; + uint32_t T0; + uint32_t MaxT; + uint8_t WindowSize; + uint8_t Tdep_count; + uint8_t temp_select; + uint8_t DTE_mode; + uint8_t T_limits[SMC_SISLANDS_DTE_MAX_TEMPERATURE_DEPENDENT_ARRAY_SIZE]; + uint32_t Tdep_tau[SMC_SISLANDS_DTE_MAX_TEMPERATURE_DEPENDENT_ARRAY_SIZE]; + uint32_t Tdep_R[SMC_SISLANDS_DTE_MAX_TEMPERATURE_DEPENDENT_ARRAY_SIZE]; + uint32_t Tthreshold; +}; + +typedef struct Smc_SIslands_DTE_Configuration Smc_SIslands_DTE_Configuration; + +#define SMC_SISLANDS_DTE_STATUS_FLAG_DTE_ON 1 + +#define SISLANDS_SMC_FIRMWARE_HEADER_LOCATION 0x10000 + +#define SISLANDS_SMC_FIRMWARE_HEADER_version 0x0 +#define SISLANDS_SMC_FIRMWARE_HEADER_flags 0x4 +#define SISLANDS_SMC_FIRMWARE_HEADER_softRegisters 0xC +#define SISLANDS_SMC_FIRMWARE_HEADER_stateTable 0x10 +#define SISLANDS_SMC_FIRMWARE_HEADER_fanTable 0x14 +#define SISLANDS_SMC_FIRMWARE_HEADER_CacConfigTable 0x18 +#define SISLANDS_SMC_FIRMWARE_HEADER_mcRegisterTable 0x24 +#define SISLANDS_SMC_FIRMWARE_HEADER_mcArbDramAutoRefreshTable 0x30 +#define SISLANDS_SMC_FIRMWARE_HEADER_spllTable 0x38 +#define SISLANDS_SMC_FIRMWARE_HEADER_DteConfiguration 0x40 +#define SISLANDS_SMC_FIRMWARE_HEADER_PAPMParameters 0x48 + +#pragma pack(pop) + +int amdgpu_si_copy_bytes_to_smc(struct amdgpu_device *adev, + u32 smc_start_address, + const u8 *src, u32 byte_count, u32 limit); +void amdgpu_si_start_smc(struct amdgpu_device *adev); +void amdgpu_si_reset_smc(struct amdgpu_device *adev); +int amdgpu_si_program_jump_on_start(struct amdgpu_device *adev); +void amdgpu_si_smc_clock(struct amdgpu_device *adev, bool enable); +bool amdgpu_si_is_smc_running(struct amdgpu_device *adev); +PPSMC_Result amdgpu_si_send_msg_to_smc(struct amdgpu_device *adev, PPSMC_Msg msg); +PPSMC_Result amdgpu_si_wait_for_smc_inactive(struct amdgpu_device *adev); +int amdgpu_si_load_smc_ucode(struct amdgpu_device *adev, u32 limit); +int amdgpu_si_read_smc_sram_dword(struct amdgpu_device *adev, u32 smc_address, + u32 *value, u32 limit); +int amdgpu_si_write_smc_sram_dword(struct amdgpu_device *adev, u32 smc_address, + u32 value, u32 limit); + +#endif + diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c b/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c index 10c0407dcb6e..f6c941550b8f 100644 --- a/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c +++ b/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c @@ -526,6 +526,20 @@ static void uvd_v4_2_ring_emit_ib(struct amdgpu_ring *ring, amdgpu_ring_write(ring, ib->length_dw); } +static unsigned uvd_v4_2_ring_get_emit_ib_size(struct amdgpu_ring *ring) +{ + return + 4; /* uvd_v4_2_ring_emit_ib */ +} + +static unsigned uvd_v4_2_ring_get_dma_frame_size(struct amdgpu_ring *ring) +{ + return + 2 + /* uvd_v4_2_ring_emit_hdp_flush */ + 2 + /* uvd_v4_2_ring_emit_hdp_invalidate */ + 14; /* uvd_v4_2_ring_emit_fence x1 no user fence */ +} + /** * uvd_v4_2_mc_resume - memory controller programming * @@ -756,6 +770,8 @@ static const struct amdgpu_ring_funcs uvd_v4_2_ring_funcs = { .pad_ib = amdgpu_ring_generic_pad_ib, .begin_use = amdgpu_uvd_ring_begin_use, .end_use = amdgpu_uvd_ring_end_use, + .get_emit_ib_size = uvd_v4_2_ring_get_emit_ib_size, + .get_dma_frame_size = uvd_v4_2_ring_get_dma_frame_size, }; static void uvd_v4_2_set_ring_funcs(struct amdgpu_device *adev) diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c index 8513376062c1..400c16fe579e 100644 --- a/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c +++ b/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c @@ -577,6 +577,20 @@ static void uvd_v5_0_ring_emit_ib(struct amdgpu_ring *ring, amdgpu_ring_write(ring, ib->length_dw); } +static unsigned uvd_v5_0_ring_get_emit_ib_size(struct amdgpu_ring *ring) +{ + return + 6; /* uvd_v5_0_ring_emit_ib */ +} + +static unsigned uvd_v5_0_ring_get_dma_frame_size(struct amdgpu_ring *ring) +{ + return + 2 + /* uvd_v5_0_ring_emit_hdp_flush */ + 2 + /* uvd_v5_0_ring_emit_hdp_invalidate */ + 14; /* uvd_v5_0_ring_emit_fence x1 no user fence */ +} + static bool uvd_v5_0_is_idle(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; @@ -807,6 +821,8 @@ static const struct amdgpu_ring_funcs uvd_v5_0_ring_funcs = { .pad_ib = amdgpu_ring_generic_pad_ib, .begin_use = amdgpu_uvd_ring_begin_use, .end_use = amdgpu_uvd_ring_end_use, + .get_emit_ib_size = uvd_v5_0_ring_get_emit_ib_size, + .get_dma_frame_size = uvd_v5_0_ring_get_dma_frame_size, }; static void uvd_v5_0_set_ring_funcs(struct amdgpu_device *adev) diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c index 2abe8a93c99f..e0fd9f21ed95 100644 --- a/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c +++ b/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c @@ -725,6 +725,31 @@ static void uvd_v6_0_ring_emit_pipeline_sync(struct amdgpu_ring *ring) amdgpu_ring_write(ring, 0xE); } +static unsigned uvd_v6_0_ring_get_emit_ib_size(struct amdgpu_ring *ring) +{ + return + 8; /* uvd_v6_0_ring_emit_ib */ +} + +static unsigned uvd_v6_0_ring_get_dma_frame_size(struct amdgpu_ring *ring) +{ + return + 2 + /* uvd_v6_0_ring_emit_hdp_flush */ + 2 + /* uvd_v6_0_ring_emit_hdp_invalidate */ + 10 + /* uvd_v6_0_ring_emit_pipeline_sync */ + 14; /* uvd_v6_0_ring_emit_fence x1 no user fence */ +} + +static unsigned uvd_v6_0_ring_get_dma_frame_size_vm(struct amdgpu_ring *ring) +{ + return + 2 + /* uvd_v6_0_ring_emit_hdp_flush */ + 2 + /* uvd_v6_0_ring_emit_hdp_invalidate */ + 10 + /* uvd_v6_0_ring_emit_pipeline_sync */ + 20 + /* uvd_v6_0_ring_emit_vm_flush */ + 14 + 14; /* uvd_v6_0_ring_emit_fence x2 vm fence */ +} + static bool uvd_v6_0_is_idle(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; @@ -1037,6 +1062,8 @@ static const struct amdgpu_ring_funcs uvd_v6_0_ring_phys_funcs = { .pad_ib = amdgpu_ring_generic_pad_ib, .begin_use = amdgpu_uvd_ring_begin_use, .end_use = amdgpu_uvd_ring_end_use, + .get_emit_ib_size = uvd_v6_0_ring_get_emit_ib_size, + .get_dma_frame_size = uvd_v6_0_ring_get_dma_frame_size, }; static const struct amdgpu_ring_funcs uvd_v6_0_ring_vm_funcs = { @@ -1056,6 +1083,8 @@ static const struct amdgpu_ring_funcs uvd_v6_0_ring_vm_funcs = { .pad_ib = amdgpu_ring_generic_pad_ib, .begin_use = amdgpu_uvd_ring_begin_use, .end_use = amdgpu_uvd_ring_end_use, + .get_emit_ib_size = uvd_v6_0_ring_get_emit_ib_size, + .get_dma_frame_size = uvd_v6_0_ring_get_dma_frame_size_vm, }; static void uvd_v6_0_set_ring_funcs(struct amdgpu_device *adev) diff --git a/drivers/gpu/drm/amd/amdgpu/vce_v2_0.c b/drivers/gpu/drm/amd/amdgpu/vce_v2_0.c index 5fa55b52c00e..76e64ad04a53 100644 --- a/drivers/gpu/drm/amd/amdgpu/vce_v2_0.c +++ b/drivers/gpu/drm/amd/amdgpu/vce_v2_0.c @@ -30,10 +30,10 @@ #include "amdgpu.h" #include "amdgpu_vce.h" #include "cikd.h" - #include "vce/vce_2_0_d.h" #include "vce/vce_2_0_sh_mask.h" - +#include "smu/smu_7_0_1_d.h" +#include "smu/smu_7_0_1_sh_mask.h" #include "oss/oss_2_0_d.h" #include "oss/oss_2_0_sh_mask.h" @@ -193,6 +193,8 @@ static int vce_v2_0_early_init(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; + adev->vce.num_rings = 2; + vce_v2_0_set_ring_funcs(adev); vce_v2_0_set_irq_funcs(adev); @@ -202,7 +204,7 @@ static int vce_v2_0_early_init(void *handle) static int vce_v2_0_sw_init(void *handle) { struct amdgpu_ring *ring; - int r; + int r, i; struct amdgpu_device *adev = (struct amdgpu_device *)handle; /* VCE */ @@ -219,19 +221,14 @@ static int vce_v2_0_sw_init(void *handle) if (r) return r; - ring = &adev->vce.ring[0]; - sprintf(ring->name, "vce0"); - r = amdgpu_ring_init(adev, ring, 512, VCE_CMD_NO_OP, 0xf, - &adev->vce.irq, 0, AMDGPU_RING_TYPE_VCE); - if (r) - return r; - - ring = &adev->vce.ring[1]; - sprintf(ring->name, "vce1"); - r = amdgpu_ring_init(adev, ring, 512, VCE_CMD_NO_OP, 0xf, - &adev->vce.irq, 0, AMDGPU_RING_TYPE_VCE); - if (r) - return r; + for (i = 0; i < adev->vce.num_rings; i++) { + ring = &adev->vce.ring[i]; + sprintf(ring->name, "vce%d", i); + r = amdgpu_ring_init(adev, ring, 512, VCE_CMD_NO_OP, 0xf, + &adev->vce.irq, 0, AMDGPU_RING_TYPE_VCE); + if (r) + return r; + } return r; } @@ -254,29 +251,23 @@ static int vce_v2_0_sw_fini(void *handle) static int vce_v2_0_hw_init(void *handle) { - struct amdgpu_ring *ring; - int r; + int r, i; struct amdgpu_device *adev = (struct amdgpu_device *)handle; r = vce_v2_0_start(adev); + /* this error mean vcpu not in running state, so just skip ring test, not stop driver initialize */ if (r) -/* this error mean vcpu not in running state, so just skip ring test, not stop driver initialize */ return 0; - ring = &adev->vce.ring[0]; - ring->ready = true; - r = amdgpu_ring_test_ring(ring); - if (r) { - ring->ready = false; - return r; - } + for (i = 0; i < adev->vce.num_rings; i++) + adev->vce.ring[i].ready = false; - ring = &adev->vce.ring[1]; - ring->ready = true; - r = amdgpu_ring_test_ring(ring); - if (r) { - ring->ready = false; - return r; + for (i = 0; i < adev->vce.num_rings; i++) { + r = amdgpu_ring_test_ring(&adev->vce.ring[i]); + if (r) + return r; + else + adev->vce.ring[i].ready = true; } DRM_INFO("VCE initialized successfully.\n"); @@ -548,11 +539,28 @@ static int vce_v2_0_process_interrupt(struct amdgpu_device *adev, return 0; } +static void vce_v2_0_set_bypass_mode(struct amdgpu_device *adev, bool enable) +{ + u32 tmp = RREG32_SMC(ixGCK_DFS_BYPASS_CNTL); + + if (enable) + tmp |= GCK_DFS_BYPASS_CNTL__BYPASSECLK_MASK; + else + tmp &= ~GCK_DFS_BYPASS_CNTL__BYPASSECLK_MASK; + + WREG32_SMC(ixGCK_DFS_BYPASS_CNTL, tmp); +} + + static int vce_v2_0_set_clockgating_state(void *handle, enum amd_clockgating_state state) { bool gate = false; struct amdgpu_device *adev = (struct amdgpu_device *)handle; + bool enable = (state == AMD_CG_STATE_GATE) ? true : false; + + + vce_v2_0_set_bypass_mode(adev, enable); if (state == AMD_CG_STATE_GATE) gate = true; @@ -614,12 +622,16 @@ static const struct amdgpu_ring_funcs vce_v2_0_ring_funcs = { .pad_ib = amdgpu_ring_generic_pad_ib, .begin_use = amdgpu_vce_ring_begin_use, .end_use = amdgpu_vce_ring_end_use, + .get_emit_ib_size = amdgpu_vce_ring_get_emit_ib_size, + .get_dma_frame_size = amdgpu_vce_ring_get_dma_frame_size, }; static void vce_v2_0_set_ring_funcs(struct amdgpu_device *adev) { - adev->vce.ring[0].funcs = &vce_v2_0_ring_funcs; - adev->vce.ring[1].funcs = &vce_v2_0_ring_funcs; + int i; + + for (i = 0; i < adev->vce.num_rings; i++) + adev->vce.ring[i].funcs = &vce_v2_0_ring_funcs; } static const struct amdgpu_irq_src_funcs vce_v2_0_irq_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c b/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c index 615b8b16ad04..a6b4e27bee89 100644 --- a/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c +++ b/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c @@ -70,8 +70,10 @@ static uint32_t vce_v3_0_ring_get_rptr(struct amdgpu_ring *ring) if (ring == &adev->vce.ring[0]) return RREG32(mmVCE_RB_RPTR); - else + else if (ring == &adev->vce.ring[1]) return RREG32(mmVCE_RB_RPTR2); + else + return RREG32(mmVCE_RB_RPTR3); } /** @@ -87,8 +89,10 @@ static uint32_t vce_v3_0_ring_get_wptr(struct amdgpu_ring *ring) if (ring == &adev->vce.ring[0]) return RREG32(mmVCE_RB_WPTR); - else + else if (ring == &adev->vce.ring[1]) return RREG32(mmVCE_RB_WPTR2); + else + return RREG32(mmVCE_RB_WPTR3); } /** @@ -104,8 +108,10 @@ static void vce_v3_0_ring_set_wptr(struct amdgpu_ring *ring) if (ring == &adev->vce.ring[0]) WREG32(mmVCE_RB_WPTR, ring->wptr); - else + else if (ring == &adev->vce.ring[1]) WREG32(mmVCE_RB_WPTR2, ring->wptr); + else + WREG32(mmVCE_RB_WPTR3, ring->wptr); } static void vce_v3_0_override_vce_clock_gating(struct amdgpu_device *adev, bool override) @@ -229,6 +235,13 @@ static int vce_v3_0_start(struct amdgpu_device *adev) WREG32(mmVCE_RB_BASE_HI2, upper_32_bits(ring->gpu_addr)); WREG32(mmVCE_RB_SIZE2, ring->ring_size / 4); + ring = &adev->vce.ring[2]; + WREG32(mmVCE_RB_RPTR3, ring->wptr); + WREG32(mmVCE_RB_WPTR3, ring->wptr); + WREG32(mmVCE_RB_BASE_LO3, ring->gpu_addr); + WREG32(mmVCE_RB_BASE_HI3, upper_32_bits(ring->gpu_addr)); + WREG32(mmVCE_RB_SIZE3, ring->ring_size / 4); + mutex_lock(&adev->grbm_idx_mutex); for (idx = 0; idx < 2; ++idx) { if (adev->vce.harvest_config & (1 << idx)) @@ -345,6 +358,8 @@ static int vce_v3_0_early_init(void *handle) (AMDGPU_VCE_HARVEST_VCE0 | AMDGPU_VCE_HARVEST_VCE1)) return -ENOENT; + adev->vce.num_rings = 3; + vce_v3_0_set_ring_funcs(adev); vce_v3_0_set_irq_funcs(adev); @@ -355,7 +370,7 @@ static int vce_v3_0_sw_init(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; struct amdgpu_ring *ring; - int r; + int r, i; /* VCE */ r = amdgpu_irq_add_id(adev, 167, &adev->vce.irq); @@ -371,19 +386,14 @@ static int vce_v3_0_sw_init(void *handle) if (r) return r; - ring = &adev->vce.ring[0]; - sprintf(ring->name, "vce0"); - r = amdgpu_ring_init(adev, ring, 512, VCE_CMD_NO_OP, 0xf, - &adev->vce.irq, 0, AMDGPU_RING_TYPE_VCE); - if (r) - return r; - - ring = &adev->vce.ring[1]; - sprintf(ring->name, "vce1"); - r = amdgpu_ring_init(adev, ring, 512, VCE_CMD_NO_OP, 0xf, - &adev->vce.irq, 0, AMDGPU_RING_TYPE_VCE); - if (r) - return r; + for (i = 0; i < adev->vce.num_rings; i++) { + ring = &adev->vce.ring[i]; + sprintf(ring->name, "vce%d", i); + r = amdgpu_ring_init(adev, ring, 512, VCE_CMD_NO_OP, 0xf, + &adev->vce.irq, 0, AMDGPU_RING_TYPE_VCE); + if (r) + return r; + } return r; } @@ -413,10 +423,10 @@ static int vce_v3_0_hw_init(void *handle) if (r) return r; - adev->vce.ring[0].ready = false; - adev->vce.ring[1].ready = false; + for (i = 0; i < adev->vce.num_rings; i++) + adev->vce.ring[i].ready = false; - for (i = 0; i < 2; i++) { + for (i = 0; i < adev->vce.num_rings; i++) { r = amdgpu_ring_test_ring(&adev->vce.ring[i]); if (r) return r; @@ -674,6 +684,7 @@ static int vce_v3_0_process_interrupt(struct amdgpu_device *adev, switch (entry->src_data) { case 0: case 1: + case 2: amdgpu_fence_process(&adev->vce.ring[entry->src_data]); break; default: @@ -685,7 +696,7 @@ static int vce_v3_0_process_interrupt(struct amdgpu_device *adev, return 0; } -static void vce_v3_set_bypass_mode(struct amdgpu_device *adev, bool enable) +static void vce_v3_0_set_bypass_mode(struct amdgpu_device *adev, bool enable) { u32 tmp = RREG32_SMC(ixGCK_DFS_BYPASS_CNTL); @@ -704,8 +715,9 @@ static int vce_v3_0_set_clockgating_state(void *handle, bool enable = (state == AMD_CG_STATE_GATE) ? true : false; int i; - if (adev->asic_type == CHIP_POLARIS10) - vce_v3_set_bypass_mode(adev, enable); + if ((adev->asic_type == CHIP_POLARIS10) || + (adev->asic_type == CHIP_TONGA)) + vce_v3_0_set_bypass_mode(adev, enable); if (!(adev->cg_flags & AMD_CG_SUPPORT_VCE_MGCG)) return 0; @@ -763,6 +775,60 @@ static int vce_v3_0_set_powergating_state(void *handle, return vce_v3_0_start(adev); } +static void vce_v3_0_ring_emit_ib(struct amdgpu_ring *ring, + struct amdgpu_ib *ib, unsigned int vm_id, bool ctx_switch) +{ + amdgpu_ring_write(ring, VCE_CMD_IB_VM); + amdgpu_ring_write(ring, vm_id); + amdgpu_ring_write(ring, lower_32_bits(ib->gpu_addr)); + amdgpu_ring_write(ring, upper_32_bits(ib->gpu_addr)); + amdgpu_ring_write(ring, ib->length_dw); +} + +static void vce_v3_0_emit_vm_flush(struct amdgpu_ring *ring, + unsigned int vm_id, uint64_t pd_addr) +{ + amdgpu_ring_write(ring, VCE_CMD_UPDATE_PTB); + amdgpu_ring_write(ring, vm_id); + amdgpu_ring_write(ring, pd_addr >> 12); + + amdgpu_ring_write(ring, VCE_CMD_FLUSH_TLB); + amdgpu_ring_write(ring, vm_id); + amdgpu_ring_write(ring, VCE_CMD_END); +} + +static void vce_v3_0_emit_pipeline_sync(struct amdgpu_ring *ring) +{ + uint32_t seq = ring->fence_drv.sync_seq; + uint64_t addr = ring->fence_drv.gpu_addr; + + amdgpu_ring_write(ring, VCE_CMD_WAIT_GE); + amdgpu_ring_write(ring, lower_32_bits(addr)); + amdgpu_ring_write(ring, upper_32_bits(addr)); + amdgpu_ring_write(ring, seq); +} + +static unsigned vce_v3_0_ring_get_emit_ib_size(struct amdgpu_ring *ring) +{ + return + 5; /* vce_v3_0_ring_emit_ib */ +} + +static unsigned vce_v3_0_ring_get_dma_frame_size(struct amdgpu_ring *ring) +{ + return + 4 + /* vce_v3_0_emit_pipeline_sync */ + 6; /* amdgpu_vce_ring_emit_fence x1 no user fence */ +} + +static unsigned vce_v3_0_ring_get_dma_frame_size_vm(struct amdgpu_ring *ring) +{ + return + 6 + /* vce_v3_0_emit_vm_flush */ + 4 + /* vce_v3_0_emit_pipeline_sync */ + 6 + 6; /* amdgpu_vce_ring_emit_fence x2 vm fence */ +} + const struct amd_ip_funcs vce_v3_0_ip_funcs = { .name = "vce_v3_0", .early_init = vce_v3_0_early_init, @@ -783,7 +849,7 @@ const struct amd_ip_funcs vce_v3_0_ip_funcs = { .set_powergating_state = vce_v3_0_set_powergating_state, }; -static const struct amdgpu_ring_funcs vce_v3_0_ring_funcs = { +static const struct amdgpu_ring_funcs vce_v3_0_ring_phys_funcs = { .get_rptr = vce_v3_0_ring_get_rptr, .get_wptr = vce_v3_0_ring_get_wptr, .set_wptr = vce_v3_0_ring_set_wptr, @@ -796,12 +862,42 @@ static const struct amdgpu_ring_funcs vce_v3_0_ring_funcs = { .pad_ib = amdgpu_ring_generic_pad_ib, .begin_use = amdgpu_vce_ring_begin_use, .end_use = amdgpu_vce_ring_end_use, + .get_emit_ib_size = vce_v3_0_ring_get_emit_ib_size, + .get_dma_frame_size = vce_v3_0_ring_get_dma_frame_size, +}; + +static const struct amdgpu_ring_funcs vce_v3_0_ring_vm_funcs = { + .get_rptr = vce_v3_0_ring_get_rptr, + .get_wptr = vce_v3_0_ring_get_wptr, + .set_wptr = vce_v3_0_ring_set_wptr, + .parse_cs = NULL, + .emit_ib = vce_v3_0_ring_emit_ib, + .emit_vm_flush = vce_v3_0_emit_vm_flush, + .emit_pipeline_sync = vce_v3_0_emit_pipeline_sync, + .emit_fence = amdgpu_vce_ring_emit_fence, + .test_ring = amdgpu_vce_ring_test_ring, + .test_ib = amdgpu_vce_ring_test_ib, + .insert_nop = amdgpu_ring_insert_nop, + .pad_ib = amdgpu_ring_generic_pad_ib, + .begin_use = amdgpu_vce_ring_begin_use, + .end_use = amdgpu_vce_ring_end_use, + .get_emit_ib_size = vce_v3_0_ring_get_emit_ib_size, + .get_dma_frame_size = vce_v3_0_ring_get_dma_frame_size_vm, }; static void vce_v3_0_set_ring_funcs(struct amdgpu_device *adev) { - adev->vce.ring[0].funcs = &vce_v3_0_ring_funcs; - adev->vce.ring[1].funcs = &vce_v3_0_ring_funcs; + int i; + + if (adev->asic_type >= CHIP_STONEY) { + for (i = 0; i < adev->vce.num_rings; i++) + adev->vce.ring[i].funcs = &vce_v3_0_ring_vm_funcs; + DRM_INFO("VCE enabled in VM mode\n"); + } else { + for (i = 0; i < adev->vce.num_rings; i++) + adev->vce.ring[i].funcs = &vce_v3_0_ring_phys_funcs; + DRM_INFO("VCE enabled in physical mode\n"); + } } static const struct amdgpu_irq_src_funcs vce_v3_0_irq_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/vi.c b/drivers/gpu/drm/amd/amdgpu/vi.c index f2e8aa1a0dbd..b688e2f77419 100644 --- a/drivers/gpu/drm/amd/amdgpu/vi.c +++ b/drivers/gpu/drm/amd/amdgpu/vi.c @@ -1650,7 +1650,7 @@ static int vi_common_early_init(void *handle) AMD_PG_SUPPORT_GFX_PIPELINE | AMD_PG_SUPPORT_UVD | AMD_PG_SUPPORT_VCE; - adev->external_rev_id = adev->rev_id + 0x1; + adev->external_rev_id = adev->rev_id + 0x61; break; default: /* FIXME: not supported yet */ diff --git a/drivers/gpu/drm/amd/amdgpu/vid.h b/drivers/gpu/drm/amd/amdgpu/vid.h index 062ee1676480..f62b261660d4 100644 --- a/drivers/gpu/drm/amd/amdgpu/vid.h +++ b/drivers/gpu/drm/amd/amdgpu/vid.h @@ -369,4 +369,8 @@ #define VCE_CMD_IB_AUTO 0x00000005 #define VCE_CMD_SEMAPHORE 0x00000006 +#define VCE_CMD_IB_VM 0x00000102 +#define VCE_CMD_WAIT_GE 0x00000106 +#define VCE_CMD_UPDATE_PTB 0x00000107 +#define VCE_CMD_FLUSH_TLB 0x00000108 #endif diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_doorbell.c b/drivers/gpu/drm/amd/amdkfd/kfd_doorbell.c index e621eba63126..ad494b875311 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_doorbell.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_doorbell.c @@ -142,13 +142,15 @@ int kfd_doorbell_mmap(struct kfd_process *process, struct vm_area_struct *vma) vma->vm_page_prot = pgprot_noncached(vma->vm_page_prot); - pr_debug("mapping doorbell page:\n"); - pr_debug(" target user address == 0x%08llX\n", - (unsigned long long) vma->vm_start); - pr_debug(" physical address == 0x%08llX\n", address); - pr_debug(" vm_flags == 0x%04lX\n", vma->vm_flags); - pr_debug(" size == 0x%04lX\n", - doorbell_process_allocation()); + pr_debug("kfd: mapping doorbell page in %s\n" + " target user address == 0x%08llX\n" + " physical address == 0x%08llX\n" + " vm_flags == 0x%04lX\n" + " size == 0x%04lX\n", + __func__, + (unsigned long long) vma->vm_start, address, vma->vm_flags, + doorbell_process_allocation()); + return io_remap_pfn_range(vma, vma->vm_start, diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.c b/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.c index 9beae87aadd5..d135cd002a95 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.c @@ -47,6 +47,9 @@ static bool initialize(struct kernel_queue *kq, struct kfd_dev *dev, pr_debug("amdkfd: In func %s initializing queue type %d size %d\n", __func__, KFD_QUEUE_TYPE_HIQ, queue_size); + memset(&prop, 0, sizeof(prop)); + memset(&nop, 0, sizeof(nop)); + nop.opcode = IT_NOP; nop.type = PM4_TYPE_3; nop.u32all |= PM4_COUNT_ZERO; @@ -121,7 +124,7 @@ static bool initialize(struct kernel_queue *kq, struct kfd_dev *dev, prop.eop_ring_buffer_address = kq->eop_gpu_addr; prop.eop_ring_buffer_size = PAGE_SIZE; - if (init_queue(&kq->queue, prop) != 0) + if (init_queue(&kq->queue, &prop) != 0) goto err_init_queue; kq->queue->device = dev; diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h index 80113c335966..4750cabe4252 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h +++ b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h @@ -619,7 +619,7 @@ int kfd_init_apertures(struct kfd_process *process); /* Queue Context Management */ struct cik_sdma_rlc_registers *get_sdma_mqd(void *mqd); -int init_queue(struct queue **q, struct queue_properties properties); +int init_queue(struct queue **q, const struct queue_properties *properties); void uninit_queue(struct queue *q); void print_queue_properties(struct queue_properties *q); void print_queue(struct queue *q); diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_process.c b/drivers/gpu/drm/amd/amdkfd/kfd_process.c index 4f3849ac8c07..ef7c8de7060e 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_process.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_process.c @@ -404,58 +404,47 @@ void kfd_unbind_process_from_device(struct kfd_dev *dev, unsigned int pasid) { struct kfd_process *p; struct kfd_process_device *pdd; - int idx, i; BUG_ON(dev == NULL); - idx = srcu_read_lock(&kfd_processes_srcu); - /* * Look for the process that matches the pasid. If there is no such * process, we either released it in amdkfd's own notifier, or there * is a bug. Unfortunately, there is no way to tell... */ - hash_for_each_rcu(kfd_processes_table, i, p, kfd_processes) - if (p->pasid == pasid) { - - srcu_read_unlock(&kfd_processes_srcu, idx); - - pr_debug("Unbinding process %d from IOMMU\n", pasid); + p = kfd_lookup_process_by_pasid(pasid); + if (!p) + return; - mutex_lock(&p->mutex); - - if ((dev->dbgmgr) && (dev->dbgmgr->pasid == p->pasid)) - kfd_dbgmgr_destroy(dev->dbgmgr); - - pqm_uninit(&p->pqm); + pr_debug("Unbinding process %d from IOMMU\n", pasid); - pdd = kfd_get_process_device_data(dev, p); + if ((dev->dbgmgr) && (dev->dbgmgr->pasid == p->pasid)) + kfd_dbgmgr_destroy(dev->dbgmgr); - if (!pdd) { - mutex_unlock(&p->mutex); - return; - } + pqm_uninit(&p->pqm); - if (pdd->reset_wavefronts) { - dbgdev_wave_reset_wavefronts(pdd->dev, p); - pdd->reset_wavefronts = false; - } + pdd = kfd_get_process_device_data(dev, p); - /* - * Just mark pdd as unbound, because we still need it - * to call amd_iommu_unbind_pasid() in when the - * process exits. - * We don't call amd_iommu_unbind_pasid() here - * because the IOMMU called us. - */ - pdd->bound = false; + if (!pdd) { + mutex_unlock(&p->mutex); + return; + } - mutex_unlock(&p->mutex); + if (pdd->reset_wavefronts) { + dbgdev_wave_reset_wavefronts(pdd->dev, p); + pdd->reset_wavefronts = false; + } - return; - } + /* + * Just mark pdd as unbound, because we still need it + * to call amd_iommu_unbind_pasid() in when the + * process exits. + * We don't call amd_iommu_unbind_pasid() here + * because the IOMMU called us. + */ + pdd->bound = false; - srcu_read_unlock(&kfd_processes_srcu, idx); + mutex_unlock(&p->mutex); } struct kfd_process_device *kfd_get_first_process_device_data(struct kfd_process *p) diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c b/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c index 7b69070f7ecc..e1fb40b84c72 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c @@ -129,7 +129,7 @@ static int create_cp_queue(struct process_queue_manager *pqm, q_properties->vmid = 0; q_properties->queue_id = qid; - retval = init_queue(q, *q_properties); + retval = init_queue(q, q_properties); if (retval != 0) goto err_init_queue; diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_queue.c b/drivers/gpu/drm/amd/amdkfd/kfd_queue.c index 9a0c90b0702e..0ab197077f2d 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_queue.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_queue.c @@ -63,7 +63,7 @@ void print_queue(struct queue *q) pr_debug("Queue Device Address: 0x%p\n", q->device); } -int init_queue(struct queue **q, struct queue_properties properties) +int init_queue(struct queue **q, const struct queue_properties *properties) { struct queue *tmp; @@ -73,7 +73,7 @@ int init_queue(struct queue **q, struct queue_properties properties) if (!tmp) return -ENOMEM; - memcpy(&tmp->properties, &properties, sizeof(struct queue_properties)); + memcpy(&tmp->properties, properties, sizeof(struct queue_properties)); *q = tmp; return 0; diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_topology.c b/drivers/gpu/drm/amd/amdkfd/kfd_topology.c index 884c96f50c3d..1e5064749959 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_topology.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_topology.c @@ -1090,19 +1090,21 @@ static uint32_t kfd_generate_gpu_id(struct kfd_dev *gpu) { uint32_t hashout; uint32_t buf[7]; + uint64_t local_mem_size; int i; if (!gpu) return 0; + local_mem_size = gpu->kfd2kgd->get_vmem_size(gpu->kgd); + buf[0] = gpu->pdev->devfn; buf[1] = gpu->pdev->subsystem_vendor; buf[2] = gpu->pdev->subsystem_device; buf[3] = gpu->pdev->device; buf[4] = gpu->pdev->bus->number; - buf[5] = (uint32_t)(gpu->kfd2kgd->get_vmem_size(gpu->kgd) - & 0xffffffff); - buf[6] = (uint32_t)(gpu->kfd2kgd->get_vmem_size(gpu->kgd) >> 32); + buf[5] = lower_32_bits(local_mem_size); + buf[6] = upper_32_bits(local_mem_size); for (i = 0, hashout = 0; i < 7; i++) hashout ^= hash_32(buf[i], KFD_GPU_ID_HASH_WIDTH); diff --git a/drivers/gpu/drm/amd/include/amd_shared.h b/drivers/gpu/drm/amd/include/amd_shared.h index db710418f35f..c934b78c9e2f 100644 --- a/drivers/gpu/drm/amd/include/amd_shared.h +++ b/drivers/gpu/drm/amd/include/amd_shared.h @@ -29,7 +29,12 @@ * Supported ASIC types */ enum amd_asic_type { - CHIP_BONAIRE = 0, + CHIP_TAHITI = 0, + CHIP_PITCAIRN, + CHIP_VERDE, + CHIP_OLAND, + CHIP_HAINAN, + CHIP_BONAIRE, CHIP_KAVERI, CHIP_KABINI, CHIP_HAWAII, diff --git a/drivers/gpu/drm/amd/include/asic_reg/si/clearstate_si.h b/drivers/gpu/drm/amd/include/asic_reg/si/clearstate_si.h new file mode 100644 index 000000000000..66e39cdb5cb0 --- /dev/null +++ b/drivers/gpu/drm/amd/include/asic_reg/si/clearstate_si.h @@ -0,0 +1,941 @@ +/* + * Copyright 2013 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + */ + +static const u32 si_SECT_CONTEXT_def_1[] = +{ + 0x00000000, // DB_RENDER_CONTROL + 0x00000000, // DB_COUNT_CONTROL + 0x00000000, // DB_DEPTH_VIEW + 0x00000000, // DB_RENDER_OVERRIDE + 0x00000000, // DB_RENDER_OVERRIDE2 + 0x00000000, // DB_HTILE_DATA_BASE + 0, // HOLE + 0, // HOLE + 0x00000000, // DB_DEPTH_BOUNDS_MIN + 0x00000000, // DB_DEPTH_BOUNDS_MAX + 0x00000000, // DB_STENCIL_CLEAR + 0x00000000, // DB_DEPTH_CLEAR + 0x00000000, // PA_SC_SCREEN_SCISSOR_TL + 0x40004000, // PA_SC_SCREEN_SCISSOR_BR + 0, // HOLE + 0x00000000, // DB_DEPTH_INFO + 0x00000000, // DB_Z_INFO + 0x00000000, // DB_STENCIL_INFO + 0x00000000, // DB_Z_READ_BASE + 0x00000000, // DB_STENCIL_READ_BASE + 0x00000000, // DB_Z_WRITE_BASE + 0x00000000, // DB_STENCIL_WRITE_BASE + 0x00000000, // DB_DEPTH_SIZE + 0x00000000, // DB_DEPTH_SLICE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0x00000000, // TA_BC_BASE_ADDR + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0x00000000, // COHER_DEST_BASE_2 + 0x00000000, // COHER_DEST_BASE_3 + 0x00000000, // PA_SC_WINDOW_OFFSET + 0x80000000, // PA_SC_WINDOW_SCISSOR_TL + 0x40004000, // PA_SC_WINDOW_SCISSOR_BR + 0x0000ffff, // PA_SC_CLIPRECT_RULE + 0x00000000, // PA_SC_CLIPRECT_0_TL + 0x40004000, // PA_SC_CLIPRECT_0_BR + 0x00000000, // PA_SC_CLIPRECT_1_TL + 0x40004000, // PA_SC_CLIPRECT_1_BR + 0x00000000, // PA_SC_CLIPRECT_2_TL + 0x40004000, // PA_SC_CLIPRECT_2_BR + 0x00000000, // PA_SC_CLIPRECT_3_TL + 0x40004000, // PA_SC_CLIPRECT_3_BR + 0xaa99aaaa, // PA_SC_EDGERULE + 0x00000000, // PA_SU_HARDWARE_SCREEN_OFFSET + 0xffffffff, // CB_TARGET_MASK + 0xffffffff, // CB_SHADER_MASK + 0x80000000, // PA_SC_GENERIC_SCISSOR_TL + 0x40004000, // PA_SC_GENERIC_SCISSOR_BR + 0x00000000, // COHER_DEST_BASE_0 + 0x00000000, // COHER_DEST_BASE_1 + 0x80000000, // PA_SC_VPORT_SCISSOR_0_TL + 0x40004000, // PA_SC_VPORT_SCISSOR_0_BR + 0x80000000, // PA_SC_VPORT_SCISSOR_1_TL + 0x40004000, // PA_SC_VPORT_SCISSOR_1_BR + 0x80000000, // PA_SC_VPORT_SCISSOR_2_TL + 0x40004000, // PA_SC_VPORT_SCISSOR_2_BR + 0x80000000, // PA_SC_VPORT_SCISSOR_3_TL + 0x40004000, // PA_SC_VPORT_SCISSOR_3_BR + 0x80000000, // PA_SC_VPORT_SCISSOR_4_TL + 0x40004000, // PA_SC_VPORT_SCISSOR_4_BR + 0x80000000, // PA_SC_VPORT_SCISSOR_5_TL + 0x40004000, // PA_SC_VPORT_SCISSOR_5_BR + 0x80000000, // PA_SC_VPORT_SCISSOR_6_TL + 0x40004000, // PA_SC_VPORT_SCISSOR_6_BR + 0x80000000, // PA_SC_VPORT_SCISSOR_7_TL + 0x40004000, // PA_SC_VPORT_SCISSOR_7_BR + 0x80000000, // PA_SC_VPORT_SCISSOR_8_TL + 0x40004000, // PA_SC_VPORT_SCISSOR_8_BR + 0x80000000, // PA_SC_VPORT_SCISSOR_9_TL + 0x40004000, // PA_SC_VPORT_SCISSOR_9_BR + 0x80000000, // PA_SC_VPORT_SCISSOR_10_TL + 0x40004000, // PA_SC_VPORT_SCISSOR_10_BR + 0x80000000, // PA_SC_VPORT_SCISSOR_11_TL + 0x40004000, // PA_SC_VPORT_SCISSOR_11_BR + 0x80000000, // PA_SC_VPORT_SCISSOR_12_TL + 0x40004000, // PA_SC_VPORT_SCISSOR_12_BR + 0x80000000, // PA_SC_VPORT_SCISSOR_13_TL + 0x40004000, // PA_SC_VPORT_SCISSOR_13_BR + 0x80000000, // PA_SC_VPORT_SCISSOR_14_TL + 0x40004000, // PA_SC_VPORT_SCISSOR_14_BR + 0x80000000, // PA_SC_VPORT_SCISSOR_15_TL + 0x40004000, // PA_SC_VPORT_SCISSOR_15_BR + 0x00000000, // PA_SC_VPORT_ZMIN_0 + 0x3f800000, // PA_SC_VPORT_ZMAX_0 + 0x00000000, // PA_SC_VPORT_ZMIN_1 + 0x3f800000, // PA_SC_VPORT_ZMAX_1 + 0x00000000, // PA_SC_VPORT_ZMIN_2 + 0x3f800000, // PA_SC_VPORT_ZMAX_2 + 0x00000000, // PA_SC_VPORT_ZMIN_3 + 0x3f800000, // PA_SC_VPORT_ZMAX_3 + 0x00000000, // PA_SC_VPORT_ZMIN_4 + 0x3f800000, // PA_SC_VPORT_ZMAX_4 + 0x00000000, // PA_SC_VPORT_ZMIN_5 + 0x3f800000, // PA_SC_VPORT_ZMAX_5 + 0x00000000, // PA_SC_VPORT_ZMIN_6 + 0x3f800000, // PA_SC_VPORT_ZMAX_6 + 0x00000000, // PA_SC_VPORT_ZMIN_7 + 0x3f800000, // PA_SC_VPORT_ZMAX_7 + 0x00000000, // PA_SC_VPORT_ZMIN_8 + 0x3f800000, // PA_SC_VPORT_ZMAX_8 + 0x00000000, // PA_SC_VPORT_ZMIN_9 + 0x3f800000, // PA_SC_VPORT_ZMAX_9 + 0x00000000, // PA_SC_VPORT_ZMIN_10 + 0x3f800000, // PA_SC_VPORT_ZMAX_10 + 0x00000000, // PA_SC_VPORT_ZMIN_11 + 0x3f800000, // PA_SC_VPORT_ZMAX_11 + 0x00000000, // PA_SC_VPORT_ZMIN_12 + 0x3f800000, // PA_SC_VPORT_ZMAX_12 + 0x00000000, // PA_SC_VPORT_ZMIN_13 + 0x3f800000, // PA_SC_VPORT_ZMAX_13 + 0x00000000, // PA_SC_VPORT_ZMIN_14 + 0x3f800000, // PA_SC_VPORT_ZMAX_14 + 0x00000000, // PA_SC_VPORT_ZMIN_15 + 0x3f800000, // PA_SC_VPORT_ZMAX_15 +}; +static const u32 si_SECT_CONTEXT_def_2[] = +{ + 0x00000000, // CP_PERFMON_CNTX_CNTL + 0x00000000, // CP_RINGID + 0x00000000, // CP_VMID + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0xffffffff, // VGT_MAX_VTX_INDX + 0x00000000, // VGT_MIN_VTX_INDX + 0x00000000, // VGT_INDX_OFFSET + 0x00000000, // VGT_MULTI_PRIM_IB_RESET_INDX + 0, // HOLE + 0x00000000, // CB_BLEND_RED + 0x00000000, // CB_BLEND_GREEN + 0x00000000, // CB_BLEND_BLUE + 0x00000000, // CB_BLEND_ALPHA + 0, // HOLE + 0, // HOLE + 0x00000000, // DB_STENCIL_CONTROL + 0x00000000, // DB_STENCILREFMASK + 0x00000000, // DB_STENCILREFMASK_BF + 0, // HOLE + 0x00000000, // PA_CL_VPORT_XSCALE + 0x00000000, // PA_CL_VPORT_XOFFSET + 0x00000000, // PA_CL_VPORT_YSCALE + 0x00000000, // PA_CL_VPORT_YOFFSET + 0x00000000, // PA_CL_VPORT_ZSCALE + 0x00000000, // PA_CL_VPORT_ZOFFSET + 0x00000000, // PA_CL_VPORT_XSCALE_1 + 0x00000000, // PA_CL_VPORT_XOFFSET_1 + 0x00000000, // PA_CL_VPORT_YSCALE_1 + 0x00000000, // PA_CL_VPORT_YOFFSET_1 + 0x00000000, // PA_CL_VPORT_ZSCALE_1 + 0x00000000, // PA_CL_VPORT_ZOFFSET_1 + 0x00000000, // PA_CL_VPORT_XSCALE_2 + 0x00000000, // PA_CL_VPORT_XOFFSET_2 + 0x00000000, // PA_CL_VPORT_YSCALE_2 + 0x00000000, // PA_CL_VPORT_YOFFSET_2 + 0x00000000, // PA_CL_VPORT_ZSCALE_2 + 0x00000000, // PA_CL_VPORT_ZOFFSET_2 + 0x00000000, // PA_CL_VPORT_XSCALE_3 + 0x00000000, // PA_CL_VPORT_XOFFSET_3 + 0x00000000, // PA_CL_VPORT_YSCALE_3 + 0x00000000, // PA_CL_VPORT_YOFFSET_3 + 0x00000000, // PA_CL_VPORT_ZSCALE_3 + 0x00000000, // PA_CL_VPORT_ZOFFSET_3 + 0x00000000, // PA_CL_VPORT_XSCALE_4 + 0x00000000, // PA_CL_VPORT_XOFFSET_4 + 0x00000000, // PA_CL_VPORT_YSCALE_4 + 0x00000000, // PA_CL_VPORT_YOFFSET_4 + 0x00000000, // PA_CL_VPORT_ZSCALE_4 + 0x00000000, // PA_CL_VPORT_ZOFFSET_4 + 0x00000000, // PA_CL_VPORT_XSCALE_5 + 0x00000000, // PA_CL_VPORT_XOFFSET_5 + 0x00000000, // PA_CL_VPORT_YSCALE_5 + 0x00000000, // PA_CL_VPORT_YOFFSET_5 + 0x00000000, // PA_CL_VPORT_ZSCALE_5 + 0x00000000, // PA_CL_VPORT_ZOFFSET_5 + 0x00000000, // PA_CL_VPORT_XSCALE_6 + 0x00000000, // PA_CL_VPORT_XOFFSET_6 + 0x00000000, // PA_CL_VPORT_YSCALE_6 + 0x00000000, // PA_CL_VPORT_YOFFSET_6 + 0x00000000, // PA_CL_VPORT_ZSCALE_6 + 0x00000000, // PA_CL_VPORT_ZOFFSET_6 + 0x00000000, // PA_CL_VPORT_XSCALE_7 + 0x00000000, // PA_CL_VPORT_XOFFSET_7 + 0x00000000, // PA_CL_VPORT_YSCALE_7 + 0x00000000, // PA_CL_VPORT_YOFFSET_7 + 0x00000000, // PA_CL_VPORT_ZSCALE_7 + 0x00000000, // PA_CL_VPORT_ZOFFSET_7 + 0x00000000, // PA_CL_VPORT_XSCALE_8 + 0x00000000, // PA_CL_VPORT_XOFFSET_8 + 0x00000000, // PA_CL_VPORT_YSCALE_8 + 0x00000000, // PA_CL_VPORT_YOFFSET_8 + 0x00000000, // PA_CL_VPORT_ZSCALE_8 + 0x00000000, // PA_CL_VPORT_ZOFFSET_8 + 0x00000000, // PA_CL_VPORT_XSCALE_9 + 0x00000000, // PA_CL_VPORT_XOFFSET_9 + 0x00000000, // PA_CL_VPORT_YSCALE_9 + 0x00000000, // PA_CL_VPORT_YOFFSET_9 + 0x00000000, // PA_CL_VPORT_ZSCALE_9 + 0x00000000, // PA_CL_VPORT_ZOFFSET_9 + 0x00000000, // PA_CL_VPORT_XSCALE_10 + 0x00000000, // PA_CL_VPORT_XOFFSET_10 + 0x00000000, // PA_CL_VPORT_YSCALE_10 + 0x00000000, // PA_CL_VPORT_YOFFSET_10 + 0x00000000, // PA_CL_VPORT_ZSCALE_10 + 0x00000000, // PA_CL_VPORT_ZOFFSET_10 + 0x00000000, // PA_CL_VPORT_XSCALE_11 + 0x00000000, // PA_CL_VPORT_XOFFSET_11 + 0x00000000, // PA_CL_VPORT_YSCALE_11 + 0x00000000, // PA_CL_VPORT_YOFFSET_11 + 0x00000000, // PA_CL_VPORT_ZSCALE_11 + 0x00000000, // PA_CL_VPORT_ZOFFSET_11 + 0x00000000, // PA_CL_VPORT_XSCALE_12 + 0x00000000, // PA_CL_VPORT_XOFFSET_12 + 0x00000000, // PA_CL_VPORT_YSCALE_12 + 0x00000000, // PA_CL_VPORT_YOFFSET_12 + 0x00000000, // PA_CL_VPORT_ZSCALE_12 + 0x00000000, // PA_CL_VPORT_ZOFFSET_12 + 0x00000000, // PA_CL_VPORT_XSCALE_13 + 0x00000000, // PA_CL_VPORT_XOFFSET_13 + 0x00000000, // PA_CL_VPORT_YSCALE_13 + 0x00000000, // PA_CL_VPORT_YOFFSET_13 + 0x00000000, // PA_CL_VPORT_ZSCALE_13 + 0x00000000, // PA_CL_VPORT_ZOFFSET_13 + 0x00000000, // PA_CL_VPORT_XSCALE_14 + 0x00000000, // PA_CL_VPORT_XOFFSET_14 + 0x00000000, // PA_CL_VPORT_YSCALE_14 + 0x00000000, // PA_CL_VPORT_YOFFSET_14 + 0x00000000, // PA_CL_VPORT_ZSCALE_14 + 0x00000000, // PA_CL_VPORT_ZOFFSET_14 + 0x00000000, // PA_CL_VPORT_XSCALE_15 + 0x00000000, // PA_CL_VPORT_XOFFSET_15 + 0x00000000, // PA_CL_VPORT_YSCALE_15 + 0x00000000, // PA_CL_VPORT_YOFFSET_15 + 0x00000000, // PA_CL_VPORT_ZSCALE_15 + 0x00000000, // PA_CL_VPORT_ZOFFSET_15 + 0x00000000, // PA_CL_UCP_0_X + 0x00000000, // PA_CL_UCP_0_Y + 0x00000000, // PA_CL_UCP_0_Z + 0x00000000, // PA_CL_UCP_0_W + 0x00000000, // PA_CL_UCP_1_X + 0x00000000, // PA_CL_UCP_1_Y + 0x00000000, // PA_CL_UCP_1_Z + 0x00000000, // PA_CL_UCP_1_W + 0x00000000, // PA_CL_UCP_2_X + 0x00000000, // PA_CL_UCP_2_Y + 0x00000000, // PA_CL_UCP_2_Z + 0x00000000, // PA_CL_UCP_2_W + 0x00000000, // PA_CL_UCP_3_X + 0x00000000, // PA_CL_UCP_3_Y + 0x00000000, // PA_CL_UCP_3_Z + 0x00000000, // PA_CL_UCP_3_W + 0x00000000, // PA_CL_UCP_4_X + 0x00000000, // PA_CL_UCP_4_Y + 0x00000000, // PA_CL_UCP_4_Z + 0x00000000, // PA_CL_UCP_4_W + 0x00000000, // PA_CL_UCP_5_X + 0x00000000, // PA_CL_UCP_5_Y + 0x00000000, // PA_CL_UCP_5_Z + 0x00000000, // PA_CL_UCP_5_W + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0x00000000, // SPI_PS_INPUT_CNTL_0 + 0x00000000, // SPI_PS_INPUT_CNTL_1 + 0x00000000, // SPI_PS_INPUT_CNTL_2 + 0x00000000, // SPI_PS_INPUT_CNTL_3 + 0x00000000, // SPI_PS_INPUT_CNTL_4 + 0x00000000, // SPI_PS_INPUT_CNTL_5 + 0x00000000, // SPI_PS_INPUT_CNTL_6 + 0x00000000, // SPI_PS_INPUT_CNTL_7 + 0x00000000, // SPI_PS_INPUT_CNTL_8 + 0x00000000, // SPI_PS_INPUT_CNTL_9 + 0x00000000, // SPI_PS_INPUT_CNTL_10 + 0x00000000, // SPI_PS_INPUT_CNTL_11 + 0x00000000, // SPI_PS_INPUT_CNTL_12 + 0x00000000, // SPI_PS_INPUT_CNTL_13 + 0x00000000, // SPI_PS_INPUT_CNTL_14 + 0x00000000, // SPI_PS_INPUT_CNTL_15 + 0x00000000, // SPI_PS_INPUT_CNTL_16 + 0x00000000, // SPI_PS_INPUT_CNTL_17 + 0x00000000, // SPI_PS_INPUT_CNTL_18 + 0x00000000, // SPI_PS_INPUT_CNTL_19 + 0x00000000, // SPI_PS_INPUT_CNTL_20 + 0x00000000, // SPI_PS_INPUT_CNTL_21 + 0x00000000, // SPI_PS_INPUT_CNTL_22 + 0x00000000, // SPI_PS_INPUT_CNTL_23 + 0x00000000, // SPI_PS_INPUT_CNTL_24 + 0x00000000, // SPI_PS_INPUT_CNTL_25 + 0x00000000, // SPI_PS_INPUT_CNTL_26 + 0x00000000, // SPI_PS_INPUT_CNTL_27 + 0x00000000, // SPI_PS_INPUT_CNTL_28 + 0x00000000, // SPI_PS_INPUT_CNTL_29 + 0x00000000, // SPI_PS_INPUT_CNTL_30 + 0x00000000, // SPI_PS_INPUT_CNTL_31 + 0x00000000, // SPI_VS_OUT_CONFIG + 0, // HOLE + 0x00000000, // SPI_PS_INPUT_ENA + 0x00000000, // SPI_PS_INPUT_ADDR + 0x00000000, // SPI_INTERP_CONTROL_0 + 0x00000002, // SPI_PS_IN_CONTROL + 0, // HOLE + 0x00000000, // SPI_BARYC_CNTL + 0, // HOLE + 0x00000000, // SPI_TMPRING_SIZE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0x00000000, // SPI_WAVE_MGMT_1 + 0x00000000, // SPI_WAVE_MGMT_2 + 0x00000000, // SPI_SHADER_POS_FORMAT + 0x00000000, // SPI_SHADER_Z_FORMAT + 0x00000000, // SPI_SHADER_COL_FORMAT + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0x00000000, // CB_BLEND0_CONTROL + 0x00000000, // CB_BLEND1_CONTROL + 0x00000000, // CB_BLEND2_CONTROL + 0x00000000, // CB_BLEND3_CONTROL + 0x00000000, // CB_BLEND4_CONTROL + 0x00000000, // CB_BLEND5_CONTROL + 0x00000000, // CB_BLEND6_CONTROL + 0x00000000, // CB_BLEND7_CONTROL +}; +static const u32 si_SECT_CONTEXT_def_3[] = +{ + 0x00000000, // PA_CL_POINT_X_RAD + 0x00000000, // PA_CL_POINT_Y_RAD + 0x00000000, // PA_CL_POINT_SIZE + 0x00000000, // PA_CL_POINT_CULL_RAD + 0x00000000, // VGT_DMA_BASE_HI + 0x00000000, // VGT_DMA_BASE +}; +static const u32 si_SECT_CONTEXT_def_4[] = +{ + 0x00000000, // DB_DEPTH_CONTROL + 0x00000000, // DB_EQAA + 0x00000000, // CB_COLOR_CONTROL + 0x00000000, // DB_SHADER_CONTROL + 0x00090000, // PA_CL_CLIP_CNTL + 0x00000004, // PA_SU_SC_MODE_CNTL + 0x00000000, // PA_CL_VTE_CNTL + 0x00000000, // PA_CL_VS_OUT_CNTL + 0x00000000, // PA_CL_NANINF_CNTL + 0x00000000, // PA_SU_LINE_STIPPLE_CNTL + 0x00000000, // PA_SU_LINE_STIPPLE_SCALE + 0x00000000, // PA_SU_PRIM_FILTER_CNTL + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0x00000000, // PA_SU_POINT_SIZE + 0x00000000, // PA_SU_POINT_MINMAX + 0x00000000, // PA_SU_LINE_CNTL + 0x00000000, // PA_SC_LINE_STIPPLE + 0x00000000, // VGT_OUTPUT_PATH_CNTL + 0x00000000, // VGT_HOS_CNTL + 0x00000000, // VGT_HOS_MAX_TESS_LEVEL + 0x00000000, // VGT_HOS_MIN_TESS_LEVEL + 0x00000000, // VGT_HOS_REUSE_DEPTH + 0x00000000, // VGT_GROUP_PRIM_TYPE + 0x00000000, // VGT_GROUP_FIRST_DECR + 0x00000000, // VGT_GROUP_DECR + 0x00000000, // VGT_GROUP_VECT_0_CNTL + 0x00000000, // VGT_GROUP_VECT_1_CNTL + 0x00000000, // VGT_GROUP_VECT_0_FMT_CNTL + 0x00000000, // VGT_GROUP_VECT_1_FMT_CNTL + 0x00000000, // VGT_GS_MODE + 0, // HOLE + 0x00000000, // PA_SC_MODE_CNTL_0 + 0x00000000, // PA_SC_MODE_CNTL_1 + 0x00000000, // VGT_ENHANCE + 0x00000100, // VGT_GS_PER_ES + 0x00000080, // VGT_ES_PER_GS + 0x00000002, // VGT_GS_PER_VS + 0x00000000, // VGT_GSVS_RING_OFFSET_1 + 0x00000000, // VGT_GSVS_RING_OFFSET_2 + 0x00000000, // VGT_GSVS_RING_OFFSET_3 + 0x00000000, // VGT_GS_OUT_PRIM_TYPE + 0x00000000, // IA_ENHANCE +}; +static const u32 si_SECT_CONTEXT_def_5[] = +{ + 0x00000000, // VGT_PRIMITIVEID_EN +}; +static const u32 si_SECT_CONTEXT_def_6[] = +{ + 0x00000000, // VGT_PRIMITIVEID_RESET +}; +static const u32 si_SECT_CONTEXT_def_7[] = +{ + 0x00000000, // VGT_MULTI_PRIM_IB_RESET_EN + 0, // HOLE + 0, // HOLE + 0x00000000, // VGT_INSTANCE_STEP_RATE_0 + 0x00000000, // VGT_INSTANCE_STEP_RATE_1 + 0x000000ff, // IA_MULTI_VGT_PARAM + 0x00000000, // VGT_ESGS_RING_ITEMSIZE + 0x00000000, // VGT_GSVS_RING_ITEMSIZE + 0x00000000, // VGT_REUSE_OFF + 0x00000000, // VGT_VTX_CNT_EN + 0x00000000, // DB_HTILE_SURFACE + 0x00000000, // DB_SRESULTS_COMPARE_STATE0 + 0x00000000, // DB_SRESULTS_COMPARE_STATE1 + 0x00000000, // DB_PRELOAD_CONTROL + 0, // HOLE + 0x00000000, // VGT_STRMOUT_BUFFER_SIZE_0 + 0x00000000, // VGT_STRMOUT_VTX_STRIDE_0 + 0, // HOLE + 0x00000000, // VGT_STRMOUT_BUFFER_OFFSET_0 + 0x00000000, // VGT_STRMOUT_BUFFER_SIZE_1 + 0x00000000, // VGT_STRMOUT_VTX_STRIDE_1 + 0, // HOLE + 0x00000000, // VGT_STRMOUT_BUFFER_OFFSET_1 + 0x00000000, // VGT_STRMOUT_BUFFER_SIZE_2 + 0x00000000, // VGT_STRMOUT_VTX_STRIDE_2 + 0, // HOLE + 0x00000000, // VGT_STRMOUT_BUFFER_OFFSET_2 + 0x00000000, // VGT_STRMOUT_BUFFER_SIZE_3 + 0x00000000, // VGT_STRMOUT_VTX_STRIDE_3 + 0, // HOLE + 0x00000000, // VGT_STRMOUT_BUFFER_OFFSET_3 + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0x00000000, // VGT_STRMOUT_DRAW_OPAQUE_OFFSET + 0x00000000, // VGT_STRMOUT_DRAW_OPAQUE_BUFFER_FILLED_SIZE + 0x00000000, // VGT_STRMOUT_DRAW_OPAQUE_VERTEX_STRIDE + 0, // HOLE + 0x00000000, // VGT_GS_MAX_VERT_OUT + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0x00000000, // VGT_SHADER_STAGES_EN + 0x00000000, // VGT_LS_HS_CONFIG + 0x00000000, // VGT_GS_VERT_ITEMSIZE + 0x00000000, // VGT_GS_VERT_ITEMSIZE_1 + 0x00000000, // VGT_GS_VERT_ITEMSIZE_2 + 0x00000000, // VGT_GS_VERT_ITEMSIZE_3 + 0x00000000, // VGT_TF_PARAM + 0x00000000, // DB_ALPHA_TO_MASK + 0, // HOLE + 0x00000000, // PA_SU_POLY_OFFSET_DB_FMT_CNTL + 0x00000000, // PA_SU_POLY_OFFSET_CLAMP + 0x00000000, // PA_SU_POLY_OFFSET_FRONT_SCALE + 0x00000000, // PA_SU_POLY_OFFSET_FRONT_OFFSET + 0x00000000, // PA_SU_POLY_OFFSET_BACK_SCALE + 0x00000000, // PA_SU_POLY_OFFSET_BACK_OFFSET + 0x00000000, // VGT_GS_INSTANCE_CNT + 0x00000000, // VGT_STRMOUT_CONFIG + 0x00000000, // VGT_STRMOUT_BUFFER_CONFIG + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0x00000000, // PA_SC_CENTROID_PRIORITY_0 + 0x00000000, // PA_SC_CENTROID_PRIORITY_1 + 0x00001000, // PA_SC_LINE_CNTL + 0x00000000, // PA_SC_AA_CONFIG + 0x00000005, // PA_SU_VTX_CNTL + 0x3f800000, // PA_CL_GB_VERT_CLIP_ADJ + 0x3f800000, // PA_CL_GB_VERT_DISC_ADJ + 0x3f800000, // PA_CL_GB_HORZ_CLIP_ADJ + 0x3f800000, // PA_CL_GB_HORZ_DISC_ADJ + 0x00000000, // PA_SC_AA_SAMPLE_LOCS_PIXEL_X0Y0_0 + 0x00000000, // PA_SC_AA_SAMPLE_LOCS_PIXEL_X0Y0_1 + 0x00000000, // PA_SC_AA_SAMPLE_LOCS_PIXEL_X0Y0_2 + 0x00000000, // PA_SC_AA_SAMPLE_LOCS_PIXEL_X0Y0_3 + 0x00000000, // PA_SC_AA_SAMPLE_LOCS_PIXEL_X1Y0_0 + 0x00000000, // PA_SC_AA_SAMPLE_LOCS_PIXEL_X1Y0_1 + 0x00000000, // PA_SC_AA_SAMPLE_LOCS_PIXEL_X1Y0_2 + 0x00000000, // PA_SC_AA_SAMPLE_LOCS_PIXEL_X1Y0_3 + 0x00000000, // PA_SC_AA_SAMPLE_LOCS_PIXEL_X0Y1_0 + 0x00000000, // PA_SC_AA_SAMPLE_LOCS_PIXEL_X0Y1_1 + 0x00000000, // PA_SC_AA_SAMPLE_LOCS_PIXEL_X0Y1_2 + 0x00000000, // PA_SC_AA_SAMPLE_LOCS_PIXEL_X0Y1_3 + 0x00000000, // PA_SC_AA_SAMPLE_LOCS_PIXEL_X1Y1_0 + 0x00000000, // PA_SC_AA_SAMPLE_LOCS_PIXEL_X1Y1_1 + 0x00000000, // PA_SC_AA_SAMPLE_LOCS_PIXEL_X1Y1_2 + 0x00000000, // PA_SC_AA_SAMPLE_LOCS_PIXEL_X1Y1_3 + 0xffffffff, // PA_SC_AA_MASK_X0Y0_X1Y0 + 0xffffffff, // PA_SC_AA_MASK_X0Y1_X1Y1 + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0, // HOLE + 0x0000000e, // VGT_VERTEX_REUSE_BLOCK_CNTL + 0x00000010, // VGT_OUT_DEALLOC_CNTL + 0x00000000, // CB_COLOR0_BASE + 0x00000000, // CB_COLOR0_PITCH + 0x00000000, // CB_COLOR0_SLICE + 0x00000000, // CB_COLOR0_VIEW + 0x00000000, // CB_COLOR0_INFO + 0x00000000, // CB_COLOR0_ATTRIB + 0, // HOLE + 0x00000000, // CB_COLOR0_CMASK + 0x00000000, // CB_COLOR0_CMASK_SLICE + 0x00000000, // CB_COLOR0_FMASK + 0x00000000, // CB_COLOR0_FMASK_SLICE + 0x00000000, // CB_COLOR0_CLEAR_WORD0 + 0x00000000, // CB_COLOR0_CLEAR_WORD1 + 0, // HOLE + 0, // HOLE + 0x00000000, // CB_COLOR1_BASE + 0x00000000, // CB_COLOR1_PITCH + 0x00000000, // CB_COLOR1_SLICE + 0x00000000, // CB_COLOR1_VIEW + 0x00000000, // CB_COLOR1_INFO + 0x00000000, // CB_COLOR1_ATTRIB + 0, // HOLE + 0x00000000, // CB_COLOR1_CMASK + 0x00000000, // CB_COLOR1_CMASK_SLICE + 0x00000000, // CB_COLOR1_FMASK + 0x00000000, // CB_COLOR1_FMASK_SLICE + 0x00000000, // CB_COLOR1_CLEAR_WORD0 + 0x00000000, // CB_COLOR1_CLEAR_WORD1 + 0, // HOLE + 0, // HOLE + 0x00000000, // CB_COLOR2_BASE + 0x00000000, // CB_COLOR2_PITCH + 0x00000000, // CB_COLOR2_SLICE + 0x00000000, // CB_COLOR2_VIEW + 0x00000000, // CB_COLOR2_INFO + 0x00000000, // CB_COLOR2_ATTRIB + 0, // HOLE + 0x00000000, // CB_COLOR2_CMASK + 0x00000000, // CB_COLOR2_CMASK_SLICE + 0x00000000, // CB_COLOR2_FMASK + 0x00000000, // CB_COLOR2_FMASK_SLICE + 0x00000000, // CB_COLOR2_CLEAR_WORD0 + 0x00000000, // CB_COLOR2_CLEAR_WORD1 + 0, // HOLE + 0, // HOLE + 0x00000000, // CB_COLOR3_BASE + 0x00000000, // CB_COLOR3_PITCH + 0x00000000, // CB_COLOR3_SLICE + 0x00000000, // CB_COLOR3_VIEW + 0x00000000, // CB_COLOR3_INFO + 0x00000000, // CB_COLOR3_ATTRIB + 0, // HOLE + 0x00000000, // CB_COLOR3_CMASK + 0x00000000, // CB_COLOR3_CMASK_SLICE + 0x00000000, // CB_COLOR3_FMASK + 0x00000000, // CB_COLOR3_FMASK_SLICE + 0x00000000, // CB_COLOR3_CLEAR_WORD0 + 0x00000000, // CB_COLOR3_CLEAR_WORD1 + 0, // HOLE + 0, // HOLE + 0x00000000, // CB_COLOR4_BASE + 0x00000000, // CB_COLOR4_PITCH + 0x00000000, // CB_COLOR4_SLICE + 0x00000000, // CB_COLOR4_VIEW + 0x00000000, // CB_COLOR4_INFO + 0x00000000, // CB_COLOR4_ATTRIB + 0, // HOLE + 0x00000000, // CB_COLOR4_CMASK + 0x00000000, // CB_COLOR4_CMASK_SLICE + 0x00000000, // CB_COLOR4_FMASK + 0x00000000, // CB_COLOR4_FMASK_SLICE + 0x00000000, // CB_COLOR4_CLEAR_WORD0 + 0x00000000, // CB_COLOR4_CLEAR_WORD1 + 0, // HOLE + 0, // HOLE + 0x00000000, // CB_COLOR5_BASE + 0x00000000, // CB_COLOR5_PITCH + 0x00000000, // CB_COLOR5_SLICE + 0x00000000, // CB_COLOR5_VIEW + 0x00000000, // CB_COLOR5_INFO + 0x00000000, // CB_COLOR5_ATTRIB + 0, // HOLE + 0x00000000, // CB_COLOR5_CMASK + 0x00000000, // CB_COLOR5_CMASK_SLICE + 0x00000000, // CB_COLOR5_FMASK + 0x00000000, // CB_COLOR5_FMASK_SLICE + 0x00000000, // CB_COLOR5_CLEAR_WORD0 + 0x00000000, // CB_COLOR5_CLEAR_WORD1 + 0, // HOLE + 0, // HOLE + 0x00000000, // CB_COLOR6_BASE + 0x00000000, // CB_COLOR6_PITCH + 0x00000000, // CB_COLOR6_SLICE + 0x00000000, // CB_COLOR6_VIEW + 0x00000000, // CB_COLOR6_INFO + 0x00000000, // CB_COLOR6_ATTRIB + 0, // HOLE + 0x00000000, // CB_COLOR6_CMASK + 0x00000000, // CB_COLOR6_CMASK_SLICE + 0x00000000, // CB_COLOR6_FMASK + 0x00000000, // CB_COLOR6_FMASK_SLICE + 0x00000000, // CB_COLOR6_CLEAR_WORD0 + 0x00000000, // CB_COLOR6_CLEAR_WORD1 + 0, // HOLE + 0, // HOLE + 0x00000000, // CB_COLOR7_BASE + 0x00000000, // CB_COLOR7_PITCH + 0x00000000, // CB_COLOR7_SLICE + 0x00000000, // CB_COLOR7_VIEW + 0x00000000, // CB_COLOR7_INFO + 0x00000000, // CB_COLOR7_ATTRIB + 0, // HOLE + 0x00000000, // CB_COLOR7_CMASK + 0x00000000, // CB_COLOR7_CMASK_SLICE + 0x00000000, // CB_COLOR7_FMASK + 0x00000000, // CB_COLOR7_FMASK_SLICE + 0x00000000, // CB_COLOR7_CLEAR_WORD0 + 0x00000000, // CB_COLOR7_CLEAR_WORD1 +}; +static const struct cs_extent_def si_SECT_CONTEXT_defs[] = +{ + {si_SECT_CONTEXT_def_1, 0x0000a000, 212 }, + {si_SECT_CONTEXT_def_2, 0x0000a0d8, 272 }, + {si_SECT_CONTEXT_def_3, 0x0000a1f5, 6 }, + {si_SECT_CONTEXT_def_4, 0x0000a200, 157 }, + {si_SECT_CONTEXT_def_5, 0x0000a2a1, 1 }, + {si_SECT_CONTEXT_def_6, 0x0000a2a3, 1 }, + {si_SECT_CONTEXT_def_7, 0x0000a2a5, 233 }, + { NULL, 0, 0 } +}; +static const struct cs_section_def si_cs_data[] = { + { si_SECT_CONTEXT_defs, SECT_CONTEXT }, + { NULL, SECT_NONE } +}; diff --git a/drivers/gpu/drm/amd/include/asic_reg/si/si_reg.h b/drivers/gpu/drm/amd/include/asic_reg/si/si_reg.h new file mode 100644 index 000000000000..895c8e2353e3 --- /dev/null +++ b/drivers/gpu/drm/amd/include/asic_reg/si/si_reg.h @@ -0,0 +1,105 @@ +/* + * Copyright 2010 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + * Authors: Alex Deucher + */ +#ifndef __SI_REG_H__ +#define __SI_REG_H__ + +/* SI */ +#define SI_DC_GPIO_HPD_MASK 0x196c +#define SI_DC_GPIO_HPD_A 0x196d +#define SI_DC_GPIO_HPD_EN 0x196e +#define SI_DC_GPIO_HPD_Y 0x196f + +#define SI_GRPH_CONTROL 0x1a01 +# define SI_GRPH_DEPTH(x) (((x) & 0x3) << 0) +# define SI_GRPH_DEPTH_8BPP 0 +# define SI_GRPH_DEPTH_16BPP 1 +# define SI_GRPH_DEPTH_32BPP 2 +# define SI_GRPH_NUM_BANKS(x) (((x) & 0x3) << 2) +# define SI_ADDR_SURF_2_BANK 0 +# define SI_ADDR_SURF_4_BANK 1 +# define SI_ADDR_SURF_8_BANK 2 +# define SI_ADDR_SURF_16_BANK 3 +# define SI_GRPH_Z(x) (((x) & 0x3) << 4) +# define SI_GRPH_BANK_WIDTH(x) (((x) & 0x3) << 6) +# define SI_ADDR_SURF_BANK_WIDTH_1 0 +# define SI_ADDR_SURF_BANK_WIDTH_2 1 +# define SI_ADDR_SURF_BANK_WIDTH_4 2 +# define SI_ADDR_SURF_BANK_WIDTH_8 3 +# define SI_GRPH_FORMAT(x) (((x) & 0x7) << 8) +/* 8 BPP */ +# define SI_GRPH_FORMAT_INDEXED 0 +/* 16 BPP */ +# define SI_GRPH_FORMAT_ARGB1555 0 +# define SI_GRPH_FORMAT_ARGB565 1 +# define SI_GRPH_FORMAT_ARGB4444 2 +# define SI_GRPH_FORMAT_AI88 3 +# define SI_GRPH_FORMAT_MONO16 4 +# define SI_GRPH_FORMAT_BGRA5551 5 +/* 32 BPP */ +# define SI_GRPH_FORMAT_ARGB8888 0 +# define SI_GRPH_FORMAT_ARGB2101010 1 +# define SI_GRPH_FORMAT_32BPP_DIG 2 +# define SI_GRPH_FORMAT_8B_ARGB2101010 3 +# define SI_GRPH_FORMAT_BGRA1010102 4 +# define SI_GRPH_FORMAT_8B_BGRA1010102 5 +# define SI_GRPH_FORMAT_RGB111110 6 +# define SI_GRPH_FORMAT_BGR101111 7 +# define SI_GRPH_BANK_HEIGHT(x) (((x) & 0x3) << 11) +# define SI_ADDR_SURF_BANK_HEIGHT_1 0 +# define SI_ADDR_SURF_BANK_HEIGHT_2 1 +# define SI_ADDR_SURF_BANK_HEIGHT_4 2 +# define SI_ADDR_SURF_BANK_HEIGHT_8 3 +# define SI_GRPH_TILE_SPLIT(x) (((x) & 0x7) << 13) +# define SI_ADDR_SURF_TILE_SPLIT_64B 0 +# define SI_ADDR_SURF_TILE_SPLIT_128B 1 +# define SI_ADDR_SURF_TILE_SPLIT_256B 2 +# define SI_ADDR_SURF_TILE_SPLIT_512B 3 +# define SI_ADDR_SURF_TILE_SPLIT_1KB 4 +# define SI_ADDR_SURF_TILE_SPLIT_2KB 5 +# define SI_ADDR_SURF_TILE_SPLIT_4KB 6 +# define SI_GRPH_MACRO_TILE_ASPECT(x) (((x) & 0x3) << 18) +# define SI_ADDR_SURF_MACRO_TILE_ASPECT_1 0 +# define SI_ADDR_SURF_MACRO_TILE_ASPECT_2 1 +# define SI_ADDR_SURF_MACRO_TILE_ASPECT_4 2 +# define SI_ADDR_SURF_MACRO_TILE_ASPECT_8 3 +# define SI_GRPH_ARRAY_MODE(x) (((x) & 0x7) << 20) +# define SI_GRPH_ARRAY_LINEAR_GENERAL 0 +# define SI_GRPH_ARRAY_LINEAR_ALIGNED 1 +# define SI_GRPH_ARRAY_1D_TILED_THIN1 2 +# define SI_GRPH_ARRAY_2D_TILED_THIN1 4 +# define SI_GRPH_PIPE_CONFIG(x) (((x) & 0x1f) << 24) +# define SI_ADDR_SURF_P2 0 +# define SI_ADDR_SURF_P4_8x16 4 +# define SI_ADDR_SURF_P4_16x16 5 +# define SI_ADDR_SURF_P4_16x32 6 +# define SI_ADDR_SURF_P4_32x32 7 +# define SI_ADDR_SURF_P8_16x16_8x16 8 +# define SI_ADDR_SURF_P8_16x32_8x16 9 +# define SI_ADDR_SURF_P8_32x32_8x16 10 +# define SI_ADDR_SURF_P8_16x32_16x16 11 +# define SI_ADDR_SURF_P8_32x32_16x16 12 +# define SI_ADDR_SURF_P8_32x32_16x32 13 +# define SI_ADDR_SURF_P8_32x64_32x32 14 + +#endif diff --git a/drivers/gpu/drm/amd/include/asic_reg/si/sid.h b/drivers/gpu/drm/amd/include/asic_reg/si/sid.h new file mode 100644 index 000000000000..8c5608a4d526 --- /dev/null +++ b/drivers/gpu/drm/amd/include/asic_reg/si/sid.h @@ -0,0 +1,2426 @@ +/* + * Copyright 2011 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + * Authors: Alex Deucher + */ +#ifndef SI_H +#define SI_H + +#define TAHITI_RB_BITMAP_WIDTH_PER_SH 2 + +#define TAHITI_GB_ADDR_CONFIG_GOLDEN 0x12011003 +#define VERDE_GB_ADDR_CONFIG_GOLDEN 0x12010002 +#define HAINAN_GB_ADDR_CONFIG_GOLDEN 0x02010001 + +#define SI_MAX_SH_GPRS 256 +#define SI_MAX_TEMP_GPRS 16 +#define SI_MAX_SH_THREADS 256 +#define SI_MAX_SH_STACK_ENTRIES 4096 +#define SI_MAX_FRC_EOV_CNT 16384 +#define SI_MAX_BACKENDS 8 +#define SI_MAX_BACKENDS_MASK 0xFF +#define SI_MAX_BACKENDS_PER_SE_MASK 0x0F +#define SI_MAX_SIMDS 12 +#define SI_MAX_SIMDS_MASK 0x0FFF +#define SI_MAX_SIMDS_PER_SE_MASK 0x00FF +#define SI_MAX_PIPES 8 +#define SI_MAX_PIPES_MASK 0xFF +#define SI_MAX_PIPES_PER_SIMD_MASK 0x3F +#define SI_MAX_LDS_NUM 0xFFFF +#define SI_MAX_TCC 16 +#define SI_MAX_TCC_MASK 0xFFFF + +#define AMDGPU_NUM_OF_VMIDS 8 + +/* SMC IND accessor regs */ +#define SMC_IND_INDEX_0 0x80 +#define SMC_IND_DATA_0 0x81 + +#define SMC_IND_ACCESS_CNTL 0x8A +# define AUTO_INCREMENT_IND_0 (1 << 0) +#define SMC_MESSAGE_0 0x8B +#define SMC_RESP_0 0x8C + +/* CG IND registers are accessed via SMC indirect space + SMC_CG_IND_START */ +#define SMC_CG_IND_START 0xc0030000 +#define SMC_CG_IND_END 0xc0040000 + +#define CG_CGTT_LOCAL_0 0x400 +#define CG_CGTT_LOCAL_1 0x401 + +/* SMC IND registers */ +#define SMC_SYSCON_RESET_CNTL 0x80000000 +# define RST_REG (1 << 0) +#define SMC_SYSCON_CLOCK_CNTL_0 0x80000004 +# define CK_DISABLE (1 << 0) +# define CKEN (1 << 24) + +#define VGA_HDP_CONTROL 0xCA +#define VGA_MEMORY_DISABLE (1 << 4) + +#define DCCG_DISP_SLOW_SELECT_REG 0x13F +#define DCCG_DISP1_SLOW_SELECT(x) ((x) << 0) +#define DCCG_DISP1_SLOW_SELECT_MASK (7 << 0) +#define DCCG_DISP1_SLOW_SELECT_SHIFT 0 +#define DCCG_DISP2_SLOW_SELECT(x) ((x) << 4) +#define DCCG_DISP2_SLOW_SELECT_MASK (7 << 4) +#define DCCG_DISP2_SLOW_SELECT_SHIFT 4 + +#define CG_SPLL_FUNC_CNTL 0x180 +#define SPLL_RESET (1 << 0) +#define SPLL_SLEEP (1 << 1) +#define SPLL_BYPASS_EN (1 << 3) +#define SPLL_REF_DIV(x) ((x) << 4) +#define SPLL_REF_DIV_MASK (0x3f << 4) +#define SPLL_PDIV_A(x) ((x) << 20) +#define SPLL_PDIV_A_MASK (0x7f << 20) +#define SPLL_PDIV_A_SHIFT 20 +#define CG_SPLL_FUNC_CNTL_2 0x181 +#define SCLK_MUX_SEL(x) ((x) << 0) +#define SCLK_MUX_SEL_MASK (0x1ff << 0) +#define SPLL_CTLREQ_CHG (1 << 23) +#define SCLK_MUX_UPDATE (1 << 26) +#define CG_SPLL_FUNC_CNTL_3 0x182 +#define SPLL_FB_DIV(x) ((x) << 0) +#define SPLL_FB_DIV_MASK (0x3ffffff << 0) +#define SPLL_FB_DIV_SHIFT 0 +#define SPLL_DITHEN (1 << 28) +#define CG_SPLL_FUNC_CNTL_4 0x183 + +#define SPLL_STATUS 0x185 +#define SPLL_CHG_STATUS (1 << 1) +#define SPLL_CNTL_MODE 0x186 +#define SPLL_SW_DIR_CONTROL (1 << 0) +# define SPLL_REFCLK_SEL(x) ((x) << 26) +# define SPLL_REFCLK_SEL_MASK (3 << 26) + +#define CG_SPLL_SPREAD_SPECTRUM 0x188 +#define SSEN (1 << 0) +#define CLK_S(x) ((x) << 4) +#define CLK_S_MASK (0xfff << 4) +#define CLK_S_SHIFT 4 +#define CG_SPLL_SPREAD_SPECTRUM_2 0x189 +#define CLK_V(x) ((x) << 0) +#define CLK_V_MASK (0x3ffffff << 0) +#define CLK_V_SHIFT 0 + +#define CG_SPLL_AUTOSCALE_CNTL 0x18b +# define AUTOSCALE_ON_SS_CLEAR (1 << 9) + +/* discrete uvd clocks */ +#define CG_UPLL_FUNC_CNTL 0x18d +# define UPLL_RESET_MASK 0x00000001 +# define UPLL_SLEEP_MASK 0x00000002 +# define UPLL_BYPASS_EN_MASK 0x00000004 +# define UPLL_CTLREQ_MASK 0x00000008 +# define UPLL_VCO_MODE_MASK 0x00000600 +# define UPLL_REF_DIV_MASK 0x003F0000 +# define UPLL_CTLACK_MASK 0x40000000 +# define UPLL_CTLACK2_MASK 0x80000000 +#define CG_UPLL_FUNC_CNTL_2 0x18e +# define UPLL_PDIV_A(x) ((x) << 0) +# define UPLL_PDIV_A_MASK 0x0000007F +# define UPLL_PDIV_B(x) ((x) << 8) +# define UPLL_PDIV_B_MASK 0x00007F00 +# define VCLK_SRC_SEL(x) ((x) << 20) +# define VCLK_SRC_SEL_MASK 0x01F00000 +# define DCLK_SRC_SEL(x) ((x) << 25) +# define DCLK_SRC_SEL_MASK 0x3E000000 +#define CG_UPLL_FUNC_CNTL_3 0x18f +# define UPLL_FB_DIV(x) ((x) << 0) +# define UPLL_FB_DIV_MASK 0x01FFFFFF +#define CG_UPLL_FUNC_CNTL_4 0x191 +# define UPLL_SPARE_ISPARE9 0x00020000 +#define CG_UPLL_FUNC_CNTL_5 0x192 +# define RESET_ANTI_MUX_MASK 0x00000200 +#define CG_UPLL_SPREAD_SPECTRUM 0x194 +# define SSEN_MASK 0x00000001 + +#define MPLL_BYPASSCLK_SEL 0x197 +# define MPLL_CLKOUT_SEL(x) ((x) << 8) +# define MPLL_CLKOUT_SEL_MASK 0xFF00 + +#define CG_CLKPIN_CNTL 0x198 +# define XTALIN_DIVIDE (1 << 1) +# define BCLK_AS_XCLK (1 << 2) +#define CG_CLKPIN_CNTL_2 0x199 +# define FORCE_BIF_REFCLK_EN (1 << 3) +# define MUX_TCLK_TO_XCLK (1 << 8) + +#define THM_CLK_CNTL 0x19b +# define CMON_CLK_SEL(x) ((x) << 0) +# define CMON_CLK_SEL_MASK 0xFF +# define TMON_CLK_SEL(x) ((x) << 8) +# define TMON_CLK_SEL_MASK 0xFF00 +#define MISC_CLK_CNTL 0x19c +# define DEEP_SLEEP_CLK_SEL(x) ((x) << 0) +# define DEEP_SLEEP_CLK_SEL_MASK 0xFF +# define ZCLK_SEL(x) ((x) << 8) +# define ZCLK_SEL_MASK 0xFF00 + +#define CG_THERMAL_CTRL 0x1c0 +#define DPM_EVENT_SRC(x) ((x) << 0) +#define DPM_EVENT_SRC_MASK (7 << 0) +#define DIG_THERM_DPM(x) ((x) << 14) +#define DIG_THERM_DPM_MASK 0x003FC000 +#define DIG_THERM_DPM_SHIFT 14 +#define CG_THERMAL_STATUS 0x1c1 +#define FDO_PWM_DUTY(x) ((x) << 9) +#define FDO_PWM_DUTY_MASK (0xff << 9) +#define FDO_PWM_DUTY_SHIFT 9 +#define CG_THERMAL_INT 0x1c2 +#define DIG_THERM_INTH(x) ((x) << 8) +#define DIG_THERM_INTH_MASK 0x0000FF00 +#define DIG_THERM_INTH_SHIFT 8 +#define DIG_THERM_INTL(x) ((x) << 16) +#define DIG_THERM_INTL_MASK 0x00FF0000 +#define DIG_THERM_INTL_SHIFT 16 +#define THERM_INT_MASK_HIGH (1 << 24) +#define THERM_INT_MASK_LOW (1 << 25) + +#define CG_MULT_THERMAL_CTRL 0x1c4 +#define TEMP_SEL(x) ((x) << 20) +#define TEMP_SEL_MASK (0xff << 20) +#define TEMP_SEL_SHIFT 20 +#define CG_MULT_THERMAL_STATUS 0x1c5 +#define ASIC_MAX_TEMP(x) ((x) << 0) +#define ASIC_MAX_TEMP_MASK 0x000001ff +#define ASIC_MAX_TEMP_SHIFT 0 +#define CTF_TEMP(x) ((x) << 9) +#define CTF_TEMP_MASK 0x0003fe00 +#define CTF_TEMP_SHIFT 9 + +#define CG_FDO_CTRL0 0x1d5 +#define FDO_STATIC_DUTY(x) ((x) << 0) +#define FDO_STATIC_DUTY_MASK 0x000000FF +#define FDO_STATIC_DUTY_SHIFT 0 +#define CG_FDO_CTRL1 0x1d6 +#define FMAX_DUTY100(x) ((x) << 0) +#define FMAX_DUTY100_MASK 0x000000FF +#define FMAX_DUTY100_SHIFT 0 +#define CG_FDO_CTRL2 0x1d7 +#define TMIN(x) ((x) << 0) +#define TMIN_MASK 0x000000FF +#define TMIN_SHIFT 0 +#define FDO_PWM_MODE(x) ((x) << 11) +#define FDO_PWM_MODE_MASK (7 << 11) +#define FDO_PWM_MODE_SHIFT 11 +#define TACH_PWM_RESP_RATE(x) ((x) << 25) +#define TACH_PWM_RESP_RATE_MASK (0x7f << 25) +#define TACH_PWM_RESP_RATE_SHIFT 25 + +#define CG_TACH_CTRL 0x1dc +# define EDGE_PER_REV(x) ((x) << 0) +# define EDGE_PER_REV_MASK (0x7 << 0) +# define EDGE_PER_REV_SHIFT 0 +# define TARGET_PERIOD(x) ((x) << 3) +# define TARGET_PERIOD_MASK 0xfffffff8 +# define TARGET_PERIOD_SHIFT 3 +#define CG_TACH_STATUS 0x1dd +# define TACH_PERIOD(x) ((x) << 0) +# define TACH_PERIOD_MASK 0xffffffff +# define TACH_PERIOD_SHIFT 0 + +#define GENERAL_PWRMGT 0x1e0 +# define GLOBAL_PWRMGT_EN (1 << 0) +# define STATIC_PM_EN (1 << 1) +# define THERMAL_PROTECTION_DIS (1 << 2) +# define THERMAL_PROTECTION_TYPE (1 << 3) +# define SW_SMIO_INDEX(x) ((x) << 6) +# define SW_SMIO_INDEX_MASK (1 << 6) +# define SW_SMIO_INDEX_SHIFT 6 +# define VOLT_PWRMGT_EN (1 << 10) +# define DYN_SPREAD_SPECTRUM_EN (1 << 23) +#define CG_TPC 0x1e1 +#define SCLK_PWRMGT_CNTL 0x1e2 +# define SCLK_PWRMGT_OFF (1 << 0) +# define SCLK_LOW_D1 (1 << 1) +# define FIR_RESET (1 << 4) +# define FIR_FORCE_TREND_SEL (1 << 5) +# define FIR_TREND_MODE (1 << 6) +# define DYN_GFX_CLK_OFF_EN (1 << 7) +# define GFX_CLK_FORCE_ON (1 << 8) +# define GFX_CLK_REQUEST_OFF (1 << 9) +# define GFX_CLK_FORCE_OFF (1 << 10) +# define GFX_CLK_OFF_ACPI_D1 (1 << 11) +# define GFX_CLK_OFF_ACPI_D2 (1 << 12) +# define GFX_CLK_OFF_ACPI_D3 (1 << 13) +# define DYN_LIGHT_SLEEP_EN (1 << 14) + +#define TARGET_AND_CURRENT_PROFILE_INDEX 0x1e6 +# define CURRENT_STATE_INDEX_MASK (0xf << 4) +# define CURRENT_STATE_INDEX_SHIFT 4 + +#define CG_FTV 0x1ef + +#define CG_FFCT_0 0x1f0 +# define UTC_0(x) ((x) << 0) +# define UTC_0_MASK (0x3ff << 0) +# define DTC_0(x) ((x) << 10) +# define DTC_0_MASK (0x3ff << 10) + +#define CG_BSP 0x1ff +# define BSP(x) ((x) << 0) +# define BSP_MASK (0xffff << 0) +# define BSU(x) ((x) << 16) +# define BSU_MASK (0xf << 16) +#define CG_AT 0x200 +# define CG_R(x) ((x) << 0) +# define CG_R_MASK (0xffff << 0) +# define CG_L(x) ((x) << 16) +# define CG_L_MASK (0xffff << 16) + +#define CG_GIT 0x201 +# define CG_GICST(x) ((x) << 0) +# define CG_GICST_MASK (0xffff << 0) +# define CG_GIPOT(x) ((x) << 16) +# define CG_GIPOT_MASK (0xffff << 16) + +#define CG_SSP 0x203 +# define SST(x) ((x) << 0) +# define SST_MASK (0xffff << 0) +# define SSTU(x) ((x) << 16) +# define SSTU_MASK (0xf << 16) + +#define CG_DISPLAY_GAP_CNTL 0x20a +# define DISP1_GAP(x) ((x) << 0) +# define DISP1_GAP_MASK (3 << 0) +# define DISP2_GAP(x) ((x) << 2) +# define DISP2_GAP_MASK (3 << 2) +# define VBI_TIMER_COUNT(x) ((x) << 4) +# define VBI_TIMER_COUNT_MASK (0x3fff << 4) +# define VBI_TIMER_UNIT(x) ((x) << 20) +# define VBI_TIMER_UNIT_MASK (7 << 20) +# define DISP1_GAP_MCHG(x) ((x) << 24) +# define DISP1_GAP_MCHG_MASK (3 << 24) +# define DISP2_GAP_MCHG(x) ((x) << 26) +# define DISP2_GAP_MCHG_MASK (3 << 26) + +#define CG_ULV_CONTROL 0x21e +#define CG_ULV_PARAMETER 0x21f + +#define SMC_SCRATCH0 0x221 + +#define CG_CAC_CTRL 0x22e +# define CAC_WINDOW(x) ((x) << 0) +# define CAC_WINDOW_MASK 0x00ffffff + +#define DMIF_ADDR_CONFIG 0x2F5 + +#define DMIF_ADDR_CALC 0x300 + +#define PIPE0_DMIF_BUFFER_CONTROL 0x0328 +# define DMIF_BUFFERS_ALLOCATED(x) ((x) << 0) +# define DMIF_BUFFERS_ALLOCATED_COMPLETED (1 << 4) + +#define SRBM_STATUS 0x394 +#define GRBM_RQ_PENDING (1 << 5) +#define VMC_BUSY (1 << 8) +#define MCB_BUSY (1 << 9) +#define MCB_NON_DISPLAY_BUSY (1 << 10) +#define MCC_BUSY (1 << 11) +#define MCD_BUSY (1 << 12) +#define SEM_BUSY (1 << 14) +#define IH_BUSY (1 << 17) + +#define SRBM_SOFT_RESET 0x398 +#define SOFT_RESET_BIF (1 << 1) +#define SOFT_RESET_DC (1 << 5) +#define SOFT_RESET_DMA1 (1 << 6) +#define SOFT_RESET_GRBM (1 << 8) +#define SOFT_RESET_HDP (1 << 9) +#define SOFT_RESET_IH (1 << 10) +#define SOFT_RESET_MC (1 << 11) +#define SOFT_RESET_ROM (1 << 14) +#define SOFT_RESET_SEM (1 << 15) +#define SOFT_RESET_VMC (1 << 17) +#define SOFT_RESET_DMA (1 << 20) +#define SOFT_RESET_TST (1 << 21) +#define SOFT_RESET_REGBB (1 << 22) +#define SOFT_RESET_ORB (1 << 23) + +#define CC_SYS_RB_BACKEND_DISABLE 0x3A0 +#define GC_USER_SYS_RB_BACKEND_DISABLE 0x3A1 + +#define SRBM_READ_ERROR 0x3A6 +#define SRBM_INT_CNTL 0x3A8 +#define SRBM_INT_ACK 0x3AA + +#define SRBM_STATUS2 0x3B1 +#define DMA_BUSY (1 << 5) +#define DMA1_BUSY (1 << 6) + +#define VM_L2_CNTL 0x500 +#define ENABLE_L2_CACHE (1 << 0) +#define ENABLE_L2_FRAGMENT_PROCESSING (1 << 1) +#define L2_CACHE_PTE_ENDIAN_SWAP_MODE(x) ((x) << 2) +#define L2_CACHE_PDE_ENDIAN_SWAP_MODE(x) ((x) << 4) +#define ENABLE_L2_PTE_CACHE_LRU_UPDATE_BY_WRITE (1 << 9) +#define ENABLE_L2_PDE0_CACHE_LRU_UPDATE_BY_WRITE (1 << 10) +#define EFFECTIVE_L2_QUEUE_SIZE(x) (((x) & 7) << 15) +#define CONTEXT1_IDENTITY_ACCESS_MODE(x) (((x) & 3) << 19) +#define VM_L2_CNTL2 0x501 +#define INVALIDATE_ALL_L1_TLBS (1 << 0) +#define INVALIDATE_L2_CACHE (1 << 1) +#define INVALIDATE_CACHE_MODE(x) ((x) << 26) +#define INVALIDATE_PTE_AND_PDE_CACHES 0 +#define INVALIDATE_ONLY_PTE_CACHES 1 +#define INVALIDATE_ONLY_PDE_CACHES 2 +#define VM_L2_CNTL3 0x502 +#define BANK_SELECT(x) ((x) << 0) +#define L2_CACHE_UPDATE_MODE(x) ((x) << 6) +#define L2_CACHE_BIGK_FRAGMENT_SIZE(x) ((x) << 15) +#define L2_CACHE_BIGK_ASSOCIATIVITY (1 << 20) +#define VM_L2_STATUS 0x503 +#define L2_BUSY (1 << 0) +#define VM_CONTEXT0_CNTL 0x504 +#define ENABLE_CONTEXT (1 << 0) +#define PAGE_TABLE_DEPTH(x) (((x) & 3) << 1) +#define RANGE_PROTECTION_FAULT_ENABLE_INTERRUPT (1 << 3) +#define RANGE_PROTECTION_FAULT_ENABLE_DEFAULT (1 << 4) +#define DUMMY_PAGE_PROTECTION_FAULT_ENABLE_INTERRUPT (1 << 6) +#define DUMMY_PAGE_PROTECTION_FAULT_ENABLE_DEFAULT (1 << 7) +#define PDE0_PROTECTION_FAULT_ENABLE_INTERRUPT (1 << 9) +#define PDE0_PROTECTION_FAULT_ENABLE_DEFAULT (1 << 10) +#define VALID_PROTECTION_FAULT_ENABLE_INTERRUPT (1 << 12) +#define VALID_PROTECTION_FAULT_ENABLE_DEFAULT (1 << 13) +#define READ_PROTECTION_FAULT_ENABLE_INTERRUPT (1 << 15) +#define READ_PROTECTION_FAULT_ENABLE_DEFAULT (1 << 16) +#define WRITE_PROTECTION_FAULT_ENABLE_INTERRUPT (1 << 18) +#define WRITE_PROTECTION_FAULT_ENABLE_DEFAULT (1 << 19) +#define PAGE_TABLE_BLOCK_SIZE(x) (((x) & 0xF) << 24) +#define VM_CONTEXT1_CNTL 0x505 +#define VM_CONTEXT0_CNTL2 0x50C +#define VM_CONTEXT1_CNTL2 0x50D +#define VM_CONTEXT8_PAGE_TABLE_BASE_ADDR 0x50E +#define VM_CONTEXT9_PAGE_TABLE_BASE_ADDR 0x50F +#define VM_CONTEXT10_PAGE_TABLE_BASE_ADDR 0x510 +#define VM_CONTEXT11_PAGE_TABLE_BASE_ADDR 0x511 +#define VM_CONTEXT12_PAGE_TABLE_BASE_ADDR 0x512 +#define VM_CONTEXT13_PAGE_TABLE_BASE_ADDR 0x513 +#define VM_CONTEXT14_PAGE_TABLE_BASE_ADDR 0x514 +#define VM_CONTEXT15_PAGE_TABLE_BASE_ADDR 0x515 + +#define VM_CONTEXT1_PROTECTION_FAULT_ADDR 0x53f +#define VM_CONTEXT1_PROTECTION_FAULT_STATUS 0x537 +#define PROTECTIONS_MASK (0xf << 0) +#define PROTECTIONS_SHIFT 0 + /* bit 0: range + * bit 1: pde0 + * bit 2: valid + * bit 3: read + * bit 4: write + */ +#define MEMORY_CLIENT_ID_MASK (0xff << 12) +#define MEMORY_CLIENT_ID_SHIFT 12 +#define MEMORY_CLIENT_RW_MASK (1 << 24) +#define MEMORY_CLIENT_RW_SHIFT 24 +#define FAULT_VMID_MASK (0xf << 25) +#define FAULT_VMID_SHIFT 25 + +#define VM_INVALIDATE_REQUEST 0x51E +#define VM_INVALIDATE_RESPONSE 0x51F + +#define VM_CONTEXT0_PROTECTION_FAULT_DEFAULT_ADDR 0x546 +#define VM_CONTEXT1_PROTECTION_FAULT_DEFAULT_ADDR 0x547 + +#define VM_CONTEXT0_PAGE_TABLE_BASE_ADDR 0x54F +#define VM_CONTEXT1_PAGE_TABLE_BASE_ADDR 0x550 +#define VM_CONTEXT2_PAGE_TABLE_BASE_ADDR 0x551 +#define VM_CONTEXT3_PAGE_TABLE_BASE_ADDR 0x552 +#define VM_CONTEXT4_PAGE_TABLE_BASE_ADDR 0x553 +#define VM_CONTEXT5_PAGE_TABLE_BASE_ADDR 0x554 +#define VM_CONTEXT6_PAGE_TABLE_BASE_ADDR 0x555 +#define VM_CONTEXT7_PAGE_TABLE_BASE_ADDR 0x556 +#define VM_CONTEXT0_PAGE_TABLE_START_ADDR 0x557 +#define VM_CONTEXT1_PAGE_TABLE_START_ADDR 0x558 + +#define VM_CONTEXT0_PAGE_TABLE_END_ADDR 0x55F +#define VM_CONTEXT1_PAGE_TABLE_END_ADDR 0x560 + +#define VM_L2_CG 0x570 +#define MC_CG_ENABLE (1 << 18) +#define MC_LS_ENABLE (1 << 19) + +#define MC_SHARED_CHMAP 0x801 +#define NOOFCHAN_SHIFT 12 +#define NOOFCHAN_MASK 0x0000f000 +#define MC_SHARED_CHREMAP 0x802 + +#define MC_VM_FB_LOCATION 0x809 +#define MC_VM_AGP_TOP 0x80A +#define MC_VM_AGP_BOT 0x80B +#define MC_VM_AGP_BASE 0x80C +#define MC_VM_SYSTEM_APERTURE_LOW_ADDR 0x80D +#define MC_VM_SYSTEM_APERTURE_HIGH_ADDR 0x80E +#define MC_VM_SYSTEM_APERTURE_DEFAULT_ADDR 0x80F + +#define MC_VM_MX_L1_TLB_CNTL 0x819 +#define ENABLE_L1_TLB (1 << 0) +#define ENABLE_L1_FRAGMENT_PROCESSING (1 << 1) +#define SYSTEM_ACCESS_MODE_PA_ONLY (0 << 3) +#define SYSTEM_ACCESS_MODE_USE_SYS_MAP (1 << 3) +#define SYSTEM_ACCESS_MODE_IN_SYS (2 << 3) +#define SYSTEM_ACCESS_MODE_NOT_IN_SYS (3 << 3) +#define SYSTEM_APERTURE_UNMAPPED_ACCESS_PASS_THRU (0 << 5) +#define ENABLE_ADVANCED_DRIVER_MODEL (1 << 6) + +#define MC_SHARED_BLACKOUT_CNTL 0x82B + +#define MC_HUB_MISC_HUB_CG 0x82E +#define MC_HUB_MISC_VM_CG 0x82F + +#define MC_HUB_MISC_SIP_CG 0x830 + +#define MC_XPB_CLK_GAT 0x91E + +#define MC_CITF_MISC_RD_CG 0x992 +#define MC_CITF_MISC_WR_CG 0x993 +#define MC_CITF_MISC_VM_CG 0x994 + +#define MC_ARB_RAMCFG 0x9D8 +#define NOOFBANK_SHIFT 0 +#define NOOFBANK_MASK 0x00000003 +#define NOOFRANK_SHIFT 2 +#define NOOFRANK_MASK 0x00000004 +#define NOOFROWS_SHIFT 3 +#define NOOFROWS_MASK 0x00000038 +#define NOOFCOLS_SHIFT 6 +#define NOOFCOLS_MASK 0x000000C0 +#define CHANSIZE_SHIFT 8 +#define CHANSIZE_MASK 0x00000100 +#define CHANSIZE_OVERRIDE (1 << 11) +#define NOOFGROUPS_SHIFT 12 +#define NOOFGROUPS_MASK 0x00001000 + +#define MC_ARB_DRAM_TIMING 0x9DD +#define MC_ARB_DRAM_TIMING2 0x9DE + +#define MC_ARB_BURST_TIME 0xA02 +#define STATE0(x) ((x) << 0) +#define STATE0_MASK (0x1f << 0) +#define STATE0_SHIFT 0 +#define STATE1(x) ((x) << 5) +#define STATE1_MASK (0x1f << 5) +#define STATE1_SHIFT 5 +#define STATE2(x) ((x) << 10) +#define STATE2_MASK (0x1f << 10) +#define STATE2_SHIFT 10 +#define STATE3(x) ((x) << 15) +#define STATE3_MASK (0x1f << 15) +#define STATE3_SHIFT 15 + +#define MC_SEQ_TRAIN_WAKEUP_CNTL 0xA3A +#define TRAIN_DONE_D0 (1 << 30) +#define TRAIN_DONE_D1 (1 << 31) + +#define MC_SEQ_SUP_CNTL 0xA32 +#define RUN_MASK (1 << 0) +#define MC_SEQ_SUP_PGM 0xA33 +#define MC_PMG_AUTO_CMD 0xA34 + +#define MC_IO_PAD_CNTL_D0 0xA74 +#define MEM_FALL_OUT_CMD (1 << 8) + +#define MC_SEQ_RAS_TIMING 0xA28 +#define MC_SEQ_CAS_TIMING 0xA29 +#define MC_SEQ_MISC_TIMING 0xA2A +#define MC_SEQ_MISC_TIMING2 0xA2B +#define MC_SEQ_PMG_TIMING 0xA2C +#define MC_SEQ_RD_CTL_D0 0xA2D +#define MC_SEQ_RD_CTL_D1 0xA2E +#define MC_SEQ_WR_CTL_D0 0xA2F +#define MC_SEQ_WR_CTL_D1 0xA30 + +#define MC_SEQ_MISC0 0xA80 +#define MC_SEQ_MISC0_VEN_ID_SHIFT 8 +#define MC_SEQ_MISC0_VEN_ID_MASK 0x00000f00 +#define MC_SEQ_MISC0_VEN_ID_VALUE 3 +#define MC_SEQ_MISC0_REV_ID_SHIFT 12 +#define MC_SEQ_MISC0_REV_ID_MASK 0x0000f000 +#define MC_SEQ_MISC0_REV_ID_VALUE 1 +#define MC_SEQ_MISC0_GDDR5_SHIFT 28 +#define MC_SEQ_MISC0_GDDR5_MASK 0xf0000000 +#define MC_SEQ_MISC0_GDDR5_VALUE 5 +#define MC_SEQ_MISC1 0xA81 +#define MC_SEQ_RESERVE_M 0xA82 +#define MC_PMG_CMD_EMRS 0xA83 + +#define MC_SEQ_IO_DEBUG_INDEX 0xA91 +#define MC_SEQ_IO_DEBUG_DATA 0xA92 + +#define MC_SEQ_MISC5 0xA95 +#define MC_SEQ_MISC6 0xA96 + +#define MC_SEQ_MISC7 0xA99 + +#define MC_SEQ_RAS_TIMING_LP 0xA9B +#define MC_SEQ_CAS_TIMING_LP 0xA9C +#define MC_SEQ_MISC_TIMING_LP 0xA9D +#define MC_SEQ_MISC_TIMING2_LP 0xA9E +#define MC_SEQ_WR_CTL_D0_LP 0xA9F +#define MC_SEQ_WR_CTL_D1_LP 0xAA0 +#define MC_SEQ_PMG_CMD_EMRS_LP 0xAA1 +#define MC_SEQ_PMG_CMD_MRS_LP 0xAA2 + +#define MC_PMG_CMD_MRS 0xAAB + +#define MC_SEQ_RD_CTL_D0_LP 0xAC7 +#define MC_SEQ_RD_CTL_D1_LP 0xAC8 + +#define MC_PMG_CMD_MRS1 0xAD1 +#define MC_SEQ_PMG_CMD_MRS1_LP 0xAD2 +#define MC_SEQ_PMG_TIMING_LP 0xAD3 + +#define MC_SEQ_WR_CTL_2 0xAD5 +#define MC_SEQ_WR_CTL_2_LP 0xAD6 +#define MC_PMG_CMD_MRS2 0xAD7 +#define MC_SEQ_PMG_CMD_MRS2_LP 0xAD8 + +#define MCLK_PWRMGT_CNTL 0xAE8 +# define DLL_SPEED(x) ((x) << 0) +# define DLL_SPEED_MASK (0x1f << 0) +# define DLL_READY (1 << 6) +# define MC_INT_CNTL (1 << 7) +# define MRDCK0_PDNB (1 << 8) +# define MRDCK1_PDNB (1 << 9) +# define MRDCK0_RESET (1 << 16) +# define MRDCK1_RESET (1 << 17) +# define DLL_READY_READ (1 << 24) +#define DLL_CNTL 0xAE9 +# define MRDCK0_BYPASS (1 << 24) +# define MRDCK1_BYPASS (1 << 25) + +#define MPLL_CNTL_MODE 0xAEC +# define MPLL_MCLK_SEL (1 << 11) +#define MPLL_FUNC_CNTL 0xAED +#define BWCTRL(x) ((x) << 20) +#define BWCTRL_MASK (0xff << 20) +#define MPLL_FUNC_CNTL_1 0xAEE +#define VCO_MODE(x) ((x) << 0) +#define VCO_MODE_MASK (3 << 0) +#define CLKFRAC(x) ((x) << 4) +#define CLKFRAC_MASK (0xfff << 4) +#define CLKF(x) ((x) << 16) +#define CLKF_MASK (0xfff << 16) +#define MPLL_FUNC_CNTL_2 0xAEF +#define MPLL_AD_FUNC_CNTL 0xAF0 +#define YCLK_POST_DIV(x) ((x) << 0) +#define YCLK_POST_DIV_MASK (7 << 0) +#define MPLL_DQ_FUNC_CNTL 0xAF1 +#define YCLK_SEL(x) ((x) << 4) +#define YCLK_SEL_MASK (1 << 4) + +#define MPLL_SS1 0xAF3 +#define CLKV(x) ((x) << 0) +#define CLKV_MASK (0x3ffffff << 0) +#define MPLL_SS2 0xAF4 +#define CLKS(x) ((x) << 0) +#define CLKS_MASK (0xfff << 0) + +#define HDP_HOST_PATH_CNTL 0xB00 +#define CLOCK_GATING_DIS (1 << 23) +#define HDP_NONSURFACE_BASE 0xB01 +#define HDP_NONSURFACE_INFO 0xB02 +#define HDP_NONSURFACE_SIZE 0xB03 + +#define HDP_DEBUG0 0xBCC + +#define HDP_ADDR_CONFIG 0xBD2 +#define HDP_MISC_CNTL 0xBD3 +#define HDP_FLUSH_INVALIDATE_CACHE (1 << 0) +#define HDP_MEM_POWER_LS 0xBD4 +#define HDP_LS_ENABLE (1 << 0) + +#define ATC_MISC_CG 0xCD4 + +#define IH_RB_CNTL 0xF80 +# define IH_RB_ENABLE (1 << 0) +# define IH_IB_SIZE(x) ((x) << 1) /* log2 */ +# define IH_RB_FULL_DRAIN_ENABLE (1 << 6) +# define IH_WPTR_WRITEBACK_ENABLE (1 << 8) +# define IH_WPTR_WRITEBACK_TIMER(x) ((x) << 9) /* log2 */ +# define IH_WPTR_OVERFLOW_ENABLE (1 << 16) +# define IH_WPTR_OVERFLOW_CLEAR (1 << 31) +#define IH_RB_BASE 0xF81 +#define IH_RB_RPTR 0xF82 +#define IH_RB_WPTR 0xF83 +# define RB_OVERFLOW (1 << 0) +# define WPTR_OFFSET_MASK 0x3fffc +#define IH_RB_WPTR_ADDR_HI 0xF84 +#define IH_RB_WPTR_ADDR_LO 0xF85 +#define IH_CNTL 0xF86 +# define ENABLE_INTR (1 << 0) +# define IH_MC_SWAP(x) ((x) << 1) +# define IH_MC_SWAP_NONE 0 +# define IH_MC_SWAP_16BIT 1 +# define IH_MC_SWAP_32BIT 2 +# define IH_MC_SWAP_64BIT 3 +# define RPTR_REARM (1 << 4) +# define MC_WRREQ_CREDIT(x) ((x) << 15) +# define MC_WR_CLEAN_CNT(x) ((x) << 20) +# define MC_VMID(x) ((x) << 25) + +#define CONFIG_MEMSIZE 0x150A + +#define INTERRUPT_CNTL 0x151A +# define IH_DUMMY_RD_OVERRIDE (1 << 0) +# define IH_DUMMY_RD_EN (1 << 1) +# define IH_REQ_NONSNOOP_EN (1 << 3) +# define GEN_IH_INT_EN (1 << 8) +#define INTERRUPT_CNTL2 0x151B + +#define HDP_MEM_COHERENCY_FLUSH_CNTL 0x1520 + +#define BIF_FB_EN 0x1524 +#define FB_READ_EN (1 << 0) +#define FB_WRITE_EN (1 << 1) + +#define HDP_REG_COHERENCY_FLUSH_CNTL 0x1528 + +/* DCE6 ELD audio interface */ +#define AZ_F0_CODEC_ENDPOINT_INDEX 0x1780 +# define AZ_ENDPOINT_REG_INDEX(x) (((x) & 0xff) << 0) +# define AZ_ENDPOINT_REG_WRITE_EN (1 << 8) +#define AZ_F0_CODEC_ENDPOINT_DATA 0x1781 + +#define AZ_F0_CODEC_PIN_CONTROL_CHANNEL_SPEAKER 0x25 +#define SPEAKER_ALLOCATION(x) (((x) & 0x7f) << 0) +#define SPEAKER_ALLOCATION_MASK (0x7f << 0) +#define SPEAKER_ALLOCATION_SHIFT 0 +#define HDMI_CONNECTION (1 << 16) +#define DP_CONNECTION (1 << 17) + +#define AZ_F0_CODEC_PIN_CONTROL_AUDIO_DESCRIPTOR0 0x28 /* LPCM */ +#define AZ_F0_CODEC_PIN_CONTROL_AUDIO_DESCRIPTOR1 0x29 /* AC3 */ +#define AZ_F0_CODEC_PIN_CONTROL_AUDIO_DESCRIPTOR2 0x2A /* MPEG1 */ +#define AZ_F0_CODEC_PIN_CONTROL_AUDIO_DESCRIPTOR3 0x2B /* MP3 */ +#define AZ_F0_CODEC_PIN_CONTROL_AUDIO_DESCRIPTOR4 0x2C /* MPEG2 */ +#define AZ_F0_CODEC_PIN_CONTROL_AUDIO_DESCRIPTOR5 0x2D /* AAC */ +#define AZ_F0_CODEC_PIN_CONTROL_AUDIO_DESCRIPTOR6 0x2E /* DTS */ +#define AZ_F0_CODEC_PIN_CONTROL_AUDIO_DESCRIPTOR7 0x2F /* ATRAC */ +#define AZ_F0_CODEC_PIN_CONTROL_AUDIO_DESCRIPTOR8 0x30 /* one bit audio - leave at 0 (default) */ +#define AZ_F0_CODEC_PIN_CONTROL_AUDIO_DESCRIPTOR9 0x31 /* Dolby Digital */ +#define AZ_F0_CODEC_PIN_CONTROL_AUDIO_DESCRIPTOR10 0x32 /* DTS-HD */ +#define AZ_F0_CODEC_PIN_CONTROL_AUDIO_DESCRIPTOR11 0x33 /* MAT-MLP */ +#define AZ_F0_CODEC_PIN_CONTROL_AUDIO_DESCRIPTOR12 0x34 /* DTS */ +#define AZ_F0_CODEC_PIN_CONTROL_AUDIO_DESCRIPTOR13 0x35 /* WMA Pro */ +# define MAX_CHANNELS(x) (((x) & 0x7) << 0) +/* max channels minus one. 7 = 8 channels */ +# define SUPPORTED_FREQUENCIES(x) (((x) & 0xff) << 8) +# define DESCRIPTOR_BYTE_2(x) (((x) & 0xff) << 16) +# define SUPPORTED_FREQUENCIES_STEREO(x) (((x) & 0xff) << 24) /* LPCM only */ +/* SUPPORTED_FREQUENCIES, SUPPORTED_FREQUENCIES_STEREO + * bit0 = 32 kHz + * bit1 = 44.1 kHz + * bit2 = 48 kHz + * bit3 = 88.2 kHz + * bit4 = 96 kHz + * bit5 = 176.4 kHz + * bit6 = 192 kHz + */ + +#define AZ_F0_CODEC_PIN_CONTROL_RESPONSE_LIPSYNC 0x37 +# define VIDEO_LIPSYNC(x) (((x) & 0xff) << 0) +# define AUDIO_LIPSYNC(x) (((x) & 0xff) << 8) +/* VIDEO_LIPSYNC, AUDIO_LIPSYNC + * 0 = invalid + * x = legal delay value + * 255 = sync not supported + */ +#define AZ_F0_CODEC_PIN_CONTROL_RESPONSE_HBR 0x38 +# define HBR_CAPABLE (1 << 0) /* enabled by default */ + +#define AZ_F0_CODEC_PIN_CONTROL_SINK_INFO0 0x3a +# define MANUFACTURER_ID(x) (((x) & 0xffff) << 0) +# define PRODUCT_ID(x) (((x) & 0xffff) << 16) +#define AZ_F0_CODEC_PIN_CONTROL_SINK_INFO1 0x3b +# define SINK_DESCRIPTION_LEN(x) (((x) & 0xff) << 0) +#define AZ_F0_CODEC_PIN_CONTROL_SINK_INFO2 0x3c +# define PORT_ID0(x) (((x) & 0xffffffff) << 0) +#define AZ_F0_CODEC_PIN_CONTROL_SINK_INFO3 0x3d +# define PORT_ID1(x) (((x) & 0xffffffff) << 0) +#define AZ_F0_CODEC_PIN_CONTROL_SINK_INFO4 0x3e +# define DESCRIPTION0(x) (((x) & 0xff) << 0) +# define DESCRIPTION1(x) (((x) & 0xff) << 8) +# define DESCRIPTION2(x) (((x) & 0xff) << 16) +# define DESCRIPTION3(x) (((x) & 0xff) << 24) +#define AZ_F0_CODEC_PIN_CONTROL_SINK_INFO5 0x3f +# define DESCRIPTION4(x) (((x) & 0xff) << 0) +# define DESCRIPTION5(x) (((x) & 0xff) << 8) +# define DESCRIPTION6(x) (((x) & 0xff) << 16) +# define DESCRIPTION7(x) (((x) & 0xff) << 24) +#define AZ_F0_CODEC_PIN_CONTROL_SINK_INFO6 0x40 +# define DESCRIPTION8(x) (((x) & 0xff) << 0) +# define DESCRIPTION9(x) (((x) & 0xff) << 8) +# define DESCRIPTION10(x) (((x) & 0xff) << 16) +# define DESCRIPTION11(x) (((x) & 0xff) << 24) +#define AZ_F0_CODEC_PIN_CONTROL_SINK_INFO7 0x41 +# define DESCRIPTION12(x) (((x) & 0xff) << 0) +# define DESCRIPTION13(x) (((x) & 0xff) << 8) +# define DESCRIPTION14(x) (((x) & 0xff) << 16) +# define DESCRIPTION15(x) (((x) & 0xff) << 24) +#define AZ_F0_CODEC_PIN_CONTROL_SINK_INFO8 0x42 +# define DESCRIPTION16(x) (((x) & 0xff) << 0) +# define DESCRIPTION17(x) (((x) & 0xff) << 8) + +#define AZ_F0_CODEC_PIN_CONTROL_HOT_PLUG_CONTROL 0x54 +# define AUDIO_ENABLED (1 << 31) + +#define AZ_F0_CODEC_PIN_CONTROL_RESPONSE_CONFIGURATION_DEFAULT 0x56 +#define PORT_CONNECTIVITY_MASK (3 << 30) +#define PORT_CONNECTIVITY_SHIFT 30 + +#define DC_LB_MEMORY_SPLIT 0x1AC3 +#define DC_LB_MEMORY_CONFIG(x) ((x) << 20) + +#define PRIORITY_A_CNT 0x1AC6 +#define PRIORITY_MARK_MASK 0x7fff +#define PRIORITY_OFF (1 << 16) +#define PRIORITY_ALWAYS_ON (1 << 20) +#define PRIORITY_B_CNT 0x1AC7 + +#define DPG_PIPE_ARBITRATION_CONTROL3 0x1B32 +# define LATENCY_WATERMARK_MASK(x) ((x) << 16) +#define DPG_PIPE_LATENCY_CONTROL 0x1B33 +# define LATENCY_LOW_WATERMARK(x) ((x) << 0) +# define LATENCY_HIGH_WATERMARK(x) ((x) << 16) + +/* 0x6bb8, 0x77b8, 0x103b8, 0x10fb8, 0x11bb8, 0x127b8 */ +#define VLINE_STATUS 0x1AEE +# define VLINE_OCCURRED (1 << 0) +# define VLINE_ACK (1 << 4) +# define VLINE_STAT (1 << 12) +# define VLINE_INTERRUPT (1 << 16) +# define VLINE_INTERRUPT_TYPE (1 << 17) +/* 0x6bbc, 0x77bc, 0x103bc, 0x10fbc, 0x11bbc, 0x127bc */ +#define VBLANK_STATUS 0x1AEF +# define VBLANK_OCCURRED (1 << 0) +# define VBLANK_ACK (1 << 4) +# define VBLANK_STAT (1 << 12) +# define VBLANK_INTERRUPT (1 << 16) +# define VBLANK_INTERRUPT_TYPE (1 << 17) + +/* 0x6b40, 0x7740, 0x10340, 0x10f40, 0x11b40, 0x12740 */ +#define INT_MASK 0x1AD0 +# define VBLANK_INT_MASK (1 << 0) +# define VLINE_INT_MASK (1 << 4) + +#define DISP_INTERRUPT_STATUS 0x183D +# define LB_D1_VLINE_INTERRUPT (1 << 2) +# define LB_D1_VBLANK_INTERRUPT (1 << 3) +# define DC_HPD1_INTERRUPT (1 << 17) +# define DC_HPD1_RX_INTERRUPT (1 << 18) +# define DACA_AUTODETECT_INTERRUPT (1 << 22) +# define DACB_AUTODETECT_INTERRUPT (1 << 23) +# define DC_I2C_SW_DONE_INTERRUPT (1 << 24) +# define DC_I2C_HW_DONE_INTERRUPT (1 << 25) +#define DISP_INTERRUPT_STATUS_CONTINUE 0x183E +# define LB_D2_VLINE_INTERRUPT (1 << 2) +# define LB_D2_VBLANK_INTERRUPT (1 << 3) +# define DC_HPD2_INTERRUPT (1 << 17) +# define DC_HPD2_RX_INTERRUPT (1 << 18) +# define DISP_TIMER_INTERRUPT (1 << 24) +#define DISP_INTERRUPT_STATUS_CONTINUE2 0x183F +# define LB_D3_VLINE_INTERRUPT (1 << 2) +# define LB_D3_VBLANK_INTERRUPT (1 << 3) +# define DC_HPD3_INTERRUPT (1 << 17) +# define DC_HPD3_RX_INTERRUPT (1 << 18) +#define DISP_INTERRUPT_STATUS_CONTINUE3 0x1840 +# define LB_D4_VLINE_INTERRUPT (1 << 2) +# define LB_D4_VBLANK_INTERRUPT (1 << 3) +# define DC_HPD4_INTERRUPT (1 << 17) +# define DC_HPD4_RX_INTERRUPT (1 << 18) +#define DISP_INTERRUPT_STATUS_CONTINUE4 0x1853 +# define LB_D5_VLINE_INTERRUPT (1 << 2) +# define LB_D5_VBLANK_INTERRUPT (1 << 3) +# define DC_HPD5_INTERRUPT (1 << 17) +# define DC_HPD5_RX_INTERRUPT (1 << 18) +#define DISP_INTERRUPT_STATUS_CONTINUE5 0x1854 +# define LB_D6_VLINE_INTERRUPT (1 << 2) +# define LB_D6_VBLANK_INTERRUPT (1 << 3) +# define DC_HPD6_INTERRUPT (1 << 17) +# define DC_HPD6_RX_INTERRUPT (1 << 18) + +/* 0x6858, 0x7458, 0x10058, 0x10c58, 0x11858, 0x12458 */ +#define GRPH_INT_STATUS 0x1A16 +# define GRPH_PFLIP_INT_OCCURRED (1 << 0) +# define GRPH_PFLIP_INT_CLEAR (1 << 8) +/* 0x685c, 0x745c, 0x1005c, 0x10c5c, 0x1185c, 0x1245c */ +#define GRPH_INT_CONTROL 0x1A17 +# define GRPH_PFLIP_INT_MASK (1 << 0) +# define GRPH_PFLIP_INT_TYPE (1 << 8) + +#define DAC_AUTODETECT_INT_CONTROL 0x19F2 + +#define DC_HPD1_INT_STATUS 0x1807 +#define DC_HPD2_INT_STATUS 0x180A +#define DC_HPD3_INT_STATUS 0x180D +#define DC_HPD4_INT_STATUS 0x1810 +#define DC_HPD5_INT_STATUS 0x1813 +#define DC_HPD6_INT_STATUS 0x1816 +# define DC_HPDx_INT_STATUS (1 << 0) +# define DC_HPDx_SENSE (1 << 1) +# define DC_HPDx_RX_INT_STATUS (1 << 8) + +#define DC_HPD1_INT_CONTROL 0x1808 +#define DC_HPD2_INT_CONTROL 0x180B +#define DC_HPD3_INT_CONTROL 0x180E +#define DC_HPD4_INT_CONTROL 0x1811 +#define DC_HPD5_INT_CONTROL 0x1814 +#define DC_HPD6_INT_CONTROL 0x1817 +# define DC_HPDx_INT_ACK (1 << 0) +# define DC_HPDx_INT_POLARITY (1 << 8) +# define DC_HPDx_INT_EN (1 << 16) +# define DC_HPDx_RX_INT_ACK (1 << 20) +# define DC_HPDx_RX_INT_EN (1 << 24) + +#define DC_HPD1_CONTROL 0x1809 +#define DC_HPD2_CONTROL 0x180C +#define DC_HPD3_CONTROL 0x180F +#define DC_HPD4_CONTROL 0x1812 +#define DC_HPD5_CONTROL 0x1815 +#define DC_HPD6_CONTROL 0x1818 +# define DC_HPDx_CONNECTION_TIMER(x) ((x) << 0) +# define DC_HPDx_RX_INT_TIMER(x) ((x) << 16) +# define DC_HPDx_EN (1 << 28) + +#define DPG_PIPE_STUTTER_CONTROL 0x1B35 +# define STUTTER_ENABLE (1 << 0) + +/* 0x6e98, 0x7a98, 0x10698, 0x11298, 0x11e98, 0x12a98 */ +#define CRTC_STATUS_FRAME_COUNT 0x1BA6 + +/* Audio clocks */ +#define DCCG_AUDIO_DTO_SOURCE 0x05ac +# define DCCG_AUDIO_DTO0_SOURCE_SEL(x) ((x) << 0) /* crtc0 - crtc5 */ +# define DCCG_AUDIO_DTO_SEL (1 << 4) /* 0=dto0 1=dto1 */ + +#define DCCG_AUDIO_DTO0_PHASE 0x05b0 +#define DCCG_AUDIO_DTO0_MODULE 0x05b4 +#define DCCG_AUDIO_DTO1_PHASE 0x05c0 +#define DCCG_AUDIO_DTO1_MODULE 0x05c4 + +#define AFMT_AUDIO_SRC_CONTROL 0x1c4f +#define AFMT_AUDIO_SRC_SELECT(x) (((x) & 7) << 0) +/* AFMT_AUDIO_SRC_SELECT + * 0 = stream0 + * 1 = stream1 + * 2 = stream2 + * 3 = stream3 + * 4 = stream4 + * 5 = stream5 + */ + +#define GRBM_CNTL 0x2000 +#define GRBM_READ_TIMEOUT(x) ((x) << 0) + +#define GRBM_STATUS2 0x2002 +#define RLC_RQ_PENDING (1 << 0) +#define RLC_BUSY (1 << 8) +#define TC_BUSY (1 << 9) + +#define GRBM_STATUS 0x2004 +#define CMDFIFO_AVAIL_MASK 0x0000000F +#define RING2_RQ_PENDING (1 << 4) +#define SRBM_RQ_PENDING (1 << 5) +#define RING1_RQ_PENDING (1 << 6) +#define CF_RQ_PENDING (1 << 7) +#define PF_RQ_PENDING (1 << 8) +#define GDS_DMA_RQ_PENDING (1 << 9) +#define GRBM_EE_BUSY (1 << 10) +#define DB_CLEAN (1 << 12) +#define CB_CLEAN (1 << 13) +#define TA_BUSY (1 << 14) +#define GDS_BUSY (1 << 15) +#define VGT_BUSY (1 << 17) +#define IA_BUSY_NO_DMA (1 << 18) +#define IA_BUSY (1 << 19) +#define SX_BUSY (1 << 20) +#define SPI_BUSY (1 << 22) +#define BCI_BUSY (1 << 23) +#define SC_BUSY (1 << 24) +#define PA_BUSY (1 << 25) +#define DB_BUSY (1 << 26) +#define CP_COHERENCY_BUSY (1 << 28) +#define CP_BUSY (1 << 29) +#define CB_BUSY (1 << 30) +#define GUI_ACTIVE (1 << 31) +#define GRBM_STATUS_SE0 0x2005 +#define GRBM_STATUS_SE1 0x2006 +#define SE_DB_CLEAN (1 << 1) +#define SE_CB_CLEAN (1 << 2) +#define SE_BCI_BUSY (1 << 22) +#define SE_VGT_BUSY (1 << 23) +#define SE_PA_BUSY (1 << 24) +#define SE_TA_BUSY (1 << 25) +#define SE_SX_BUSY (1 << 26) +#define SE_SPI_BUSY (1 << 27) +#define SE_SC_BUSY (1 << 29) +#define SE_DB_BUSY (1 << 30) +#define SE_CB_BUSY (1 << 31) + +#define GRBM_SOFT_RESET 0x2008 +#define SOFT_RESET_CP (1 << 0) +#define SOFT_RESET_CB (1 << 1) +#define SOFT_RESET_RLC (1 << 2) +#define SOFT_RESET_DB (1 << 3) +#define SOFT_RESET_GDS (1 << 4) +#define SOFT_RESET_PA (1 << 5) +#define SOFT_RESET_SC (1 << 6) +#define SOFT_RESET_BCI (1 << 7) +#define SOFT_RESET_SPI (1 << 8) +#define SOFT_RESET_SX (1 << 10) +#define SOFT_RESET_TC (1 << 11) +#define SOFT_RESET_TA (1 << 12) +#define SOFT_RESET_VGT (1 << 14) +#define SOFT_RESET_IA (1 << 15) + +#define GRBM_GFX_INDEX 0x200B +#define INSTANCE_INDEX(x) ((x) << 0) +#define SH_INDEX(x) ((x) << 8) +#define SE_INDEX(x) ((x) << 16) +#define SH_BROADCAST_WRITES (1 << 29) +#define INSTANCE_BROADCAST_WRITES (1 << 30) +#define SE_BROADCAST_WRITES (1 << 31) + +#define GRBM_INT_CNTL 0x2018 +# define RDERR_INT_ENABLE (1 << 0) +# define GUI_IDLE_INT_ENABLE (1 << 19) + +#define CP_STRMOUT_CNTL 0x213F +#define SCRATCH_REG0 0x2140 +#define SCRATCH_REG1 0x2141 +#define SCRATCH_REG2 0x2142 +#define SCRATCH_REG3 0x2143 +#define SCRATCH_REG4 0x2144 +#define SCRATCH_REG5 0x2145 +#define SCRATCH_REG6 0x2146 +#define SCRATCH_REG7 0x2147 + +#define SCRATCH_UMSK 0x2150 +#define SCRATCH_ADDR 0x2151 + +#define CP_SEM_WAIT_TIMER 0x216F + +#define CP_SEM_INCOMPLETE_TIMER_CNTL 0x2172 + +#define CP_ME_CNTL 0x21B6 +#define CP_CE_HALT (1 << 24) +#define CP_PFP_HALT (1 << 26) +#define CP_ME_HALT (1 << 28) + +#define CP_COHER_CNTL2 0x217A + +#define CP_RB2_RPTR 0x21BE +#define CP_RB1_RPTR 0x21BF +#define CP_RB0_RPTR 0x21C0 +#define CP_RB_WPTR_DELAY 0x21C1 + +#define CP_QUEUE_THRESHOLDS 0x21D8 +#define ROQ_IB1_START(x) ((x) << 0) +#define ROQ_IB2_START(x) ((x) << 8) +#define CP_MEQ_THRESHOLDS 0x21D9 +#define MEQ1_START(x) ((x) << 0) +#define MEQ2_START(x) ((x) << 8) + +#define CP_PERFMON_CNTL 0x21FF + +#define VGT_VTX_VECT_EJECT_REG 0x222C + +#define VGT_CACHE_INVALIDATION 0x2231 +#define CACHE_INVALIDATION(x) ((x) << 0) +#define VC_ONLY 0 +#define TC_ONLY 1 +#define VC_AND_TC 2 +#define AUTO_INVLD_EN(x) ((x) << 6) +#define NO_AUTO 0 +#define ES_AUTO 1 +#define GS_AUTO 2 +#define ES_AND_GS_AUTO 3 +#define VGT_ESGS_RING_SIZE 0x2232 +#define VGT_GSVS_RING_SIZE 0x2233 + +#define VGT_GS_VERTEX_REUSE 0x2235 + +#define VGT_PRIMITIVE_TYPE 0x2256 +#define VGT_INDEX_TYPE 0x2257 + +#define VGT_NUM_INDICES 0x225C +#define VGT_NUM_INSTANCES 0x225D + +#define VGT_TF_RING_SIZE 0x2262 + +#define VGT_HS_OFFCHIP_PARAM 0x226C + +#define VGT_TF_MEMORY_BASE 0x226E + +#define CC_GC_SHADER_ARRAY_CONFIG 0x226F +#define INACTIVE_CUS_MASK 0xFFFF0000 +#define INACTIVE_CUS_SHIFT 16 +#define GC_USER_SHADER_ARRAY_CONFIG 0x2270 + +#define PA_CL_ENHANCE 0x2285 +#define CLIP_VTX_REORDER_ENA (1 << 0) +#define NUM_CLIP_SEQ(x) ((x) << 1) + +#define PA_SU_LINE_STIPPLE_VALUE 0x2298 + +#define PA_SC_LINE_STIPPLE_STATE 0x22C4 + +#define PA_SC_FORCE_EOV_MAX_CNTS 0x22C9 +#define FORCE_EOV_MAX_CLK_CNT(x) ((x) << 0) +#define FORCE_EOV_MAX_REZ_CNT(x) ((x) << 16) + +#define PA_SC_FIFO_SIZE 0x22F3 +#define SC_FRONTEND_PRIM_FIFO_SIZE(x) ((x) << 0) +#define SC_BACKEND_PRIM_FIFO_SIZE(x) ((x) << 6) +#define SC_HIZ_TILE_FIFO_SIZE(x) ((x) << 15) +#define SC_EARLYZ_TILE_FIFO_SIZE(x) ((x) << 23) + +#define PA_SC_ENHANCE 0x22FC + +#define SQ_CONFIG 0x2300 + +#define SQC_CACHES 0x2302 + +#define SQ_POWER_THROTTLE 0x2396 +#define MIN_POWER(x) ((x) << 0) +#define MIN_POWER_MASK (0x3fff << 0) +#define MIN_POWER_SHIFT 0 +#define MAX_POWER(x) ((x) << 16) +#define MAX_POWER_MASK (0x3fff << 16) +#define MAX_POWER_SHIFT 0 +#define SQ_POWER_THROTTLE2 0x2397 +#define MAX_POWER_DELTA(x) ((x) << 0) +#define MAX_POWER_DELTA_MASK (0x3fff << 0) +#define MAX_POWER_DELTA_SHIFT 0 +#define STI_SIZE(x) ((x) << 16) +#define STI_SIZE_MASK (0x3ff << 16) +#define STI_SIZE_SHIFT 16 +#define LTI_RATIO(x) ((x) << 27) +#define LTI_RATIO_MASK (0xf << 27) +#define LTI_RATIO_SHIFT 27 + +#define SX_DEBUG_1 0x2418 + +#define SPI_STATIC_THREAD_MGMT_1 0x2438 +#define SPI_STATIC_THREAD_MGMT_2 0x2439 +#define SPI_STATIC_THREAD_MGMT_3 0x243A +#define SPI_PS_MAX_WAVE_ID 0x243B + +#define SPI_CONFIG_CNTL 0x2440 + +#define SPI_CONFIG_CNTL_1 0x244F +#define VTX_DONE_DELAY(x) ((x) << 0) +#define INTERP_ONE_PRIM_PER_ROW (1 << 4) + +#define CGTS_TCC_DISABLE 0x2452 +#define CGTS_USER_TCC_DISABLE 0x2453 +#define TCC_DISABLE_MASK 0xFFFF0000 +#define TCC_DISABLE_SHIFT 16 +#define CGTS_SM_CTRL_REG 0x2454 +#define OVERRIDE (1 << 21) +#define LS_OVERRIDE (1 << 22) + +#define SPI_LB_CU_MASK 0x24D5 + +#define TA_CNTL_AUX 0x2542 + +#define CC_RB_BACKEND_DISABLE 0x263D +#define BACKEND_DISABLE(x) ((x) << 16) +#define GB_ADDR_CONFIG 0x263E +#define NUM_PIPES(x) ((x) << 0) +#define NUM_PIPES_MASK 0x00000007 +#define NUM_PIPES_SHIFT 0 +#define PIPE_INTERLEAVE_SIZE(x) ((x) << 4) +#define PIPE_INTERLEAVE_SIZE_MASK 0x00000070 +#define PIPE_INTERLEAVE_SIZE_SHIFT 4 +#define NUM_SHADER_ENGINES(x) ((x) << 12) +#define NUM_SHADER_ENGINES_MASK 0x00003000 +#define NUM_SHADER_ENGINES_SHIFT 12 +#define SHADER_ENGINE_TILE_SIZE(x) ((x) << 16) +#define SHADER_ENGINE_TILE_SIZE_MASK 0x00070000 +#define SHADER_ENGINE_TILE_SIZE_SHIFT 16 +#define NUM_GPUS(x) ((x) << 20) +#define NUM_GPUS_MASK 0x00700000 +#define NUM_GPUS_SHIFT 20 +#define MULTI_GPU_TILE_SIZE(x) ((x) << 24) +#define MULTI_GPU_TILE_SIZE_MASK 0x03000000 +#define MULTI_GPU_TILE_SIZE_SHIFT 24 +#define ROW_SIZE(x) ((x) << 28) +#define ROW_SIZE_MASK 0x30000000 +#define ROW_SIZE_SHIFT 28 + +#define GB_TILE_MODE0 0x2644 +# define MICRO_TILE_MODE(x) ((x) << 0) +# define ADDR_SURF_DISPLAY_MICRO_TILING 0 +# define ADDR_SURF_THIN_MICRO_TILING 1 +# define ADDR_SURF_DEPTH_MICRO_TILING 2 +# define ARRAY_MODE(x) ((x) << 2) +# define ARRAY_LINEAR_GENERAL 0 +# define ARRAY_LINEAR_ALIGNED 1 +# define ARRAY_1D_TILED_THIN1 2 +# define ARRAY_2D_TILED_THIN1 4 +# define PIPE_CONFIG(x) ((x) << 6) +# define ADDR_SURF_P2 0 +# define ADDR_SURF_P4_8x16 4 +# define ADDR_SURF_P4_16x16 5 +# define ADDR_SURF_P4_16x32 6 +# define ADDR_SURF_P4_32x32 7 +# define ADDR_SURF_P8_16x16_8x16 8 +# define ADDR_SURF_P8_16x32_8x16 9 +# define ADDR_SURF_P8_32x32_8x16 10 +# define ADDR_SURF_P8_16x32_16x16 11 +# define ADDR_SURF_P8_32x32_16x16 12 +# define ADDR_SURF_P8_32x32_16x32 13 +# define ADDR_SURF_P8_32x64_32x32 14 +# define TILE_SPLIT(x) ((x) << 11) +# define ADDR_SURF_TILE_SPLIT_64B 0 +# define ADDR_SURF_TILE_SPLIT_128B 1 +# define ADDR_SURF_TILE_SPLIT_256B 2 +# define ADDR_SURF_TILE_SPLIT_512B 3 +# define ADDR_SURF_TILE_SPLIT_1KB 4 +# define ADDR_SURF_TILE_SPLIT_2KB 5 +# define ADDR_SURF_TILE_SPLIT_4KB 6 +# define BANK_WIDTH(x) ((x) << 14) +# define ADDR_SURF_BANK_WIDTH_1 0 +# define ADDR_SURF_BANK_WIDTH_2 1 +# define ADDR_SURF_BANK_WIDTH_4 2 +# define ADDR_SURF_BANK_WIDTH_8 3 +# define BANK_HEIGHT(x) ((x) << 16) +# define ADDR_SURF_BANK_HEIGHT_1 0 +# define ADDR_SURF_BANK_HEIGHT_2 1 +# define ADDR_SURF_BANK_HEIGHT_4 2 +# define ADDR_SURF_BANK_HEIGHT_8 3 +# define MACRO_TILE_ASPECT(x) ((x) << 18) +# define ADDR_SURF_MACRO_ASPECT_1 0 +# define ADDR_SURF_MACRO_ASPECT_2 1 +# define ADDR_SURF_MACRO_ASPECT_4 2 +# define ADDR_SURF_MACRO_ASPECT_8 3 +# define NUM_BANKS(x) ((x) << 20) +# define ADDR_SURF_2_BANK 0 +# define ADDR_SURF_4_BANK 1 +# define ADDR_SURF_8_BANK 2 +# define ADDR_SURF_16_BANK 3 +#define GB_TILE_MODE1 0x2645 +#define GB_TILE_MODE2 0x2646 +#define GB_TILE_MODE3 0x2647 +#define GB_TILE_MODE4 0x2648 +#define GB_TILE_MODE5 0x2649 +#define GB_TILE_MODE6 0x264a +#define GB_TILE_MODE7 0x264b +#define GB_TILE_MODE8 0x264c +#define GB_TILE_MODE9 0x264d +#define GB_TILE_MODE10 0x264e +#define GB_TILE_MODE11 0x264f +#define GB_TILE_MODE12 0x2650 +#define GB_TILE_MODE13 0x2651 +#define GB_TILE_MODE14 0x2652 +#define GB_TILE_MODE15 0x2653 +#define GB_TILE_MODE16 0x2654 +#define GB_TILE_MODE17 0x2655 +#define GB_TILE_MODE18 0x2656 +#define GB_TILE_MODE19 0x2657 +#define GB_TILE_MODE20 0x2658 +#define GB_TILE_MODE21 0x2659 +#define GB_TILE_MODE22 0x265a +#define GB_TILE_MODE23 0x265b +#define GB_TILE_MODE24 0x265c +#define GB_TILE_MODE25 0x265d +#define GB_TILE_MODE26 0x265e +#define GB_TILE_MODE27 0x265f +#define GB_TILE_MODE28 0x2660 +#define GB_TILE_MODE29 0x2661 +#define GB_TILE_MODE30 0x2662 +#define GB_TILE_MODE31 0x2663 + +#define CB_PERFCOUNTER0_SELECT0 0x2688 +#define CB_PERFCOUNTER0_SELECT1 0x2689 +#define CB_PERFCOUNTER1_SELECT0 0x268A +#define CB_PERFCOUNTER1_SELECT1 0x268B +#define CB_PERFCOUNTER2_SELECT0 0x268C +#define CB_PERFCOUNTER2_SELECT1 0x268D +#define CB_PERFCOUNTER3_SELECT0 0x268E +#define CB_PERFCOUNTER3_SELECT1 0x268F + +#define CB_CGTT_SCLK_CTRL 0x2698 + +#define GC_USER_RB_BACKEND_DISABLE 0x26DF +#define BACKEND_DISABLE_MASK 0x00FF0000 +#define BACKEND_DISABLE_SHIFT 16 + +#define TCP_CHAN_STEER_LO 0x2B03 +#define TCP_CHAN_STEER_HI 0x2B94 + +#define CP_RB0_BASE 0x3040 +#define CP_RB0_CNTL 0x3041 +#define RB_BUFSZ(x) ((x) << 0) +#define RB_BLKSZ(x) ((x) << 8) +#define BUF_SWAP_32BIT (2 << 16) +#define RB_NO_UPDATE (1 << 27) +#define RB_RPTR_WR_ENA (1 << 31) + +#define CP_RB0_RPTR_ADDR 0x3043 +#define CP_RB0_RPTR_ADDR_HI 0x3044 +#define CP_RB0_WPTR 0x3045 + +#define CP_PFP_UCODE_ADDR 0x3054 +#define CP_PFP_UCODE_DATA 0x3055 +#define CP_ME_RAM_RADDR 0x3056 +#define CP_ME_RAM_WADDR 0x3057 +#define CP_ME_RAM_DATA 0x3058 + +#define CP_CE_UCODE_ADDR 0x305A +#define CP_CE_UCODE_DATA 0x305B + +#define CP_RB1_BASE 0x3060 +#define CP_RB1_CNTL 0x3061 +#define CP_RB1_RPTR_ADDR 0x3062 +#define CP_RB1_RPTR_ADDR_HI 0x3063 +#define CP_RB1_WPTR 0x3064 +#define CP_RB2_BASE 0x3065 +#define CP_RB2_CNTL 0x3066 +#define CP_RB2_RPTR_ADDR 0x3067 +#define CP_RB2_RPTR_ADDR_HI 0x3068 +#define CP_RB2_WPTR 0x3069 +#define CP_INT_CNTL_RING0 0x306A +#define CP_INT_CNTL_RING1 0x306B +#define CP_INT_CNTL_RING2 0x306C +# define CNTX_BUSY_INT_ENABLE (1 << 19) +# define CNTX_EMPTY_INT_ENABLE (1 << 20) +# define WAIT_MEM_SEM_INT_ENABLE (1 << 21) +# define TIME_STAMP_INT_ENABLE (1 << 26) +# define CP_RINGID2_INT_ENABLE (1 << 29) +# define CP_RINGID1_INT_ENABLE (1 << 30) +# define CP_RINGID0_INT_ENABLE (1 << 31) +#define CP_INT_STATUS_RING0 0x306D +#define CP_INT_STATUS_RING1 0x306E +#define CP_INT_STATUS_RING2 0x306F +# define WAIT_MEM_SEM_INT_STAT (1 << 21) +# define TIME_STAMP_INT_STAT (1 << 26) +# define CP_RINGID2_INT_STAT (1 << 29) +# define CP_RINGID1_INT_STAT (1 << 30) +# define CP_RINGID0_INT_STAT (1 << 31) + +#define CP_MEM_SLP_CNTL 0x3079 +# define CP_MEM_LS_EN (1 << 0) + +#define CP_DEBUG 0x307F + +#define RLC_CNTL 0x30C0 +# define RLC_ENABLE (1 << 0) +#define RLC_RL_BASE 0x30C1 +#define RLC_RL_SIZE 0x30C2 +#define RLC_LB_CNTL 0x30C3 +# define LOAD_BALANCE_ENABLE (1 << 0) +#define RLC_SAVE_AND_RESTORE_BASE 0x30C4 +#define RLC_LB_CNTR_MAX 0x30C5 +#define RLC_LB_CNTR_INIT 0x30C6 + +#define RLC_CLEAR_STATE_RESTORE_BASE 0x30C8 + +#define RLC_UCODE_ADDR 0x30CB +#define RLC_UCODE_DATA 0x30CC + +#define RLC_GPU_CLOCK_COUNT_LSB 0x30CE +#define RLC_GPU_CLOCK_COUNT_MSB 0x30CF +#define RLC_CAPTURE_GPU_CLOCK_COUNT 0x30D0 +#define RLC_MC_CNTL 0x30D1 +#define RLC_UCODE_CNTL 0x30D2 +#define RLC_STAT 0x30D3 +# define RLC_BUSY_STATUS (1 << 0) +# define GFX_POWER_STATUS (1 << 1) +# define GFX_CLOCK_STATUS (1 << 2) +# define GFX_LS_STATUS (1 << 3) + +#define RLC_PG_CNTL 0x30D7 +# define GFX_PG_ENABLE (1 << 0) +# define GFX_PG_SRC (1 << 1) + +#define RLC_CGTT_MGCG_OVERRIDE 0x3100 +#define RLC_CGCG_CGLS_CTRL 0x3101 +# define CGCG_EN (1 << 0) +# define CGLS_EN (1 << 1) + +#define RLC_TTOP_D 0x3105 +# define RLC_PUD(x) ((x) << 0) +# define RLC_PUD_MASK (0xff << 0) +# define RLC_PDD(x) ((x) << 8) +# define RLC_PDD_MASK (0xff << 8) +# define RLC_TTPD(x) ((x) << 16) +# define RLC_TTPD_MASK (0xff << 16) +# define RLC_MSD(x) ((x) << 24) +# define RLC_MSD_MASK (0xff << 24) + +#define RLC_LB_INIT_CU_MASK 0x3107 + +#define RLC_PG_AO_CU_MASK 0x310B +#define RLC_MAX_PG_CU 0x310C +# define MAX_PU_CU(x) ((x) << 0) +# define MAX_PU_CU_MASK (0xff << 0) +#define RLC_AUTO_PG_CTRL 0x310C +# define AUTO_PG_EN (1 << 0) +# define GRBM_REG_SGIT(x) ((x) << 3) +# define GRBM_REG_SGIT_MASK (0xffff << 3) +# define PG_AFTER_GRBM_REG_ST(x) ((x) << 19) +# define PG_AFTER_GRBM_REG_ST_MASK (0x1fff << 19) + +#define RLC_SERDES_WR_MASTER_MASK_0 0x3115 +#define RLC_SERDES_WR_MASTER_MASK_1 0x3116 +#define RLC_SERDES_WR_CTRL 0x3117 + +#define RLC_SERDES_MASTER_BUSY_0 0x3119 +#define RLC_SERDES_MASTER_BUSY_1 0x311A + +#define RLC_GCPM_GENERAL_3 0x311E + +#define DB_RENDER_CONTROL 0xA000 + +#define DB_DEPTH_INFO 0xA00F + +#define PA_SC_RASTER_CONFIG 0xA0D4 +# define RASTER_CONFIG_RB_MAP_0 0 +# define RASTER_CONFIG_RB_MAP_1 1 +# define RASTER_CONFIG_RB_MAP_2 2 +# define RASTER_CONFIG_RB_MAP_3 3 + +#define VGT_EVENT_INITIATOR 0xA2A4 +# define SAMPLE_STREAMOUTSTATS1 (1 << 0) +# define SAMPLE_STREAMOUTSTATS2 (2 << 0) +# define SAMPLE_STREAMOUTSTATS3 (3 << 0) +# define CACHE_FLUSH_TS (4 << 0) +# define CACHE_FLUSH (6 << 0) +# define CS_PARTIAL_FLUSH (7 << 0) +# define VGT_STREAMOUT_RESET (10 << 0) +# define END_OF_PIPE_INCR_DE (11 << 0) +# define END_OF_PIPE_IB_END (12 << 0) +# define RST_PIX_CNT (13 << 0) +# define VS_PARTIAL_FLUSH (15 << 0) +# define PS_PARTIAL_FLUSH (16 << 0) +# define CACHE_FLUSH_AND_INV_TS_EVENT (20 << 0) +# define ZPASS_DONE (21 << 0) +# define CACHE_FLUSH_AND_INV_EVENT (22 << 0) +# define PERFCOUNTER_START (23 << 0) +# define PERFCOUNTER_STOP (24 << 0) +# define PIPELINESTAT_START (25 << 0) +# define PIPELINESTAT_STOP (26 << 0) +# define PERFCOUNTER_SAMPLE (27 << 0) +# define SAMPLE_PIPELINESTAT (30 << 0) +# define SAMPLE_STREAMOUTSTATS (32 << 0) +# define RESET_VTX_CNT (33 << 0) +# define VGT_FLUSH (36 << 0) +# define BOTTOM_OF_PIPE_TS (40 << 0) +# define DB_CACHE_FLUSH_AND_INV (42 << 0) +# define FLUSH_AND_INV_DB_DATA_TS (43 << 0) +# define FLUSH_AND_INV_DB_META (44 << 0) +# define FLUSH_AND_INV_CB_DATA_TS (45 << 0) +# define FLUSH_AND_INV_CB_META (46 << 0) +# define CS_DONE (47 << 0) +# define PS_DONE (48 << 0) +# define FLUSH_AND_INV_CB_PIXEL_DATA (49 << 0) +# define THREAD_TRACE_START (51 << 0) +# define THREAD_TRACE_STOP (52 << 0) +# define THREAD_TRACE_FLUSH (54 << 0) +# define THREAD_TRACE_FINISH (55 << 0) + +/* PIF PHY0 registers idx/data 0x8/0xc */ +#define PB0_PIF_CNTL 0x10 +# define LS2_EXIT_TIME(x) ((x) << 17) +# define LS2_EXIT_TIME_MASK (0x7 << 17) +# define LS2_EXIT_TIME_SHIFT 17 +#define PB0_PIF_PAIRING 0x11 +# define MULTI_PIF (1 << 25) +#define PB0_PIF_PWRDOWN_0 0x12 +# define PLL_POWER_STATE_IN_TXS2_0(x) ((x) << 7) +# define PLL_POWER_STATE_IN_TXS2_0_MASK (0x7 << 7) +# define PLL_POWER_STATE_IN_TXS2_0_SHIFT 7 +# define PLL_POWER_STATE_IN_OFF_0(x) ((x) << 10) +# define PLL_POWER_STATE_IN_OFF_0_MASK (0x7 << 10) +# define PLL_POWER_STATE_IN_OFF_0_SHIFT 10 +# define PLL_RAMP_UP_TIME_0(x) ((x) << 24) +# define PLL_RAMP_UP_TIME_0_MASK (0x7 << 24) +# define PLL_RAMP_UP_TIME_0_SHIFT 24 +#define PB0_PIF_PWRDOWN_1 0x13 +# define PLL_POWER_STATE_IN_TXS2_1(x) ((x) << 7) +# define PLL_POWER_STATE_IN_TXS2_1_MASK (0x7 << 7) +# define PLL_POWER_STATE_IN_TXS2_1_SHIFT 7 +# define PLL_POWER_STATE_IN_OFF_1(x) ((x) << 10) +# define PLL_POWER_STATE_IN_OFF_1_MASK (0x7 << 10) +# define PLL_POWER_STATE_IN_OFF_1_SHIFT 10 +# define PLL_RAMP_UP_TIME_1(x) ((x) << 24) +# define PLL_RAMP_UP_TIME_1_MASK (0x7 << 24) +# define PLL_RAMP_UP_TIME_1_SHIFT 24 + +#define PB0_PIF_PWRDOWN_2 0x17 +# define PLL_POWER_STATE_IN_TXS2_2(x) ((x) << 7) +# define PLL_POWER_STATE_IN_TXS2_2_MASK (0x7 << 7) +# define PLL_POWER_STATE_IN_TXS2_2_SHIFT 7 +# define PLL_POWER_STATE_IN_OFF_2(x) ((x) << 10) +# define PLL_POWER_STATE_IN_OFF_2_MASK (0x7 << 10) +# define PLL_POWER_STATE_IN_OFF_2_SHIFT 10 +# define PLL_RAMP_UP_TIME_2(x) ((x) << 24) +# define PLL_RAMP_UP_TIME_2_MASK (0x7 << 24) +# define PLL_RAMP_UP_TIME_2_SHIFT 24 +#define PB0_PIF_PWRDOWN_3 0x18 +# define PLL_POWER_STATE_IN_TXS2_3(x) ((x) << 7) +# define PLL_POWER_STATE_IN_TXS2_3_MASK (0x7 << 7) +# define PLL_POWER_STATE_IN_TXS2_3_SHIFT 7 +# define PLL_POWER_STATE_IN_OFF_3(x) ((x) << 10) +# define PLL_POWER_STATE_IN_OFF_3_MASK (0x7 << 10) +# define PLL_POWER_STATE_IN_OFF_3_SHIFT 10 +# define PLL_RAMP_UP_TIME_3(x) ((x) << 24) +# define PLL_RAMP_UP_TIME_3_MASK (0x7 << 24) +# define PLL_RAMP_UP_TIME_3_SHIFT 24 +/* PIF PHY1 registers idx/data 0x10/0x14 */ +#define PB1_PIF_CNTL 0x10 +#define PB1_PIF_PAIRING 0x11 +#define PB1_PIF_PWRDOWN_0 0x12 +#define PB1_PIF_PWRDOWN_1 0x13 + +#define PB1_PIF_PWRDOWN_2 0x17 +#define PB1_PIF_PWRDOWN_3 0x18 +/* PCIE registers idx/data 0x30/0x34 */ +#define PCIE_CNTL2 0x1c /* PCIE */ +# define SLV_MEM_LS_EN (1 << 16) +# define SLV_MEM_AGGRESSIVE_LS_EN (1 << 17) +# define MST_MEM_LS_EN (1 << 18) +# define REPLAY_MEM_LS_EN (1 << 19) +#define PCIE_LC_STATUS1 0x28 /* PCIE */ +# define LC_REVERSE_RCVR (1 << 0) +# define LC_REVERSE_XMIT (1 << 1) +# define LC_OPERATING_LINK_WIDTH_MASK (0x7 << 2) +# define LC_OPERATING_LINK_WIDTH_SHIFT 2 +# define LC_DETECTED_LINK_WIDTH_MASK (0x7 << 5) +# define LC_DETECTED_LINK_WIDTH_SHIFT 5 + +#define PCIE_P_CNTL 0x40 /* PCIE */ +# define P_IGNORE_EDB_ERR (1 << 6) + +/* PCIE PORT registers idx/data 0x38/0x3c */ +#define PCIE_LC_CNTL 0xa0 +# define LC_L0S_INACTIVITY(x) ((x) << 8) +# define LC_L0S_INACTIVITY_MASK (0xf << 8) +# define LC_L0S_INACTIVITY_SHIFT 8 +# define LC_L1_INACTIVITY(x) ((x) << 12) +# define LC_L1_INACTIVITY_MASK (0xf << 12) +# define LC_L1_INACTIVITY_SHIFT 12 +# define LC_PMI_TO_L1_DIS (1 << 16) +# define LC_ASPM_TO_L1_DIS (1 << 24) +#define PCIE_LC_LINK_WIDTH_CNTL 0xa2 /* PCIE_P */ +# define LC_LINK_WIDTH_SHIFT 0 +# define LC_LINK_WIDTH_MASK 0x7 +# define LC_LINK_WIDTH_X0 0 +# define LC_LINK_WIDTH_X1 1 +# define LC_LINK_WIDTH_X2 2 +# define LC_LINK_WIDTH_X4 3 +# define LC_LINK_WIDTH_X8 4 +# define LC_LINK_WIDTH_X16 6 +# define LC_LINK_WIDTH_RD_SHIFT 4 +# define LC_LINK_WIDTH_RD_MASK 0x70 +# define LC_RECONFIG_ARC_MISSING_ESCAPE (1 << 7) +# define LC_RECONFIG_NOW (1 << 8) +# define LC_RENEGOTIATION_SUPPORT (1 << 9) +# define LC_RENEGOTIATE_EN (1 << 10) +# define LC_SHORT_RECONFIG_EN (1 << 11) +# define LC_UPCONFIGURE_SUPPORT (1 << 12) +# define LC_UPCONFIGURE_DIS (1 << 13) +# define LC_DYN_LANES_PWR_STATE(x) ((x) << 21) +# define LC_DYN_LANES_PWR_STATE_MASK (0x3 << 21) +# define LC_DYN_LANES_PWR_STATE_SHIFT 21 +#define PCIE_LC_N_FTS_CNTL 0xa3 /* PCIE_P */ +# define LC_XMIT_N_FTS(x) ((x) << 0) +# define LC_XMIT_N_FTS_MASK (0xff << 0) +# define LC_XMIT_N_FTS_SHIFT 0 +# define LC_XMIT_N_FTS_OVERRIDE_EN (1 << 8) +# define LC_N_FTS_MASK (0xff << 24) +#define PCIE_LC_SPEED_CNTL 0xa4 /* PCIE_P */ +# define LC_GEN2_EN_STRAP (1 << 0) +# define LC_GEN3_EN_STRAP (1 << 1) +# define LC_TARGET_LINK_SPEED_OVERRIDE_EN (1 << 2) +# define LC_TARGET_LINK_SPEED_OVERRIDE_MASK (0x3 << 3) +# define LC_TARGET_LINK_SPEED_OVERRIDE_SHIFT 3 +# define LC_FORCE_EN_SW_SPEED_CHANGE (1 << 5) +# define LC_FORCE_DIS_SW_SPEED_CHANGE (1 << 6) +# define LC_FORCE_EN_HW_SPEED_CHANGE (1 << 7) +# define LC_FORCE_DIS_HW_SPEED_CHANGE (1 << 8) +# define LC_INITIATE_LINK_SPEED_CHANGE (1 << 9) +# define LC_SPEED_CHANGE_ATTEMPTS_ALLOWED_MASK (0x3 << 10) +# define LC_SPEED_CHANGE_ATTEMPTS_ALLOWED_SHIFT 10 +# define LC_CURRENT_DATA_RATE_MASK (0x3 << 13) /* 0/1/2 = gen1/2/3 */ +# define LC_CURRENT_DATA_RATE_SHIFT 13 +# define LC_CLR_FAILED_SPD_CHANGE_CNT (1 << 16) +# define LC_OTHER_SIDE_EVER_SENT_GEN2 (1 << 18) +# define LC_OTHER_SIDE_SUPPORTS_GEN2 (1 << 19) +# define LC_OTHER_SIDE_EVER_SENT_GEN3 (1 << 20) +# define LC_OTHER_SIDE_SUPPORTS_GEN3 (1 << 21) + +#define PCIE_LC_CNTL2 0xb1 +# define LC_ALLOW_PDWN_IN_L1 (1 << 17) +# define LC_ALLOW_PDWN_IN_L23 (1 << 18) + +#define PCIE_LC_CNTL3 0xb5 /* PCIE_P */ +# define LC_GO_TO_RECOVERY (1 << 30) +#define PCIE_LC_CNTL4 0xb6 /* PCIE_P */ +# define LC_REDO_EQ (1 << 5) +# define LC_SET_QUIESCE (1 << 13) + +/* + * UVD + */ +#define UVD_UDEC_ADDR_CONFIG 0x3bd3 +#define UVD_UDEC_DB_ADDR_CONFIG 0x3bd4 +#define UVD_UDEC_DBW_ADDR_CONFIG 0x3bd5 +#define UVD_RBC_RB_RPTR 0x3da4 +#define UVD_RBC_RB_WPTR 0x3da5 +#define UVD_STATUS 0x3daf + +#define UVD_CGC_CTRL 0x3dc2 +# define DCM (1 << 0) +# define CG_DT(x) ((x) << 2) +# define CG_DT_MASK (0xf << 2) +# define CLK_OD(x) ((x) << 6) +# define CLK_OD_MASK (0x1f << 6) + + /* UVD CTX indirect */ +#define UVD_CGC_MEM_CTRL 0xC0 +#define UVD_CGC_CTRL2 0xC1 +# define DYN_OR_EN (1 << 0) +# define DYN_RR_EN (1 << 1) +# define G_DIV_ID(x) ((x) << 2) +# define G_DIV_ID_MASK (0x7 << 2) + +/* + * PM4 + */ +#define PACKET0(reg, n) ((RADEON_PACKET_TYPE0 << 30) | \ + (((reg) >> 2) & 0xFFFF) | \ + ((n) & 0x3FFF) << 16) +#define CP_PACKET2 0x80000000 +#define PACKET2_PAD_SHIFT 0 +#define PACKET2_PAD_MASK (0x3fffffff << 0) + +#define PACKET2(v) (CP_PACKET2 | REG_SET(PACKET2_PAD, (v))) +#define RADEON_PACKET_TYPE3 3 +#define PACKET3(op, n) ((RADEON_PACKET_TYPE3 << 30) | \ + (((op) & 0xFF) << 8) | \ + ((n) & 0x3FFF) << 16) + +#define PACKET3_COMPUTE(op, n) (PACKET3(op, n) | 1 << 1) + +/* Packet 3 types */ +#define PACKET3_NOP 0x10 +#define PACKET3_SET_BASE 0x11 +#define PACKET3_BASE_INDEX(x) ((x) << 0) +#define GDS_PARTITION_BASE 2 +#define CE_PARTITION_BASE 3 +#define PACKET3_CLEAR_STATE 0x12 +#define PACKET3_INDEX_BUFFER_SIZE 0x13 +#define PACKET3_DISPATCH_DIRECT 0x15 +#define PACKET3_DISPATCH_INDIRECT 0x16 +#define PACKET3_ALLOC_GDS 0x1B +#define PACKET3_WRITE_GDS_RAM 0x1C +#define PACKET3_ATOMIC_GDS 0x1D +#define PACKET3_ATOMIC 0x1E +#define PACKET3_OCCLUSION_QUERY 0x1F +#define PACKET3_SET_PREDICATION 0x20 +#define PACKET3_REG_RMW 0x21 +#define PACKET3_COND_EXEC 0x22 +#define PACKET3_PRED_EXEC 0x23 +#define PACKET3_DRAW_INDIRECT 0x24 +#define PACKET3_DRAW_INDEX_INDIRECT 0x25 +#define PACKET3_INDEX_BASE 0x26 +#define PACKET3_DRAW_INDEX_2 0x27 +#define PACKET3_CONTEXT_CONTROL 0x28 +#define PACKET3_INDEX_TYPE 0x2A +#define PACKET3_DRAW_INDIRECT_MULTI 0x2C +#define PACKET3_DRAW_INDEX_AUTO 0x2D +#define PACKET3_DRAW_INDEX_IMMD 0x2E +#define PACKET3_NUM_INSTANCES 0x2F +#define PACKET3_DRAW_INDEX_MULTI_AUTO 0x30 +#define PACKET3_INDIRECT_BUFFER_CONST 0x31 +#define PACKET3_INDIRECT_BUFFER 0x3F +#define PACKET3_STRMOUT_BUFFER_UPDATE 0x34 +#define PACKET3_DRAW_INDEX_OFFSET_2 0x35 +#define PACKET3_DRAW_INDEX_MULTI_ELEMENT 0x36 +#define PACKET3_WRITE_DATA 0x37 +#define WRITE_DATA_DST_SEL(x) ((x) << 8) + /* 0 - register + * 1 - memory (sync - via GRBM) + * 2 - tc/l2 + * 3 - gds + * 4 - reserved + * 5 - memory (async - direct) + */ +#define WR_ONE_ADDR (1 << 16) +#define WR_CONFIRM (1 << 20) +#define WRITE_DATA_ENGINE_SEL(x) ((x) << 30) + /* 0 - me + * 1 - pfp + * 2 - ce + */ +#define PACKET3_DRAW_INDEX_INDIRECT_MULTI 0x38 +#define PACKET3_MEM_SEMAPHORE 0x39 +#define PACKET3_MPEG_INDEX 0x3A +#define PACKET3_COPY_DW 0x3B +#define PACKET3_WAIT_REG_MEM 0x3C +#define WAIT_REG_MEM_FUNCTION(x) ((x) << 0) + /* 0 - always + * 1 - < + * 2 - <= + * 3 - == + * 4 - != + * 5 - >= + * 6 - > + */ +#define WAIT_REG_MEM_MEM_SPACE(x) ((x) << 4) + /* 0 - reg + * 1 - mem + */ +#define WAIT_REG_MEM_ENGINE(x) ((x) << 8) + /* 0 - me + * 1 - pfp + */ +#define PACKET3_MEM_WRITE 0x3D +#define PACKET3_COPY_DATA 0x40 +#define PACKET3_CP_DMA 0x41 +/* 1. header + * 2. SRC_ADDR_LO or DATA [31:0] + * 3. CP_SYNC [31] | SRC_SEL [30:29] | ENGINE [27] | DST_SEL [21:20] | + * SRC_ADDR_HI [7:0] + * 4. DST_ADDR_LO [31:0] + * 5. DST_ADDR_HI [7:0] + * 6. COMMAND [30:21] | BYTE_COUNT [20:0] + */ +# define PACKET3_CP_DMA_DST_SEL(x) ((x) << 20) + /* 0 - DST_ADDR + * 1 - GDS + */ +# define PACKET3_CP_DMA_ENGINE(x) ((x) << 27) + /* 0 - ME + * 1 - PFP + */ +# define PACKET3_CP_DMA_SRC_SEL(x) ((x) << 29) + /* 0 - SRC_ADDR + * 1 - GDS + * 2 - DATA + */ +# define PACKET3_CP_DMA_CP_SYNC (1 << 31) +/* COMMAND */ +# define PACKET3_CP_DMA_DIS_WC (1 << 21) +# define PACKET3_CP_DMA_CMD_SRC_SWAP(x) ((x) << 22) + /* 0 - none + * 1 - 8 in 16 + * 2 - 8 in 32 + * 3 - 8 in 64 + */ +# define PACKET3_CP_DMA_CMD_DST_SWAP(x) ((x) << 24) + /* 0 - none + * 1 - 8 in 16 + * 2 - 8 in 32 + * 3 - 8 in 64 + */ +# define PACKET3_CP_DMA_CMD_SAS (1 << 26) + /* 0 - memory + * 1 - register + */ +# define PACKET3_CP_DMA_CMD_DAS (1 << 27) + /* 0 - memory + * 1 - register + */ +# define PACKET3_CP_DMA_CMD_SAIC (1 << 28) +# define PACKET3_CP_DMA_CMD_DAIC (1 << 29) +# define PACKET3_CP_DMA_CMD_RAW_WAIT (1 << 30) +#define PACKET3_PFP_SYNC_ME 0x42 +#define PACKET3_SURFACE_SYNC 0x43 +# define PACKET3_DEST_BASE_0_ENA (1 << 0) +# define PACKET3_DEST_BASE_1_ENA (1 << 1) +# define PACKET3_CB0_DEST_BASE_ENA (1 << 6) +# define PACKET3_CB1_DEST_BASE_ENA (1 << 7) +# define PACKET3_CB2_DEST_BASE_ENA (1 << 8) +# define PACKET3_CB3_DEST_BASE_ENA (1 << 9) +# define PACKET3_CB4_DEST_BASE_ENA (1 << 10) +# define PACKET3_CB5_DEST_BASE_ENA (1 << 11) +# define PACKET3_CB6_DEST_BASE_ENA (1 << 12) +# define PACKET3_CB7_DEST_BASE_ENA (1 << 13) +# define PACKET3_DB_DEST_BASE_ENA (1 << 14) +# define PACKET3_DEST_BASE_2_ENA (1 << 19) +# define PACKET3_DEST_BASE_3_ENA (1 << 21) +# define PACKET3_TCL1_ACTION_ENA (1 << 22) +# define PACKET3_TC_ACTION_ENA (1 << 23) +# define PACKET3_CB_ACTION_ENA (1 << 25) +# define PACKET3_DB_ACTION_ENA (1 << 26) +# define PACKET3_SH_KCACHE_ACTION_ENA (1 << 27) +# define PACKET3_SH_ICACHE_ACTION_ENA (1 << 29) +#define PACKET3_ME_INITIALIZE 0x44 +#define PACKET3_ME_INITIALIZE_DEVICE_ID(x) ((x) << 16) +#define PACKET3_COND_WRITE 0x45 +#define PACKET3_EVENT_WRITE 0x46 +#define EVENT_TYPE(x) ((x) << 0) +#define EVENT_INDEX(x) ((x) << 8) + /* 0 - any non-TS event + * 1 - ZPASS_DONE + * 2 - SAMPLE_PIPELINESTAT + * 3 - SAMPLE_STREAMOUTSTAT* + * 4 - *S_PARTIAL_FLUSH + * 5 - EOP events + * 6 - EOS events + * 7 - CACHE_FLUSH, CACHE_FLUSH_AND_INV_EVENT + */ +#define INV_L2 (1 << 20) + /* INV TC L2 cache when EVENT_INDEX = 7 */ +#define PACKET3_EVENT_WRITE_EOP 0x47 +#define DATA_SEL(x) ((x) << 29) + /* 0 - discard + * 1 - send low 32bit data + * 2 - send 64bit data + * 3 - send 64bit counter value + */ +#define INT_SEL(x) ((x) << 24) + /* 0 - none + * 1 - interrupt only (DATA_SEL = 0) + * 2 - interrupt when data write is confirmed + */ +#define PACKET3_EVENT_WRITE_EOS 0x48 +#define PACKET3_PREAMBLE_CNTL 0x4A +# define PACKET3_PREAMBLE_BEGIN_CLEAR_STATE (2 << 28) +# define PACKET3_PREAMBLE_END_CLEAR_STATE (3 << 28) +#define PACKET3_ONE_REG_WRITE 0x57 +#define PACKET3_LOAD_CONFIG_REG 0x5F +#define PACKET3_LOAD_CONTEXT_REG 0x60 +#define PACKET3_LOAD_SH_REG 0x61 +#define PACKET3_SET_CONFIG_REG 0x68 +#define PACKET3_SET_CONFIG_REG_START 0x00002000 +#define PACKET3_SET_CONFIG_REG_END 0x00002c00 +#define PACKET3_SET_CONTEXT_REG 0x69 +#define PACKET3_SET_CONTEXT_REG_START 0x000a000 +#define PACKET3_SET_CONTEXT_REG_END 0x000a400 +#define PACKET3_SET_CONTEXT_REG_INDIRECT 0x73 +#define PACKET3_SET_RESOURCE_INDIRECT 0x74 +#define PACKET3_SET_SH_REG 0x76 +#define PACKET3_SET_SH_REG_START 0x00002c00 +#define PACKET3_SET_SH_REG_END 0x00003000 +#define PACKET3_SET_SH_REG_OFFSET 0x77 +#define PACKET3_ME_WRITE 0x7A +#define PACKET3_SCRATCH_RAM_WRITE 0x7D +#define PACKET3_SCRATCH_RAM_READ 0x7E +#define PACKET3_CE_WRITE 0x7F +#define PACKET3_LOAD_CONST_RAM 0x80 +#define PACKET3_WRITE_CONST_RAM 0x81 +#define PACKET3_WRITE_CONST_RAM_OFFSET 0x82 +#define PACKET3_DUMP_CONST_RAM 0x83 +#define PACKET3_INCREMENT_CE_COUNTER 0x84 +#define PACKET3_INCREMENT_DE_COUNTER 0x85 +#define PACKET3_WAIT_ON_CE_COUNTER 0x86 +#define PACKET3_WAIT_ON_DE_COUNTER 0x87 +#define PACKET3_WAIT_ON_DE_COUNTER_DIFF 0x88 +#define PACKET3_SET_CE_DE_COUNTERS 0x89 +#define PACKET3_WAIT_ON_AVAIL_BUFFER 0x8A +#define PACKET3_SWITCH_BUFFER 0x8B + +/* ASYNC DMA - first instance at 0xd000, second at 0xd800 */ +#define DMA0_REGISTER_OFFSET 0x0 /* not a register */ +#define DMA1_REGISTER_OFFSET 0x200 /* not a register */ + +#define DMA_RB_CNTL 0x3400 +# define DMA_RB_ENABLE (1 << 0) +# define DMA_RB_SIZE(x) ((x) << 1) /* log2 */ +# define DMA_RB_SWAP_ENABLE (1 << 9) /* 8IN32 */ +# define DMA_RPTR_WRITEBACK_ENABLE (1 << 12) +# define DMA_RPTR_WRITEBACK_SWAP_ENABLE (1 << 13) /* 8IN32 */ +# define DMA_RPTR_WRITEBACK_TIMER(x) ((x) << 16) /* log2 */ +#define DMA_RB_BASE 0x3401 +#define DMA_RB_RPTR 0x3402 +#define DMA_RB_WPTR 0x3403 + +#define DMA_RB_RPTR_ADDR_HI 0x3407 +#define DMA_RB_RPTR_ADDR_LO 0x3408 + +#define DMA_IB_CNTL 0x3409 +# define DMA_IB_ENABLE (1 << 0) +# define DMA_IB_SWAP_ENABLE (1 << 4) +# define CMD_VMID_FORCE (1 << 31) +#define DMA_IB_RPTR 0x340a +#define DMA_CNTL 0x340b +# define TRAP_ENABLE (1 << 0) +# define SEM_INCOMPLETE_INT_ENABLE (1 << 1) +# define SEM_WAIT_INT_ENABLE (1 << 2) +# define DATA_SWAP_ENABLE (1 << 3) +# define FENCE_SWAP_ENABLE (1 << 4) +# define CTXEMPTY_INT_ENABLE (1 << 28) +#define DMA_STATUS_REG 0x340d +# define DMA_IDLE (1 << 0) +#define DMA_TILING_CONFIG 0x342e + +#define DMA_POWER_CNTL 0x342f +# define MEM_POWER_OVERRIDE (1 << 8) +#define DMA_CLK_CTRL 0x3430 + +#define DMA_PG 0x3435 +# define PG_CNTL_ENABLE (1 << 0) +#define DMA_PGFSM_CONFIG 0x3436 +#define DMA_PGFSM_WRITE 0x3437 + +#define DMA_PACKET(cmd, b, t, s, n) ((((cmd) & 0xF) << 28) | \ + (((b) & 0x1) << 26) | \ + (((t) & 0x1) << 23) | \ + (((s) & 0x1) << 22) | \ + (((n) & 0xFFFFF) << 0)) + +#define DMA_IB_PACKET(cmd, vmid, n) ((((cmd) & 0xF) << 28) | \ + (((vmid) & 0xF) << 20) | \ + (((n) & 0xFFFFF) << 0)) + +#define DMA_PTE_PDE_PACKET(n) ((2 << 28) | \ + (1 << 26) | \ + (1 << 21) | \ + (((n) & 0xFFFFF) << 0)) + +/* async DMA Packet types */ +#define DMA_PACKET_WRITE 0x2 +#define DMA_PACKET_COPY 0x3 +#define DMA_PACKET_INDIRECT_BUFFER 0x4 +#define DMA_PACKET_SEMAPHORE 0x5 +#define DMA_PACKET_FENCE 0x6 +#define DMA_PACKET_TRAP 0x7 +#define DMA_PACKET_SRBM_WRITE 0x9 +#define DMA_PACKET_CONSTANT_FILL 0xd +#define DMA_PACKET_POLL_REG_MEM 0xe +#define DMA_PACKET_NOP 0xf + +#define VCE_STATUS 0x20004 +#define VCE_VCPU_CNTL 0x20014 +#define VCE_CLK_EN (1 << 0) +#define VCE_VCPU_CACHE_OFFSET0 0x20024 +#define VCE_VCPU_CACHE_SIZE0 0x20028 +#define VCE_VCPU_CACHE_OFFSET1 0x2002c +#define VCE_VCPU_CACHE_SIZE1 0x20030 +#define VCE_VCPU_CACHE_OFFSET2 0x20034 +#define VCE_VCPU_CACHE_SIZE2 0x20038 +#define VCE_SOFT_RESET 0x20120 +#define VCE_ECPU_SOFT_RESET (1 << 0) +#define VCE_FME_SOFT_RESET (1 << 2) +#define VCE_RB_BASE_LO2 0x2016c +#define VCE_RB_BASE_HI2 0x20170 +#define VCE_RB_SIZE2 0x20174 +#define VCE_RB_RPTR2 0x20178 +#define VCE_RB_WPTR2 0x2017c +#define VCE_RB_BASE_LO 0x20180 +#define VCE_RB_BASE_HI 0x20184 +#define VCE_RB_SIZE 0x20188 +#define VCE_RB_RPTR 0x2018c +#define VCE_RB_WPTR 0x20190 +#define VCE_CLOCK_GATING_A 0x202f8 +#define VCE_CLOCK_GATING_B 0x202fc +#define VCE_UENC_CLOCK_GATING 0x205bc +#define VCE_UENC_REG_CLOCK_GATING 0x205c0 +#define VCE_FW_REG_STATUS 0x20e10 +# define VCE_FW_REG_STATUS_BUSY (1 << 0) +# define VCE_FW_REG_STATUS_PASS (1 << 3) +# define VCE_FW_REG_STATUS_DONE (1 << 11) +#define VCE_LMI_FW_START_KEYSEL 0x20e18 +#define VCE_LMI_FW_PERIODIC_CTRL 0x20e20 +#define VCE_LMI_CTRL2 0x20e74 +#define VCE_LMI_CTRL 0x20e98 +#define VCE_LMI_VM_CTRL 0x20ea0 +#define VCE_LMI_SWAP_CNTL 0x20eb4 +#define VCE_LMI_SWAP_CNTL1 0x20eb8 +#define VCE_LMI_CACHE_CTRL 0x20ef4 + +#define VCE_CMD_NO_OP 0x00000000 +#define VCE_CMD_END 0x00000001 +#define VCE_CMD_IB 0x00000002 +#define VCE_CMD_FENCE 0x00000003 +#define VCE_CMD_TRAP 0x00000004 +#define VCE_CMD_IB_AUTO 0x00000005 +#define VCE_CMD_SEMAPHORE 0x00000006 + + +//#dce stupp +/* display controller offsets used for crtc/cur/lut/grph/viewport/etc. */ +#define SI_CRTC0_REGISTER_OFFSET 0 //(0x6df0 - 0x6df0)/4 +#define SI_CRTC1_REGISTER_OFFSET 0x300 //(0x79f0 - 0x6df0)/4 +#define SI_CRTC2_REGISTER_OFFSET 0x2600 //(0x105f0 - 0x6df0)/4 +#define SI_CRTC3_REGISTER_OFFSET 0x2900 //(0x111f0 - 0x6df0)/4 +#define SI_CRTC4_REGISTER_OFFSET 0x2c00 //(0x11df0 - 0x6df0)/4 +#define SI_CRTC5_REGISTER_OFFSET 0x2f00 //(0x129f0 - 0x6df0)/4 + +#define CURSOR_WIDTH 64 +#define CURSOR_HEIGHT 64 +#define AMDGPU_MM_INDEX 0x0000 +#define AMDGPU_MM_DATA 0x0001 + +#define VERDE_NUM_CRTC 6 +#define BLACKOUT_MODE_MASK 0x00000007 +#define VGA_RENDER_CONTROL 0xC0 +#define R_000300_VGA_RENDER_CONTROL 0xC0 +#define C_000300_VGA_VSTATUS_CNTL 0xFFFCFFFF +#define EVERGREEN_CRTC_STATUS 0x1BA3 +#define EVERGREEN_CRTC_V_BLANK (1 << 0) +#define EVERGREEN_CRTC_STATUS_POSITION 0x1BA4 +/* CRTC blocks at 0x6df0, 0x79f0, 0x105f0, 0x111f0, 0x11df0, 0x129f0 */ +#define EVERGREEN_CRTC_V_BLANK_START_END 0x1b8d +#define EVERGREEN_CRTC_CONTROL 0x1b9c +#define EVERGREEN_CRTC_MASTER_EN (1 << 0) +#define EVERGREEN_CRTC_DISP_READ_REQUEST_DISABLE (1 << 24) +#define EVERGREEN_CRTC_BLANK_CONTROL 0x1b9d +#define EVERGREEN_CRTC_BLANK_DATA_EN (1 << 8) +#define EVERGREEN_CRTC_V_BLANK (1 << 0) +#define EVERGREEN_CRTC_STATUS_HV_COUNT 0x1ba8 +#define EVERGREEN_CRTC_UPDATE_LOCK 0x1bb5 +#define EVERGREEN_MASTER_UPDATE_LOCK 0x1bbd +#define EVERGREEN_MASTER_UPDATE_MODE 0x1bbe +#define EVERGREEN_GRPH_UPDATE_LOCK (1 << 16) +#define EVERGREEN_GRPH_PRIMARY_SURFACE_ADDRESS_HIGH 0x1a07 +#define EVERGREEN_GRPH_SECONDARY_SURFACE_ADDRESS_HIGH 0x1a08 +#define EVERGREEN_GRPH_PRIMARY_SURFACE_ADDRESS 0x1a04 +#define EVERGREEN_GRPH_SECONDARY_SURFACE_ADDRESS 0x1a05 +#define EVERGREEN_GRPH_UPDATE 0x1a11 +#define EVERGREEN_VGA_MEMORY_BASE_ADDRESS 0xc4 +#define EVERGREEN_VGA_MEMORY_BASE_ADDRESS_HIGH 0xc9 +#define EVERGREEN_GRPH_SURFACE_UPDATE_PENDING (1 << 2) + +#define EVERGREEN_DATA_FORMAT 0x1ac0 +# define EVERGREEN_INTERLEAVE_EN (1 << 0) + +#define MC_SHARED_CHMAP__NOOFCHAN_MASK 0xf000 +#define MC_SHARED_CHMAP__NOOFCHAN__SHIFT 0xc + +#define R600_D1GRPH_ARRAY_MODE_LINEAR_GENERAL (0 << 20) +#define R600_D1GRPH_ARRAY_MODE_LINEAR_ALIGNED (1 << 20) +#define R600_D1GRPH_ARRAY_MODE_1D_TILED_THIN1 (2 << 20) +#define R600_D1GRPH_ARRAY_MODE_2D_TILED_THIN1 (4 << 20) + +#define R700_D1GRPH_PRIMARY_SURFACE_ADDRESS_HIGH 0x1a45 +#define R700_D2GRPH_PRIMARY_SURFACE_ADDRESS_HIGH 0x1845 + +#define R700_D2GRPH_SECONDARY_SURFACE_ADDRESS_HIGH 0x1847 +#define R700_D1GRPH_SECONDARY_SURFACE_ADDRESS_HIGH 0x1a47 + +#define DISP_INTERRUPT_STATUS__LB_D1_VBLANK_INTERRUPT_MASK 0x8 +#define DISP_INTERRUPT_STATUS_CONTINUE__LB_D2_VBLANK_INTERRUPT_MASK 0x8 +#define DISP_INTERRUPT_STATUS_CONTINUE2__LB_D3_VBLANK_INTERRUPT_MASK 0x8 +#define DISP_INTERRUPT_STATUS_CONTINUE3__LB_D4_VBLANK_INTERRUPT_MASK 0x8 +#define DISP_INTERRUPT_STATUS_CONTINUE4__LB_D5_VBLANK_INTERRUPT_MASK 0x8 +#define DISP_INTERRUPT_STATUS_CONTINUE5__LB_D6_VBLANK_INTERRUPT_MASK 0x8 + +#define DISP_INTERRUPT_STATUS__LB_D1_VLINE_INTERRUPT_MASK 0x4 +#define DISP_INTERRUPT_STATUS_CONTINUE__LB_D2_VLINE_INTERRUPT_MASK 0x4 +#define DISP_INTERRUPT_STATUS_CONTINUE2__LB_D3_VLINE_INTERRUPT_MASK 0x4 +#define DISP_INTERRUPT_STATUS_CONTINUE3__LB_D4_VLINE_INTERRUPT_MASK 0x4 +#define DISP_INTERRUPT_STATUS_CONTINUE4__LB_D5_VLINE_INTERRUPT_MASK 0x4 +#define DISP_INTERRUPT_STATUS_CONTINUE5__LB_D6_VLINE_INTERRUPT_MASK 0x4 + +#define DISP_INTERRUPT_STATUS__DC_HPD1_INTERRUPT_MASK 0x20000 +#define DISP_INTERRUPT_STATUS_CONTINUE__DC_HPD2_INTERRUPT_MASK 0x20000 +#define DISP_INTERRUPT_STATUS_CONTINUE2__DC_HPD3_INTERRUPT_MASK 0x20000 +#define DISP_INTERRUPT_STATUS_CONTINUE3__DC_HPD4_INTERRUPT_MASK 0x20000 +#define DISP_INTERRUPT_STATUS_CONTINUE4__DC_HPD5_INTERRUPT_MASK 0x20000 +#define DISP_INTERRUPT_STATUS_CONTINUE5__DC_HPD6_INTERRUPT_MASK 0x20000 + +#define GRPH_INTERRUPT_STATUS__GRPH_PFLIP_INT_OCCURRED_MASK 0x1 +#define GRPH_INTERRUPT_STATUS__GRPH_PFLIP_INT_CLEAR_MASK 0x100 + +#define DC_HPD1_INT_CONTROL__DC_HPD1_INT_ACK_MASK 0x1 + +#define R600_D1GRPH_SWAP_CONTROL 0x1843 +#define R600_D1GRPH_SWAP_ENDIAN_NONE (0 << 0) +#define R600_D1GRPH_SWAP_ENDIAN_16BIT (1 << 0) +#define R600_D1GRPH_SWAP_ENDIAN_32BIT (2 << 0) +#define R600_D1GRPH_SWAP_ENDIAN_64BIT (3 << 0) + +#define AVIVO_D1VGA_CONTROL 0x00cc +# define AVIVO_DVGA_CONTROL_MODE_ENABLE (1 << 0) +# define AVIVO_DVGA_CONTROL_TIMING_SELECT (1 << 8) +# define AVIVO_DVGA_CONTROL_SYNC_POLARITY_SELECT (1 << 9) +# define AVIVO_DVGA_CONTROL_OVERSCAN_TIMING_SELECT (1 << 10) +# define AVIVO_DVGA_CONTROL_OVERSCAN_COLOR_EN (1 << 16) +# define AVIVO_DVGA_CONTROL_ROTATE (1 << 24) +#define AVIVO_D2VGA_CONTROL 0x00ce + +#define R600_BUS_CNTL 0x1508 +# define R600_BIOS_ROM_DIS (1 << 1) + +#define R600_ROM_CNTL 0x580 +# define R600_SCK_OVERWRITE (1 << 1) +# define R600_SCK_PRESCALE_CRYSTAL_CLK_SHIFT 28 +# define R600_SCK_PRESCALE_CRYSTAL_CLK_MASK (0xf << 28) + +#define GRPH_INTERRUPT_CONTROL__GRPH_PFLIP_INT_MASK_MASK 0x1 + +#define FMT_BIT_DEPTH_CONTROL 0x1bf2 +#define FMT_TRUNCATE_EN (1 << 0) +#define FMT_TRUNCATE_DEPTH (1 << 4) +#define FMT_SPATIAL_DITHER_EN (1 << 8) +#define FMT_SPATIAL_DITHER_MODE(x) ((x) << 9) +#define FMT_SPATIAL_DITHER_DEPTH (1 << 12) +#define FMT_FRAME_RANDOM_ENABLE (1 << 13) +#define FMT_RGB_RANDOM_ENABLE (1 << 14) +#define FMT_HIGHPASS_RANDOM_ENABLE (1 << 15) +#define FMT_TEMPORAL_DITHER_EN (1 << 16) +#define FMT_TEMPORAL_DITHER_DEPTH (1 << 20) +#define FMT_TEMPORAL_DITHER_OFFSET(x) ((x) << 21) +#define FMT_TEMPORAL_LEVEL (1 << 24) +#define FMT_TEMPORAL_DITHER_RESET (1 << 25) +#define FMT_25FRC_SEL(x) ((x) << 26) +#define FMT_50FRC_SEL(x) ((x) << 28) +#define FMT_75FRC_SEL(x) ((x) << 30) + +#define EVERGREEN_DC_LUT_CONTROL 0x1a80 +#define EVERGREEN_DC_LUT_BLACK_OFFSET_BLUE 0x1a81 +#define EVERGREEN_DC_LUT_BLACK_OFFSET_GREEN 0x1a82 +#define EVERGREEN_DC_LUT_BLACK_OFFSET_RED 0x1a83 +#define EVERGREEN_DC_LUT_WHITE_OFFSET_BLUE 0x1a84 +#define EVERGREEN_DC_LUT_WHITE_OFFSET_GREEN 0x1a85 +#define EVERGREEN_DC_LUT_WHITE_OFFSET_RED 0x1a86 +#define EVERGREEN_DC_LUT_30_COLOR 0x1a7c +#define EVERGREEN_DC_LUT_RW_INDEX 0x1a79 +#define EVERGREEN_DC_LUT_WRITE_EN_MASK 0x1a7e +#define EVERGREEN_DC_LUT_RW_MODE 0x1a78 + +#define EVERGREEN_GRPH_ENABLE 0x1a00 +#define EVERGREEN_GRPH_CONTROL 0x1a01 +#define EVERGREEN_GRPH_DEPTH(x) (((x) & 0x3) << 0) +#define EVERGREEN_GRPH_DEPTH_8BPP 0 +#define EVERGREEN_GRPH_DEPTH_16BPP 1 +#define EVERGREEN_GRPH_DEPTH_32BPP 2 +#define EVERGREEN_GRPH_NUM_BANKS(x) (((x) & 0x3) << 2) +#define EVERGREEN_ADDR_SURF_2_BANK 0 +#define EVERGREEN_ADDR_SURF_4_BANK 1 +#define EVERGREEN_ADDR_SURF_8_BANK 2 +#define EVERGREEN_ADDR_SURF_16_BANK 3 +#define EVERGREEN_GRPH_Z(x) (((x) & 0x3) << 4) +#define EVERGREEN_GRPH_BANK_WIDTH(x) (((x) & 0x3) << 6) +#define EVERGREEN_ADDR_SURF_BANK_WIDTH_1 0 +#define EVERGREEN_ADDR_SURF_BANK_WIDTH_2 1 +#define EVERGREEN_ADDR_SURF_BANK_WIDTH_4 2 +#define EVERGREEN_ADDR_SURF_BANK_WIDTH_8 3 +#define EVERGREEN_GRPH_FORMAT(x) (((x) & 0x7) << 8) + +#define EVERGREEN_GRPH_FORMAT_INDEXED 0 +#define EVERGREEN_GRPH_FORMAT_ARGB1555 0 +#define EVERGREEN_GRPH_FORMAT_ARGB565 1 +#define EVERGREEN_GRPH_FORMAT_ARGB4444 2 +#define EVERGREEN_GRPH_FORMAT_AI88 3 +#define EVERGREEN_GRPH_FORMAT_MONO16 4 +#define EVERGREEN_GRPH_FORMAT_BGRA5551 5 + +/* 32 BPP */ +#define EVERGREEN_GRPH_FORMAT_ARGB8888 0 +#define EVERGREEN_GRPH_FORMAT_ARGB2101010 1 +#define EVERGREEN_GRPH_FORMAT_32BPP_DIG 2 +#define EVERGREEN_GRPH_FORMAT_8B_ARGB2101010 3 +#define EVERGREEN_GRPH_FORMAT_BGRA1010102 4 +#define EVERGREEN_GRPH_FORMAT_8B_BGRA1010102 5 +#define EVERGREEN_GRPH_FORMAT_RGB111110 6 +#define EVERGREEN_GRPH_FORMAT_BGR101111 7 +#define EVERGREEN_GRPH_BANK_HEIGHT(x) (((x) & 0x3) << 11) +#define EVERGREEN_ADDR_SURF_BANK_HEIGHT_1 0 +#define EVERGREEN_ADDR_SURF_BANK_HEIGHT_2 1 +#define EVERGREEN_ADDR_SURF_BANK_HEIGHT_4 2 +#define EVERGREEN_ADDR_SURF_BANK_HEIGHT_8 3 +#define EVERGREEN_GRPH_TILE_SPLIT(x) (((x) & 0x7) << 13) +#define EVERGREEN_ADDR_SURF_TILE_SPLIT_64B 0 +#define EVERGREEN_ADDR_SURF_TILE_SPLIT_128B 1 +#define EVERGREEN_ADDR_SURF_TILE_SPLIT_256B 2 +#define EVERGREEN_ADDR_SURF_TILE_SPLIT_512B 3 +#define EVERGREEN_ADDR_SURF_TILE_SPLIT_1KB 4 +#define EVERGREEN_ADDR_SURF_TILE_SPLIT_2KB 5 +#define EVERGREEN_ADDR_SURF_TILE_SPLIT_4KB 6 +#define EVERGREEN_GRPH_MACRO_TILE_ASPECT(x) (((x) & 0x3) << 18) +#define EVERGREEN_ADDR_SURF_MACRO_TILE_ASPECT_1 0 +#define EVERGREEN_ADDR_SURF_MACRO_TILE_ASPECT_2 1 +#define EVERGREEN_ADDR_SURF_MACRO_TILE_ASPECT_4 2 +#define EVERGREEN_ADDR_SURF_MACRO_TILE_ASPECT_8 3 +#define EVERGREEN_GRPH_ARRAY_MODE(x) (((x) & 0x7) << 20) +#define EVERGREEN_GRPH_ARRAY_LINEAR_GENERAL 0 +#define EVERGREEN_GRPH_ARRAY_LINEAR_ALIGNED 1 +#define EVERGREEN_GRPH_ARRAY_1D_TILED_THIN1 2 +#define EVERGREEN_GRPH_ARRAY_2D_TILED_THIN1 4 +#define EVERGREEN_ADDR_SURF_MACRO_TILE_ASPECT_1 0 +#define EVERGREEN_ADDR_SURF_MACRO_TILE_ASPECT_2 1 +#define EVERGREEN_ADDR_SURF_MACRO_TILE_ASPECT_4 2 +#define EVERGREEN_ADDR_SURF_MACRO_TILE_ASPECT_8 3 + +#define EVERGREEN_GRPH_SWAP_CONTROL 0x1a03 +#define EVERGREEN_GRPH_ENDIAN_SWAP(x) (((x) & 0x3) << 0) +# define EVERGREEN_GRPH_ENDIAN_NONE 0 +# define EVERGREEN_GRPH_ENDIAN_8IN16 1 +# define EVERGREEN_GRPH_ENDIAN_8IN32 2 +# define EVERGREEN_GRPH_ENDIAN_8IN64 3 + +#define EVERGREEN_D3VGA_CONTROL 0xf8 +#define EVERGREEN_D4VGA_CONTROL 0xf9 +#define EVERGREEN_D5VGA_CONTROL 0xfa +#define EVERGREEN_D6VGA_CONTROL 0xfb + +#define EVERGREEN_GRPH_SURFACE_ADDRESS_MASK 0xffffff00 + +#define EVERGREEN_GRPH_LUT_10BIT_BYPASS_CONTROL 0x1a02 +#define EVERGREEN_LUT_10BIT_BYPASS_EN (1 << 8) + +#define EVERGREEN_GRPH_PITCH 0x1a06 +#define EVERGREEN_GRPH_PRIMARY_SURFACE_ADDRESS_HIGH 0x1a07 +#define EVERGREEN_GRPH_SECONDARY_SURFACE_ADDRESS_HIGH 0x1a08 +#define EVERGREEN_GRPH_SURFACE_OFFSET_X 0x1a09 +#define EVERGREEN_GRPH_SURFACE_OFFSET_Y 0x1a0a +#define EVERGREEN_GRPH_X_START 0x1a0b +#define EVERGREEN_GRPH_Y_START 0x1a0c +#define EVERGREEN_GRPH_X_END 0x1a0d +#define EVERGREEN_GRPH_Y_END 0x1a0e +#define EVERGREEN_GRPH_UPDATE 0x1a11 +#define EVERGREEN_GRPH_SURFACE_UPDATE_PENDING (1 << 2) +#define EVERGREEN_GRPH_UPDATE_LOCK (1 << 16) +#define EVERGREEN_GRPH_FLIP_CONTROL 0x1a12 +#define EVERGREEN_GRPH_SURFACE_UPDATE_H_RETRACE_EN (1 << 0) + +#define EVERGREEN_VIEWPORT_START 0x1b5c +#define EVERGREEN_VIEWPORT_SIZE 0x1b5d +#define EVERGREEN_DESKTOP_HEIGHT 0x1ac1 + +/* CUR blocks at 0x6998, 0x7598, 0x10198, 0x10d98, 0x11998, 0x12598 */ +#define EVERGREEN_CUR_CONTROL 0x1a66 +# define EVERGREEN_CURSOR_EN (1 << 0) +# define EVERGREEN_CURSOR_MODE(x) (((x) & 0x3) << 8) +# define EVERGREEN_CURSOR_MONO 0 +# define EVERGREEN_CURSOR_24_1 1 +# define EVERGREEN_CURSOR_24_8_PRE_MULT 2 +# define EVERGREEN_CURSOR_24_8_UNPRE_MULT 3 +# define EVERGREEN_CURSOR_2X_MAGNIFY (1 << 16) +# define EVERGREEN_CURSOR_FORCE_MC_ON (1 << 20) +# define EVERGREEN_CURSOR_URGENT_CONTROL(x) (((x) & 0x7) << 24) +# define EVERGREEN_CURSOR_URGENT_ALWAYS 0 +# define EVERGREEN_CURSOR_URGENT_1_8 1 +# define EVERGREEN_CURSOR_URGENT_1_4 2 +# define EVERGREEN_CURSOR_URGENT_3_8 3 +# define EVERGREEN_CURSOR_URGENT_1_2 4 +#define EVERGREEN_CUR_SURFACE_ADDRESS 0x1a67 +# define EVERGREEN_CUR_SURFACE_ADDRESS_MASK 0xfffff000 +#define EVERGREEN_CUR_SIZE 0x1a68 +#define EVERGREEN_CUR_SURFACE_ADDRESS_HIGH 0x1a69 +#define EVERGREEN_CUR_POSITION 0x1a6a +#define EVERGREEN_CUR_HOT_SPOT 0x1a6b +#define EVERGREEN_CUR_COLOR1 0x1a6c +#define EVERGREEN_CUR_COLOR2 0x1a6d +#define EVERGREEN_CUR_UPDATE 0x1a6e +# define EVERGREEN_CURSOR_UPDATE_PENDING (1 << 0) +# define EVERGREEN_CURSOR_UPDATE_TAKEN (1 << 1) +# define EVERGREEN_CURSOR_UPDATE_LOCK (1 << 16) +# define EVERGREEN_CURSOR_DISABLE_MULTIPLE_UPDATE (1 << 24) + + +#define NI_INPUT_CSC_CONTROL 0x1a35 +# define NI_INPUT_CSC_GRPH_MODE(x) (((x) & 0x3) << 0) +# define NI_INPUT_CSC_BYPASS 0 +# define NI_INPUT_CSC_PROG_COEFF 1 +# define NI_INPUT_CSC_PROG_SHARED_MATRIXA 2 +# define NI_INPUT_CSC_OVL_MODE(x) (((x) & 0x3) << 4) + +#define NI_OUTPUT_CSC_CONTROL 0x1a3c +# define NI_OUTPUT_CSC_GRPH_MODE(x) (((x) & 0x7) << 0) +# define NI_OUTPUT_CSC_BYPASS 0 +# define NI_OUTPUT_CSC_TV_RGB 1 +# define NI_OUTPUT_CSC_YCBCR_601 2 +# define NI_OUTPUT_CSC_YCBCR_709 3 +# define NI_OUTPUT_CSC_PROG_COEFF 4 +# define NI_OUTPUT_CSC_PROG_SHARED_MATRIXB 5 +# define NI_OUTPUT_CSC_OVL_MODE(x) (((x) & 0x7) << 4) + +#define NI_DEGAMMA_CONTROL 0x1a58 +# define NI_GRPH_DEGAMMA_MODE(x) (((x) & 0x3) << 0) +# define NI_DEGAMMA_BYPASS 0 +# define NI_DEGAMMA_SRGB_24 1 +# define NI_DEGAMMA_XVYCC_222 2 +# define NI_OVL_DEGAMMA_MODE(x) (((x) & 0x3) << 4) +# define NI_ICON_DEGAMMA_MODE(x) (((x) & 0x3) << 8) +# define NI_CURSOR_DEGAMMA_MODE(x) (((x) & 0x3) << 12) + +#define NI_GAMUT_REMAP_CONTROL 0x1a59 +# define NI_GRPH_GAMUT_REMAP_MODE(x) (((x) & 0x3) << 0) +# define NI_GAMUT_REMAP_BYPASS 0 +# define NI_GAMUT_REMAP_PROG_COEFF 1 +# define NI_GAMUT_REMAP_PROG_SHARED_MATRIXA 2 +# define NI_GAMUT_REMAP_PROG_SHARED_MATRIXB 3 +# define NI_OVL_GAMUT_REMAP_MODE(x) (((x) & 0x3) << 4) + +#define NI_REGAMMA_CONTROL 0x1aa0 +# define NI_GRPH_REGAMMA_MODE(x) (((x) & 0x7) << 0) +# define NI_REGAMMA_BYPASS 0 +# define NI_REGAMMA_SRGB_24 1 +# define NI_REGAMMA_XVYCC_222 2 +# define NI_REGAMMA_PROG_A 3 +# define NI_REGAMMA_PROG_B 4 +# define NI_OVL_REGAMMA_MODE(x) (((x) & 0x7) << 4) + + +#define NI_PRESCALE_GRPH_CONTROL 0x1a2d +# define NI_GRPH_PRESCALE_BYPASS (1 << 4) + +#define NI_PRESCALE_OVL_CONTROL 0x1a31 +# define NI_OVL_PRESCALE_BYPASS (1 << 4) + +#define NI_INPUT_GAMMA_CONTROL 0x1a10 +# define NI_GRPH_INPUT_GAMMA_MODE(x) (((x) & 0x3) << 0) +# define NI_INPUT_GAMMA_USE_LUT 0 +# define NI_INPUT_GAMMA_BYPASS 1 +# define NI_INPUT_GAMMA_SRGB_24 2 +# define NI_INPUT_GAMMA_XVYCC_222 3 +# define NI_OVL_INPUT_GAMMA_MODE(x) (((x) & 0x3) << 4) + +#define IH_RB_WPTR__RB_OVERFLOW_MASK 0x1 +#define IH_RB_CNTL__WPTR_OVERFLOW_CLEAR_MASK 0x80000000 +#define SRBM_STATUS__IH_BUSY_MASK 0x20000 +#define SRBM_SOFT_RESET__SOFT_RESET_IH_MASK 0x400 + +#define BLACKOUT_MODE_MASK 0x00000007 +#define VGA_RENDER_CONTROL 0xC0 +#define R_000300_VGA_RENDER_CONTROL 0xC0 +#define C_000300_VGA_VSTATUS_CNTL 0xFFFCFFFF +#define EVERGREEN_CRTC_STATUS 0x1BA3 +#define EVERGREEN_CRTC_V_BLANK (1 << 0) +#define EVERGREEN_CRTC_STATUS_POSITION 0x1BA4 +/* CRTC blocks at 0x6df0, 0x79f0, 0x105f0, 0x111f0, 0x11df0, 0x129f0 */ +#define EVERGREEN_CRTC_V_BLANK_START_END 0x1b8d +#define EVERGREEN_CRTC_CONTROL 0x1b9c +# define EVERGREEN_CRTC_MASTER_EN (1 << 0) +# define EVERGREEN_CRTC_DISP_READ_REQUEST_DISABLE (1 << 24) +#define EVERGREEN_CRTC_BLANK_CONTROL 0x1b9d +# define EVERGREEN_CRTC_BLANK_DATA_EN (1 << 8) +# define EVERGREEN_CRTC_V_BLANK (1 << 0) +#define EVERGREEN_CRTC_STATUS_HV_COUNT 0x1ba8 +#define EVERGREEN_CRTC_UPDATE_LOCK 0x1bb5 +#define EVERGREEN_MASTER_UPDATE_LOCK 0x1bbd +#define EVERGREEN_MASTER_UPDATE_MODE 0x1bbe +#define EVERGREEN_GRPH_UPDATE_LOCK (1 << 16) +#define EVERGREEN_GRPH_PRIMARY_SURFACE_ADDRESS_HIGH 0x1a07 +#define EVERGREEN_GRPH_SECONDARY_SURFACE_ADDRESS_HIGH 0x1a08 +#define EVERGREEN_GRPH_PRIMARY_SURFACE_ADDRESS 0x1a04 +#define EVERGREEN_GRPH_SECONDARY_SURFACE_ADDRESS 0x1a05 +#define EVERGREEN_GRPH_UPDATE 0x1a11 +#define EVERGREEN_VGA_MEMORY_BASE_ADDRESS 0xc4 +#define EVERGREEN_VGA_MEMORY_BASE_ADDRESS_HIGH 0xc9 +#define EVERGREEN_GRPH_SURFACE_UPDATE_PENDING (1 << 2) + +#define mmVM_CONTEXT1_CNTL__xxRANGE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x10 +#define mmVM_CONTEXT1_CNTL__xxRANGE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x4 +#define mmVM_CONTEXT1_CNTL__xxDUMMY_PAGE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x80 +#define mmVM_CONTEXT1_CNTL__xxDUMMY_PAGE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x7 +#define mmVM_CONTEXT1_CNTL__xxPDE0_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x400 +#define mmVM_CONTEXT1_CNTL__xxPDE0_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0xa +#define mmVM_CONTEXT1_CNTL__xxVALID_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x2000 +#define mmVM_CONTEXT1_CNTL__xxVALID_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0xd +#define mmVM_CONTEXT1_CNTL__xxREAD_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x10000 +#define mmVM_CONTEXT1_CNTL__xxREAD_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x10 +#define mmVM_CONTEXT1_CNTL__xxWRITE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x80000 +#define mmVM_CONTEXT1_CNTL__xxWRITE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x13 + +#define mmVM_CONTEXT1_PROTECTION_FAULT_STATUS__xxVMID_MASK 0x1e000000 +#define mmVM_CONTEXT1_PROTECTION_FAULT_STATUS__xxVMID__SHIFT 0x19 +#define mmVM_CONTEXT1_PROTECTION_FAULT_STATUS__xxPROTECTIONS_MASK 0xff +#define mmVM_CONTEXT1_PROTECTION_FAULT_STATUS__xxPROTECTIONS__SHIFT 0x0 +#define mmVM_CONTEXT1_PROTECTION_FAULT_STATUS__xxMEMORY_CLIENT_ID_MASK 0xff000 +#define mmVM_CONTEXT1_PROTECTION_FAULT_STATUS__xxMEMORY_CLIENT_ID__SHIFT 0xc +#define mmVM_CONTEXT1_PROTECTION_FAULT_STATUS__xxMEMORY_CLIENT_RW_MASK 0x1000000 +#define mmVM_CONTEXT1_PROTECTION_FAULT_STATUS__xxMEMORY_CLIENT_RW__SHIFT 0x18 + +#define mmMC_SHARED_BLACKOUT_CNTL__xxBLACKOUT_MODE_MASK 0x7 +#define mmMC_SHARED_BLACKOUT_CNTL__xxBLACKOUT_MODE__SHIFT 0x0 + +#define mmBIF_FB_EN__xxFB_READ_EN_MASK 0x1 +#define mmBIF_FB_EN__xxFB_READ_EN__SHIFT 0x0 +#define mmBIF_FB_EN__xxFB_WRITE_EN_MASK 0x2 +#define mmBIF_FB_EN__xxFB_WRITE_EN__SHIFT 0x1 + +#define mmSRBM_SOFT_RESET__xxSOFT_RESET_VMC_MASK 0x20000 +#define mmSRBM_SOFT_RESET__xxSOFT_RESET_VMC__SHIFT 0x11 +#define mmSRBM_SOFT_RESET__xxSOFT_RESET_MC_MASK 0x800 +#define mmSRBM_SOFT_RESET__xxSOFT_RESET_MC__SHIFT 0xb + +#define VM_CONTEXT1_CNTL__RANGE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x8 +#define VM_CONTEXT1_CNTL__RANGE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x3 +#define VM_CONTEXT1_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x40 +#define VM_CONTEXT1_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x6 +#define VM_CONTEXT1_CNTL__PDE0_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x200 +#define VM_CONTEXT1_CNTL__PDE0_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x9 +#define VM_CONTEXT1_CNTL__VALID_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x1000 +#define VM_CONTEXT1_CNTL__VALID_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0xc +#define VM_CONTEXT1_CNTL__READ_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x8000 +#define VM_CONTEXT1_CNTL__READ_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0xf +#define VM_CONTEXT1_CNTL__WRITE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x40000 +#define VM_CONTEXT1_CNTL__WRITE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x12 + +#define MC_SEQ_MISC0__MT__MASK 0xf0000000 +#define MC_SEQ_MISC0__MT__GDDR1 0x10000000 +#define MC_SEQ_MISC0__MT__DDR2 0x20000000 +#define MC_SEQ_MISC0__MT__GDDR3 0x30000000 +#define MC_SEQ_MISC0__MT__GDDR4 0x40000000 +#define MC_SEQ_MISC0__MT__GDDR5 0x50000000 +#define MC_SEQ_MISC0__MT__HBM 0x60000000 +#define MC_SEQ_MISC0__MT__DDR3 0xB0000000 + +#define SRBM_STATUS__MCB_BUSY_MASK 0x200 +#define SRBM_STATUS__MCB_BUSY__SHIFT 0x9 +#define SRBM_STATUS__MCB_NON_DISPLAY_BUSY_MASK 0x400 +#define SRBM_STATUS__MCB_NON_DISPLAY_BUSY__SHIFT 0xa +#define SRBM_STATUS__MCC_BUSY_MASK 0x800 +#define SRBM_STATUS__MCC_BUSY__SHIFT 0xb +#define SRBM_STATUS__MCD_BUSY_MASK 0x1000 +#define SRBM_STATUS__MCD_BUSY__SHIFT 0xc +#define SRBM_STATUS__VMC_BUSY_MASK 0x100 +#define SRBM_STATUS__VMC_BUSY__SHIFT 0x8 + + +#define GRBM_STATUS__GUI_ACTIVE_MASK 0x80000000 +#define CP_INT_CNTL_RING__TIME_STAMP_INT_ENABLE_MASK 0x4000000 +#define CP_INT_CNTL_RING0__PRIV_REG_INT_ENABLE_MASK 0x800000 +#define CP_INT_CNTL_RING0__PRIV_INSTR_INT_ENABLE_MASK 0x400000 +#define PACKET3_SEM_WAIT_ON_SIGNAL (0x1 << 12) +#define PACKET3_SEM_SEL_SIGNAL (0x6 << 29) +#define PACKET3_SEM_SEL_WAIT (0x7 << 29) + +#define CONFIG_CNTL 0x1509 +#define CC_DRM_ID_STRAPS 0X1559 +#define AMDGPU_PCIE_INDEX 0xc +#define AMDGPU_PCIE_DATA 0xd + +#define DMA_SEM_INCOMPLETE_TIMER_CNTL 0x3411 +#define DMA_SEM_WAIT_FAIL_TIMER_CNTL 0x3412 +#define DMA_MODE 0x342f +#define DMA_RB_RPTR_ADDR_HI 0x3407 +#define DMA_RB_RPTR_ADDR_LO 0x3408 +#define DMA_BUSY_MASK 0x20 +#define DMA1_BUSY_MASK 0X40 +#define SDMA_MAX_INSTANCE 2 + +#define PCIE_BUS_CLK 10000 +#define TCLK (PCIE_BUS_CLK / 10) +#define CC_DRM_ID_STRAPS__ATI_REV_ID_MASK 0xf0000000 +#define CC_DRM_ID_STRAPS__ATI_REV_ID__SHIFT 0x1c +#define PCIE_PORT_INDEX 0xe +#define PCIE_PORT_DATA 0xf +#define EVERGREEN_PIF_PHY0_INDEX 0x8 +#define EVERGREEN_PIF_PHY0_DATA 0xc +#define EVERGREEN_PIF_PHY1_INDEX 0x10 +#define EVERGREEN_PIF_PHY1_DATA 0x14 + +#define MC_VM_FB_OFFSET 0x81a + +#endif diff --git a/drivers/gpu/drm/amd/include/atombios.h b/drivers/gpu/drm/amd/include/atombios.h index 3493da5c8f0e..4a4d3797a6d3 100644 --- a/drivers/gpu/drm/amd/include/atombios.h +++ b/drivers/gpu/drm/amd/include/atombios.h @@ -494,6 +494,7 @@ typedef struct _COMPUTE_MEMORY_ENGINE_PLL_PARAMETERS_V3 union { ATOM_COMPUTE_CLOCK_FREQ ulClock; //Input Parameter + ULONG ulClockParams; //ULONG access for BE ATOM_S_MPLL_FB_DIVIDER ulFbDiv; //Output Parameter }; UCHAR ucRefDiv; //Output Parameter @@ -526,6 +527,7 @@ typedef struct _COMPUTE_MEMORY_ENGINE_PLL_PARAMETERS_V5 union { ATOM_COMPUTE_CLOCK_FREQ ulClock; //Input Parameter + ULONG ulClockParams; //ULONG access for BE ATOM_S_MPLL_FB_DIVIDER ulFbDiv; //Output Parameter }; UCHAR ucRefDiv; //Output Parameter diff --git a/drivers/gpu/drm/amd/include/cgs_common.h b/drivers/gpu/drm/amd/include/cgs_common.h index b86aba9d019f..6aa8938fd826 100644 --- a/drivers/gpu/drm/amd/include/cgs_common.h +++ b/drivers/gpu/drm/amd/include/cgs_common.h @@ -119,6 +119,8 @@ enum cgs_system_info_id { CGS_SYSTEM_INFO_PG_FLAGS, CGS_SYSTEM_INFO_GFX_CU_INFO, CGS_SYSTEM_INFO_GFX_SE_INFO, + CGS_SYSTEM_INFO_PCIE_SUB_SYS_ID, + CGS_SYSTEM_INFO_PCIE_SUB_SYS_VENDOR_ID, CGS_SYSTEM_INFO_ID_MAXIMUM, }; diff --git a/drivers/gpu/drm/amd/powerplay/amd_powerplay.c b/drivers/gpu/drm/amd/powerplay/amd_powerplay.c index 2de34a5a85c2..b1d19409bf86 100644 --- a/drivers/gpu/drm/amd/powerplay/amd_powerplay.c +++ b/drivers/gpu/drm/amd/powerplay/amd_powerplay.c @@ -538,7 +538,6 @@ int pp_dpm_dispatch_tasks(void *handle, enum amd_pp_event event_id, void *input, ret = pem_handle_event(pp_handle->eventmgr, event_id, &data); break; case AMD_PP_EVENT_READJUST_POWER_STATE: - pp_handle->hwmgr->current_ps = pp_handle->hwmgr->boot_ps; ret = pem_handle_event(pp_handle->eventmgr, event_id, &data); break; default: @@ -765,15 +764,12 @@ static int pp_dpm_set_pp_table(void *handle, const char *buf, size_t size) PP_CHECK_HW(hwmgr); if (!hwmgr->hardcode_pp_table) { - hwmgr->hardcode_pp_table = - kzalloc(hwmgr->soft_pp_table_size, GFP_KERNEL); + hwmgr->hardcode_pp_table = kmemdup(hwmgr->soft_pp_table, + hwmgr->soft_pp_table_size, + GFP_KERNEL); if (!hwmgr->hardcode_pp_table) return -ENOMEM; - - /* to avoid powerplay crash when hardcode pptable is empty */ - memcpy(hwmgr->hardcode_pp_table, hwmgr->soft_pp_table, - hwmgr->soft_pp_table_size); } memcpy(hwmgr->hardcode_pp_table, buf, size); diff --git a/drivers/gpu/drm/amd/powerplay/eventmgr/psm.c b/drivers/gpu/drm/amd/powerplay/eventmgr/psm.c index a46225c0fc01..1d1875a7cb2d 100644 --- a/drivers/gpu/drm/amd/powerplay/eventmgr/psm.c +++ b/drivers/gpu/drm/amd/powerplay/eventmgr/psm.c @@ -70,11 +70,12 @@ int psm_set_states(struct pp_eventmgr *eventmgr, unsigned long *state_id) int i; table_entries = hwmgr->num_ps; + state = hwmgr->ps; for (i = 0; i < table_entries; i++) { if (state->id == *state_id) { - hwmgr->request_ps = state; + memcpy(hwmgr->request_ps, state, hwmgr->ps_size); return 0; } state = (struct pp_power_state *)((unsigned long)state + hwmgr->ps_size); @@ -106,7 +107,7 @@ int psm_adjust_power_state_dynamic(struct pp_eventmgr *eventmgr, bool skip) if (!equal || phm_check_smc_update_required_for_display_configuration(hwmgr)) { phm_apply_state_adjust_rules(hwmgr, requested, pcurrent); phm_set_power_state(hwmgr, &pcurrent->hardware, &requested->hardware); - hwmgr->current_ps = requested; + memcpy(hwmgr->current_ps, hwmgr->request_ps, hwmgr->ps_size); } return 0; } diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/Makefile b/drivers/gpu/drm/amd/powerplay/hwmgr/Makefile index abbcbc9f6eca..6e359c90dfda 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/Makefile +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/Makefile @@ -5,7 +5,7 @@ HARDWARE_MGR = hwmgr.o processpptables.o functiontables.o \ hardwaremanager.o pp_acpi.o cz_hwmgr.o \ cz_clockpowergating.o tonga_powertune.o\ - tonga_processpptables.o ppatomctrl.o \ + process_pptables_v1_0.o ppatomctrl.o \ tonga_hwmgr.o pppcielanes.o tonga_thermal.o\ fiji_powertune.o fiji_hwmgr.o tonga_clockpowergating.o \ fiji_clockpowergating.o fiji_thermal.o \ diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/fiji_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/fiji_hwmgr.c index 9368e21f5695..74300d6ef686 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/fiji_hwmgr.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/fiji_hwmgr.c @@ -44,8 +44,8 @@ #include "dce/dce_10_0_sh_mask.h" #include "pppcielanes.h" #include "fiji_hwmgr.h" -#include "tonga_processpptables.h" -#include "tonga_pptable.h" +#include "process_pptables_v1_0.h" +#include "pptable_v1_0.h" #include "pp_debug.h" #include "pp_acpi.h" #include "amd_pcie_helpers.h" @@ -112,7 +112,7 @@ static const uint8_t fiji_clock_stretch_amount_conversion[2][6] = static const unsigned long PhwFiji_Magic = (unsigned long)(PHM_VIslands_Magic); -struct fiji_power_state *cast_phw_fiji_power_state( +static struct fiji_power_state *cast_phw_fiji_power_state( struct pp_hw_power_state *hw_ps) { PP_ASSERT_WITH_CODE((PhwFiji_Magic == hw_ps->magic), @@ -122,7 +122,8 @@ struct fiji_power_state *cast_phw_fiji_power_state( return (struct fiji_power_state *)hw_ps; } -const struct fiji_power_state *cast_const_phw_fiji_power_state( +static const struct +fiji_power_state *cast_const_phw_fiji_power_state( const struct pp_hw_power_state *hw_ps) { PP_ASSERT_WITH_CODE((PhwFiji_Magic == hw_ps->magic), @@ -1626,7 +1627,7 @@ static int fiji_setup_default_dpm_tables(struct pp_hwmgr *hwmgr) * @param voltage - voltage to look for * @return 0 on success */ -uint8_t fiji_get_voltage_index( +static uint8_t fiji_get_voltage_index( struct phm_ppt_v1_voltage_lookup_table *lookup_table, uint16_t voltage) { uint8_t count = (uint8_t) (lookup_table->count); @@ -1690,7 +1691,7 @@ static int fiji_populate_cac_table(struct pp_hwmgr *hwmgr, * @return always 0 */ -int fiji_populate_smc_voltage_tables(struct pp_hwmgr *hwmgr, +static int fiji_populate_smc_voltage_tables(struct pp_hwmgr *hwmgr, struct SMU73_Discrete_DpmTable *table) { int result; @@ -2301,7 +2302,7 @@ static int fiji_populate_all_memory_levels(struct pp_hwmgr *hwmgr) * @param mclk the MCLK value to be used in the decision if MVDD should be high or low. * @param voltage the SMC VOLTAGE structure to be populated */ -int fiji_populate_mvdd_value(struct pp_hwmgr *hwmgr, +static int fiji_populate_mvdd_value(struct pp_hwmgr *hwmgr, uint32_t mclk, SMIO_Pattern *smio_pat) { const struct fiji_hwmgr *data = (struct fiji_hwmgr *)(hwmgr->backend); @@ -4005,7 +4006,7 @@ static int fiji_get_pp_table_entry(struct pp_hwmgr *hwmgr, ps = (struct fiji_power_state *)(&state->hardware); - result = tonga_get_powerplay_table_entry(hwmgr, entry_index, state, + result = get_powerplay_table_entry_v1_0(hwmgr, entry_index, state, fiji_get_pp_table_entry_callback_func); /* This is the earliest time we have all the dependency table and the VBIOS boot state @@ -4622,7 +4623,7 @@ static int fiji_generate_dpm_level_enable_mask( return 0; } -int fiji_enable_disable_uvd_dpm(struct pp_hwmgr *hwmgr, bool enable) +static int fiji_enable_disable_uvd_dpm(struct pp_hwmgr *hwmgr, bool enable) { return smum_send_msg_to_smc(hwmgr->smumgr, enable ? (PPSMC_Msg)PPSMC_MSG_UVDDPM_Enable : @@ -4636,14 +4637,14 @@ int fiji_enable_disable_vce_dpm(struct pp_hwmgr *hwmgr, bool enable) PPSMC_MSG_VCEDPM_Disable); } -int fiji_enable_disable_samu_dpm(struct pp_hwmgr *hwmgr, bool enable) +static int fiji_enable_disable_samu_dpm(struct pp_hwmgr *hwmgr, bool enable) { return smum_send_msg_to_smc(hwmgr->smumgr, enable? PPSMC_MSG_SAMUDPM_Enable : PPSMC_MSG_SAMUDPM_Disable); } -int fiji_enable_disable_acp_dpm(struct pp_hwmgr *hwmgr, bool enable) +static int fiji_enable_disable_acp_dpm(struct pp_hwmgr *hwmgr, bool enable) { return smum_send_msg_to_smc(hwmgr->smumgr, enable? PPSMC_MSG_ACPDPM_Enable : @@ -4880,7 +4881,7 @@ static void fiji_apply_dal_minimum_voltage_request(struct pp_hwmgr *hwmgr) return; } -int fiji_upload_dpm_level_enable_mask(struct pp_hwmgr *hwmgr) +static int fiji_upload_dpm_level_enable_mask(struct pp_hwmgr *hwmgr) { int result; struct fiji_hwmgr *data = (struct fiji_hwmgr *)(hwmgr->backend); @@ -5156,7 +5157,7 @@ static int fiji_program_display_gap(struct pp_hwmgr *hwmgr) return 0; } -int fiji_display_configuration_changed_task(struct pp_hwmgr *hwmgr) +static int fiji_display_configuration_changed_task(struct pp_hwmgr *hwmgr) { return fiji_program_display_gap(hwmgr); } @@ -5187,7 +5188,7 @@ static int fiji_set_max_fan_rpm_output(struct pp_hwmgr *hwmgr, PPSMC_MSG_SetFanRpmMax, us_max_fan_rpm); } -int fiji_dpm_set_interrupt_state(void *private_data, +static int fiji_dpm_set_interrupt_state(void *private_data, unsigned src_id, unsigned type, int enabled) { @@ -5235,7 +5236,7 @@ int fiji_dpm_set_interrupt_state(void *private_data, return 0; } -int fiji_register_internal_thermal_interrupt(struct pp_hwmgr *hwmgr, +static int fiji_register_internal_thermal_interrupt(struct pp_hwmgr *hwmgr, const void *thermal_interrupt_info) { int result; @@ -5405,7 +5406,10 @@ static inline bool fiji_are_power_levels_equal(const struct fiji_performance_lev (pl1->pcie_lane == pl2->pcie_lane)); } -int fiji_check_states_equal(struct pp_hwmgr *hwmgr, const struct pp_hw_power_state *pstate1, const struct pp_hw_power_state *pstate2, bool *equal) +static int +fiji_check_states_equal(struct pp_hwmgr *hwmgr, + const struct pp_hw_power_state *pstate1, + const struct pp_hw_power_state *pstate2, bool *equal) { const struct fiji_power_state *psa = cast_const_phw_fiji_power_state(pstate1); const struct fiji_power_state *psb = cast_const_phw_fiji_power_state(pstate2); @@ -5437,7 +5441,8 @@ int fiji_check_states_equal(struct pp_hwmgr *hwmgr, const struct pp_hw_power_sta return 0; } -bool fiji_check_smc_update_required_for_display_configuration(struct pp_hwmgr *hwmgr) +static bool +fiji_check_smc_update_required_for_display_configuration(struct pp_hwmgr *hwmgr) { struct fiji_hwmgr *data = (struct fiji_hwmgr *)(hwmgr->backend); bool is_update_required = false; @@ -5547,7 +5552,7 @@ static const struct pp_hwmgr_func fiji_hwmgr_funcs = { .dynamic_state_management_enable = &fiji_enable_dpm_tasks, .dynamic_state_management_disable = &fiji_disable_dpm_tasks, .force_dpm_level = &fiji_dpm_force_dpm_level, - .get_num_of_pp_table_entries = &tonga_get_number_of_powerplay_table_entries, + .get_num_of_pp_table_entries = &get_number_of_powerplay_table_entries_v1_0, .get_power_state_size = &fiji_get_power_state_size, .get_pp_table_entry = &fiji_get_pp_table_entry, .patch_boot_state = &fiji_patch_boot_state, @@ -5589,7 +5594,7 @@ static const struct pp_hwmgr_func fiji_hwmgr_funcs = { int fiji_hwmgr_init(struct pp_hwmgr *hwmgr) { hwmgr->hwmgr_func = &fiji_hwmgr_funcs; - hwmgr->pptable_func = &tonga_pptable_funcs; + hwmgr->pptable_func = &pptable_v1_0_funcs; pp_fiji_thermal_initialize(hwmgr); return 0; } diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/fiji_thermal.c b/drivers/gpu/drm/amd/powerplay/hwmgr/fiji_thermal.c index 92976b68d6fd..7f431e762262 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/fiji_thermal.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/fiji_thermal.c @@ -152,7 +152,7 @@ int fiji_fan_ctrl_set_default_mode(struct pp_hwmgr *hwmgr) return 0; } -int fiji_fan_ctrl_start_smc_fan_control(struct pp_hwmgr *hwmgr) +static int fiji_fan_ctrl_start_smc_fan_control(struct pp_hwmgr *hwmgr) { int result; @@ -421,7 +421,7 @@ int fiji_thermal_stop_thermal_controller(struct pp_hwmgr *hwmgr) * @param Result the last failure code * @return result from set temperature range routine */ -int tf_fiji_thermal_setup_fan_table(struct pp_hwmgr *hwmgr, +static int tf_fiji_thermal_setup_fan_table(struct pp_hwmgr *hwmgr, void *input, void *output, void *storage, int result) { struct fiji_hwmgr *data = (struct fiji_hwmgr *)(hwmgr->backend); @@ -533,7 +533,7 @@ int tf_fiji_thermal_setup_fan_table(struct pp_hwmgr *hwmgr, * @param Result the last failure code * @return result from set temperature range routine */ -int tf_fiji_thermal_start_smc_fan_control(struct pp_hwmgr *hwmgr, +static int tf_fiji_thermal_start_smc_fan_control(struct pp_hwmgr *hwmgr, void *input, void *output, void *storage, int result) { /* If the fantable setup has failed we could have disabled diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/hardwaremanager.c b/drivers/gpu/drm/amd/powerplay/hwmgr/hardwaremanager.c index 789f98ad2615..14f8c1f4da3d 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/hardwaremanager.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/hardwaremanager.c @@ -24,8 +24,6 @@ #include "hwmgr.h" #include "hardwaremanager.h" #include "power_state.h" -#include "pp_acpi.h" -#include "amd_acpi.h" #include "pp_debug.h" #define PHM_FUNC_CHECK(hw) \ @@ -34,38 +32,6 @@ return -EINVAL; \ } while (0) -void phm_init_dynamic_caps(struct pp_hwmgr *hwmgr) -{ - phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_DisableVoltageTransition); - phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_DisableEngineTransition); - phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_DisableMemoryTransition); - phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_DisableMGClockGating); - phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_DisableMGCGTSSM); - phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_DisableLSClockGating); - phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_Force3DClockSupport); - phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_DisableLightSleep); - phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_DisableMCLS); - phm_cap_set(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_DisablePowerGating); - - phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_DisableDPM); - phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_DisableSMUUVDHandshake); - phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_ThermalAutoThrottling); - - phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_PCIEPerformanceRequest); - - phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_NoOD5Support); - phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_UserMaxClockForMultiDisplays); - - phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_VpuRecoveryInProgress); - - phm_cap_set(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_UVDDPM); - phm_cap_set(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_VCEDPM); - - if (acpi_atcs_functions_supported(hwmgr->device, ATCS_FUNCTION_PCIE_PERFORMANCE_REQUEST) && - acpi_atcs_functions_supported(hwmgr->device, ATCS_FUNCTION_PCIE_DEVICE_READY_NOTIFICATION)) - phm_cap_set(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_PCIEPerformanceRequest); -} - bool phm_is_hw_access_blocked(struct pp_hwmgr *hwmgr) { return hwmgr->block_hw_access; diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c index d829076ed9ea..524d0dd4f0e9 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c @@ -32,8 +32,8 @@ #include "pp_debug.h" #include "ppatomctrl.h" #include "ppsmc.h" - -#define VOLTAGE_SCALE 4 +#include "pp_acpi.h" +#include "amd_acpi.h" extern int cz_hwmgr_init(struct pp_hwmgr *hwmgr); extern int tonga_hwmgr_init(struct pp_hwmgr *hwmgr); @@ -41,23 +41,12 @@ extern int fiji_hwmgr_init(struct pp_hwmgr *hwmgr); extern int polaris10_hwmgr_init(struct pp_hwmgr *hwmgr); extern int iceland_hwmgr_init(struct pp_hwmgr *hwmgr); -static int hwmgr_set_features_platform_caps(struct pp_hwmgr *hwmgr) -{ - if (amdgpu_sclk_deep_sleep_en) - phm_cap_set(hwmgr->platform_descriptor.platformCaps, - PHM_PlatformCaps_SclkDeepSleep); - else - phm_cap_unset(hwmgr->platform_descriptor.platformCaps, - PHM_PlatformCaps_SclkDeepSleep); - - if (amdgpu_powercontainment) - phm_cap_set(hwmgr->platform_descriptor.platformCaps, - PHM_PlatformCaps_PowerContainment); - else - phm_cap_unset(hwmgr->platform_descriptor.platformCaps, - PHM_PlatformCaps_PowerContainment); +static void hwmgr_init_default_caps(struct pp_hwmgr *hwmgr); +static int hwmgr_set_user_specify_caps(struct pp_hwmgr *hwmgr); - return 0; +uint8_t convert_to_vid(uint16_t vddc) +{ + return (uint8_t) ((6200 - (vddc * VOLTAGE_SCALE)) / 25); } int hwmgr_init(struct amd_pp_init *pp_init, struct pp_instance *handle) @@ -76,13 +65,12 @@ int hwmgr_init(struct amd_pp_init *pp_init, struct pp_instance *handle) hwmgr->device = pp_init->device; hwmgr->chip_family = pp_init->chip_family; hwmgr->chip_id = pp_init->chip_id; - hwmgr->hw_revision = pp_init->rev_id; - hwmgr->sub_sys_id = pp_init->sub_sys_id; - hwmgr->sub_vendor_id = pp_init->sub_vendor_id; hwmgr->usec_timeout = AMD_MAX_USEC_TIMEOUT; hwmgr->power_source = PP_PowerSource_AC; + hwmgr->pp_table_version = PP_TABLE_V1; - hwmgr_set_features_platform_caps(hwmgr); + hwmgr_init_default_caps(hwmgr); + hwmgr_set_user_specify_caps(hwmgr); switch (hwmgr->chip_family) { case AMDGPU_FAMILY_CZ: @@ -111,8 +99,6 @@ int hwmgr_init(struct amd_pp_init *pp_init, struct pp_instance *handle) return -EINVAL; } - phm_init_dynamic_caps(hwmgr); - return 0; } @@ -131,6 +117,8 @@ int hwmgr_fini(struct pp_hwmgr *hwmgr) kfree(hwmgr->set_temperature_range.function_list); kfree(hwmgr->ps); + kfree(hwmgr->current_ps); + kfree(hwmgr->request_ps); kfree(hwmgr); return 0; } @@ -155,10 +143,17 @@ int hw_init_power_state_table(struct pp_hwmgr *hwmgr) sizeof(struct pp_power_state); hwmgr->ps = kzalloc(size * table_entries, GFP_KERNEL); - if (hwmgr->ps == NULL) return -ENOMEM; + hwmgr->request_ps = kzalloc(size, GFP_KERNEL); + if (hwmgr->request_ps == NULL) + return -ENOMEM; + + hwmgr->current_ps = kzalloc(size, GFP_KERNEL); + if (hwmgr->current_ps == NULL) + return -ENOMEM; + state = hwmgr->ps; for (i = 0; i < table_entries; i++) { @@ -166,7 +161,8 @@ int hw_init_power_state_table(struct pp_hwmgr *hwmgr) if (state->classification.flags & PP_StateClassificationFlag_Boot) { hwmgr->boot_ps = state; - hwmgr->current_ps = hwmgr->request_ps = state; + memcpy(hwmgr->current_ps, state, size); + memcpy(hwmgr->request_ps, state, size); } state->id = i + 1; /* assigned unique num for every power state id */ @@ -176,6 +172,7 @@ int hw_init_power_state_table(struct pp_hwmgr *hwmgr) state = (struct pp_power_state *)((unsigned long)state + size); } + return 0; } @@ -209,8 +206,6 @@ int phm_wait_on_register(struct pp_hwmgr *hwmgr, uint32_t index, } - - /** * Returns once the part of the register indicated by the mask has * reached the given value.The indirect space is described by giving @@ -452,6 +447,27 @@ uint8_t phm_get_voltage_index( return i - 1; } +uint8_t phm_get_voltage_id(pp_atomctrl_voltage_table *voltage_table, + uint32_t voltage) +{ + uint8_t count = (uint8_t) (voltage_table->count); + uint8_t i = 0; + + PP_ASSERT_WITH_CODE((NULL != voltage_table), + "Voltage Table empty.", return 0;); + PP_ASSERT_WITH_CODE((0 != count), + "Voltage Table empty.", return 0;); + + for (i = 0; i < count; i++) { + /* find first voltage bigger than requested */ + if (voltage_table->entries[i].value >= voltage) + return i; + } + + /* voltage is bigger than max voltage in the table */ + return i - 1; +} + uint16_t phm_find_closest_vddci(struct pp_atomctrl_voltage_table *vddci_table, uint16_t vddci) { uint32_t i; @@ -539,7 +555,8 @@ int phm_initializa_dynamic_state_adjustment_rule_settings(struct pp_hwmgr *hwmgr table_clk_vlt->entries[2].v = 810; table_clk_vlt->entries[3].clk = PP_DAL_POWERLEVEL_PERFORMANCE; table_clk_vlt->entries[3].v = 900; - pptable_info->vddc_dep_on_dal_pwrl = table_clk_vlt; + if (pptable_info != NULL) + pptable_info->vddc_dep_on_dal_pwrl = table_clk_vlt; hwmgr->dyn_state.vddc_dep_on_dal_pwrl = table_clk_vlt; } @@ -605,3 +622,94 @@ void phm_apply_dal_min_voltage_request(struct pp_hwmgr *hwmgr) printk(KERN_ERR "DAL requested level can not" " found a available voltage in VDDC DPM Table \n"); } + +void hwmgr_init_default_caps(struct pp_hwmgr *hwmgr) +{ + phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_DisableVoltageTransition); + phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_DisableEngineTransition); + phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_DisableMemoryTransition); + phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_DisableMGClockGating); + phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_DisableMGCGTSSM); + phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_DisableLSClockGating); + phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_Force3DClockSupport); + phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_DisableLightSleep); + phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_DisableMCLS); + phm_cap_set(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_DisablePowerGating); + + phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_DisableDPM); + phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_DisableSMUUVDHandshake); + phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_ThermalAutoThrottling); + + phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_PCIEPerformanceRequest); + + phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_NoOD5Support); + phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_UserMaxClockForMultiDisplays); + + phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_VpuRecoveryInProgress); + + phm_cap_set(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_UVDDPM); + phm_cap_set(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_VCEDPM); + + if (acpi_atcs_functions_supported(hwmgr->device, ATCS_FUNCTION_PCIE_PERFORMANCE_REQUEST) && + acpi_atcs_functions_supported(hwmgr->device, ATCS_FUNCTION_PCIE_DEVICE_READY_NOTIFICATION)) + phm_cap_set(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_PCIEPerformanceRequest); + + phm_cap_set(hwmgr->platform_descriptor.platformCaps, + PHM_PlatformCaps_DynamicPatchPowerState); + + phm_cap_set(hwmgr->platform_descriptor.platformCaps, + PHM_PlatformCaps_EnableSMU7ThermalManagement); + + phm_cap_set(hwmgr->platform_descriptor.platformCaps, + PHM_PlatformCaps_DynamicPowerManagement); + + phm_cap_set(hwmgr->platform_descriptor.platformCaps, + PHM_PlatformCaps_SMC); + + phm_cap_set(hwmgr->platform_descriptor.platformCaps, + PHM_PlatformCaps_DynamicUVDState); + + phm_cap_set(hwmgr->platform_descriptor.platformCaps, + PHM_PlatformCaps_FanSpeedInTableIsRPM); + + return; +} + +int hwmgr_set_user_specify_caps(struct pp_hwmgr *hwmgr) +{ + if (amdgpu_sclk_deep_sleep_en) + phm_cap_set(hwmgr->platform_descriptor.platformCaps, + PHM_PlatformCaps_SclkDeepSleep); + else + phm_cap_unset(hwmgr->platform_descriptor.platformCaps, + PHM_PlatformCaps_SclkDeepSleep); + + if (amdgpu_powercontainment) + phm_cap_set(hwmgr->platform_descriptor.platformCaps, + PHM_PlatformCaps_PowerContainment); + else + phm_cap_unset(hwmgr->platform_descriptor.platformCaps, + PHM_PlatformCaps_PowerContainment); + + hwmgr->feature_mask = amdgpu_pp_feature_mask; + + return 0; +} + +int phm_get_voltage_evv_on_sclk(struct pp_hwmgr *hwmgr, uint8_t voltage_type, + uint32_t sclk, uint16_t id, uint16_t *voltage) +{ + uint32_t vol; + int ret = 0; + + if (hwmgr->chip_id < CHIP_POLARIS10) { + atomctrl_get_voltage_evv_on_sclk(hwmgr, voltage_type, sclk, id, voltage); + if (*voltage >= 2000 || *voltage == 0) + *voltage = 1150; + } else { + ret = atomctrl_get_voltage_evv_on_sclk_ai(hwmgr, voltage_type, sclk, id, &vol); + *voltage = (uint16_t)vol/100; + } + return ret; +} + diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/iceland_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/iceland_hwmgr.c index 8a7ada50551c..5abe43360ec0 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/iceland_hwmgr.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/iceland_hwmgr.c @@ -781,7 +781,7 @@ static int iceland_upload_firmware(struct pp_hwmgr *hwmgr) * @param hwmgr the address of the powerplay hardware manager. * @return always 0 */ -int iceland_process_firmware_header(struct pp_hwmgr *hwmgr) +static int iceland_process_firmware_header(struct pp_hwmgr *hwmgr) { iceland_hwmgr *data = (iceland_hwmgr *)(hwmgr->backend); @@ -1355,14 +1355,6 @@ static int iceland_populate_smc_mvdd_table(struct pp_hwmgr *hwmgr, return 0; } -/** - * Convert a voltage value in mv unit to VID number required by SMU firmware - */ -static uint8_t convert_to_vid(uint16_t vddc) -{ - return (uint8_t) ((6200 - (vddc * VOLTAGE_SCALE)) / 25); -} - int iceland_populate_bapm_vddc_vid_sidd(struct pp_hwmgr *hwmgr) { int i; @@ -2606,7 +2598,7 @@ static int iceland_populate_smc_initial_state(struct pp_hwmgr *hwmgr) * @param pInput the pointer to input data (PowerState) * @return always 0 */ -int iceland_init_smc_table(struct pp_hwmgr *hwmgr) +static int iceland_init_smc_table(struct pp_hwmgr *hwmgr) { int result; iceland_hwmgr *data = (iceland_hwmgr *)(hwmgr->backend); @@ -4629,7 +4621,7 @@ static int iceland_update_vce_dpm(struct pp_hwmgr *hwmgr, const void *input) return 0; } -int iceland_update_sclk_threshold(struct pp_hwmgr *hwmgr) +static int iceland_update_sclk_threshold(struct pp_hwmgr *hwmgr) { iceland_hwmgr *data = (iceland_hwmgr *)(hwmgr->backend); diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/polaris10_clockpowergating.c b/drivers/gpu/drm/amd/powerplay/hwmgr/polaris10_clockpowergating.c index b5edb5105986..7e405b04c2c5 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/polaris10_clockpowergating.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/polaris10_clockpowergating.c @@ -31,7 +31,7 @@ int polaris10_phm_powerdown_uvd(struct pp_hwmgr *hwmgr) return 0; } -int polaris10_phm_powerup_uvd(struct pp_hwmgr *hwmgr) +static int polaris10_phm_powerup_uvd(struct pp_hwmgr *hwmgr) { if (phm_cf_want_uvd_power_gating(hwmgr)) { if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, @@ -47,7 +47,7 @@ int polaris10_phm_powerup_uvd(struct pp_hwmgr *hwmgr) return 0; } -int polaris10_phm_powerdown_vce(struct pp_hwmgr *hwmgr) +static int polaris10_phm_powerdown_vce(struct pp_hwmgr *hwmgr) { if (phm_cf_want_vce_power_gating(hwmgr)) return smum_send_msg_to_smc(hwmgr->smumgr, @@ -55,7 +55,7 @@ int polaris10_phm_powerdown_vce(struct pp_hwmgr *hwmgr) return 0; } -int polaris10_phm_powerup_vce(struct pp_hwmgr *hwmgr) +static int polaris10_phm_powerup_vce(struct pp_hwmgr *hwmgr) { if (phm_cf_want_vce_power_gating(hwmgr)) return smum_send_msg_to_smc(hwmgr->smumgr, @@ -63,7 +63,7 @@ int polaris10_phm_powerup_vce(struct pp_hwmgr *hwmgr) return 0; } -int polaris10_phm_powerdown_samu(struct pp_hwmgr *hwmgr) +static int polaris10_phm_powerdown_samu(struct pp_hwmgr *hwmgr) { if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_SamuPowerGating)) @@ -72,7 +72,7 @@ int polaris10_phm_powerdown_samu(struct pp_hwmgr *hwmgr) return 0; } -int polaris10_phm_powerup_samu(struct pp_hwmgr *hwmgr) +static int polaris10_phm_powerup_samu(struct pp_hwmgr *hwmgr) { if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_SamuPowerGating)) diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/polaris10_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/polaris10_hwmgr.c index b69132296672..191ed504effb 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/polaris10_hwmgr.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/polaris10_hwmgr.c @@ -33,11 +33,11 @@ #include "pp_debug.h" #include "ppatomctrl.h" #include "atombios.h" -#include "tonga_pptable.h" +#include "pptable_v1_0.h" #include "pppcielanes.h" #include "amd_pcie_helpers.h" #include "hardwaremanager.h" -#include "tonga_processpptables.h" +#include "process_pptables_v1_0.h" #include "cgs_common.h" #include "smu74.h" #include "smu_ucode_xfer_vi.h" @@ -108,7 +108,7 @@ enum DPM_EVENT_SRC { static const unsigned long PhwPolaris10_Magic = (unsigned long)(PHM_VIslands_Magic); -struct polaris10_power_state *cast_phw_polaris10_power_state( +static struct polaris10_power_state *cast_phw_polaris10_power_state( struct pp_hw_power_state *hw_ps) { PP_ASSERT_WITH_CODE((PhwPolaris10_Magic == hw_ps->magic), @@ -118,7 +118,8 @@ struct polaris10_power_state *cast_phw_polaris10_power_state( return (struct polaris10_power_state *)hw_ps; } -const struct polaris10_power_state *cast_const_phw_polaris10_power_state( +static const struct polaris10_power_state * +cast_const_phw_polaris10_power_state( const struct pp_hw_power_state *hw_ps) { PP_ASSERT_WITH_CODE((PhwPolaris10_Magic == hw_ps->magic), @@ -141,7 +142,7 @@ static bool polaris10_is_dpm_running(struct pp_hwmgr *hwmgr) * @param hwmgr the address of the powerplay hardware manager. * @return always 0 */ -int phm_get_mc_microcode_version (struct pp_hwmgr *hwmgr) +static int phm_get_mc_microcode_version(struct pp_hwmgr *hwmgr) { cgs_write_register(hwmgr->device, mmMC_SEQ_IO_DEBUG_INDEX, 0x9F); @@ -150,7 +151,7 @@ int phm_get_mc_microcode_version (struct pp_hwmgr *hwmgr) return 0; } -uint16_t phm_get_current_pcie_speed(struct pp_hwmgr *hwmgr) +static uint16_t phm_get_current_pcie_speed(struct pp_hwmgr *hwmgr) { uint32_t speedCntl = 0; @@ -161,7 +162,7 @@ uint16_t phm_get_current_pcie_speed(struct pp_hwmgr *hwmgr) PCIE_LC_SPEED_CNTL, LC_CURRENT_DATA_RATE)); } -int phm_get_current_pcie_lane_number(struct pp_hwmgr *hwmgr) +static int phm_get_current_pcie_lane_number(struct pp_hwmgr *hwmgr) { uint32_t link_width; @@ -181,7 +182,7 @@ int phm_get_current_pcie_lane_number(struct pp_hwmgr *hwmgr) * @param pHwMgr the address of the powerplay hardware manager. * @return always PP_Result_OK */ -int polaris10_enable_smc_voltage_controller(struct pp_hwmgr *hwmgr) +static int polaris10_enable_smc_voltage_controller(struct pp_hwmgr *hwmgr) { PP_ASSERT_WITH_CODE( (hwmgr->smumgr->smumgr_funcs->send_msg_to_smc(hwmgr->smumgr, PPSMC_MSG_Voltage_Cntl_Enable) == 0), @@ -661,7 +662,7 @@ static int polaris10_setup_default_pcie_table(struct pp_hwmgr *hwmgr) * on the power policy or external client requests, * such as UVD request, etc. */ -int polaris10_setup_default_dpm_tables(struct pp_hwmgr *hwmgr) +static int polaris10_setup_default_dpm_tables(struct pp_hwmgr *hwmgr) { struct polaris10_hwmgr *data = (struct polaris10_hwmgr *)(hwmgr->backend); struct phm_ppt_v1_information *table_info = @@ -735,11 +736,6 @@ int polaris10_setup_default_dpm_tables(struct pp_hwmgr *hwmgr) return 0; } -uint8_t convert_to_vid(uint16_t vddc) -{ - return (uint8_t) ((6200 - (vddc * VOLTAGE_SCALE)) / 25); -} - /** * Mvdd table preparation for SMC. * @@ -840,7 +836,7 @@ static int polaris10_populate_cac_table(struct pp_hwmgr *hwmgr, * @return always 0 */ -int polaris10_populate_smc_voltage_tables(struct pp_hwmgr *hwmgr, +static int polaris10_populate_smc_voltage_tables(struct pp_hwmgr *hwmgr, struct SMU74_Discrete_DpmTable *table) { polaris10_populate_smc_vddci_table(hwmgr, table); @@ -1417,7 +1413,7 @@ static int polaris10_populate_all_memory_levels(struct pp_hwmgr *hwmgr) * @param mclk the MCLK value to be used in the decision if MVDD should be high or low. * @param voltage the SMC VOLTAGE structure to be populated */ -int polaris10_populate_mvdd_value(struct pp_hwmgr *hwmgr, +static int polaris10_populate_mvdd_value(struct pp_hwmgr *hwmgr, uint32_t mclk, SMIO_Pattern *smio_pat) { const struct polaris10_hwmgr *data = (struct polaris10_hwmgr *)(hwmgr->backend); @@ -1931,7 +1927,7 @@ static int polaris10_populate_vr_config(struct pp_hwmgr *hwmgr, } -int polaris10_populate_avfs_parameters(struct pp_hwmgr *hwmgr) +static int polaris10_populate_avfs_parameters(struct pp_hwmgr *hwmgr) { struct polaris10_hwmgr *data = (struct polaris10_hwmgr *)(hwmgr->backend); SMU74_Discrete_DpmTable *table = &(data->smc_state_table); @@ -2560,7 +2556,7 @@ static int polaris10_disable_thermal_auto_throttle(struct pp_hwmgr *hwmgr) return polaris10_disable_auto_throttle_source(hwmgr, PHM_AutoThrottleSource_Thermal); } -int polaris10_pcie_performance_request(struct pp_hwmgr *hwmgr) +static int polaris10_pcie_performance_request(struct pp_hwmgr *hwmgr) { struct polaris10_hwmgr *data = (struct polaris10_hwmgr *)(hwmgr->backend); data->pcie_performance_request = true; @@ -2568,7 +2564,7 @@ int polaris10_pcie_performance_request(struct pp_hwmgr *hwmgr) return 0; } -int polaris10_enable_dpm_tasks(struct pp_hwmgr *hwmgr) +static int polaris10_enable_dpm_tasks(struct pp_hwmgr *hwmgr) { int tmp_result, result = 0; tmp_result = (!polaris10_is_dpm_running(hwmgr)) ? 0 : -1; @@ -2749,12 +2745,12 @@ int polaris10_reset_asic_tasks(struct pp_hwmgr *hwmgr) return 0; } -int polaris10_hwmgr_backend_fini(struct pp_hwmgr *hwmgr) +static int polaris10_hwmgr_backend_fini(struct pp_hwmgr *hwmgr) { return phm_hwmgr_backend_fini(hwmgr); } -int polaris10_set_features_platform_caps(struct pp_hwmgr *hwmgr) +static int polaris10_set_features_platform_caps(struct pp_hwmgr *hwmgr) { struct polaris10_hwmgr *data = (struct polaris10_hwmgr *)(hwmgr->backend); @@ -3109,7 +3105,7 @@ static int polaris10_set_private_data_based_on_pptable(struct pp_hwmgr *hwmgr) return 0; } -int polaris10_patch_voltage_workaround(struct pp_hwmgr *hwmgr) +static int polaris10_patch_voltage_workaround(struct pp_hwmgr *hwmgr) { struct phm_ppt_v1_information *table_info = (struct phm_ppt_v1_information *)(hwmgr->pptable); @@ -3118,11 +3114,27 @@ int polaris10_patch_voltage_workaround(struct pp_hwmgr *hwmgr) struct phm_ppt_v1_voltage_lookup_table *lookup_table = table_info->vddc_lookup_table; uint32_t i; + uint32_t hw_revision, sub_vendor_id, sub_sys_id; + struct cgs_system_info sys_info = {0}; + + sys_info.size = sizeof(struct cgs_system_info); - if (hwmgr->chip_id == CHIP_POLARIS10 && hwmgr->hw_revision == 0xC7 && - ((hwmgr->sub_sys_id == 0xb37 && hwmgr->sub_vendor_id == 0x1002) || - (hwmgr->sub_sys_id == 0x4a8 && hwmgr->sub_vendor_id == 0x1043) || - (hwmgr->sub_sys_id == 0x9480 && hwmgr->sub_vendor_id == 0x1682))) { + sys_info.info_id = CGS_SYSTEM_INFO_PCIE_REV; + cgs_query_system_info(hwmgr->device, &sys_info); + hw_revision = (uint32_t)sys_info.value; + + sys_info.info_id = CGS_SYSTEM_INFO_PCIE_SUB_SYS_ID; + cgs_query_system_info(hwmgr->device, &sys_info); + sub_sys_id = (uint32_t)sys_info.value; + + sys_info.info_id = CGS_SYSTEM_INFO_PCIE_SUB_SYS_VENDOR_ID; + cgs_query_system_info(hwmgr->device, &sys_info); + sub_vendor_id = (uint32_t)sys_info.value; + + if (hwmgr->chip_id == CHIP_POLARIS10 && hw_revision == 0xC7 && + ((sub_sys_id == 0xb37 && sub_vendor_id == 0x1002) || + (sub_sys_id == 0x4a8 && sub_vendor_id == 0x1043) || + (sub_sys_id == 0x9480 && sub_vendor_id == 0x1682))) { if (lookup_table->entries[dep_mclk_table->entries[dep_mclk_table->count-1].vddInd].us_vdd >= 1000) return 0; @@ -3137,7 +3149,7 @@ int polaris10_patch_voltage_workaround(struct pp_hwmgr *hwmgr) } -int polaris10_hwmgr_backend_init(struct pp_hwmgr *hwmgr) +static int polaris10_hwmgr_backend_init(struct pp_hwmgr *hwmgr) { struct polaris10_hwmgr *data; struct pp_atomctrl_gpio_pin_assignment gpio_pin_assignment; @@ -3880,7 +3892,7 @@ static int polaris10_get_pp_table_entry(struct pp_hwmgr *hwmgr, ps = (struct polaris10_power_state *)(&state->hardware); - result = tonga_get_powerplay_table_entry(hwmgr, entry_index, state, + result = get_powerplay_table_entry_v1_0(hwmgr, entry_index, state, polaris10_get_pp_table_entry_callback_func); /* This is the earliest time we have all the dependency table and the VBIOS boot state @@ -4347,7 +4359,8 @@ static int polaris10_generate_dpm_level_enable_mask( return 0; } -int polaris10_enable_disable_uvd_dpm(struct pp_hwmgr *hwmgr, bool enable) +static int +polaris10_enable_disable_uvd_dpm(struct pp_hwmgr *hwmgr, bool enable) { return smum_send_msg_to_smc(hwmgr->smumgr, enable ? PPSMC_MSG_UVDDPM_Enable : @@ -4361,7 +4374,8 @@ int polaris10_enable_disable_vce_dpm(struct pp_hwmgr *hwmgr, bool enable) PPSMC_MSG_VCEDPM_Disable); } -int polaris10_enable_disable_samu_dpm(struct pp_hwmgr *hwmgr, bool enable) +static int +polaris10_enable_disable_samu_dpm(struct pp_hwmgr *hwmgr, bool enable) { return smum_send_msg_to_smc(hwmgr->smumgr, enable? PPSMC_MSG_SAMUDPM_Enable : @@ -4675,14 +4689,16 @@ static int polaris10_set_max_fan_pwm_output(struct pp_hwmgr *hwmgr, uint16_t us_ } -int polaris10_notify_smc_display_change(struct pp_hwmgr *hwmgr, bool has_display) +static int +polaris10_notify_smc_display_change(struct pp_hwmgr *hwmgr, bool has_display) { PPSMC_Msg msg = has_display ? (PPSMC_Msg)PPSMC_HasDisplay : (PPSMC_Msg)PPSMC_NoDisplay; return (smum_send_msg_to_smc(hwmgr->smumgr, msg) == 0) ? 0 : -1; } -int polaris10_notify_smc_display_config_after_ps_adjustment(struct pp_hwmgr *hwmgr) +static int +polaris10_notify_smc_display_config_after_ps_adjustment(struct pp_hwmgr *hwmgr) { uint32_t num_active_displays = 0; struct cgs_display_info info = {0}; @@ -4705,7 +4721,7 @@ int polaris10_notify_smc_display_config_after_ps_adjustment(struct pp_hwmgr *hwm * @param hwmgr the address of the powerplay hardware manager. * @return always OK */ -int polaris10_program_display_gap(struct pp_hwmgr *hwmgr) +static int polaris10_program_display_gap(struct pp_hwmgr *hwmgr) { struct polaris10_hwmgr *data = (struct polaris10_hwmgr *)(hwmgr->backend); uint32_t num_active_displays = 0; @@ -4750,7 +4766,7 @@ int polaris10_program_display_gap(struct pp_hwmgr *hwmgr) } -int polaris10_display_configuration_changed_task(struct pp_hwmgr *hwmgr) +static int polaris10_display_configuration_changed_task(struct pp_hwmgr *hwmgr) { return polaris10_program_display_gap(hwmgr); } @@ -4774,13 +4790,15 @@ static int polaris10_set_max_fan_rpm_output(struct pp_hwmgr *hwmgr, uint16_t us_ PPSMC_MSG_SetFanRpmMax, us_max_fan_rpm); } -int polaris10_register_internal_thermal_interrupt(struct pp_hwmgr *hwmgr, +static int +polaris10_register_internal_thermal_interrupt(struct pp_hwmgr *hwmgr, const void *thermal_interrupt_info) { return 0; } -bool polaris10_check_smc_update_required_for_display_configuration(struct pp_hwmgr *hwmgr) +static bool polaris10_check_smc_update_required_for_display_configuration( + struct pp_hwmgr *hwmgr) { struct polaris10_hwmgr *data = (struct polaris10_hwmgr *)(hwmgr->backend); bool is_update_required = false; @@ -4810,7 +4828,9 @@ static inline bool polaris10_are_power_levels_equal(const struct polaris10_perfo (pl1->pcie_lane == pl2->pcie_lane)); } -int polaris10_check_states_equal(struct pp_hwmgr *hwmgr, const struct pp_hw_power_state *pstate1, const struct pp_hw_power_state *pstate2, bool *equal) +static int polaris10_check_states_equal(struct pp_hwmgr *hwmgr, + const struct pp_hw_power_state *pstate1, + const struct pp_hw_power_state *pstate2, bool *equal) { const struct polaris10_power_state *psa = cast_const_phw_polaris10_power_state(pstate1); const struct polaris10_power_state *psb = cast_const_phw_polaris10_power_state(pstate2); @@ -4841,7 +4861,7 @@ int polaris10_check_states_equal(struct pp_hwmgr *hwmgr, const struct pp_hw_powe return 0; } -int polaris10_upload_mc_firmware(struct pp_hwmgr *hwmgr) +static int polaris10_upload_mc_firmware(struct pp_hwmgr *hwmgr) { struct polaris10_hwmgr *data = (struct polaris10_hwmgr *)(hwmgr->backend); @@ -4954,7 +4974,7 @@ static int polaris10_init_sclk_threshold(struct pp_hwmgr *hwmgr) return 0; } -int polaris10_setup_asic_task(struct pp_hwmgr *hwmgr) +static int polaris10_setup_asic_task(struct pp_hwmgr *hwmgr) { int tmp_result, result = 0; @@ -5225,7 +5245,7 @@ static const struct pp_hwmgr_func polaris10_hwmgr_funcs = { .get_sclk = polaris10_dpm_get_sclk, .patch_boot_state = polaris10_dpm_patch_boot_state, .get_pp_table_entry = polaris10_get_pp_table_entry, - .get_num_of_pp_table_entries = tonga_get_number_of_powerplay_table_entries, + .get_num_of_pp_table_entries = get_number_of_powerplay_table_entries_v1_0, .print_current_perforce_level = polaris10_print_current_perforce_level, .powerdown_uvd = polaris10_phm_powerdown_uvd, .powergate_uvd = polaris10_phm_powergate_uvd, @@ -5262,7 +5282,7 @@ static const struct pp_hwmgr_func polaris10_hwmgr_funcs = { int polaris10_hwmgr_init(struct pp_hwmgr *hwmgr) { hwmgr->hwmgr_func = &polaris10_hwmgr_funcs; - hwmgr->pptable_func = &tonga_pptable_funcs; + hwmgr->pptable_func = &pptable_v1_0_funcs; pp_polaris10_thermal_initialize(hwmgr); return 0; diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/polaris10_hwmgr.h b/drivers/gpu/drm/amd/powerplay/hwmgr/polaris10_hwmgr.h index 33c33947e827..378ab342c257 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/polaris10_hwmgr.h +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/polaris10_hwmgr.h @@ -30,6 +30,7 @@ #include "ppatomctrl.h" #include "polaris10_ppsmc.h" #include "polaris10_powertune.h" +#include "polaris10_smumgr.h" #define POLARIS10_MAX_HARDWARE_POWERLEVELS 2 @@ -165,10 +166,6 @@ struct polaris10_pcie_perf_range { uint16_t max; uint16_t min; }; -struct polaris10_range_table { - uint32_t trans_lower_frequency; /* in 10khz */ - uint32_t trans_upper_frequency; -}; struct polaris10_hwmgr { struct polaris10_dpm_table dpm_table; diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/polaris10_powertune.h b/drivers/gpu/drm/amd/powerplay/hwmgr/polaris10_powertune.h index bc78e28f010d..329119d6cc71 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/polaris10_powertune.h +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/polaris10_powertune.h @@ -66,19 +66,6 @@ struct polaris10_pt_config_reg { enum polaris10_pt_config_reg_type type; }; -struct polaris10_pt_defaults { - uint8_t SviLoadLineEn; - uint8_t SviLoadLineVddC; - uint8_t TDC_VDDC_ThrottleReleaseLimitPerc; - uint8_t TDC_MAWt; - uint8_t TdcWaterfallCtl; - uint8_t DTEAmbientTempBase; - - uint32_t DisplayCac; - uint32_t BAPM_TEMP_GRADIENT; - uint16_t BAPMTI_R[SMU74_DTE_ITERATIONS * SMU74_DTE_SOURCES * SMU74_DTE_SINKS]; - uint16_t BAPMTI_RC[SMU74_DTE_ITERATIONS * SMU74_DTE_SOURCES * SMU74_DTE_SINKS]; -}; void polaris10_initialize_power_tune_defaults(struct pp_hwmgr *hwmgr); int polaris10_populate_bapm_parameters_in_dpm_table(struct pp_hwmgr *hwmgr); diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/polaris10_thermal.c b/drivers/gpu/drm/amd/powerplay/hwmgr/polaris10_thermal.c index b206632d4650..41f835adba91 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/polaris10_thermal.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/polaris10_thermal.c @@ -152,7 +152,7 @@ int polaris10_fan_ctrl_set_default_mode(struct pp_hwmgr *hwmgr) return 0; } -int polaris10_fan_ctrl_start_smc_fan_control(struct pp_hwmgr *hwmgr) +static int polaris10_fan_ctrl_start_smc_fan_control(struct pp_hwmgr *hwmgr) { int result; @@ -425,7 +425,7 @@ int polaris10_thermal_stop_thermal_controller(struct pp_hwmgr *hwmgr) * @param Result the last failure code * @return result from set temperature range routine */ -int tf_polaris10_thermal_setup_fan_table(struct pp_hwmgr *hwmgr, +static int tf_polaris10_thermal_setup_fan_table(struct pp_hwmgr *hwmgr, void *input, void *output, void *storage, int result) { struct polaris10_hwmgr *data = (struct polaris10_hwmgr *)(hwmgr->backend); @@ -537,7 +537,7 @@ int tf_polaris10_thermal_setup_fan_table(struct pp_hwmgr *hwmgr, * @param Result the last failure code * @return result from set temperature range routine */ -int tf_polaris10_thermal_start_smc_fan_control(struct pp_hwmgr *hwmgr, +static int tf_polaris10_thermal_start_smc_fan_control(struct pp_hwmgr *hwmgr, void *input, void *output, void *storage, int result) { /* If the fantable setup has failed we could have disabled diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/tonga_pptable.h b/drivers/gpu/drm/amd/powerplay/hwmgr/pptable_v1_0.h index f127198aafc4..1e870f58dd12 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/tonga_pptable.h +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/pptable_v1_0.h @@ -164,7 +164,7 @@ typedef struct _ATOM_Tonga_State { typedef struct _ATOM_Tonga_State_Array { UCHAR ucRevId; UCHAR ucNumEntries; /* Number of entries. */ - ATOM_Tonga_State states[1]; /* Dynamically allocate entries. */ + ATOM_Tonga_State entries[1]; /* Dynamically allocate entries. */ } ATOM_Tonga_State_Array; typedef struct _ATOM_Tonga_MCLK_Dependency_Record { diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/tonga_processpptables.c b/drivers/gpu/drm/amd/powerplay/hwmgr/process_pptables_v1_0.c index a0ffd4a73d8c..7de701d8a450 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/tonga_processpptables.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/process_pptables_v1_0.c @@ -23,13 +23,13 @@ #include <linux/module.h> #include <linux/slab.h> -#include "tonga_processpptables.h" +#include "process_pptables_v1_0.h" #include "ppatomctrl.h" #include "atombios.h" #include "pp_debug.h" #include "hwmgr.h" #include "cgs_common.h" -#include "tonga_pptable.h" +#include "pptable_v1_0.h" /** * Private Function used during initialization. @@ -153,12 +153,14 @@ const void *get_powerplay_table(struct pp_hwmgr *hwmgr) static int get_vddc_lookup_table( struct pp_hwmgr *hwmgr, phm_ppt_v1_voltage_lookup_table **lookup_table, - const ATOM_Tonga_Voltage_Lookup_Table *vddc_lookup_pp_tables, - uint32_t max_levels + const ATOM_Tonga_Voltage_Lookup_Table *vddc_lookup_pp_tables, + uint32_t max_levels ) { uint32_t table_size, i; phm_ppt_v1_voltage_lookup_table *table; + phm_ppt_v1_voltage_lookup_record *record; + ATOM_Tonga_Voltage_Lookup_Record *atom_record; PP_ASSERT_WITH_CODE((0 != vddc_lookup_pp_tables->ucNumEntries), "Invalid CAC Leakage PowerPlay Table!", return 1); @@ -176,15 +178,17 @@ static int get_vddc_lookup_table( table->count = vddc_lookup_pp_tables->ucNumEntries; for (i = 0; i < vddc_lookup_pp_tables->ucNumEntries; i++) { - table->entries[i].us_calculated = 0; - table->entries[i].us_vdd = - vddc_lookup_pp_tables->entries[i].usVdd; - table->entries[i].us_cac_low = - vddc_lookup_pp_tables->entries[i].usCACLow; - table->entries[i].us_cac_mid = - vddc_lookup_pp_tables->entries[i].usCACMid; - table->entries[i].us_cac_high = - vddc_lookup_pp_tables->entries[i].usCACHigh; + record = GET_FLEXIBLE_ARRAY_MEMBER_ADDR( + phm_ppt_v1_voltage_lookup_record, + entries, table, i); + atom_record = GET_FLEXIBLE_ARRAY_MEMBER_ADDR( + ATOM_Tonga_Voltage_Lookup_Record, + entries, vddc_lookup_pp_tables, i); + record->us_calculated = 0; + record->us_vdd = atom_record->usVdd; + record->us_cac_low = atom_record->usCACLow; + record->us_cac_mid = atom_record->usCACMid; + record->us_cac_high = atom_record->usCACHigh; } *lookup_table = table; @@ -313,11 +317,12 @@ static int init_dpm_2_parameters( static int get_valid_clk( struct pp_hwmgr *hwmgr, struct phm_clock_array **clk_table, - const phm_ppt_v1_clock_voltage_dependency_table * clk_volt_pp_table + phm_ppt_v1_clock_voltage_dependency_table const *clk_volt_pp_table ) { uint32_t table_size, i; struct phm_clock_array *table; + phm_ppt_v1_clock_voltage_dependency_record *dep_record; PP_ASSERT_WITH_CODE((0 != clk_volt_pp_table->count), "Invalid PowerPlay Table!", return -1); @@ -334,9 +339,12 @@ static int get_valid_clk( table->count = (uint32_t)clk_volt_pp_table->count; - for (i = 0; i < table->count; i++) - table->values[i] = (uint32_t)clk_volt_pp_table->entries[i].clk; - + for (i = 0; i < table->count; i++) { + dep_record = GET_FLEXIBLE_ARRAY_MEMBER_ADDR( + phm_ppt_v1_clock_voltage_dependency_record, + entries, clk_volt_pp_table, i); + table->values[i] = (uint32_t)dep_record->clk; + } *clk_table = table; return 0; @@ -345,7 +353,7 @@ static int get_valid_clk( static int get_hard_limits( struct pp_hwmgr *hwmgr, struct phm_clock_and_voltage_limits *limits, - const ATOM_Tonga_Hard_Limit_Table * limitable + ATOM_Tonga_Hard_Limit_Table const *limitable ) { PP_ASSERT_WITH_CODE((0 != limitable->ucNumEntries), "Invalid PowerPlay Table!", return -1); @@ -363,11 +371,13 @@ static int get_hard_limits( static int get_mclk_voltage_dependency_table( struct pp_hwmgr *hwmgr, phm_ppt_v1_clock_voltage_dependency_table **pp_tonga_mclk_dep_table, - const ATOM_Tonga_MCLK_Dependency_Table * mclk_dep_table + ATOM_Tonga_MCLK_Dependency_Table const *mclk_dep_table ) { uint32_t table_size, i; phm_ppt_v1_clock_voltage_dependency_table *mclk_table; + phm_ppt_v1_clock_voltage_dependency_record *mclk_table_record; + ATOM_Tonga_MCLK_Dependency_Record *mclk_dep_record; PP_ASSERT_WITH_CODE((0 != mclk_dep_table->ucNumEntries), "Invalid PowerPlay Table!", return -1); @@ -385,16 +395,17 @@ static int get_mclk_voltage_dependency_table( mclk_table->count = (uint32_t)mclk_dep_table->ucNumEntries; for (i = 0; i < mclk_dep_table->ucNumEntries; i++) { - mclk_table->entries[i].vddInd = - mclk_dep_table->entries[i].ucVddcInd; - mclk_table->entries[i].vdd_offset = - mclk_dep_table->entries[i].usVddgfxOffset; - mclk_table->entries[i].vddci = - mclk_dep_table->entries[i].usVddci; - mclk_table->entries[i].mvdd = - mclk_dep_table->entries[i].usMvdd; - mclk_table->entries[i].clk = - mclk_dep_table->entries[i].ulMclk; + mclk_table_record = GET_FLEXIBLE_ARRAY_MEMBER_ADDR( + phm_ppt_v1_clock_voltage_dependency_record, + entries, mclk_table, i); + mclk_dep_record = GET_FLEXIBLE_ARRAY_MEMBER_ADDR( + ATOM_Tonga_MCLK_Dependency_Record, + entries, mclk_dep_table, i); + mclk_table_record->vddInd = mclk_dep_record->ucVddcInd; + mclk_table_record->vdd_offset = mclk_dep_record->usVddgfxOffset; + mclk_table_record->vddci = mclk_dep_record->usVddci; + mclk_table_record->mvdd = mclk_dep_record->usMvdd; + mclk_table_record->clk = mclk_dep_record->ulMclk; } *pp_tonga_mclk_dep_table = mclk_table; @@ -405,15 +416,17 @@ static int get_mclk_voltage_dependency_table( static int get_sclk_voltage_dependency_table( struct pp_hwmgr *hwmgr, phm_ppt_v1_clock_voltage_dependency_table **pp_tonga_sclk_dep_table, - const PPTable_Generic_SubTable_Header *sclk_dep_table + PPTable_Generic_SubTable_Header const *sclk_dep_table ) { uint32_t table_size, i; phm_ppt_v1_clock_voltage_dependency_table *sclk_table; + phm_ppt_v1_clock_voltage_dependency_record *sclk_table_record; if (sclk_dep_table->ucRevId < 1) { const ATOM_Tonga_SCLK_Dependency_Table *tonga_table = (ATOM_Tonga_SCLK_Dependency_Table *)sclk_dep_table; + ATOM_Tonga_SCLK_Dependency_Record *sclk_dep_record; PP_ASSERT_WITH_CODE((0 != tonga_table->ucNumEntries), "Invalid PowerPlay Table!", return -1); @@ -431,20 +444,23 @@ static int get_sclk_voltage_dependency_table( sclk_table->count = (uint32_t)tonga_table->ucNumEntries; for (i = 0; i < tonga_table->ucNumEntries; i++) { - sclk_table->entries[i].vddInd = - tonga_table->entries[i].ucVddInd; - sclk_table->entries[i].vdd_offset = - tonga_table->entries[i].usVddcOffset; - sclk_table->entries[i].clk = - tonga_table->entries[i].ulSclk; - sclk_table->entries[i].cks_enable = - (((tonga_table->entries[i].ucCKSVOffsetandDisable & 0x80) >> 7) == 0) ? 1 : 0; - sclk_table->entries[i].cks_voffset = - (tonga_table->entries[i].ucCKSVOffsetandDisable & 0x7F); + sclk_dep_record = GET_FLEXIBLE_ARRAY_MEMBER_ADDR( + ATOM_Tonga_SCLK_Dependency_Record, + entries, tonga_table, i); + sclk_table_record = GET_FLEXIBLE_ARRAY_MEMBER_ADDR( + phm_ppt_v1_clock_voltage_dependency_record, + entries, sclk_table, i); + sclk_table_record->vddInd = sclk_dep_record->ucVddInd; + sclk_table_record->vdd_offset = sclk_dep_record->usVddcOffset; + sclk_table_record->clk = sclk_dep_record->ulSclk; + sclk_table_record->cks_enable = + (((sclk_dep_record->ucCKSVOffsetandDisable & 0x80) >> 7) == 0) ? 1 : 0; + sclk_table_record->cks_voffset = (sclk_dep_record->ucCKSVOffsetandDisable & 0x7F); } } else { const ATOM_Polaris_SCLK_Dependency_Table *polaris_table = (ATOM_Polaris_SCLK_Dependency_Table *)sclk_dep_table; + ATOM_Polaris_SCLK_Dependency_Record *sclk_dep_record; PP_ASSERT_WITH_CODE((0 != polaris_table->ucNumEntries), "Invalid PowerPlay Table!", return -1); @@ -462,17 +478,19 @@ static int get_sclk_voltage_dependency_table( sclk_table->count = (uint32_t)polaris_table->ucNumEntries; for (i = 0; i < polaris_table->ucNumEntries; i++) { - sclk_table->entries[i].vddInd = - polaris_table->entries[i].ucVddInd; - sclk_table->entries[i].vdd_offset = - polaris_table->entries[i].usVddcOffset; - sclk_table->entries[i].clk = - polaris_table->entries[i].ulSclk; - sclk_table->entries[i].cks_enable = - (((polaris_table->entries[i].ucCKSVOffsetandDisable & 0x80) >> 7) == 0) ? 1 : 0; - sclk_table->entries[i].cks_voffset = - (polaris_table->entries[i].ucCKSVOffsetandDisable & 0x7F); - sclk_table->entries[i].sclk_offset = polaris_table->entries[i].ulSclkOffset; + sclk_dep_record = GET_FLEXIBLE_ARRAY_MEMBER_ADDR( + ATOM_Polaris_SCLK_Dependency_Record, + entries, polaris_table, i); + sclk_table_record = GET_FLEXIBLE_ARRAY_MEMBER_ADDR( + phm_ppt_v1_clock_voltage_dependency_record, + entries, sclk_table, i); + sclk_table_record->vddInd = sclk_dep_record->ucVddInd; + sclk_table_record->vdd_offset = sclk_dep_record->usVddcOffset; + sclk_table_record->clk = sclk_dep_record->ulSclk; + sclk_table_record->cks_enable = + (((sclk_dep_record->ucCKSVOffsetandDisable & 0x80) >> 7) == 0) ? 1 : 0; + sclk_table_record->cks_voffset = (sclk_dep_record->ucCKSVOffsetandDisable & 0x7F); + sclk_table_record->sclk_offset = sclk_dep_record->ulSclkOffset; } } *pp_tonga_sclk_dep_table = sclk_table; @@ -483,16 +501,19 @@ static int get_sclk_voltage_dependency_table( static int get_pcie_table( struct pp_hwmgr *hwmgr, phm_ppt_v1_pcie_table **pp_tonga_pcie_table, - const PPTable_Generic_SubTable_Header * pTable + PPTable_Generic_SubTable_Header const *ptable ) { uint32_t table_size, i, pcie_count; phm_ppt_v1_pcie_table *pcie_table; struct phm_ppt_v1_information *pp_table_information = (struct phm_ppt_v1_information *)(hwmgr->pptable); + phm_ppt_v1_pcie_record *pcie_record; + + if (ptable->ucRevId < 1) { + const ATOM_Tonga_PCIE_Table *atom_pcie_table = (ATOM_Tonga_PCIE_Table *)ptable; + ATOM_Tonga_PCIE_Record *atom_pcie_record; - if (pTable->ucRevId < 1) { - const ATOM_Tonga_PCIE_Table *atom_pcie_table = (ATOM_Tonga_PCIE_Table *)pTable; PP_ASSERT_WITH_CODE((atom_pcie_table->ucNumEntries != 0), "Invalid PowerPlay Table!", return -1); @@ -518,18 +539,23 @@ static int get_pcie_table( Disregarding the excess entries... \n"); pcie_table->count = pcie_count; - for (i = 0; i < pcie_count; i++) { - pcie_table->entries[i].gen_speed = - atom_pcie_table->entries[i].ucPCIEGenSpeed; - pcie_table->entries[i].lane_width = - atom_pcie_table->entries[i].usPCIELaneWidth; + pcie_record = GET_FLEXIBLE_ARRAY_MEMBER_ADDR( + phm_ppt_v1_pcie_record, + entries, pcie_table, i); + atom_pcie_record = GET_FLEXIBLE_ARRAY_MEMBER_ADDR( + ATOM_Tonga_PCIE_Record, + entries, atom_pcie_table, i); + pcie_record->gen_speed = atom_pcie_record->ucPCIEGenSpeed; + pcie_record->lane_width = atom_pcie_record->usPCIELaneWidth; } *pp_tonga_pcie_table = pcie_table; } else { /* Polaris10/Polaris11 and newer. */ - const ATOM_Polaris10_PCIE_Table *atom_pcie_table = (ATOM_Polaris10_PCIE_Table *)pTable; + const ATOM_Polaris10_PCIE_Table *atom_pcie_table = (ATOM_Polaris10_PCIE_Table *)ptable; + ATOM_Polaris10_PCIE_Record *atom_pcie_record; + PP_ASSERT_WITH_CODE((atom_pcie_table->ucNumEntries != 0), "Invalid PowerPlay Table!", return -1); @@ -557,12 +583,15 @@ static int get_pcie_table( pcie_table->count = pcie_count; for (i = 0; i < pcie_count; i++) { - pcie_table->entries[i].gen_speed = - atom_pcie_table->entries[i].ucPCIEGenSpeed; - pcie_table->entries[i].lane_width = - atom_pcie_table->entries[i].usPCIELaneWidth; - pcie_table->entries[i].pcie_sclk = - atom_pcie_table->entries[i].ulPCIE_Sclk; + pcie_record = GET_FLEXIBLE_ARRAY_MEMBER_ADDR( + phm_ppt_v1_pcie_record, + entries, pcie_table, i); + atom_pcie_record = GET_FLEXIBLE_ARRAY_MEMBER_ADDR( + ATOM_Polaris10_PCIE_Record, + entries, atom_pcie_table, i); + pcie_record->gen_speed = atom_pcie_record->ucPCIEGenSpeed; + pcie_record->lane_width = atom_pcie_record->usPCIELaneWidth; + pcie_record->pcie_sclk = atom_pcie_record->ulPCIE_Sclk; } *pp_tonga_pcie_table = pcie_table; @@ -684,6 +713,7 @@ static int get_mm_clock_voltage_table( uint32_t table_size, i; const ATOM_Tonga_MM_Dependency_Record *mm_dependency_record; phm_ppt_v1_mm_clock_voltage_dependency_table *mm_table; + phm_ppt_v1_mm_clock_voltage_dependency_record *mm_table_record; PP_ASSERT_WITH_CODE((0 != mm_dependency_table->ucNumEntries), "Invalid PowerPlay Table!", return -1); @@ -700,14 +730,19 @@ static int get_mm_clock_voltage_table( mm_table->count = mm_dependency_table->ucNumEntries; for (i = 0; i < mm_dependency_table->ucNumEntries; i++) { - mm_dependency_record = &mm_dependency_table->entries[i]; - mm_table->entries[i].vddcInd = mm_dependency_record->ucVddcInd; - mm_table->entries[i].vddgfx_offset = mm_dependency_record->usVddgfxOffset; - mm_table->entries[i].aclk = mm_dependency_record->ulAClk; - mm_table->entries[i].samclock = mm_dependency_record->ulSAMUClk; - mm_table->entries[i].eclk = mm_dependency_record->ulEClk; - mm_table->entries[i].vclk = mm_dependency_record->ulVClk; - mm_table->entries[i].dclk = mm_dependency_record->ulDClk; + mm_dependency_record = GET_FLEXIBLE_ARRAY_MEMBER_ADDR( + ATOM_Tonga_MM_Dependency_Record, + entries, mm_dependency_table, i); + mm_table_record = GET_FLEXIBLE_ARRAY_MEMBER_ADDR( + phm_ppt_v1_mm_clock_voltage_dependency_record, + entries, mm_table, i); + mm_table_record->vddcInd = mm_dependency_record->ucVddcInd; + mm_table_record->vddgfx_offset = mm_dependency_record->usVddgfxOffset; + mm_table_record->aclk = mm_dependency_record->ulAClk; + mm_table_record->samclock = mm_dependency_record->ulSAMUClk; + mm_table_record->eclk = mm_dependency_record->ulEClk; + mm_table_record->vclk = mm_dependency_record->ulVClk; + mm_table_record->dclk = mm_dependency_record->ulDClk; } *tonga_mm_table = mm_table; @@ -1014,7 +1049,7 @@ static int check_powerplay_tables( return 0; } -int tonga_pp_tables_initialize(struct pp_hwmgr *hwmgr) +int pp_tables_v1_0_initialize(struct pp_hwmgr *hwmgr) { int result = 0; const ATOM_Tonga_POWERPLAYTABLE *powerplay_table; @@ -1065,7 +1100,7 @@ int tonga_pp_tables_initialize(struct pp_hwmgr *hwmgr) return result; } -int tonga_pp_tables_uninitialize(struct pp_hwmgr *hwmgr) +int pp_tables_v1_0_uninitialize(struct pp_hwmgr *hwmgr) { struct phm_ppt_v1_information *pp_table_information = (struct phm_ppt_v1_information *)(hwmgr->pptable); @@ -1109,14 +1144,14 @@ int tonga_pp_tables_uninitialize(struct pp_hwmgr *hwmgr) return 0; } -const struct pp_table_func tonga_pptable_funcs = { - .pptable_init = tonga_pp_tables_initialize, - .pptable_fini = tonga_pp_tables_uninitialize, +const struct pp_table_func pptable_v1_0_funcs = { + .pptable_init = pp_tables_v1_0_initialize, + .pptable_fini = pp_tables_v1_0_uninitialize, }; -int tonga_get_number_of_powerplay_table_entries(struct pp_hwmgr *hwmgr) +int get_number_of_powerplay_table_entries_v1_0(struct pp_hwmgr *hwmgr) { - const ATOM_Tonga_State_Array * state_arrays; + ATOM_Tonga_State_Array const *state_arrays; const ATOM_Tonga_POWERPLAYTABLE *pp_table = get_powerplay_table(hwmgr); PP_ASSERT_WITH_CODE((NULL != pp_table), @@ -1163,6 +1198,71 @@ static uint32_t make_classification_flags(struct pp_hwmgr *hwmgr, return result; } +static int ppt_get_num_of_vce_state_table_entries_v1_0(struct pp_hwmgr *hwmgr) +{ + const ATOM_Tonga_POWERPLAYTABLE *pp_table = get_powerplay_table(hwmgr); + const ATOM_Tonga_VCE_State_Table *vce_state_table = + (ATOM_Tonga_VCE_State_Table *)(((unsigned long)pp_table) + le16_to_cpu(pp_table->usVCEStateTableOffset)); + + if (vce_state_table == NULL) + return 0; + + return vce_state_table->ucNumEntries; +} + +static int ppt_get_vce_state_table_entry_v1_0(struct pp_hwmgr *hwmgr, uint32_t i, + struct pp_vce_state *vce_state, void **clock_info, uint32_t *flag) +{ + const ATOM_Tonga_VCE_State_Record *vce_state_record; + ATOM_Tonga_SCLK_Dependency_Record *sclk_dep_record; + ATOM_Tonga_MCLK_Dependency_Record *mclk_dep_record; + ATOM_Tonga_MM_Dependency_Record *mm_dep_record; + const ATOM_Tonga_POWERPLAYTABLE *pptable = get_powerplay_table(hwmgr); + const ATOM_Tonga_VCE_State_Table *vce_state_table = (ATOM_Tonga_VCE_State_Table *)(((unsigned long)pptable) + + le16_to_cpu(pptable->usVCEStateTableOffset)); + const ATOM_Tonga_SCLK_Dependency_Table *sclk_dep_table = (ATOM_Tonga_SCLK_Dependency_Table *)(((unsigned long)pptable) + + le16_to_cpu(pptable->usSclkDependencyTableOffset)); + const ATOM_Tonga_MCLK_Dependency_Table *mclk_dep_table = (ATOM_Tonga_MCLK_Dependency_Table *)(((unsigned long)pptable) + + le16_to_cpu(pptable->usMclkDependencyTableOffset)); + const ATOM_Tonga_MM_Dependency_Table *mm_dep_table = (ATOM_Tonga_MM_Dependency_Table *)(((unsigned long)pptable) + + le16_to_cpu(pptable->usMMDependencyTableOffset)); + + PP_ASSERT_WITH_CODE((i < vce_state_table->ucNumEntries), + "Requested state entry ID is out of range!", + return -EINVAL); + + vce_state_record = GET_FLEXIBLE_ARRAY_MEMBER_ADDR( + ATOM_Tonga_VCE_State_Record, + entries, vce_state_table, i); + sclk_dep_record = GET_FLEXIBLE_ARRAY_MEMBER_ADDR( + ATOM_Tonga_SCLK_Dependency_Record, + entries, sclk_dep_table, + vce_state_record->ucSCLKIndex); + mm_dep_record = GET_FLEXIBLE_ARRAY_MEMBER_ADDR( + ATOM_Tonga_MM_Dependency_Record, + entries, mm_dep_table, + vce_state_record->ucVCEClockIndex); + *flag = vce_state_record->ucFlag; + + vce_state->evclk = mm_dep_record->ulEClk; + vce_state->ecclk = mm_dep_record->ulEClk; + vce_state->sclk = sclk_dep_record->ulSclk; + + if (vce_state_record->ucMCLKIndex >= mclk_dep_table->ucNumEntries) + mclk_dep_record = GET_FLEXIBLE_ARRAY_MEMBER_ADDR( + ATOM_Tonga_MCLK_Dependency_Record, + entries, mclk_dep_table, + mclk_dep_table->ucNumEntries - 1); + else + mclk_dep_record = GET_FLEXIBLE_ARRAY_MEMBER_ADDR( + ATOM_Tonga_MCLK_Dependency_Record, + entries, mclk_dep_table, + vce_state_record->ucMCLKIndex); + + vce_state->mclk = mclk_dep_record->ulMclk; + return 0; +} + /** * Create a Power State out of an entry in the PowerPlay table. * This function is called by the hardware back-end. @@ -1171,15 +1271,17 @@ static uint32_t make_classification_flags(struct pp_hwmgr *hwmgr, * @param power_state The address of the PowerState instance being created. * @return -1 if the entry cannot be retrieved. */ -int tonga_get_powerplay_table_entry(struct pp_hwmgr *hwmgr, +int get_powerplay_table_entry_v1_0(struct pp_hwmgr *hwmgr, uint32_t entry_index, struct pp_power_state *power_state, int (*call_back_func)(struct pp_hwmgr *, void *, struct pp_power_state *, void *, uint32_t)) { int result = 0; - const ATOM_Tonga_State_Array * state_arrays; + const ATOM_Tonga_State_Array *state_arrays; const ATOM_Tonga_State *state_entry; const ATOM_Tonga_POWERPLAYTABLE *pp_table = get_powerplay_table(hwmgr); + int i, j; + uint32_t flags = 0; PP_ASSERT_WITH_CODE((NULL != pp_table), "Missing PowerPlay Table!", return -1;); power_state->classification.bios_index = entry_index; @@ -1196,7 +1298,9 @@ int tonga_get_powerplay_table_entry(struct pp_hwmgr *hwmgr, PP_ASSERT_WITH_CODE((entry_index <= state_arrays->ucNumEntries), "Invalid PowerPlay Table State Array Entry.", return -1); - state_entry = &(state_arrays->states[entry_index]); + state_entry = GET_FLEXIBLE_ARRAY_MEMBER_ADDR( + ATOM_Tonga_State, entries, + state_arrays, entry_index); result = call_back_func(hwmgr, (void *)state_entry, power_state, (void *)pp_table, @@ -1209,5 +1313,13 @@ int tonga_get_powerplay_table_entry(struct pp_hwmgr *hwmgr, PP_StateClassificationFlag_Boot)) result = hwmgr->hwmgr_func->patch_boot_state(hwmgr, &(power_state->hardware)); + hwmgr->num_vce_state_tables = i = ppt_get_num_of_vce_state_table_entries_v1_0(hwmgr); + + if ((i != 0) && (i <= PP_MAX_VCE_LEVELS)) { + for (j = 0; j < i; j++) + ppt_get_vce_state_table_entry_v1_0(hwmgr, j, &(hwmgr->vce_states[j]), NULL, &flags); + } + return result; } + diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/tonga_processpptables.h b/drivers/gpu/drm/amd/powerplay/hwmgr/process_pptables_v1_0.h index d24b8887f466..b9710abdff01 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/tonga_processpptables.h +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/process_pptables_v1_0.h @@ -20,14 +20,14 @@ * OTHER DEALINGS IN THE SOFTWARE. * */ -#ifndef TONGA_PROCESSPPTABLES_H -#define TONGA_PROCESSPPTABLES_H +#ifndef _PROCESSPPTABLES_V1_0_H +#define _PROCESSPPTABLES_V1_0_H #include "hwmgr.h" -extern const struct pp_table_func tonga_pptable_funcs; -extern int tonga_get_number_of_powerplay_table_entries(struct pp_hwmgr *hwmgr); -extern int tonga_get_powerplay_table_entry(struct pp_hwmgr *hwmgr, uint32_t entry_index, +extern const struct pp_table_func pptable_v1_0_funcs; +extern int get_number_of_powerplay_table_entries_v1_0(struct pp_hwmgr *hwmgr); +extern int get_powerplay_table_entry_v1_0(struct pp_hwmgr *hwmgr, uint32_t entry_index, struct pp_power_state *power_state, int (*call_back_func)(struct pp_hwmgr *, void *, struct pp_power_state *, void *, uint32_t)); diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/processpptables.c b/drivers/gpu/drm/amd/powerplay/hwmgr/processpptables.c index 6c321b0d8a1e..ccf7ebeaf892 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/processpptables.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/processpptables.c @@ -1523,7 +1523,7 @@ int get_number_of_vce_state_table_entries( int get_vce_state_table_entry(struct pp_hwmgr *hwmgr, unsigned long i, - struct PP_VCEState *vce_state, + struct pp_vce_state *vce_state, void **clock_info, unsigned long *flag) { diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/tonga_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/tonga_hwmgr.c index 7f9ba7f15e19..582d04aed346 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/tonga_hwmgr.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/tonga_hwmgr.c @@ -29,8 +29,8 @@ #include "tonga_hwmgr.h" #include "pptable.h" #include "processpptables.h" -#include "tonga_processpptables.h" -#include "tonga_pptable.h" +#include "process_pptables_v1_0.h" +#include "pptable_v1_0.h" #include "pp_debug.h" #include "tonga_ppsmc.h" #include "cgs_common.h" @@ -202,6 +202,7 @@ uint8_t tonga_get_voltage_id(pp_atomctrl_voltage_table *voltage_table, return i - 1; } + /** * @brief PhwTonga_GetVoltageOrder * Returns index of requested voltage record in lookup(table) @@ -229,7 +230,7 @@ uint8_t tonga_get_voltage_index(phm_ppt_v1_voltage_lookup_table *look_up_table, return i-1; } -bool tonga_is_dpm_running(struct pp_hwmgr *hwmgr) +static bool tonga_is_dpm_running(struct pp_hwmgr *hwmgr) { /* * We return the status of Voltage Control instead of checking SCLK/MCLK DPM @@ -334,7 +335,7 @@ void tonga_initialize_dpm_defaults(struct pp_hwmgr *hwmgr) } -int tonga_update_sclk_threshold(struct pp_hwmgr *hwmgr) +static int tonga_update_sclk_threshold(struct pp_hwmgr *hwmgr) { tonga_hwmgr *data = (tonga_hwmgr *)(hwmgr->backend); @@ -771,7 +772,7 @@ int tonga_set_boot_state(struct pp_hwmgr *hwmgr) * @param hwmgr the address of the powerplay hardware manager. * @return always 0 */ -int tonga_process_firmware_header(struct pp_hwmgr *hwmgr) +static int tonga_process_firmware_header(struct pp_hwmgr *hwmgr) { tonga_hwmgr *data = (tonga_hwmgr *)(hwmgr->backend); struct tonga_smumgr *tonga_smu = (struct tonga_smumgr *)(hwmgr->smumgr->backend); @@ -1315,15 +1316,6 @@ static int tonga_populate_smc_mvdd_table(struct pp_hwmgr *hwmgr, } /** - * Convert a voltage value in mv unit to VID number required by SMU firmware - */ -static uint8_t convert_to_vid(uint16_t vddc) -{ - return (uint8_t) ((6200 - (vddc * VOLTAGE_SCALE)) / 25); -} - - -/** * Preparation of vddc and vddgfx CAC tables for SMC. * * @param hwmgr the address of the hardware manager @@ -2894,7 +2886,7 @@ int tonga_populate_smc_initial_state(struct pp_hwmgr *hwmgr, * @param pInput the pointer to input data (PowerState) * @return always 0 */ -int tonga_init_smc_table(struct pp_hwmgr *hwmgr) +static int tonga_init_smc_table(struct pp_hwmgr *hwmgr) { int result; tonga_hwmgr *data = (tonga_hwmgr *)(hwmgr->backend); @@ -3989,7 +3981,7 @@ int tonga_set_valid_flag(phw_tonga_mc_reg_table *table) return 0; } -int tonga_initialize_mc_reg_table(struct pp_hwmgr *hwmgr) +static int tonga_initialize_mc_reg_table(struct pp_hwmgr *hwmgr) { int result; tonga_hwmgr *data = (tonga_hwmgr *)(hwmgr->backend); @@ -4326,6 +4318,79 @@ int tonga_program_voting_clients(struct pp_hwmgr *hwmgr) return 0; } +static void tonga_set_dpm_event_sources(struct pp_hwmgr *hwmgr, uint32_t sources) +{ + bool protection; + enum DPM_EVENT_SRC src; + + switch (sources) { + default: + printk(KERN_ERR "Unknown throttling event sources."); + /* fall through */ + case 0: + protection = false; + /* src is unused */ + break; + case (1 << PHM_AutoThrottleSource_Thermal): + protection = true; + src = DPM_EVENT_SRC_DIGITAL; + break; + case (1 << PHM_AutoThrottleSource_External): + protection = true; + src = DPM_EVENT_SRC_EXTERNAL; + break; + case (1 << PHM_AutoThrottleSource_External) | + (1 << PHM_AutoThrottleSource_Thermal): + protection = true; + src = DPM_EVENT_SRC_DIGITAL_OR_EXTERNAL; + break; + } + /* Order matters - don't enable thermal protection for the wrong source. */ + if (protection) { + PHM_WRITE_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, CG_THERMAL_CTRL, + DPM_EVENT_SRC, src); + PHM_WRITE_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, GENERAL_PWRMGT, + THERMAL_PROTECTION_DIS, + !phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, + PHM_PlatformCaps_ThermalController)); + } else + PHM_WRITE_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, GENERAL_PWRMGT, + THERMAL_PROTECTION_DIS, 1); +} + +static int tonga_enable_auto_throttle_source(struct pp_hwmgr *hwmgr, + PHM_AutoThrottleSource source) +{ + struct tonga_hwmgr *data = (struct tonga_hwmgr *)(hwmgr->backend); + + if (!(data->active_auto_throttle_sources & (1 << source))) { + data->active_auto_throttle_sources |= 1 << source; + tonga_set_dpm_event_sources(hwmgr, data->active_auto_throttle_sources); + } + return 0; +} + +static int tonga_enable_thermal_auto_throttle(struct pp_hwmgr *hwmgr) +{ + return tonga_enable_auto_throttle_source(hwmgr, PHM_AutoThrottleSource_Thermal); +} + +static int tonga_disable_auto_throttle_source(struct pp_hwmgr *hwmgr, + PHM_AutoThrottleSource source) +{ + struct tonga_hwmgr *data = (struct tonga_hwmgr *)(hwmgr->backend); + + if (data->active_auto_throttle_sources & (1 << source)) { + data->active_auto_throttle_sources &= ~(1 << source); + tonga_set_dpm_event_sources(hwmgr, data->active_auto_throttle_sources); + } + return 0; +} + +static int tonga_disable_thermal_auto_throttle(struct pp_hwmgr *hwmgr) +{ + return tonga_disable_auto_throttle_source(hwmgr, PHM_AutoThrottleSource_Thermal); +} int tonga_enable_dpm_tasks(struct pp_hwmgr *hwmgr) { @@ -4409,6 +4474,10 @@ int tonga_enable_dpm_tasks(struct pp_hwmgr *hwmgr) PP_ASSERT_WITH_CODE((tmp_result == 0), "Failed to power control set level!", result = tmp_result); + tmp_result = tonga_enable_thermal_auto_throttle(hwmgr); + PP_ASSERT_WITH_CODE((0 == tmp_result), + "Failed to enable thermal auto throttle!", result = tmp_result); + return result; } @@ -4420,6 +4489,10 @@ int tonga_disable_dpm_tasks(struct pp_hwmgr *hwmgr) PP_ASSERT_WITH_CODE((0 == tmp_result), "SMC is still running!", return 0); + tmp_result = tonga_disable_thermal_auto_throttle(hwmgr); + PP_ASSERT_WITH_CODE((tmp_result == 0), + "Failed to disable thermal auto throttle!", result = tmp_result); + tmp_result = tonga_stop_dpm(hwmgr); PP_ASSERT_WITH_CODE((0 == tmp_result), "Failed to stop DPM!", result = tmp_result); @@ -5090,7 +5163,7 @@ static int tonga_get_pp_table_entry(struct pp_hwmgr *hwmgr, tonga_ps = cast_phw_tonga_power_state(&(ps->hardware)); - result = tonga_get_powerplay_table_entry(hwmgr, entry_index, ps, + result = get_powerplay_table_entry_v1_0(hwmgr, entry_index, ps, tonga_get_pp_table_entry_callback_func); /* This is the earliest time we have all the dependency table and the VBIOS boot state @@ -6254,7 +6327,7 @@ static const struct pp_hwmgr_func tonga_hwmgr_funcs = { .get_sclk = tonga_dpm_get_sclk, .patch_boot_state = tonga_dpm_patch_boot_state, .get_pp_table_entry = tonga_get_pp_table_entry, - .get_num_of_pp_table_entries = tonga_get_number_of_powerplay_table_entries, + .get_num_of_pp_table_entries = get_number_of_powerplay_table_entries_v1_0, .print_current_perforce_level = tonga_print_current_perforce_level, .powerdown_uvd = tonga_phm_powerdown_uvd, .powergate_uvd = tonga_phm_powergate_uvd, @@ -6290,7 +6363,7 @@ static const struct pp_hwmgr_func tonga_hwmgr_funcs = { int tonga_hwmgr_init(struct pp_hwmgr *hwmgr) { hwmgr->hwmgr_func = &tonga_hwmgr_funcs; - hwmgr->pptable_func = &tonga_pptable_funcs; + hwmgr->pptable_func = &pptable_v1_0_funcs; pp_tonga_thermal_initialize(hwmgr); return 0; } diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/tonga_powertune.c b/drivers/gpu/drm/amd/powerplay/hwmgr/tonga_powertune.c index 9496ade3247e..24d9a05e7997 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/tonga_powertune.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/tonga_powertune.c @@ -56,9 +56,6 @@ void tonga_initialize_power_tune_defaults(struct pp_hwmgr *hwmgr) else tonga_hwmgr->power_tune_defaults = &tonga_power_tune_data_set_array[0]; - /* Assume disabled */ - phm_cap_unset(hwmgr->platform_descriptor.platformCaps, - PHM_PlatformCaps_PowerContainment); phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_CAC); phm_cap_unset(hwmgr->platform_descriptor.platformCaps, diff --git a/drivers/gpu/drm/amd/powerplay/inc/amd_powerplay.h b/drivers/gpu/drm/amd/powerplay/inc/amd_powerplay.h index 3f8172f545b0..18f39e89a7aa 100644 --- a/drivers/gpu/drm/amd/powerplay/inc/amd_powerplay.h +++ b/drivers/gpu/drm/amd/powerplay/inc/amd_powerplay.h @@ -131,9 +131,6 @@ struct amd_pp_init { struct cgs_device *device; uint32_t chip_family; uint32_t chip_id; - uint32_t rev_id; - uint16_t sub_sys_id; - uint16_t sub_vendor_id; }; enum amd_pp_display_config_type{ diff --git a/drivers/gpu/drm/amd/powerplay/inc/hardwaremanager.h b/drivers/gpu/drm/amd/powerplay/inc/hardwaremanager.h index 962cb5385951..d4495839c64c 100644 --- a/drivers/gpu/drm/amd/powerplay/inc/hardwaremanager.h +++ b/drivers/gpu/drm/amd/powerplay/inc/hardwaremanager.h @@ -341,7 +341,6 @@ extern int phm_powerdown_uvd(struct pp_hwmgr *hwmgr); extern int phm_setup_asic(struct pp_hwmgr *hwmgr); extern int phm_enable_dynamic_state_management(struct pp_hwmgr *hwmgr); extern int phm_disable_dynamic_state_management(struct pp_hwmgr *hwmgr); -extern void phm_init_dynamic_caps(struct pp_hwmgr *hwmgr); extern bool phm_is_hw_access_blocked(struct pp_hwmgr *hwmgr); extern int phm_block_hw_access(struct pp_hwmgr *hwmgr, bool block); extern int phm_set_power_state(struct pp_hwmgr *hwmgr, diff --git a/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h b/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h index 36b4ec9c9cb1..e98748344801 100644 --- a/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h +++ b/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h @@ -31,18 +31,20 @@ #include "hwmgr_ppt.h" #include "ppatomctrl.h" #include "hwmgr_ppt.h" +#include "power_state.h" struct pp_instance; struct pp_hwmgr; -struct pp_hw_power_state; -struct pp_power_state; -struct PP_VCEState; struct phm_fan_speed_info; struct pp_atomctrl_voltage_table; - extern int amdgpu_powercontainment; extern int amdgpu_sclk_deep_sleep_en; +extern unsigned amdgpu_pp_feature_mask; + +#define VOLTAGE_SCALE 4 + +uint8_t convert_to_vid(uint16_t vddc); enum DISPLAY_GAP { DISPLAY_GAP_VBLANK_OR_WM = 0, /* Wait for vblank or MCHG watermark. */ @@ -52,7 +54,6 @@ enum DISPLAY_GAP { }; typedef enum DISPLAY_GAP DISPLAY_GAP; - struct vi_dpm_level { bool enabled; uint32_t value; @@ -74,6 +75,19 @@ enum PP_Result { #define PCIE_PERF_REQ_GEN2 3 #define PCIE_PERF_REQ_GEN3 4 +enum PP_FEATURE_MASK { + PP_SCLK_DPM_MASK = 0x1, + PP_MCLK_DPM_MASK = 0x2, + PP_PCIE_DPM_MASK = 0x4, + PP_SCLK_DEEP_SLEEP_MASK = 0x8, + PP_POWER_CONTAINMENT_MASK = 0x10, + PP_UVD_HANDSHAKE_MASK = 0x20, + PP_SMC_VOLTAGE_CONTROL_MASK = 0x40, + PP_VBI_TIME_SUPPORT_MASK = 0x80, + PP_ULV_MASK = 0x100, + PP_ENABLE_GFX_CG_THRU_SMU = 0x200 +}; + enum PHM_BackEnd_Magic { PHM_Dummy_Magic = 0xAA5555AA, PHM_RV770_Magic = 0xDCBAABCD, @@ -354,7 +368,7 @@ struct pp_table_func { int (*pptable_get_vce_state_table_entry)( struct pp_hwmgr *hwmgr, unsigned long i, - struct PP_VCEState *vce_state, + struct pp_vce_state *vce_state, void **clock_info, unsigned long *flag); }; @@ -573,22 +587,43 @@ struct phm_microcode_version_info { uint32_t NB; }; +#define PP_MAX_VCE_LEVELS 6 + +enum PP_VCE_LEVEL { + PP_VCE_LEVEL_AC_ALL = 0, /* AC, All cases */ + PP_VCE_LEVEL_DC_EE = 1, /* DC, entropy encoding */ + PP_VCE_LEVEL_DC_LL_LOW = 2, /* DC, low latency queue, res <= 720 */ + PP_VCE_LEVEL_DC_LL_HIGH = 3, /* DC, low latency queue, 1080 >= res > 720 */ + PP_VCE_LEVEL_DC_GP_LOW = 4, /* DC, general purpose queue, res <= 720 */ + PP_VCE_LEVEL_DC_GP_HIGH = 5, /* DC, general purpose queue, 1080 >= res > 720 */ +}; + + +enum PP_TABLE_VERSION { + PP_TABLE_V0 = 0, + PP_TABLE_V1, + PP_TABLE_V2, + PP_TABLE_MAX +}; + /** * The main hardware manager structure. */ struct pp_hwmgr { uint32_t chip_family; uint32_t chip_id; - uint32_t hw_revision; - uint32_t sub_sys_id; - uint32_t sub_vendor_id; + uint32_t pp_table_version; void *device; struct pp_smumgr *smumgr; const void *soft_pp_table; uint32_t soft_pp_table_size; void *hardcode_pp_table; bool need_pp_table_upload; + + struct pp_vce_state vce_states[PP_MAX_VCE_LEVELS]; + uint32_t num_vce_state_tables; + enum amd_dpm_forced_level dpm_level; bool block_hw_access; struct phm_gfx_arbiter gfx_arbiter; @@ -626,6 +661,7 @@ struct pp_hwmgr { struct pp_power_state *boot_ps; struct pp_power_state *uvd_ps; struct amd_pp_display_configuration display_config; + uint32_t feature_mask; }; @@ -661,6 +697,8 @@ extern void phm_trim_voltage_table_to_fit_state_table(uint32_t max_vol_steps, st extern int phm_reset_single_dpm_table(void *table, uint32_t count, int max); extern void phm_setup_pcie_table_entry(void *table, uint32_t index, uint32_t pcie_gen, uint32_t pcie_lanes); extern int32_t phm_get_dpm_level_enable_mask_value(void *table); +extern uint8_t phm_get_voltage_id(struct pp_atomctrl_voltage_table *voltage_table, + uint32_t voltage); extern uint8_t phm_get_voltage_index(struct phm_ppt_v1_voltage_lookup_table *lookup_table, uint16_t voltage); extern uint16_t phm_find_closest_vddci(struct pp_atomctrl_voltage_table *vddci_table, uint16_t vddci); extern int phm_find_boot_level(void *table, uint32_t value, uint32_t *boot_level); @@ -671,6 +709,9 @@ extern int phm_hwmgr_backend_fini(struct pp_hwmgr *hwmgr); extern uint32_t phm_get_lowest_enabled_level(struct pp_hwmgr *hwmgr, uint32_t mask); extern void phm_apply_dal_min_voltage_request(struct pp_hwmgr *hwmgr); +extern int phm_get_voltage_evv_on_sclk(struct pp_hwmgr *hwmgr, uint8_t voltage_type, + uint32_t sclk, uint16_t id, uint16_t *voltage); + #define PHM_ENTIRE_REGISTER_MASK 0xFFFFFFFFU #define PHM_FIELD_SHIFT(reg, field) reg##__##field##__SHIFT @@ -685,8 +726,6 @@ extern void phm_apply_dal_min_voltage_request(struct pp_hwmgr *hwmgr); PHM_FIELD_SHIFT(reg, field)) - - /* Operations on named fields. */ #define PHM_READ_FIELD(device, reg, field) \ diff --git a/drivers/gpu/drm/amd/powerplay/inc/power_state.h b/drivers/gpu/drm/amd/powerplay/inc/power_state.h index a3f0ce4d5835..9ceaed9ac52a 100644 --- a/drivers/gpu/drm/amd/powerplay/inc/power_state.h +++ b/drivers/gpu/drm/amd/powerplay/inc/power_state.h @@ -158,7 +158,7 @@ struct pp_power_state { /*Structure to hold a VCE state entry*/ -struct PP_VCEState { +struct pp_vce_state { uint32_t evclk; uint32_t ecclk; uint32_t sclk; @@ -171,30 +171,28 @@ enum PP_MMProfilingState { PP_MMProfilingState_Stopped }; -struct PP_Clock_Engine_Request { - unsigned long clientType; - unsigned long ctxid; +struct pp_clock_engine_request { + unsigned long client_type; + unsigned long ctx_id; uint64_t context_handle; unsigned long sclk; - unsigned long sclkHardMin; + unsigned long sclk_hard_min; unsigned long mclk; unsigned long iclk; unsigned long evclk; unsigned long ecclk; - unsigned long ecclkHardMin; + unsigned long ecclk_hard_min; unsigned long vclk; unsigned long dclk; - unsigned long samclk; - unsigned long acpclk; - unsigned long sclkOverdrive; - unsigned long mclkOverdrive; + unsigned long sclk_over_drive; + unsigned long mclk_over_drive; unsigned long sclk_threshold; unsigned long flag; unsigned long vclk_ceiling; unsigned long dclk_ceiling; unsigned long num_cus; - unsigned long pmflag; - enum PP_MMProfilingState MMProfilingState; + unsigned long pm_flag; + enum PP_MMProfilingState mm_profiling_state; }; #endif diff --git a/drivers/gpu/drm/amd/powerplay/inc/pp_debug.h b/drivers/gpu/drm/amd/powerplay/inc/pp_debug.h index d7d83b7c7f95..bfdbec10cdd5 100644 --- a/drivers/gpu/drm/amd/powerplay/inc/pp_debug.h +++ b/drivers/gpu/drm/amd/powerplay/inc/pp_debug.h @@ -43,5 +43,8 @@ } while (0) +#define GET_FLEXIBLE_ARRAY_MEMBER_ADDR(type, member, ptr, n) \ + (type *)((char *)&(ptr)->member + (sizeof(type) * (n))) + #endif /* PP_DEBUG_H */ diff --git a/drivers/gpu/drm/amd/powerplay/inc/smumgr.h b/drivers/gpu/drm/amd/powerplay/inc/smumgr.h index 3c235f0177cd..34abfd2cde53 100644 --- a/drivers/gpu/drm/amd/powerplay/inc/smumgr.h +++ b/drivers/gpu/drm/amd/powerplay/inc/smumgr.h @@ -74,7 +74,6 @@ struct pp_smumgr_func { struct pp_smumgr { uint32_t chip_family; uint32_t chip_id; - uint32_t hw_revision; void *device; void *backend; uint32_t usec_timeout; @@ -122,6 +121,12 @@ extern int smu_allocate_memory(void *device, uint32_t size, extern int smu_free_memory(void *device, void *handle); +extern int cz_smum_init(struct pp_smumgr *smumgr); +extern int iceland_smum_init(struct pp_smumgr *smumgr); +extern int tonga_smum_init(struct pp_smumgr *smumgr); +extern int fiji_smum_init(struct pp_smumgr *smumgr); +extern int polaris10_smum_init(struct pp_smumgr *smumgr); + #define SMUM_FIELD_SHIFT(reg, field) reg##__##field##__SHIFT #define SMUM_FIELD_MASK(reg, field) reg##__##field##_MASK diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/cz_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/cz_smumgr.c index 87c023e518ab..5a44485526d2 100644 --- a/drivers/gpu/drm/amd/powerplay/smumgr/cz_smumgr.c +++ b/drivers/gpu/drm/amd/powerplay/smumgr/cz_smumgr.c @@ -89,13 +89,8 @@ static int cz_send_msg_to_smc(struct pp_smumgr *smumgr, uint16_t msg) if (result != 0) return result; - result = SMUM_WAIT_FIELD_UNEQUAL(smumgr, + return SMUM_WAIT_FIELD_UNEQUAL(smumgr, SMU_MP1_SRBM2P_RESP_0, CONTENT, 0); - - if (result != 0) - return result; - - return 0; } static int cz_set_smc_sram_address(struct pp_smumgr *smumgr, @@ -106,12 +101,12 @@ static int cz_set_smc_sram_address(struct pp_smumgr *smumgr, if (0 != (3 & smc_address)) { printk(KERN_ERR "[ powerplay ] SMC address must be 4 byte aligned\n"); - return -1; + return -EINVAL; } if (limit <= (smc_address + 3)) { printk(KERN_ERR "[ powerplay ] SMC address beyond the SMC RAM area\n"); - return -1; + return -EINVAL; } cgs_write_register(smumgr->device, mmMP0PUB_IND_INDEX_0, @@ -129,9 +124,10 @@ static int cz_write_smc_sram_dword(struct pp_smumgr *smumgr, return -EINVAL; result = cz_set_smc_sram_address(smumgr, smc_address, limit); - cgs_write_register(smumgr->device, mmMP0PUB_IND_DATA_0, value); + if (!result) + cgs_write_register(smumgr->device, mmMP0PUB_IND_DATA_0, value); - return 0; + return result; } static int cz_send_msg_to_smc_with_parameter(struct pp_smumgr *smumgr, @@ -148,7 +144,6 @@ static int cz_send_msg_to_smc_with_parameter(struct pp_smumgr *smumgr, static int cz_request_smu_load_fw(struct pp_smumgr *smumgr) { struct cz_smumgr *cz_smu = (struct cz_smumgr *)(smumgr->backend); - int result = 0; uint32_t smc_address; if (!smumgr->reload_fw) { @@ -177,11 +172,9 @@ static int cz_request_smu_load_fw(struct pp_smumgr *smumgr) cz_send_msg_to_smc_with_parameter(smumgr, PPSMC_MSG_ExecuteJob, cz_smu->toc_entry_power_profiling_index); - result = cz_send_msg_to_smc_with_parameter(smumgr, + return cz_send_msg_to_smc_with_parameter(smumgr, PPSMC_MSG_ExecuteJob, cz_smu->toc_entry_initialize_index); - - return result; } static int cz_check_fw_load_finish(struct pp_smumgr *smumgr, @@ -195,9 +188,6 @@ static int cz_check_fw_load_finish(struct pp_smumgr *smumgr, if (smumgr == NULL || smumgr->device == NULL) return -EINVAL; - return cgs_read_register(smumgr->device, - mmSMU_MP1_SRBM2P_ARG_0); - cgs_write_register(smumgr->device, mmMP0PUB_IND_INDEX, index); for (i = 0; i < smumgr->usec_timeout; i++) { @@ -275,7 +265,10 @@ static int cz_start_smu(struct pp_smumgr *smumgr) if (smumgr->chip_id == CHIP_STONEY) fw_to_check &= ~(UCODE_ID_SDMA1_MASK | UCODE_ID_CP_MEC_JT2_MASK); - cz_request_smu_load_fw(smumgr); + ret = cz_request_smu_load_fw(smumgr); + if (ret) + printk(KERN_ERR "[ powerplay] SMU firmware load failed\n"); + cz_check_fw_load_finish(smumgr, fw_to_check); ret = cz_load_mec_firmware(smumgr); @@ -566,10 +559,7 @@ static int cz_smu_construct_toc_for_bootup(struct pp_smumgr *smumgr) cz_smu_populate_single_ucode_load_task(smumgr, CZ_SCRATCH_ENTRY_UCODE_ID_SDMA0, false); - if (smumgr->chip_id == CHIP_STONEY) - cz_smu_populate_single_ucode_load_task(smumgr, - CZ_SCRATCH_ENTRY_UCODE_ID_SDMA0, false); - else + if (smumgr->chip_id != CHIP_STONEY) cz_smu_populate_single_ucode_load_task(smumgr, CZ_SCRATCH_ENTRY_UCODE_ID_SDMA1, false); cz_smu_populate_single_ucode_load_task(smumgr, @@ -580,10 +570,7 @@ static int cz_smu_construct_toc_for_bootup(struct pp_smumgr *smumgr) CZ_SCRATCH_ENTRY_UCODE_ID_CP_ME, false); cz_smu_populate_single_ucode_load_task(smumgr, CZ_SCRATCH_ENTRY_UCODE_ID_CP_MEC_JT1, false); - if (smumgr->chip_id == CHIP_STONEY) - cz_smu_populate_single_ucode_load_task(smumgr, - CZ_SCRATCH_ENTRY_UCODE_ID_CP_MEC_JT1, false); - else + if (smumgr->chip_id != CHIP_STONEY) cz_smu_populate_single_ucode_load_task(smumgr, CZ_SCRATCH_ENTRY_UCODE_ID_CP_MEC_JT2, false); cz_smu_populate_single_ucode_load_task(smumgr, @@ -610,19 +597,12 @@ static int cz_smu_construct_toc(struct pp_smumgr *smumgr) struct cz_smumgr *cz_smu = (struct cz_smumgr *)smumgr->backend; cz_smu->toc_entry_used_count = 0; - cz_smu_initialize_toc_empty_job_list(smumgr); - cz_smu_construct_toc_for_rlc_aram_save(smumgr); - cz_smu_construct_toc_for_vddgfx_enter(smumgr); - cz_smu_construct_toc_for_vddgfx_exit(smumgr); - cz_smu_construct_toc_for_power_profiling(smumgr); - cz_smu_construct_toc_for_bootup(smumgr); - cz_smu_construct_toc_for_clock_table(smumgr); return 0; diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smumgr.c index 704ff4cc0023..8047ad221e74 100644 --- a/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smumgr.c +++ b/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smumgr.c @@ -40,7 +40,6 @@ #include "cgs_common.h" #define POLARIS10_SMC_SIZE 0x20000 -#define VOLTAGE_SCALE 4 /* Microcode file is stored in this buffer */ #define BUFFER_SIZE 80000 diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smumgr.h b/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smumgr.h index e5377aec057f..7c2445f1f043 100644 --- a/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smumgr.h +++ b/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smumgr.h @@ -26,12 +26,27 @@ #include <polaris10_ppsmc.h> #include <pp_endian.h> +#include "smu74.h" struct polaris10_avfs { enum AVFS_BTC_STATUS avfs_btc_status; uint32_t avfs_btc_param; }; +struct polaris10_pt_defaults { + uint8_t SviLoadLineEn; + uint8_t SviLoadLineVddC; + uint8_t TDC_VDDC_ThrottleReleaseLimitPerc; + uint8_t TDC_MAWt; + uint8_t TdcWaterfallCtl; + uint8_t DTEAmbientTempBase; + + uint32_t DisplayCac; + uint32_t BAPM_TEMP_GRADIENT; + uint16_t BAPMTI_R[SMU74_DTE_ITERATIONS * SMU74_DTE_SOURCES * SMU74_DTE_SINKS]; + uint16_t BAPMTI_RC[SMU74_DTE_ITERATIONS * SMU74_DTE_SOURCES * SMU74_DTE_SINKS]; +}; + struct polaris10_buffer_entry { uint32_t data_size; uint32_t mc_addr_low; @@ -40,6 +55,11 @@ struct polaris10_buffer_entry { unsigned long handle; }; +struct polaris10_range_table { + uint32_t trans_lower_frequency; /* in 10khz */ + uint32_t trans_upper_frequency; +}; + struct polaris10_smumgr { uint8_t *header; uint8_t *mec_image; diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/smumgr.c index cf3cabee8918..bbeb786db003 100644 --- a/drivers/gpu/drm/amd/powerplay/smumgr/smumgr.c +++ b/drivers/gpu/drm/amd/powerplay/smumgr/smumgr.c @@ -28,11 +28,7 @@ #include "smumgr.h" #include "cgs_common.h" #include "linux/delay.h" -#include "cz_smumgr.h" -#include "tonga_smumgr.h" -#include "iceland_smumgr.h" -#include "fiji_smumgr.h" -#include "polaris10_smumgr.h" + int smum_init(struct amd_pp_init *pp_init, struct pp_instance *handle) { @@ -48,7 +44,6 @@ int smum_init(struct amd_pp_init *pp_init, struct pp_instance *handle) smumgr->device = pp_init->device; smumgr->chip_family = pp_init->chip_family; smumgr->chip_id = pp_init->chip_id; - smumgr->hw_revision = pp_init->rev_id; smumgr->usec_timeout = AMD_MAX_USEC_TIMEOUT; smumgr->reload_fw = 1; handle->smu_mgr = smumgr; diff --git a/drivers/gpu/drm/bridge/dw-hdmi.c b/drivers/gpu/drm/bridge/dw-hdmi.c index cdf39aa3943c..66ad8e6fb11e 100644 --- a/drivers/gpu/drm/bridge/dw-hdmi.c +++ b/drivers/gpu/drm/bridge/dw-hdmi.c @@ -1813,9 +1813,6 @@ void dw_hdmi_unbind(struct device *dev, struct device *master, void *data) /* Disable all interrupts */ hdmi_writeb(hdmi, ~0, HDMI_IH_MUTE_PHY_STAT0); - hdmi->connector.funcs->destroy(&hdmi->connector); - hdmi->encoder->funcs->destroy(hdmi->encoder); - clk_disable_unprepare(hdmi->iahb_clk); clk_disable_unprepare(hdmi->isfr_clk); i2c_put_adapter(hdmi->ddc); diff --git a/drivers/gpu/drm/drm_dp_helper.c b/drivers/gpu/drm/drm_dp_helper.c index 48c122da27a6..ac3924a877e0 100644 --- a/drivers/gpu/drm/drm_dp_helper.c +++ b/drivers/gpu/drm/drm_dp_helper.c @@ -440,6 +440,179 @@ int drm_dp_link_configure(struct drm_dp_aux *aux, struct drm_dp_link *link) } EXPORT_SYMBOL(drm_dp_link_configure); +/** + * drm_dp_downstream_max_clock() - extract branch device max + * pixel rate for legacy VGA + * converter or max TMDS clock + * rate for others + * @dpcd: DisplayPort configuration data + * @port_cap: port capabilities + * + * Returns max clock in kHz on success or 0 if max clock not defined + */ +int drm_dp_downstream_max_clock(const u8 dpcd[DP_RECEIVER_CAP_SIZE], + const u8 port_cap[4]) +{ + int type = port_cap[0] & DP_DS_PORT_TYPE_MASK; + bool detailed_cap_info = dpcd[DP_DOWNSTREAMPORT_PRESENT] & + DP_DETAILED_CAP_INFO_AVAILABLE; + + if (!detailed_cap_info) + return 0; + + switch (type) { + case DP_DS_PORT_TYPE_VGA: + return port_cap[1] * 8 * 1000; + case DP_DS_PORT_TYPE_DVI: + case DP_DS_PORT_TYPE_HDMI: + case DP_DS_PORT_TYPE_DP_DUALMODE: + return port_cap[1] * 2500; + default: + return 0; + } +} +EXPORT_SYMBOL(drm_dp_downstream_max_clock); + +/** + * drm_dp_downstream_max_bpc() - extract branch device max + * bits per component + * @dpcd: DisplayPort configuration data + * @port_cap: port capabilities + * + * Returns max bpc on success or 0 if max bpc not defined + */ +int drm_dp_downstream_max_bpc(const u8 dpcd[DP_RECEIVER_CAP_SIZE], + const u8 port_cap[4]) +{ + int type = port_cap[0] & DP_DS_PORT_TYPE_MASK; + bool detailed_cap_info = dpcd[DP_DOWNSTREAMPORT_PRESENT] & + DP_DETAILED_CAP_INFO_AVAILABLE; + int bpc; + + if (!detailed_cap_info) + return 0; + + switch (type) { + case DP_DS_PORT_TYPE_VGA: + case DP_DS_PORT_TYPE_DVI: + case DP_DS_PORT_TYPE_HDMI: + case DP_DS_PORT_TYPE_DP_DUALMODE: + bpc = port_cap[2] & DP_DS_MAX_BPC_MASK; + + switch (bpc) { + case DP_DS_8BPC: + return 8; + case DP_DS_10BPC: + return 10; + case DP_DS_12BPC: + return 12; + case DP_DS_16BPC: + return 16; + } + default: + return 0; + } +} +EXPORT_SYMBOL(drm_dp_downstream_max_bpc); + +/** + * drm_dp_downstream_id() - identify branch device + * @aux: DisplayPort AUX channel + * @id: DisplayPort branch device id + * + * Returns branch device id on success or NULL on failure + */ +int drm_dp_downstream_id(struct drm_dp_aux *aux, char id[6]) +{ + return drm_dp_dpcd_read(aux, DP_BRANCH_ID, id, 6); +} +EXPORT_SYMBOL(drm_dp_downstream_id); + +/** + * drm_dp_downstream_debug() - debug DP branch devices + * @m: pointer for debugfs file + * @dpcd: DisplayPort configuration data + * @port_cap: port capabilities + * @aux: DisplayPort AUX channel + * + */ +void drm_dp_downstream_debug(struct seq_file *m, + const u8 dpcd[DP_RECEIVER_CAP_SIZE], + const u8 port_cap[4], struct drm_dp_aux *aux) +{ + bool detailed_cap_info = dpcd[DP_DOWNSTREAMPORT_PRESENT] & + DP_DETAILED_CAP_INFO_AVAILABLE; + int clk; + int bpc; + char id[6]; + int len; + uint8_t rev[2]; + int type = port_cap[0] & DP_DS_PORT_TYPE_MASK; + bool branch_device = dpcd[DP_DOWNSTREAMPORT_PRESENT] & + DP_DWN_STRM_PORT_PRESENT; + + seq_printf(m, "\tDP branch device present: %s\n", + branch_device ? "yes" : "no"); + + if (!branch_device) + return; + + switch (type) { + case DP_DS_PORT_TYPE_DP: + seq_puts(m, "\t\tType: DisplayPort\n"); + break; + case DP_DS_PORT_TYPE_VGA: + seq_puts(m, "\t\tType: VGA\n"); + break; + case DP_DS_PORT_TYPE_DVI: + seq_puts(m, "\t\tType: DVI\n"); + break; + case DP_DS_PORT_TYPE_HDMI: + seq_puts(m, "\t\tType: HDMI\n"); + break; + case DP_DS_PORT_TYPE_NON_EDID: + seq_puts(m, "\t\tType: others without EDID support\n"); + break; + case DP_DS_PORT_TYPE_DP_DUALMODE: + seq_puts(m, "\t\tType: DP++\n"); + break; + case DP_DS_PORT_TYPE_WIRELESS: + seq_puts(m, "\t\tType: Wireless\n"); + break; + default: + seq_puts(m, "\t\tType: N/A\n"); + } + + drm_dp_downstream_id(aux, id); + seq_printf(m, "\t\tID: %s\n", id); + + len = drm_dp_dpcd_read(aux, DP_BRANCH_HW_REV, &rev[0], 1); + if (len > 0) + seq_printf(m, "\t\tHW: %d.%d\n", + (rev[0] & 0xf0) >> 4, rev[0] & 0xf); + + len = drm_dp_dpcd_read(aux, DP_BRANCH_SW_REV, &rev, 2); + if (len > 0) + seq_printf(m, "\t\tSW: %d.%d\n", rev[0], rev[1]); + + if (detailed_cap_info) { + clk = drm_dp_downstream_max_clock(dpcd, port_cap); + + if (clk > 0) { + if (type == DP_DS_PORT_TYPE_VGA) + seq_printf(m, "\t\tMax dot clock: %d kHz\n", clk); + else + seq_printf(m, "\t\tMax TMDS clock: %d kHz\n", clk); + } + + bpc = drm_dp_downstream_max_bpc(dpcd, port_cap); + + if (bpc > 0) + seq_printf(m, "\t\tMax bpc: %d\n", bpc); + } +} +EXPORT_SYMBOL(drm_dp_downstream_debug); + /* * I2C-over-AUX implementation */ diff --git a/drivers/gpu/drm/drm_global.c b/drivers/gpu/drm/drm_global.c index 3d2e91c4d78e..b404287abb97 100644 --- a/drivers/gpu/drm/drm_global.c +++ b/drivers/gpu/drm/drm_global.c @@ -65,30 +65,34 @@ void drm_global_release(void) int drm_global_item_ref(struct drm_global_reference *ref) { - int ret; + int ret = 0; struct drm_global_item *item = &glob[ref->global_type]; mutex_lock(&item->mutex); if (item->refcount == 0) { - item->object = kzalloc(ref->size, GFP_KERNEL); - if (unlikely(item->object == NULL)) { + ref->object = kzalloc(ref->size, GFP_KERNEL); + if (unlikely(ref->object == NULL)) { ret = -ENOMEM; - goto out_err; + goto error_unlock; } - - ref->object = item->object; ret = ref->init(ref); if (unlikely(ret != 0)) - goto out_err; + goto error_free; + item->object = ref->object; + } else { + ref->object = item->object; } + ++item->refcount; - ref->object = item->object; mutex_unlock(&item->mutex); return 0; -out_err: + +error_free: + kfree(ref->object); + ref->object = NULL; +error_unlock: mutex_unlock(&item->mutex); - item->object = NULL; return ret; } EXPORT_SYMBOL(drm_global_item_ref); diff --git a/drivers/gpu/drm/fsl-dcu/fsl_dcu_drm_drv.c b/drivers/gpu/drm/fsl-dcu/fsl_dcu_drm_drv.c index 0d2ae94e4c67..0884c45aefe8 100644 --- a/drivers/gpu/drm/fsl-dcu/fsl_dcu_drm_drv.c +++ b/drivers/gpu/drm/fsl-dcu/fsl_dcu_drm_drv.c @@ -270,7 +270,7 @@ static int fsl_dcu_drm_pm_resume(struct device *dev) ret = clk_prepare_enable(fsl_dev->pix_clk); if (ret < 0) { dev_err(dev, "failed to enable pix clk\n"); - return ret; + goto disable_dcu_clk; } fsl_dcu_drm_init_planes(fsl_dev->drm); @@ -284,6 +284,10 @@ static int fsl_dcu_drm_pm_resume(struct device *dev) enable_irq(fsl_dev->irq); return 0; + +disable_dcu_clk: + clk_disable_unprepare(fsl_dev->clk); + return ret; } #endif @@ -330,6 +334,7 @@ static int fsl_dcu_drm_probe(struct platform_device *pdev) const char *pix_clk_in_name; const struct of_device_id *id; int ret; + u8 div_ratio_shift = 0; fsl_dev = devm_kzalloc(dev, sizeof(*fsl_dev), GFP_KERNEL); if (!fsl_dev) @@ -382,11 +387,14 @@ static int fsl_dcu_drm_probe(struct platform_device *pdev) pix_clk_in = fsl_dev->clk; } + if (of_property_read_bool(dev->of_node, "big-endian")) + div_ratio_shift = 24; + pix_clk_in_name = __clk_get_name(pix_clk_in); snprintf(pix_clk_name, sizeof(pix_clk_name), "%s_pix", pix_clk_in_name); fsl_dev->pix_clk = clk_register_divider(dev, pix_clk_name, pix_clk_in_name, 0, base + DCU_DIV_RATIO, - 0, 8, CLK_DIVIDER_ROUND_CLOSEST, NULL); + div_ratio_shift, 8, CLK_DIVIDER_ROUND_CLOSEST, NULL); if (IS_ERR(fsl_dev->pix_clk)) { dev_err(dev, "failed to register pix clk\n"); ret = PTR_ERR(fsl_dev->pix_clk); diff --git a/drivers/gpu/drm/fsl-dcu/fsl_tcon.c b/drivers/gpu/drm/fsl-dcu/fsl_tcon.c index bca09ea24632..3194e544ee27 100644 --- a/drivers/gpu/drm/fsl-dcu/fsl_tcon.c +++ b/drivers/gpu/drm/fsl-dcu/fsl_tcon.c @@ -57,10 +57,7 @@ static int fsl_tcon_init_regmap(struct device *dev, tcon->regs = devm_regmap_init_mmio(dev, regs, &fsl_tcon_regmap_config); - if (IS_ERR(tcon->regs)) - return PTR_ERR(tcon->regs); - - return 0; + return PTR_ERR_OR_ZERO(tcon->regs); } struct fsl_tcon *fsl_tcon_init(struct device *dev) diff --git a/drivers/gpu/drm/i915/Makefile b/drivers/gpu/drm/i915/Makefile index a7da24640e88..a998c2bce70a 100644 --- a/drivers/gpu/drm/i915/Makefile +++ b/drivers/gpu/drm/i915/Makefile @@ -16,6 +16,7 @@ i915-y := i915_drv.o \ i915_params.o \ i915_pci.o \ i915_suspend.o \ + i915_sw_fence.o \ i915_sysfs.o \ intel_csr.o \ intel_device_info.o \ diff --git a/drivers/gpu/drm/i915/i915_cmd_parser.c b/drivers/gpu/drm/i915/i915_cmd_parser.c index 3c72b3b103e7..70980f82a15b 100644 --- a/drivers/gpu/drm/i915/i915_cmd_parser.c +++ b/drivers/gpu/drm/i915/i915_cmd_parser.c @@ -984,7 +984,7 @@ static u32 *copy_batch(struct drm_i915_gem_object *dst_obj, src = ERR_PTR(-ENODEV); if (src_needs_clflush && - i915_memcpy_from_wc((void *)(uintptr_t)batch_start_offset, 0, 0)) { + i915_memcpy_from_wc((void *)(uintptr_t)batch_start_offset, NULL, 0)) { src = i915_gem_object_pin_map(src_obj, I915_MAP_WC); if (!IS_ERR(src)) { i915_memcpy_from_wc(dst, diff --git a/drivers/gpu/drm/i915/i915_debugfs.c b/drivers/gpu/drm/i915/i915_debugfs.c index a95d7bc81fb9..27b0e34dadec 100644 --- a/drivers/gpu/drm/i915/i915_debugfs.c +++ b/drivers/gpu/drm/i915/i915_debugfs.c @@ -40,6 +40,11 @@ #include <drm/i915_drm.h> #include "i915_drv.h" +static inline struct drm_i915_private *node_to_i915(struct drm_info_node *node) +{ + return to_i915(node->minor->dev); +} + /* As the drm_debugfs_init() routines are called before dev->dev_private is * allocated we need to hook into the minor for release. */ static int @@ -57,7 +62,7 @@ drm_add_fake_info_node(struct drm_minor *minor, node->minor = minor; node->dent = ent; - node->info_ent = (void *) key; + node->info_ent = (void *)key; mutex_lock(&minor->debugfs_lock); list_add(&node->list, &minor->debugfs_list); @@ -68,12 +73,11 @@ drm_add_fake_info_node(struct drm_minor *minor, static int i915_capabilities(struct seq_file *m, void *data) { - struct drm_info_node *node = m->private; - struct drm_device *dev = node->minor->dev; - const struct intel_device_info *info = INTEL_INFO(dev); + struct drm_i915_private *dev_priv = node_to_i915(m->private); + const struct intel_device_info *info = INTEL_INFO(dev_priv); - seq_printf(m, "gen: %d\n", info->gen); - seq_printf(m, "pch: %d\n", INTEL_PCH_TYPE(dev)); + seq_printf(m, "gen: %d\n", INTEL_GEN(dev_priv)); + seq_printf(m, "pch: %d\n", INTEL_PCH_TYPE(dev_priv)); #define PRINT_FLAG(x) seq_printf(m, #x ": %s\n", yesno(info->x)) #define SEP_SEMICOLON ; DEV_INFO_FOR_EACH_FLAG(PRINT_FLAG, SEP_SEMICOLON); @@ -155,7 +159,7 @@ describe_obj(struct seq_file *m, struct drm_i915_gem_object *obj) seq_printf(m, "] %x %s%s%s", i915_gem_active_get_seqno(&obj->last_write, &obj->base.dev->struct_mutex), - i915_cache_level_str(to_i915(obj->base.dev), obj->cache_level), + i915_cache_level_str(dev_priv, obj->cache_level), obj->dirty ? " dirty" : "", obj->madv == I915_MADV_DONTNEED ? " purgeable" : ""); if (obj->base.name) @@ -195,7 +199,7 @@ describe_obj(struct seq_file *m, struct drm_i915_gem_object *obj) } engine = i915_gem_active_get_engine(&obj->last_write, - &obj->base.dev->struct_mutex); + &dev_priv->drm.struct_mutex); if (engine) seq_printf(m, " (%s)", engine->name); @@ -221,9 +225,8 @@ static int obj_rank_by_stolen(void *priv, static int i915_gem_stolen_list_info(struct seq_file *m, void *data) { - struct drm_info_node *node = m->private; - struct drm_device *dev = node->minor->dev; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = node_to_i915(m->private); + struct drm_device *dev = &dev_priv->drm; struct drm_i915_gem_object *obj; u64 total_obj_size, total_gtt_size; LIST_HEAD(stolen); @@ -365,29 +368,29 @@ static int per_file_ctx_stats(int id, void *ptr, void *data) static void print_context_stats(struct seq_file *m, struct drm_i915_private *dev_priv) { + struct drm_device *dev = &dev_priv->drm; struct file_stats stats; struct drm_file *file; memset(&stats, 0, sizeof(stats)); - mutex_lock(&dev_priv->drm.struct_mutex); + mutex_lock(&dev->struct_mutex); if (dev_priv->kernel_context) per_file_ctx_stats(0, dev_priv->kernel_context, &stats); - list_for_each_entry(file, &dev_priv->drm.filelist, lhead) { + list_for_each_entry(file, &dev->filelist, lhead) { struct drm_i915_file_private *fpriv = file->driver_priv; idr_for_each(&fpriv->context_idr, per_file_ctx_stats, &stats); } - mutex_unlock(&dev_priv->drm.struct_mutex); + mutex_unlock(&dev->struct_mutex); print_file_stats(m, "[k]contexts", stats); } -static int i915_gem_object_info(struct seq_file *m, void* data) +static int i915_gem_object_info(struct seq_file *m, void *data) { - struct drm_info_node *node = m->private; - struct drm_device *dev = node->minor->dev; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = node_to_i915(m->private); + struct drm_device *dev = &dev_priv->drm; struct i915_ggtt *ggtt = &dev_priv->ggtt; u32 count, mapped_count, purgeable_count, dpy_count; u64 size, mapped_size, purgeable_size, dpy_size; @@ -497,9 +500,9 @@ static int i915_gem_object_info(struct seq_file *m, void* data) static int i915_gem_gtt_info(struct seq_file *m, void *data) { struct drm_info_node *node = m->private; - struct drm_device *dev = node->minor->dev; - struct drm_i915_private *dev_priv = to_i915(dev); - bool show_pin_display_only = !!data; + struct drm_i915_private *dev_priv = node_to_i915(node); + struct drm_device *dev = &dev_priv->drm; + bool show_pin_display_only = !!node->info_ent->data; struct drm_i915_gem_object *obj; u64 total_obj_size, total_gtt_size; int count, ret; @@ -531,9 +534,8 @@ static int i915_gem_gtt_info(struct seq_file *m, void *data) static int i915_gem_pageflip_info(struct seq_file *m, void *data) { - struct drm_info_node *node = m->private; - struct drm_device *dev = node->minor->dev; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = node_to_i915(m->private); + struct drm_device *dev = &dev_priv->drm; struct intel_crtc *crtc; int ret; @@ -580,7 +582,7 @@ static int i915_gem_pageflip_info(struct seq_file *m, void *data) intel_crtc_get_vblank_counter(crtc)); seq_printf(m, "%d prepares\n", atomic_read(&work->pending)); - if (INTEL_INFO(dev)->gen >= 4) + if (INTEL_GEN(dev_priv) >= 4) addr = I915_HI_DISPBASE(I915_READ(DSPSURF(crtc->plane))); else addr = I915_READ(DSPADDR(crtc->plane)); @@ -601,9 +603,8 @@ static int i915_gem_pageflip_info(struct seq_file *m, void *data) static int i915_gem_batch_pool_info(struct seq_file *m, void *data) { - struct drm_info_node *node = m->private; - struct drm_device *dev = node->minor->dev; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = node_to_i915(m->private); + struct drm_device *dev = &dev_priv->drm; struct drm_i915_gem_object *obj; struct intel_engine_cs *engine; int total = 0; @@ -646,9 +647,8 @@ static int i915_gem_batch_pool_info(struct seq_file *m, void *data) static int i915_gem_request_info(struct seq_file *m, void *data) { - struct drm_info_node *node = m->private; - struct drm_device *dev = node->minor->dev; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = node_to_i915(m->private); + struct drm_device *dev = &dev_priv->drm; struct intel_engine_cs *engine; struct drm_i915_gem_request *req; int ret, any; @@ -713,41 +713,25 @@ static void i915_ring_seqno_info(struct seq_file *m, static int i915_gem_seqno_info(struct seq_file *m, void *data) { - struct drm_info_node *node = m->private; - struct drm_device *dev = node->minor->dev; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = node_to_i915(m->private); struct intel_engine_cs *engine; - int ret; - - ret = mutex_lock_interruptible(&dev->struct_mutex); - if (ret) - return ret; - intel_runtime_pm_get(dev_priv); for_each_engine(engine, dev_priv) i915_ring_seqno_info(m, engine); - intel_runtime_pm_put(dev_priv); - mutex_unlock(&dev->struct_mutex); - return 0; } static int i915_interrupt_info(struct seq_file *m, void *data) { - struct drm_info_node *node = m->private; - struct drm_device *dev = node->minor->dev; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = node_to_i915(m->private); struct intel_engine_cs *engine; - int ret, i, pipe; + int i, pipe; - ret = mutex_lock_interruptible(&dev->struct_mutex); - if (ret) - return ret; intel_runtime_pm_get(dev_priv); - if (IS_CHERRYVIEW(dev)) { + if (IS_CHERRYVIEW(dev_priv)) { seq_printf(m, "Master Interrupt Control:\t%08x\n", I915_READ(GEN8_MASTER_IRQ)); @@ -786,7 +770,7 @@ static int i915_interrupt_info(struct seq_file *m, void *data) I915_READ(GEN8_PCU_IIR)); seq_printf(m, "PCU interrupt enable:\t%08x\n", I915_READ(GEN8_PCU_IER)); - } else if (INTEL_INFO(dev)->gen >= 8) { + } else if (INTEL_GEN(dev_priv) >= 8) { seq_printf(m, "Master Interrupt Control:\t%08x\n", I915_READ(GEN8_MASTER_IRQ)); @@ -842,7 +826,7 @@ static int i915_interrupt_info(struct seq_file *m, void *data) I915_READ(GEN8_PCU_IIR)); seq_printf(m, "PCU interrupt enable:\t%08x\n", I915_READ(GEN8_PCU_IER)); - } else if (IS_VALLEYVIEW(dev)) { + } else if (IS_VALLEYVIEW(dev_priv)) { seq_printf(m, "Display IER:\t%08x\n", I915_READ(VLV_IER)); seq_printf(m, "Display IIR:\t%08x\n", @@ -880,7 +864,7 @@ static int i915_interrupt_info(struct seq_file *m, void *data) seq_printf(m, "DPINVGTT:\t%08x\n", I915_READ(DPINVGTT)); - } else if (!HAS_PCH_SPLIT(dev)) { + } else if (!HAS_PCH_SPLIT(dev_priv)) { seq_printf(m, "Interrupt enable: %08x\n", I915_READ(IER)); seq_printf(m, "Interrupt identity: %08x\n", @@ -912,7 +896,7 @@ static int i915_interrupt_info(struct seq_file *m, void *data) I915_READ(GTIMR)); } for_each_engine(engine, dev_priv) { - if (INTEL_INFO(dev)->gen >= 6) { + if (INTEL_GEN(dev_priv) >= 6) { seq_printf(m, "Graphics Interrupt mask (%s): %08x\n", engine->name, I915_READ_IMR(engine)); @@ -920,16 +904,14 @@ static int i915_interrupt_info(struct seq_file *m, void *data) i915_ring_seqno_info(m, engine); } intel_runtime_pm_put(dev_priv); - mutex_unlock(&dev->struct_mutex); return 0; } static int i915_gem_fence_regs_info(struct seq_file *m, void *data) { - struct drm_info_node *node = m->private; - struct drm_device *dev = node->minor->dev; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = node_to_i915(m->private); + struct drm_device *dev = &dev_priv->drm; int i, ret; ret = mutex_lock_interruptible(&dev->struct_mutex); @@ -956,8 +938,7 @@ static int i915_gem_fence_regs_info(struct seq_file *m, void *data) static int i915_hws_info(struct seq_file *m, void *data) { struct drm_info_node *node = m->private; - struct drm_device *dev = node->minor->dev; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = node_to_i915(node); struct intel_engine_cs *engine; const u32 *hws; int i; @@ -982,33 +963,25 @@ i915_error_state_write(struct file *filp, loff_t *ppos) { struct i915_error_state_file_priv *error_priv = filp->private_data; - struct drm_device *dev = error_priv->dev; - int ret; DRM_DEBUG_DRIVER("Resetting error state\n"); - - ret = mutex_lock_interruptible(&dev->struct_mutex); - if (ret) - return ret; - - i915_destroy_error_state(dev); - mutex_unlock(&dev->struct_mutex); + i915_destroy_error_state(error_priv->dev); return cnt; } static int i915_error_state_open(struct inode *inode, struct file *file) { - struct drm_device *dev = inode->i_private; + struct drm_i915_private *dev_priv = inode->i_private; struct i915_error_state_file_priv *error_priv; error_priv = kzalloc(sizeof(*error_priv), GFP_KERNEL); if (!error_priv) return -ENOMEM; - error_priv->dev = dev; + error_priv->dev = &dev_priv->drm; - i915_error_state_get(dev, error_priv); + i915_error_state_get(&dev_priv->drm, error_priv); file->private_data = error_priv; @@ -1034,7 +1007,8 @@ static ssize_t i915_error_state_read(struct file *file, char __user *userbuf, ssize_t ret_count = 0; int ret; - ret = i915_error_state_buf_init(&error_str, to_i915(error_priv->dev), count, *pos); + ret = i915_error_state_buf_init(&error_str, + to_i915(error_priv->dev), count, *pos); if (ret) return ret; @@ -1067,16 +1041,15 @@ static const struct file_operations i915_error_state_fops = { static int i915_next_seqno_get(void *data, u64 *val) { - struct drm_device *dev = data; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = data; int ret; - ret = mutex_lock_interruptible(&dev->struct_mutex); + ret = mutex_lock_interruptible(&dev_priv->drm.struct_mutex); if (ret) return ret; *val = dev_priv->next_seqno; - mutex_unlock(&dev->struct_mutex); + mutex_unlock(&dev_priv->drm.struct_mutex); return 0; } @@ -1084,7 +1057,8 @@ i915_next_seqno_get(void *data, u64 *val) static int i915_next_seqno_set(void *data, u64 val) { - struct drm_device *dev = data; + struct drm_i915_private *dev_priv = data; + struct drm_device *dev = &dev_priv->drm; int ret; ret = mutex_lock_interruptible(&dev->struct_mutex); @@ -1103,14 +1077,13 @@ DEFINE_SIMPLE_ATTRIBUTE(i915_next_seqno_fops, static int i915_frequency_info(struct seq_file *m, void *unused) { - struct drm_info_node *node = m->private; - struct drm_device *dev = node->minor->dev; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = node_to_i915(m->private); + struct drm_device *dev = &dev_priv->drm; int ret = 0; intel_runtime_pm_get(dev_priv); - if (IS_GEN5(dev)) { + if (IS_GEN5(dev_priv)) { u16 rgvswctl = I915_READ16(MEMSWCTL); u16 rgvstat = I915_READ16(MEMSTAT_ILK); @@ -1120,7 +1093,7 @@ static int i915_frequency_info(struct seq_file *m, void *unused) MEMSTAT_VID_SHIFT); seq_printf(m, "Current P-state: %d\n", (rgvstat & MEMSTAT_PSTATE_MASK) >> MEMSTAT_PSTATE_SHIFT); - } else if (IS_VALLEYVIEW(dev) || IS_CHERRYVIEW(dev)) { + } else if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv)) { u32 freq_sts; mutex_lock(&dev_priv->rps.hw_lock); @@ -1147,7 +1120,7 @@ static int i915_frequency_info(struct seq_file *m, void *unused) "efficient (RPe) frequency: %d MHz\n", intel_gpu_freq(dev_priv, dev_priv->rps.efficient_freq)); mutex_unlock(&dev_priv->rps.hw_lock); - } else if (INTEL_INFO(dev)->gen >= 6) { + } else if (INTEL_GEN(dev_priv) >= 6) { u32 rp_state_limits; u32 gt_perf_status; u32 rp_state_cap; @@ -1159,7 +1132,7 @@ static int i915_frequency_info(struct seq_file *m, void *unused) int max_freq; rp_state_limits = I915_READ(GEN6_RP_STATE_LIMITS); - if (IS_BROXTON(dev)) { + if (IS_BROXTON(dev_priv)) { rp_state_cap = I915_READ(BXT_RP_STATE_CAP); gt_perf_status = I915_READ(BXT_GT_PERF_STATUS); } else { @@ -1175,11 +1148,11 @@ static int i915_frequency_info(struct seq_file *m, void *unused) intel_uncore_forcewake_get(dev_priv, FORCEWAKE_ALL); reqf = I915_READ(GEN6_RPNSWREQ); - if (IS_GEN9(dev)) + if (IS_GEN9(dev_priv)) reqf >>= 23; else { reqf &= ~GEN6_TURBO_DISABLE; - if (IS_HASWELL(dev) || IS_BROADWELL(dev)) + if (IS_HASWELL(dev_priv) || IS_BROADWELL(dev_priv)) reqf >>= 24; else reqf >>= 25; @@ -1197,9 +1170,9 @@ static int i915_frequency_info(struct seq_file *m, void *unused) rpdownei = I915_READ(GEN6_RP_CUR_DOWN_EI) & GEN6_CURIAVG_MASK; rpcurdown = I915_READ(GEN6_RP_CUR_DOWN) & GEN6_CURBSYTAVG_MASK; rpprevdown = I915_READ(GEN6_RP_PREV_DOWN) & GEN6_CURBSYTAVG_MASK; - if (IS_GEN9(dev)) + if (IS_GEN9(dev_priv)) cagf = (rpstat & GEN9_CAGF_MASK) >> GEN9_CAGF_SHIFT; - else if (IS_HASWELL(dev) || IS_BROADWELL(dev)) + else if (IS_HASWELL(dev_priv) || IS_BROADWELL(dev_priv)) cagf = (rpstat & HSW_CAGF_MASK) >> HSW_CAGF_SHIFT; else cagf = (rpstat & GEN6_CAGF_MASK) >> GEN6_CAGF_SHIFT; @@ -1208,7 +1181,7 @@ static int i915_frequency_info(struct seq_file *m, void *unused) intel_uncore_forcewake_put(dev_priv, FORCEWAKE_ALL); mutex_unlock(&dev->struct_mutex); - if (IS_GEN6(dev) || IS_GEN7(dev)) { + if (IS_GEN6(dev_priv) || IS_GEN7(dev_priv)) { pm_ier = I915_READ(GEN6_PMIER); pm_imr = I915_READ(GEN6_PMIMR); pm_isr = I915_READ(GEN6_PMISR); @@ -1226,7 +1199,7 @@ static int i915_frequency_info(struct seq_file *m, void *unused) seq_printf(m, "pm_intr_keep: 0x%08x\n", dev_priv->rps.pm_intr_keep); seq_printf(m, "GT_PERF_STATUS: 0x%08x\n", gt_perf_status); seq_printf(m, "Render p-state ratio: %d\n", - (gt_perf_status & (IS_GEN9(dev) ? 0x1ff00 : 0xff00)) >> 8); + (gt_perf_status & (IS_GEN9(dev_priv) ? 0x1ff00 : 0xff00)) >> 8); seq_printf(m, "Render p-state VID: %d\n", gt_perf_status & 0xff); seq_printf(m, "Render p-state limit: %d\n", @@ -1255,22 +1228,22 @@ static int i915_frequency_info(struct seq_file *m, void *unused) seq_printf(m, "Down threshold: %d%%\n", dev_priv->rps.down_threshold); - max_freq = (IS_BROXTON(dev) ? rp_state_cap >> 0 : + max_freq = (IS_BROXTON(dev_priv) ? rp_state_cap >> 0 : rp_state_cap >> 16) & 0xff; - max_freq *= (IS_SKYLAKE(dev) || IS_KABYLAKE(dev) ? + max_freq *= (IS_SKYLAKE(dev_priv) || IS_KABYLAKE(dev_priv) ? GEN9_FREQ_SCALER : 1); seq_printf(m, "Lowest (RPN) frequency: %dMHz\n", intel_gpu_freq(dev_priv, max_freq)); max_freq = (rp_state_cap & 0xff00) >> 8; - max_freq *= (IS_SKYLAKE(dev) || IS_KABYLAKE(dev) ? + max_freq *= (IS_SKYLAKE(dev_priv) || IS_KABYLAKE(dev_priv) ? GEN9_FREQ_SCALER : 1); seq_printf(m, "Nominal (RP1) frequency: %dMHz\n", intel_gpu_freq(dev_priv, max_freq)); - max_freq = (IS_BROXTON(dev) ? rp_state_cap >> 16 : + max_freq = (IS_BROXTON(dev_priv) ? rp_state_cap >> 16 : rp_state_cap >> 0) & 0xff; - max_freq *= (IS_SKYLAKE(dev) || IS_KABYLAKE(dev) ? + max_freq *= (IS_SKYLAKE(dev_priv) || IS_KABYLAKE(dev_priv) ? GEN9_FREQ_SCALER : 1); seq_printf(m, "Max non-overclocked (RP0) frequency: %dMHz\n", intel_gpu_freq(dev_priv, max_freq)); @@ -1306,9 +1279,7 @@ out: static int i915_hangcheck_info(struct seq_file *m, void *unused) { - struct drm_info_node *node = m->private; - struct drm_device *dev = node->minor->dev; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = node_to_i915(m->private); struct intel_engine_cs *engine; u64 acthd[I915_NUM_ENGINES]; u32 seqno[I915_NUM_ENGINES]; @@ -1316,6 +1287,15 @@ static int i915_hangcheck_info(struct seq_file *m, void *unused) enum intel_engine_id id; int j; + if (test_bit(I915_WEDGED, &dev_priv->gpu_error.flags)) + seq_printf(m, "Wedged\n"); + if (test_bit(I915_RESET_IN_PROGRESS, &dev_priv->gpu_error.flags)) + seq_printf(m, "Reset in progress\n"); + if (waitqueue_active(&dev_priv->gpu_error.wait_queue)) + seq_printf(m, "Waiter holding struct mutex\n"); + if (waitqueue_active(&dev_priv->gpu_error.reset_queue)) + seq_printf(m, "struct_mutex blocked for reset\n"); + if (!i915.enable_hangcheck) { seq_printf(m, "Hangcheck disabled\n"); return 0; @@ -1376,9 +1356,8 @@ static int i915_hangcheck_info(struct seq_file *m, void *unused) static int ironlake_drpc_info(struct seq_file *m) { - struct drm_info_node *node = m->private; - struct drm_device *dev = node->minor->dev; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = node_to_i915(m->private); + struct drm_device *dev = &dev_priv->drm; u32 rgvmodectl, rstdbyctl; u16 crstandvid; int ret; @@ -1444,9 +1423,7 @@ static int ironlake_drpc_info(struct seq_file *m) static int i915_forcewake_domains(struct seq_file *m, void *data) { - struct drm_info_node *node = m->private; - struct drm_device *dev = node->minor->dev; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = node_to_i915(m->private); struct intel_uncore_forcewake_domain *fw_domain; spin_lock_irq(&dev_priv->uncore.lock); @@ -1462,9 +1439,7 @@ static int i915_forcewake_domains(struct seq_file *m, void *data) static int vlv_drpc_info(struct seq_file *m) { - struct drm_info_node *node = m->private; - struct drm_device *dev = node->minor->dev; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = node_to_i915(m->private); u32 rpmodectl1, rcctl1, pw_status; intel_runtime_pm_get(dev_priv); @@ -1502,9 +1477,8 @@ static int vlv_drpc_info(struct seq_file *m) static int gen6_drpc_info(struct seq_file *m) { - struct drm_info_node *node = m->private; - struct drm_device *dev = node->minor->dev; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = node_to_i915(m->private); + struct drm_device *dev = &dev_priv->drm; u32 rpmodectl1, gt_core_status, rcctl1, rc6vids = 0; u32 gen9_powergate_enable = 0, gen9_powergate_status = 0; unsigned forcewake_count; @@ -1534,7 +1508,7 @@ static int gen6_drpc_info(struct seq_file *m) rpmodectl1 = I915_READ(GEN6_RP_CONTROL); rcctl1 = I915_READ(GEN6_RC_CONTROL); - if (INTEL_INFO(dev)->gen >= 9) { + if (INTEL_GEN(dev_priv) >= 9) { gen9_powergate_enable = I915_READ(GEN9_PG_ENABLE); gen9_powergate_status = I915_READ(GEN9_PWRGT_DOMAIN_STATUS); } @@ -1556,7 +1530,7 @@ static int gen6_drpc_info(struct seq_file *m) yesno(rcctl1 & GEN6_RC_CTL_RC1e_ENABLE)); seq_printf(m, "RC6 Enabled: %s\n", yesno(rcctl1 & GEN6_RC_CTL_RC6_ENABLE)); - if (INTEL_INFO(dev)->gen >= 9) { + if (INTEL_GEN(dev_priv) >= 9) { seq_printf(m, "Render Well Gating Enabled: %s\n", yesno(gen9_powergate_enable & GEN9_RENDER_PG_ENABLE)); seq_printf(m, "Media Well Gating Enabled: %s\n", @@ -1590,7 +1564,7 @@ static int gen6_drpc_info(struct seq_file *m) seq_printf(m, "Core Power Down: %s\n", yesno(gt_core_status & GEN6_CORE_CPD_STATE_MASK)); - if (INTEL_INFO(dev)->gen >= 9) { + if (INTEL_GEN(dev_priv) >= 9) { seq_printf(m, "Render Power Well: %s\n", (gen9_powergate_status & GEN9_PWRGT_RENDER_STATUS_MASK) ? "Up" : "Down"); @@ -1620,12 +1594,11 @@ static int gen6_drpc_info(struct seq_file *m) static int i915_drpc_info(struct seq_file *m, void *unused) { - struct drm_info_node *node = m->private; - struct drm_device *dev = node->minor->dev; + struct drm_i915_private *dev_priv = node_to_i915(m->private); - if (IS_VALLEYVIEW(dev) || IS_CHERRYVIEW(dev)) + if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv)) return vlv_drpc_info(m); - else if (INTEL_INFO(dev)->gen >= 6) + else if (INTEL_GEN(dev_priv) >= 6) return gen6_drpc_info(m); else return ironlake_drpc_info(m); @@ -1633,9 +1606,7 @@ static int i915_drpc_info(struct seq_file *m, void *unused) static int i915_frontbuffer_tracking(struct seq_file *m, void *unused) { - struct drm_info_node *node = m->private; - struct drm_device *dev = node->minor->dev; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = node_to_i915(m->private); seq_printf(m, "FB tracking busy bits: 0x%08x\n", dev_priv->fb_tracking.busy_bits); @@ -1648,11 +1619,9 @@ static int i915_frontbuffer_tracking(struct seq_file *m, void *unused) static int i915_fbc_status(struct seq_file *m, void *unused) { - struct drm_info_node *node = m->private; - struct drm_device *dev = node->minor->dev; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = node_to_i915(m->private); - if (!HAS_FBC(dev)) { + if (!HAS_FBC(dev_priv)) { seq_puts(m, "FBC unsupported on this chipset\n"); return 0; } @@ -1666,7 +1635,7 @@ static int i915_fbc_status(struct seq_file *m, void *unused) seq_printf(m, "FBC disabled: %s\n", dev_priv->fbc.no_fbc_reason); - if (INTEL_INFO(dev_priv)->gen >= 7) + if (INTEL_GEN(dev_priv) >= 7) seq_printf(m, "Compressing: %s\n", yesno(I915_READ(FBC_STATUS2) & FBC_COMPRESSION_MASK)); @@ -1679,10 +1648,9 @@ static int i915_fbc_status(struct seq_file *m, void *unused) static int i915_fbc_fc_get(void *data, u64 *val) { - struct drm_device *dev = data; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = data; - if (INTEL_INFO(dev)->gen < 7 || !HAS_FBC(dev)) + if (INTEL_GEN(dev_priv) < 7 || !HAS_FBC(dev_priv)) return -ENODEV; *val = dev_priv->fbc.false_color; @@ -1692,11 +1660,10 @@ static int i915_fbc_fc_get(void *data, u64 *val) static int i915_fbc_fc_set(void *data, u64 val) { - struct drm_device *dev = data; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = data; u32 reg; - if (INTEL_INFO(dev)->gen < 7 || !HAS_FBC(dev)) + if (INTEL_GEN(dev_priv) < 7 || !HAS_FBC(dev_priv)) return -ENODEV; mutex_lock(&dev_priv->fbc.lock); @@ -1718,11 +1685,9 @@ DEFINE_SIMPLE_ATTRIBUTE(i915_fbc_fc_fops, static int i915_ips_status(struct seq_file *m, void *unused) { - struct drm_info_node *node = m->private; - struct drm_device *dev = node->minor->dev; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = node_to_i915(m->private); - if (!HAS_IPS(dev)) { + if (!HAS_IPS(dev_priv)) { seq_puts(m, "not supported\n"); return 0; } @@ -1732,7 +1697,7 @@ static int i915_ips_status(struct seq_file *m, void *unused) seq_printf(m, "Enabled by kernel parameter: %s\n", yesno(i915.enable_ips)); - if (INTEL_INFO(dev)->gen >= 8) { + if (INTEL_GEN(dev_priv) >= 8) { seq_puts(m, "Currently: unknown\n"); } else { if (I915_READ(IPS_CTL) & IPS_ENABLE) @@ -1748,23 +1713,21 @@ static int i915_ips_status(struct seq_file *m, void *unused) static int i915_sr_status(struct seq_file *m, void *unused) { - struct drm_info_node *node = m->private; - struct drm_device *dev = node->minor->dev; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = node_to_i915(m->private); bool sr_enabled = false; intel_runtime_pm_get(dev_priv); - if (HAS_PCH_SPLIT(dev)) + if (HAS_PCH_SPLIT(dev_priv)) sr_enabled = I915_READ(WM1_LP_ILK) & WM1_LP_SR_EN; - else if (IS_CRESTLINE(dev) || IS_G4X(dev) || - IS_I945G(dev) || IS_I945GM(dev)) + else if (IS_CRESTLINE(dev_priv) || IS_G4X(dev_priv) || + IS_I945G(dev_priv) || IS_I945GM(dev_priv)) sr_enabled = I915_READ(FW_BLC_SELF) & FW_BLC_SELF_EN; - else if (IS_I915GM(dev)) + else if (IS_I915GM(dev_priv)) sr_enabled = I915_READ(INSTPM) & INSTPM_SELF_EN; - else if (IS_PINEVIEW(dev)) + else if (IS_PINEVIEW(dev_priv)) sr_enabled = I915_READ(DSPFW3) & PINEVIEW_SELF_REFRESH_EN; - else if (IS_VALLEYVIEW(dev) || IS_CHERRYVIEW(dev)) + else if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv)) sr_enabled = I915_READ(FW_BLC_SELF_VLV) & FW_CSPWRDWNEN; intel_runtime_pm_put(dev_priv); @@ -1777,13 +1740,12 @@ static int i915_sr_status(struct seq_file *m, void *unused) static int i915_emon_status(struct seq_file *m, void *unused) { - struct drm_info_node *node = m->private; - struct drm_device *dev = node->minor->dev; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = node_to_i915(m->private); + struct drm_device *dev = &dev_priv->drm; unsigned long temp, chipset, gfx; int ret; - if (!IS_GEN5(dev)) + if (!IS_GEN5(dev_priv)) return -ENODEV; ret = mutex_lock_interruptible(&dev->struct_mutex); @@ -1805,14 +1767,12 @@ static int i915_emon_status(struct seq_file *m, void *unused) static int i915_ring_freq_table(struct seq_file *m, void *unused) { - struct drm_info_node *node = m->private; - struct drm_device *dev = node->minor->dev; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = node_to_i915(m->private); int ret = 0; int gpu_freq, ia_freq; unsigned int max_gpu_freq, min_gpu_freq; - if (!HAS_CORE_RING_FREQ(dev)) { + if (!HAS_LLC(dev_priv)) { seq_puts(m, "unsupported on this chipset\n"); return 0; } @@ -1823,7 +1783,7 @@ static int i915_ring_freq_table(struct seq_file *m, void *unused) if (ret) goto out; - if (IS_SKYLAKE(dev) || IS_KABYLAKE(dev)) { + if (IS_SKYLAKE(dev_priv) || IS_KABYLAKE(dev_priv)) { /* Convert GT frequency to 50 HZ units */ min_gpu_freq = dev_priv->rps.min_freq_softlimit / GEN9_FREQ_SCALER; @@ -1843,7 +1803,7 @@ static int i915_ring_freq_table(struct seq_file *m, void *unused) &ia_freq); seq_printf(m, "%d\t\t%d\t\t\t\t%d\n", intel_gpu_freq(dev_priv, (gpu_freq * - (IS_SKYLAKE(dev) || IS_KABYLAKE(dev) ? + (IS_SKYLAKE(dev_priv) || IS_KABYLAKE(dev_priv) ? GEN9_FREQ_SCALER : 1))), ((ia_freq >> 0) & 0xff) * 100, ((ia_freq >> 8) & 0xff) * 100); @@ -1858,9 +1818,8 @@ out: static int i915_opregion(struct seq_file *m, void *unused) { - struct drm_info_node *node = m->private; - struct drm_device *dev = node->minor->dev; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = node_to_i915(m->private); + struct drm_device *dev = &dev_priv->drm; struct intel_opregion *opregion = &dev_priv->opregion; int ret; @@ -1879,10 +1838,7 @@ out: static int i915_vbt(struct seq_file *m, void *unused) { - struct drm_info_node *node = m->private; - struct drm_device *dev = node->minor->dev; - struct drm_i915_private *dev_priv = to_i915(dev); - struct intel_opregion *opregion = &dev_priv->opregion; + struct intel_opregion *opregion = &node_to_i915(m->private)->opregion; if (opregion->vbt) seq_write(m, opregion->vbt, opregion->vbt_size); @@ -1892,8 +1848,8 @@ static int i915_vbt(struct seq_file *m, void *unused) static int i915_gem_framebuffer_info(struct seq_file *m, void *data) { - struct drm_info_node *node = m->private; - struct drm_device *dev = node->minor->dev; + struct drm_i915_private *dev_priv = node_to_i915(m->private); + struct drm_device *dev = &dev_priv->drm; struct intel_framebuffer *fbdev_fb = NULL; struct drm_framebuffer *drm_fb; int ret; @@ -1903,8 +1859,8 @@ static int i915_gem_framebuffer_info(struct seq_file *m, void *data) return ret; #ifdef CONFIG_DRM_FBDEV_EMULATION - if (to_i915(dev)->fbdev) { - fbdev_fb = to_intel_framebuffer(to_i915(dev)->fbdev->helper.fb); + if (dev_priv->fbdev) { + fbdev_fb = to_intel_framebuffer(dev_priv->fbdev->helper.fb); seq_printf(m, "fbcon size: %d x %d, depth %d, %d bpp, modifier 0x%llx, refcount %d, obj ", fbdev_fb->base.width, @@ -1949,9 +1905,8 @@ static void describe_ctx_ring(struct seq_file *m, struct intel_ring *ring) static int i915_context_status(struct seq_file *m, void *unused) { - struct drm_info_node *node = m->private; - struct drm_device *dev = node->minor->dev; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = node_to_i915(m->private); + struct drm_device *dev = &dev_priv->drm; struct intel_engine_cs *engine; struct i915_gem_context *ctx; int ret; @@ -2043,9 +1998,8 @@ static void i915_dump_lrc_obj(struct seq_file *m, static int i915_dump_lrc(struct seq_file *m, void *unused) { - struct drm_info_node *node = (struct drm_info_node *) m->private; - struct drm_device *dev = node->minor->dev; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = node_to_i915(m->private); + struct drm_device *dev = &dev_priv->drm; struct intel_engine_cs *engine; struct i915_gem_context *ctx; int ret; @@ -2070,9 +2024,8 @@ static int i915_dump_lrc(struct seq_file *m, void *unused) static int i915_execlists(struct seq_file *m, void *data) { - struct drm_info_node *node = (struct drm_info_node *)m->private; - struct drm_device *dev = node->minor->dev; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = node_to_i915(m->private); + struct drm_device *dev = &dev_priv->drm; struct intel_engine_cs *engine; u32 status_pointer; u8 read_pointer; @@ -2107,7 +2060,7 @@ static int i915_execlists(struct seq_file *m, void *data) status_pointer = I915_READ(RING_CONTEXT_STATUS_PTR(engine)); seq_printf(m, "\tStatus pointer: 0x%08X\n", status_pointer); - read_pointer = engine->next_context_status_buffer; + read_pointer = GEN8_CSB_READ_PTR(status_pointer); write_pointer = GEN8_CSB_WRITE_PTR(status_pointer); if (read_pointer > write_pointer) write_pointer += GEN8_CSB_ENTRIES; @@ -2173,9 +2126,8 @@ static const char *swizzle_string(unsigned swizzle) static int i915_swizzle_info(struct seq_file *m, void *data) { - struct drm_info_node *node = m->private; - struct drm_device *dev = node->minor->dev; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = node_to_i915(m->private); + struct drm_device *dev = &dev_priv->drm; int ret; ret = mutex_lock_interruptible(&dev->struct_mutex); @@ -2188,7 +2140,7 @@ static int i915_swizzle_info(struct seq_file *m, void *data) seq_printf(m, "bit6 swizzle for Y-tiling = %s\n", swizzle_string(dev_priv->mm.bit_6_swizzle_y)); - if (IS_GEN3(dev) || IS_GEN4(dev)) { + if (IS_GEN3(dev_priv) || IS_GEN4(dev_priv)) { seq_printf(m, "DDC = 0x%08x\n", I915_READ(DCC)); seq_printf(m, "DDC2 = 0x%08x\n", @@ -2197,7 +2149,7 @@ static int i915_swizzle_info(struct seq_file *m, void *data) I915_READ16(C0DRB3)); seq_printf(m, "C1DRB3 = 0x%04x\n", I915_READ16(C1DRB3)); - } else if (INTEL_INFO(dev)->gen >= 6) { + } else if (INTEL_GEN(dev_priv) >= 6) { seq_printf(m, "MAD_DIMM_C0 = 0x%08x\n", I915_READ(MAD_DIMM_C0)); seq_printf(m, "MAD_DIMM_C1 = 0x%08x\n", @@ -2206,7 +2158,7 @@ static int i915_swizzle_info(struct seq_file *m, void *data) I915_READ(MAD_DIMM_C2)); seq_printf(m, "TILECTL = 0x%08x\n", I915_READ(TILECTL)); - if (INTEL_INFO(dev)->gen >= 8) + if (INTEL_GEN(dev_priv) >= 8) seq_printf(m, "GAMTARBMODE = 0x%08x\n", I915_READ(GAMTARBMODE)); else @@ -2246,9 +2198,9 @@ static int per_file_ctx(int id, void *ptr, void *data) return 0; } -static void gen8_ppgtt_info(struct seq_file *m, struct drm_device *dev) +static void gen8_ppgtt_info(struct seq_file *m, + struct drm_i915_private *dev_priv) { - struct drm_i915_private *dev_priv = to_i915(dev); struct intel_engine_cs *engine; struct i915_hw_ppgtt *ppgtt = dev_priv->mm.aliasing_ppgtt; int i; @@ -2267,9 +2219,9 @@ static void gen8_ppgtt_info(struct seq_file *m, struct drm_device *dev) } } -static void gen6_ppgtt_info(struct seq_file *m, struct drm_device *dev) +static void gen6_ppgtt_info(struct seq_file *m, + struct drm_i915_private *dev_priv) { - struct drm_i915_private *dev_priv = to_i915(dev); struct intel_engine_cs *engine; if (IS_GEN6(dev_priv)) @@ -2301,22 +2253,23 @@ static void gen6_ppgtt_info(struct seq_file *m, struct drm_device *dev) static int i915_ppgtt_info(struct seq_file *m, void *data) { - struct drm_info_node *node = m->private; - struct drm_device *dev = node->minor->dev; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = node_to_i915(m->private); + struct drm_device *dev = &dev_priv->drm; struct drm_file *file; + int ret; - int ret = mutex_lock_interruptible(&dev->struct_mutex); + mutex_lock(&dev->filelist_mutex); + ret = mutex_lock_interruptible(&dev->struct_mutex); if (ret) - return ret; + goto out_unlock; + intel_runtime_pm_get(dev_priv); - if (INTEL_INFO(dev)->gen >= 8) - gen8_ppgtt_info(m, dev); - else if (INTEL_INFO(dev)->gen >= 6) - gen6_ppgtt_info(m, dev); + if (INTEL_GEN(dev_priv) >= 8) + gen8_ppgtt_info(m, dev_priv); + else if (INTEL_GEN(dev_priv) >= 6) + gen6_ppgtt_info(m, dev_priv); - mutex_lock(&dev->filelist_mutex); list_for_each_entry_reverse(file, &dev->filelist, lhead) { struct drm_i915_file_private *file_priv = file->driver_priv; struct task_struct *task; @@ -2324,19 +2277,19 @@ static int i915_ppgtt_info(struct seq_file *m, void *data) task = get_pid_task(file->pid, PIDTYPE_PID); if (!task) { ret = -ESRCH; - goto out_unlock; + goto out_rpm; } seq_printf(m, "\nproc: %s\n", task->comm); put_task_struct(task); idr_for_each(&file_priv->context_idr, per_file_ctx, (void *)(unsigned long)m); } -out_unlock: - mutex_unlock(&dev->filelist_mutex); +out_rpm: intel_runtime_pm_put(dev_priv); mutex_unlock(&dev->struct_mutex); - +out_unlock: + mutex_unlock(&dev->filelist_mutex); return ret; } @@ -2367,9 +2320,8 @@ static const char *rps_power_to_str(unsigned int power) static int i915_rps_boost_info(struct seq_file *m, void *data) { - struct drm_info_node *node = m->private; - struct drm_device *dev = node->minor->dev; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = node_to_i915(m->private); + struct drm_device *dev = &dev_priv->drm; struct drm_file *file; seq_printf(m, "RPS enabled? %d\n", dev_priv->rps.enabled); @@ -2437,12 +2389,10 @@ static int i915_rps_boost_info(struct seq_file *m, void *data) static int i915_llc(struct seq_file *m, void *data) { - struct drm_info_node *node = m->private; - struct drm_device *dev = node->minor->dev; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = node_to_i915(m->private); const bool edram = INTEL_GEN(dev_priv) > 8; - seq_printf(m, "LLC: %s\n", yesno(HAS_LLC(dev))); + seq_printf(m, "LLC: %s\n", yesno(HAS_LLC(dev_priv))); seq_printf(m, "%s: %lluMB\n", edram ? "eDRAM" : "eLLC", intel_uncore_edram_size(dev_priv)/1024/1024); @@ -2451,8 +2401,7 @@ static int i915_llc(struct seq_file *m, void *data) static int i915_guc_load_status_info(struct seq_file *m, void *data) { - struct drm_info_node *node = m->private; - struct drm_i915_private *dev_priv = to_i915(node->minor->dev); + struct drm_i915_private *dev_priv = node_to_i915(m->private); struct intel_guc_fw *guc_fw = &dev_priv->guc.guc_fw; u32 tmp, i; @@ -2523,9 +2472,8 @@ static void i915_guc_client_info(struct seq_file *m, static int i915_guc_info(struct seq_file *m, void *data) { - struct drm_info_node *node = m->private; - struct drm_device *dev = node->minor->dev; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = node_to_i915(m->private); + struct drm_device *dev = &dev_priv->drm; struct intel_guc guc; struct i915_guc_client client = {}; struct intel_engine_cs *engine; @@ -2574,9 +2522,7 @@ static int i915_guc_info(struct seq_file *m, void *data) static int i915_guc_log_dump(struct seq_file *m, void *data) { - struct drm_info_node *node = m->private; - struct drm_device *dev = node->minor->dev; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = node_to_i915(m->private); struct drm_i915_gem_object *obj; int i = 0, pg; @@ -2602,15 +2548,13 @@ static int i915_guc_log_dump(struct seq_file *m, void *data) static int i915_edp_psr_status(struct seq_file *m, void *data) { - struct drm_info_node *node = m->private; - struct drm_device *dev = node->minor->dev; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = node_to_i915(m->private); u32 psrperf = 0; u32 stat[3]; enum pipe pipe; bool enabled = false; - if (!HAS_PSR(dev)) { + if (!HAS_PSR(dev_priv)) { seq_puts(m, "PSR not supported\n"); return 0; } @@ -2627,7 +2571,7 @@ static int i915_edp_psr_status(struct seq_file *m, void *data) seq_printf(m, "Re-enable work scheduled: %s\n", yesno(work_busy(&dev_priv->psr.work.work))); - if (HAS_DDI(dev)) + if (HAS_DDI(dev_priv)) enabled = I915_READ(EDP_PSR_CTL) & EDP_PSR_ENABLE; else { for_each_pipe(dev_priv, pipe) { @@ -2644,7 +2588,7 @@ static int i915_edp_psr_status(struct seq_file *m, void *data) seq_printf(m, "HW Enabled & Active bit: %s", yesno(enabled)); - if (!HAS_DDI(dev)) + if (!HAS_DDI(dev_priv)) for_each_pipe(dev_priv, pipe) { if ((stat[pipe] == VLV_EDP_PSR_ACTIVE_NORFB_UP) || (stat[pipe] == VLV_EDP_PSR_ACTIVE_SF_UPDATE)) @@ -2656,7 +2600,7 @@ static int i915_edp_psr_status(struct seq_file *m, void *data) * VLV/CHV PSR has no kind of performance counter * SKL+ Perf counter is reset to 0 everytime DC state is entered */ - if (IS_HASWELL(dev) || IS_BROADWELL(dev)) { + if (IS_HASWELL(dev_priv) || IS_BROADWELL(dev_priv)) { psrperf = I915_READ(EDP_PSR_PERF_CNT) & EDP_PSR_PERF_CNT_MASK; @@ -2670,8 +2614,8 @@ static int i915_edp_psr_status(struct seq_file *m, void *data) static int i915_sink_crc(struct seq_file *m, void *data) { - struct drm_info_node *node = m->private; - struct drm_device *dev = node->minor->dev; + struct drm_i915_private *dev_priv = node_to_i915(m->private); + struct drm_device *dev = &dev_priv->drm; struct intel_connector *connector; struct intel_dp *intel_dp = NULL; int ret; @@ -2710,13 +2654,11 @@ out: static int i915_energy_uJ(struct seq_file *m, void *data) { - struct drm_info_node *node = m->private; - struct drm_device *dev = node->minor->dev; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = node_to_i915(m->private); u64 power; u32 units; - if (INTEL_INFO(dev)->gen < 6) + if (INTEL_GEN(dev_priv) < 6) return -ENODEV; intel_runtime_pm_get(dev_priv); @@ -2736,9 +2678,8 @@ static int i915_energy_uJ(struct seq_file *m, void *data) static int i915_runtime_pm_status(struct seq_file *m, void *unused) { - struct drm_info_node *node = m->private; - struct drm_device *dev = node->minor->dev; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = node_to_i915(m->private); + struct pci_dev *pdev = dev_priv->drm.pdev; if (!HAS_RUNTIME_PM(dev_priv)) seq_puts(m, "Runtime power management not supported\n"); @@ -2748,22 +2689,20 @@ static int i915_runtime_pm_status(struct seq_file *m, void *unused) yesno(!intel_irqs_enabled(dev_priv))); #ifdef CONFIG_PM seq_printf(m, "Usage count: %d\n", - atomic_read(&dev->dev->power.usage_count)); + atomic_read(&dev_priv->drm.dev->power.usage_count)); #else seq_printf(m, "Device Power Management (CONFIG_PM) disabled\n"); #endif seq_printf(m, "PCI device power state: %s [%d]\n", - pci_power_name(dev_priv->drm.pdev->current_state), - dev_priv->drm.pdev->current_state); + pci_power_name(pdev->current_state), + pdev->current_state); return 0; } static int i915_power_domain_info(struct seq_file *m, void *unused) { - struct drm_info_node *node = m->private; - struct drm_device *dev = node->minor->dev; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = node_to_i915(m->private); struct i915_power_domains *power_domains = &dev_priv->power_domains; int i; @@ -2796,12 +2735,10 @@ static int i915_power_domain_info(struct seq_file *m, void *unused) static int i915_dmc_info(struct seq_file *m, void *unused) { - struct drm_info_node *node = m->private; - struct drm_device *dev = node->minor->dev; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = node_to_i915(m->private); struct intel_csr *csr; - if (!HAS_CSR(dev)) { + if (!HAS_CSR(dev_priv)) { seq_puts(m, "not supported\n"); return 0; } @@ -2819,12 +2756,12 @@ static int i915_dmc_info(struct seq_file *m, void *unused) seq_printf(m, "version: %d.%d\n", CSR_VERSION_MAJOR(csr->version), CSR_VERSION_MINOR(csr->version)); - if (IS_SKYLAKE(dev) && csr->version >= CSR_VERSION(1, 6)) { + if (IS_SKYLAKE(dev_priv) && csr->version >= CSR_VERSION(1, 6)) { seq_printf(m, "DC3 -> DC5 count: %d\n", I915_READ(SKL_CSR_DC3_DC5_COUNT)); seq_printf(m, "DC5 -> DC6 count: %d\n", I915_READ(SKL_CSR_DC5_DC6_COUNT)); - } else if (IS_BROXTON(dev) && csr->version >= CSR_VERSION(1, 4)) { + } else if (IS_BROXTON(dev_priv) && csr->version >= CSR_VERSION(1, 4)) { seq_printf(m, "DC3 -> DC5 count: %d\n", I915_READ(BXT_CSR_DC3_DC5_COUNT)); } @@ -2861,8 +2798,8 @@ static void intel_encoder_info(struct seq_file *m, struct intel_crtc *intel_crtc, struct intel_encoder *intel_encoder) { - struct drm_info_node *node = m->private; - struct drm_device *dev = node->minor->dev; + struct drm_i915_private *dev_priv = node_to_i915(m->private); + struct drm_device *dev = &dev_priv->drm; struct drm_crtc *crtc = &intel_crtc->base; struct intel_connector *intel_connector; struct drm_encoder *encoder; @@ -2888,8 +2825,8 @@ static void intel_encoder_info(struct seq_file *m, static void intel_crtc_info(struct seq_file *m, struct intel_crtc *intel_crtc) { - struct drm_info_node *node = m->private; - struct drm_device *dev = node->minor->dev; + struct drm_i915_private *dev_priv = node_to_i915(m->private); + struct drm_device *dev = &dev_priv->drm; struct drm_crtc *crtc = &intel_crtc->base; struct intel_encoder *intel_encoder; struct drm_plane_state *plane_state = crtc->primary->state; @@ -2923,6 +2860,9 @@ static void intel_dp_info(struct seq_file *m, seq_printf(m, "\taudio support: %s\n", yesno(intel_dp->has_audio)); if (intel_connector->base.connector_type == DRM_MODE_CONNECTOR_eDP) intel_panel_info(m, &intel_connector->panel); + + drm_dp_downstream_debug(m, intel_dp->dpcd, intel_dp->downstream_ports, + &intel_dp->aux); } static void intel_hdmi_info(struct seq_file *m, @@ -2987,12 +2927,11 @@ static void intel_connector_info(struct seq_file *m, intel_seq_print_mode(m, 2, mode); } -static bool cursor_active(struct drm_device *dev, int pipe) +static bool cursor_active(struct drm_i915_private *dev_priv, int pipe) { - struct drm_i915_private *dev_priv = to_i915(dev); u32 state; - if (IS_845G(dev) || IS_I865G(dev)) + if (IS_845G(dev_priv) || IS_I865G(dev_priv)) state = I915_READ(CURCNTR(PIPE_A)) & CURSOR_ENABLE; else state = I915_READ(CURCNTR(pipe)) & CURSOR_MODE; @@ -3000,9 +2939,9 @@ static bool cursor_active(struct drm_device *dev, int pipe) return state; } -static bool cursor_position(struct drm_device *dev, int pipe, int *x, int *y) +static bool cursor_position(struct drm_i915_private *dev_priv, + int pipe, int *x, int *y) { - struct drm_i915_private *dev_priv = to_i915(dev); u32 pos; pos = I915_READ(CURPOS(pipe)); @@ -3015,7 +2954,7 @@ static bool cursor_position(struct drm_device *dev, int pipe, int *x, int *y) if (pos & (CURSOR_POS_SIGN << CURSOR_Y_SHIFT)) *y = -*y; - return cursor_active(dev, pipe); + return cursor_active(dev_priv, pipe); } static const char *plane_type(enum drm_plane_type type) @@ -3058,8 +2997,8 @@ static const char *plane_rotation(unsigned int rotation) static void intel_plane_info(struct seq_file *m, struct intel_crtc *intel_crtc) { - struct drm_info_node *node = m->private; - struct drm_device *dev = node->minor->dev; + struct drm_i915_private *dev_priv = node_to_i915(m->private); + struct drm_device *dev = &dev_priv->drm; struct intel_plane *intel_plane; for_each_intel_plane_on_crtc(dev, intel_crtc, intel_plane) { @@ -3130,9 +3069,8 @@ static void intel_scaler_info(struct seq_file *m, struct intel_crtc *intel_crtc) static int i915_display_info(struct seq_file *m, void *unused) { - struct drm_info_node *node = m->private; - struct drm_device *dev = node->minor->dev; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = node_to_i915(m->private); + struct drm_device *dev = &dev_priv->drm; struct intel_crtc *crtc; struct drm_connector *connector; @@ -3156,7 +3094,7 @@ static int i915_display_info(struct seq_file *m, void *unused) if (pipe_config->base.active) { intel_crtc_info(m, crtc); - active = cursor_position(dev, crtc->pipe, &x, &y); + active = cursor_position(dev_priv, crtc->pipe, &x, &y); seq_printf(m, "\tcursor visible? %s, position (%d, %d), size %dx%d, addr 0x%08x, active? %s\n", yesno(crtc->cursor_base), x, y, crtc->base.cursor->state->crtc_w, @@ -3185,11 +3123,10 @@ static int i915_display_info(struct seq_file *m, void *unused) static int i915_semaphore_status(struct seq_file *m, void *unused) { - struct drm_info_node *node = (struct drm_info_node *) m->private; - struct drm_device *dev = node->minor->dev; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = node_to_i915(m->private); + struct drm_device *dev = &dev_priv->drm; struct intel_engine_cs *engine; - int num_rings = INTEL_INFO(dev)->num_rings; + int num_rings = INTEL_INFO(dev_priv)->num_rings; enum intel_engine_id id; int j, ret; @@ -3203,7 +3140,7 @@ static int i915_semaphore_status(struct seq_file *m, void *unused) return ret; intel_runtime_pm_get(dev_priv); - if (IS_BROADWELL(dev)) { + if (IS_BROADWELL(dev_priv)) { struct page *page; uint64_t *seqno; @@ -3258,9 +3195,8 @@ static int i915_semaphore_status(struct seq_file *m, void *unused) static int i915_shared_dplls_info(struct seq_file *m, void *unused) { - struct drm_info_node *node = (struct drm_info_node *) m->private; - struct drm_device *dev = node->minor->dev; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = node_to_i915(m->private); + struct drm_device *dev = &dev_priv->drm; int i; drm_modeset_lock_all(dev); @@ -3288,9 +3224,8 @@ static int i915_wa_registers(struct seq_file *m, void *unused) int i; int ret; struct intel_engine_cs *engine; - struct drm_info_node *node = (struct drm_info_node *) m->private; - struct drm_device *dev = node->minor->dev; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = node_to_i915(m->private); + struct drm_device *dev = &dev_priv->drm; struct i915_workarounds *workarounds = &dev_priv->workarounds; enum intel_engine_id id; @@ -3326,15 +3261,14 @@ static int i915_wa_registers(struct seq_file *m, void *unused) static int i915_ddb_info(struct seq_file *m, void *unused) { - struct drm_info_node *node = m->private; - struct drm_device *dev = node->minor->dev; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = node_to_i915(m->private); + struct drm_device *dev = &dev_priv->drm; struct skl_ddb_allocation *ddb; struct skl_ddb_entry *entry; enum pipe pipe; int plane; - if (INTEL_INFO(dev)->gen < 9) + if (INTEL_GEN(dev_priv) < 9) return 0; drm_modeset_lock_all(dev); @@ -3364,7 +3298,8 @@ static int i915_ddb_info(struct seq_file *m, void *unused) } static void drrs_status_per_crtc(struct seq_file *m, - struct drm_device *dev, struct intel_crtc *intel_crtc) + struct drm_device *dev, + struct intel_crtc *intel_crtc) { struct drm_i915_private *dev_priv = to_i915(dev); struct i915_drrs *drrs = &dev_priv->drrs; @@ -3433,8 +3368,8 @@ static void drrs_status_per_crtc(struct seq_file *m, static int i915_drrs_status(struct seq_file *m, void *unused) { - struct drm_info_node *node = m->private; - struct drm_device *dev = node->minor->dev; + struct drm_i915_private *dev_priv = node_to_i915(m->private); + struct drm_device *dev = &dev_priv->drm; struct intel_crtc *intel_crtc; int active_crtc_cnt = 0; @@ -3457,14 +3392,14 @@ static int i915_drrs_status(struct seq_file *m, void *unused) struct pipe_crc_info { const char *name; - struct drm_device *dev; + struct drm_i915_private *dev_priv; enum pipe pipe; }; static int i915_dp_mst_info(struct seq_file *m, void *unused) { - struct drm_info_node *node = (struct drm_info_node *) m->private; - struct drm_device *dev = node->minor->dev; + struct drm_i915_private *dev_priv = node_to_i915(m->private); + struct drm_device *dev = &dev_priv->drm; struct intel_encoder *intel_encoder; struct intel_digital_port *intel_dig_port; struct drm_connector *connector; @@ -3493,10 +3428,10 @@ static int i915_dp_mst_info(struct seq_file *m, void *unused) static int i915_pipe_crc_open(struct inode *inode, struct file *filep) { struct pipe_crc_info *info = inode->i_private; - struct drm_i915_private *dev_priv = to_i915(info->dev); + struct drm_i915_private *dev_priv = info->dev_priv; struct intel_pipe_crc *pipe_crc = &dev_priv->pipe_crc[info->pipe]; - if (info->pipe >= INTEL_INFO(info->dev)->num_pipes) + if (info->pipe >= INTEL_INFO(dev_priv)->num_pipes) return -ENODEV; spin_lock_irq(&pipe_crc->lock); @@ -3517,7 +3452,7 @@ static int i915_pipe_crc_open(struct inode *inode, struct file *filep) static int i915_pipe_crc_release(struct inode *inode, struct file *filep) { struct pipe_crc_info *info = inode->i_private; - struct drm_i915_private *dev_priv = to_i915(info->dev); + struct drm_i915_private *dev_priv = info->dev_priv; struct intel_pipe_crc *pipe_crc = &dev_priv->pipe_crc[info->pipe]; spin_lock_irq(&pipe_crc->lock); @@ -3544,8 +3479,7 @@ i915_pipe_crc_read(struct file *filep, char __user *user_buf, size_t count, loff_t *pos) { struct pipe_crc_info *info = filep->private_data; - struct drm_device *dev = info->dev; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = info->dev_priv; struct intel_pipe_crc *pipe_crc = &dev_priv->pipe_crc[info->pipe]; char buf[PIPE_CRC_BUFFER_LEN]; int n_entries; @@ -3641,11 +3575,11 @@ static struct pipe_crc_info i915_pipe_crc_data[I915_MAX_PIPES] = { static int i915_pipe_crc_create(struct dentry *root, struct drm_minor *minor, enum pipe pipe) { - struct drm_device *dev = minor->dev; + struct drm_i915_private *dev_priv = to_i915(minor->dev); struct dentry *ent; struct pipe_crc_info *info = &i915_pipe_crc_data[pipe]; - info->dev = dev; + info->dev_priv = dev_priv; ent = debugfs_create_file(info->name, S_IRUGO, root, info, &i915_pipe_crc_fops); if (!ent) @@ -3675,8 +3609,7 @@ static const char *pipe_crc_source_name(enum intel_pipe_crc_source source) static int display_crc_ctl_show(struct seq_file *m, void *data) { - struct drm_device *dev = m->private; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = m->private; int i; for (i = 0; i < I915_MAX_PIPES; i++) @@ -3688,9 +3621,7 @@ static int display_crc_ctl_show(struct seq_file *m, void *data) static int display_crc_ctl_open(struct inode *inode, struct file *file) { - struct drm_device *dev = inode->i_private; - - return single_open(file, display_crc_ctl_show, dev); + return single_open(file, display_crc_ctl_show, inode->i_private); } static int i8xx_pipe_crc_ctl_reg(enum intel_pipe_crc_source *source, @@ -3713,9 +3644,11 @@ static int i8xx_pipe_crc_ctl_reg(enum intel_pipe_crc_source *source, return 0; } -static int i9xx_pipe_crc_auto_source(struct drm_device *dev, enum pipe pipe, +static int i9xx_pipe_crc_auto_source(struct drm_i915_private *dev_priv, + enum pipe pipe, enum intel_pipe_crc_source *source) { + struct drm_device *dev = &dev_priv->drm; struct intel_encoder *encoder; struct intel_crtc *crtc; struct intel_digital_port *dig_port; @@ -3765,16 +3698,15 @@ static int i9xx_pipe_crc_auto_source(struct drm_device *dev, enum pipe pipe, return ret; } -static int vlv_pipe_crc_ctl_reg(struct drm_device *dev, +static int vlv_pipe_crc_ctl_reg(struct drm_i915_private *dev_priv, enum pipe pipe, enum intel_pipe_crc_source *source, uint32_t *val) { - struct drm_i915_private *dev_priv = to_i915(dev); bool need_stable_symbols = false; if (*source == INTEL_PIPE_CRC_SOURCE_AUTO) { - int ret = i9xx_pipe_crc_auto_source(dev, pipe, source); + int ret = i9xx_pipe_crc_auto_source(dev_priv, pipe, source); if (ret) return ret; } @@ -3792,7 +3724,7 @@ static int vlv_pipe_crc_ctl_reg(struct drm_device *dev, need_stable_symbols = true; break; case INTEL_PIPE_CRC_SOURCE_DP_D: - if (!IS_CHERRYVIEW(dev)) + if (!IS_CHERRYVIEW(dev_priv)) return -EINVAL; *val = PIPE_CRC_ENABLE | PIPE_CRC_SOURCE_DP_D_VLV; need_stable_symbols = true; @@ -3836,16 +3768,15 @@ static int vlv_pipe_crc_ctl_reg(struct drm_device *dev, return 0; } -static int i9xx_pipe_crc_ctl_reg(struct drm_device *dev, +static int i9xx_pipe_crc_ctl_reg(struct drm_i915_private *dev_priv, enum pipe pipe, enum intel_pipe_crc_source *source, uint32_t *val) { - struct drm_i915_private *dev_priv = to_i915(dev); bool need_stable_symbols = false; if (*source == INTEL_PIPE_CRC_SOURCE_AUTO) { - int ret = i9xx_pipe_crc_auto_source(dev, pipe, source); + int ret = i9xx_pipe_crc_auto_source(dev_priv, pipe, source); if (ret) return ret; } @@ -3855,24 +3786,24 @@ static int i9xx_pipe_crc_ctl_reg(struct drm_device *dev, *val = PIPE_CRC_ENABLE | PIPE_CRC_SOURCE_PIPE_I9XX; break; case INTEL_PIPE_CRC_SOURCE_TV: - if (!SUPPORTS_TV(dev)) + if (!SUPPORTS_TV(dev_priv)) return -EINVAL; *val = PIPE_CRC_ENABLE | PIPE_CRC_SOURCE_TV_PRE; break; case INTEL_PIPE_CRC_SOURCE_DP_B: - if (!IS_G4X(dev)) + if (!IS_G4X(dev_priv)) return -EINVAL; *val = PIPE_CRC_ENABLE | PIPE_CRC_SOURCE_DP_B_G4X; need_stable_symbols = true; break; case INTEL_PIPE_CRC_SOURCE_DP_C: - if (!IS_G4X(dev)) + if (!IS_G4X(dev_priv)) return -EINVAL; *val = PIPE_CRC_ENABLE | PIPE_CRC_SOURCE_DP_C_G4X; need_stable_symbols = true; break; case INTEL_PIPE_CRC_SOURCE_DP_D: - if (!IS_G4X(dev)) + if (!IS_G4X(dev_priv)) return -EINVAL; *val = PIPE_CRC_ENABLE | PIPE_CRC_SOURCE_DP_D_G4X; need_stable_symbols = true; @@ -3896,7 +3827,7 @@ static int i9xx_pipe_crc_ctl_reg(struct drm_device *dev, if (need_stable_symbols) { uint32_t tmp = I915_READ(PORT_DFT2_G4X); - WARN_ON(!IS_G4X(dev)); + WARN_ON(!IS_G4X(dev_priv)); I915_WRITE(PORT_DFT_I9XX, I915_READ(PORT_DFT_I9XX) | DC_BALANCE_RESET); @@ -3912,10 +3843,9 @@ static int i9xx_pipe_crc_ctl_reg(struct drm_device *dev, return 0; } -static void vlv_undo_pipe_scramble_reset(struct drm_device *dev, +static void vlv_undo_pipe_scramble_reset(struct drm_i915_private *dev_priv, enum pipe pipe) { - struct drm_i915_private *dev_priv = to_i915(dev); uint32_t tmp = I915_READ(PORT_DFT2_G4X); switch (pipe) { @@ -3937,10 +3867,9 @@ static void vlv_undo_pipe_scramble_reset(struct drm_device *dev, } -static void g4x_undo_pipe_scramble_reset(struct drm_device *dev, +static void g4x_undo_pipe_scramble_reset(struct drm_i915_private *dev_priv, enum pipe pipe) { - struct drm_i915_private *dev_priv = to_i915(dev); uint32_t tmp = I915_READ(PORT_DFT2_G4X); if (pipe == PIPE_A) @@ -3981,9 +3910,10 @@ static int ilk_pipe_crc_ctl_reg(enum intel_pipe_crc_source *source, return 0; } -static void hsw_trans_edp_pipe_A_crc_wa(struct drm_device *dev, bool enable) +static void hsw_trans_edp_pipe_A_crc_wa(struct drm_i915_private *dev_priv, + bool enable) { - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_device *dev = &dev_priv->drm; struct intel_crtc *crtc = to_intel_crtc(dev_priv->pipe_to_crtc_mapping[PIPE_A]); struct intel_crtc_state *pipe_config; @@ -4017,7 +3947,7 @@ out: drm_atomic_state_free(state); } -static int ivb_pipe_crc_ctl_reg(struct drm_device *dev, +static int ivb_pipe_crc_ctl_reg(struct drm_i915_private *dev_priv, enum pipe pipe, enum intel_pipe_crc_source *source, uint32_t *val) @@ -4033,8 +3963,8 @@ static int ivb_pipe_crc_ctl_reg(struct drm_device *dev, *val = PIPE_CRC_ENABLE | PIPE_CRC_SOURCE_SPRITE_IVB; break; case INTEL_PIPE_CRC_SOURCE_PF: - if (IS_HASWELL(dev) && pipe == PIPE_A) - hsw_trans_edp_pipe_A_crc_wa(dev, true); + if (IS_HASWELL(dev_priv) && pipe == PIPE_A) + hsw_trans_edp_pipe_A_crc_wa(dev_priv, true); *val = PIPE_CRC_ENABLE | PIPE_CRC_SOURCE_PF_IVB; break; @@ -4048,13 +3978,14 @@ static int ivb_pipe_crc_ctl_reg(struct drm_device *dev, return 0; } -static int pipe_crc_set_source(struct drm_device *dev, enum pipe pipe, +static int pipe_crc_set_source(struct drm_i915_private *dev_priv, + enum pipe pipe, enum intel_pipe_crc_source source) { - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_device *dev = &dev_priv->drm; struct intel_pipe_crc *pipe_crc = &dev_priv->pipe_crc[pipe]; - struct intel_crtc *crtc = to_intel_crtc(intel_get_crtc_for_pipe(dev, - pipe)); + struct intel_crtc *crtc = + to_intel_crtc(intel_get_crtc_for_pipe(dev, pipe)); enum intel_display_power_domain power_domain; u32 val = 0; /* shut up gcc */ int ret; @@ -4072,16 +4003,16 @@ static int pipe_crc_set_source(struct drm_device *dev, enum pipe pipe, return -EIO; } - if (IS_GEN2(dev)) + if (IS_GEN2(dev_priv)) ret = i8xx_pipe_crc_ctl_reg(&source, &val); - else if (INTEL_INFO(dev)->gen < 5) - ret = i9xx_pipe_crc_ctl_reg(dev, pipe, &source, &val); - else if (IS_VALLEYVIEW(dev) || IS_CHERRYVIEW(dev)) - ret = vlv_pipe_crc_ctl_reg(dev, pipe, &source, &val); - else if (IS_GEN5(dev) || IS_GEN6(dev)) + else if (INTEL_GEN(dev_priv) < 5) + ret = i9xx_pipe_crc_ctl_reg(dev_priv, pipe, &source, &val); + else if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv)) + ret = vlv_pipe_crc_ctl_reg(dev_priv, pipe, &source, &val); + else if (IS_GEN5(dev_priv) || IS_GEN6(dev_priv)) ret = ilk_pipe_crc_ctl_reg(&source, &val); else - ret = ivb_pipe_crc_ctl_reg(dev, pipe, &source, &val); + ret = ivb_pipe_crc_ctl_reg(dev_priv, pipe, &source, &val); if (ret != 0) goto out; @@ -4145,12 +4076,12 @@ static int pipe_crc_set_source(struct drm_device *dev, enum pipe pipe, kfree(entries); - if (IS_G4X(dev)) - g4x_undo_pipe_scramble_reset(dev, pipe); - else if (IS_VALLEYVIEW(dev) || IS_CHERRYVIEW(dev)) - vlv_undo_pipe_scramble_reset(dev, pipe); - else if (IS_HASWELL(dev) && pipe == PIPE_A) - hsw_trans_edp_pipe_A_crc_wa(dev, false); + if (IS_G4X(dev_priv)) + g4x_undo_pipe_scramble_reset(dev_priv, pipe); + else if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv)) + vlv_undo_pipe_scramble_reset(dev_priv, pipe); + else if (IS_HASWELL(dev_priv) && pipe == PIPE_A) + hsw_trans_edp_pipe_A_crc_wa(dev_priv, false); hsw_enable_ips(crtc); } @@ -4254,7 +4185,8 @@ display_crc_ctl_parse_source(const char *buf, enum intel_pipe_crc_source *s) return -EINVAL; } -static int display_crc_ctl_parse(struct drm_device *dev, char *buf, size_t len) +static int display_crc_ctl_parse(struct drm_i915_private *dev_priv, + char *buf, size_t len) { #define N_WORDS 3 int n_words; @@ -4285,14 +4217,14 @@ static int display_crc_ctl_parse(struct drm_device *dev, char *buf, size_t len) return -EINVAL; } - return pipe_crc_set_source(dev, pipe, source); + return pipe_crc_set_source(dev_priv, pipe, source); } static ssize_t display_crc_ctl_write(struct file *file, const char __user *ubuf, size_t len, loff_t *offp) { struct seq_file *m = file->private_data; - struct drm_device *dev = m->private; + struct drm_i915_private *dev_priv = m->private; char *tmpbuf; int ret; @@ -4315,7 +4247,7 @@ static ssize_t display_crc_ctl_write(struct file *file, const char __user *ubuf, } tmpbuf[len] = '\0'; - ret = display_crc_ctl_parse(dev, tmpbuf, len); + ret = display_crc_ctl_parse(dev_priv, tmpbuf, len); out: kfree(tmpbuf); @@ -4336,8 +4268,8 @@ static const struct file_operations i915_display_crc_ctl_fops = { }; static ssize_t i915_displayport_test_active_write(struct file *file, - const char __user *ubuf, - size_t len, loff_t *offp) + const char __user *ubuf, + size_t len, loff_t *offp) { char *input_buffer; int status = 0; @@ -4367,7 +4299,6 @@ static ssize_t i915_displayport_test_active_write(struct file *file, DRM_DEBUG_DRIVER("Copied %d bytes from user\n", (unsigned int)len); list_for_each_entry(connector, connector_list, head) { - if (connector->connector_type != DRM_MODE_CONNECTOR_DisplayPort) continue; @@ -4405,7 +4336,6 @@ static int i915_displayport_test_active_show(struct seq_file *m, void *data) struct intel_dp *intel_dp; list_for_each_entry(connector, connector_list, head) { - if (connector->connector_type != DRM_MODE_CONNECTOR_DisplayPort) continue; @@ -4425,11 +4355,12 @@ static int i915_displayport_test_active_show(struct seq_file *m, void *data) } static int i915_displayport_test_active_open(struct inode *inode, - struct file *file) + struct file *file) { - struct drm_device *dev = inode->i_private; + struct drm_i915_private *dev_priv = inode->i_private; - return single_open(file, i915_displayport_test_active_show, dev); + return single_open(file, i915_displayport_test_active_show, + &dev_priv->drm); } static const struct file_operations i915_displayport_test_active_fops = { @@ -4449,7 +4380,6 @@ static int i915_displayport_test_data_show(struct seq_file *m, void *data) struct intel_dp *intel_dp; list_for_each_entry(connector, connector_list, head) { - if (connector->connector_type != DRM_MODE_CONNECTOR_DisplayPort) continue; @@ -4465,11 +4395,12 @@ static int i915_displayport_test_data_show(struct seq_file *m, void *data) return 0; } static int i915_displayport_test_data_open(struct inode *inode, - struct file *file) + struct file *file) { - struct drm_device *dev = inode->i_private; + struct drm_i915_private *dev_priv = inode->i_private; - return single_open(file, i915_displayport_test_data_show, dev); + return single_open(file, i915_displayport_test_data_show, + &dev_priv->drm); } static const struct file_operations i915_displayport_test_data_fops = { @@ -4488,7 +4419,6 @@ static int i915_displayport_test_type_show(struct seq_file *m, void *data) struct intel_dp *intel_dp; list_for_each_entry(connector, connector_list, head) { - if (connector->connector_type != DRM_MODE_CONNECTOR_DisplayPort) continue; @@ -4507,9 +4437,10 @@ static int i915_displayport_test_type_show(struct seq_file *m, void *data) static int i915_displayport_test_type_open(struct inode *inode, struct file *file) { - struct drm_device *dev = inode->i_private; + struct drm_i915_private *dev_priv = inode->i_private; - return single_open(file, i915_displayport_test_type_show, dev); + return single_open(file, i915_displayport_test_type_show, + &dev_priv->drm); } static const struct file_operations i915_displayport_test_type_fops = { @@ -4522,13 +4453,14 @@ static const struct file_operations i915_displayport_test_type_fops = { static void wm_latency_show(struct seq_file *m, const uint16_t wm[8]) { - struct drm_device *dev = m->private; + struct drm_i915_private *dev_priv = m->private; + struct drm_device *dev = &dev_priv->drm; int level; int num_levels; - if (IS_CHERRYVIEW(dev)) + if (IS_CHERRYVIEW(dev_priv)) num_levels = 3; - else if (IS_VALLEYVIEW(dev)) + else if (IS_VALLEYVIEW(dev_priv)) num_levels = 1; else num_levels = ilk_wm_max_level(dev) + 1; @@ -4542,8 +4474,8 @@ static void wm_latency_show(struct seq_file *m, const uint16_t wm[8]) * - WM1+ latency values in 0.5us units * - latencies are in us on gen9/vlv/chv */ - if (INTEL_INFO(dev)->gen >= 9 || IS_VALLEYVIEW(dev) || - IS_CHERRYVIEW(dev)) + if (INTEL_GEN(dev_priv) >= 9 || IS_VALLEYVIEW(dev_priv) || + IS_CHERRYVIEW(dev_priv)) latency *= 10; else if (level > 0) latency *= 5; @@ -4557,14 +4489,13 @@ static void wm_latency_show(struct seq_file *m, const uint16_t wm[8]) static int pri_wm_latency_show(struct seq_file *m, void *data) { - struct drm_device *dev = m->private; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = m->private; const uint16_t *latencies; - if (INTEL_INFO(dev)->gen >= 9) + if (INTEL_GEN(dev_priv) >= 9) latencies = dev_priv->wm.skl_latency; else - latencies = to_i915(dev)->wm.pri_latency; + latencies = dev_priv->wm.pri_latency; wm_latency_show(m, latencies); @@ -4573,14 +4504,13 @@ static int pri_wm_latency_show(struct seq_file *m, void *data) static int spr_wm_latency_show(struct seq_file *m, void *data) { - struct drm_device *dev = m->private; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = m->private; const uint16_t *latencies; - if (INTEL_INFO(dev)->gen >= 9) + if (INTEL_GEN(dev_priv) >= 9) latencies = dev_priv->wm.skl_latency; else - latencies = to_i915(dev)->wm.spr_latency; + latencies = dev_priv->wm.spr_latency; wm_latency_show(m, latencies); @@ -4589,14 +4519,13 @@ static int spr_wm_latency_show(struct seq_file *m, void *data) static int cur_wm_latency_show(struct seq_file *m, void *data) { - struct drm_device *dev = m->private; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = m->private; const uint16_t *latencies; - if (INTEL_INFO(dev)->gen >= 9) + if (INTEL_GEN(dev_priv) >= 9) latencies = dev_priv->wm.skl_latency; else - latencies = to_i915(dev)->wm.cur_latency; + latencies = dev_priv->wm.cur_latency; wm_latency_show(m, latencies); @@ -4605,48 +4534,49 @@ static int cur_wm_latency_show(struct seq_file *m, void *data) static int pri_wm_latency_open(struct inode *inode, struct file *file) { - struct drm_device *dev = inode->i_private; + struct drm_i915_private *dev_priv = inode->i_private; - if (INTEL_INFO(dev)->gen < 5) + if (INTEL_GEN(dev_priv) < 5) return -ENODEV; - return single_open(file, pri_wm_latency_show, dev); + return single_open(file, pri_wm_latency_show, dev_priv); } static int spr_wm_latency_open(struct inode *inode, struct file *file) { - struct drm_device *dev = inode->i_private; + struct drm_i915_private *dev_priv = inode->i_private; - if (HAS_GMCH_DISPLAY(dev)) + if (HAS_GMCH_DISPLAY(dev_priv)) return -ENODEV; - return single_open(file, spr_wm_latency_show, dev); + return single_open(file, spr_wm_latency_show, dev_priv); } static int cur_wm_latency_open(struct inode *inode, struct file *file) { - struct drm_device *dev = inode->i_private; + struct drm_i915_private *dev_priv = inode->i_private; - if (HAS_GMCH_DISPLAY(dev)) + if (HAS_GMCH_DISPLAY(dev_priv)) return -ENODEV; - return single_open(file, cur_wm_latency_show, dev); + return single_open(file, cur_wm_latency_show, dev_priv); } static ssize_t wm_latency_write(struct file *file, const char __user *ubuf, size_t len, loff_t *offp, uint16_t wm[8]) { struct seq_file *m = file->private_data; - struct drm_device *dev = m->private; + struct drm_i915_private *dev_priv = m->private; + struct drm_device *dev = &dev_priv->drm; uint16_t new[8] = { 0 }; int num_levels; int level; int ret; char tmp[32]; - if (IS_CHERRYVIEW(dev)) + if (IS_CHERRYVIEW(dev_priv)) num_levels = 3; - else if (IS_VALLEYVIEW(dev)) + else if (IS_VALLEYVIEW(dev_priv)) num_levels = 1; else num_levels = ilk_wm_max_level(dev) + 1; @@ -4680,14 +4610,13 @@ static ssize_t pri_wm_latency_write(struct file *file, const char __user *ubuf, size_t len, loff_t *offp) { struct seq_file *m = file->private_data; - struct drm_device *dev = m->private; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = m->private; uint16_t *latencies; - if (INTEL_INFO(dev)->gen >= 9) + if (INTEL_GEN(dev_priv) >= 9) latencies = dev_priv->wm.skl_latency; else - latencies = to_i915(dev)->wm.pri_latency; + latencies = dev_priv->wm.pri_latency; return wm_latency_write(file, ubuf, len, offp, latencies); } @@ -4696,14 +4625,13 @@ static ssize_t spr_wm_latency_write(struct file *file, const char __user *ubuf, size_t len, loff_t *offp) { struct seq_file *m = file->private_data; - struct drm_device *dev = m->private; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = m->private; uint16_t *latencies; - if (INTEL_INFO(dev)->gen >= 9) + if (INTEL_GEN(dev_priv) >= 9) latencies = dev_priv->wm.skl_latency; else - latencies = to_i915(dev)->wm.spr_latency; + latencies = dev_priv->wm.spr_latency; return wm_latency_write(file, ubuf, len, offp, latencies); } @@ -4712,14 +4640,13 @@ static ssize_t cur_wm_latency_write(struct file *file, const char __user *ubuf, size_t len, loff_t *offp) { struct seq_file *m = file->private_data; - struct drm_device *dev = m->private; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = m->private; uint16_t *latencies; - if (INTEL_INFO(dev)->gen >= 9) + if (INTEL_GEN(dev_priv) >= 9) latencies = dev_priv->wm.skl_latency; else - latencies = to_i915(dev)->wm.cur_latency; + latencies = dev_priv->wm.cur_latency; return wm_latency_write(file, ubuf, len, offp, latencies); } @@ -4754,8 +4681,7 @@ static const struct file_operations i915_cur_wm_latency_fops = { static int i915_wedged_get(void *data, u64 *val) { - struct drm_device *dev = data; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = data; *val = i915_terminally_wedged(&dev_priv->gpu_error); @@ -4765,8 +4691,7 @@ i915_wedged_get(void *data, u64 *val) static int i915_wedged_set(void *data, u64 val) { - struct drm_device *dev = data; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = data; /* * There is no safeguard against this debugfs entry colliding @@ -4796,8 +4721,7 @@ DEFINE_SIMPLE_ATTRIBUTE(i915_wedged_fops, static int i915_ring_missed_irq_get(void *data, u64 *val) { - struct drm_device *dev = data; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = data; *val = dev_priv->gpu_error.missed_irq_rings; return 0; @@ -4806,8 +4730,8 @@ i915_ring_missed_irq_get(void *data, u64 *val) static int i915_ring_missed_irq_set(void *data, u64 val) { - struct drm_device *dev = data; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = data; + struct drm_device *dev = &dev_priv->drm; int ret; /* Lock against concurrent debugfs callers */ @@ -4827,8 +4751,7 @@ DEFINE_SIMPLE_ATTRIBUTE(i915_ring_missed_irq_fops, static int i915_ring_test_irq_get(void *data, u64 *val) { - struct drm_device *dev = data; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = data; *val = dev_priv->gpu_error.test_irq_rings; @@ -4838,8 +4761,7 @@ i915_ring_test_irq_get(void *data, u64 *val) static int i915_ring_test_irq_set(void *data, u64 val) { - struct drm_device *dev = data; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = data; val &= INTEL_INFO(dev_priv)->ring_mask; DRM_DEBUG_DRIVER("Masking interrupts on rings 0x%08llx\n", val); @@ -4871,8 +4793,8 @@ i915_drop_caches_get(void *data, u64 *val) static int i915_drop_caches_set(void *data, u64 val) { - struct drm_device *dev = data; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = data; + struct drm_device *dev = &dev_priv->drm; int ret; DRM_DEBUG("Dropping caches: 0x%08llx\n", val); @@ -4884,7 +4806,9 @@ i915_drop_caches_set(void *data, u64 val) return ret; if (val & DROP_ACTIVE) { - ret = i915_gem_wait_for_idle(dev_priv, true); + ret = i915_gem_wait_for_idle(dev_priv, + I915_WAIT_INTERRUPTIBLE | + I915_WAIT_LOCKED); if (ret) goto unlock; } @@ -4911,10 +4835,9 @@ DEFINE_SIMPLE_ATTRIBUTE(i915_drop_caches_fops, static int i915_max_freq_get(void *data, u64 *val) { - struct drm_device *dev = data; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = data; - if (INTEL_INFO(dev)->gen < 6) + if (INTEL_GEN(dev_priv) < 6) return -ENODEV; *val = intel_gpu_freq(dev_priv, dev_priv->rps.max_freq_softlimit); @@ -4924,12 +4847,11 @@ i915_max_freq_get(void *data, u64 *val) static int i915_max_freq_set(void *data, u64 val) { - struct drm_device *dev = data; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = data; u32 hw_max, hw_min; int ret; - if (INTEL_INFO(dev)->gen < 6) + if (INTEL_GEN(dev_priv) < 6) return -ENODEV; DRM_DEBUG_DRIVER("Manually setting max freq to %llu\n", val); @@ -4967,8 +4889,7 @@ DEFINE_SIMPLE_ATTRIBUTE(i915_max_freq_fops, static int i915_min_freq_get(void *data, u64 *val) { - struct drm_device *dev = data; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = data; if (INTEL_GEN(dev_priv) < 6) return -ENODEV; @@ -4980,8 +4901,7 @@ i915_min_freq_get(void *data, u64 *val) static int i915_min_freq_set(void *data, u64 val) { - struct drm_device *dev = data; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = data; u32 hw_max, hw_min; int ret; @@ -5002,7 +4922,8 @@ i915_min_freq_set(void *data, u64 val) hw_max = dev_priv->rps.max_freq; hw_min = dev_priv->rps.min_freq; - if (val < hw_min || val > hw_max || val > dev_priv->rps.max_freq_softlimit) { + if (val < hw_min || + val > hw_max || val > dev_priv->rps.max_freq_softlimit) { mutex_unlock(&dev_priv->rps.hw_lock); return -EINVAL; } @@ -5023,12 +4944,12 @@ DEFINE_SIMPLE_ATTRIBUTE(i915_min_freq_fops, static int i915_cache_sharing_get(void *data, u64 *val) { - struct drm_device *dev = data; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = data; + struct drm_device *dev = &dev_priv->drm; u32 snpcr; int ret; - if (!(IS_GEN6(dev) || IS_GEN7(dev))) + if (!(IS_GEN6(dev_priv) || IS_GEN7(dev_priv))) return -ENODEV; ret = mutex_lock_interruptible(&dev->struct_mutex); @@ -5039,7 +4960,7 @@ i915_cache_sharing_get(void *data, u64 *val) snpcr = I915_READ(GEN6_MBCUNIT_SNPCR); intel_runtime_pm_put(dev_priv); - mutex_unlock(&dev_priv->drm.struct_mutex); + mutex_unlock(&dev->struct_mutex); *val = (snpcr & GEN6_MBC_SNPCR_MASK) >> GEN6_MBC_SNPCR_SHIFT; @@ -5049,11 +4970,10 @@ i915_cache_sharing_get(void *data, u64 *val) static int i915_cache_sharing_set(void *data, u64 val) { - struct drm_device *dev = data; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = data; u32 snpcr; - if (!(IS_GEN6(dev) || IS_GEN7(dev))) + if (!(IS_GEN6(dev_priv) || IS_GEN7(dev_priv))) return -ENODEV; if (val > 3) @@ -5076,18 +4996,9 @@ DEFINE_SIMPLE_ATTRIBUTE(i915_cache_sharing_fops, i915_cache_sharing_get, i915_cache_sharing_set, "%llu\n"); -struct sseu_dev_status { - unsigned int slice_total; - unsigned int subslice_total; - unsigned int subslice_per_slice; - unsigned int eu_total; - unsigned int eu_per_subslice; -}; - -static void cherryview_sseu_device_status(struct drm_device *dev, - struct sseu_dev_status *stat) +static void cherryview_sseu_device_status(struct drm_i915_private *dev_priv, + struct sseu_dev_info *sseu) { - struct drm_i915_private *dev_priv = to_i915(dev); int ss_max = 2; int ss; u32 sig1[ss_max], sig2[ss_max]; @@ -5104,28 +5015,27 @@ static void cherryview_sseu_device_status(struct drm_device *dev, /* skip disabled subslice */ continue; - stat->slice_total = 1; - stat->subslice_per_slice++; + sseu->slice_mask = BIT(0); + sseu->subslice_mask |= BIT(ss); eu_cnt = ((sig1[ss] & CHV_EU08_PG_ENABLE) ? 0 : 2) + ((sig1[ss] & CHV_EU19_PG_ENABLE) ? 0 : 2) + ((sig1[ss] & CHV_EU210_PG_ENABLE) ? 0 : 2) + ((sig2[ss] & CHV_EU311_PG_ENABLE) ? 0 : 2); - stat->eu_total += eu_cnt; - stat->eu_per_subslice = max(stat->eu_per_subslice, eu_cnt); + sseu->eu_total += eu_cnt; + sseu->eu_per_subslice = max_t(unsigned int, + sseu->eu_per_subslice, eu_cnt); } - stat->subslice_total = stat->subslice_per_slice; } -static void gen9_sseu_device_status(struct drm_device *dev, - struct sseu_dev_status *stat) +static void gen9_sseu_device_status(struct drm_i915_private *dev_priv, + struct sseu_dev_info *sseu) { - struct drm_i915_private *dev_priv = to_i915(dev); int s_max = 3, ss_max = 4; int s, ss; u32 s_reg[s_max], eu_reg[2*s_max], eu_mask[2]; /* BXT has a single slice and at most 3 subslices. */ - if (IS_BROXTON(dev)) { + if (IS_BROXTON(dev_priv)) { s_max = 1; ss_max = 3; } @@ -5146,133 +5056,134 @@ static void gen9_sseu_device_status(struct drm_device *dev, GEN9_PGCTL_SSB_EU311_ACK; for (s = 0; s < s_max; s++) { - unsigned int ss_cnt = 0; - if ((s_reg[s] & GEN9_PGCTL_SLICE_ACK) == 0) /* skip disabled slice */ continue; - stat->slice_total++; + sseu->slice_mask |= BIT(s); - if (IS_SKYLAKE(dev) || IS_KABYLAKE(dev)) - ss_cnt = INTEL_INFO(dev)->subslice_per_slice; + if (IS_SKYLAKE(dev_priv) || IS_KABYLAKE(dev_priv)) + sseu->subslice_mask = + INTEL_INFO(dev_priv)->sseu.subslice_mask; for (ss = 0; ss < ss_max; ss++) { unsigned int eu_cnt; - if (IS_BROXTON(dev) && - !(s_reg[s] & (GEN9_PGCTL_SS_ACK(ss)))) - /* skip disabled subslice */ - continue; + if (IS_BROXTON(dev_priv)) { + if (!(s_reg[s] & (GEN9_PGCTL_SS_ACK(ss)))) + /* skip disabled subslice */ + continue; - if (IS_BROXTON(dev)) - ss_cnt++; + sseu->subslice_mask |= BIT(ss); + } eu_cnt = 2 * hweight32(eu_reg[2*s + ss/2] & eu_mask[ss%2]); - stat->eu_total += eu_cnt; - stat->eu_per_subslice = max(stat->eu_per_subslice, - eu_cnt); + sseu->eu_total += eu_cnt; + sseu->eu_per_subslice = max_t(unsigned int, + sseu->eu_per_subslice, + eu_cnt); } - - stat->subslice_total += ss_cnt; - stat->subslice_per_slice = max(stat->subslice_per_slice, - ss_cnt); } } -static void broadwell_sseu_device_status(struct drm_device *dev, - struct sseu_dev_status *stat) +static void broadwell_sseu_device_status(struct drm_i915_private *dev_priv, + struct sseu_dev_info *sseu) { - struct drm_i915_private *dev_priv = to_i915(dev); - int s; u32 slice_info = I915_READ(GEN8_GT_SLICE_INFO); + int s; - stat->slice_total = hweight32(slice_info & GEN8_LSLICESTAT_MASK); + sseu->slice_mask = slice_info & GEN8_LSLICESTAT_MASK; - if (stat->slice_total) { - stat->subslice_per_slice = INTEL_INFO(dev)->subslice_per_slice; - stat->subslice_total = stat->slice_total * - stat->subslice_per_slice; - stat->eu_per_subslice = INTEL_INFO(dev)->eu_per_subslice; - stat->eu_total = stat->eu_per_subslice * stat->subslice_total; + if (sseu->slice_mask) { + sseu->subslice_mask = INTEL_INFO(dev_priv)->sseu.subslice_mask; + sseu->eu_per_subslice = + INTEL_INFO(dev_priv)->sseu.eu_per_subslice; + sseu->eu_total = sseu->eu_per_subslice * + sseu_subslice_total(sseu); /* subtract fused off EU(s) from enabled slice(s) */ - for (s = 0; s < stat->slice_total; s++) { - u8 subslice_7eu = INTEL_INFO(dev)->subslice_7eu[s]; + for (s = 0; s < fls(sseu->slice_mask); s++) { + u8 subslice_7eu = + INTEL_INFO(dev_priv)->sseu.subslice_7eu[s]; - stat->eu_total -= hweight8(subslice_7eu); + sseu->eu_total -= hweight8(subslice_7eu); } } } -static int i915_sseu_status(struct seq_file *m, void *unused) +static void i915_print_sseu_info(struct seq_file *m, bool is_available_info, + const struct sseu_dev_info *sseu) { - struct drm_info_node *node = (struct drm_info_node *) m->private; - struct drm_i915_private *dev_priv = to_i915(node->minor->dev); - struct drm_device *dev = &dev_priv->drm; - struct sseu_dev_status stat; + struct drm_i915_private *dev_priv = node_to_i915(m->private); + const char *type = is_available_info ? "Available" : "Enabled"; - if (INTEL_INFO(dev)->gen < 8) - return -ENODEV; + seq_printf(m, " %s Slice Mask: %04x\n", type, + sseu->slice_mask); + seq_printf(m, " %s Slice Total: %u\n", type, + hweight8(sseu->slice_mask)); + seq_printf(m, " %s Subslice Total: %u\n", type, + sseu_subslice_total(sseu)); + seq_printf(m, " %s Subslice Mask: %04x\n", type, + sseu->subslice_mask); + seq_printf(m, " %s Subslice Per Slice: %u\n", type, + hweight8(sseu->subslice_mask)); + seq_printf(m, " %s EU Total: %u\n", type, + sseu->eu_total); + seq_printf(m, " %s EU Per Subslice: %u\n", type, + sseu->eu_per_subslice); + + if (!is_available_info) + return; + + seq_printf(m, " Has Pooled EU: %s\n", yesno(HAS_POOLED_EU(dev_priv))); + if (HAS_POOLED_EU(dev_priv)) + seq_printf(m, " Min EU in pool: %u\n", sseu->min_eu_in_pool); - seq_puts(m, "SSEU Device Info\n"); - seq_printf(m, " Available Slice Total: %u\n", - INTEL_INFO(dev)->slice_total); - seq_printf(m, " Available Subslice Total: %u\n", - INTEL_INFO(dev)->subslice_total); - seq_printf(m, " Available Subslice Per Slice: %u\n", - INTEL_INFO(dev)->subslice_per_slice); - seq_printf(m, " Available EU Total: %u\n", - INTEL_INFO(dev)->eu_total); - seq_printf(m, " Available EU Per Subslice: %u\n", - INTEL_INFO(dev)->eu_per_subslice); - seq_printf(m, " Has Pooled EU: %s\n", yesno(HAS_POOLED_EU(dev))); - if (HAS_POOLED_EU(dev)) - seq_printf(m, " Min EU in pool: %u\n", - INTEL_INFO(dev)->min_eu_in_pool); seq_printf(m, " Has Slice Power Gating: %s\n", - yesno(INTEL_INFO(dev)->has_slice_pg)); + yesno(sseu->has_slice_pg)); seq_printf(m, " Has Subslice Power Gating: %s\n", - yesno(INTEL_INFO(dev)->has_subslice_pg)); + yesno(sseu->has_subslice_pg)); seq_printf(m, " Has EU Power Gating: %s\n", - yesno(INTEL_INFO(dev)->has_eu_pg)); + yesno(sseu->has_eu_pg)); +} + +static int i915_sseu_status(struct seq_file *m, void *unused) +{ + struct drm_i915_private *dev_priv = node_to_i915(m->private); + struct sseu_dev_info sseu; + + if (INTEL_GEN(dev_priv) < 8) + return -ENODEV; + + seq_puts(m, "SSEU Device Info\n"); + i915_print_sseu_info(m, true, &INTEL_INFO(dev_priv)->sseu); seq_puts(m, "SSEU Device Status\n"); - memset(&stat, 0, sizeof(stat)); + memset(&sseu, 0, sizeof(sseu)); intel_runtime_pm_get(dev_priv); - if (IS_CHERRYVIEW(dev)) { - cherryview_sseu_device_status(dev, &stat); - } else if (IS_BROADWELL(dev)) { - broadwell_sseu_device_status(dev, &stat); - } else if (INTEL_INFO(dev)->gen >= 9) { - gen9_sseu_device_status(dev, &stat); + if (IS_CHERRYVIEW(dev_priv)) { + cherryview_sseu_device_status(dev_priv, &sseu); + } else if (IS_BROADWELL(dev_priv)) { + broadwell_sseu_device_status(dev_priv, &sseu); + } else if (INTEL_GEN(dev_priv) >= 9) { + gen9_sseu_device_status(dev_priv, &sseu); } intel_runtime_pm_put(dev_priv); - seq_printf(m, " Enabled Slice Total: %u\n", - stat.slice_total); - seq_printf(m, " Enabled Subslice Total: %u\n", - stat.subslice_total); - seq_printf(m, " Enabled Subslice Per Slice: %u\n", - stat.subslice_per_slice); - seq_printf(m, " Enabled EU Total: %u\n", - stat.eu_total); - seq_printf(m, " Enabled EU Per Subslice: %u\n", - stat.eu_per_subslice); + i915_print_sseu_info(m, false, &sseu); return 0; } static int i915_forcewake_open(struct inode *inode, struct file *file) { - struct drm_device *dev = inode->i_private; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = inode->i_private; - if (INTEL_INFO(dev)->gen < 6) + if (INTEL_GEN(dev_priv) < 6) return 0; intel_runtime_pm_get(dev_priv); @@ -5283,10 +5194,9 @@ static int i915_forcewake_open(struct inode *inode, struct file *file) static int i915_forcewake_release(struct inode *inode, struct file *file) { - struct drm_device *dev = inode->i_private; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = inode->i_private; - if (INTEL_INFO(dev)->gen < 6) + if (INTEL_GEN(dev_priv) < 6) return 0; intel_uncore_forcewake_put(dev_priv, FORCEWAKE_ALL); @@ -5303,12 +5213,11 @@ static const struct file_operations i915_forcewake_fops = { static int i915_forcewake_create(struct dentry *root, struct drm_minor *minor) { - struct drm_device *dev = minor->dev; struct dentry *ent; ent = debugfs_create_file("i915_forcewake_user", S_IRUSR, - root, dev, + root, to_i915(minor->dev), &i915_forcewake_fops); if (!ent) return -ENOMEM; @@ -5321,12 +5230,11 @@ static int i915_debugfs_create(struct dentry *root, const char *name, const struct file_operations *fops) { - struct drm_device *dev = minor->dev; struct dentry *ent; ent = debugfs_create_file(name, S_IRUGO | S_IWUSR, - root, dev, + root, to_i915(minor->dev), fops); if (!ent) return -ENOMEM; @@ -5413,9 +5321,8 @@ static const struct i915_debugfs_files { {"i915_dp_test_active", &i915_displayport_test_active_fops} }; -void intel_display_crc_init(struct drm_device *dev) +void intel_display_crc_init(struct drm_i915_private *dev_priv) { - struct drm_i915_private *dev_priv = to_i915(dev); enum pipe pipe; for_each_pipe(dev_priv, pipe) { @@ -5463,7 +5370,7 @@ void i915_debugfs_unregister(struct drm_i915_private *dev_priv) drm_debugfs_remove_files(i915_debugfs_list, I915_DEBUGFS_ENTRIES, minor); - drm_debugfs_remove_files((struct drm_info_list *) &i915_forcewake_fops, + drm_debugfs_remove_files((struct drm_info_list *)&i915_forcewake_fops, 1, minor); for (i = 0; i < ARRAY_SIZE(i915_pipe_crc_data); i++) { @@ -5475,7 +5382,7 @@ void i915_debugfs_unregister(struct drm_i915_private *dev_priv) for (i = 0; i < ARRAY_SIZE(i915_debugfs_files); i++) { struct drm_info_list *info_list = - (struct drm_info_list *) i915_debugfs_files[i].fops; + (struct drm_info_list *)i915_debugfs_files[i].fops; drm_debugfs_remove_files(info_list, 1, minor); } @@ -5555,6 +5462,40 @@ static const struct file_operations i915_dpcd_fops = { .release = single_release, }; +static int i915_panel_show(struct seq_file *m, void *data) +{ + struct drm_connector *connector = m->private; + struct intel_dp *intel_dp = + enc_to_intel_dp(&intel_attached_encoder(connector)->base); + + if (connector->status != connector_status_connected) + return -ENODEV; + + seq_printf(m, "Panel power up delay: %d\n", + intel_dp->panel_power_up_delay); + seq_printf(m, "Panel power down delay: %d\n", + intel_dp->panel_power_down_delay); + seq_printf(m, "Backlight on delay: %d\n", + intel_dp->backlight_on_delay); + seq_printf(m, "Backlight off delay: %d\n", + intel_dp->backlight_off_delay); + + return 0; +} + +static int i915_panel_open(struct inode *inode, struct file *file) +{ + return single_open(file, i915_panel_show, inode->i_private); +} + +static const struct file_operations i915_panel_fops = { + .owner = THIS_MODULE, + .open = i915_panel_open, + .read = seq_read, + .llseek = seq_lseek, + .release = single_release, +}; + /** * i915_debugfs_connector_add - add i915 specific connector debugfs files * @connector: pointer to a registered drm_connector @@ -5574,8 +5515,12 @@ int i915_debugfs_connector_add(struct drm_connector *connector) if (connector->connector_type == DRM_MODE_CONNECTOR_DisplayPort || connector->connector_type == DRM_MODE_CONNECTOR_eDP) - debugfs_create_file("i915_dpcd", S_IRUGO, root, connector, - &i915_dpcd_fops); + debugfs_create_file("i915_dpcd", S_IRUGO, root, + connector, &i915_dpcd_fops); + + if (connector->connector_type == DRM_MODE_CONNECTOR_eDP) + debugfs_create_file("i915_panel_timings", S_IRUGO, root, + connector, &i915_panel_fops); return 0; } diff --git a/drivers/gpu/drm/i915/i915_drv.c b/drivers/gpu/drm/i915/i915_drv.c index 13ae340ef1f3..7f4e8adec8a8 100644 --- a/drivers/gpu/drm/i915/i915_drv.c +++ b/drivers/gpu/drm/i915/i915_drv.c @@ -77,7 +77,7 @@ __i915_printk(struct drm_i915_private *dev_priv, const char *level, const char *fmt, ...) { static bool shown_bug_once; - struct device *dev = dev_priv->drm.dev; + struct device *kdev = dev_priv->drm.dev; bool is_error = level[1] <= KERN_ERR[1]; bool is_debug = level[1] == KERN_DEBUG[1]; struct va_format vaf; @@ -91,11 +91,11 @@ __i915_printk(struct drm_i915_private *dev_priv, const char *level, vaf.fmt = fmt; vaf.va = &args; - dev_printk(level, dev, "[" DRM_NAME ":%ps] %pV", + dev_printk(level, kdev, "[" DRM_NAME ":%ps] %pV", __builtin_return_address(0), &vaf); if (is_error && !shown_bug_once) { - dev_notice(dev, "%s", FDO_BUG_MSG); + dev_notice(kdev, "%s", FDO_BUG_MSG); shown_bug_once = true; } @@ -232,6 +232,7 @@ static int i915_getparam(struct drm_device *dev, void *data, struct drm_file *file_priv) { struct drm_i915_private *dev_priv = to_i915(dev); + struct pci_dev *pdev = dev_priv->drm.pdev; drm_i915_getparam_t *param = data; int value; @@ -242,13 +243,10 @@ static int i915_getparam(struct drm_device *dev, void *data, /* Reject all old ums/dri params. */ return -ENODEV; case I915_PARAM_CHIPSET_ID: - value = dev->pdev->device; + value = pdev->device; break; case I915_PARAM_REVISION: - value = dev->pdev->revision; - break; - case I915_PARAM_HAS_GEM: - value = 1; + value = pdev->revision; break; case I915_PARAM_NUM_FENCES_AVAIL: value = dev_priv->num_fence_regs; @@ -256,13 +254,6 @@ static int i915_getparam(struct drm_device *dev, void *data, case I915_PARAM_HAS_OVERLAY: value = dev_priv->overlay ? 1 : 0; break; - case I915_PARAM_HAS_PAGEFLIPPING: - value = 1; - break; - case I915_PARAM_HAS_EXECBUF2: - /* depends on GEM */ - value = 1; - break; case I915_PARAM_HAS_BSD: value = intel_engine_initialized(&dev_priv->engine[VCS]); break; @@ -275,67 +266,34 @@ static int i915_getparam(struct drm_device *dev, void *data, case I915_PARAM_HAS_BSD2: value = intel_engine_initialized(&dev_priv->engine[VCS2]); break; - case I915_PARAM_HAS_RELAXED_FENCING: - value = 1; - break; - case I915_PARAM_HAS_COHERENT_RINGS: - value = 1; - break; case I915_PARAM_HAS_EXEC_CONSTANTS: - value = INTEL_INFO(dev)->gen >= 4; - break; - case I915_PARAM_HAS_RELAXED_DELTA: - value = 1; - break; - case I915_PARAM_HAS_GEN7_SOL_RESET: - value = 1; + value = INTEL_GEN(dev_priv) >= 4; break; case I915_PARAM_HAS_LLC: - value = HAS_LLC(dev); + value = HAS_LLC(dev_priv); break; case I915_PARAM_HAS_WT: - value = HAS_WT(dev); + value = HAS_WT(dev_priv); break; case I915_PARAM_HAS_ALIASING_PPGTT: - value = USES_PPGTT(dev); - break; - case I915_PARAM_HAS_WAIT_TIMEOUT: - value = 1; + value = USES_PPGTT(dev_priv); break; case I915_PARAM_HAS_SEMAPHORES: value = i915.semaphores; break; - case I915_PARAM_HAS_PRIME_VMAP_FLUSH: - value = 1; - break; case I915_PARAM_HAS_SECURE_BATCHES: value = capable(CAP_SYS_ADMIN); break; - case I915_PARAM_HAS_PINNED_BATCHES: - value = 1; - break; - case I915_PARAM_HAS_EXEC_NO_RELOC: - value = 1; - break; - case I915_PARAM_HAS_EXEC_HANDLE_LUT: - value = 1; - break; case I915_PARAM_CMD_PARSER_VERSION: value = i915_cmd_parser_get_version(dev_priv); break; - case I915_PARAM_HAS_COHERENT_PHYS_GTT: - value = 1; - break; - case I915_PARAM_MMAP_VERSION: - value = 1; - break; case I915_PARAM_SUBSLICE_TOTAL: - value = INTEL_INFO(dev)->subslice_total; + value = sseu_subslice_total(&INTEL_INFO(dev_priv)->sseu); if (!value) return -ENODEV; break; case I915_PARAM_EU_TOTAL: - value = INTEL_INFO(dev)->eu_total; + value = INTEL_INFO(dev_priv)->sseu.eu_total; if (!value) return -ENODEV; break; @@ -343,16 +301,43 @@ static int i915_getparam(struct drm_device *dev, void *data, value = i915.enable_hangcheck && intel_has_gpu_reset(dev_priv); break; case I915_PARAM_HAS_RESOURCE_STREAMER: - value = HAS_RESOURCE_STREAMER(dev); - break; - case I915_PARAM_HAS_EXEC_SOFTPIN: - value = 1; + value = HAS_RESOURCE_STREAMER(dev_priv); break; case I915_PARAM_HAS_POOLED_EU: - value = HAS_POOLED_EU(dev); + value = HAS_POOLED_EU(dev_priv); break; case I915_PARAM_MIN_EU_IN_POOL: - value = INTEL_INFO(dev)->min_eu_in_pool; + value = INTEL_INFO(dev_priv)->sseu.min_eu_in_pool; + break; + case I915_PARAM_MMAP_GTT_VERSION: + /* Though we've started our numbering from 1, and so class all + * earlier versions as 0, in effect their value is undefined as + * the ioctl will report EINVAL for the unknown param! + */ + value = i915_gem_mmap_gtt_version(); + break; + case I915_PARAM_MMAP_VERSION: + /* Remember to bump this if the version changes! */ + case I915_PARAM_HAS_GEM: + case I915_PARAM_HAS_PAGEFLIPPING: + case I915_PARAM_HAS_EXECBUF2: /* depends on GEM */ + case I915_PARAM_HAS_RELAXED_FENCING: + case I915_PARAM_HAS_COHERENT_RINGS: + case I915_PARAM_HAS_RELAXED_DELTA: + case I915_PARAM_HAS_GEN7_SOL_RESET: + case I915_PARAM_HAS_WAIT_TIMEOUT: + case I915_PARAM_HAS_PRIME_VMAP_FLUSH: + case I915_PARAM_HAS_PINNED_BATCHES: + case I915_PARAM_HAS_EXEC_NO_RELOC: + case I915_PARAM_HAS_EXEC_HANDLE_LUT: + case I915_PARAM_HAS_COHERENT_PHYS_GTT: + case I915_PARAM_HAS_EXEC_SOFTPIN: + /* For the time being all of these are always true; + * if some supported hardware does not have one of these + * features this value needs to be provided from + * INTEL_INFO(), a feature macro, or similar. + */ + value = 1; break; default: DRM_DEBUG("Unknown parameter %d\n", param->param); @@ -516,7 +501,7 @@ static void i915_switcheroo_set_state(struct pci_dev *pdev, enum vga_switcheroo_ pr_info("switched on\n"); dev->switch_power_state = DRM_SWITCH_POWER_CHANGING; /* i915 resume handler doesn't set to D0 */ - pci_set_power_state(dev->pdev, PCI_D0); + pci_set_power_state(pdev, PCI_D0); i915_resume_switcheroo(dev); dev->switch_power_state = DRM_SWITCH_POWER_ON; } else { @@ -574,7 +559,6 @@ static void i915_gem_fini(struct drm_device *dev) } mutex_lock(&dev->struct_mutex); - i915_gem_reset(dev); i915_gem_cleanup_engines(dev); i915_gem_context_fini(dev); mutex_unlock(&dev->struct_mutex); @@ -585,6 +569,7 @@ static void i915_gem_fini(struct drm_device *dev) static int i915_load_modeset_init(struct drm_device *dev) { struct drm_i915_private *dev_priv = to_i915(dev); + struct pci_dev *pdev = dev_priv->drm.pdev; int ret; if (i915_inject_load_failure()) @@ -601,13 +586,13 @@ static int i915_load_modeset_init(struct drm_device *dev) * then we do not take part in VGA arbitration and the * vga_client_register() fails with -ENODEV. */ - ret = vga_client_register(dev->pdev, dev, NULL, i915_vga_set_decode); + ret = vga_client_register(pdev, dev, NULL, i915_vga_set_decode); if (ret && ret != -ENODEV) goto out; intel_register_dsm_handler(); - ret = vga_switcheroo_register_client(dev->pdev, &i915_switcheroo_ops, false); + ret = vga_switcheroo_register_client(pdev, &i915_switcheroo_ops, false); if (ret) goto cleanup_vga_client; @@ -659,9 +644,9 @@ cleanup_irq: cleanup_csr: intel_csr_ucode_fini(dev_priv); intel_power_domains_fini(dev_priv); - vga_switcheroo_unregister_client(dev->pdev); + vga_switcheroo_unregister_client(pdev); cleanup_vga_client: - vga_client_register(dev->pdev, NULL, NULL, NULL); + vga_client_register(pdev, NULL, NULL, NULL); out: return ret; } @@ -849,7 +834,7 @@ static int i915_driver_init_early(struct drm_i915_private *dev_priv, intel_init_audio_hooks(dev_priv); i915_gem_load_init(&dev_priv->drm); - intel_display_crc_init(&dev_priv->drm); + intel_display_crc_init(dev_priv); intel_device_info_dump(dev_priv); @@ -881,6 +866,7 @@ static void i915_driver_cleanup_early(struct drm_i915_private *dev_priv) static int i915_mmio_setup(struct drm_device *dev) { struct drm_i915_private *dev_priv = to_i915(dev); + struct pci_dev *pdev = dev_priv->drm.pdev; int mmio_bar; int mmio_size; @@ -897,7 +883,7 @@ static int i915_mmio_setup(struct drm_device *dev) mmio_size = 512 * 1024; else mmio_size = 2 * 1024 * 1024; - dev_priv->regs = pci_iomap(dev->pdev, mmio_bar, mmio_size); + dev_priv->regs = pci_iomap(pdev, mmio_bar, mmio_size); if (dev_priv->regs == NULL) { DRM_ERROR("failed to map registers\n"); @@ -913,9 +899,10 @@ static int i915_mmio_setup(struct drm_device *dev) static void i915_mmio_cleanup(struct drm_device *dev) { struct drm_i915_private *dev_priv = to_i915(dev); + struct pci_dev *pdev = dev_priv->drm.pdev; intel_teardown_mchbar(dev); - pci_iounmap(dev->pdev, dev_priv->regs); + pci_iounmap(pdev, dev_priv->regs); } /** @@ -994,6 +981,7 @@ static void intel_sanitize_options(struct drm_i915_private *dev_priv) */ static int i915_driver_init_hw(struct drm_i915_private *dev_priv) { + struct pci_dev *pdev = dev_priv->drm.pdev; struct drm_device *dev = &dev_priv->drm; int ret; @@ -1032,11 +1020,11 @@ static int i915_driver_init_hw(struct drm_i915_private *dev_priv) goto out_ggtt; } - pci_set_master(dev->pdev); + pci_set_master(pdev); /* overlay on gen2 is broken and can't address above 1G */ if (IS_GEN2(dev)) { - ret = dma_set_coherent_mask(&dev->pdev->dev, DMA_BIT_MASK(30)); + ret = dma_set_coherent_mask(&pdev->dev, DMA_BIT_MASK(30)); if (ret) { DRM_ERROR("failed to set DMA mask\n"); @@ -1053,7 +1041,7 @@ static int i915_driver_init_hw(struct drm_i915_private *dev_priv) * which also needs to be handled carefully. */ if (IS_BROADWATER(dev) || IS_CRESTLINE(dev)) { - ret = dma_set_coherent_mask(&dev->pdev->dev, DMA_BIT_MASK(32)); + ret = dma_set_coherent_mask(&pdev->dev, DMA_BIT_MASK(32)); if (ret) { DRM_ERROR("failed to set DMA mask\n"); @@ -1083,7 +1071,7 @@ static int i915_driver_init_hw(struct drm_i915_private *dev_priv) * stuck interrupts on some machines. */ if (!IS_I945G(dev) && !IS_I945GM(dev)) { - if (pci_enable_msi(dev->pdev) < 0) + if (pci_enable_msi(pdev) < 0) DRM_DEBUG_DRIVER("can't enable MSI"); } @@ -1101,10 +1089,10 @@ out_ggtt: */ static void i915_driver_cleanup_hw(struct drm_i915_private *dev_priv) { - struct drm_device *dev = &dev_priv->drm; + struct pci_dev *pdev = dev_priv->drm.pdev; - if (dev->pdev->msi_enabled) - pci_disable_msi(dev->pdev); + if (pdev->msi_enabled) + pci_disable_msi(pdev); pm_qos_remove_request(&dev_priv->pm_qos); i915_ggtt_cleanup_hw(dev_priv); @@ -1133,7 +1121,7 @@ static void i915_driver_register(struct drm_i915_private *dev_priv) /* Reveal our presence to userspace */ if (drm_dev_register(dev, 0) == 0) { i915_debugfs_register(dev_priv); - i915_setup_sysfs(dev); + i915_setup_sysfs(dev_priv); } else DRM_ERROR("Failed to register driver for userspace access!\n"); @@ -1170,7 +1158,7 @@ static void i915_driver_unregister(struct drm_i915_private *dev_priv) acpi_video_unregister(); intel_opregion_unregister(dev_priv); - i915_teardown_sysfs(&dev_priv->drm); + i915_teardown_sysfs(dev_priv); i915_debugfs_unregister(dev_priv); drm_dev_unregister(&dev_priv->drm); @@ -1250,6 +1238,11 @@ int i915_driver_load(struct pci_dev *pdev, const struct pci_device_id *ent) intel_runtime_pm_enable(dev_priv); + /* Everything is in place, we can now relax! */ + DRM_INFO("Initialized %s %d.%d.%d %s for %s on minor %d\n", + driver.name, driver.major, driver.minor, driver.patchlevel, + driver.date, pci_name(pdev), dev_priv->drm.primary->index); + intel_runtime_pm_put(dev_priv); return 0; @@ -1274,6 +1267,7 @@ out_free_priv: void i915_driver_unload(struct drm_device *dev) { struct drm_i915_private *dev_priv = to_i915(dev); + struct pci_dev *pdev = dev_priv->drm.pdev; intel_fbdev_fini(dev); @@ -1302,8 +1296,8 @@ void i915_driver_unload(struct drm_device *dev) kfree(dev_priv->vbt.lfp_lvds_vbt_mode); dev_priv->vbt.lfp_lvds_vbt_mode = NULL; - vga_switcheroo_unregister_client(dev->pdev); - vga_client_register(dev->pdev, NULL, NULL, NULL); + vga_switcheroo_unregister_client(pdev); + vga_client_register(pdev, NULL, NULL, NULL); intel_csr_ucode_fini(dev_priv); @@ -1400,6 +1394,7 @@ static bool suspend_to_idle(struct drm_i915_private *dev_priv) static int i915_drm_suspend(struct drm_device *dev) { struct drm_i915_private *dev_priv = to_i915(dev); + struct pci_dev *pdev = dev_priv->drm.pdev; pci_power_t opregion_target_state; int error; @@ -1416,11 +1411,11 @@ static int i915_drm_suspend(struct drm_device *dev) drm_kms_helper_poll_disable(dev); - pci_save_state(dev->pdev); + pci_save_state(pdev); error = i915_gem_suspend(dev); if (error) { - dev_err(&dev->pdev->dev, + dev_err(&pdev->dev, "GEM idle failed, resume might fail\n"); goto out; } @@ -1462,9 +1457,10 @@ out: return error; } -static int i915_drm_suspend_late(struct drm_device *drm_dev, bool hibernation) +static int i915_drm_suspend_late(struct drm_device *dev, bool hibernation) { - struct drm_i915_private *dev_priv = to_i915(drm_dev); + struct drm_i915_private *dev_priv = to_i915(dev); + struct pci_dev *pdev = dev_priv->drm.pdev; bool fw_csr; int ret; @@ -1498,7 +1494,7 @@ static int i915_drm_suspend_late(struct drm_device *drm_dev, bool hibernation) goto out; } - pci_disable_device(drm_dev->pdev); + pci_disable_device(pdev); /* * During hibernation on some platforms the BIOS may try to access * the device even though it's already in D3 and hang the machine. So @@ -1512,7 +1508,7 @@ static int i915_drm_suspend_late(struct drm_device *drm_dev, bool hibernation) * Acer Aspire 1830T */ if (!(hibernation && INTEL_INFO(dev_priv)->gen < 6)) - pci_set_power_state(drm_dev->pdev, PCI_D3hot); + pci_set_power_state(pdev, PCI_D3hot); dev_priv->suspended_to_idle = suspend_to_idle(dev_priv); @@ -1552,6 +1548,7 @@ static int i915_drm_resume(struct drm_device *dev) int ret; disable_rpm_wakeref_asserts(dev_priv); + intel_sanitize_gt_powersave(dev_priv); ret = i915_ggtt_enable_hw(dev_priv); if (ret) @@ -1581,7 +1578,7 @@ static int i915_drm_resume(struct drm_device *dev) mutex_lock(&dev->struct_mutex); if (i915_gem_init_hw(dev)) { DRM_ERROR("failed to re-initialize GPU, declaring wedged!\n"); - atomic_or(I915_WEDGED, &dev_priv->gpu_error.reset_counter); + i915_gem_set_wedged(dev_priv); } mutex_unlock(&dev->struct_mutex); @@ -1629,6 +1626,7 @@ static int i915_drm_resume(struct drm_device *dev) static int i915_drm_resume_early(struct drm_device *dev) { struct drm_i915_private *dev_priv = to_i915(dev); + struct pci_dev *pdev = dev_priv->drm.pdev; int ret; /* @@ -1651,7 +1649,7 @@ static int i915_drm_resume_early(struct drm_device *dev) * the device powered we can also remove the following set power state * call. */ - ret = pci_set_power_state(dev->pdev, PCI_D0); + ret = pci_set_power_state(pdev, PCI_D0); if (ret) { DRM_ERROR("failed to set PCI D0 power state (%d)\n", ret); goto out; @@ -1670,12 +1668,12 @@ static int i915_drm_resume_early(struct drm_device *dev) * depend on the device enable refcount we can't anyway depend on them * disabling/enabling the device. */ - if (pci_enable_device(dev->pdev)) { + if (pci_enable_device(pdev)) { ret = -EIO; goto out; } - pci_set_master(dev->pdev); + pci_set_master(pdev); disable_rpm_wakeref_asserts(dev_priv); @@ -1727,8 +1725,10 @@ int i915_resume_switcheroo(struct drm_device *dev) * i915_reset - reset chip after a hang * @dev: drm device to reset * - * Reset the chip. Useful if a hang is detected. Returns zero on successful - * reset or otherwise an error code. + * Reset the chip. Useful if a hang is detected. Marks the device as wedged + * on failure. + * + * Caller must hold the struct_mutex. * * Procedure is fairly simple: * - reset the chip using the reset reg @@ -1738,29 +1738,22 @@ int i915_resume_switcheroo(struct drm_device *dev) * - re-init interrupt state * - re-init display */ -int i915_reset(struct drm_i915_private *dev_priv) +void i915_reset(struct drm_i915_private *dev_priv) { struct drm_device *dev = &dev_priv->drm; struct i915_gpu_error *error = &dev_priv->gpu_error; - unsigned reset_counter; int ret; - mutex_lock(&dev->struct_mutex); + lockdep_assert_held(&dev->struct_mutex); - /* Clear any previous failed attempts at recovery. Time to try again. */ - atomic_andnot(I915_WEDGED, &error->reset_counter); + if (!test_and_clear_bit(I915_RESET_IN_PROGRESS, &error->flags)) + return; - /* Clear the reset-in-progress flag and increment the reset epoch. */ - reset_counter = atomic_inc_return(&error->reset_counter); - if (WARN_ON(__i915_reset_in_progress(reset_counter))) { - ret = -EIO; - goto error; - } + /* Clear any previous failed attempts at recovery. Time to try again. */ + __clear_bit(I915_WEDGED, &error->flags); + error->reset_count++; pr_notice("drm/i915: Resetting chip after gpu hang\n"); - - i915_gem_reset(dev); - ret = intel_gpu_reset(dev_priv, ALL_ENGINES); if (ret) { if (ret != -ENODEV) @@ -1770,6 +1763,7 @@ int i915_reset(struct drm_i915_private *dev_priv) goto error; } + i915_gem_reset(dev_priv); intel_overlay_reset(dev_priv); /* Ok, now get things going again... */ @@ -1792,43 +1786,43 @@ int i915_reset(struct drm_i915_private *dev_priv) goto error; } - mutex_unlock(&dev->struct_mutex); - /* * rps/rc6 re-init is necessary to restore state lost after the * reset and the re-install of gt irqs. Skip for ironlake per * previous concerns that it doesn't respond well to some forms * of re-init after reset. */ + intel_sanitize_gt_powersave(dev_priv); intel_autoenable_gt_powersave(dev_priv); - return 0; +wakeup: + wake_up_bit(&error->flags, I915_RESET_IN_PROGRESS); + return; error: - atomic_or(I915_WEDGED, &error->reset_counter); - mutex_unlock(&dev->struct_mutex); - return ret; + i915_gem_set_wedged(dev_priv); + goto wakeup; } -static int i915_pm_suspend(struct device *dev) +static int i915_pm_suspend(struct device *kdev) { - struct pci_dev *pdev = to_pci_dev(dev); - struct drm_device *drm_dev = pci_get_drvdata(pdev); + struct pci_dev *pdev = to_pci_dev(kdev); + struct drm_device *dev = pci_get_drvdata(pdev); - if (!drm_dev) { - dev_err(dev, "DRM not initialized, aborting suspend.\n"); + if (!dev) { + dev_err(kdev, "DRM not initialized, aborting suspend.\n"); return -ENODEV; } - if (drm_dev->switch_power_state == DRM_SWITCH_POWER_OFF) + if (dev->switch_power_state == DRM_SWITCH_POWER_OFF) return 0; - return i915_drm_suspend(drm_dev); + return i915_drm_suspend(dev); } -static int i915_pm_suspend_late(struct device *dev) +static int i915_pm_suspend_late(struct device *kdev) { - struct drm_device *drm_dev = &dev_to_i915(dev)->drm; + struct drm_device *dev = &kdev_to_i915(kdev)->drm; /* * We have a suspend ordering issue with the snd-hda driver also @@ -1839,57 +1833,57 @@ static int i915_pm_suspend_late(struct device *dev) * FIXME: This should be solved with a special hdmi sink device or * similar so that power domains can be employed. */ - if (drm_dev->switch_power_state == DRM_SWITCH_POWER_OFF) + if (dev->switch_power_state == DRM_SWITCH_POWER_OFF) return 0; - return i915_drm_suspend_late(drm_dev, false); + return i915_drm_suspend_late(dev, false); } -static int i915_pm_poweroff_late(struct device *dev) +static int i915_pm_poweroff_late(struct device *kdev) { - struct drm_device *drm_dev = &dev_to_i915(dev)->drm; + struct drm_device *dev = &kdev_to_i915(kdev)->drm; - if (drm_dev->switch_power_state == DRM_SWITCH_POWER_OFF) + if (dev->switch_power_state == DRM_SWITCH_POWER_OFF) return 0; - return i915_drm_suspend_late(drm_dev, true); + return i915_drm_suspend_late(dev, true); } -static int i915_pm_resume_early(struct device *dev) +static int i915_pm_resume_early(struct device *kdev) { - struct drm_device *drm_dev = &dev_to_i915(dev)->drm; + struct drm_device *dev = &kdev_to_i915(kdev)->drm; - if (drm_dev->switch_power_state == DRM_SWITCH_POWER_OFF) + if (dev->switch_power_state == DRM_SWITCH_POWER_OFF) return 0; - return i915_drm_resume_early(drm_dev); + return i915_drm_resume_early(dev); } -static int i915_pm_resume(struct device *dev) +static int i915_pm_resume(struct device *kdev) { - struct drm_device *drm_dev = &dev_to_i915(dev)->drm; + struct drm_device *dev = &kdev_to_i915(kdev)->drm; - if (drm_dev->switch_power_state == DRM_SWITCH_POWER_OFF) + if (dev->switch_power_state == DRM_SWITCH_POWER_OFF) return 0; - return i915_drm_resume(drm_dev); + return i915_drm_resume(dev); } /* freeze: before creating the hibernation_image */ -static int i915_pm_freeze(struct device *dev) +static int i915_pm_freeze(struct device *kdev) { - return i915_pm_suspend(dev); + return i915_pm_suspend(kdev); } -static int i915_pm_freeze_late(struct device *dev) +static int i915_pm_freeze_late(struct device *kdev) { int ret; - ret = i915_pm_suspend_late(dev); + ret = i915_pm_suspend_late(kdev); if (ret) return ret; - ret = i915_gem_freeze_late(dev_to_i915(dev)); + ret = i915_gem_freeze_late(kdev_to_i915(kdev)); if (ret) return ret; @@ -1897,25 +1891,25 @@ static int i915_pm_freeze_late(struct device *dev) } /* thaw: called after creating the hibernation image, but before turning off. */ -static int i915_pm_thaw_early(struct device *dev) +static int i915_pm_thaw_early(struct device *kdev) { - return i915_pm_resume_early(dev); + return i915_pm_resume_early(kdev); } -static int i915_pm_thaw(struct device *dev) +static int i915_pm_thaw(struct device *kdev) { - return i915_pm_resume(dev); + return i915_pm_resume(kdev); } /* restore: called after loading the hibernation image. */ -static int i915_pm_restore_early(struct device *dev) +static int i915_pm_restore_early(struct device *kdev) { - return i915_pm_resume_early(dev); + return i915_pm_resume_early(kdev); } -static int i915_pm_restore(struct device *dev) +static int i915_pm_restore(struct device *kdev) { - return i915_pm_resume(dev); + return i915_pm_resume(kdev); } /* @@ -2277,9 +2271,9 @@ static int vlv_resume_prepare(struct drm_i915_private *dev_priv, return ret; } -static int intel_runtime_suspend(struct device *device) +static int intel_runtime_suspend(struct device *kdev) { - struct pci_dev *pdev = to_pci_dev(device); + struct pci_dev *pdev = to_pci_dev(kdev); struct drm_device *dev = pci_get_drvdata(pdev); struct drm_i915_private *dev_priv = to_i915(dev); int ret; @@ -2305,7 +2299,7 @@ static int intel_runtime_suspend(struct device *device) * Bump the expiration timestamp, otherwise the suspend won't * be rescheduled. */ - pm_runtime_mark_last_busy(device); + pm_runtime_mark_last_busy(kdev); return -EAGAIN; } @@ -2384,9 +2378,9 @@ static int intel_runtime_suspend(struct device *device) return 0; } -static int intel_runtime_resume(struct device *device) +static int intel_runtime_resume(struct device *kdev) { - struct pci_dev *pdev = to_pci_dev(device); + struct pci_dev *pdev = to_pci_dev(kdev); struct drm_device *dev = pci_get_drvdata(pdev); struct drm_i915_private *dev_priv = to_i915(dev); int ret = 0; diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h index e6069057eb98..4dd307ed4336 100644 --- a/drivers/gpu/drm/i915/i915_drv.h +++ b/drivers/gpu/drm/i915/i915_drv.h @@ -70,7 +70,7 @@ #define DRIVER_NAME "i915" #define DRIVER_DESC "Intel Graphics" -#define DRIVER_DATE "20160822" +#define DRIVER_DATE "20160919" #undef WARN_ON /* Many gcc seem to no see through this and fall over :( */ @@ -510,8 +510,12 @@ struct drm_i915_display_funcs { struct intel_initial_plane_config *); int (*crtc_compute_clock)(struct intel_crtc *crtc, struct intel_crtc_state *crtc_state); - void (*crtc_enable)(struct drm_crtc *crtc); - void (*crtc_disable)(struct drm_crtc *crtc); + void (*crtc_enable)(struct intel_crtc_state *pipe_config, + struct drm_atomic_state *old_state); + void (*crtc_disable)(struct intel_crtc_state *old_crtc_state, + struct drm_atomic_state *old_state); + void (*update_crtcs)(struct drm_atomic_state *state, + unsigned int *crtc_vblank_mask); void (*audio_codec_enable)(struct drm_connector *connector, struct intel_encoder *encoder, const struct drm_display_mode *adjusted_mode); @@ -575,8 +579,6 @@ struct intel_uncore_funcs { uint16_t val, bool trace); void (*mmio_writel)(struct drm_i915_private *dev_priv, i915_reg_t r, uint32_t val, bool trace); - void (*mmio_writeq)(struct drm_i915_private *dev_priv, i915_reg_t r, - uint64_t val, bool trace); }; struct intel_uncore { @@ -637,7 +639,7 @@ struct intel_csr { func(is_i915g) sep \ func(is_i945gm) sep \ func(is_g33) sep \ - func(need_gfx_hws) sep \ + func(hws_needs_physical) sep \ func(is_g4x) sep \ func(is_pineview) sep \ func(is_broadwater) sep \ @@ -652,6 +654,19 @@ struct intel_csr { func(is_kabylake) sep \ func(is_preliminary) sep \ func(has_fbc) sep \ + func(has_psr) sep \ + func(has_runtime_pm) sep \ + func(has_csr) sep \ + func(has_resource_streamer) sep \ + func(has_rc6) sep \ + func(has_rc6p) sep \ + func(has_dp_mst) sep \ + func(has_gmbus_irq) sep \ + func(has_hw_contexts) sep \ + func(has_logical_ring_contexts) sep \ + func(has_l3_dpf) sep \ + func(has_gmch_display) sep \ + func(has_guc) sep \ func(has_pipe_cxsr) sep \ func(has_hotplug) sep \ func(cursor_needs_physical) sep \ @@ -667,6 +682,24 @@ struct intel_csr { #define DEFINE_FLAG(name) u8 name:1 #define SEP_SEMICOLON ; +struct sseu_dev_info { + u8 slice_mask; + u8 subslice_mask; + u8 eu_total; + u8 eu_per_subslice; + u8 min_eu_in_pool; + /* For each slice, which subslice(s) has(have) 7 EUs (bitfield)? */ + u8 subslice_7eu[3]; + u8 has_slice_pg:1; + u8 has_subslice_pg:1; + u8 has_eu_pg:1; +}; + +static inline unsigned int sseu_subslice_total(const struct sseu_dev_info *sseu) +{ + return hweight8(sseu->slice_mask) * hweight8(sseu->subslice_mask); +} + struct intel_device_info { u32 display_mmio_offset; u16 device_id; @@ -677,6 +710,7 @@ struct intel_device_info { u8 ring_mask; /* Rings supported by the HW */ u8 num_rings; DEV_INFO_FOR_EACH_FLAG(DEFINE_FLAG, SEP_SEMICOLON); + u16 ddb_size; /* in blocks */ /* Register offsets for the various display pipes and transcoders */ int pipe_offsets[I915_MAX_TRANSCODERS]; int trans_offsets[I915_MAX_TRANSCODERS]; @@ -684,17 +718,7 @@ struct intel_device_info { int cursor_offsets[I915_MAX_PIPES]; /* Slice/subslice/EU info */ - u8 slice_total; - u8 subslice_total; - u8 subslice_per_slice; - u8 eu_total; - u8 eu_per_subslice; - u8 min_eu_in_pool; - /* For each slice, which subslice(s) has(have) 7 EUs (bitfield)? */ - u8 subslice_7eu[3]; - u8 has_slice_pg:1; - u8 has_subslice_pg:1; - u8 has_eu_pg:1; + struct sseu_dev_info sseu; struct color_luts { u16 degamma_lut_size; @@ -1161,6 +1185,7 @@ struct intel_gen6_power_mgmt { bool interrupts_enabled; u32 pm_iir; + /* PM interrupt bits that should never be masked */ u32 pm_intr_keep; /* Frequencies are stored in potentially platform dependent multiples. @@ -1334,7 +1359,7 @@ struct i915_gem_mm { bool interruptible; /* the indicator for dispatch video commands on two BSD rings */ - unsigned int bsd_engine_dispatch_index; + atomic_t bsd_engine_dispatch_index; /** Bit 6 swizzling required for X tiling */ uint32_t bit_6_swizzle_x; @@ -1382,9 +1407,10 @@ struct i915_gpu_error { * State variable controlling the reset flow and count * * This is a counter which gets incremented when reset is triggered, - * and again when reset has been handled. So odd values (lowest bit set) - * means that reset is in progress and even values that - * (reset_counter >> 1):th reset was successfully completed. + * + * Before the reset commences, the I915_RESET_IN_PROGRESS bit is set + * meaning that any waiters holding onto the struct_mutex should + * relinquish the lock immediately in order for the reset to start. * * If reset is not completed succesfully, the I915_WEDGE bit is * set meaning that hardware is terminally sour and there is no @@ -1399,10 +1425,11 @@ struct i915_gpu_error { * naturally enforces the correct ordering between the bail-out of the * waiter and the gpu reset work code. */ - atomic_t reset_counter; + unsigned long reset_count; -#define I915_RESET_IN_PROGRESS_FLAG 1 -#define I915_WEDGED (1 << 31) + unsigned long flags; +#define I915_RESET_IN_PROGRESS 0 +#define I915_WEDGED (BITS_PER_LONG - 1) /** * Waitqueue to signal when a hang is detected. Used to for waiters @@ -1956,6 +1983,13 @@ struct drm_i915_private { struct i915_suspend_saved_registers regfile; struct vlv_s0ix_state vlv_s0ix_state; + enum { + I915_SKL_SAGV_UNKNOWN = 0, + I915_SKL_SAGV_DISABLED, + I915_SKL_SAGV_ENABLED, + I915_SKL_SAGV_NOT_CONTROLLED + } skl_sagv_status; + struct { /* * Raw watermark latency values: @@ -2010,6 +2044,7 @@ struct drm_i915_private { /* Abstract the submission mechanism (legacy ringbuffer or execlists) away */ struct { + void (*resume)(struct drm_i915_private *); void (*cleanup_engine)(struct intel_engine_cs *engine); /** @@ -2057,9 +2092,9 @@ static inline struct drm_i915_private *to_i915(const struct drm_device *dev) return container_of(dev, struct drm_i915_private, drm); } -static inline struct drm_i915_private *dev_to_i915(struct device *dev) +static inline struct drm_i915_private *kdev_to_i915(struct device *kdev) { - return to_i915(dev_get_drvdata(dev)); + return to_i915(dev_get_drvdata(kdev)); } static inline struct drm_i915_private *guc_to_i915(struct intel_guc *guc) @@ -2082,13 +2117,16 @@ static inline struct drm_i915_private *guc_to_i915(struct intel_guc *guc) for_each_if (((id__) = (engine__)->id, \ intel_engine_initialized(engine__))) +#define __mask_next_bit(mask) ({ \ + int __idx = ffs(mask) - 1; \ + mask &= ~BIT(__idx); \ + __idx; \ +}) + /* Iterator over subset of engines selected by mask */ -#define for_each_engine_masked(engine__, dev_priv__, mask__) \ - for ((engine__) = &(dev_priv__)->engine[0]; \ - (engine__) < &(dev_priv__)->engine[I915_NUM_ENGINES]; \ - (engine__)++) \ - for_each_if (((mask__) & intel_engine_flag(engine__)) && \ - intel_engine_initialized(engine__)) +#define for_each_engine_masked(engine__, dev_priv__, mask__, tmp__) \ + for (tmp__ = mask__ & INTEL_INFO(dev_priv__)->ring_mask; \ + tmp__ ? (engine__ = &(dev_priv__)->engine[__mask_next_bit(tmp__)]), 1 : 0; ) enum hdmi_force_audio { HDMI_AUDIO_OFF_DVI = -2, /* no aux data for HDMI-DVI converter */ @@ -2549,7 +2587,7 @@ struct drm_i915_cmd_table { BUILD_BUG(); \ __p; \ }) -#define INTEL_INFO(p) (&__I915__(p)->info) +#define INTEL_INFO(p) (&__I915__(p)->info) #define INTEL_GEN(p) (INTEL_INFO(p)->gen) #define INTEL_DEVID(p) (INTEL_INFO(p)->device_id) @@ -2716,10 +2754,10 @@ struct drm_i915_cmd_table { #define HAS_EDRAM(dev) (!!(__I915__(dev)->edram_cap & EDRAM_ENABLED)) #define HAS_WT(dev) ((IS_HASWELL(dev) || IS_BROADWELL(dev)) && \ HAS_EDRAM(dev)) -#define I915_NEED_GFX_HWS(dev) (INTEL_INFO(dev)->need_gfx_hws) +#define HWS_NEEDS_PHYSICAL(dev) (INTEL_INFO(dev)->hws_needs_physical) -#define HAS_HW_CONTEXTS(dev) (INTEL_INFO(dev)->gen >= 6) -#define HAS_LOGICAL_RING_CONTEXTS(dev) (INTEL_INFO(dev)->gen >= 8) +#define HAS_HW_CONTEXTS(dev) (INTEL_INFO(dev)->has_hw_contexts) +#define HAS_LOGICAL_RING_CONTEXTS(dev) (INTEL_INFO(dev)->has_logical_ring_contexts) #define USES_PPGTT(dev) (i915.enable_ppgtt) #define USES_FULL_PPGTT(dev) (i915.enable_ppgtt >= 2) #define USES_FULL_48BIT_PPGTT(dev) (i915.enable_ppgtt == 3) @@ -2743,7 +2781,7 @@ struct drm_i915_cmd_table { * interrupt source and so prevents the other device from working properly. */ #define HAS_AUX_IRQ(dev) (INTEL_INFO(dev)->gen >= 5) -#define HAS_GMBUS_IRQ(dev) (INTEL_INFO(dev)->gen >= 5) +#define HAS_GMBUS_IRQ(dev) (INTEL_INFO(dev)->has_gmbus_irq) /* With the 945 and later, Y tiling got adjusted so that it was 32 128-byte * rows, which changed the alignment requirements and fence programming. @@ -2759,38 +2797,27 @@ struct drm_i915_cmd_table { #define HAS_IPS(dev) (IS_HSW_ULT(dev) || IS_BROADWELL(dev)) -#define HAS_DP_MST(dev) (IS_HASWELL(dev) || IS_BROADWELL(dev) || \ - INTEL_INFO(dev)->gen >= 9) +#define HAS_DP_MST(dev) (INTEL_INFO(dev)->has_dp_mst) #define HAS_DDI(dev) (INTEL_INFO(dev)->has_ddi) #define HAS_FPGA_DBG_UNCLAIMED(dev) (INTEL_INFO(dev)->has_fpga_dbg) -#define HAS_PSR(dev) (IS_HASWELL(dev) || IS_BROADWELL(dev) || \ - IS_VALLEYVIEW(dev) || IS_CHERRYVIEW(dev) || \ - IS_SKYLAKE(dev) || IS_KABYLAKE(dev)) -#define HAS_RUNTIME_PM(dev) (IS_GEN6(dev) || IS_HASWELL(dev) || \ - IS_BROADWELL(dev) || IS_VALLEYVIEW(dev) || \ - IS_CHERRYVIEW(dev) || IS_SKYLAKE(dev) || \ - IS_KABYLAKE(dev) || IS_BROXTON(dev)) -#define HAS_RC6(dev) (INTEL_INFO(dev)->gen >= 6) -#define HAS_RC6p(dev) (IS_GEN6(dev) || IS_IVYBRIDGE(dev)) - -#define HAS_CSR(dev) (IS_GEN9(dev)) +#define HAS_PSR(dev) (INTEL_INFO(dev)->has_psr) +#define HAS_RUNTIME_PM(dev) (INTEL_INFO(dev)->has_runtime_pm) +#define HAS_RC6(dev) (INTEL_INFO(dev)->has_rc6) +#define HAS_RC6p(dev) (INTEL_INFO(dev)->has_rc6p) + +#define HAS_CSR(dev) (INTEL_INFO(dev)->has_csr) /* * For now, anything with a GuC requires uCode loading, and then supports * command submission once loaded. But these are logically independent * properties, so we have separate macros to test them. */ -#define HAS_GUC(dev) (IS_GEN9(dev)) +#define HAS_GUC(dev) (INTEL_INFO(dev)->has_guc) #define HAS_GUC_UCODE(dev) (HAS_GUC(dev)) #define HAS_GUC_SCHED(dev) (HAS_GUC(dev)) -#define HAS_RESOURCE_STREAMER(dev) (IS_HASWELL(dev) || \ - INTEL_INFO(dev)->gen >= 8) - -#define HAS_CORE_RING_FREQ(dev) (INTEL_INFO(dev)->gen >= 6 && \ - !IS_VALLEYVIEW(dev) && !IS_CHERRYVIEW(dev) && \ - !IS_BROXTON(dev)) +#define HAS_RESOURCE_STREAMER(dev) (INTEL_INFO(dev)->has_resource_streamer) #define HAS_POOLED_EU(dev) (INTEL_INFO(dev)->has_pooled_eu) @@ -2818,11 +2845,10 @@ struct drm_i915_cmd_table { #define HAS_PCH_NOP(dev) (INTEL_PCH_TYPE(dev) == PCH_NOP) #define HAS_PCH_SPLIT(dev) (INTEL_PCH_TYPE(dev) != PCH_NONE) -#define HAS_GMCH_DISPLAY(dev) (INTEL_INFO(dev)->gen < 5 || \ - IS_VALLEYVIEW(dev) || IS_CHERRYVIEW(dev)) +#define HAS_GMCH_DISPLAY(dev) (INTEL_INFO(dev)->has_gmch_display) /* DPF == dynamic parity feature */ -#define HAS_L3_DPF(dev) (IS_IVYBRIDGE(dev) || IS_HASWELL(dev)) +#define HAS_L3_DPF(dev) (INTEL_INFO(dev)->has_l3_dpf) #define NUM_L3_SLICES(dev) (IS_HSW_GT3(dev) ? 2 : HAS_L3_DPF(dev)) #define GT_FREQUENCY_MULTIPLIER 50 @@ -2843,7 +2869,7 @@ extern int i915_suspend_switcheroo(struct drm_device *dev, pm_message_t state); extern int i915_resume_switcheroo(struct drm_device *dev); int intel_sanitize_enable_ppgtt(struct drm_i915_private *dev_priv, - int enable_ppgtt); + int enable_ppgtt); bool intel_sanitize_semaphores(struct drm_i915_private *dev_priv, int value); @@ -2861,7 +2887,7 @@ extern long i915_compat_ioctl(struct file *filp, unsigned int cmd, #endif extern int intel_gpu_reset(struct drm_i915_private *dev_priv, u32 engine_mask); extern bool intel_has_gpu_reset(struct drm_i915_private *dev_priv); -extern int i915_reset(struct drm_i915_private *dev_priv); +extern void i915_reset(struct drm_i915_private *dev_priv); extern int intel_guc_reset(struct drm_i915_private *dev_priv); extern void intel_engine_init_hangcheck(struct intel_engine_cs *engine); extern unsigned long i915_chipset_val(struct drm_i915_private *dev_priv); @@ -3197,8 +3223,6 @@ i915_gem_obj_finish_shmem_access(struct drm_i915_gem_object *obj) } int __must_check i915_mutex_lock_interruptible(struct drm_device *dev); -int i915_gem_object_sync(struct drm_i915_gem_object *obj, - struct drm_i915_gem_request *to); void i915_vma_move_to_active(struct i915_vma *vma, struct drm_i915_gem_request *req, unsigned int flags); @@ -3207,6 +3231,7 @@ int i915_gem_dumb_create(struct drm_file *file_priv, struct drm_mode_create_dumb *args); int i915_gem_mmap_gtt(struct drm_file *file_priv, struct drm_device *dev, uint32_t handle, uint64_t *offset); +int i915_gem_mmap_gtt_version(void); void i915_gem_track_fb(struct drm_i915_gem_object *old, struct drm_i915_gem_object *new, @@ -3219,54 +3244,35 @@ i915_gem_find_active_request(struct intel_engine_cs *engine); void i915_gem_retire_requests(struct drm_i915_private *dev_priv); -static inline u32 i915_reset_counter(struct i915_gpu_error *error) -{ - return atomic_read(&error->reset_counter); -} - -static inline bool __i915_reset_in_progress(u32 reset) -{ - return unlikely(reset & I915_RESET_IN_PROGRESS_FLAG); -} - -static inline bool __i915_reset_in_progress_or_wedged(u32 reset) -{ - return unlikely(reset & (I915_RESET_IN_PROGRESS_FLAG | I915_WEDGED)); -} - -static inline bool __i915_terminally_wedged(u32 reset) -{ - return unlikely(reset & I915_WEDGED); -} - static inline bool i915_reset_in_progress(struct i915_gpu_error *error) { - return __i915_reset_in_progress(i915_reset_counter(error)); + return unlikely(test_bit(I915_RESET_IN_PROGRESS, &error->flags)); } -static inline bool i915_reset_in_progress_or_wedged(struct i915_gpu_error *error) +static inline bool i915_terminally_wedged(struct i915_gpu_error *error) { - return __i915_reset_in_progress_or_wedged(i915_reset_counter(error)); + return unlikely(test_bit(I915_WEDGED, &error->flags)); } -static inline bool i915_terminally_wedged(struct i915_gpu_error *error) +static inline bool i915_reset_in_progress_or_wedged(struct i915_gpu_error *error) { - return __i915_terminally_wedged(i915_reset_counter(error)); + return i915_reset_in_progress(error) | i915_terminally_wedged(error); } static inline u32 i915_reset_count(struct i915_gpu_error *error) { - return ((i915_reset_counter(error) & ~I915_WEDGED) + 1) / 2; + return READ_ONCE(error->reset_count); } -void i915_gem_reset(struct drm_device *dev); +void i915_gem_reset(struct drm_i915_private *dev_priv); +void i915_gem_set_wedged(struct drm_i915_private *dev_priv); bool i915_gem_clflush_object(struct drm_i915_gem_object *obj, bool force); int __must_check i915_gem_init(struct drm_device *dev); int __must_check i915_gem_init_hw(struct drm_device *dev); void i915_gem_init_swizzling(struct drm_device *dev); void i915_gem_cleanup_engines(struct drm_device *dev); int __must_check i915_gem_wait_for_idle(struct drm_i915_private *dev_priv, - bool interruptible); + unsigned int flags); int __must_check i915_gem_suspend(struct drm_device *dev); void i915_gem_resume(struct drm_device *dev); int i915_gem_fault(struct vm_area_struct *vma, struct vm_fault *vmf); @@ -3388,7 +3394,6 @@ void i915_gem_object_save_bit_17_swizzle(struct drm_i915_gem_object *obj); int __must_check i915_gem_context_init(struct drm_device *dev); void i915_gem_context_lost(struct drm_i915_private *dev_priv); void i915_gem_context_fini(struct drm_device *dev); -void i915_gem_context_reset(struct drm_device *dev); int i915_gem_context_open(struct drm_device *dev, struct drm_file *file); void i915_gem_context_close(struct drm_device *dev, struct drm_file *file); int i915_switch_context(struct drm_i915_gem_request *req); @@ -3507,13 +3512,13 @@ static inline bool i915_gem_object_needs_bit17_swizzle(struct drm_i915_gem_objec int i915_debugfs_register(struct drm_i915_private *dev_priv); void i915_debugfs_unregister(struct drm_i915_private *dev_priv); int i915_debugfs_connector_add(struct drm_connector *connector); -void intel_display_crc_init(struct drm_device *dev); +void intel_display_crc_init(struct drm_i915_private *dev_priv); #else static inline int i915_debugfs_register(struct drm_i915_private *dev_priv) {return 0;} static inline void i915_debugfs_unregister(struct drm_i915_private *dev_priv) {} static inline int i915_debugfs_connector_add(struct drm_connector *connector) { return 0; } -static inline void intel_display_crc_init(struct drm_device *dev) {} +static inline void intel_display_crc_init(struct drm_i915_private *dev_priv) {} #endif /* i915_gpu_error.c */ @@ -3557,8 +3562,8 @@ extern int i915_save_state(struct drm_device *dev); extern int i915_restore_state(struct drm_device *dev); /* i915_sysfs.c */ -void i915_setup_sysfs(struct drm_device *dev_priv); -void i915_teardown_sysfs(struct drm_device *dev_priv); +void i915_setup_sysfs(struct drm_i915_private *dev_priv); +void i915_teardown_sysfs(struct drm_i915_private *dev_priv); /* intel_i2c.c */ extern int intel_setup_gmbus(struct drm_device *dev); @@ -3735,9 +3740,16 @@ int intel_freq_opcode(struct drm_i915_private *dev_priv, int val); * will be implemented using 2 32-bit writes in an arbitrary order with * an arbitrary delay between them. This can cause the hardware to * act upon the intermediate value, possibly leading to corruption and - * machine death. You have been warned. + * machine death. For this reason we do not support I915_WRITE64, or + * dev_priv->uncore.funcs.mmio_writeq. + * + * When reading a 64-bit value as two 32-bit values, the delay may cause + * the two reads to mismatch, e.g. a timestamp overflowing. Also note that + * occasionally a 64-bit register does not actualy support a full readq + * and must be read using two 32-bit reads. + * + * You have been warned. */ -#define I915_WRITE64(reg, val) dev_priv->uncore.funcs.mmio_writeq(dev_priv, (reg), (val), true) #define I915_READ64(reg) dev_priv->uncore.funcs.mmio_readq(dev_priv, (reg), true) #define I915_READ64_2x32(lower_reg, upper_reg) ({ \ @@ -3780,7 +3792,7 @@ __raw_write(64, q) #undef __raw_write /* These are untraced mmio-accessors that are only valid to be used inside - * criticial sections inside IRQ handlers where forcewake is explicitly + * critical sections inside IRQ handlers where forcewake is explicitly * controlled. * Think twice, and think again, before using these. * Note: Should only be used between intel_uncore_forcewake_irqlock() and @@ -3852,7 +3864,9 @@ wait_remaining_ms_from_jiffies(unsigned long timestamp_jiffies, int to_wait_ms) schedule_timeout_uninterruptible(remaining_jiffies); } } -static inline bool __i915_request_irq_complete(struct drm_i915_gem_request *req) + +static inline bool +__i915_request_irq_complete(struct drm_i915_gem_request *req) { struct intel_engine_cs *engine = req->engine; @@ -3914,17 +3928,6 @@ static inline bool __i915_request_irq_complete(struct drm_i915_gem_request *req) return true; } - /* We need to check whether any gpu reset happened in between - * the request being submitted and now. If a reset has occurred, - * the seqno will have been advance past ours and our request - * is complete. If we are in the process of handling a reset, - * the request is effectively complete as the rendering will - * be discarded, but we need to return in order to drop the - * struct_mutex. - */ - if (i915_reset_in_progress(&req->i915->gpu_error)) - return true; - return false; } diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c index 04607d4115d6..c8bd02277b7d 100644 --- a/drivers/gpu/drm/i915/i915_gem.c +++ b/drivers/gpu/drm/i915/i915_gem.c @@ -386,7 +386,8 @@ __unsafe_wait_rendering(struct drm_i915_gem_object *obj, int ret; ret = i915_gem_active_wait_unlocked(&active[idx], - true, NULL, rps); + I915_WAIT_INTERRUPTIBLE, + NULL, rps); if (ret) return ret; } @@ -1680,6 +1681,56 @@ static unsigned int tile_row_pages(struct drm_i915_gem_object *obj) } /** + * i915_gem_mmap_gtt_version - report the current feature set for GTT mmaps + * + * A history of the GTT mmap interface: + * + * 0 - Everything had to fit into the GTT. Both parties of a memcpy had to + * aligned and suitable for fencing, and still fit into the available + * mappable space left by the pinned display objects. A classic problem + * we called the page-fault-of-doom where we would ping-pong between + * two objects that could not fit inside the GTT and so the memcpy + * would page one object in at the expense of the other between every + * single byte. + * + * 1 - Objects can be any size, and have any compatible fencing (X Y, or none + * as set via i915_gem_set_tiling() [DRM_I915_GEM_SET_TILING]). If the + * object is too large for the available space (or simply too large + * for the mappable aperture!), a view is created instead and faulted + * into userspace. (This view is aligned and sized appropriately for + * fenced access.) + * + * Restrictions: + * + * * snoopable objects cannot be accessed via the GTT. It can cause machine + * hangs on some architectures, corruption on others. An attempt to service + * a GTT page fault from a snoopable object will generate a SIGBUS. + * + * * the object must be able to fit into RAM (physical memory, though no + * limited to the mappable aperture). + * + * + * Caveats: + * + * * a new GTT page fault will synchronize rendering from the GPU and flush + * all data to system memory. Subsequent access will not be synchronized. + * + * * all mappings are revoked on runtime device suspend. + * + * * there are only 8, 16 or 32 fence registers to share between all users + * (older machines require fence register for display and blitter access + * as well). Contention of the fence registers will cause the previous users + * to be unmapped and any new access will generate new page faults. + * + * * running out of memory while servicing a fault may generate a SIGBUS, + * rather than the expected SIGSEGV. + */ +int i915_gem_mmap_gtt_version(void) +{ + return 1; +} + +/** * i915_gem_fault - fault a page into the GTT * @area: CPU VMA in question * @vmf: fault info @@ -1694,6 +1745,9 @@ static unsigned int tile_row_pages(struct drm_i915_gem_object *obj) * from the GTT and/or fence registers to make room. So performance may * suffer if the GTT working set is large or there are few fence registers * left. + * + * The current feature set supported by i915_gem_fault() and thus GTT mmaps + * is exposed via I915_PARAM_MMAP_GTT_VERSION (see i915_gem_mmap_gtt_version). */ int i915_gem_fault(struct vm_area_struct *area, struct vm_fault *vmf) { @@ -1973,7 +2027,7 @@ static int i915_gem_object_create_mmap_offset(struct drm_i915_gem_object *obj) * to claim that space for ourselves, we need to take the big * struct_mutex to free the requests+objects and allocate our slot. */ - err = i915_gem_wait_for_idle(dev_priv, true); + err = i915_gem_wait_for_idle(dev_priv, I915_WAIT_INTERRUPTIBLE); if (err) return err; @@ -2495,32 +2549,94 @@ i915_gem_find_active_request(struct intel_engine_cs *engine) if (i915_gem_request_completed(request)) continue; + if (!i915_sw_fence_done(&request->submit)) + break; + return request; } return NULL; } -static void i915_gem_reset_engine_status(struct intel_engine_cs *engine) +static void reset_request(struct drm_i915_gem_request *request) +{ + void *vaddr = request->ring->vaddr; + u32 head; + + /* As this request likely depends on state from the lost + * context, clear out all the user operations leaving the + * breadcrumb at the end (so we get the fence notifications). + */ + head = request->head; + if (request->postfix < head) { + memset(vaddr + head, 0, request->ring->size - head); + head = 0; + } + memset(vaddr + head, 0, request->postfix - head); +} + +static void i915_gem_reset_engine(struct intel_engine_cs *engine) { struct drm_i915_gem_request *request; + struct i915_gem_context *incomplete_ctx; bool ring_hung; + /* Ensure irq handler finishes, and not run again. */ + tasklet_kill(&engine->irq_tasklet); + if (engine->irq_seqno_barrier) + engine->irq_seqno_barrier(engine); + request = i915_gem_find_active_request(engine); - if (request == NULL) + if (!request) return; ring_hung = engine->hangcheck.score >= HANGCHECK_SCORE_RING_HUNG; - i915_set_reset_status(request->ctx, ring_hung); + if (!ring_hung) + return; + + DRM_DEBUG_DRIVER("resetting %s to restart from tail of request 0x%x\n", + engine->name, request->fence.seqno); + + /* Setup the CS to resume from the breadcrumb of the hung request */ + engine->reset_hw(engine, request); + + /* Users of the default context do not rely on logical state + * preserved between batches. They have to emit full state on + * every batch and so it is safe to execute queued requests following + * the hang. + * + * Other contexts preserve state, now corrupt. We want to skip all + * queued requests that reference the corrupt context. + */ + incomplete_ctx = request->ctx; + if (i915_gem_context_is_default(incomplete_ctx)) + return; + list_for_each_entry_continue(request, &engine->request_list, link) - i915_set_reset_status(request->ctx, false); + if (request->ctx == incomplete_ctx) + reset_request(request); } -static void i915_gem_reset_engine_cleanup(struct intel_engine_cs *engine) +void i915_gem_reset(struct drm_i915_private *dev_priv) { - struct drm_i915_gem_request *request; - struct intel_ring *ring; + struct intel_engine_cs *engine; + + i915_gem_retire_requests(dev_priv); + + for_each_engine(engine, dev_priv) + i915_gem_reset_engine(engine); + + i915_gem_restore_fences(&dev_priv->drm); +} + +static void nop_submit_request(struct drm_i915_gem_request *request) +{ +} + +static void i915_gem_cleanup_engine(struct intel_engine_cs *engine) +{ + engine->submit_request = nop_submit_request; /* Mark all pending requests as complete so that any concurrent * (lockless) lookup doesn't try and wait upon the request as we @@ -2535,60 +2651,30 @@ static void i915_gem_reset_engine_cleanup(struct intel_engine_cs *engine) */ if (i915.enable_execlists) { - /* Ensure irq handler finishes or is cancelled. */ - tasklet_kill(&engine->irq_tasklet); - - intel_execlists_cancel_requests(engine); - } - - /* - * We must free the requests after all the corresponding objects have - * been moved off active lists. Which is the same order as the normal - * retire_requests function does. This is important if object hold - * implicit references on things like e.g. ppgtt address spaces through - * the request. - */ - request = i915_gem_active_raw(&engine->last_request, - &engine->i915->drm.struct_mutex); - if (request) - i915_gem_request_retire_upto(request); - GEM_BUG_ON(intel_engine_is_active(engine)); - - /* Having flushed all requests from all queues, we know that all - * ringbuffers must now be empty. However, since we do not reclaim - * all space when retiring the request (to prevent HEADs colliding - * with rapid ringbuffer wraparound) the amount of available space - * upon reset is less than when we start. Do one more pass over - * all the ringbuffers to reset last_retired_head. - */ - list_for_each_entry(ring, &engine->buffers, link) { - ring->last_retired_head = ring->tail; - intel_ring_update_space(ring); + spin_lock(&engine->execlist_lock); + INIT_LIST_HEAD(&engine->execlist_queue); + i915_gem_request_put(engine->execlist_port[0].request); + i915_gem_request_put(engine->execlist_port[1].request); + memset(engine->execlist_port, 0, sizeof(engine->execlist_port)); + spin_unlock(&engine->execlist_lock); } engine->i915->gt.active_engines &= ~intel_engine_flag(engine); } -void i915_gem_reset(struct drm_device *dev) +void i915_gem_set_wedged(struct drm_i915_private *dev_priv) { - struct drm_i915_private *dev_priv = to_i915(dev); struct intel_engine_cs *engine; - /* - * Before we free the objects from the requests, we need to inspect - * them for finding the guilty party. As the requests only borrow - * their reference to the objects, the inspection must be done first. - */ - for_each_engine(engine, dev_priv) - i915_gem_reset_engine_status(engine); + lockdep_assert_held(&dev_priv->drm.struct_mutex); + set_bit(I915_WEDGED, &dev_priv->gpu_error.flags); + i915_gem_context_lost(dev_priv); for_each_engine(engine, dev_priv) - i915_gem_reset_engine_cleanup(engine); + i915_gem_cleanup_engine(engine); mod_delayed_work(dev_priv->wq, &dev_priv->gt.idle_work, 0); - i915_gem_context_reset(dev); - - i915_gem_restore_fences(dev); + i915_gem_retire_requests(dev_priv); } static void @@ -2721,7 +2807,8 @@ i915_gem_wait_ioctl(struct drm_device *dev, void *data, struct drm_file *file) active = __I915_BO_ACTIVE(obj); for_each_active(active, idx) { s64 *timeout = args->timeout_ns >= 0 ? &args->timeout_ns : NULL; - ret = i915_gem_active_wait_unlocked(&obj->last_read[idx], true, + ret = i915_gem_active_wait_unlocked(&obj->last_read[idx], + I915_WAIT_INTERRUPTIBLE, timeout, rps); if (ret) break; @@ -2731,96 +2818,6 @@ i915_gem_wait_ioctl(struct drm_device *dev, void *data, struct drm_file *file) return ret; } -static int -__i915_gem_object_sync(struct drm_i915_gem_request *to, - struct drm_i915_gem_request *from) -{ - int ret; - - if (to->engine == from->engine) - return 0; - - if (!i915.semaphores) { - ret = i915_wait_request(from, - from->i915->mm.interruptible, - NULL, - NO_WAITBOOST); - if (ret) - return ret; - } else { - int idx = intel_engine_sync_index(from->engine, to->engine); - if (from->fence.seqno <= from->engine->semaphore.sync_seqno[idx]) - return 0; - - trace_i915_gem_ring_sync_to(to, from); - ret = to->engine->semaphore.sync_to(to, from); - if (ret) - return ret; - - from->engine->semaphore.sync_seqno[idx] = from->fence.seqno; - } - - return 0; -} - -/** - * i915_gem_object_sync - sync an object to a ring. - * - * @obj: object which may be in use on another ring. - * @to: request we are wishing to use - * - * This code is meant to abstract object synchronization with the GPU. - * Conceptually we serialise writes between engines inside the GPU. - * We only allow one engine to write into a buffer at any time, but - * multiple readers. To ensure each has a coherent view of memory, we must: - * - * - If there is an outstanding write request to the object, the new - * request must wait for it to complete (either CPU or in hw, requests - * on the same ring will be naturally ordered). - * - * - If we are a write request (pending_write_domain is set), the new - * request must wait for outstanding read requests to complete. - * - * Returns 0 if successful, else propagates up the lower layer error. - */ -int -i915_gem_object_sync(struct drm_i915_gem_object *obj, - struct drm_i915_gem_request *to) -{ - struct i915_gem_active *active; - unsigned long active_mask; - int idx; - - lockdep_assert_held(&obj->base.dev->struct_mutex); - - active_mask = i915_gem_object_get_active(obj); - if (!active_mask) - return 0; - - if (obj->base.pending_write_domain) { - active = obj->last_read; - } else { - active_mask = 1; - active = &obj->last_write; - } - - for_each_active(active_mask, idx) { - struct drm_i915_gem_request *request; - int ret; - - request = i915_gem_active_peek(&active[idx], - &obj->base.dev->struct_mutex); - if (!request) - continue; - - ret = __i915_gem_object_sync(to, request); - if (ret) - return ret; - } - - return 0; -} - static void __i915_vma_iounmap(struct i915_vma *vma) { GEM_BUG_ON(i915_vma_is_pinned(vma)); @@ -2924,7 +2921,7 @@ destroy: } int i915_gem_wait_for_idle(struct drm_i915_private *dev_priv, - bool interruptible) + unsigned int flags) { struct intel_engine_cs *engine; int ret; @@ -2933,7 +2930,7 @@ int i915_gem_wait_for_idle(struct drm_i915_private *dev_priv, if (engine->last_context == NULL) continue; - ret = intel_engine_idle(engine, interruptible); + ret = intel_engine_idle(engine, flags); if (ret) return ret; } @@ -3688,7 +3685,7 @@ i915_gem_ring_throttle(struct drm_device *dev, struct drm_file *file) if (target == NULL) return 0; - ret = i915_wait_request(target, true, NULL, NULL); + ret = i915_wait_request(target, I915_WAIT_INTERRUPTIBLE, NULL, NULL); i915_gem_request_put(target); return ret; @@ -4244,7 +4241,9 @@ int i915_gem_suspend(struct drm_device *dev) if (ret) goto err; - ret = i915_gem_wait_for_idle(dev_priv, true); + ret = i915_gem_wait_for_idle(dev_priv, + I915_WAIT_INTERRUPTIBLE | + I915_WAIT_LOCKED); if (ret) goto err; @@ -4280,8 +4279,7 @@ void i915_gem_resume(struct drm_device *dev) * guarantee that the context image is complete. So let's just reset * it and start again. */ - if (i915.enable_execlists) - intel_lr_context_reset(dev_priv, dev_priv->kernel_context); + dev_priv->gt.resume(dev_priv); mutex_unlock(&dev->struct_mutex); } @@ -4433,8 +4431,10 @@ int i915_gem_init(struct drm_device *dev) mutex_lock(&dev->struct_mutex); if (!i915.enable_execlists) { + dev_priv->gt.resume = intel_legacy_submission_resume; dev_priv->gt.cleanup_engine = intel_engine_cleanup; } else { + dev_priv->gt.resume = intel_lr_context_resume; dev_priv->gt.cleanup_engine = intel_logical_ring_cleanup; } @@ -4467,7 +4467,7 @@ int i915_gem_init(struct drm_device *dev) * for all other failure, such as an allocation failure, bail. */ DRM_ERROR("Failed to initialize GPU, declaring it wedged\n"); - atomic_or(I915_WEDGED, &dev_priv->gpu_error.reset_counter); + i915_gem_set_wedged(dev_priv); ret = 0; } @@ -4569,6 +4569,8 @@ i915_gem_load_init(struct drm_device *dev) dev_priv->mm.interruptible = true; + atomic_set(&dev_priv->mm.bsd_engine_dispatch_index, 0); + spin_lock_init(&dev_priv->fb_tracking.lock); } @@ -4587,6 +4589,11 @@ void i915_gem_load_cleanup(struct drm_device *dev) int i915_gem_freeze_late(struct drm_i915_private *dev_priv) { struct drm_i915_gem_object *obj; + struct list_head *phases[] = { + &dev_priv->mm.unbound_list, + &dev_priv->mm.bound_list, + NULL + }, **p; /* Called just before we write the hibernation image. * @@ -4597,16 +4604,18 @@ int i915_gem_freeze_late(struct drm_i915_private *dev_priv) * * To make sure the hibernation image contains the latest state, * we update that state just before writing out the image. + * + * To try and reduce the hibernation image, we manually shrink + * the objects as well. */ - list_for_each_entry(obj, &dev_priv->mm.unbound_list, global_list) { - obj->base.read_domains = I915_GEM_DOMAIN_CPU; - obj->base.write_domain = I915_GEM_DOMAIN_CPU; - } + i915_gem_shrink_all(dev_priv); - list_for_each_entry(obj, &dev_priv->mm.bound_list, global_list) { - obj->base.read_domains = I915_GEM_DOMAIN_CPU; - obj->base.write_domain = I915_GEM_DOMAIN_CPU; + for (p = phases; *p; p++) { + list_for_each_entry(obj, *p, global_list) { + obj->base.read_domains = I915_GEM_DOMAIN_CPU; + obj->base.write_domain = I915_GEM_DOMAIN_CPU; + } } return 0; diff --git a/drivers/gpu/drm/i915/i915_gem_context.c b/drivers/gpu/drm/i915/i915_gem_context.c index 35950ee46a1d..df10f4e95736 100644 --- a/drivers/gpu/drm/i915/i915_gem_context.c +++ b/drivers/gpu/drm/i915/i915_gem_context.c @@ -420,22 +420,6 @@ static void i915_gem_context_unpin(struct i915_gem_context *ctx, } } -void i915_gem_context_reset(struct drm_device *dev) -{ - struct drm_i915_private *dev_priv = to_i915(dev); - - lockdep_assert_held(&dev->struct_mutex); - - if (i915.enable_execlists) { - struct i915_gem_context *ctx; - - list_for_each_entry(ctx, &dev_priv->context_list, link) - intel_lr_context_reset(dev_priv, ctx); - } - - i915_gem_context_lost(dev_priv); -} - int i915_gem_context_init(struct drm_device *dev) { struct drm_i915_private *dev_priv = to_i915(dev); diff --git a/drivers/gpu/drm/i915/i915_gem_evict.c b/drivers/gpu/drm/i915/i915_gem_evict.c index 815d5fbe07ac..5b6f81c1dbca 100644 --- a/drivers/gpu/drm/i915/i915_gem_evict.c +++ b/drivers/gpu/drm/i915/i915_gem_evict.c @@ -170,7 +170,9 @@ search_again: if (ret) return ret; - ret = i915_gem_wait_for_idle(dev_priv, true); + ret = i915_gem_wait_for_idle(dev_priv, + I915_WAIT_INTERRUPTIBLE | + I915_WAIT_LOCKED); if (ret) return ret; @@ -275,7 +277,9 @@ int i915_gem_evict_vm(struct i915_address_space *vm, bool do_idle) return ret; } - ret = i915_gem_wait_for_idle(dev_priv, true); + ret = i915_gem_wait_for_idle(dev_priv, + I915_WAIT_INTERRUPTIBLE | + I915_WAIT_LOCKED); if (ret) return ret; diff --git a/drivers/gpu/drm/i915/i915_gem_execbuffer.c b/drivers/gpu/drm/i915/i915_gem_execbuffer.c index 601156c353cc..33c85227643d 100644 --- a/drivers/gpu/drm/i915/i915_gem_execbuffer.c +++ b/drivers/gpu/drm/i915/i915_gem_execbuffer.c @@ -1131,13 +1131,25 @@ i915_gem_execbuffer_move_to_gpu(struct drm_i915_gem_request *req, list_for_each_entry(vma, vmas, exec_list) { struct drm_i915_gem_object *obj = vma->obj; + struct reservation_object *resv; if (obj->flags & other_rings) { - ret = i915_gem_object_sync(obj, req); + ret = i915_gem_request_await_object + (req, obj, obj->base.pending_write_domain); if (ret) return ret; } + resv = i915_gem_object_get_dmabuf_resv(obj); + if (resv) { + ret = i915_sw_fence_await_reservation + (&req->submit, resv, &i915_fence_ops, + obj->base.pending_write_domain, 10*HZ, + GFP_KERNEL | __GFP_NOWARN); + if (ret < 0) + return ret; + } + if (obj->base.write_domain & I915_GEM_DOMAIN_CPU) i915_gem_clflush_object(obj, false); } @@ -1253,12 +1265,9 @@ static struct i915_gem_context * i915_gem_validate_context(struct drm_device *dev, struct drm_file *file, struct intel_engine_cs *engine, const u32 ctx_id) { - struct i915_gem_context *ctx = NULL; + struct i915_gem_context *ctx; struct i915_ctx_hang_stats *hs; - if (engine->id != RCS && ctx_id != DEFAULT_CONTEXT_HANDLE) - return ERR_PTR(-EINVAL); - ctx = i915_gem_context_lookup(file->driver_priv, ctx_id); if (IS_ERR(ctx)) return ctx; @@ -1538,13 +1547,9 @@ gen8_dispatch_bsd_engine(struct drm_i915_private *dev_priv, struct drm_i915_file_private *file_priv = file->driver_priv; /* Check whether the file_priv has already selected one ring. */ - if ((int)file_priv->bsd_engine < 0) { - /* If not, use the ping-pong mechanism to select one. */ - mutex_lock(&dev_priv->drm.struct_mutex); - file_priv->bsd_engine = dev_priv->mm.bsd_engine_dispatch_index; - dev_priv->mm.bsd_engine_dispatch_index ^= 1; - mutex_unlock(&dev_priv->drm.struct_mutex); - } + if ((int)file_priv->bsd_engine < 0) + file_priv->bsd_engine = atomic_fetch_xor(1, + &dev_priv->mm.bsd_engine_dispatch_index); return file_priv->bsd_engine; } diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.c b/drivers/gpu/drm/i915/i915_gem_gtt.c index b90fdcee992a..0bb4232f66bc 100644 --- a/drivers/gpu/drm/i915/i915_gem_gtt.c +++ b/drivers/gpu/drm/i915/i915_gem_gtt.c @@ -32,6 +32,8 @@ #include "i915_trace.h" #include "intel_drv.h" +#define I915_GFP_DMA (GFP_KERNEL | __GFP_HIGHMEM) + /** * DOC: Global GTT views * @@ -122,8 +124,11 @@ int intel_sanitize_enable_ppgtt(struct drm_i915_private *dev_priv, has_full_48bit_ppgtt = IS_BROADWELL(dev_priv) || INTEL_GEN(dev_priv) >= 9; - if (intel_vgpu_active(dev_priv)) - has_full_ppgtt = false; /* emulation is too hard */ + if (intel_vgpu_active(dev_priv)) { + /* emulation is too hard */ + has_full_ppgtt = false; + has_full_48bit_ppgtt = false; + } if (!has_aliasing_ppgtt) return 0; @@ -158,7 +163,7 @@ int intel_sanitize_enable_ppgtt(struct drm_i915_private *dev_priv, return 0; } - if (INTEL_GEN(dev_priv) >= 8 && i915.enable_execlists) + if (INTEL_GEN(dev_priv) >= 8 && i915.enable_execlists && has_full_ppgtt) return has_full_48bit_ppgtt ? 3 : 2; else return has_aliasing_ppgtt ? 1 : 0; @@ -326,16 +331,16 @@ static gen6_pte_t iris_pte_encode(dma_addr_t addr, static int __setup_page_dma(struct drm_device *dev, struct i915_page_dma *p, gfp_t flags) { - struct device *device = &dev->pdev->dev; + struct device *kdev = &dev->pdev->dev; p->page = alloc_page(flags); if (!p->page) return -ENOMEM; - p->daddr = dma_map_page(device, + p->daddr = dma_map_page(kdev, p->page, 0, 4096, PCI_DMA_BIDIRECTIONAL); - if (dma_mapping_error(device, p->daddr)) { + if (dma_mapping_error(kdev, p->daddr)) { __free_page(p->page); return -EINVAL; } @@ -345,15 +350,17 @@ static int __setup_page_dma(struct drm_device *dev, static int setup_page_dma(struct drm_device *dev, struct i915_page_dma *p) { - return __setup_page_dma(dev, p, GFP_KERNEL); + return __setup_page_dma(dev, p, I915_GFP_DMA); } static void cleanup_page_dma(struct drm_device *dev, struct i915_page_dma *p) { + struct pci_dev *pdev = dev->pdev; + if (WARN_ON(!p->page)) return; - dma_unmap_page(&dev->pdev->dev, p->daddr, 4096, PCI_DMA_BIDIRECTIONAL); + dma_unmap_page(&pdev->dev, p->daddr, 4096, PCI_DMA_BIDIRECTIONAL); __free_page(p->page); memset(p, 0, sizeof(*p)); } @@ -407,33 +414,18 @@ static void fill_page_dma_32(struct drm_device *dev, struct i915_page_dma *p, fill_page_dma(dev, p, v); } -static struct i915_page_scratch *alloc_scratch_page(struct drm_device *dev) +static int +setup_scratch_page(struct drm_device *dev, + struct i915_page_dma *scratch, + gfp_t gfp) { - struct i915_page_scratch *sp; - int ret; - - sp = kzalloc(sizeof(*sp), GFP_KERNEL); - if (sp == NULL) - return ERR_PTR(-ENOMEM); - - ret = __setup_page_dma(dev, px_base(sp), GFP_DMA32 | __GFP_ZERO); - if (ret) { - kfree(sp); - return ERR_PTR(ret); - } - - set_pages_uc(px_page(sp), 1); - - return sp; + return __setup_page_dma(dev, scratch, gfp | __GFP_ZERO); } -static void free_scratch_page(struct drm_device *dev, - struct i915_page_scratch *sp) +static void cleanup_scratch_page(struct drm_device *dev, + struct i915_page_dma *scratch) { - set_pages_wb(px_page(sp), 1); - - cleanup_px(dev, sp); - kfree(sp); + cleanup_page_dma(dev, scratch); } static struct i915_page_table *alloc_pt(struct drm_device *dev) @@ -479,7 +471,7 @@ static void gen8_initialize_pt(struct i915_address_space *vm, { gen8_pte_t scratch_pte; - scratch_pte = gen8_pte_encode(px_dma(vm->scratch_page), + scratch_pte = gen8_pte_encode(vm->scratch_page.daddr, I915_CACHE_LLC, true); fill_px(vm->dev, pt, scratch_pte); @@ -490,9 +482,9 @@ static void gen6_initialize_pt(struct i915_address_space *vm, { gen6_pte_t scratch_pte; - WARN_ON(px_dma(vm->scratch_page) == 0); + WARN_ON(vm->scratch_page.daddr == 0); - scratch_pte = vm->pte_encode(px_dma(vm->scratch_page), + scratch_pte = vm->pte_encode(vm->scratch_page.daddr, I915_CACHE_LLC, true, 0); fill32_px(vm->dev, pt, scratch_pte); @@ -776,7 +768,7 @@ static void gen8_ppgtt_clear_range(struct i915_address_space *vm, bool use_scratch) { struct i915_hw_ppgtt *ppgtt = i915_vm_to_ppgtt(vm); - gen8_pte_t scratch_pte = gen8_pte_encode(px_dma(vm->scratch_page), + gen8_pte_t scratch_pte = gen8_pte_encode(vm->scratch_page.daddr, I915_CACHE_LLC, use_scratch); if (!USES_FULL_48BIT_PPGTT(vm->dev)) { @@ -882,9 +874,9 @@ static int gen8_init_scratch(struct i915_address_space *vm) struct drm_device *dev = vm->dev; int ret; - vm->scratch_page = alloc_scratch_page(dev); - if (IS_ERR(vm->scratch_page)) - return PTR_ERR(vm->scratch_page); + ret = setup_scratch_page(dev, &vm->scratch_page, I915_GFP_DMA); + if (ret) + return ret; vm->scratch_pt = alloc_pt(dev); if (IS_ERR(vm->scratch_pt)) { @@ -918,7 +910,7 @@ free_pd: free_pt: free_pt(dev, vm->scratch_pt); free_scratch_page: - free_scratch_page(dev, vm->scratch_page); + cleanup_scratch_page(dev, &vm->scratch_page); return ret; } @@ -962,7 +954,7 @@ static void gen8_free_scratch(struct i915_address_space *vm) free_pdp(dev, vm->scratch_pdp); free_pd(dev, vm->scratch_pd); free_pt(dev, vm->scratch_pt); - free_scratch_page(dev, vm->scratch_page); + cleanup_scratch_page(dev, &vm->scratch_page); } static void gen8_ppgtt_cleanup_3lvl(struct drm_device *dev, @@ -1459,7 +1451,7 @@ static void gen8_dump_ppgtt(struct i915_hw_ppgtt *ppgtt, struct seq_file *m) struct i915_address_space *vm = &ppgtt->base; uint64_t start = ppgtt->base.start; uint64_t length = ppgtt->base.total; - gen8_pte_t scratch_pte = gen8_pte_encode(px_dma(vm->scratch_page), + gen8_pte_t scratch_pte = gen8_pte_encode(vm->scratch_page.daddr, I915_CACHE_LLC, true); if (!USES_FULL_48BIT_PPGTT(vm->dev)) { @@ -1576,7 +1568,7 @@ static void gen6_dump_ppgtt(struct i915_hw_ppgtt *ppgtt, struct seq_file *m) uint32_t pte, pde; uint32_t start = ppgtt->base.start, length = ppgtt->base.total; - scratch_pte = vm->pte_encode(px_dma(vm->scratch_page), + scratch_pte = vm->pte_encode(vm->scratch_page.daddr, I915_CACHE_LLC, true, 0); gen6_for_each_pde(unused, &ppgtt->pd, start, length, pde) { @@ -1801,7 +1793,7 @@ static void gen6_ppgtt_clear_range(struct i915_address_space *vm, unsigned first_pte = first_entry % GEN6_PTES; unsigned last_pte, i; - scratch_pte = vm->pte_encode(px_dma(vm->scratch_page), + scratch_pte = vm->pte_encode(vm->scratch_page.daddr, I915_CACHE_LLC, true, 0); while (num_entries) { @@ -1947,14 +1939,15 @@ unwind_out: static int gen6_init_scratch(struct i915_address_space *vm) { struct drm_device *dev = vm->dev; + int ret; - vm->scratch_page = alloc_scratch_page(dev); - if (IS_ERR(vm->scratch_page)) - return PTR_ERR(vm->scratch_page); + ret = setup_scratch_page(dev, &vm->scratch_page, I915_GFP_DMA); + if (ret) + return ret; vm->scratch_pt = alloc_pt(dev); if (IS_ERR(vm->scratch_pt)) { - free_scratch_page(dev, vm->scratch_page); + cleanup_scratch_page(dev, &vm->scratch_page); return PTR_ERR(vm->scratch_pt); } @@ -1968,7 +1961,7 @@ static void gen6_free_scratch(struct i915_address_space *vm) struct drm_device *dev = vm->dev; free_pt(dev, vm->scratch_pt); - free_scratch_page(dev, vm->scratch_page); + cleanup_scratch_page(dev, &vm->scratch_page); } static void gen6_ppgtt_cleanup(struct i915_address_space *vm) @@ -2311,12 +2304,7 @@ int i915_gem_gtt_prepare_object(struct drm_i915_gem_object *obj) static void gen8_set_pte(void __iomem *addr, gen8_pte_t pte) { -#ifdef writeq writeq(pte, addr); -#else - iowrite32((u32)pte, addr); - iowrite32(pte >> 32, addr + 4); -#endif } static void gen8_ggtt_insert_page(struct i915_address_space *vm, @@ -2509,7 +2497,7 @@ static void gen8_ggtt_clear_range(struct i915_address_space *vm, first_entry, num_entries, max_entries)) num_entries = max_entries; - scratch_pte = gen8_pte_encode(px_dma(vm->scratch_page), + scratch_pte = gen8_pte_encode(vm->scratch_page.daddr, I915_CACHE_LLC, use_scratch); for (i = 0; i < num_entries; i++) @@ -2541,7 +2529,7 @@ static void gen6_ggtt_clear_range(struct i915_address_space *vm, first_entry, num_entries, max_entries)) num_entries = max_entries; - scratch_pte = vm->pte_encode(px_dma(vm->scratch_page), + scratch_pte = vm->pte_encode(vm->scratch_page.daddr, I915_CACHE_LLC, use_scratch, 0); for (i = 0; i < num_entries; i++) @@ -2685,19 +2673,19 @@ static void ggtt_unbind_vma(struct i915_vma *vma) void i915_gem_gtt_finish_object(struct drm_i915_gem_object *obj) { - struct drm_device *dev = obj->base.dev; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = to_i915(obj->base.dev); + struct device *kdev = &dev_priv->drm.pdev->dev; struct i915_ggtt *ggtt = &dev_priv->ggtt; if (unlikely(ggtt->do_idle_maps)) { - if (i915_gem_wait_for_idle(dev_priv, false)) { + if (i915_gem_wait_for_idle(dev_priv, I915_WAIT_LOCKED)) { DRM_ERROR("Failed to wait for idle; VT'd may hang.\n"); /* Wait a bit, in hopes it avoids the hang */ udelay(10); } } - dma_unmap_sg(&dev->pdev->dev, obj->pages->sgl, obj->pages->nents, + dma_unmap_sg(kdev, obj->pages->sgl, obj->pages->nents, PCI_DMA_BIDIRECTIONAL); } @@ -2894,8 +2882,8 @@ static size_t gen9_get_stolen_size(u16 gen9_gmch_ctl) static int ggtt_probe_common(struct i915_ggtt *ggtt, u64 size) { struct pci_dev *pdev = ggtt->base.dev->pdev; - struct i915_page_scratch *scratch_page; phys_addr_t phys_addr; + int ret; /* For Modern GENs the PTEs and register space are split in the BAR */ phys_addr = pci_resource_start(pdev, 0) + pci_resource_len(pdev, 0) / 2; @@ -2916,16 +2904,16 @@ static int ggtt_probe_common(struct i915_ggtt *ggtt, u64 size) return -ENOMEM; } - scratch_page = alloc_scratch_page(ggtt->base.dev); - if (IS_ERR(scratch_page)) { + ret = setup_scratch_page(ggtt->base.dev, + &ggtt->base.scratch_page, + GFP_DMA32); + if (ret) { DRM_ERROR("Scratch setup failed\n"); /* iounmap will also get called at remove, but meh */ iounmap(ggtt->gsm); - return PTR_ERR(scratch_page); + return ret; } - ggtt->base.scratch_page = scratch_page; - return 0; } @@ -3007,7 +2995,7 @@ static void gen6_gmch_remove(struct i915_address_space *vm) struct i915_ggtt *ggtt = i915_vm_to_ggtt(vm); iounmap(ggtt->gsm); - free_scratch_page(vm->dev, vm->scratch_page); + cleanup_scratch_page(vm->dev, &vm->scratch_page); } static int gen8_gmch_probe(struct i915_ggtt *ggtt) @@ -3244,8 +3232,7 @@ void i915_gem_restore_gtt_mappings(struct drm_device *dev) { struct drm_i915_private *dev_priv = to_i915(dev); struct i915_ggtt *ggtt = &dev_priv->ggtt; - struct drm_i915_gem_object *obj; - struct i915_vma *vma; + struct drm_i915_gem_object *obj, *on; i915_check_and_clear_faults(dev_priv); @@ -3253,20 +3240,32 @@ void i915_gem_restore_gtt_mappings(struct drm_device *dev) ggtt->base.clear_range(&ggtt->base, ggtt->base.start, ggtt->base.total, true); - /* Cache flush objects bound into GGTT and rebind them. */ - list_for_each_entry(obj, &dev_priv->mm.bound_list, global_list) { + ggtt->base.closed = true; /* skip rewriting PTE on VMA unbind */ + + /* clflush objects bound into the GGTT and rebind them. */ + list_for_each_entry_safe(obj, on, + &dev_priv->mm.bound_list, global_list) { + bool ggtt_bound = false; + struct i915_vma *vma; + list_for_each_entry(vma, &obj->vma_list, obj_link) { if (vma->vm != &ggtt->base) continue; + if (!i915_vma_unbind(vma)) + continue; + WARN_ON(i915_vma_bind(vma, obj->cache_level, PIN_UPDATE)); + ggtt_bound = true; } - if (obj->pin_display) + if (ggtt_bound) WARN_ON(i915_gem_object_set_to_gtt_domain(obj, false)); } + ggtt->base.closed = false; + if (INTEL_INFO(dev)->gen >= 8) { if (IS_CHERRYVIEW(dev) || IS_BROXTON(dev)) chv_setup_private_ppat(dev_priv); diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.h b/drivers/gpu/drm/i915/i915_gem_gtt.h index a9aec25535ac..ec78be2f8c77 100644 --- a/drivers/gpu/drm/i915/i915_gem_gtt.h +++ b/drivers/gpu/drm/i915/i915_gem_gtt.h @@ -312,10 +312,6 @@ struct i915_page_dma { #define px_page(px) (px_base(px)->page) #define px_dma(px) (px_base(px)->daddr) -struct i915_page_scratch { - struct i915_page_dma base; -}; - struct i915_page_table { struct i915_page_dma base; @@ -361,7 +357,7 @@ struct i915_address_space { bool closed; - struct i915_page_scratch *scratch_page; + struct i915_page_dma scratch_page; struct i915_page_table *scratch_pt; struct i915_page_directory *scratch_pd; struct i915_page_directory_pointer *scratch_pdp; /* GEN8+ & 48b PPGTT */ diff --git a/drivers/gpu/drm/i915/i915_gem_request.c b/drivers/gpu/drm/i915/i915_gem_request.c index 1a215320cefb..40978bc12ceb 100644 --- a/drivers/gpu/drm/i915/i915_gem_request.c +++ b/drivers/gpu/drm/i915/i915_gem_request.c @@ -233,16 +233,18 @@ void i915_gem_request_retire_upto(struct drm_i915_gem_request *req) } while (tmp != req); } -static int i915_gem_check_wedge(unsigned int reset_counter, bool interruptible) +static int i915_gem_check_wedge(struct drm_i915_private *dev_priv) { - if (__i915_terminally_wedged(reset_counter)) + struct i915_gpu_error *error = &dev_priv->gpu_error; + + if (i915_terminally_wedged(error)) return -EIO; - if (__i915_reset_in_progress(reset_counter)) { + if (i915_reset_in_progress(error)) { /* Non-interruptible callers can't handle -EAGAIN, hence return * -EIO unconditionally for these. */ - if (!interruptible) + if (!dev_priv->mm.interruptible) return -EIO; return -EAGAIN; @@ -258,7 +260,9 @@ static int i915_gem_init_seqno(struct drm_i915_private *dev_priv, u32 seqno) /* Carefully retire all requests without writing to the rings */ for_each_engine(engine, dev_priv) { - ret = intel_engine_idle(engine, true); + ret = intel_engine_idle(engine, + I915_WAIT_INTERRUPTIBLE | + I915_WAIT_LOCKED); if (ret) return ret; } @@ -314,6 +318,26 @@ static int i915_gem_get_seqno(struct drm_i915_private *dev_priv, u32 *seqno) return 0; } +static int __i915_sw_fence_call +submit_notify(struct i915_sw_fence *fence, enum i915_sw_fence_notify state) +{ + struct drm_i915_gem_request *request = + container_of(fence, typeof(*request), submit); + + /* Will be called from irq-context when using foreign DMA fences */ + + switch (state) { + case FENCE_COMPLETE: + request->engine->submit_request(request); + break; + + case FENCE_FREE: + break; + } + + return NOTIFY_DONE; +} + /** * i915_gem_request_alloc - allocate a request structure * @@ -331,7 +355,6 @@ i915_gem_request_alloc(struct intel_engine_cs *engine, struct i915_gem_context *ctx) { struct drm_i915_private *dev_priv = engine->i915; - unsigned int reset_counter = i915_reset_counter(&dev_priv->gpu_error); struct drm_i915_gem_request *req; u32 seqno; int ret; @@ -340,7 +363,7 @@ i915_gem_request_alloc(struct intel_engine_cs *engine, * EIO if the GPU is already wedged, or EAGAIN to drop the struct_mutex * and restart. */ - ret = i915_gem_check_wedge(reset_counter, dev_priv->mm.interruptible); + ret = i915_gem_check_wedge(dev_priv); if (ret) return ERR_PTR(ret); @@ -393,6 +416,8 @@ i915_gem_request_alloc(struct intel_engine_cs *engine, engine->fence_context, seqno); + i915_sw_fence_init(&req->submit, submit_notify); + INIT_LIST_HEAD(&req->active_list); req->i915 = dev_priv; req->engine = engine; @@ -402,7 +427,6 @@ i915_gem_request_alloc(struct intel_engine_cs *engine, req->previous_context = NULL; req->file_priv = NULL; req->batch = NULL; - req->elsp_submitted = 0; /* * Reserve space in the ring buffer for all the commands required to @@ -436,6 +460,94 @@ err: return ERR_PTR(ret); } +static int +i915_gem_request_await_request(struct drm_i915_gem_request *to, + struct drm_i915_gem_request *from) +{ + int idx, ret; + + GEM_BUG_ON(to == from); + + if (to->engine == from->engine) + return 0; + + idx = intel_engine_sync_index(from->engine, to->engine); + if (from->fence.seqno <= from->engine->semaphore.sync_seqno[idx]) + return 0; + + trace_i915_gem_ring_sync_to(to, from); + if (!i915.semaphores) { + if (!i915_spin_request(from, TASK_INTERRUPTIBLE, 2)) { + ret = i915_sw_fence_await_dma_fence(&to->submit, + &from->fence, 0, + GFP_KERNEL); + if (ret < 0) + return ret; + } + } else { + ret = to->engine->semaphore.sync_to(to, from); + if (ret) + return ret; + } + + from->engine->semaphore.sync_seqno[idx] = from->fence.seqno; + return 0; +} + +/** + * i915_gem_request_await_object - set this request to (async) wait upon a bo + * + * @to: request we are wishing to use + * @obj: object which may be in use on another ring. + * + * This code is meant to abstract object synchronization with the GPU. + * Conceptually we serialise writes between engines inside the GPU. + * We only allow one engine to write into a buffer at any time, but + * multiple readers. To ensure each has a coherent view of memory, we must: + * + * - If there is an outstanding write request to the object, the new + * request must wait for it to complete (either CPU or in hw, requests + * on the same ring will be naturally ordered). + * + * - If we are a write request (pending_write_domain is set), the new + * request must wait for outstanding read requests to complete. + * + * Returns 0 if successful, else propagates up the lower layer error. + */ +int +i915_gem_request_await_object(struct drm_i915_gem_request *to, + struct drm_i915_gem_object *obj, + bool write) +{ + struct i915_gem_active *active; + unsigned long active_mask; + int idx; + + if (write) { + active_mask = i915_gem_object_get_active(obj); + active = obj->last_read; + } else { + active_mask = 1; + active = &obj->last_write; + } + + for_each_active(active_mask, idx) { + struct drm_i915_gem_request *request; + int ret; + + request = i915_gem_active_peek(&active[idx], + &obj->base.dev->struct_mutex); + if (!request) + continue; + + ret = i915_gem_request_await_request(to, request); + if (ret) + return ret; + } + + return 0; +} + static void i915_gem_mark_busy(const struct intel_engine_cs *engine) { struct drm_i915_private *dev_priv = engine->i915; @@ -466,10 +578,13 @@ void __i915_add_request(struct drm_i915_gem_request *request, bool flush_caches) { struct intel_engine_cs *engine = request->engine; struct intel_ring *ring = request->ring; + struct drm_i915_gem_request *prev; u32 request_start; u32 reserved_tail; int ret; + trace_i915_gem_request_add(request); + /* * To ensure that this call will not fail, space for its emissions * should already have been reserved in the ring buffer. Let the ring @@ -493,20 +608,6 @@ void __i915_add_request(struct drm_i915_gem_request *request, bool flush_caches) WARN(ret, "engine->emit_flush() failed: %d!\n", ret); } - trace_i915_gem_request_add(request); - - /* Seal the request and mark it as pending execution. Note that - * we may inspect this state, without holding any locks, during - * hangcheck. Hence we apply the barrier to ensure that we do not - * see a more recent value in the hws than we are tracking. - */ - request->emitted_jiffies = jiffies; - request->previous_seqno = engine->last_submitted_seqno; - engine->last_submitted_seqno = request->fence.seqno; - i915_gem_active_set(&engine->last_request, request); - list_add_tail(&request->link, &engine->request_list); - list_add_tail(&request->ring_link, &ring->request_list); - /* Record the position of the start of the breadcrumb so that * should we detect the updated seqno part-way through the * GPU processing the request, we never over-estimate the @@ -527,8 +628,40 @@ void __i915_add_request(struct drm_i915_gem_request *request, bool flush_caches) "for adding the request (%d bytes)\n", reserved_tail, ret); + /* Seal the request and mark it as pending execution. Note that + * we may inspect this state, without holding any locks, during + * hangcheck. Hence we apply the barrier to ensure that we do not + * see a more recent value in the hws than we are tracking. + */ + + prev = i915_gem_active_raw(&engine->last_request, + &request->i915->drm.struct_mutex); + if (prev) + i915_sw_fence_await_sw_fence(&request->submit, &prev->submit, + &request->submitq); + + request->emitted_jiffies = jiffies; + request->previous_seqno = engine->last_submitted_seqno; + engine->last_submitted_seqno = request->fence.seqno; + i915_gem_active_set(&engine->last_request, request); + list_add_tail(&request->link, &engine->request_list); + list_add_tail(&request->ring_link, &ring->request_list); + i915_gem_mark_busy(engine); - engine->submit_request(request); + + local_bh_disable(); + i915_sw_fence_commit(&request->submit); + local_bh_enable(); /* Kick the execlists tasklet if just scheduled */ +} + +static void reset_wait_queue(wait_queue_head_t *q, wait_queue_t *wait) +{ + unsigned long flags; + + spin_lock_irqsave(&q->lock, flags); + if (list_empty(&wait->task_list)) + __add_wait_queue(q, wait); + spin_unlock_irqrestore(&q->lock, flags); } static unsigned long local_clock_us(unsigned int *cpu) @@ -598,7 +731,7 @@ bool __i915_spin_request(const struct drm_i915_gem_request *req, /** * i915_wait_request - wait until execution of request has finished * @req: duh! - * @interruptible: do an interruptible wait (normally yes) + * @flags: how to wait * @timeout: in - how long to wait (NULL forever); out - how much time remaining * @rps: client to charge for RPS boosting * @@ -613,17 +746,22 @@ bool __i915_spin_request(const struct drm_i915_gem_request *req, * errno with remaining time filled in timeout argument. */ int i915_wait_request(struct drm_i915_gem_request *req, - bool interruptible, + unsigned int flags, s64 *timeout, struct intel_rps_client *rps) { - int state = interruptible ? TASK_INTERRUPTIBLE : TASK_UNINTERRUPTIBLE; + const int state = flags & I915_WAIT_INTERRUPTIBLE ? + TASK_INTERRUPTIBLE : TASK_UNINTERRUPTIBLE; DEFINE_WAIT(reset); struct intel_wait wait; unsigned long timeout_remain; int ret = 0; might_sleep(); +#if IS_ENABLED(CONFIG_LOCKDEP) + GEM_BUG_ON(!!lockdep_is_held(&req->i915->drm.struct_mutex) != + !!(flags & I915_WAIT_LOCKED)); +#endif if (i915_gem_request_completed(req)) return 0; @@ -666,7 +804,8 @@ int i915_wait_request(struct drm_i915_gem_request *req, goto complete; set_current_state(state); - add_wait_queue(&req->i915->gpu_error.wait_queue, &reset); + if (flags & I915_WAIT_LOCKED) + add_wait_queue(&req->i915->gpu_error.wait_queue, &reset); intel_wait_init(&wait, req->fence.seqno); if (intel_engine_add_wait(req->engine, &wait)) @@ -702,14 +841,35 @@ wakeup: if (__i915_request_irq_complete(req)) break; + /* If the GPU is hung, and we hold the lock, reset the GPU + * and then check for completion. On a full reset, the engine's + * HW seqno will be advanced passed us and we are complete. + * If we do a partial reset, we have to wait for the GPU to + * resume and update the breadcrumb. + * + * If we don't hold the mutex, we can just wait for the worker + * to come along and update the breadcrumb (either directly + * itself, or indirectly by recovering the GPU). + */ + if (flags & I915_WAIT_LOCKED && + i915_reset_in_progress(&req->i915->gpu_error)) { + __set_current_state(TASK_RUNNING); + i915_reset(req->i915); + reset_wait_queue(&req->i915->gpu_error.wait_queue, + &reset); + continue; + } + /* Only spin if we know the GPU is processing this request */ if (i915_spin_request(req, state, 2)) break; } - remove_wait_queue(&req->i915->gpu_error.wait_queue, &reset); intel_engine_remove_wait(req->engine, &wait); + if (flags & I915_WAIT_LOCKED) + remove_wait_queue(&req->i915->gpu_error.wait_queue, &reset); __set_current_state(TASK_RUNNING); + complete: trace_i915_gem_request_wait_end(req); @@ -749,21 +909,24 @@ complete: return ret; } -static void engine_retire_requests(struct intel_engine_cs *engine) +static bool engine_retire_requests(struct intel_engine_cs *engine) { struct drm_i915_gem_request *request, *next; list_for_each_entry_safe(request, next, &engine->request_list, link) { if (!i915_gem_request_completed(request)) - break; + return false; i915_gem_request_retire(request); } + + return true; } void i915_gem_retire_requests(struct drm_i915_private *dev_priv) { struct intel_engine_cs *engine; + unsigned int tmp; lockdep_assert_held(&dev_priv->drm.struct_mutex); @@ -772,11 +935,9 @@ void i915_gem_retire_requests(struct drm_i915_private *dev_priv) GEM_BUG_ON(!dev_priv->gt.awake); - for_each_engine(engine, dev_priv) { - engine_retire_requests(engine); - if (!intel_engine_is_active(engine)) + for_each_engine_masked(engine, dev_priv, dev_priv->gt.active_engines, tmp) + if (engine_retire_requests(engine)) dev_priv->gt.active_engines &= ~intel_engine_flag(engine); - } if (dev_priv->gt.active_engines == 0) queue_delayed_work(dev_priv->wq, diff --git a/drivers/gpu/drm/i915/i915_gem_request.h b/drivers/gpu/drm/i915/i915_gem_request.h index 6c72bd8d9423..974bd7bcc801 100644 --- a/drivers/gpu/drm/i915/i915_gem_request.h +++ b/drivers/gpu/drm/i915/i915_gem_request.h @@ -28,6 +28,7 @@ #include <linux/fence.h> #include "i915_gem.h" +#include "i915_sw_fence.h" struct intel_wait { struct rb_node node; @@ -82,26 +83,32 @@ struct drm_i915_gem_request { struct intel_ring *ring; struct intel_signal_node signaling; + struct i915_sw_fence submit; + wait_queue_t submitq; + /** GEM sequence number associated with the previous request, * when the HWS breadcrumb is equal to this the GPU is processing * this request. */ u32 previous_seqno; - /** Position in the ringbuffer of the start of the request */ + /** Position in the ring of the start of the request */ u32 head; /** - * Position in the ringbuffer of the start of the postfix. - * This is required to calculate the maximum available ringbuffer - * space without overwriting the postfix. + * Position in the ring of the start of the postfix. + * This is required to calculate the maximum available ring space + * without overwriting the postfix. */ u32 postfix; - /** Position in the ringbuffer of the end of the whole request */ + /** Position in the ring of the end of the whole request */ u32 tail; - /** Preallocate space in the ringbuffer for the emitting the request */ + /** Position in the ring of the end of any workarounds after the tail */ + u32 wa_tail; + + /** Preallocate space in the ring for the emitting the request */ u32 reserved_space; /** @@ -134,27 +141,8 @@ struct drm_i915_gem_request { /** file_priv list entry for this request */ struct list_head client_list; - /** - * The ELSP only accepts two elements at a time, so we queue - * context/tail pairs on a given queue (ring->execlist_queue) until the - * hardware is available. The queue serves a double purpose: we also use - * it to keep track of the up to 2 contexts currently in the hardware - * (usually one in execution and the other queued up by the GPU): We - * only remove elements from the head of the queue when the hardware - * informs us that an element has been completed. - * - * All accesses to the queue are mediated by a spinlock - * (ring->execlist_lock). - */ - - /** Execlist link in the submission queue.*/ + /** Link in the execlist submission queue, guarded by execlist_lock. */ struct list_head execlist_link; - - /** Execlists no. of times this request has been sent to the ELSP */ - int elsp_submitted; - - /** Execlists context hardware id. */ - unsigned int ctx_hw_id; }; extern const struct fence_ops i915_fence_ops; @@ -222,6 +210,11 @@ static inline void i915_gem_request_assign(struct drm_i915_gem_request **pdst, *pdst = src; } +int +i915_gem_request_await_object(struct drm_i915_gem_request *to, + struct drm_i915_gem_object *obj, + bool write); + void __i915_add_request(struct drm_i915_gem_request *req, bool flush_caches); #define i915_add_request(req) \ __i915_add_request(req, true) @@ -234,10 +227,12 @@ struct intel_rps_client; #define IS_RPS_USER(p) (!IS_ERR_OR_NULL(p)) int i915_wait_request(struct drm_i915_gem_request *req, - bool interruptible, + unsigned int flags, s64 *timeout, struct intel_rps_client *rps) __attribute__((nonnull(1))); +#define I915_WAIT_INTERRUPTIBLE BIT(0) +#define I915_WAIT_LOCKED BIT(1) /* struct_mutex held, handle GPU reset */ static inline u32 intel_engine_get_seqno(struct intel_engine_cs *engine); @@ -472,6 +467,19 @@ __i915_gem_active_get_rcu(const struct i915_gem_active *active) if (!request || i915_gem_request_completed(request)) return NULL; + /* An especially silly compiler could decide to recompute the + * result of i915_gem_request_completed, more specifically + * re-emit the load for request->fence.seqno. A race would catch + * a later seqno value, which could flip the result from true to + * false. Which means part of the instructions below might not + * be executed, while later on instructions are executed. Due to + * barriers within the refcounting the inconsistency can't reach + * past the call to i915_gem_request_get_rcu, but not executing + * that while still executing i915_gem_request_put() creates + * havoc enough. Prevent this with a compiler barrier. + */ + barrier(); + request = i915_gem_request_get_rcu(request); /* What stops the following rcu_access_pointer() from occurring @@ -578,13 +586,15 @@ i915_gem_active_wait(const struct i915_gem_active *active, struct mutex *mutex) if (!request) return 0; - return i915_wait_request(request, true, NULL, NULL); + return i915_wait_request(request, + I915_WAIT_INTERRUPTIBLE | I915_WAIT_LOCKED, + NULL, NULL); } /** * i915_gem_active_wait_unlocked - waits until the request is completed * @active - the active request on which to wait - * @interruptible - whether the wait can be woken by a userspace signal + * @flags - how to wait * @timeout - how long to wait at most * @rps - userspace client to charge for a waitboost * @@ -605,7 +615,7 @@ i915_gem_active_wait(const struct i915_gem_active *active, struct mutex *mutex) */ static inline int i915_gem_active_wait_unlocked(const struct i915_gem_active *active, - bool interruptible, + unsigned int flags, s64 *timeout, struct intel_rps_client *rps) { @@ -614,7 +624,7 @@ i915_gem_active_wait_unlocked(const struct i915_gem_active *active, request = i915_gem_active_get_unlocked(active); if (request) { - ret = i915_wait_request(request, interruptible, timeout, rps); + ret = i915_wait_request(request, flags, timeout, rps); i915_gem_request_put(request); } @@ -641,7 +651,9 @@ i915_gem_active_retire(struct i915_gem_active *active, if (!request) return 0; - ret = i915_wait_request(request, true, NULL, NULL); + ret = i915_wait_request(request, + I915_WAIT_INTERRUPTIBLE | I915_WAIT_LOCKED, + NULL, NULL); if (ret) return ret; diff --git a/drivers/gpu/drm/i915/i915_gem_shrinker.c b/drivers/gpu/drm/i915/i915_gem_shrinker.c index b80802b35353..1c237d02f30b 100644 --- a/drivers/gpu/drm/i915/i915_gem_shrinker.c +++ b/drivers/gpu/drm/i915/i915_gem_shrinker.c @@ -323,7 +323,7 @@ i915_gem_shrinker_lock_uninterruptible(struct drm_i915_private *dev_priv, unsigned long timeout = jiffies + msecs_to_jiffies_timeout(timeout_ms); do { - if (i915_gem_wait_for_idle(dev_priv, false) == 0 && + if (i915_gem_wait_for_idle(dev_priv, 0) == 0 && i915_gem_shrinker_lock(&dev_priv->drm, &slu->unlock)) break; @@ -414,7 +414,7 @@ i915_gem_shrinker_vmap(struct notifier_block *nb, unsigned long event, void *ptr return NOTIFY_DONE; /* Force everything onto the inactive lists */ - ret = i915_gem_wait_for_idle(dev_priv, false); + ret = i915_gem_wait_for_idle(dev_priv, I915_WAIT_LOCKED); if (ret) goto out; diff --git a/drivers/gpu/drm/i915/i915_gem_stolen.c b/drivers/gpu/drm/i915/i915_gem_stolen.c index aa050fa1e558..59989e8ee5dc 100644 --- a/drivers/gpu/drm/i915/i915_gem_stolen.c +++ b/drivers/gpu/drm/i915/i915_gem_stolen.c @@ -92,6 +92,7 @@ void i915_gem_stolen_remove_node(struct drm_i915_private *dev_priv, static unsigned long i915_stolen_to_physical(struct drm_device *dev) { struct drm_i915_private *dev_priv = to_i915(dev); + struct pci_dev *pdev = dev_priv->drm.pdev; struct i915_ggtt *ggtt = &dev_priv->ggtt; struct resource *r; u32 base; @@ -111,7 +112,7 @@ static unsigned long i915_stolen_to_physical(struct drm_device *dev) if (INTEL_INFO(dev)->gen >= 3) { u32 bsm; - pci_read_config_dword(dev->pdev, INTEL_BSM, &bsm); + pci_read_config_dword(pdev, INTEL_BSM, &bsm); base = bsm & INTEL_BSM_MASK; } else if (IS_I865G(dev)) { @@ -119,7 +120,7 @@ static unsigned long i915_stolen_to_physical(struct drm_device *dev) u16 toud = 0; u8 tmp; - pci_bus_read_config_byte(dev->pdev->bus, PCI_DEVFN(0, 0), + pci_bus_read_config_byte(pdev->bus, PCI_DEVFN(0, 0), I845_ESMRAMC, &tmp); if (tmp & TSEG_ENABLE) { @@ -133,7 +134,7 @@ static unsigned long i915_stolen_to_physical(struct drm_device *dev) } } - pci_bus_read_config_word(dev->pdev->bus, PCI_DEVFN(0, 0), + pci_bus_read_config_word(pdev->bus, PCI_DEVFN(0, 0), I865_TOUD, &toud); base = (toud << 16) + tseg_size; @@ -142,13 +143,13 @@ static unsigned long i915_stolen_to_physical(struct drm_device *dev) u32 tom; u8 tmp; - pci_bus_read_config_byte(dev->pdev->bus, PCI_DEVFN(0, 0), + pci_bus_read_config_byte(pdev->bus, PCI_DEVFN(0, 0), I85X_ESMRAMC, &tmp); if (tmp & TSEG_ENABLE) tseg_size = MB(1); - pci_bus_read_config_byte(dev->pdev->bus, PCI_DEVFN(0, 1), + pci_bus_read_config_byte(pdev->bus, PCI_DEVFN(0, 1), I85X_DRB3, &tmp); tom = tmp * MB(32); @@ -158,7 +159,7 @@ static unsigned long i915_stolen_to_physical(struct drm_device *dev) u32 tom; u8 tmp; - pci_bus_read_config_byte(dev->pdev->bus, PCI_DEVFN(0, 0), + pci_bus_read_config_byte(pdev->bus, PCI_DEVFN(0, 0), I845_ESMRAMC, &tmp); if (tmp & TSEG_ENABLE) { @@ -172,7 +173,7 @@ static unsigned long i915_stolen_to_physical(struct drm_device *dev) } } - pci_bus_read_config_byte(dev->pdev->bus, PCI_DEVFN(0, 0), + pci_bus_read_config_byte(pdev->bus, PCI_DEVFN(0, 0), I830_DRB3, &tmp); tom = tmp * MB(32); @@ -182,7 +183,7 @@ static unsigned long i915_stolen_to_physical(struct drm_device *dev) u32 tom; u8 tmp; - pci_bus_read_config_byte(dev->pdev->bus, PCI_DEVFN(0, 0), + pci_bus_read_config_byte(pdev->bus, PCI_DEVFN(0, 0), I830_ESMRAMC, &tmp); if (tmp & TSEG_ENABLE) { @@ -192,7 +193,7 @@ static unsigned long i915_stolen_to_physical(struct drm_device *dev) tseg_size = KB(512); } - pci_bus_read_config_byte(dev->pdev->bus, PCI_DEVFN(0, 0), + pci_bus_read_config_byte(pdev->bus, PCI_DEVFN(0, 0), I830_DRB3, &tmp); tom = tmp * MB(32); diff --git a/drivers/gpu/drm/i915/i915_gem_userptr.c b/drivers/gpu/drm/i915/i915_gem_userptr.c index be54825ef3e8..e537930c64b5 100644 --- a/drivers/gpu/drm/i915/i915_gem_userptr.c +++ b/drivers/gpu/drm/i915/i915_gem_userptr.c @@ -68,7 +68,7 @@ static void wait_rendering(struct drm_i915_gem_object *obj) for_each_active(active, idx) i915_gem_active_wait_unlocked(&obj->last_read[idx], - false, NULL, NULL); + 0, NULL, NULL); } static void cancel_userptr(struct work_struct *work) diff --git a/drivers/gpu/drm/i915/i915_gpu_error.c b/drivers/gpu/drm/i915/i915_gpu_error.c index 41ec7a183c73..334f15df7c8d 100644 --- a/drivers/gpu/drm/i915/i915_gpu_error.c +++ b/drivers/gpu/drm/i915/i915_gpu_error.c @@ -336,6 +336,7 @@ int i915_error_state_to_str(struct drm_i915_error_state_buf *m, { struct drm_device *dev = error_priv->dev; struct drm_i915_private *dev_priv = to_i915(dev); + struct pci_dev *pdev = dev_priv->drm.pdev; struct drm_i915_error_state *error = error_priv->error; struct drm_i915_error_object *obj; int i, j, offset, elt; @@ -367,11 +368,11 @@ int i915_error_state_to_str(struct drm_i915_error_state_buf *m, } err_printf(m, "Reset count: %u\n", error->reset_count); err_printf(m, "Suspend count: %u\n", error->suspend_count); - err_printf(m, "PCI ID: 0x%04x\n", dev->pdev->device); - err_printf(m, "PCI Revision: 0x%02x\n", dev->pdev->revision); + err_printf(m, "PCI ID: 0x%04x\n", pdev->device); + err_printf(m, "PCI Revision: 0x%02x\n", pdev->revision); err_printf(m, "PCI Subsystem: %04x:%04x\n", - dev->pdev->subsystem_vendor, - dev->pdev->subsystem_device); + pdev->subsystem_vendor, + pdev->subsystem_device); err_printf(m, "IOMMU enabled?: %d\n", error->iommu); if (HAS_CSR(dev)) { @@ -488,7 +489,10 @@ int i915_error_state_to_str(struct drm_i915_error_state_buf *m, } } - if (ee->num_waiters) { + if (IS_ERR(ee->waiters)) { + err_printf(m, "%s --- ? waiters [unable to acquire spinlock]\n", + dev_priv->engine[i].name); + } else if (ee->num_waiters) { err_printf(m, "%s --- %d waiters\n", dev_priv->engine[i].name, ee->num_waiters); @@ -647,7 +651,8 @@ static void i915_error_state_free(struct kref *error_ref) i915_error_object_free(ee->wa_ctx); kfree(ee->requests); - kfree(ee->waiters); + if (!IS_ERR_OR_NULL(ee->waiters)) + kfree(ee->waiters); } i915_error_object_free(error->semaphore); @@ -932,7 +937,14 @@ static void error_record_engine_waiters(struct intel_engine_cs *engine, ee->num_waiters = 0; ee->waiters = NULL; - spin_lock(&b->lock); + if (RB_EMPTY_ROOT(&b->waiters)) + return; + + if (!spin_trylock(&b->lock)) { + ee->waiters = ERR_PTR(-EDEADLK); + return; + } + count = 0; for (rb = rb_first(&b->waiters); rb != NULL; rb = rb_next(rb)) count++; @@ -946,9 +958,13 @@ static void error_record_engine_waiters(struct intel_engine_cs *engine, if (!waiter) return; - ee->waiters = waiter; + if (!spin_trylock(&b->lock)) { + kfree(waiter); + ee->waiters = ERR_PTR(-EDEADLK); + return; + } - spin_lock(&b->lock); + ee->waiters = waiter; for (rb = rb_first(&b->waiters); rb; rb = rb_next(rb)) { struct intel_wait *w = container_of(rb, typeof(*w), node); @@ -1009,7 +1025,7 @@ static void error_record_engine_registers(struct drm_i915_error_state *error, if (INTEL_GEN(dev_priv) > 2) ee->mode = I915_READ_MODE(engine); - if (I915_NEED_GFX_HWS(dev_priv)) { + if (!HWS_NEEDS_PHYSICAL(dev_priv)) { i915_reg_t mmio; if (IS_GEN7(dev_priv)) { diff --git a/drivers/gpu/drm/i915/i915_guc_reg.h b/drivers/gpu/drm/i915/i915_guc_reg.h index cf5a65be4fe0..a47e1e4aec03 100644 --- a/drivers/gpu/drm/i915/i915_guc_reg.h +++ b/drivers/gpu/drm/i915/i915_guc_reg.h @@ -103,9 +103,6 @@ #define HOST2GUC_INTERRUPT _MMIO(0xc4c8) #define HOST2GUC_TRIGGER (1<<0) -#define DRBMISC1 0x1984 -#define DOORBELL_ENABLE (1<<0) - #define GEN8_DRBREGL(x) _MMIO(0x1000 + (x) * 8) #define GEN8_DRB_VALID (1<<0) #define GEN8_DRBREGU(x) _MMIO(0x1000 + (x) * 8 + 4) diff --git a/drivers/gpu/drm/i915/i915_guc_submission.c b/drivers/gpu/drm/i915/i915_guc_submission.c index e4369411f07d..43358e18d34c 100644 --- a/drivers/gpu/drm/i915/i915_guc_submission.c +++ b/drivers/gpu/drm/i915/i915_guc_submission.c @@ -59,7 +59,7 @@ * WQ_TYPE_INORDER is needed to support legacy submission via GuC, which * represents in-order queue. The kernel driver packs ring tail pointer and an * ELSP context descriptor dword into Work Item. - * See guc_add_workqueue_item() + * See guc_wq_item_append() * */ @@ -114,10 +114,8 @@ static int host2guc_action(struct intel_guc *guc, u32 *data, u32 len) if (ret != -ETIMEDOUT) ret = -EIO; - DRM_ERROR("GUC: host2guc action 0x%X failed. ret=%d " - "status=0x%08X response=0x%08X\n", - data[0], ret, status, - I915_READ(SOFT_SCRATCH(15))); + DRM_WARN("Action 0x%X failed; ret=%d status=0x%08X response=0x%08X\n", + data[0], ret, status, I915_READ(SOFT_SCRATCH(15))); dev_priv->guc.action_fail += 1; dev_priv->guc.action_err = ret; @@ -290,7 +288,7 @@ static uint32_t select_doorbell_cacheline(struct intel_guc *guc) /* * Initialise the process descriptor shared with the GuC firmware. */ -static void guc_init_proc_desc(struct intel_guc *guc, +static void guc_proc_desc_init(struct intel_guc *guc, struct i915_guc_client *client) { struct guc_process_desc *desc; @@ -322,7 +320,7 @@ static void guc_init_proc_desc(struct intel_guc *guc, * write queue, etc). */ -static void guc_init_ctx_desc(struct intel_guc *guc, +static void guc_ctx_desc_init(struct intel_guc *guc, struct i915_guc_client *client) { struct drm_i915_private *dev_priv = guc_to_i915(guc); @@ -330,6 +328,7 @@ static void guc_init_ctx_desc(struct intel_guc *guc, struct i915_gem_context *ctx = client->owner; struct guc_context_desc desc; struct sg_table *sg; + unsigned int tmp; u32 gfx_addr; memset(&desc, 0, sizeof(desc)); @@ -339,7 +338,7 @@ static void guc_init_ctx_desc(struct intel_guc *guc, desc.priority = client->priority; desc.db_id = client->doorbell_id; - for_each_engine_masked(engine, dev_priv, client->engines) { + for_each_engine_masked(engine, dev_priv, client->engines, tmp) { struct intel_context *ce = &ctx->engine[engine->id]; uint32_t guc_engine_id = engine->guc_id; struct guc_execlist_context *lrc = &desc.lrc[guc_engine_id]; @@ -400,7 +399,7 @@ static void guc_init_ctx_desc(struct intel_guc *guc, sizeof(desc) * client->ctx_index); } -static void guc_fini_ctx_desc(struct intel_guc *guc, +static void guc_ctx_desc_fini(struct intel_guc *guc, struct i915_guc_client *client) { struct guc_context_desc desc; @@ -414,7 +413,7 @@ static void guc_fini_ctx_desc(struct intel_guc *guc, } /** - * i915_guc_wq_check_space() - check that the GuC can accept a request + * i915_guc_wq_reserve() - reserve space in the GuC's workqueue * @request: request associated with the commands * * Return: 0 if space is available @@ -422,35 +421,39 @@ static void guc_fini_ctx_desc(struct intel_guc *guc, * * This function must be called (and must return 0) before a request * is submitted to the GuC via i915_guc_submit() below. Once a result - * of 0 has been returned, it remains valid until (but only until) - * the next call to submit(). + * of 0 has been returned, it must be balanced by a corresponding + * call to submit(). * - * This precheck allows the caller to determine in advance that space + * Reservation allows the caller to determine in advance that space * will be available for the next submission before committing resources * to it, and helps avoid late failures with complicated recovery paths. */ -int i915_guc_wq_check_space(struct drm_i915_gem_request *request) +int i915_guc_wq_reserve(struct drm_i915_gem_request *request) { const size_t wqi_size = sizeof(struct guc_wq_item); struct i915_guc_client *gc = request->i915->guc.execbuf_client; - struct guc_process_desc *desc; + struct guc_process_desc *desc = gc->client_base + gc->proc_desc_offset; u32 freespace; + int ret; - GEM_BUG_ON(gc == NULL); - - desc = gc->client_base + gc->proc_desc_offset; - + spin_lock(&gc->wq_lock); freespace = CIRC_SPACE(gc->wq_tail, desc->head, gc->wq_size); - if (likely(freespace >= wqi_size)) - return 0; - - gc->no_wq_space += 1; + freespace -= gc->wq_rsvd; + if (likely(freespace >= wqi_size)) { + gc->wq_rsvd += wqi_size; + ret = 0; + } else { + gc->no_wq_space++; + ret = -EAGAIN; + } + spin_unlock(&gc->wq_lock); - return -EAGAIN; + return ret; } -static void guc_add_workqueue_item(struct i915_guc_client *gc, - struct drm_i915_gem_request *rq) +/* Construct a Work Item and append it to the GuC's Work Queue */ +static void guc_wq_item_append(struct i915_guc_client *gc, + struct drm_i915_gem_request *rq) { /* wqi_len is in DWords, and does not include the one-word header */ const size_t wqi_size = sizeof(struct guc_wq_item); @@ -463,7 +466,7 @@ static void guc_add_workqueue_item(struct i915_guc_client *gc, desc = gc->client_base + gc->proc_desc_offset; - /* Free space is guaranteed, see i915_guc_wq_check_space() above */ + /* Free space is guaranteed, see i915_guc_wq_reserve() above */ freespace = CIRC_SPACE(gc->wq_tail, desc->head, gc->wq_size); GEM_BUG_ON(freespace < wqi_size); @@ -481,12 +484,14 @@ static void guc_add_workqueue_item(struct i915_guc_client *gc, * workqueue buffer dw by dw. */ BUILD_BUG_ON(wqi_size != 16); + GEM_BUG_ON(gc->wq_rsvd < wqi_size); /* postincrement WQ tail for next time */ wq_off = gc->wq_tail; + GEM_BUG_ON(wq_off & (wqi_size - 1)); gc->wq_tail += wqi_size; gc->wq_tail &= gc->wq_size - 1; - GEM_BUG_ON(wq_off & (wqi_size - 1)); + gc->wq_rsvd -= wqi_size; /* WQ starts from the page after doorbell / process_desc */ wq_page = (wq_off + GUC_DB_SIZE) >> PAGE_SHIFT; @@ -551,8 +556,8 @@ static int guc_ring_doorbell(struct i915_guc_client *gc) if (db_ret.db_status == GUC_DOORBELL_DISABLED) break; - DRM_ERROR("Cookie mismatch. Expected %d, returned %d\n", - db_cmp.cookie, db_ret.cookie); + DRM_WARN("Cookie mismatch. Expected %d, found %d\n", + db_cmp.cookie, db_ret.cookie); /* update the cookie to newly read cookie from GuC */ db_cmp.cookie = db_ret.cookie; @@ -571,14 +576,13 @@ static int guc_ring_doorbell(struct i915_guc_client *gc) * Return: 0 on success, otherwise an errno. * (Note: nonzero really shouldn't happen!) * - * The caller must have already called i915_guc_wq_check_space() above - * with a result of 0 (success) since the last request submission. This - * guarantees that there is space in the work queue for the new request, - * so enqueuing the item cannot fail. + * The caller must have already called i915_guc_wq_reserve() above with + * a result of 0 (success), guaranteeing that there is space in the work + * queue for the new request, so enqueuing the item cannot fail. * * Bad Things Will Happen if the caller violates this protocol e.g. calls - * submit() when check() says there's no space, or calls submit() multiple - * times with no intervening check(). + * submit() when _reserve() says there's no space, or calls _submit() + * a different number of times from (successful) calls to _reserve(). * * The only error here arises if the doorbell hardware isn't functioning * as expected, which really shouln't happen. @@ -590,7 +594,8 @@ static void i915_guc_submit(struct drm_i915_gem_request *rq) struct i915_guc_client *client = guc->execbuf_client; int b_ret; - guc_add_workqueue_item(client, rq); + spin_lock(&client->wq_lock); + guc_wq_item_append(client, rq); b_ret = guc_ring_doorbell(client); client->submissions[engine_id] += 1; @@ -600,6 +605,7 @@ static void i915_guc_submit(struct drm_i915_gem_request *rq) guc->submissions[engine_id] += 1; guc->last_seqno[engine_id] = rq->fence.seqno; + spin_unlock(&client->wq_lock); } /* @@ -680,7 +686,7 @@ guc_client_free(struct drm_i915_private *dev_priv, i915_vma_unpin_and_release(&client->vma); if (client->ctx_index != GUC_INVALID_CTX_ID) { - guc_fini_ctx_desc(guc, client); + guc_ctx_desc_fini(guc, client); ida_simple_remove(&guc->ctx_ids, client->ctx_index); } @@ -733,8 +739,8 @@ static void guc_init_doorbell_hw(struct intel_guc *guc) /* Restore to original value */ err = guc_update_doorbell_id(guc, client, db_id); if (err) - DRM_ERROR("Failed to restore doorbell to %d, err %d\n", - db_id, err); + DRM_WARN("Failed to restore doorbell to %d, err %d\n", + db_id, err); /* Read back & verify all doorbell registers */ for (i = 0; i < GUC_MAX_DOORBELLS; ++i) @@ -790,6 +796,8 @@ guc_client_alloc(struct drm_i915_private *dev_priv, /* We'll keep just the first (doorbell/proc) page permanently kmap'd. */ client->vma = vma; client->client_base = kmap(i915_vma_first_page(vma)); + + spin_lock_init(&client->wq_lock); client->wq_offset = GUC_DB_SIZE; client->wq_size = GUC_WQ_SIZE; @@ -810,8 +818,8 @@ guc_client_alloc(struct drm_i915_private *dev_priv, else client->proc_desc_offset = (GUC_DB_SIZE / 2); - guc_init_proc_desc(guc, client); - guc_init_ctx_desc(guc, client); + guc_proc_desc_init(guc, client); + guc_ctx_desc_init(guc, client); if (guc_init_doorbell(guc, client, db_id)) goto err; @@ -823,13 +831,11 @@ guc_client_alloc(struct drm_i915_private *dev_priv, return client; err: - DRM_ERROR("FAILED to create priority %u GuC client!\n", priority); - guc_client_free(dev_priv, client); return NULL; } -static void guc_create_log(struct intel_guc *guc) +static void guc_log_create(struct intel_guc *guc) { struct i915_vma *vma; unsigned long offset; @@ -869,7 +875,7 @@ static void guc_create_log(struct intel_guc *guc) guc->log_flags = (offset << GUC_LOG_BUF_ADDR_SHIFT) | flags; } -static void init_guc_policies(struct guc_policies *policies) +static void guc_policies_init(struct guc_policies *policies) { struct guc_policy *policy; u32 p, i; @@ -891,7 +897,7 @@ static void init_guc_policies(struct guc_policies *policies) policies->is_valid = 1; } -static void guc_create_ads(struct intel_guc *guc) +static void guc_addon_create(struct intel_guc *guc) { struct drm_i915_private *dev_priv = guc_to_i915(guc); struct i915_vma *vma; @@ -934,7 +940,7 @@ static void guc_create_ads(struct intel_guc *guc) /* GuC scheduling policies */ policies = (void *)ads + sizeof(struct guc_ads); - init_guc_policies(policies); + guc_policies_init(policies); ads->scheduler_policies = i915_ggtt_offset(vma) + sizeof(struct guc_ads); @@ -965,9 +971,11 @@ static void guc_create_ads(struct intel_guc *guc) */ int i915_guc_submission_init(struct drm_i915_private *dev_priv) { + const size_t ctxsize = sizeof(struct guc_context_desc); + const size_t poolsize = GUC_MAX_GPU_CONTEXTS * ctxsize; + const size_t gemsize = round_up(poolsize, PAGE_SIZE); struct intel_guc *guc = &dev_priv->guc; struct i915_vma *vma; - u32 size; /* Wipe bitmap & delete client in case of reinitialisation */ bitmap_clear(guc->doorbell_bitmap, 0, GUC_MAX_DOORBELLS); @@ -979,15 +987,14 @@ int i915_guc_submission_init(struct drm_i915_private *dev_priv) if (guc->ctx_pool_vma) return 0; /* already allocated */ - size = PAGE_ALIGN(GUC_MAX_GPU_CONTEXTS*sizeof(struct guc_context_desc)); - vma = guc_allocate_vma(guc, size); + vma = guc_allocate_vma(guc, gemsize); if (IS_ERR(vma)) return PTR_ERR(vma); guc->ctx_pool_vma = vma; ida_init(&guc->ctx_ids); - guc_create_log(guc); - guc_create_ads(guc); + guc_log_create(guc); + guc_addon_create(guc); return 0; } @@ -997,6 +1004,7 @@ int i915_guc_submission_enable(struct drm_i915_private *dev_priv) struct intel_guc *guc = &dev_priv->guc; struct i915_guc_client *client; struct intel_engine_cs *engine; + struct drm_i915_gem_request *request; /* client for execbuf submission */ client = guc_client_alloc(dev_priv, @@ -1004,7 +1012,7 @@ int i915_guc_submission_enable(struct drm_i915_private *dev_priv) GUC_CTX_PRIORITY_KMD_NORMAL, dev_priv->kernel_context); if (!client) { - DRM_ERROR("Failed to create execbuf guc_client\n"); + DRM_ERROR("Failed to create normal GuC client!\n"); return -ENOMEM; } @@ -1013,9 +1021,17 @@ int i915_guc_submission_enable(struct drm_i915_private *dev_priv) guc_init_doorbell_hw(guc); /* Take over from manual control of ELSP (execlists) */ - for_each_engine(engine, dev_priv) + for_each_engine(engine, dev_priv) { engine->submit_request = i915_guc_submit; + /* Replay the current set of previously submitted requests */ + list_for_each_entry(request, &engine->request_list, link) { + client->wq_rsvd += sizeof(struct guc_wq_item); + if (i915_sw_fence_done(&request->submit)) + i915_guc_submit(request); + } + } + return 0; } diff --git a/drivers/gpu/drm/i915/i915_irq.c b/drivers/gpu/drm/i915/i915_irq.c index ebb83d5a448b..c128fdbd24e4 100644 --- a/drivers/gpu/drm/i915/i915_irq.c +++ b/drivers/gpu/drm/i915/i915_irq.c @@ -371,7 +371,7 @@ void gen6_disable_rps_interrupts(struct drm_i915_private *dev_priv) spin_lock_irq(&dev_priv->irq_lock); dev_priv->rps.interrupts_enabled = false; - I915_WRITE(GEN6_PMINTRMSK, gen6_sanitize_rps_pm_mask(dev_priv, ~0)); + I915_WRITE(GEN6_PMINTRMSK, gen6_sanitize_rps_pm_mask(dev_priv, ~0u)); __gen6_disable_pm_irq(dev_priv, dev_priv->pm_rps_events); I915_WRITE(gen6_pm_ier(dev_priv), I915_READ(gen6_pm_ier(dev_priv)) & @@ -2497,57 +2497,52 @@ static void i915_reset_and_wakeup(struct drm_i915_private *dev_priv) char *error_event[] = { I915_ERROR_UEVENT "=1", NULL }; char *reset_event[] = { I915_RESET_UEVENT "=1", NULL }; char *reset_done_event[] = { I915_ERROR_UEVENT "=0", NULL }; - int ret; kobject_uevent_env(kobj, KOBJ_CHANGE, error_event); + DRM_DEBUG_DRIVER("resetting chip\n"); + kobject_uevent_env(kobj, KOBJ_CHANGE, reset_event); + /* - * Note that there's only one work item which does gpu resets, so we - * need not worry about concurrent gpu resets potentially incrementing - * error->reset_counter twice. We only need to take care of another - * racing irq/hangcheck declaring the gpu dead for a second time. A - * quick check for that is good enough: schedule_work ensures the - * correct ordering between hang detection and this work item, and since - * the reset in-progress bit is only ever set by code outside of this - * work we don't need to worry about any other races. + * In most cases it's guaranteed that we get here with an RPM + * reference held, for example because there is a pending GPU + * request that won't finish until the reset is done. This + * isn't the case at least when we get here by doing a + * simulated reset via debugs, so get an RPM reference. */ - if (i915_reset_in_progress(&dev_priv->gpu_error)) { - DRM_DEBUG_DRIVER("resetting chip\n"); - kobject_uevent_env(kobj, KOBJ_CHANGE, reset_event); - - /* - * In most cases it's guaranteed that we get here with an RPM - * reference held, for example because there is a pending GPU - * request that won't finish until the reset is done. This - * isn't the case at least when we get here by doing a - * simulated reset via debugs, so get an RPM reference. - */ - intel_runtime_pm_get(dev_priv); - - intel_prepare_reset(dev_priv); + intel_runtime_pm_get(dev_priv); + intel_prepare_reset(dev_priv); + do { /* * All state reset _must_ be completed before we update the * reset counter, for otherwise waiters might miss the reset * pending state and not properly drop locks, resulting in * deadlocks with the reset work. */ - ret = i915_reset(dev_priv); + if (mutex_trylock(&dev_priv->drm.struct_mutex)) { + i915_reset(dev_priv); + mutex_unlock(&dev_priv->drm.struct_mutex); + } - intel_finish_reset(dev_priv); + /* We need to wait for anyone holding the lock to wakeup */ + } while (wait_on_bit_timeout(&dev_priv->gpu_error.flags, + I915_RESET_IN_PROGRESS, + TASK_UNINTERRUPTIBLE, + HZ)); - intel_runtime_pm_put(dev_priv); + intel_finish_reset(dev_priv); + intel_runtime_pm_put(dev_priv); - if (ret == 0) - kobject_uevent_env(kobj, - KOBJ_CHANGE, reset_done_event); + if (!test_bit(I915_WEDGED, &dev_priv->gpu_error.flags)) + kobject_uevent_env(kobj, + KOBJ_CHANGE, reset_done_event); - /* - * Note: The wake_up also serves as a memory barrier so that - * waiters see the update value of the reset counter atomic_t. - */ - wake_up_all(&dev_priv->gpu_error.reset_queue); - } + /* + * Note: The wake_up also serves as a memory barrier so that + * waiters see the updated value of the dev_priv->gpu_error. + */ + wake_up_all(&dev_priv->gpu_error.reset_queue); } static void i915_report_and_clear_eir(struct drm_i915_private *dev_priv) @@ -2666,25 +2661,26 @@ void i915_handle_error(struct drm_i915_private *dev_priv, i915_capture_error_state(dev_priv, engine_mask, error_msg); i915_report_and_clear_eir(dev_priv); - if (engine_mask) { - atomic_or(I915_RESET_IN_PROGRESS_FLAG, - &dev_priv->gpu_error.reset_counter); + if (!engine_mask) + return; + + if (test_and_set_bit(I915_RESET_IN_PROGRESS, + &dev_priv->gpu_error.flags)) + return; - /* - * Wakeup waiting processes so that the reset function - * i915_reset_and_wakeup doesn't deadlock trying to grab - * various locks. By bumping the reset counter first, the woken - * processes will see a reset in progress and back off, - * releasing their locks and then wait for the reset completion. - * We must do this for _all_ gpu waiters that might hold locks - * that the reset work needs to acquire. - * - * Note: The wake_up serves as the required memory barrier to - * ensure that the waiters see the updated value of the reset - * counter atomic_t. - */ - i915_error_wake_up(dev_priv); - } + /* + * Wakeup waiting processes so that the reset function + * i915_reset_and_wakeup doesn't deadlock trying to grab + * various locks. By bumping the reset counter first, the woken + * processes will see a reset in progress and back off, + * releasing their locks and then wait for the reset completion. + * We must do this for _all_ gpu waiters that might hold locks + * that the reset work needs to acquire. + * + * Note: The wake_up also provides a memory barrier to ensure that the + * waiters see the updated value of the reset flags. + */ + i915_error_wake_up(dev_priv); i915_reset_and_wakeup(dev_priv); } @@ -2835,10 +2831,10 @@ semaphore_wait_to_signaller_ring(struct intel_engine_cs *engine, u32 ipehr, } } - DRM_ERROR("No signaller ring found for ring %i, ipehr 0x%08x, offset 0x%016llx\n", - engine->id, ipehr, offset); + DRM_DEBUG_DRIVER("No signaller ring found for ring %i, ipehr 0x%08x, offset 0x%016llx\n", + engine->id, ipehr, offset); - return NULL; + return ERR_PTR(-ENODEV); } static struct intel_engine_cs * @@ -2926,6 +2922,9 @@ static int semaphore_passed(struct intel_engine_cs *engine) if (signaller == NULL) return -1; + if (IS_ERR(signaller)) + return 0; + /* Prevent pathological recursion due to driver bugs */ if (signaller->hangcheck.deadlock >= I915_NUM_ENGINES) return -1; @@ -3079,6 +3078,7 @@ static void i915_hangcheck_elapsed(struct work_struct *work) bool busy = intel_engine_has_waiter(engine); u64 acthd; u32 seqno; + u32 submit; semaphore_clear_deadlocks(dev_priv); @@ -3094,14 +3094,11 @@ static void i915_hangcheck_elapsed(struct work_struct *work) acthd = intel_engine_get_active_head(engine); seqno = intel_engine_get_seqno(engine); + submit = READ_ONCE(engine->last_submitted_seqno); if (engine->hangcheck.seqno == seqno) { - if (!intel_engine_is_active(engine)) { + if (i915_seqno_passed(seqno, submit)) { engine->hangcheck.action = HANGCHECK_IDLE; - if (busy) { - /* Safeguard against driver failure */ - engine->hangcheck.score += BUSY; - } } else { /* We always increment the hangcheck score * if the engine is busy and still processing @@ -3167,6 +3164,7 @@ static void i915_hangcheck_elapsed(struct work_struct *work) if (hung) { char msg[80]; + unsigned int tmp; int len; /* If some rings hung but others were still busy, only @@ -3176,7 +3174,7 @@ static void i915_hangcheck_elapsed(struct work_struct *work) hung &= ~stuck; len = scnprintf(msg, sizeof(msg), "%s on ", stuck == hung ? "No progress" : "Hang"); - for_each_engine_masked(engine, dev_priv, hung) + for_each_engine_masked(engine, dev_priv, hung, tmp) len += scnprintf(msg + len, sizeof(msg) - len, "%s, ", engine->name); msg[len-2] = '\0'; @@ -4502,7 +4500,7 @@ void intel_irq_init(struct drm_i915_private *dev_priv) dev_priv->rps.pm_intr_keep |= GEN6_PM_RP_UP_EI_EXPIRED; if (INTEL_INFO(dev_priv)->gen >= 8) - dev_priv->rps.pm_intr_keep |= GEN8_PMINTR_REDIRECT_TO_NON_DISP; + dev_priv->rps.pm_intr_keep |= GEN8_PMINTR_REDIRECT_TO_GUC; INIT_DELAYED_WORK(&dev_priv->gpu_error.hangcheck_work, i915_hangcheck_elapsed); diff --git a/drivers/gpu/drm/i915/i915_pci.c b/drivers/gpu/drm/i915/i915_pci.c index 2587b1bd41f4..687c768833b3 100644 --- a/drivers/gpu/drm/i915/i915_pci.c +++ b/drivers/gpu/drm/i915/i915_pci.c @@ -54,208 +54,216 @@ #define CHV_COLORS \ .color = { .degamma_lut_size = 65, .gamma_lut_size = 257 } +#define GEN2_FEATURES \ + .gen = 2, .num_pipes = 1, \ + .has_overlay = 1, .overlay_needs_physical = 1, \ + .has_gmch_display = 1, \ + .hws_needs_physical = 1, \ + .ring_mask = RENDER_RING, \ + GEN_DEFAULT_PIPEOFFSETS, \ + CURSOR_OFFSETS + static const struct intel_device_info intel_i830_info = { - .gen = 2, .is_mobile = 1, .cursor_needs_physical = 1, .num_pipes = 2, - .has_overlay = 1, .overlay_needs_physical = 1, - .ring_mask = RENDER_RING, - GEN_DEFAULT_PIPEOFFSETS, - CURSOR_OFFSETS, + GEN2_FEATURES, + .is_mobile = 1, .cursor_needs_physical = 1, + .num_pipes = 2, /* legal, last one wins */ }; static const struct intel_device_info intel_845g_info = { - .gen = 2, .num_pipes = 1, - .has_overlay = 1, .overlay_needs_physical = 1, - .ring_mask = RENDER_RING, - GEN_DEFAULT_PIPEOFFSETS, - CURSOR_OFFSETS, + GEN2_FEATURES, }; static const struct intel_device_info intel_i85x_info = { - .gen = 2, .is_i85x = 1, .is_mobile = 1, .num_pipes = 2, + GEN2_FEATURES, + .is_i85x = 1, .is_mobile = 1, + .num_pipes = 2, /* legal, last one wins */ .cursor_needs_physical = 1, - .has_overlay = 1, .overlay_needs_physical = 1, .has_fbc = 1, - .ring_mask = RENDER_RING, - GEN_DEFAULT_PIPEOFFSETS, - CURSOR_OFFSETS, }; static const struct intel_device_info intel_i865g_info = { - .gen = 2, .num_pipes = 1, - .has_overlay = 1, .overlay_needs_physical = 1, - .ring_mask = RENDER_RING, - GEN_DEFAULT_PIPEOFFSETS, - CURSOR_OFFSETS, + GEN2_FEATURES, }; +#define GEN3_FEATURES \ + .gen = 3, .num_pipes = 2, \ + .has_gmch_display = 1, \ + .ring_mask = RENDER_RING, \ + GEN_DEFAULT_PIPEOFFSETS, \ + CURSOR_OFFSETS + static const struct intel_device_info intel_i915g_info = { - .gen = 3, .is_i915g = 1, .cursor_needs_physical = 1, .num_pipes = 2, + GEN3_FEATURES, + .is_i915g = 1, .cursor_needs_physical = 1, .has_overlay = 1, .overlay_needs_physical = 1, - .ring_mask = RENDER_RING, - GEN_DEFAULT_PIPEOFFSETS, - CURSOR_OFFSETS, + .hws_needs_physical = 1, }; static const struct intel_device_info intel_i915gm_info = { - .gen = 3, .is_mobile = 1, .num_pipes = 2, + GEN3_FEATURES, + .is_mobile = 1, .cursor_needs_physical = 1, .has_overlay = 1, .overlay_needs_physical = 1, .supports_tv = 1, .has_fbc = 1, - .ring_mask = RENDER_RING, - GEN_DEFAULT_PIPEOFFSETS, - CURSOR_OFFSETS, + .hws_needs_physical = 1, }; static const struct intel_device_info intel_i945g_info = { - .gen = 3, .has_hotplug = 1, .cursor_needs_physical = 1, .num_pipes = 2, + GEN3_FEATURES, + .has_hotplug = 1, .cursor_needs_physical = 1, .has_overlay = 1, .overlay_needs_physical = 1, - .ring_mask = RENDER_RING, - GEN_DEFAULT_PIPEOFFSETS, - CURSOR_OFFSETS, + .hws_needs_physical = 1, }; static const struct intel_device_info intel_i945gm_info = { - .gen = 3, .is_i945gm = 1, .is_mobile = 1, .num_pipes = 2, + GEN3_FEATURES, + .is_i945gm = 1, .is_mobile = 1, .has_hotplug = 1, .cursor_needs_physical = 1, .has_overlay = 1, .overlay_needs_physical = 1, .supports_tv = 1, .has_fbc = 1, - .ring_mask = RENDER_RING, - GEN_DEFAULT_PIPEOFFSETS, - CURSOR_OFFSETS, + .hws_needs_physical = 1, }; +#define GEN4_FEATURES \ + .gen = 4, .num_pipes = 2, \ + .has_hotplug = 1, \ + .has_gmch_display = 1, \ + .ring_mask = RENDER_RING, \ + GEN_DEFAULT_PIPEOFFSETS, \ + CURSOR_OFFSETS + static const struct intel_device_info intel_i965g_info = { - .gen = 4, .is_broadwater = 1, .num_pipes = 2, - .has_hotplug = 1, + GEN4_FEATURES, + .is_broadwater = 1, .has_overlay = 1, - .ring_mask = RENDER_RING, - GEN_DEFAULT_PIPEOFFSETS, - CURSOR_OFFSETS, + .hws_needs_physical = 1, }; static const struct intel_device_info intel_i965gm_info = { - .gen = 4, .is_crestline = 1, .num_pipes = 2, - .is_mobile = 1, .has_fbc = 1, .has_hotplug = 1, + GEN4_FEATURES, + .is_crestline = 1, + .is_mobile = 1, .has_fbc = 1, .has_overlay = 1, .supports_tv = 1, - .ring_mask = RENDER_RING, - GEN_DEFAULT_PIPEOFFSETS, - CURSOR_OFFSETS, + .hws_needs_physical = 1, }; static const struct intel_device_info intel_g33_info = { - .gen = 3, .is_g33 = 1, .num_pipes = 2, - .need_gfx_hws = 1, .has_hotplug = 1, + GEN3_FEATURES, + .is_g33 = 1, + .has_hotplug = 1, .has_overlay = 1, - .ring_mask = RENDER_RING, - GEN_DEFAULT_PIPEOFFSETS, - CURSOR_OFFSETS, }; static const struct intel_device_info intel_g45_info = { - .gen = 4, .is_g4x = 1, .need_gfx_hws = 1, .num_pipes = 2, - .has_pipe_cxsr = 1, .has_hotplug = 1, + GEN4_FEATURES, + .is_g4x = 1, + .has_pipe_cxsr = 1, .ring_mask = RENDER_RING | BSD_RING, - GEN_DEFAULT_PIPEOFFSETS, - CURSOR_OFFSETS, }; static const struct intel_device_info intel_gm45_info = { - .gen = 4, .is_g4x = 1, .num_pipes = 2, - .is_mobile = 1, .need_gfx_hws = 1, .has_fbc = 1, - .has_pipe_cxsr = 1, .has_hotplug = 1, + GEN4_FEATURES, + .is_g4x = 1, + .is_mobile = 1, .has_fbc = 1, + .has_pipe_cxsr = 1, .supports_tv = 1, .ring_mask = RENDER_RING | BSD_RING, - GEN_DEFAULT_PIPEOFFSETS, - CURSOR_OFFSETS, }; static const struct intel_device_info intel_pineview_info = { - .gen = 3, .is_g33 = 1, .is_pineview = 1, .is_mobile = 1, .num_pipes = 2, - .need_gfx_hws = 1, .has_hotplug = 1, + GEN3_FEATURES, + .is_g33 = 1, .is_pineview = 1, .is_mobile = 1, + .has_hotplug = 1, .has_overlay = 1, - .ring_mask = RENDER_RING, - GEN_DEFAULT_PIPEOFFSETS, - CURSOR_OFFSETS, }; +#define GEN5_FEATURES \ + .gen = 5, .num_pipes = 2, \ + .has_hotplug = 1, \ + .has_gmbus_irq = 1, \ + .ring_mask = RENDER_RING | BSD_RING, \ + GEN_DEFAULT_PIPEOFFSETS, \ + CURSOR_OFFSETS + static const struct intel_device_info intel_ironlake_d_info = { - .gen = 5, .num_pipes = 2, - .need_gfx_hws = 1, .has_hotplug = 1, - .ring_mask = RENDER_RING | BSD_RING, - GEN_DEFAULT_PIPEOFFSETS, - CURSOR_OFFSETS, + GEN5_FEATURES, }; static const struct intel_device_info intel_ironlake_m_info = { - .gen = 5, .is_mobile = 1, .num_pipes = 2, - .need_gfx_hws = 1, .has_hotplug = 1, - .has_fbc = 1, - .ring_mask = RENDER_RING | BSD_RING, - GEN_DEFAULT_PIPEOFFSETS, - CURSOR_OFFSETS, + GEN5_FEATURES, + .is_mobile = 1, }; +#define GEN6_FEATURES \ + .gen = 6, .num_pipes = 2, \ + .has_hotplug = 1, \ + .has_fbc = 1, \ + .ring_mask = RENDER_RING | BSD_RING | BLT_RING, \ + .has_llc = 1, \ + .has_rc6 = 1, \ + .has_rc6p = 1, \ + .has_gmbus_irq = 1, \ + .has_hw_contexts = 1, \ + GEN_DEFAULT_PIPEOFFSETS, \ + CURSOR_OFFSETS + static const struct intel_device_info intel_sandybridge_d_info = { - .gen = 6, .num_pipes = 2, - .need_gfx_hws = 1, .has_hotplug = 1, - .has_fbc = 1, - .ring_mask = RENDER_RING | BSD_RING | BLT_RING, - .has_llc = 1, - GEN_DEFAULT_PIPEOFFSETS, - CURSOR_OFFSETS, + GEN6_FEATURES, }; static const struct intel_device_info intel_sandybridge_m_info = { - .gen = 6, .is_mobile = 1, .num_pipes = 2, - .need_gfx_hws = 1, .has_hotplug = 1, - .has_fbc = 1, - .ring_mask = RENDER_RING | BSD_RING | BLT_RING, - .has_llc = 1, - GEN_DEFAULT_PIPEOFFSETS, - CURSOR_OFFSETS, + GEN6_FEATURES, + .is_mobile = 1, }; #define GEN7_FEATURES \ .gen = 7, .num_pipes = 3, \ - .need_gfx_hws = 1, .has_hotplug = 1, \ + .has_hotplug = 1, \ .has_fbc = 1, \ .ring_mask = RENDER_RING | BSD_RING | BLT_RING, \ .has_llc = 1, \ + .has_rc6 = 1, \ + .has_rc6p = 1, \ + .has_gmbus_irq = 1, \ + .has_hw_contexts = 1, \ GEN_DEFAULT_PIPEOFFSETS, \ IVB_CURSOR_OFFSETS static const struct intel_device_info intel_ivybridge_d_info = { GEN7_FEATURES, .is_ivybridge = 1, + .has_l3_dpf = 1, }; static const struct intel_device_info intel_ivybridge_m_info = { GEN7_FEATURES, .is_ivybridge = 1, .is_mobile = 1, + .has_l3_dpf = 1, }; static const struct intel_device_info intel_ivybridge_q_info = { GEN7_FEATURES, .is_ivybridge = 1, .num_pipes = 0, /* legal, last one wins */ + .has_l3_dpf = 1, }; #define VLV_FEATURES \ .gen = 7, .num_pipes = 2, \ - .need_gfx_hws = 1, .has_hotplug = 1, \ + .has_psr = 1, \ + .has_runtime_pm = 1, \ + .has_rc6 = 1, \ + .has_gmbus_irq = 1, \ + .has_hw_contexts = 1, \ + .has_gmch_display = 1, \ + .has_hotplug = 1, \ .ring_mask = RENDER_RING | BSD_RING | BLT_RING, \ .display_mmio_offset = VLV_DISPLAY_BASE, \ GEN_DEFAULT_PIPEOFFSETS, \ CURSOR_OFFSETS -static const struct intel_device_info intel_valleyview_m_info = { - VLV_FEATURES, - .is_valleyview = 1, - .is_mobile = 1, -}; - -static const struct intel_device_info intel_valleyview_d_info = { +static const struct intel_device_info intel_valleyview_info = { VLV_FEATURES, .is_valleyview = 1, }; @@ -264,54 +272,50 @@ static const struct intel_device_info intel_valleyview_d_info = { GEN7_FEATURES, \ .ring_mask = RENDER_RING | BSD_RING | BLT_RING | VEBOX_RING, \ .has_ddi = 1, \ - .has_fpga_dbg = 1 - -static const struct intel_device_info intel_haswell_d_info = { - HSW_FEATURES, - .is_haswell = 1, -}; - -static const struct intel_device_info intel_haswell_m_info = { + .has_fpga_dbg = 1, \ + .has_psr = 1, \ + .has_resource_streamer = 1, \ + .has_dp_mst = 1, \ + .has_rc6p = 0 /* RC6p removed-by HSW */, \ + .has_runtime_pm = 1 + +static const struct intel_device_info intel_haswell_info = { HSW_FEATURES, .is_haswell = 1, - .is_mobile = 1, + .has_l3_dpf = 1, }; #define BDW_FEATURES \ HSW_FEATURES, \ - BDW_COLORS + BDW_COLORS, \ + .has_logical_ring_contexts = 1 -static const struct intel_device_info intel_broadwell_d_info = { +static const struct intel_device_info intel_broadwell_info = { BDW_FEATURES, .gen = 8, .is_broadwell = 1, }; -static const struct intel_device_info intel_broadwell_m_info = { - BDW_FEATURES, - .gen = 8, .is_mobile = 1, - .is_broadwell = 1, -}; - -static const struct intel_device_info intel_broadwell_gt3d_info = { +static const struct intel_device_info intel_broadwell_gt3_info = { BDW_FEATURES, .gen = 8, .is_broadwell = 1, .ring_mask = RENDER_RING | BSD_RING | BLT_RING | VEBOX_RING | BSD2_RING, }; -static const struct intel_device_info intel_broadwell_gt3m_info = { - BDW_FEATURES, - .gen = 8, .is_mobile = 1, - .is_broadwell = 1, - .ring_mask = RENDER_RING | BSD_RING | BLT_RING | VEBOX_RING | BSD2_RING, -}; - static const struct intel_device_info intel_cherryview_info = { .gen = 8, .num_pipes = 3, - .need_gfx_hws = 1, .has_hotplug = 1, + .has_hotplug = 1, .ring_mask = RENDER_RING | BSD_RING | BLT_RING | VEBOX_RING, .is_cherryview = 1, + .has_psr = 1, + .has_runtime_pm = 1, + .has_resource_streamer = 1, + .has_rc6 = 1, + .has_gmbus_irq = 1, + .has_hw_contexts = 1, + .has_logical_ring_contexts = 1, + .has_gmch_display = 1, .display_mmio_offset = VLV_DISPLAY_BASE, GEN_CHV_PIPEOFFSETS, CURSOR_OFFSETS, @@ -322,25 +326,41 @@ static const struct intel_device_info intel_skylake_info = { BDW_FEATURES, .is_skylake = 1, .gen = 9, + .has_csr = 1, + .has_guc = 1, + .ddb_size = 896, }; static const struct intel_device_info intel_skylake_gt3_info = { BDW_FEATURES, .is_skylake = 1, .gen = 9, + .has_csr = 1, + .has_guc = 1, + .ddb_size = 896, .ring_mask = RENDER_RING | BSD_RING | BLT_RING | VEBOX_RING | BSD2_RING, }; static const struct intel_device_info intel_broxton_info = { .is_broxton = 1, .gen = 9, - .need_gfx_hws = 1, .has_hotplug = 1, + .has_hotplug = 1, .ring_mask = RENDER_RING | BSD_RING | BLT_RING | VEBOX_RING, .num_pipes = 3, .has_ddi = 1, .has_fpga_dbg = 1, .has_fbc = 1, + .has_runtime_pm = 1, .has_pooled_eu = 0, + .has_csr = 1, + .has_resource_streamer = 1, + .has_rc6 = 1, + .has_dp_mst = 1, + .has_gmbus_irq = 1, + .has_hw_contexts = 1, + .has_logical_ring_contexts = 1, + .has_guc = 1, + .ddb_size = 512, GEN_DEFAULT_PIPEOFFSETS, IVB_CURSOR_OFFSETS, BDW_COLORS, @@ -350,12 +370,18 @@ static const struct intel_device_info intel_kabylake_info = { BDW_FEATURES, .is_kabylake = 1, .gen = 9, + .has_csr = 1, + .has_guc = 1, + .ddb_size = 896, }; static const struct intel_device_info intel_kabylake_gt3_info = { BDW_FEATURES, .is_kabylake = 1, .gen = 9, + .has_csr = 1, + .has_guc = 1, + .ddb_size = 896, .ring_mask = RENDER_RING | BSD_RING | BLT_RING | VEBOX_RING | BSD2_RING, }; @@ -387,14 +413,10 @@ static const struct pci_device_id pciidlist[] = { INTEL_IVB_Q_IDS(&intel_ivybridge_q_info), /* must be first IVB */ INTEL_IVB_M_IDS(&intel_ivybridge_m_info), INTEL_IVB_D_IDS(&intel_ivybridge_d_info), - INTEL_HSW_D_IDS(&intel_haswell_d_info), - INTEL_HSW_M_IDS(&intel_haswell_m_info), - INTEL_VLV_M_IDS(&intel_valleyview_m_info), - INTEL_VLV_D_IDS(&intel_valleyview_d_info), - INTEL_BDW_GT12M_IDS(&intel_broadwell_m_info), - INTEL_BDW_GT12D_IDS(&intel_broadwell_d_info), - INTEL_BDW_GT3M_IDS(&intel_broadwell_gt3m_info), - INTEL_BDW_GT3D_IDS(&intel_broadwell_gt3d_info), + INTEL_HSW_IDS(&intel_haswell_info), + INTEL_VLV_IDS(&intel_valleyview_info), + INTEL_BDW_GT12_IDS(&intel_broadwell_info), + INTEL_BDW_GT3_IDS(&intel_broadwell_gt3_info), INTEL_CHV_IDS(&intel_cherryview_info), INTEL_SKL_GT1_IDS(&intel_skylake_info), INTEL_SKL_GT2_IDS(&intel_skylake_info), diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h index d4adf2806c50..70d96162def6 100644 --- a/drivers/gpu/drm/i915/i915_reg.h +++ b/drivers/gpu/drm/i915/i915_reg.h @@ -7067,7 +7067,7 @@ enum { #define VLV_RCEDATA _MMIO(0xA0BC) #define GEN6_RC6pp_THRESHOLD _MMIO(0xA0C0) #define GEN6_PMINTRMSK _MMIO(0xA168) -#define GEN8_PMINTR_REDIRECT_TO_NON_DISP (1<<31) +#define GEN8_PMINTR_REDIRECT_TO_GUC (1<<31) #define GEN8_MISC_CTRL0 _MMIO(0xA180) #define VLV_PWRDWNUPCTL _MMIO(0xA294) #define GEN9_MEDIA_PG_IDLE_HYSTERESIS _MMIO(0xA0C4) @@ -7123,6 +7123,15 @@ enum { #define GEN6_PCODE_MAILBOX _MMIO(0x138124) #define GEN6_PCODE_READY (1<<31) +#define GEN6_PCODE_ERROR_MASK 0xFF +#define GEN6_PCODE_SUCCESS 0x0 +#define GEN6_PCODE_ILLEGAL_CMD 0x1 +#define GEN6_PCODE_MIN_FREQ_TABLE_GT_RATIO_OUT_OF_RANGE 0x2 +#define GEN6_PCODE_TIMEOUT 0x3 +#define GEN6_PCODE_UNIMPLEMENTED_CMD 0xFF +#define GEN7_PCODE_TIMEOUT 0x2 +#define GEN7_PCODE_ILLEGAL_DATA 0x3 +#define GEN7_PCODE_MIN_FREQ_TABLE_GT_RATIO_OUT_OF_RANGE 0x10 #define GEN6_PCODE_WRITE_RC6VIDS 0x4 #define GEN6_PCODE_READ_RC6VIDS 0x5 #define GEN6_ENCODE_RC6_VID(mv) (((mv) - 245) / 5) @@ -7144,6 +7153,10 @@ enum { #define HSW_PCODE_DE_WRITE_FREQ_REQ 0x17 #define DISPLAY_IPS_CONTROL 0x19 #define HSW_PCODE_DYNAMIC_DUTY_CYCLE_CONTROL 0x1A +#define GEN9_PCODE_SAGV_CONTROL 0x21 +#define GEN9_SAGV_DISABLE 0x0 +#define GEN9_SAGV_IS_DISABLED 0x1 +#define GEN9_SAGV_ENABLE 0x3 #define GEN6_PCODE_DATA _MMIO(0x138128) #define GEN6_PCODE_FREQ_IA_RATIO_SHIFT 8 #define GEN6_PCODE_FREQ_RING_RATIO_SHIFT 16 diff --git a/drivers/gpu/drm/i915/i915_suspend.c b/drivers/gpu/drm/i915/i915_suspend.c index 4f272777b4f4..a0af170062b1 100644 --- a/drivers/gpu/drm/i915/i915_suspend.c +++ b/drivers/gpu/drm/i915/i915_suspend.c @@ -63,6 +63,7 @@ static void i915_restore_display(struct drm_device *dev) int i915_save_state(struct drm_device *dev) { struct drm_i915_private *dev_priv = to_i915(dev); + struct pci_dev *pdev = dev_priv->drm.pdev; int i; mutex_lock(&dev->struct_mutex); @@ -70,7 +71,7 @@ int i915_save_state(struct drm_device *dev) i915_save_display(dev); if (IS_GEN4(dev)) - pci_read_config_word(dev->pdev, GCDGMBUS, + pci_read_config_word(pdev, GCDGMBUS, &dev_priv->regfile.saveGCDGMBUS); /* Cache mode state */ @@ -108,6 +109,7 @@ int i915_save_state(struct drm_device *dev) int i915_restore_state(struct drm_device *dev) { struct drm_i915_private *dev_priv = to_i915(dev); + struct pci_dev *pdev = dev_priv->drm.pdev; int i; mutex_lock(&dev->struct_mutex); @@ -115,7 +117,7 @@ int i915_restore_state(struct drm_device *dev) i915_gem_restore_fences(dev); if (IS_GEN4(dev)) - pci_write_config_word(dev->pdev, GCDGMBUS, + pci_write_config_word(pdev, GCDGMBUS, dev_priv->regfile.saveGCDGMBUS); i915_restore_display(dev); diff --git a/drivers/gpu/drm/i915/i915_sw_fence.c b/drivers/gpu/drm/i915/i915_sw_fence.c new file mode 100644 index 000000000000..1e5cbc585ca2 --- /dev/null +++ b/drivers/gpu/drm/i915/i915_sw_fence.c @@ -0,0 +1,362 @@ +/* + * (C) Copyright 2016 Intel Corporation + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License + * as published by the Free Software Foundation; version 2 + * of the License. + */ + +#include <linux/slab.h> +#include <linux/fence.h> +#include <linux/reservation.h> + +#include "i915_sw_fence.h" + +static DEFINE_SPINLOCK(i915_sw_fence_lock); + +static int __i915_sw_fence_notify(struct i915_sw_fence *fence, + enum i915_sw_fence_notify state) +{ + i915_sw_fence_notify_t fn; + + fn = (i915_sw_fence_notify_t)(fence->flags & I915_SW_FENCE_MASK); + return fn(fence, state); +} + +static void i915_sw_fence_free(struct kref *kref) +{ + struct i915_sw_fence *fence = container_of(kref, typeof(*fence), kref); + + WARN_ON(atomic_read(&fence->pending) > 0); + + if (fence->flags & I915_SW_FENCE_MASK) + __i915_sw_fence_notify(fence, FENCE_FREE); + else + kfree(fence); +} + +static void i915_sw_fence_put(struct i915_sw_fence *fence) +{ + kref_put(&fence->kref, i915_sw_fence_free); +} + +static struct i915_sw_fence *i915_sw_fence_get(struct i915_sw_fence *fence) +{ + kref_get(&fence->kref); + return fence; +} + +static void __i915_sw_fence_wake_up_all(struct i915_sw_fence *fence, + struct list_head *continuation) +{ + wait_queue_head_t *x = &fence->wait; + wait_queue_t *pos, *next; + unsigned long flags; + + atomic_set_release(&fence->pending, -1); /* 0 -> -1 [done] */ + + /* + * To prevent unbounded recursion as we traverse the graph of + * i915_sw_fences, we move the task_list from this, the next ready + * fence, to the tail of the original fence's task_list + * (and so added to the list to be woken). + */ + + spin_lock_irqsave_nested(&x->lock, flags, 1 + !!continuation); + if (continuation) { + list_for_each_entry_safe(pos, next, &x->task_list, task_list) { + if (pos->func == autoremove_wake_function) + pos->func(pos, TASK_NORMAL, 0, continuation); + else + list_move_tail(&pos->task_list, continuation); + } + } else { + LIST_HEAD(extra); + + do { + list_for_each_entry_safe(pos, next, + &x->task_list, task_list) + pos->func(pos, TASK_NORMAL, 0, &extra); + + if (list_empty(&extra)) + break; + + list_splice_tail_init(&extra, &x->task_list); + } while (1); + } + spin_unlock_irqrestore(&x->lock, flags); +} + +static void __i915_sw_fence_complete(struct i915_sw_fence *fence, + struct list_head *continuation) +{ + if (!atomic_dec_and_test(&fence->pending)) + return; + + if (fence->flags & I915_SW_FENCE_MASK && + __i915_sw_fence_notify(fence, FENCE_COMPLETE) != NOTIFY_DONE) + return; + + __i915_sw_fence_wake_up_all(fence, continuation); +} + +static void i915_sw_fence_complete(struct i915_sw_fence *fence) +{ + if (WARN_ON(i915_sw_fence_done(fence))) + return; + + __i915_sw_fence_complete(fence, NULL); +} + +static void i915_sw_fence_await(struct i915_sw_fence *fence) +{ + WARN_ON(atomic_inc_return(&fence->pending) <= 1); +} + +void i915_sw_fence_init(struct i915_sw_fence *fence, i915_sw_fence_notify_t fn) +{ + BUG_ON((unsigned long)fn & ~I915_SW_FENCE_MASK); + + init_waitqueue_head(&fence->wait); + kref_init(&fence->kref); + atomic_set(&fence->pending, 1); + fence->flags = (unsigned long)fn; +} + +void i915_sw_fence_commit(struct i915_sw_fence *fence) +{ + i915_sw_fence_complete(fence); + i915_sw_fence_put(fence); +} + +static int i915_sw_fence_wake(wait_queue_t *wq, unsigned mode, int flags, void *key) +{ + list_del(&wq->task_list); + __i915_sw_fence_complete(wq->private, key); + i915_sw_fence_put(wq->private); + return 0; +} + +static bool __i915_sw_fence_check_if_after(struct i915_sw_fence *fence, + const struct i915_sw_fence * const signaler) +{ + wait_queue_t *wq; + + if (__test_and_set_bit(I915_SW_FENCE_CHECKED_BIT, &fence->flags)) + return false; + + if (fence == signaler) + return true; + + list_for_each_entry(wq, &fence->wait.task_list, task_list) { + if (wq->func != i915_sw_fence_wake) + continue; + + if (__i915_sw_fence_check_if_after(wq->private, signaler)) + return true; + } + + return false; +} + +static void __i915_sw_fence_clear_checked_bit(struct i915_sw_fence *fence) +{ + wait_queue_t *wq; + + if (!__test_and_clear_bit(I915_SW_FENCE_CHECKED_BIT, &fence->flags)) + return; + + list_for_each_entry(wq, &fence->wait.task_list, task_list) { + if (wq->func != i915_sw_fence_wake) + continue; + + __i915_sw_fence_clear_checked_bit(wq->private); + } +} + +static bool i915_sw_fence_check_if_after(struct i915_sw_fence *fence, + const struct i915_sw_fence * const signaler) +{ + unsigned long flags; + bool err; + + if (!IS_ENABLED(CONFIG_I915_SW_FENCE_CHECK_DAG)) + return false; + + spin_lock_irqsave(&i915_sw_fence_lock, flags); + err = __i915_sw_fence_check_if_after(fence, signaler); + __i915_sw_fence_clear_checked_bit(fence); + spin_unlock_irqrestore(&i915_sw_fence_lock, flags); + + return err; +} + +int i915_sw_fence_await_sw_fence(struct i915_sw_fence *fence, + struct i915_sw_fence *signaler, + wait_queue_t *wq) +{ + unsigned long flags; + int pending; + + if (i915_sw_fence_done(signaler)) + return 0; + + /* The dependency graph must be acyclic. */ + if (unlikely(i915_sw_fence_check_if_after(fence, signaler))) + return -EINVAL; + + INIT_LIST_HEAD(&wq->task_list); + wq->flags = 0; + wq->func = i915_sw_fence_wake; + wq->private = i915_sw_fence_get(fence); + + i915_sw_fence_await(fence); + + spin_lock_irqsave(&signaler->wait.lock, flags); + if (likely(!i915_sw_fence_done(signaler))) { + __add_wait_queue_tail(&signaler->wait, wq); + pending = 1; + } else { + i915_sw_fence_wake(wq, 0, 0, NULL); + pending = 0; + } + spin_unlock_irqrestore(&signaler->wait.lock, flags); + + return pending; +} + +struct dma_fence_cb { + struct fence_cb base; + struct i915_sw_fence *fence; + struct fence *dma; + struct timer_list timer; +}; + +static void timer_i915_sw_fence_wake(unsigned long data) +{ + struct dma_fence_cb *cb = (struct dma_fence_cb *)data; + + printk(KERN_WARNING "asynchronous wait on fence %s:%s:%x timed out\n", + cb->dma->ops->get_driver_name(cb->dma), + cb->dma->ops->get_timeline_name(cb->dma), + cb->dma->seqno); + fence_put(cb->dma); + cb->dma = NULL; + + i915_sw_fence_commit(cb->fence); + cb->timer.function = NULL; +} + +static void dma_i915_sw_fence_wake(struct fence *dma, struct fence_cb *data) +{ + struct dma_fence_cb *cb = container_of(data, typeof(*cb), base); + + del_timer_sync(&cb->timer); + if (cb->timer.function) + i915_sw_fence_commit(cb->fence); + fence_put(cb->dma); + + kfree(cb); +} + +int i915_sw_fence_await_dma_fence(struct i915_sw_fence *fence, + struct fence *dma, + unsigned long timeout, + gfp_t gfp) +{ + struct dma_fence_cb *cb; + int ret; + + if (fence_is_signaled(dma)) + return 0; + + cb = kmalloc(sizeof(*cb), gfp); + if (!cb) { + if (!gfpflags_allow_blocking(gfp)) + return -ENOMEM; + + return fence_wait(dma, false); + } + + cb->fence = i915_sw_fence_get(fence); + i915_sw_fence_await(fence); + + cb->dma = NULL; + __setup_timer(&cb->timer, + timer_i915_sw_fence_wake, (unsigned long)cb, + TIMER_IRQSAFE); + if (timeout) { + cb->dma = fence_get(dma); + mod_timer(&cb->timer, round_jiffies_up(jiffies + timeout)); + } + + ret = fence_add_callback(dma, &cb->base, dma_i915_sw_fence_wake); + if (ret == 0) { + ret = 1; + } else { + dma_i915_sw_fence_wake(dma, &cb->base); + if (ret == -ENOENT) /* fence already signaled */ + ret = 0; + } + + return ret; +} + +int i915_sw_fence_await_reservation(struct i915_sw_fence *fence, + struct reservation_object *resv, + const struct fence_ops *exclude, + bool write, + unsigned long timeout, + gfp_t gfp) +{ + struct fence *excl; + int ret = 0, pending; + + if (write) { + struct fence **shared; + unsigned int count, i; + + ret = reservation_object_get_fences_rcu(resv, + &excl, &count, &shared); + if (ret) + return ret; + + for (i = 0; i < count; i++) { + if (shared[i]->ops == exclude) + continue; + + pending = i915_sw_fence_await_dma_fence(fence, + shared[i], + timeout, + gfp); + if (pending < 0) { + ret = pending; + break; + } + + ret |= pending; + } + + for (i = 0; i < count; i++) + fence_put(shared[i]); + kfree(shared); + } else { + excl = reservation_object_get_excl_rcu(resv); + } + + if (ret >= 0 && excl && excl->ops != exclude) { + pending = i915_sw_fence_await_dma_fence(fence, + excl, + timeout, + gfp); + if (pending < 0) + ret = pending; + else + ret |= pending; + } + + fence_put(excl); + + return ret; +} diff --git a/drivers/gpu/drm/i915/i915_sw_fence.h b/drivers/gpu/drm/i915/i915_sw_fence.h new file mode 100644 index 000000000000..373141602ca4 --- /dev/null +++ b/drivers/gpu/drm/i915/i915_sw_fence.h @@ -0,0 +1,65 @@ +/* + * i915_sw_fence.h - library routines for N:M synchronisation points + * + * Copyright (C) 2016 Intel Corporation + * + * This file is released under the GPLv2. + * + */ + +#ifndef _I915_SW_FENCE_H_ +#define _I915_SW_FENCE_H_ + +#include <linux/gfp.h> +#include <linux/kref.h> +#include <linux/notifier.h> /* for NOTIFY_DONE */ +#include <linux/wait.h> + +struct completion; +struct fence; +struct fence_ops; +struct reservation_object; + +struct i915_sw_fence { + wait_queue_head_t wait; + unsigned long flags; + struct kref kref; + atomic_t pending; +}; + +#define I915_SW_FENCE_CHECKED_BIT 0 /* used internally for DAG checking */ +#define I915_SW_FENCE_PRIVATE_BIT 1 /* available for use by owner */ +#define I915_SW_FENCE_MASK (~3) + +enum i915_sw_fence_notify { + FENCE_COMPLETE, + FENCE_FREE +}; + +typedef int (*i915_sw_fence_notify_t)(struct i915_sw_fence *, + enum i915_sw_fence_notify state); +#define __i915_sw_fence_call __aligned(4) + +void i915_sw_fence_init(struct i915_sw_fence *fence, i915_sw_fence_notify_t fn); +void i915_sw_fence_commit(struct i915_sw_fence *fence); + +int i915_sw_fence_await_sw_fence(struct i915_sw_fence *fence, + struct i915_sw_fence *after, + wait_queue_t *wq); +int i915_sw_fence_await_dma_fence(struct i915_sw_fence *fence, + struct fence *dma, + unsigned long timeout, + gfp_t gfp); +int i915_sw_fence_await_reservation(struct i915_sw_fence *fence, + struct reservation_object *resv, + const struct fence_ops *exclude, + bool write, + unsigned long timeout, + gfp_t gfp); + +static inline bool i915_sw_fence_done(const struct i915_sw_fence *fence) +{ + return atomic_read(&fence->pending) < 0; +} + +#endif /* _I915_SW_FENCE_H_ */ diff --git a/drivers/gpu/drm/i915/i915_sysfs.c b/drivers/gpu/drm/i915/i915_sysfs.c index f1ffde7f7c0b..1012eeea1324 100644 --- a/drivers/gpu/drm/i915/i915_sysfs.c +++ b/drivers/gpu/drm/i915/i915_sysfs.c @@ -32,13 +32,16 @@ #include "intel_drv.h" #include "i915_drv.h" -#define dev_to_drm_minor(d) dev_get_drvdata((d)) +static inline struct drm_i915_private *kdev_minor_to_i915(struct device *kdev) +{ + struct drm_minor *minor = dev_get_drvdata(kdev); + return to_i915(minor->dev); +} #ifdef CONFIG_PM -static u32 calc_residency(struct drm_device *dev, +static u32 calc_residency(struct drm_i915_private *dev_priv, i915_reg_t reg) { - struct drm_i915_private *dev_priv = to_i915(dev); u64 raw_time; /* 32b value may overflow during fixed point math */ u64 units = 128ULL, div = 100000ULL; u32 ret; @@ -49,13 +52,13 @@ static u32 calc_residency(struct drm_device *dev, intel_runtime_pm_get(dev_priv); /* On VLV and CHV, residency time is in CZ units rather than 1.28us */ - if (IS_VALLEYVIEW(dev) || IS_CHERRYVIEW(dev)) { + if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv)) { units = 1; div = dev_priv->czclk_freq; if (I915_READ(VLV_COUNTER_CONTROL) & VLV_COUNT_RANGE_HIGH) units <<= 8; - } else if (IS_BROXTON(dev)) { + } else if (IS_BROXTON(dev_priv)) { units = 1; div = 1200; /* 833.33ns */ } @@ -76,32 +79,32 @@ show_rc6_mask(struct device *kdev, struct device_attribute *attr, char *buf) static ssize_t show_rc6_ms(struct device *kdev, struct device_attribute *attr, char *buf) { - struct drm_minor *dminor = dev_get_drvdata(kdev); - u32 rc6_residency = calc_residency(dminor->dev, GEN6_GT_GFX_RC6); + struct drm_i915_private *dev_priv = kdev_minor_to_i915(kdev); + u32 rc6_residency = calc_residency(dev_priv, GEN6_GT_GFX_RC6); return snprintf(buf, PAGE_SIZE, "%u\n", rc6_residency); } static ssize_t show_rc6p_ms(struct device *kdev, struct device_attribute *attr, char *buf) { - struct drm_minor *dminor = dev_to_drm_minor(kdev); - u32 rc6p_residency = calc_residency(dminor->dev, GEN6_GT_GFX_RC6p); + struct drm_i915_private *dev_priv = kdev_minor_to_i915(kdev); + u32 rc6p_residency = calc_residency(dev_priv, GEN6_GT_GFX_RC6p); return snprintf(buf, PAGE_SIZE, "%u\n", rc6p_residency); } static ssize_t show_rc6pp_ms(struct device *kdev, struct device_attribute *attr, char *buf) { - struct drm_minor *dminor = dev_to_drm_minor(kdev); - u32 rc6pp_residency = calc_residency(dminor->dev, GEN6_GT_GFX_RC6pp); + struct drm_i915_private *dev_priv = kdev_minor_to_i915(kdev); + u32 rc6pp_residency = calc_residency(dev_priv, GEN6_GT_GFX_RC6pp); return snprintf(buf, PAGE_SIZE, "%u\n", rc6pp_residency); } static ssize_t show_media_rc6_ms(struct device *kdev, struct device_attribute *attr, char *buf) { - struct drm_minor *dminor = dev_get_drvdata(kdev); - u32 rc6_residency = calc_residency(dminor->dev, VLV_GT_MEDIA_RC6); + struct drm_i915_private *dev_priv = kdev_minor_to_i915(kdev); + u32 rc6_residency = calc_residency(dev_priv, VLV_GT_MEDIA_RC6); return snprintf(buf, PAGE_SIZE, "%u\n", rc6_residency); } @@ -144,9 +147,9 @@ static struct attribute_group media_rc6_attr_group = { }; #endif -static int l3_access_valid(struct drm_device *dev, loff_t offset) +static int l3_access_valid(struct drm_i915_private *dev_priv, loff_t offset) { - if (!HAS_L3_DPF(dev)) + if (!HAS_L3_DPF(dev_priv)) return -EPERM; if (offset % 4 != 0) @@ -163,22 +166,21 @@ i915_l3_read(struct file *filp, struct kobject *kobj, struct bin_attribute *attr, char *buf, loff_t offset, size_t count) { - struct device *dev = kobj_to_dev(kobj); - struct drm_minor *dminor = dev_to_drm_minor(dev); - struct drm_device *drm_dev = dminor->dev; - struct drm_i915_private *dev_priv = to_i915(drm_dev); + struct device *kdev = kobj_to_dev(kobj); + struct drm_i915_private *dev_priv = kdev_minor_to_i915(kdev); + struct drm_device *dev = &dev_priv->drm; int slice = (int)(uintptr_t)attr->private; int ret; count = round_down(count, 4); - ret = l3_access_valid(drm_dev, offset); + ret = l3_access_valid(dev_priv, offset); if (ret) return ret; count = min_t(size_t, GEN7_L3LOG_SIZE - offset, count); - ret = i915_mutex_lock_interruptible(drm_dev); + ret = i915_mutex_lock_interruptible(dev); if (ret) return ret; @@ -189,7 +191,7 @@ i915_l3_read(struct file *filp, struct kobject *kobj, else memset(buf, 0, count); - mutex_unlock(&drm_dev->struct_mutex); + mutex_unlock(&dev->struct_mutex); return count; } @@ -199,30 +201,29 @@ i915_l3_write(struct file *filp, struct kobject *kobj, struct bin_attribute *attr, char *buf, loff_t offset, size_t count) { - struct device *dev = kobj_to_dev(kobj); - struct drm_minor *dminor = dev_to_drm_minor(dev); - struct drm_device *drm_dev = dminor->dev; - struct drm_i915_private *dev_priv = to_i915(drm_dev); + struct device *kdev = kobj_to_dev(kobj); + struct drm_i915_private *dev_priv = kdev_minor_to_i915(kdev); + struct drm_device *dev = &dev_priv->drm; struct i915_gem_context *ctx; u32 *temp = NULL; /* Just here to make handling failures easy */ int slice = (int)(uintptr_t)attr->private; int ret; - if (!HAS_HW_CONTEXTS(drm_dev)) + if (!HAS_HW_CONTEXTS(dev_priv)) return -ENXIO; - ret = l3_access_valid(drm_dev, offset); + ret = l3_access_valid(dev_priv, offset); if (ret) return ret; - ret = i915_mutex_lock_interruptible(drm_dev); + ret = i915_mutex_lock_interruptible(dev); if (ret) return ret; if (!dev_priv->l3_parity.remap_info[slice]) { temp = kzalloc(GEN7_L3LOG_SIZE, GFP_KERNEL); if (!temp) { - mutex_unlock(&drm_dev->struct_mutex); + mutex_unlock(&dev->struct_mutex); return -ENOMEM; } } @@ -240,7 +241,7 @@ i915_l3_write(struct file *filp, struct kobject *kobj, list_for_each_entry(ctx, &dev_priv->context_list, link) ctx->remap_slice |= (1<<slice); - mutex_unlock(&drm_dev->struct_mutex); + mutex_unlock(&dev->struct_mutex); return count; } @@ -266,9 +267,7 @@ static struct bin_attribute dpf_attrs_1 = { static ssize_t gt_act_freq_mhz_show(struct device *kdev, struct device_attribute *attr, char *buf) { - struct drm_minor *minor = dev_to_drm_minor(kdev); - struct drm_device *dev = minor->dev; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = kdev_minor_to_i915(kdev); int ret; intel_runtime_pm_get(dev_priv); @@ -298,9 +297,7 @@ static ssize_t gt_act_freq_mhz_show(struct device *kdev, static ssize_t gt_cur_freq_mhz_show(struct device *kdev, struct device_attribute *attr, char *buf) { - struct drm_minor *minor = dev_to_drm_minor(kdev); - struct drm_device *dev = minor->dev; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = kdev_minor_to_i915(kdev); return snprintf(buf, PAGE_SIZE, "%d\n", intel_gpu_freq(dev_priv, @@ -309,8 +306,7 @@ static ssize_t gt_cur_freq_mhz_show(struct device *kdev, static ssize_t gt_boost_freq_mhz_show(struct device *kdev, struct device_attribute *attr, char *buf) { - struct drm_minor *minor = dev_to_drm_minor(kdev); - struct drm_i915_private *dev_priv = to_i915(minor->dev); + struct drm_i915_private *dev_priv = kdev_minor_to_i915(kdev); return snprintf(buf, PAGE_SIZE, "%d\n", intel_gpu_freq(dev_priv, @@ -321,9 +317,7 @@ static ssize_t gt_boost_freq_mhz_store(struct device *kdev, struct device_attribute *attr, const char *buf, size_t count) { - struct drm_minor *minor = dev_to_drm_minor(kdev); - struct drm_device *dev = minor->dev; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = kdev_minor_to_i915(kdev); u32 val; ssize_t ret; @@ -346,9 +340,7 @@ static ssize_t gt_boost_freq_mhz_store(struct device *kdev, static ssize_t vlv_rpe_freq_mhz_show(struct device *kdev, struct device_attribute *attr, char *buf) { - struct drm_minor *minor = dev_to_drm_minor(kdev); - struct drm_device *dev = minor->dev; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = kdev_minor_to_i915(kdev); return snprintf(buf, PAGE_SIZE, "%d\n", intel_gpu_freq(dev_priv, @@ -357,9 +349,7 @@ static ssize_t vlv_rpe_freq_mhz_show(struct device *kdev, static ssize_t gt_max_freq_mhz_show(struct device *kdev, struct device_attribute *attr, char *buf) { - struct drm_minor *minor = dev_to_drm_minor(kdev); - struct drm_device *dev = minor->dev; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = kdev_minor_to_i915(kdev); return snprintf(buf, PAGE_SIZE, "%d\n", intel_gpu_freq(dev_priv, @@ -370,9 +360,7 @@ static ssize_t gt_max_freq_mhz_store(struct device *kdev, struct device_attribute *attr, const char *buf, size_t count) { - struct drm_minor *minor = dev_to_drm_minor(kdev); - struct drm_device *dev = minor->dev; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = kdev_minor_to_i915(kdev); u32 val; ssize_t ret; @@ -418,9 +406,7 @@ static ssize_t gt_max_freq_mhz_store(struct device *kdev, static ssize_t gt_min_freq_mhz_show(struct device *kdev, struct device_attribute *attr, char *buf) { - struct drm_minor *minor = dev_to_drm_minor(kdev); - struct drm_device *dev = minor->dev; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = kdev_minor_to_i915(kdev); return snprintf(buf, PAGE_SIZE, "%d\n", intel_gpu_freq(dev_priv, @@ -431,9 +417,7 @@ static ssize_t gt_min_freq_mhz_store(struct device *kdev, struct device_attribute *attr, const char *buf, size_t count) { - struct drm_minor *minor = dev_to_drm_minor(kdev); - struct drm_device *dev = minor->dev; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = kdev_minor_to_i915(kdev); u32 val; ssize_t ret; @@ -490,9 +474,7 @@ static DEVICE_ATTR(gt_RPn_freq_mhz, S_IRUGO, gt_rp_mhz_show, NULL); /* For now we have a static number of RP states */ static ssize_t gt_rp_mhz_show(struct device *kdev, struct device_attribute *attr, char *buf) { - struct drm_minor *minor = dev_to_drm_minor(kdev); - struct drm_device *dev = minor->dev; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = kdev_minor_to_i915(kdev); u32 val; if (attr == &dev_attr_gt_RP0_freq_mhz) @@ -538,8 +520,8 @@ static ssize_t error_state_read(struct file *filp, struct kobject *kobj, { struct device *kdev = kobj_to_dev(kobj); - struct drm_minor *minor = dev_to_drm_minor(kdev); - struct drm_device *dev = minor->dev; + struct drm_i915_private *dev_priv = kdev_minor_to_i915(kdev); + struct drm_device *dev = &dev_priv->drm; struct i915_error_state_file_priv error_priv; struct drm_i915_error_state_buf error_str; ssize_t ret_count = 0; @@ -573,18 +555,10 @@ static ssize_t error_state_write(struct file *file, struct kobject *kobj, loff_t off, size_t count) { struct device *kdev = kobj_to_dev(kobj); - struct drm_minor *minor = dev_to_drm_minor(kdev); - struct drm_device *dev = minor->dev; - int ret; + struct drm_i915_private *dev_priv = kdev_minor_to_i915(kdev); DRM_DEBUG_DRIVER("Resetting error state\n"); - - ret = mutex_lock_interruptible(&dev->struct_mutex); - if (ret) - return ret; - - i915_destroy_error_state(dev); - mutex_unlock(&dev->struct_mutex); + i915_destroy_error_state(&dev_priv->drm); return count; } @@ -597,37 +571,38 @@ static struct bin_attribute error_state_attr = { .write = error_state_write, }; -void i915_setup_sysfs(struct drm_device *dev) +void i915_setup_sysfs(struct drm_i915_private *dev_priv) { + struct device *kdev = dev_priv->drm.primary->kdev; int ret; #ifdef CONFIG_PM - if (HAS_RC6(dev)) { - ret = sysfs_merge_group(&dev->primary->kdev->kobj, + if (HAS_RC6(dev_priv)) { + ret = sysfs_merge_group(&kdev->kobj, &rc6_attr_group); if (ret) DRM_ERROR("RC6 residency sysfs setup failed\n"); } - if (HAS_RC6p(dev)) { - ret = sysfs_merge_group(&dev->primary->kdev->kobj, + if (HAS_RC6p(dev_priv)) { + ret = sysfs_merge_group(&kdev->kobj, &rc6p_attr_group); if (ret) DRM_ERROR("RC6p residency sysfs setup failed\n"); } - if (IS_VALLEYVIEW(dev) || IS_CHERRYVIEW(dev)) { - ret = sysfs_merge_group(&dev->primary->kdev->kobj, + if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv)) { + ret = sysfs_merge_group(&kdev->kobj, &media_rc6_attr_group); if (ret) DRM_ERROR("Media RC6 residency sysfs setup failed\n"); } #endif - if (HAS_L3_DPF(dev)) { - ret = device_create_bin_file(dev->primary->kdev, &dpf_attrs); + if (HAS_L3_DPF(dev_priv)) { + ret = device_create_bin_file(kdev, &dpf_attrs); if (ret) DRM_ERROR("l3 parity sysfs setup failed\n"); - if (NUM_L3_SLICES(dev) > 1) { - ret = device_create_bin_file(dev->primary->kdev, + if (NUM_L3_SLICES(dev_priv) > 1) { + ret = device_create_bin_file(kdev, &dpf_attrs_1); if (ret) DRM_ERROR("l3 parity slice 1 setup failed\n"); @@ -635,30 +610,32 @@ void i915_setup_sysfs(struct drm_device *dev) } ret = 0; - if (IS_VALLEYVIEW(dev) || IS_CHERRYVIEW(dev)) - ret = sysfs_create_files(&dev->primary->kdev->kobj, vlv_attrs); - else if (INTEL_INFO(dev)->gen >= 6) - ret = sysfs_create_files(&dev->primary->kdev->kobj, gen6_attrs); + if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv)) + ret = sysfs_create_files(&kdev->kobj, vlv_attrs); + else if (INTEL_GEN(dev_priv) >= 6) + ret = sysfs_create_files(&kdev->kobj, gen6_attrs); if (ret) DRM_ERROR("RPS sysfs setup failed\n"); - ret = sysfs_create_bin_file(&dev->primary->kdev->kobj, + ret = sysfs_create_bin_file(&kdev->kobj, &error_state_attr); if (ret) DRM_ERROR("error_state sysfs setup failed\n"); } -void i915_teardown_sysfs(struct drm_device *dev) +void i915_teardown_sysfs(struct drm_i915_private *dev_priv) { - sysfs_remove_bin_file(&dev->primary->kdev->kobj, &error_state_attr); - if (IS_VALLEYVIEW(dev) || IS_CHERRYVIEW(dev)) - sysfs_remove_files(&dev->primary->kdev->kobj, vlv_attrs); + struct device *kdev = dev_priv->drm.primary->kdev; + + sysfs_remove_bin_file(&kdev->kobj, &error_state_attr); + if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv)) + sysfs_remove_files(&kdev->kobj, vlv_attrs); else - sysfs_remove_files(&dev->primary->kdev->kobj, gen6_attrs); - device_remove_bin_file(dev->primary->kdev, &dpf_attrs_1); - device_remove_bin_file(dev->primary->kdev, &dpf_attrs); + sysfs_remove_files(&kdev->kobj, gen6_attrs); + device_remove_bin_file(kdev, &dpf_attrs_1); + device_remove_bin_file(kdev, &dpf_attrs); #ifdef CONFIG_PM - sysfs_unmerge_group(&dev->primary->kdev->kobj, &rc6_attr_group); - sysfs_unmerge_group(&dev->primary->kdev->kobj, &rc6p_attr_group); + sysfs_unmerge_group(&kdev->kobj, &rc6_attr_group); + sysfs_unmerge_group(&kdev->kobj, &rc6p_attr_group); #endif } diff --git a/drivers/gpu/drm/i915/i915_vgpu.c b/drivers/gpu/drm/i915/i915_vgpu.c index ca2e91259948..dae340cfc6c7 100644 --- a/drivers/gpu/drm/i915/i915_vgpu.c +++ b/drivers/gpu/drm/i915/i915_vgpu.c @@ -65,9 +65,6 @@ void i915_check_vgpu(struct drm_i915_private *dev_priv) BUILD_BUG_ON(sizeof(struct vgt_if) != VGT_PVINFO_SIZE); - if (!IS_HASWELL(dev_priv)) - return; - magic = __raw_i915_read64(dev_priv, vgtif_reg(magic)); if (magic != VGT_MAGIC) return; diff --git a/drivers/gpu/drm/i915/intel_audio.c b/drivers/gpu/drm/i915/intel_audio.c index 85389cdd0bec..6c70a5bfd7d8 100644 --- a/drivers/gpu/drm/i915/intel_audio.c +++ b/drivers/gpu/drm/i915/intel_audio.c @@ -359,9 +359,7 @@ static void ilk_audio_codec_disable(struct intel_encoder *encoder) { struct drm_i915_private *dev_priv = to_i915(encoder->base.dev); struct intel_crtc *intel_crtc = to_intel_crtc(encoder->base.crtc); - struct intel_digital_port *intel_dig_port = - enc_to_dig_port(&encoder->base); - enum port port = intel_dig_port->port; + enum port port = enc_to_dig_port(&encoder->base)->port; enum pipe pipe = intel_crtc->pipe; uint32_t tmp, eldv; i915_reg_t aud_config, aud_cntrl_st2; @@ -407,13 +405,10 @@ static void ilk_audio_codec_enable(struct drm_connector *connector, { struct drm_i915_private *dev_priv = to_i915(connector->dev); struct intel_crtc *intel_crtc = to_intel_crtc(encoder->base.crtc); - struct intel_digital_port *intel_dig_port = - enc_to_dig_port(&encoder->base); - enum port port = intel_dig_port->port; + enum port port = enc_to_dig_port(&encoder->base)->port; enum pipe pipe = intel_crtc->pipe; uint8_t *eld = connector->eld; - uint32_t eldv; - uint32_t tmp; + uint32_t tmp, eldv; int len, i; i915_reg_t hdmiw_hdmiedid, aud_config, aud_cntl_st, aud_cntrl_st2; @@ -581,26 +576,26 @@ void intel_init_audio_hooks(struct drm_i915_private *dev_priv) } } -static void i915_audio_component_get_power(struct device *dev) +static void i915_audio_component_get_power(struct device *kdev) { - intel_display_power_get(dev_to_i915(dev), POWER_DOMAIN_AUDIO); + intel_display_power_get(kdev_to_i915(kdev), POWER_DOMAIN_AUDIO); } -static void i915_audio_component_put_power(struct device *dev) +static void i915_audio_component_put_power(struct device *kdev) { - intel_display_power_put(dev_to_i915(dev), POWER_DOMAIN_AUDIO); + intel_display_power_put(kdev_to_i915(kdev), POWER_DOMAIN_AUDIO); } -static void i915_audio_component_codec_wake_override(struct device *dev, +static void i915_audio_component_codec_wake_override(struct device *kdev, bool enable) { - struct drm_i915_private *dev_priv = dev_to_i915(dev); + struct drm_i915_private *dev_priv = kdev_to_i915(kdev); u32 tmp; if (!IS_SKYLAKE(dev_priv) && !IS_KABYLAKE(dev_priv)) return; - i915_audio_component_get_power(dev); + i915_audio_component_get_power(kdev); /* * Enable/disable generating the codec wake signal, overriding the @@ -618,13 +613,13 @@ static void i915_audio_component_codec_wake_override(struct device *dev, usleep_range(1000, 1500); } - i915_audio_component_put_power(dev); + i915_audio_component_put_power(kdev); } /* Get CDCLK in kHz */ -static int i915_audio_component_get_cdclk_freq(struct device *dev) +static int i915_audio_component_get_cdclk_freq(struct device *kdev) { - struct drm_i915_private *dev_priv = dev_to_i915(dev); + struct drm_i915_private *dev_priv = kdev_to_i915(kdev); if (WARN_ON_ONCE(!HAS_DDI(dev_priv))) return -ENODEV; @@ -632,10 +627,10 @@ static int i915_audio_component_get_cdclk_freq(struct device *dev) return dev_priv->cdclk_freq; } -static int i915_audio_component_sync_audio_rate(struct device *dev, +static int i915_audio_component_sync_audio_rate(struct device *kdev, int port, int rate) { - struct drm_i915_private *dev_priv = dev_to_i915(dev); + struct drm_i915_private *dev_priv = kdev_to_i915(kdev); struct intel_encoder *intel_encoder; struct intel_crtc *crtc; struct drm_display_mode *mode; @@ -652,7 +647,7 @@ static int i915_audio_component_sync_audio_rate(struct device *dev, !IS_HASWELL(dev_priv)) return 0; - i915_audio_component_get_power(dev); + i915_audio_component_get_power(kdev); mutex_lock(&dev_priv->av_mutex); /* 1. get the pipe */ intel_encoder = dev_priv->dig_port_map[port]; @@ -703,15 +698,15 @@ static int i915_audio_component_sync_audio_rate(struct device *dev, unlock: mutex_unlock(&dev_priv->av_mutex); - i915_audio_component_put_power(dev); + i915_audio_component_put_power(kdev); return err; } -static int i915_audio_component_get_eld(struct device *dev, int port, +static int i915_audio_component_get_eld(struct device *kdev, int port, bool *enabled, unsigned char *buf, int max_bytes) { - struct drm_i915_private *dev_priv = dev_to_i915(dev); + struct drm_i915_private *dev_priv = kdev_to_i915(kdev); struct intel_encoder *intel_encoder; struct intel_digital_port *intel_dig_port; const u8 *eld; @@ -745,11 +740,11 @@ static const struct i915_audio_component_ops i915_audio_component_ops = { .get_eld = i915_audio_component_get_eld, }; -static int i915_audio_component_bind(struct device *i915_dev, - struct device *hda_dev, void *data) +static int i915_audio_component_bind(struct device *i915_kdev, + struct device *hda_kdev, void *data) { struct i915_audio_component *acomp = data; - struct drm_i915_private *dev_priv = dev_to_i915(i915_dev); + struct drm_i915_private *dev_priv = kdev_to_i915(i915_kdev); int i; if (WARN_ON(acomp->ops || acomp->dev)) @@ -757,7 +752,7 @@ static int i915_audio_component_bind(struct device *i915_dev, drm_modeset_lock_all(&dev_priv->drm); acomp->ops = &i915_audio_component_ops; - acomp->dev = i915_dev; + acomp->dev = i915_kdev; BUILD_BUG_ON(MAX_PORTS != I915_MAX_PORTS); for (i = 0; i < ARRAY_SIZE(acomp->aud_sample_rate); i++) acomp->aud_sample_rate[i] = 0; @@ -767,11 +762,11 @@ static int i915_audio_component_bind(struct device *i915_dev, return 0; } -static void i915_audio_component_unbind(struct device *i915_dev, - struct device *hda_dev, void *data) +static void i915_audio_component_unbind(struct device *i915_kdev, + struct device *hda_kdev, void *data) { struct i915_audio_component *acomp = data; - struct drm_i915_private *dev_priv = dev_to_i915(i915_dev); + struct drm_i915_private *dev_priv = kdev_to_i915(i915_kdev); drm_modeset_lock_all(&dev_priv->drm); acomp->ops = NULL; diff --git a/drivers/gpu/drm/i915/intel_breadcrumbs.c b/drivers/gpu/drm/i915/intel_breadcrumbs.c index 2491e4c1eaf0..9bad14d22c95 100644 --- a/drivers/gpu/drm/i915/intel_breadcrumbs.c +++ b/drivers/gpu/drm/i915/intel_breadcrumbs.c @@ -462,7 +462,10 @@ static int intel_breadcrumbs_signaler(void *arg) */ intel_engine_remove_wait(engine, &request->signaling.wait); + + local_bh_disable(); fence_signal(&request->fence); + local_bh_enable(); /* kick start the tasklets */ /* Find the next oldest signal. Note that as we have * not been holding the lock, another client may diff --git a/drivers/gpu/drm/i915/intel_color.c b/drivers/gpu/drm/i915/intel_color.c index bc0fef3d3335..95a72771eea6 100644 --- a/drivers/gpu/drm/i915/intel_color.c +++ b/drivers/gpu/drm/i915/intel_color.c @@ -100,13 +100,14 @@ static void i9xx_load_csc_matrix(struct drm_crtc_state *crtc_state) struct intel_crtc *intel_crtc = to_intel_crtc(crtc); int i, pipe = intel_crtc->pipe; uint16_t coeffs[9] = { 0, }; + struct intel_crtc_state *intel_crtc_state = to_intel_crtc_state(crtc_state); if (crtc_state->ctm) { struct drm_color_ctm *ctm = (struct drm_color_ctm *)crtc_state->ctm->data; uint64_t input[9] = { 0, }; - if (intel_crtc->config->limited_color_range) { + if (intel_crtc_state->limited_color_range) { ctm_mult_by_limited(input, ctm->matrix); } else { for (i = 0; i < ARRAY_SIZE(input); i++) @@ -158,7 +159,7 @@ static void i9xx_load_csc_matrix(struct drm_crtc_state *crtc_state) * into consideration. */ for (i = 0; i < 3; i++) { - if (intel_crtc->config->limited_color_range) + if (intel_crtc_state->limited_color_range) coeffs[i * 3 + i] = I9XX_CSC_COEFF_LIMITED_RANGE; else @@ -182,7 +183,7 @@ static void i9xx_load_csc_matrix(struct drm_crtc_state *crtc_state) if (INTEL_INFO(dev)->gen > 6) { uint16_t postoff = 0; - if (intel_crtc->config->limited_color_range) + if (intel_crtc_state->limited_color_range) postoff = (16 * (1 << 12) / 255) & 0x1fff; I915_WRITE(PIPE_CSC_POSTOFF_HI(pipe), postoff); @@ -193,7 +194,7 @@ static void i9xx_load_csc_matrix(struct drm_crtc_state *crtc_state) } else { uint32_t mode = CSC_MODE_YUV_TO_RGB; - if (intel_crtc->config->limited_color_range) + if (intel_crtc_state->limited_color_range) mode |= CSC_BLACK_SCREEN_OFFSET; I915_WRITE(PIPE_CSC_MODE(pipe), mode); @@ -263,7 +264,8 @@ void intel_color_set_csc(struct drm_crtc_state *crtc_state) /* Loads the legacy palette/gamma unit for the CRTC. */ static void i9xx_load_luts_internal(struct drm_crtc *crtc, - struct drm_property_blob *blob) + struct drm_property_blob *blob, + struct intel_crtc_state *crtc_state) { struct drm_device *dev = crtc->dev; struct drm_i915_private *dev_priv = to_i915(dev); @@ -272,7 +274,7 @@ static void i9xx_load_luts_internal(struct drm_crtc *crtc, int i; if (HAS_GMCH_DISPLAY(dev)) { - if (intel_crtc_has_type(intel_crtc->config, INTEL_OUTPUT_DSI)) + if (intel_crtc_has_type(crtc_state, INTEL_OUTPUT_DSI)) assert_dsi_pll_enabled(dev_priv); else assert_pll_enabled(dev_priv, pipe); @@ -305,7 +307,8 @@ static void i9xx_load_luts_internal(struct drm_crtc *crtc, static void i9xx_load_luts(struct drm_crtc_state *crtc_state) { - i9xx_load_luts_internal(crtc_state->crtc, crtc_state->gamma_lut); + i9xx_load_luts_internal(crtc_state->crtc, crtc_state->gamma_lut, + to_intel_crtc_state(crtc_state)); } /* Loads the legacy palette/gamma unit for the CRTC on Haswell. */ @@ -323,7 +326,7 @@ static void haswell_load_luts(struct drm_crtc_state *crtc_state) * Workaround : Do not read or write the pipe palette/gamma data while * GAMMA_MODE is configured for split gamma and IPS_CTL has IPS enabled. */ - if (IS_HASWELL(dev) && intel_crtc->config->ips_enabled && + if (IS_HASWELL(dev) && intel_crtc_state->ips_enabled && (intel_crtc_state->gamma_mode == GAMMA_MODE_MODE_SPLIT)) { hsw_disable_ips(intel_crtc); reenable_ips = true; @@ -436,7 +439,8 @@ static void cherryview_load_luts(struct drm_crtc_state *state) /* Turn off degamma/gamma on CGM block. */ I915_WRITE(CGM_PIPE_MODE(pipe), (state->ctm ? CGM_PIPE_MODE_CSC : 0)); - i9xx_load_luts_internal(crtc, state->gamma_lut); + i9xx_load_luts_internal(crtc, state->gamma_lut, + to_intel_crtc_state(state)); return; } @@ -479,7 +483,7 @@ static void cherryview_load_luts(struct drm_crtc_state *state) * Also program a linear LUT in the legacy block (behind the * CGM block). */ - i9xx_load_luts_internal(crtc, NULL); + i9xx_load_luts_internal(crtc, NULL, to_intel_crtc_state(state)); } void intel_color_load_luts(struct drm_crtc_state *crtc_state) diff --git a/drivers/gpu/drm/i915/intel_crt.c b/drivers/gpu/drm/i915/intel_crt.c index 827b6ef4e9ae..dfbcf16b41df 100644 --- a/drivers/gpu/drm/i915/intel_crt.c +++ b/drivers/gpu/drm/i915/intel_crt.c @@ -143,13 +143,15 @@ static void hsw_crt_get_config(struct intel_encoder *encoder, /* Note: The caller is required to filter out dpms modes not supported by the * platform. */ -static void intel_crt_set_dpms(struct intel_encoder *encoder, int mode) +static void intel_crt_set_dpms(struct intel_encoder *encoder, + struct intel_crtc_state *crtc_state, + int mode) { struct drm_device *dev = encoder->base.dev; struct drm_i915_private *dev_priv = to_i915(dev); struct intel_crt *crt = intel_encoder_to_crt(encoder); - struct intel_crtc *crtc = to_intel_crtc(encoder->base.crtc); - const struct drm_display_mode *adjusted_mode = &crtc->config->base.adjusted_mode; + struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc); + const struct drm_display_mode *adjusted_mode = &crtc_state->base.adjusted_mode; u32 adpa; if (INTEL_INFO(dev)->gen >= 5) @@ -193,23 +195,45 @@ static void intel_crt_set_dpms(struct intel_encoder *encoder, int mode) I915_WRITE(crt->adpa_reg, adpa); } -static void intel_disable_crt(struct intel_encoder *encoder) +static void intel_disable_crt(struct intel_encoder *encoder, + struct intel_crtc_state *old_crtc_state, + struct drm_connector_state *old_conn_state) { - intel_crt_set_dpms(encoder, DRM_MODE_DPMS_OFF); + intel_crt_set_dpms(encoder, old_crtc_state, DRM_MODE_DPMS_OFF); } -static void pch_disable_crt(struct intel_encoder *encoder) +static void pch_disable_crt(struct intel_encoder *encoder, + struct intel_crtc_state *old_crtc_state, + struct drm_connector_state *old_conn_state) { } -static void pch_post_disable_crt(struct intel_encoder *encoder) +static void pch_post_disable_crt(struct intel_encoder *encoder, + struct intel_crtc_state *old_crtc_state, + struct drm_connector_state *old_conn_state) { - intel_disable_crt(encoder); + intel_disable_crt(encoder, old_crtc_state, old_conn_state); } -static void intel_enable_crt(struct intel_encoder *encoder) +static void hsw_post_disable_crt(struct intel_encoder *encoder, + struct intel_crtc_state *old_crtc_state, + struct drm_connector_state *old_conn_state) { - intel_crt_set_dpms(encoder, DRM_MODE_DPMS_ON); + struct drm_i915_private *dev_priv = to_i915(encoder->base.dev); + + pch_post_disable_crt(encoder, old_crtc_state, old_conn_state); + + lpt_disable_pch_transcoder(dev_priv); + lpt_disable_iclkip(dev_priv); + + intel_ddi_fdi_post_disable(encoder, old_crtc_state, old_conn_state); +} + +static void intel_enable_crt(struct intel_encoder *encoder, + struct intel_crtc_state *pipe_config, + struct drm_connector_state *conn_state) +{ + intel_crt_set_dpms(encoder, pipe_config, DRM_MODE_DPMS_ON); } static enum drm_mode_status @@ -253,7 +277,8 @@ intel_crt_mode_valid(struct drm_connector *connector, } static bool intel_crt_compute_config(struct intel_encoder *encoder, - struct intel_crtc_state *pipe_config) + struct intel_crtc_state *pipe_config, + struct drm_connector_state *conn_state) { struct drm_device *dev = encoder->base.dev; @@ -894,6 +919,7 @@ void intel_crt_init(struct drm_device *dev) if (HAS_DDI(dev)) { crt->base.get_config = hsw_crt_get_config; crt->base.get_hw_state = intel_ddi_get_hw_state; + crt->base.post_disable = hsw_post_disable_crt; } else { crt->base.get_config = intel_crt_get_config; crt->base.get_hw_state = intel_crt_get_hw_state; diff --git a/drivers/gpu/drm/i915/intel_csr.c b/drivers/gpu/drm/i915/intel_csr.c index fb27d187876c..1ea0e1f43397 100644 --- a/drivers/gpu/drm/i915/intel_csr.c +++ b/drivers/gpu/drm/i915/intel_csr.c @@ -34,15 +34,15 @@ * low-power state and comes back to normal. */ -#define I915_CSR_KBL "i915/kbl_dmc_ver1.bin" +#define I915_CSR_KBL "i915/kbl_dmc_ver1_01.bin" MODULE_FIRMWARE(I915_CSR_KBL); #define KBL_CSR_VERSION_REQUIRED CSR_VERSION(1, 1) -#define I915_CSR_SKL "i915/skl_dmc_ver1.bin" +#define I915_CSR_SKL "i915/skl_dmc_ver1_26.bin" MODULE_FIRMWARE(I915_CSR_SKL); -#define SKL_CSR_VERSION_REQUIRED CSR_VERSION(1, 23) +#define SKL_CSR_VERSION_REQUIRED CSR_VERSION(1, 26) -#define I915_CSR_BXT "i915/bxt_dmc_ver1.bin" +#define I915_CSR_BXT "i915/bxt_dmc_ver1_07.bin" MODULE_FIRMWARE(I915_CSR_BXT); #define BXT_CSR_VERSION_REQUIRED CSR_VERSION(1, 7) diff --git a/drivers/gpu/drm/i915/intel_ddi.c b/drivers/gpu/drm/i915/intel_ddi.c index c2df4e429b19..15d47c87def6 100644 --- a/drivers/gpu/drm/i915/intel_ddi.c +++ b/drivers/gpu/drm/i915/intel_ddi.c @@ -546,6 +546,27 @@ static void intel_wait_ddi_buf_idle(struct drm_i915_private *dev_priv, DRM_ERROR("Timeout waiting for DDI BUF %c idle bit\n", port_name(port)); } +static uint32_t hsw_pll_to_ddi_pll_sel(struct intel_shared_dpll *pll) +{ + switch (pll->id) { + case DPLL_ID_WRPLL1: + return PORT_CLK_SEL_WRPLL1; + case DPLL_ID_WRPLL2: + return PORT_CLK_SEL_WRPLL2; + case DPLL_ID_SPLL: + return PORT_CLK_SEL_SPLL; + case DPLL_ID_LCPLL_810: + return PORT_CLK_SEL_LCPLL_810; + case DPLL_ID_LCPLL_1350: + return PORT_CLK_SEL_LCPLL_1350; + case DPLL_ID_LCPLL_2700: + return PORT_CLK_SEL_LCPLL_2700; + default: + MISSING_CASE(pll->id); + return PORT_CLK_SEL_NONE; + } +} + /* Starting with Haswell, different DDI ports can work in FDI mode for * connection to the PCH-located connectors. For this, it is necessary to train * both the DDI port and PCH receiver for the desired DDI buffer settings. @@ -561,7 +582,7 @@ void hsw_fdi_link_train(struct drm_crtc *crtc) struct drm_i915_private *dev_priv = to_i915(dev); struct intel_crtc *intel_crtc = to_intel_crtc(crtc); struct intel_encoder *encoder; - u32 temp, i, rx_ctl_val; + u32 temp, i, rx_ctl_val, ddi_pll_sel; for_each_encoder_on_crtc(dev, crtc, encoder) { WARN_ON(encoder->type != INTEL_OUTPUT_ANALOG); @@ -592,8 +613,9 @@ void hsw_fdi_link_train(struct drm_crtc *crtc) I915_WRITE(FDI_RX_CTL(PIPE_A), rx_ctl_val); /* Configure Port Clock Select */ - I915_WRITE(PORT_CLK_SEL(PORT_E), intel_crtc->config->ddi_pll_sel); - WARN_ON(intel_crtc->config->ddi_pll_sel != PORT_CLK_SEL_SPLL); + ddi_pll_sel = hsw_pll_to_ddi_pll_sel(intel_crtc->config->shared_dpll); + I915_WRITE(PORT_CLK_SEL(PORT_E), ddi_pll_sel); + WARN_ON(ddi_pll_sel != PORT_CLK_SEL_SPLL); /* Start the training iterating through available voltages and emphasis, * testing each value twice. */ @@ -870,7 +892,7 @@ static void skl_ddi_clock_get(struct intel_encoder *encoder, int link_clock = 0; uint32_t dpll_ctl1, dpll; - dpll = pipe_config->ddi_pll_sel; + dpll = intel_get_shared_dpll_id(dev_priv, pipe_config->shared_dpll); dpll_ctl1 = I915_READ(DPLL_CTRL1); @@ -918,7 +940,7 @@ static void hsw_ddi_clock_get(struct intel_encoder *encoder, int link_clock = 0; u32 val, pll; - val = pipe_config->ddi_pll_sel; + val = hsw_pll_to_ddi_pll_sel(pipe_config->shared_dpll); switch (val & PORT_CLK_SEL_MASK) { case PORT_CLK_SEL_LCPLL_810: link_clock = 81000; @@ -1586,13 +1608,15 @@ uint32_t ddi_signal_levels(struct intel_dp *intel_dp) } void intel_ddi_clk_select(struct intel_encoder *encoder, - const struct intel_crtc_state *pipe_config) + struct intel_shared_dpll *pll) { struct drm_i915_private *dev_priv = to_i915(encoder->base.dev); enum port port = intel_ddi_get_encoder_port(encoder); + if (WARN_ON(!pll)) + return; + if (IS_SKYLAKE(dev_priv) || IS_KABYLAKE(dev_priv)) { - uint32_t dpll = pipe_config->ddi_pll_sel; uint32_t val; /* DDI -> PLL mapping */ @@ -1600,70 +1624,91 @@ void intel_ddi_clk_select(struct intel_encoder *encoder, val &= ~(DPLL_CTRL2_DDI_CLK_OFF(port) | DPLL_CTRL2_DDI_CLK_SEL_MASK(port)); - val |= (DPLL_CTRL2_DDI_CLK_SEL(dpll, port) | + val |= (DPLL_CTRL2_DDI_CLK_SEL(pll->id, port) | DPLL_CTRL2_DDI_SEL_OVERRIDE(port)); I915_WRITE(DPLL_CTRL2, val); } else if (INTEL_INFO(dev_priv)->gen < 9) { - WARN_ON(pipe_config->ddi_pll_sel == PORT_CLK_SEL_NONE); - I915_WRITE(PORT_CLK_SEL(port), pipe_config->ddi_pll_sel); + I915_WRITE(PORT_CLK_SEL(port), hsw_pll_to_ddi_pll_sel(pll)); } } -static void intel_ddi_pre_enable(struct intel_encoder *intel_encoder) +static void intel_ddi_pre_enable_dp(struct intel_encoder *encoder, + int link_rate, uint32_t lane_count, + struct intel_shared_dpll *pll, + bool link_mst) { - struct drm_encoder *encoder = &intel_encoder->base; - struct drm_i915_private *dev_priv = to_i915(encoder->dev); - struct intel_crtc *crtc = to_intel_crtc(encoder->crtc); - enum port port = intel_ddi_get_encoder_port(intel_encoder); - int type = intel_encoder->type; - - if (type == INTEL_OUTPUT_HDMI) { - struct intel_hdmi *intel_hdmi = enc_to_intel_hdmi(encoder); - - intel_dp_dual_mode_set_tmds_output(intel_hdmi, true); - } + struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base); + struct drm_i915_private *dev_priv = to_i915(encoder->base.dev); + enum port port = intel_ddi_get_encoder_port(encoder); - if (type == INTEL_OUTPUT_EDP) { - struct intel_dp *intel_dp = enc_to_intel_dp(encoder); + intel_dp_set_link_params(intel_dp, link_rate, lane_count, + link_mst); + if (encoder->type == INTEL_OUTPUT_EDP) intel_edp_panel_on(intel_dp); - } - - intel_ddi_clk_select(intel_encoder, crtc->config); - if (type == INTEL_OUTPUT_DP || type == INTEL_OUTPUT_EDP) { - struct intel_dp *intel_dp = enc_to_intel_dp(encoder); - - intel_prepare_dp_ddi_buffers(intel_encoder); - - intel_dp_set_link_params(intel_dp, crtc->config); + intel_ddi_clk_select(encoder, pll); + intel_prepare_dp_ddi_buffers(encoder); + intel_ddi_init_dp_buf_reg(encoder); + intel_dp_sink_dpms(intel_dp, DRM_MODE_DPMS_ON); + intel_dp_start_link_train(intel_dp); + if (port != PORT_A || INTEL_GEN(dev_priv) >= 9) + intel_dp_stop_link_train(intel_dp); +} - intel_ddi_init_dp_buf_reg(intel_encoder); +static void intel_ddi_pre_enable_hdmi(struct intel_encoder *encoder, + bool has_hdmi_sink, + struct drm_display_mode *adjusted_mode, + struct intel_shared_dpll *pll) +{ + struct intel_hdmi *intel_hdmi = enc_to_intel_hdmi(&encoder->base); + struct drm_i915_private *dev_priv = to_i915(encoder->base.dev); + struct drm_encoder *drm_encoder = &encoder->base; + enum port port = intel_ddi_get_encoder_port(encoder); + int level = intel_ddi_hdmi_level(dev_priv, port); - intel_dp_sink_dpms(intel_dp, DRM_MODE_DPMS_ON); - intel_dp_start_link_train(intel_dp); - if (port != PORT_A || INTEL_INFO(dev_priv)->gen >= 9) - intel_dp_stop_link_train(intel_dp); - } else if (type == INTEL_OUTPUT_HDMI) { - struct intel_hdmi *intel_hdmi = enc_to_intel_hdmi(encoder); - int level = intel_ddi_hdmi_level(dev_priv, port); + intel_dp_dual_mode_set_tmds_output(intel_hdmi, true); + intel_ddi_clk_select(encoder, pll); + intel_prepare_hdmi_ddi_buffers(encoder); + if (IS_SKYLAKE(dev_priv) || IS_KABYLAKE(dev_priv)) + skl_ddi_set_iboost(encoder, level); + else if (IS_BROXTON(dev_priv)) + bxt_ddi_vswing_sequence(dev_priv, level, port, + INTEL_OUTPUT_HDMI); - intel_prepare_hdmi_ddi_buffers(intel_encoder); + intel_hdmi->set_infoframes(drm_encoder, + has_hdmi_sink, + adjusted_mode); +} - if (IS_SKYLAKE(dev_priv) || IS_KABYLAKE(dev_priv)) - skl_ddi_set_iboost(intel_encoder, level); - else if (IS_BROXTON(dev_priv)) - bxt_ddi_vswing_sequence(dev_priv, level, port, - INTEL_OUTPUT_HDMI); +static void intel_ddi_pre_enable(struct intel_encoder *intel_encoder, + struct intel_crtc_state *pipe_config, + struct drm_connector_state *conn_state) +{ + struct drm_encoder *encoder = &intel_encoder->base; + struct intel_crtc *crtc = to_intel_crtc(encoder->crtc); + int type = intel_encoder->type; - intel_hdmi->set_infoframes(encoder, - crtc->config->has_hdmi_sink, - &crtc->config->base.adjusted_mode); + if (type == INTEL_OUTPUT_DP || type == INTEL_OUTPUT_EDP) { + intel_ddi_pre_enable_dp(intel_encoder, + crtc->config->port_clock, + crtc->config->lane_count, + crtc->config->shared_dpll, + intel_crtc_has_type(crtc->config, + INTEL_OUTPUT_DP_MST)); + } + if (type == INTEL_OUTPUT_HDMI) { + intel_ddi_pre_enable_hdmi(intel_encoder, + crtc->config->has_hdmi_sink, + &crtc->config->base.adjusted_mode, + crtc->config->shared_dpll); } } -static void intel_ddi_post_disable(struct intel_encoder *intel_encoder) +static void intel_ddi_post_disable(struct intel_encoder *intel_encoder, + struct intel_crtc_state *old_crtc_state, + struct drm_connector_state *old_conn_state) { struct drm_encoder *encoder = &intel_encoder->base; struct drm_device *dev = encoder->dev; @@ -1673,6 +1718,8 @@ static void intel_ddi_post_disable(struct intel_encoder *intel_encoder) uint32_t val; bool wait = false; + /* old_crtc_state and old_conn_state are NULL when called from DP_MST */ + val = I915_READ(DDI_BUF_CTL(port)); if (val & DDI_BUF_CTL_ENABLE) { val &= ~DDI_BUF_CTL_ENABLE; @@ -1708,7 +1755,42 @@ static void intel_ddi_post_disable(struct intel_encoder *intel_encoder) } } -static void intel_enable_ddi(struct intel_encoder *intel_encoder) +void intel_ddi_fdi_post_disable(struct intel_encoder *intel_encoder, + struct intel_crtc_state *old_crtc_state, + struct drm_connector_state *old_conn_state) +{ + struct drm_i915_private *dev_priv = to_i915(intel_encoder->base.dev); + uint32_t val; + + /* + * Bspec lists this as both step 13 (before DDI_BUF_CTL disable) + * and step 18 (after clearing PORT_CLK_SEL). Based on a BUN, + * step 13 is the correct place for it. Step 18 is where it was + * originally before the BUN. + */ + val = I915_READ(FDI_RX_CTL(PIPE_A)); + val &= ~FDI_RX_ENABLE; + I915_WRITE(FDI_RX_CTL(PIPE_A), val); + + intel_ddi_post_disable(intel_encoder, old_crtc_state, old_conn_state); + + val = I915_READ(FDI_RX_MISC(PIPE_A)); + val &= ~(FDI_RX_PWRDN_LANE1_MASK | FDI_RX_PWRDN_LANE0_MASK); + val |= FDI_RX_PWRDN_LANE1_VAL(2) | FDI_RX_PWRDN_LANE0_VAL(2); + I915_WRITE(FDI_RX_MISC(PIPE_A), val); + + val = I915_READ(FDI_RX_CTL(PIPE_A)); + val &= ~FDI_PCDCLK; + I915_WRITE(FDI_RX_CTL(PIPE_A), val); + + val = I915_READ(FDI_RX_CTL(PIPE_A)); + val &= ~FDI_RX_PLL_ENABLE; + I915_WRITE(FDI_RX_CTL(PIPE_A), val); +} + +static void intel_enable_ddi(struct intel_encoder *intel_encoder, + struct intel_crtc_state *pipe_config, + struct drm_connector_state *conn_state) { struct drm_encoder *encoder = &intel_encoder->base; struct drm_crtc *crtc = encoder->crtc; @@ -1737,7 +1819,7 @@ static void intel_enable_ddi(struct intel_encoder *intel_encoder) intel_edp_backlight_on(intel_dp); intel_psr_enable(intel_dp); - intel_edp_drrs_enable(intel_dp); + intel_edp_drrs_enable(intel_dp, pipe_config); } if (intel_crtc->config->has_audio) { @@ -1746,7 +1828,9 @@ static void intel_enable_ddi(struct intel_encoder *intel_encoder) } } -static void intel_disable_ddi(struct intel_encoder *intel_encoder) +static void intel_disable_ddi(struct intel_encoder *intel_encoder, + struct intel_crtc_state *old_crtc_state, + struct drm_connector_state *old_conn_state) { struct drm_encoder *encoder = &intel_encoder->base; struct drm_crtc *crtc = encoder->crtc; @@ -1763,7 +1847,7 @@ static void intel_disable_ddi(struct intel_encoder *intel_encoder) if (type == INTEL_OUTPUT_EDP) { struct intel_dp *intel_dp = enc_to_intel_dp(encoder); - intel_edp_drrs_disable(intel_dp); + intel_edp_drrs_disable(intel_dp, old_crtc_state); intel_psr_disable(intel_dp); intel_edp_backlight_off(intel_dp); } @@ -2052,7 +2136,9 @@ bxt_ddi_phy_calc_lane_lat_optim_mask(struct intel_encoder *encoder, } } -static void bxt_ddi_pre_pll_enable(struct intel_encoder *encoder) +static void bxt_ddi_pre_pll_enable(struct intel_encoder *encoder, + struct intel_crtc_state *pipe_config, + struct drm_connector_state *conn_state) { struct intel_digital_port *dport = enc_to_dig_port(&encoder->base); struct drm_i915_private *dev_priv = to_i915(dport->base.base.dev); @@ -2141,38 +2227,6 @@ void intel_ddi_prepare_link_retrain(struct intel_dp *intel_dp) udelay(600); } -void intel_ddi_fdi_disable(struct drm_crtc *crtc) -{ - struct drm_i915_private *dev_priv = to_i915(crtc->dev); - struct intel_encoder *intel_encoder = intel_ddi_get_crtc_encoder(crtc); - uint32_t val; - - /* - * Bspec lists this as both step 13 (before DDI_BUF_CTL disable) - * and step 18 (after clearing PORT_CLK_SEL). Based on a BUN, - * step 13 is the correct place for it. Step 18 is where it was - * originally before the BUN. - */ - val = I915_READ(FDI_RX_CTL(PIPE_A)); - val &= ~FDI_RX_ENABLE; - I915_WRITE(FDI_RX_CTL(PIPE_A), val); - - intel_ddi_post_disable(intel_encoder); - - val = I915_READ(FDI_RX_MISC(PIPE_A)); - val &= ~(FDI_RX_PWRDN_LANE1_MASK | FDI_RX_PWRDN_LANE0_MASK); - val |= FDI_RX_PWRDN_LANE1_VAL(2) | FDI_RX_PWRDN_LANE0_VAL(2); - I915_WRITE(FDI_RX_MISC(PIPE_A), val); - - val = I915_READ(FDI_RX_CTL(PIPE_A)); - val &= ~FDI_PCDCLK; - I915_WRITE(FDI_RX_CTL(PIPE_A), val); - - val = I915_READ(FDI_RX_CTL(PIPE_A)); - val &= ~FDI_RX_PLL_ENABLE; - I915_WRITE(FDI_RX_CTL(PIPE_A), val); -} - void intel_ddi_get_config(struct intel_encoder *encoder, struct intel_crtc_state *pipe_config) { @@ -2272,7 +2326,8 @@ void intel_ddi_get_config(struct intel_encoder *encoder, } static bool intel_ddi_compute_config(struct intel_encoder *encoder, - struct intel_crtc_state *pipe_config) + struct intel_crtc_state *pipe_config, + struct drm_connector_state *conn_state) { struct drm_i915_private *dev_priv = to_i915(encoder->base.dev); int type = encoder->type; @@ -2285,9 +2340,9 @@ static bool intel_ddi_compute_config(struct intel_encoder *encoder, pipe_config->cpu_transcoder = TRANSCODER_EDP; if (type == INTEL_OUTPUT_HDMI) - ret = intel_hdmi_compute_config(encoder, pipe_config); + ret = intel_hdmi_compute_config(encoder, pipe_config, conn_state); else - ret = intel_dp_compute_config(encoder, pipe_config); + ret = intel_dp_compute_config(encoder, pipe_config, conn_state); if (IS_BROXTON(dev_priv) && ret) pipe_config->lane_lat_optim_mask = @@ -2338,6 +2393,45 @@ intel_ddi_init_hdmi_connector(struct intel_digital_port *intel_dig_port) return connector; } +struct intel_shared_dpll * +intel_ddi_get_link_dpll(struct intel_dp *intel_dp, int clock) +{ + struct intel_connector *connector = intel_dp->attached_connector; + struct intel_encoder *encoder = connector->encoder; + struct drm_i915_private *dev_priv = to_i915(encoder->base.dev); + struct intel_digital_port *dig_port = dp_to_dig_port(intel_dp); + struct intel_shared_dpll *pll = NULL; + struct intel_shared_dpll_config tmp_pll_config; + enum intel_dpll_id dpll_id; + + if (IS_BROXTON(dev_priv)) { + dpll_id = (enum intel_dpll_id)dig_port->port; + /* + * Select the required PLL. This works for platforms where + * there is no shared DPLL. + */ + pll = &dev_priv->shared_dplls[dpll_id]; + if (WARN_ON(pll->active_mask)) { + + DRM_ERROR("Shared DPLL in use. active_mask:%x\n", + pll->active_mask); + return NULL; + } + tmp_pll_config = pll->config; + if (!bxt_ddi_dp_set_dpll_hw_state(clock, + &pll->config.hw_state)) { + DRM_ERROR("Could not setup DPLL\n"); + pll->config = tmp_pll_config; + return NULL; + } + } else if (IS_SKYLAKE(dev_priv) || IS_KABYLAKE(dev_priv)) { + pll = skl_find_link_pll(dev_priv, clock); + } else if (IS_HASWELL(dev_priv) || IS_BROADWELL(dev_priv)) { + pll = hsw_ddi_dp_get_dpll(encoder, clock); + } + return pll; +} + void intel_ddi_init(struct drm_device *dev, enum port port) { struct drm_i915_private *dev_priv = to_i915(dev); diff --git a/drivers/gpu/drm/i915/intel_device_info.c b/drivers/gpu/drm/i915/intel_device_info.c index cba137f9ad3e..73b6858600ac 100644 --- a/drivers/gpu/drm/i915/intel_device_info.c +++ b/drivers/gpu/drm/i915/intel_device_info.c @@ -46,71 +46,70 @@ void intel_device_info_dump(struct drm_i915_private *dev_priv) static void cherryview_sseu_info_init(struct drm_i915_private *dev_priv) { - struct intel_device_info *info = mkwrite_device_info(dev_priv); + struct sseu_dev_info *sseu = &mkwrite_device_info(dev_priv)->sseu; u32 fuse, eu_dis; fuse = I915_READ(CHV_FUSE_GT); - info->slice_total = 1; + sseu->slice_mask = BIT(0); if (!(fuse & CHV_FGT_DISABLE_SS0)) { - info->subslice_per_slice++; + sseu->subslice_mask |= BIT(0); eu_dis = fuse & (CHV_FGT_EU_DIS_SS0_R0_MASK | CHV_FGT_EU_DIS_SS0_R1_MASK); - info->eu_total += 8 - hweight32(eu_dis); + sseu->eu_total += 8 - hweight32(eu_dis); } if (!(fuse & CHV_FGT_DISABLE_SS1)) { - info->subslice_per_slice++; + sseu->subslice_mask |= BIT(1); eu_dis = fuse & (CHV_FGT_EU_DIS_SS1_R0_MASK | CHV_FGT_EU_DIS_SS1_R1_MASK); - info->eu_total += 8 - hweight32(eu_dis); + sseu->eu_total += 8 - hweight32(eu_dis); } - info->subslice_total = info->subslice_per_slice; /* * CHV expected to always have a uniform distribution of EU * across subslices. */ - info->eu_per_subslice = info->subslice_total ? - info->eu_total / info->subslice_total : + sseu->eu_per_subslice = sseu_subslice_total(sseu) ? + sseu->eu_total / sseu_subslice_total(sseu) : 0; /* * CHV supports subslice power gating on devices with more than * one subslice, and supports EU power gating on devices with * more than one EU pair per subslice. */ - info->has_slice_pg = 0; - info->has_subslice_pg = (info->subslice_total > 1); - info->has_eu_pg = (info->eu_per_subslice > 2); + sseu->has_slice_pg = 0; + sseu->has_subslice_pg = sseu_subslice_total(sseu) > 1; + sseu->has_eu_pg = (sseu->eu_per_subslice > 2); } static void gen9_sseu_info_init(struct drm_i915_private *dev_priv) { struct intel_device_info *info = mkwrite_device_info(dev_priv); + struct sseu_dev_info *sseu = &info->sseu; int s_max = 3, ss_max = 4, eu_max = 8; int s, ss; - u32 fuse2, s_enable, ss_disable, eu_disable; + u32 fuse2, eu_disable; u8 eu_mask = 0xff; fuse2 = I915_READ(GEN8_FUSE2); - s_enable = (fuse2 & GEN8_F2_S_ENA_MASK) >> GEN8_F2_S_ENA_SHIFT; - ss_disable = (fuse2 & GEN9_F2_SS_DIS_MASK) >> GEN9_F2_SS_DIS_SHIFT; + sseu->slice_mask = (fuse2 & GEN8_F2_S_ENA_MASK) >> GEN8_F2_S_ENA_SHIFT; - info->slice_total = hweight32(s_enable); /* * The subslice disable field is global, i.e. it applies * to each of the enabled slices. */ - info->subslice_per_slice = ss_max - hweight32(ss_disable); - info->subslice_total = info->slice_total * info->subslice_per_slice; + sseu->subslice_mask = (1 << ss_max) - 1; + sseu->subslice_mask &= ~((fuse2 & GEN9_F2_SS_DIS_MASK) >> + GEN9_F2_SS_DIS_SHIFT); /* * Iterate through enabled slices and subslices to * count the total enabled EU. */ for (s = 0; s < s_max; s++) { - if (!(s_enable & BIT(s))) + if (!(sseu->slice_mask & BIT(s))) /* skip disabled slice */ continue; @@ -118,7 +117,7 @@ static void gen9_sseu_info_init(struct drm_i915_private *dev_priv) for (ss = 0; ss < ss_max; ss++) { int eu_per_ss; - if (ss_disable & BIT(ss)) + if (!(sseu->subslice_mask & BIT(ss))) /* skip disabled subslice */ continue; @@ -131,9 +130,9 @@ static void gen9_sseu_info_init(struct drm_i915_private *dev_priv) * subslices if they are unbalanced. */ if (eu_per_ss == 7) - info->subslice_7eu[s] |= BIT(ss); + sseu->subslice_7eu[s] |= BIT(ss); - info->eu_total += eu_per_ss; + sseu->eu_total += eu_per_ss; } } @@ -144,9 +143,9 @@ static void gen9_sseu_info_init(struct drm_i915_private *dev_priv) * recovery. BXT is expected to be perfectly uniform in EU * distribution. */ - info->eu_per_subslice = info->subslice_total ? - DIV_ROUND_UP(info->eu_total, - info->subslice_total) : 0; + sseu->eu_per_subslice = sseu_subslice_total(sseu) ? + DIV_ROUND_UP(sseu->eu_total, + sseu_subslice_total(sseu)) : 0; /* * SKL supports slice power gating on devices with more than * one slice, and supports EU power gating on devices with @@ -155,15 +154,15 @@ static void gen9_sseu_info_init(struct drm_i915_private *dev_priv) * supports EU power gating on devices with more than one EU * pair per subslice. */ - info->has_slice_pg = + sseu->has_slice_pg = (IS_SKYLAKE(dev_priv) || IS_KABYLAKE(dev_priv)) && - info->slice_total > 1; - info->has_subslice_pg = - IS_BROXTON(dev_priv) && info->subslice_total > 1; - info->has_eu_pg = info->eu_per_subslice > 2; + hweight8(sseu->slice_mask) > 1; + sseu->has_subslice_pg = + IS_BROXTON(dev_priv) && sseu_subslice_total(sseu) > 1; + sseu->has_eu_pg = sseu->eu_per_subslice > 2; if (IS_BROXTON(dev_priv)) { -#define IS_SS_DISABLED(_ss_disable, ss) (_ss_disable & BIT(ss)) +#define IS_SS_DISABLED(ss) (!(sseu->subslice_mask & BIT(ss))) /* * There is a HW issue in 2x6 fused down parts that requires * Pooled EU to be enabled as a WA. The pool configuration @@ -171,19 +170,18 @@ static void gen9_sseu_info_init(struct drm_i915_private *dev_priv) * doesn't affect if the device has all 3 subslices enabled. */ /* WaEnablePooledEuFor2x6:bxt */ - info->has_pooled_eu = ((info->subslice_per_slice == 3) || - (info->subslice_per_slice == 2 && + info->has_pooled_eu = ((hweight8(sseu->subslice_mask) == 3) || + (hweight8(sseu->subslice_mask) == 2 && INTEL_REVID(dev_priv) < BXT_REVID_C0)); - info->min_eu_in_pool = 0; + sseu->min_eu_in_pool = 0; if (info->has_pooled_eu) { - if (IS_SS_DISABLED(ss_disable, 0) || - IS_SS_DISABLED(ss_disable, 2)) - info->min_eu_in_pool = 3; - else if (IS_SS_DISABLED(ss_disable, 1)) - info->min_eu_in_pool = 6; + if (IS_SS_DISABLED(2) || IS_SS_DISABLED(0)) + sseu->min_eu_in_pool = 3; + else if (IS_SS_DISABLED(1)) + sseu->min_eu_in_pool = 6; else - info->min_eu_in_pool = 9; + sseu->min_eu_in_pool = 9; } #undef IS_SS_DISABLED } @@ -191,14 +189,20 @@ static void gen9_sseu_info_init(struct drm_i915_private *dev_priv) static void broadwell_sseu_info_init(struct drm_i915_private *dev_priv) { - struct intel_device_info *info = mkwrite_device_info(dev_priv); + struct sseu_dev_info *sseu = &mkwrite_device_info(dev_priv)->sseu; const int s_max = 3, ss_max = 3, eu_max = 8; int s, ss; - u32 fuse2, eu_disable[s_max], s_enable, ss_disable; + u32 fuse2, eu_disable[s_max]; fuse2 = I915_READ(GEN8_FUSE2); - s_enable = (fuse2 & GEN8_F2_S_ENA_MASK) >> GEN8_F2_S_ENA_SHIFT; - ss_disable = (fuse2 & GEN8_F2_SS_DIS_MASK) >> GEN8_F2_SS_DIS_SHIFT; + sseu->slice_mask = (fuse2 & GEN8_F2_S_ENA_MASK) >> GEN8_F2_S_ENA_SHIFT; + /* + * The subslice disable field is global, i.e. it applies + * to each of the enabled slices. + */ + sseu->subslice_mask = BIT(ss_max) - 1; + sseu->subslice_mask &= ~((fuse2 & GEN8_F2_SS_DIS_MASK) >> + GEN8_F2_SS_DIS_SHIFT); eu_disable[0] = I915_READ(GEN8_EU_DISABLE0) & GEN8_EU_DIS0_S0_MASK; eu_disable[1] = (I915_READ(GEN8_EU_DISABLE0) >> GEN8_EU_DIS0_S1_SHIFT) | @@ -208,28 +212,19 @@ static void broadwell_sseu_info_init(struct drm_i915_private *dev_priv) ((I915_READ(GEN8_EU_DISABLE2) & GEN8_EU_DIS2_S2_MASK) << (32 - GEN8_EU_DIS1_S2_SHIFT)); - info->slice_total = hweight32(s_enable); - - /* - * The subslice disable field is global, i.e. it applies - * to each of the enabled slices. - */ - info->subslice_per_slice = ss_max - hweight32(ss_disable); - info->subslice_total = info->slice_total * info->subslice_per_slice; - /* * Iterate through enabled slices and subslices to * count the total enabled EU. */ for (s = 0; s < s_max; s++) { - if (!(s_enable & (0x1 << s))) + if (!(sseu->slice_mask & BIT(s))) /* skip disabled slice */ continue; for (ss = 0; ss < ss_max; ss++) { u32 n_disabled; - if (ss_disable & (0x1 << ss)) + if (!(sseu->subslice_mask & BIT(ss))) /* skip disabled subslice */ continue; @@ -239,9 +234,9 @@ static void broadwell_sseu_info_init(struct drm_i915_private *dev_priv) * Record which subslices have 7 EUs. */ if (eu_max - n_disabled == 7) - info->subslice_7eu[s] |= 1 << ss; + sseu->subslice_7eu[s] |= 1 << ss; - info->eu_total += eu_max - n_disabled; + sseu->eu_total += eu_max - n_disabled; } } @@ -250,16 +245,17 @@ static void broadwell_sseu_info_init(struct drm_i915_private *dev_priv) * subslices with the exception that any one EU in any one subslice may * be fused off for die recovery. */ - info->eu_per_subslice = info->subslice_total ? - DIV_ROUND_UP(info->eu_total, info->subslice_total) : 0; + sseu->eu_per_subslice = sseu_subslice_total(sseu) ? + DIV_ROUND_UP(sseu->eu_total, + sseu_subslice_total(sseu)) : 0; /* * BDW supports slice power gating on devices with more than * one slice. */ - info->has_slice_pg = (info->slice_total > 1); - info->has_subslice_pg = 0; - info->has_eu_pg = 0; + sseu->has_slice_pg = hweight8(sseu->slice_mask) > 1; + sseu->has_subslice_pg = 0; + sseu->has_eu_pg = 0; } /* @@ -374,15 +370,19 @@ void intel_device_info_runtime_init(struct drm_i915_private *dev_priv) if (IS_BXT_REVID(dev_priv, 0, BXT_REVID_A1)) info->has_snoop = false; - DRM_DEBUG_DRIVER("slice total: %u\n", info->slice_total); - DRM_DEBUG_DRIVER("subslice total: %u\n", info->subslice_total); - DRM_DEBUG_DRIVER("subslice per slice: %u\n", info->subslice_per_slice); - DRM_DEBUG_DRIVER("EU total: %u\n", info->eu_total); - DRM_DEBUG_DRIVER("EU per subslice: %u\n", info->eu_per_subslice); + DRM_DEBUG_DRIVER("slice mask: %04x\n", info->sseu.slice_mask); + DRM_DEBUG_DRIVER("slice total: %u\n", hweight8(info->sseu.slice_mask)); + DRM_DEBUG_DRIVER("subslice total: %u\n", + sseu_subslice_total(&info->sseu)); + DRM_DEBUG_DRIVER("subslice mask %04x\n", info->sseu.subslice_mask); + DRM_DEBUG_DRIVER("subslice per slice: %u\n", + hweight8(info->sseu.subslice_mask)); + DRM_DEBUG_DRIVER("EU total: %u\n", info->sseu.eu_total); + DRM_DEBUG_DRIVER("EU per subslice: %u\n", info->sseu.eu_per_subslice); DRM_DEBUG_DRIVER("has slice power gating: %s\n", - info->has_slice_pg ? "y" : "n"); + info->sseu.has_slice_pg ? "y" : "n"); DRM_DEBUG_DRIVER("has subslice power gating: %s\n", - info->has_subslice_pg ? "y" : "n"); + info->sseu.has_subslice_pg ? "y" : "n"); DRM_DEBUG_DRIVER("has EU power gating: %s\n", - info->has_eu_pg ? "y" : "n"); + info->sseu.has_eu_pg ? "y" : "n"); } diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index d224f64836c5..497d99b88468 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -1907,7 +1907,7 @@ static void ironlake_disable_pch_transcoder(struct drm_i915_private *dev_priv, } } -static void lpt_disable_pch_transcoder(struct drm_i915_private *dev_priv) +void lpt_disable_pch_transcoder(struct drm_i915_private *dev_priv) { u32 val; @@ -3377,6 +3377,7 @@ static void skylake_update_primary_plane(struct drm_plane *plane, struct drm_i915_private *dev_priv = to_i915(dev); struct intel_crtc *intel_crtc = to_intel_crtc(crtc_state->base.crtc); struct drm_framebuffer *fb = plane_state->base.fb; + const struct skl_wm_values *wm = &dev_priv->wm.skl_results; int pipe = intel_crtc->pipe; u32 plane_ctl; unsigned int rotation = plane_state->base.rotation; @@ -3410,6 +3411,9 @@ static void skylake_update_primary_plane(struct drm_plane *plane, intel_crtc->adjusted_x = src_x; intel_crtc->adjusted_y = src_y; + if (wm->dirty_pipes & drm_crtc_mask(&intel_crtc->base)) + skl_write_plane_wm(intel_crtc, wm, 0); + I915_WRITE(PLANE_CTL(pipe, 0), plane_ctl); I915_WRITE(PLANE_OFFSET(pipe, 0), (src_y << 16) | src_x); I915_WRITE(PLANE_STRIDE(pipe, 0), stride); @@ -3441,7 +3445,15 @@ static void skylake_disable_primary_plane(struct drm_plane *primary, { struct drm_device *dev = crtc->dev; struct drm_i915_private *dev_priv = to_i915(dev); - int pipe = to_intel_crtc(crtc)->pipe; + struct intel_crtc *intel_crtc = to_intel_crtc(crtc); + int pipe = intel_crtc->pipe; + + /* + * We only populate skl_results on watermark updates, and if the + * plane's visiblity isn't actually changing neither is its watermarks. + */ + if (!crtc->primary->state->visible) + skl_write_plane_wm(intel_crtc, &dev_priv->wm.skl_results, 0); I915_WRITE(PLANE_CTL(pipe, 0), 0); I915_WRITE(PLANE_SURF(pipe, 0), 0); @@ -3634,15 +3646,26 @@ void intel_finish_reset(struct drm_i915_private *dev_priv) mutex_unlock(&dev->mode_config.mutex); } +static bool abort_flip_on_reset(struct intel_crtc *crtc) +{ + struct i915_gpu_error *error = &to_i915(crtc->base.dev)->gpu_error; + + if (i915_reset_in_progress(error)) + return true; + + if (crtc->reset_count != i915_reset_count(error)) + return true; + + return false; +} + static bool intel_crtc_has_pending_flip(struct drm_crtc *crtc) { struct drm_device *dev = crtc->dev; struct intel_crtc *intel_crtc = to_intel_crtc(crtc); - unsigned reset_counter; bool pending; - reset_counter = i915_reset_counter(&to_i915(dev)->gpu_error); - if (intel_crtc->reset_counter != reset_counter) + if (abort_flip_on_reset(intel_crtc)) return false; spin_lock_irq(&dev->event_lock); @@ -4285,7 +4308,7 @@ static int intel_crtc_wait_for_pending_flips(struct drm_crtc *crtc) return 0; } -static void lpt_disable_iclkip(struct drm_i915_private *dev_priv) +void lpt_disable_iclkip(struct drm_i915_private *dev_priv) { u32 temp; @@ -5154,15 +5177,137 @@ static void intel_crtc_disable_planes(struct drm_crtc *crtc, unsigned plane_mask intel_frontbuffer_flip(to_i915(dev), INTEL_FRONTBUFFER_ALL_MASK(pipe)); } -static void ironlake_crtc_enable(struct drm_crtc *crtc) +static void intel_encoders_pre_pll_enable(struct drm_crtc *crtc, + struct intel_crtc_state *crtc_state, + struct drm_atomic_state *old_state) +{ + struct drm_connector_state *old_conn_state; + struct drm_connector *conn; + int i; + + for_each_connector_in_state(old_state, conn, old_conn_state, i) { + struct drm_connector_state *conn_state = conn->state; + struct intel_encoder *encoder = + to_intel_encoder(conn_state->best_encoder); + + if (conn_state->crtc != crtc) + continue; + + if (encoder->pre_pll_enable) + encoder->pre_pll_enable(encoder, crtc_state, conn_state); + } +} + +static void intel_encoders_pre_enable(struct drm_crtc *crtc, + struct intel_crtc_state *crtc_state, + struct drm_atomic_state *old_state) +{ + struct drm_connector_state *old_conn_state; + struct drm_connector *conn; + int i; + + for_each_connector_in_state(old_state, conn, old_conn_state, i) { + struct drm_connector_state *conn_state = conn->state; + struct intel_encoder *encoder = + to_intel_encoder(conn_state->best_encoder); + + if (conn_state->crtc != crtc) + continue; + + if (encoder->pre_enable) + encoder->pre_enable(encoder, crtc_state, conn_state); + } +} + +static void intel_encoders_enable(struct drm_crtc *crtc, + struct intel_crtc_state *crtc_state, + struct drm_atomic_state *old_state) { + struct drm_connector_state *old_conn_state; + struct drm_connector *conn; + int i; + + for_each_connector_in_state(old_state, conn, old_conn_state, i) { + struct drm_connector_state *conn_state = conn->state; + struct intel_encoder *encoder = + to_intel_encoder(conn_state->best_encoder); + + if (conn_state->crtc != crtc) + continue; + + encoder->enable(encoder, crtc_state, conn_state); + intel_opregion_notify_encoder(encoder, true); + } +} + +static void intel_encoders_disable(struct drm_crtc *crtc, + struct intel_crtc_state *old_crtc_state, + struct drm_atomic_state *old_state) +{ + struct drm_connector_state *old_conn_state; + struct drm_connector *conn; + int i; + + for_each_connector_in_state(old_state, conn, old_conn_state, i) { + struct intel_encoder *encoder = + to_intel_encoder(old_conn_state->best_encoder); + + if (old_conn_state->crtc != crtc) + continue; + + intel_opregion_notify_encoder(encoder, false); + encoder->disable(encoder, old_crtc_state, old_conn_state); + } +} + +static void intel_encoders_post_disable(struct drm_crtc *crtc, + struct intel_crtc_state *old_crtc_state, + struct drm_atomic_state *old_state) +{ + struct drm_connector_state *old_conn_state; + struct drm_connector *conn; + int i; + + for_each_connector_in_state(old_state, conn, old_conn_state, i) { + struct intel_encoder *encoder = + to_intel_encoder(old_conn_state->best_encoder); + + if (old_conn_state->crtc != crtc) + continue; + + if (encoder->post_disable) + encoder->post_disable(encoder, old_crtc_state, old_conn_state); + } +} + +static void intel_encoders_post_pll_disable(struct drm_crtc *crtc, + struct intel_crtc_state *old_crtc_state, + struct drm_atomic_state *old_state) +{ + struct drm_connector_state *old_conn_state; + struct drm_connector *conn; + int i; + + for_each_connector_in_state(old_state, conn, old_conn_state, i) { + struct intel_encoder *encoder = + to_intel_encoder(old_conn_state->best_encoder); + + if (old_conn_state->crtc != crtc) + continue; + + if (encoder->post_pll_disable) + encoder->post_pll_disable(encoder, old_crtc_state, old_conn_state); + } +} + +static void ironlake_crtc_enable(struct intel_crtc_state *pipe_config, + struct drm_atomic_state *old_state) +{ + struct drm_crtc *crtc = pipe_config->base.crtc; struct drm_device *dev = crtc->dev; struct drm_i915_private *dev_priv = to_i915(dev); struct intel_crtc *intel_crtc = to_intel_crtc(crtc); - struct intel_encoder *encoder; int pipe = intel_crtc->pipe; - struct intel_crtc_state *pipe_config = - to_intel_crtc_state(crtc->state); if (WARN_ON(intel_crtc->active)) return; @@ -5200,9 +5345,7 @@ static void ironlake_crtc_enable(struct drm_crtc *crtc) intel_crtc->active = true; - for_each_encoder_on_crtc(dev, crtc, encoder) - if (encoder->pre_enable) - encoder->pre_enable(encoder); + intel_encoders_pre_enable(crtc, pipe_config, old_state); if (intel_crtc->config->has_pch_encoder) { /* Note: FDI PLL enabling _must_ be done before we enable the @@ -5232,8 +5375,7 @@ static void ironlake_crtc_enable(struct drm_crtc *crtc) assert_vblank_disabled(crtc); drm_crtc_vblank_on(crtc); - for_each_encoder_on_crtc(dev, crtc, encoder) - encoder->enable(encoder); + intel_encoders_enable(crtc, pipe_config, old_state); if (HAS_PCH_CPT(dev)) cpt_verify_modeset(dev, intel_crtc->pipe); @@ -5251,16 +5393,15 @@ static bool hsw_crtc_supports_ips(struct intel_crtc *crtc) return HAS_IPS(crtc->base.dev) && crtc->pipe == PIPE_A; } -static void haswell_crtc_enable(struct drm_crtc *crtc) +static void haswell_crtc_enable(struct intel_crtc_state *pipe_config, + struct drm_atomic_state *old_state) { + struct drm_crtc *crtc = pipe_config->base.crtc; struct drm_device *dev = crtc->dev; struct drm_i915_private *dev_priv = to_i915(dev); struct intel_crtc *intel_crtc = to_intel_crtc(crtc); - struct intel_encoder *encoder; int pipe = intel_crtc->pipe, hsw_workaround_pipe; enum transcoder cpu_transcoder = intel_crtc->config->cpu_transcoder; - struct intel_crtc_state *pipe_config = - to_intel_crtc_state(crtc->state); if (WARN_ON(intel_crtc->active)) return; @@ -5269,9 +5410,7 @@ static void haswell_crtc_enable(struct drm_crtc *crtc) intel_set_pch_fifo_underrun_reporting(dev_priv, TRANSCODER_A, false); - for_each_encoder_on_crtc(dev, crtc, encoder) - if (encoder->pre_pll_enable) - encoder->pre_pll_enable(encoder); + intel_encoders_pre_pll_enable(crtc, pipe_config, old_state); if (intel_crtc->config->shared_dpll) intel_enable_shared_dpll(intel_crtc); @@ -5309,10 +5448,7 @@ static void haswell_crtc_enable(struct drm_crtc *crtc) else intel_set_cpu_fifo_underrun_reporting(dev_priv, pipe, true); - for_each_encoder_on_crtc(dev, crtc, encoder) { - if (encoder->pre_enable) - encoder->pre_enable(encoder); - } + intel_encoders_pre_enable(crtc, pipe_config, old_state); if (intel_crtc->config->has_pch_encoder) dev_priv->display.fdi_link_train(crtc); @@ -5353,10 +5489,7 @@ static void haswell_crtc_enable(struct drm_crtc *crtc) assert_vblank_disabled(crtc); drm_crtc_vblank_on(crtc); - for_each_encoder_on_crtc(dev, crtc, encoder) { - encoder->enable(encoder); - intel_opregion_notify_encoder(encoder, true); - } + intel_encoders_enable(crtc, pipe_config, old_state); if (intel_crtc->config->has_pch_encoder) { intel_wait_for_vblank(dev, pipe); @@ -5390,12 +5523,13 @@ static void ironlake_pfit_disable(struct intel_crtc *crtc, bool force) } } -static void ironlake_crtc_disable(struct drm_crtc *crtc) +static void ironlake_crtc_disable(struct intel_crtc_state *old_crtc_state, + struct drm_atomic_state *old_state) { + struct drm_crtc *crtc = old_crtc_state->base.crtc; struct drm_device *dev = crtc->dev; struct drm_i915_private *dev_priv = to_i915(dev); struct intel_crtc *intel_crtc = to_intel_crtc(crtc); - struct intel_encoder *encoder; int pipe = intel_crtc->pipe; /* @@ -5408,8 +5542,7 @@ static void ironlake_crtc_disable(struct drm_crtc *crtc) intel_set_pch_fifo_underrun_reporting(dev_priv, pipe, false); } - for_each_encoder_on_crtc(dev, crtc, encoder) - encoder->disable(encoder); + intel_encoders_disable(crtc, old_crtc_state, old_state); drm_crtc_vblank_off(crtc); assert_vblank_disabled(crtc); @@ -5421,9 +5554,7 @@ static void ironlake_crtc_disable(struct drm_crtc *crtc) if (intel_crtc->config->has_pch_encoder) ironlake_fdi_disable(crtc); - for_each_encoder_on_crtc(dev, crtc, encoder) - if (encoder->post_disable) - encoder->post_disable(encoder); + intel_encoders_post_disable(crtc, old_crtc_state, old_state); if (intel_crtc->config->has_pch_encoder) { ironlake_disable_pch_transcoder(dev_priv, pipe); @@ -5453,22 +5584,20 @@ static void ironlake_crtc_disable(struct drm_crtc *crtc) intel_set_pch_fifo_underrun_reporting(dev_priv, pipe, true); } -static void haswell_crtc_disable(struct drm_crtc *crtc) +static void haswell_crtc_disable(struct intel_crtc_state *old_crtc_state, + struct drm_atomic_state *old_state) { + struct drm_crtc *crtc = old_crtc_state->base.crtc; struct drm_device *dev = crtc->dev; struct drm_i915_private *dev_priv = to_i915(dev); struct intel_crtc *intel_crtc = to_intel_crtc(crtc); - struct intel_encoder *encoder; enum transcoder cpu_transcoder = intel_crtc->config->cpu_transcoder; if (intel_crtc->config->has_pch_encoder) intel_set_pch_fifo_underrun_reporting(dev_priv, TRANSCODER_A, false); - for_each_encoder_on_crtc(dev, crtc, encoder) { - intel_opregion_notify_encoder(encoder, false); - encoder->disable(encoder); - } + intel_encoders_disable(crtc, old_crtc_state, old_state); drm_crtc_vblank_off(crtc); assert_vblank_disabled(crtc); @@ -5491,18 +5620,11 @@ static void haswell_crtc_disable(struct drm_crtc *crtc) if (!transcoder_is_dsi(cpu_transcoder)) intel_ddi_disable_pipe_clock(intel_crtc); - for_each_encoder_on_crtc(dev, crtc, encoder) - if (encoder->post_disable) - encoder->post_disable(encoder); - - if (intel_crtc->config->has_pch_encoder) { - lpt_disable_pch_transcoder(dev_priv); - lpt_disable_iclkip(dev_priv); - intel_ddi_fdi_disable(crtc); + intel_encoders_post_disable(crtc, old_crtc_state, old_state); + if (old_crtc_state->has_pch_encoder) intel_set_pch_fifo_underrun_reporting(dev_priv, TRANSCODER_A, true); - } } static void i9xx_pfit_enable(struct intel_crtc *crtc) @@ -6558,14 +6680,13 @@ static void valleyview_modeset_commit_cdclk(struct drm_atomic_state *old_state) intel_display_power_put(dev_priv, POWER_DOMAIN_PIPE_A); } -static void valleyview_crtc_enable(struct drm_crtc *crtc) +static void valleyview_crtc_enable(struct intel_crtc_state *pipe_config, + struct drm_atomic_state *old_state) { + struct drm_crtc *crtc = pipe_config->base.crtc; struct drm_device *dev = crtc->dev; struct drm_i915_private *dev_priv = to_i915(dev); struct intel_crtc *intel_crtc = to_intel_crtc(crtc); - struct intel_encoder *encoder; - struct intel_crtc_state *pipe_config = - to_intel_crtc_state(crtc->state); int pipe = intel_crtc->pipe; if (WARN_ON(intel_crtc->active)) @@ -6590,9 +6711,7 @@ static void valleyview_crtc_enable(struct drm_crtc *crtc) intel_set_cpu_fifo_underrun_reporting(dev_priv, pipe, true); - for_each_encoder_on_crtc(dev, crtc, encoder) - if (encoder->pre_pll_enable) - encoder->pre_pll_enable(encoder); + intel_encoders_pre_pll_enable(crtc, pipe_config, old_state); if (IS_CHERRYVIEW(dev)) { chv_prepare_pll(intel_crtc, intel_crtc->config); @@ -6602,9 +6721,7 @@ static void valleyview_crtc_enable(struct drm_crtc *crtc) vlv_enable_pll(intel_crtc, intel_crtc->config); } - for_each_encoder_on_crtc(dev, crtc, encoder) - if (encoder->pre_enable) - encoder->pre_enable(encoder); + intel_encoders_pre_enable(crtc, pipe_config, old_state); i9xx_pfit_enable(intel_crtc); @@ -6616,8 +6733,7 @@ static void valleyview_crtc_enable(struct drm_crtc *crtc) assert_vblank_disabled(crtc); drm_crtc_vblank_on(crtc); - for_each_encoder_on_crtc(dev, crtc, encoder) - encoder->enable(encoder); + intel_encoders_enable(crtc, pipe_config, old_state); } static void i9xx_set_pll_dividers(struct intel_crtc *crtc) @@ -6629,14 +6745,13 @@ static void i9xx_set_pll_dividers(struct intel_crtc *crtc) I915_WRITE(FP1(crtc->pipe), crtc->config->dpll_hw_state.fp1); } -static void i9xx_crtc_enable(struct drm_crtc *crtc) +static void i9xx_crtc_enable(struct intel_crtc_state *pipe_config, + struct drm_atomic_state *old_state) { + struct drm_crtc *crtc = pipe_config->base.crtc; struct drm_device *dev = crtc->dev; struct drm_i915_private *dev_priv = to_i915(dev); struct intel_crtc *intel_crtc = to_intel_crtc(crtc); - struct intel_encoder *encoder; - struct intel_crtc_state *pipe_config = - to_intel_crtc_state(crtc->state); enum pipe pipe = intel_crtc->pipe; if (WARN_ON(intel_crtc->active)) @@ -6657,9 +6772,7 @@ static void i9xx_crtc_enable(struct drm_crtc *crtc) if (!IS_GEN2(dev)) intel_set_cpu_fifo_underrun_reporting(dev_priv, pipe, true); - for_each_encoder_on_crtc(dev, crtc, encoder) - if (encoder->pre_enable) - encoder->pre_enable(encoder); + intel_encoders_pre_enable(crtc, pipe_config, old_state); i9xx_enable_pll(intel_crtc); @@ -6673,8 +6786,7 @@ static void i9xx_crtc_enable(struct drm_crtc *crtc) assert_vblank_disabled(crtc); drm_crtc_vblank_on(crtc); - for_each_encoder_on_crtc(dev, crtc, encoder) - encoder->enable(encoder); + intel_encoders_enable(crtc, pipe_config, old_state); } static void i9xx_pfit_disable(struct intel_crtc *crtc) @@ -6692,12 +6804,13 @@ static void i9xx_pfit_disable(struct intel_crtc *crtc) I915_WRITE(PFIT_CONTROL, 0); } -static void i9xx_crtc_disable(struct drm_crtc *crtc) +static void i9xx_crtc_disable(struct intel_crtc_state *old_crtc_state, + struct drm_atomic_state *old_state) { + struct drm_crtc *crtc = old_crtc_state->base.crtc; struct drm_device *dev = crtc->dev; struct drm_i915_private *dev_priv = to_i915(dev); struct intel_crtc *intel_crtc = to_intel_crtc(crtc); - struct intel_encoder *encoder; int pipe = intel_crtc->pipe; /* @@ -6707,8 +6820,7 @@ static void i9xx_crtc_disable(struct drm_crtc *crtc) if (IS_GEN2(dev)) intel_wait_for_vblank(dev, pipe); - for_each_encoder_on_crtc(dev, crtc, encoder) - encoder->disable(encoder); + intel_encoders_disable(crtc, old_crtc_state, old_state); drm_crtc_vblank_off(crtc); assert_vblank_disabled(crtc); @@ -6717,9 +6829,7 @@ static void i9xx_crtc_disable(struct drm_crtc *crtc) i9xx_pfit_disable(intel_crtc); - for_each_encoder_on_crtc(dev, crtc, encoder) - if (encoder->post_disable) - encoder->post_disable(encoder); + intel_encoders_post_disable(crtc, old_crtc_state, old_state); if (!intel_crtc_has_type(intel_crtc->config, INTEL_OUTPUT_DSI)) { if (IS_CHERRYVIEW(dev)) @@ -6730,9 +6840,7 @@ static void i9xx_crtc_disable(struct drm_crtc *crtc) i9xx_disable_pll(intel_crtc); } - for_each_encoder_on_crtc(dev, crtc, encoder) - if (encoder->post_pll_disable) - encoder->post_pll_disable(encoder); + intel_encoders_post_pll_disable(crtc, old_crtc_state, old_state); if (!IS_GEN2(dev)) intel_set_cpu_fifo_underrun_reporting(dev_priv, pipe, false); @@ -6745,6 +6853,9 @@ static void intel_crtc_disable_noatomic(struct drm_crtc *crtc) struct drm_i915_private *dev_priv = to_i915(crtc->dev); enum intel_display_power_domain domain; unsigned long domains; + struct drm_atomic_state *state; + struct intel_crtc_state *crtc_state; + int ret; if (!intel_crtc->active) return; @@ -6758,7 +6869,18 @@ static void intel_crtc_disable_noatomic(struct drm_crtc *crtc) to_intel_plane_state(crtc->primary->state)->base.visible = false; } - dev_priv->display.crtc_disable(crtc); + state = drm_atomic_state_alloc(crtc->dev); + state->acquire_ctx = crtc->dev->mode_config.acquire_ctx; + + /* Everything's already locked, -EDEADLK can't happen. */ + crtc_state = intel_atomic_get_crtc_state(state, intel_crtc); + ret = drm_atomic_add_affected_connectors(state, crtc); + + WARN_ON(IS_ERR(crtc_state) || ret); + + dev_priv->display.crtc_disable(crtc_state, state); + + drm_atomic_state_free(state); DRM_DEBUG_KMS("[CRTC:%d:%s] hw state adjusted, was enabled, now disabled\n", crtc->base.id, crtc->name); @@ -7273,9 +7395,10 @@ static int i9xx_misc_get_display_clock_speed(struct drm_device *dev) static int pnv_get_display_clock_speed(struct drm_device *dev) { + struct pci_dev *pdev = dev->pdev; u16 gcfgc = 0; - pci_read_config_word(dev->pdev, GCFGC, &gcfgc); + pci_read_config_word(pdev, GCFGC, &gcfgc); switch (gcfgc & GC_DISPLAY_CLOCK_MASK) { case GC_DISPLAY_CLOCK_267_MHZ_PNV: @@ -7297,9 +7420,10 @@ static int pnv_get_display_clock_speed(struct drm_device *dev) static int i915gm_get_display_clock_speed(struct drm_device *dev) { + struct pci_dev *pdev = dev->pdev; u16 gcfgc = 0; - pci_read_config_word(dev->pdev, GCFGC, &gcfgc); + pci_read_config_word(pdev, GCFGC, &gcfgc); if (gcfgc & GC_LOW_FREQUENCY_ENABLE) return 133333; @@ -7321,6 +7445,7 @@ static int i865_get_display_clock_speed(struct drm_device *dev) static int i85x_get_display_clock_speed(struct drm_device *dev) { + struct pci_dev *pdev = dev->pdev; u16 hpllcc = 0; /* @@ -7328,10 +7453,10 @@ static int i85x_get_display_clock_speed(struct drm_device *dev) * encoding is different :( * FIXME is this the right way to detect 852GM/852GMV? */ - if (dev->pdev->revision == 0x1) + if (pdev->revision == 0x1) return 133333; - pci_bus_read_config_word(dev->pdev->bus, + pci_bus_read_config_word(pdev->bus, PCI_DEVFN(0, 3), HPLLCC, &hpllcc); /* Assume that the hardware is in the high speed state. This @@ -7432,10 +7557,11 @@ static unsigned int intel_hpll_vco(struct drm_device *dev) static int gm45_get_display_clock_speed(struct drm_device *dev) { + struct pci_dev *pdev = dev->pdev; unsigned int cdclk_sel, vco = intel_hpll_vco(dev); uint16_t tmp = 0; - pci_read_config_word(dev->pdev, GCFGC, &tmp); + pci_read_config_word(pdev, GCFGC, &tmp); cdclk_sel = (tmp >> 12) & 0x1; @@ -7454,6 +7580,7 @@ static int gm45_get_display_clock_speed(struct drm_device *dev) static int i965gm_get_display_clock_speed(struct drm_device *dev) { + struct pci_dev *pdev = dev->pdev; static const uint8_t div_3200[] = { 16, 10, 8 }; static const uint8_t div_4000[] = { 20, 12, 10 }; static const uint8_t div_5333[] = { 24, 16, 14 }; @@ -7461,7 +7588,7 @@ static int i965gm_get_display_clock_speed(struct drm_device *dev) unsigned int cdclk_sel, vco = intel_hpll_vco(dev); uint16_t tmp = 0; - pci_read_config_word(dev->pdev, GCFGC, &tmp); + pci_read_config_word(pdev, GCFGC, &tmp); cdclk_sel = ((tmp >> 8) & 0x1f) - 1; @@ -7491,6 +7618,7 @@ fail: static int g33_get_display_clock_speed(struct drm_device *dev) { + struct pci_dev *pdev = dev->pdev; static const uint8_t div_3200[] = { 12, 10, 8, 7, 5, 16 }; static const uint8_t div_4000[] = { 14, 12, 10, 8, 6, 20 }; static const uint8_t div_4800[] = { 20, 14, 12, 10, 8, 24 }; @@ -7499,7 +7627,7 @@ static int g33_get_display_clock_speed(struct drm_device *dev) unsigned int cdclk_sel, vco = intel_hpll_vco(dev); uint16_t tmp = 0; - pci_read_config_word(dev->pdev, GCFGC, &tmp); + pci_read_config_word(pdev, GCFGC, &tmp); cdclk_sel = (tmp >> 4) & 0x7; @@ -9665,7 +9793,7 @@ skylake_get_initial_plane_config(struct intel_crtc *crtc, return; error: - kfree(fb); + kfree(intel_fb); } static void ironlake_get_pfit_config(struct intel_crtc *crtc, @@ -9910,7 +10038,7 @@ static void hsw_write_dcomp(struct drm_i915_private *dev_priv, uint32_t val) mutex_lock(&dev_priv->rps.hw_lock); if (sandybridge_pcode_write(dev_priv, GEN6_PCODE_WRITE_D_COMP, val)) - DRM_ERROR("Failed to write to D_COMP\n"); + DRM_DEBUG_KMS("Failed to write to D_COMP\n"); mutex_unlock(&dev_priv->rps.hw_lock); } else { I915_WRITE(D_COMP_BDW, val); @@ -10318,15 +10446,12 @@ static void bxt_get_ddi_pll(struct drm_i915_private *dev_priv, switch (port) { case PORT_A: - pipe_config->ddi_pll_sel = SKL_DPLL0; id = DPLL_ID_SKL_DPLL0; break; case PORT_B: - pipe_config->ddi_pll_sel = SKL_DPLL1; id = DPLL_ID_SKL_DPLL1; break; case PORT_C: - pipe_config->ddi_pll_sel = SKL_DPLL2; id = DPLL_ID_SKL_DPLL2; break; default: @@ -10345,25 +10470,10 @@ static void skylake_get_ddi_pll(struct drm_i915_private *dev_priv, u32 temp; temp = I915_READ(DPLL_CTRL2) & DPLL_CTRL2_DDI_CLK_SEL_MASK(port); - pipe_config->ddi_pll_sel = temp >> (port * 3 + 1); + id = temp >> (port * 3 + 1); - switch (pipe_config->ddi_pll_sel) { - case SKL_DPLL0: - id = DPLL_ID_SKL_DPLL0; - break; - case SKL_DPLL1: - id = DPLL_ID_SKL_DPLL1; - break; - case SKL_DPLL2: - id = DPLL_ID_SKL_DPLL2; - break; - case SKL_DPLL3: - id = DPLL_ID_SKL_DPLL3; - break; - default: - MISSING_CASE(pipe_config->ddi_pll_sel); + if (WARN_ON(id < SKL_DPLL0 || id > SKL_DPLL3)) return; - } pipe_config->shared_dpll = intel_get_shared_dpll_by_id(dev_priv, id); } @@ -10373,10 +10483,9 @@ static void haswell_get_ddi_pll(struct drm_i915_private *dev_priv, struct intel_crtc_state *pipe_config) { enum intel_dpll_id id; + uint32_t ddi_pll_sel = I915_READ(PORT_CLK_SEL(port)); - pipe_config->ddi_pll_sel = I915_READ(PORT_CLK_SEL(port)); - - switch (pipe_config->ddi_pll_sel) { + switch (ddi_pll_sel) { case PORT_CLK_SEL_WRPLL1: id = DPLL_ID_WRPLL1; break; @@ -10396,7 +10505,7 @@ static void haswell_get_ddi_pll(struct drm_i915_private *dev_priv, id = DPLL_ID_LCPLL_2700; break; default: - MISSING_CASE(pipe_config->ddi_pll_sel); + MISSING_CASE(ddi_pll_sel); /* fall through */ case PORT_CLK_SEL_NONE: return; @@ -10690,9 +10799,13 @@ static void i9xx_update_cursor(struct drm_crtc *crtc, u32 base, struct drm_device *dev = crtc->dev; struct drm_i915_private *dev_priv = to_i915(dev); struct intel_crtc *intel_crtc = to_intel_crtc(crtc); + const struct skl_wm_values *wm = &dev_priv->wm.skl_results; int pipe = intel_crtc->pipe; uint32_t cntl = 0; + if (INTEL_GEN(dev_priv) >= 9 && wm->dirty_pipes & drm_crtc_mask(crtc)) + skl_write_cursor_wm(intel_crtc, wm); + if (plane_state && plane_state->base.visible) { cntl = MCURSOR_GAMMA_ENABLE; switch (plane_state->base.crtc_w) { @@ -11431,10 +11544,8 @@ static bool __pageflip_finished_cs(struct intel_crtc *crtc, { struct drm_device *dev = crtc->base.dev; struct drm_i915_private *dev_priv = to_i915(dev); - unsigned reset_counter; - reset_counter = i915_reset_counter(&dev_priv->gpu_error); - if (crtc->reset_counter != reset_counter) + if (abort_flip_on_reset(crtc)) return true; /* @@ -11911,8 +12022,7 @@ static void intel_mmio_flip_work_func(struct work_struct *w) if (work->flip_queued_req) WARN_ON(i915_wait_request(work->flip_queued_req, - false, NULL, - NO_WAITBOOST)); + 0, NULL, NO_WAITBOOST)); /* For framebuffer backed by dmabuf, wait for fence */ resv = i915_gem_object_get_dmabuf_resv(obj); @@ -12094,17 +12204,14 @@ static int intel_crtc_page_flip(struct drm_crtc *crtc, crtc->primary->fb = fb; update_state_fb(crtc->primary); - intel_fbc_pre_update(intel_crtc, intel_crtc->config, - to_intel_plane_state(primary->state)); - work->pending_flip_obj = i915_gem_object_get(obj); ret = i915_mutex_lock_interruptible(dev); if (ret) goto cleanup; - intel_crtc->reset_counter = i915_reset_counter(&dev_priv->gpu_error); - if (__i915_reset_in_progress_or_wedged(intel_crtc->reset_counter)) { + intel_crtc->reset_count = i915_reset_count(&dev_priv->gpu_error); + if (i915_reset_in_progress_or_wedged(&dev_priv->gpu_error)) { ret = -EIO; goto cleanup; } @@ -12142,6 +12249,17 @@ static int intel_crtc_page_flip(struct drm_crtc *crtc, work->gtt_offset += intel_crtc->dspaddr_offset; work->rotation = crtc->primary->state->rotation; + /* + * There's the potential that the next frame will not be compatible with + * FBC, so we want to call pre_update() before the actual page flip. + * The problem is that pre_update() caches some information about the fb + * object, so we want to do this only after the object is pinned. Let's + * be on the safe side and do this immediately before scheduling the + * flip. + */ + intel_fbc_pre_update(intel_crtc, intel_crtc->config, + to_intel_plane_state(primary->state)); + if (mmio_flip) { INIT_WORK(&work->mmio_work, intel_mmio_flip_work_func); @@ -12155,7 +12273,7 @@ static int intel_crtc_page_flip(struct drm_crtc *crtc, goto cleanup_unpin; } - ret = i915_gem_object_sync(obj, request); + ret = i915_gem_request_await_object(request, obj, false); if (ret) goto cleanup_request; @@ -12668,10 +12786,9 @@ static void intel_dump_pipe_config(struct intel_crtc *crtc, DRM_DEBUG_KMS("double wide: %i\n", pipe_config->double_wide); if (IS_BROXTON(dev)) { - DRM_DEBUG_KMS("ddi_pll_sel: %u; dpll_hw_state: ebb0: 0x%x, ebb4: 0x%x," + DRM_DEBUG_KMS("dpll_hw_state: ebb0: 0x%x, ebb4: 0x%x," "pll0: 0x%x, pll1: 0x%x, pll2: 0x%x, pll3: 0x%x, " "pll6: 0x%x, pll8: 0x%x, pll9: 0x%x, pll10: 0x%x, pcsdw12: 0x%x\n", - pipe_config->ddi_pll_sel, pipe_config->dpll_hw_state.ebb0, pipe_config->dpll_hw_state.ebb4, pipe_config->dpll_hw_state.pll0, @@ -12684,15 +12801,13 @@ static void intel_dump_pipe_config(struct intel_crtc *crtc, pipe_config->dpll_hw_state.pll10, pipe_config->dpll_hw_state.pcsdw12); } else if (IS_SKYLAKE(dev) || IS_KABYLAKE(dev)) { - DRM_DEBUG_KMS("ddi_pll_sel: %u; dpll_hw_state: " + DRM_DEBUG_KMS("dpll_hw_state: " "ctrl1: 0x%x, cfgcr1: 0x%x, cfgcr2: 0x%x\n", - pipe_config->ddi_pll_sel, pipe_config->dpll_hw_state.ctrl1, pipe_config->dpll_hw_state.cfgcr1, pipe_config->dpll_hw_state.cfgcr2); } else if (HAS_DDI(dev)) { - DRM_DEBUG_KMS("ddi_pll_sel: 0x%x; dpll_hw_state: wrpll: 0x%x spll: 0x%x\n", - pipe_config->ddi_pll_sel, + DRM_DEBUG_KMS("dpll_hw_state: wrpll: 0x%x spll: 0x%x\n", pipe_config->dpll_hw_state.wrpll, pipe_config->dpll_hw_state.spll); } else { @@ -12805,7 +12920,6 @@ clear_intel_crtc_state(struct intel_crtc_state *crtc_state) struct intel_crtc_scaler_state scaler_state; struct intel_dpll_hw_state dpll_hw_state; struct intel_shared_dpll *shared_dpll; - uint32_t ddi_pll_sel; bool force_thru; /* FIXME: before the switch to atomic started, a new pipe_config was @@ -12817,7 +12931,6 @@ clear_intel_crtc_state(struct intel_crtc_state *crtc_state) scaler_state = crtc_state->scaler_state; shared_dpll = crtc_state->shared_dpll; dpll_hw_state = crtc_state->dpll_hw_state; - ddi_pll_sel = crtc_state->ddi_pll_sel; force_thru = crtc_state->pch_pfit.force_thru; memset(crtc_state, 0, sizeof *crtc_state); @@ -12826,7 +12939,6 @@ clear_intel_crtc_state(struct intel_crtc_state *crtc_state) crtc_state->scaler_state = scaler_state; crtc_state->shared_dpll = shared_dpll; crtc_state->dpll_hw_state = dpll_hw_state; - crtc_state->ddi_pll_sel = ddi_pll_sel; crtc_state->pch_pfit.force_thru = force_thru; } @@ -12914,7 +13026,7 @@ encoder_retry: encoder = to_intel_encoder(connector_state->best_encoder); - if (!(encoder->compute_config(encoder, pipe_config))) { + if (!(encoder->compute_config(encoder, pipe_config, connector_state))) { DRM_DEBUG_KMS("Encoder config failure\n"); goto fail; } @@ -13002,12 +13114,6 @@ static bool intel_fuzzy_clock_check(int clock1, int clock2) return false; } -#define for_each_intel_crtc_masked(dev, mask, intel_crtc) \ - list_for_each_entry((intel_crtc), \ - &(dev)->mode_config.crtc_list, \ - base.head) \ - for_each_if (mask & (1 <<(intel_crtc)->pipe)) - static bool intel_compare_m_n(unsigned int m, unsigned int n, unsigned int m2, unsigned int n2, @@ -13255,8 +13361,6 @@ intel_pipe_config_compare(struct drm_device *dev, PIPE_CONF_CHECK_I(double_wide); - PIPE_CONF_CHECK_X(ddi_pll_sel); - PIPE_CONF_CHECK_P(shared_dpll); PIPE_CONF_CHECK_X(dpll_hw_state.dpll); PIPE_CONF_CHECK_X(dpll_hw_state.dpll_md); @@ -13338,16 +13442,23 @@ static void verify_wm_state(struct drm_crtc *crtc, hw_entry->start, hw_entry->end); } - /* cursor */ - hw_entry = &hw_ddb.plane[pipe][PLANE_CURSOR]; - sw_entry = &sw_ddb->plane[pipe][PLANE_CURSOR]; - - if (!skl_ddb_entry_equal(hw_entry, sw_entry)) { - DRM_ERROR("mismatch in DDB state pipe %c cursor " - "(expected (%u,%u), found (%u,%u))\n", - pipe_name(pipe), - sw_entry->start, sw_entry->end, - hw_entry->start, hw_entry->end); + /* + * cursor + * If the cursor plane isn't active, we may not have updated it's ddb + * allocation. In that case since the ddb allocation will be updated + * once the plane becomes visible, we can skip this check + */ + if (intel_crtc->cursor_addr) { + hw_entry = &hw_ddb.plane[pipe][PLANE_CURSOR]; + sw_entry = &sw_ddb->plane[pipe][PLANE_CURSOR]; + + if (!skl_ddb_entry_equal(hw_entry, sw_entry)) { + DRM_ERROR("mismatch in DDB state pipe %c cursor " + "(expected (%u,%u), found (%u,%u))\n", + pipe_name(pipe), + sw_entry->start, sw_entry->end, + hw_entry->start, hw_entry->end); + } } } @@ -13963,7 +14074,8 @@ static int intel_atomic_prepare_commit(struct drm_device *dev, continue; ret = i915_wait_request(intel_plane_state->wait_req, - true, NULL, NULL); + I915_WAIT_INTERRUPTIBLE, + NULL, NULL); if (ret) { /* Any hang should be swallowed by the wait */ WARN_ON(ret == -EIO); @@ -14053,6 +14165,111 @@ static bool needs_vblank_wait(struct intel_crtc_state *crtc_state) return false; } +static void intel_update_crtc(struct drm_crtc *crtc, + struct drm_atomic_state *state, + struct drm_crtc_state *old_crtc_state, + unsigned int *crtc_vblank_mask) +{ + struct drm_device *dev = crtc->dev; + struct drm_i915_private *dev_priv = to_i915(dev); + struct intel_crtc *intel_crtc = to_intel_crtc(crtc); + struct intel_crtc_state *pipe_config = to_intel_crtc_state(crtc->state); + bool modeset = needs_modeset(crtc->state); + + if (modeset) { + update_scanline_offset(intel_crtc); + dev_priv->display.crtc_enable(pipe_config, state); + } else { + intel_pre_plane_update(to_intel_crtc_state(old_crtc_state)); + } + + if (drm_atomic_get_existing_plane_state(state, crtc->primary)) { + intel_fbc_enable( + intel_crtc, pipe_config, + to_intel_plane_state(crtc->primary->state)); + } + + drm_atomic_helper_commit_planes_on_crtc(old_crtc_state); + + if (needs_vblank_wait(pipe_config)) + *crtc_vblank_mask |= drm_crtc_mask(crtc); +} + +static void intel_update_crtcs(struct drm_atomic_state *state, + unsigned int *crtc_vblank_mask) +{ + struct drm_crtc *crtc; + struct drm_crtc_state *old_crtc_state; + int i; + + for_each_crtc_in_state(state, crtc, old_crtc_state, i) { + if (!crtc->state->active) + continue; + + intel_update_crtc(crtc, state, old_crtc_state, + crtc_vblank_mask); + } +} + +static void skl_update_crtcs(struct drm_atomic_state *state, + unsigned int *crtc_vblank_mask) +{ + struct drm_device *dev = state->dev; + struct drm_i915_private *dev_priv = to_i915(dev); + struct intel_atomic_state *intel_state = to_intel_atomic_state(state); + struct drm_crtc *crtc; + struct drm_crtc_state *old_crtc_state; + struct skl_ddb_allocation *new_ddb = &intel_state->wm_results.ddb; + struct skl_ddb_allocation *cur_ddb = &dev_priv->wm.skl_hw.ddb; + unsigned int updated = 0; + bool progress; + enum pipe pipe; + + /* + * Whenever the number of active pipes changes, we need to make sure we + * update the pipes in the right order so that their ddb allocations + * never overlap with eachother inbetween CRTC updates. Otherwise we'll + * cause pipe underruns and other bad stuff. + */ + do { + int i; + progress = false; + + for_each_crtc_in_state(state, crtc, old_crtc_state, i) { + bool vbl_wait = false; + unsigned int cmask = drm_crtc_mask(crtc); + pipe = to_intel_crtc(crtc)->pipe; + + if (updated & cmask || !crtc->state->active) + continue; + if (skl_ddb_allocation_overlaps(state, cur_ddb, new_ddb, + pipe)) + continue; + + updated |= cmask; + + /* + * If this is an already active pipe, it's DDB changed, + * and this isn't the last pipe that needs updating + * then we need to wait for a vblank to pass for the + * new ddb allocation to take effect. + */ + if (!skl_ddb_allocation_equals(cur_ddb, new_ddb, pipe) && + !crtc->state->active_changed && + intel_state->wm_results.dirty_pipes != updated) + vbl_wait = true; + + intel_update_crtc(crtc, state, old_crtc_state, + crtc_vblank_mask); + + if (vbl_wait) + intel_wait_for_vblank(dev, pipe); + + progress = true; + } + } while (progress); +} + static void intel_atomic_commit_tail(struct drm_atomic_state *state) { struct drm_device *dev = state->dev; @@ -14076,7 +14293,7 @@ static void intel_atomic_commit_tail(struct drm_atomic_state *state) continue; ret = i915_wait_request(intel_plane_state->wait_req, - true, NULL, NULL); + 0, NULL, NULL); /* EIO should be eaten, and we can't get interrupted in the * worker, and blocking commits have waited already. */ WARN_ON(ret); @@ -14112,7 +14329,7 @@ static void intel_atomic_commit_tail(struct drm_atomic_state *state) if (old_crtc_state->active) { intel_crtc_disable_planes(crtc, old_crtc_state->plane_mask); - dev_priv->display.crtc_disable(crtc); + dev_priv->display.crtc_disable(to_intel_crtc_state(old_crtc_state), state); intel_crtc->active = false; intel_fbc_disable(intel_crtc); intel_disable_shared_dpll(intel_crtc); @@ -14141,20 +14358,19 @@ static void intel_atomic_commit_tail(struct drm_atomic_state *state) intel_state->cdclk_pll_vco != dev_priv->cdclk_pll.vco)) dev_priv->display.modeset_commit_cdclk(state); + /* + * SKL workaround: bspec recommends we disable the SAGV when we + * have more then one pipe enabled + */ + if (IS_SKYLAKE(dev_priv) && !skl_can_enable_sagv(state)) + skl_disable_sagv(dev_priv); + intel_modeset_verify_disabled(dev); } - /* Now enable the clocks, plane, pipe, and connectors that we set up. */ + /* Complete the events for pipes that have now been disabled */ for_each_crtc_in_state(state, crtc, old_crtc_state, i) { - struct intel_crtc *intel_crtc = to_intel_crtc(crtc); bool modeset = needs_modeset(crtc->state); - struct intel_crtc_state *pipe_config = - to_intel_crtc_state(crtc->state); - - if (modeset && crtc->state->active) { - update_scanline_offset(to_intel_crtc(crtc)); - dev_priv->display.crtc_enable(crtc); - } /* Complete events for now disable pipes here. */ if (modeset && !crtc->state->active && crtc->state->event) { @@ -14164,21 +14380,11 @@ static void intel_atomic_commit_tail(struct drm_atomic_state *state) crtc->state->event = NULL; } - - if (!modeset) - intel_pre_plane_update(to_intel_crtc_state(old_crtc_state)); - - if (crtc->state->active && - drm_atomic_get_existing_plane_state(state, crtc->primary)) - intel_fbc_enable(intel_crtc, pipe_config, to_intel_plane_state(crtc->primary->state)); - - if (crtc->state->active) - drm_atomic_helper_commit_planes_on_crtc(old_crtc_state); - - if (pipe_config->base.active && needs_vblank_wait(pipe_config)) - crtc_vblank_mask |= 1 << i; } + /* Now enable the clocks, plane, pipe, and connectors that we set up. */ + dev_priv->display.update_crtcs(state, &crtc_vblank_mask); + /* FIXME: We should call drm_atomic_helper_commit_hw_done() here * already, but still need the state for the delayed optimization. To * fix this: @@ -14214,6 +14420,10 @@ static void intel_atomic_commit_tail(struct drm_atomic_state *state) intel_modeset_verify_crtc(crtc, old_crtc_state, crtc->state); } + if (IS_SKYLAKE(dev_priv) && intel_state->modeset && + skl_can_enable_sagv(state)) + skl_enable_sagv(dev_priv); + drm_atomic_helper_commit_hw_done(state); if (intel_state->modeset) @@ -14354,8 +14564,6 @@ out: drm_atomic_state_free(state); } -#undef for_each_intel_crtc_masked - /* * FIXME: Remove this once i915 is fully DRIVER_ATOMIC by calling * drm_atomic_helper_legacy_gamma_set() directly. @@ -14601,10 +14809,12 @@ static void intel_begin_crtc_commit(struct drm_crtc *crtc, struct drm_crtc_state *old_crtc_state) { struct drm_device *dev = crtc->dev; + struct drm_i915_private *dev_priv = to_i915(dev); struct intel_crtc *intel_crtc = to_intel_crtc(crtc); struct intel_crtc_state *old_intel_state = to_intel_crtc_state(old_crtc_state); bool modeset = needs_modeset(crtc->state); + enum pipe pipe = intel_crtc->pipe; /* Perform vblank evasion around commit operation */ intel_pipe_update_start(intel_crtc); @@ -14619,8 +14829,12 @@ static void intel_begin_crtc_commit(struct drm_crtc *crtc, if (to_intel_crtc_state(crtc->state)->update_pipe) intel_update_pipe_config(intel_crtc, old_intel_state); - else if (INTEL_INFO(dev)->gen >= 9) + else if (INTEL_GEN(dev_priv) >= 9) { skl_detach_scalers(intel_crtc); + + I915_WRITE(PIPE_WM_LINETIME(pipe), + dev_priv->wm.skl_hw.wm_linetime[pipe]); + } } static void intel_finish_crtc_commit(struct drm_crtc *crtc, @@ -15782,6 +15996,11 @@ void intel_init_display_hooks(struct drm_i915_private *dev_priv) skl_modeset_calc_cdclk; } + if (dev_priv->info.gen >= 9) + dev_priv->display.update_crtcs = skl_update_crtcs; + else + dev_priv->display.update_crtcs = intel_update_crtcs; + switch (INTEL_INFO(dev_priv)->gen) { case 2: dev_priv->display.queue_flip = intel_gen2_queue_flip; @@ -15983,15 +16202,16 @@ static void intel_init_quirks(struct drm_device *dev) static void i915_disable_vga(struct drm_device *dev) { struct drm_i915_private *dev_priv = to_i915(dev); + struct pci_dev *pdev = dev_priv->drm.pdev; u8 sr1; i915_reg_t vga_reg = i915_vgacntrl_reg(dev); /* WaEnableVGAAccessThroughIOPort:ctg,elk,ilk,snb,ivb,vlv,hsw */ - vga_get_uninterruptible(dev->pdev, VGA_RSRC_LEGACY_IO); + vga_get_uninterruptible(pdev, VGA_RSRC_LEGACY_IO); outb(SR01, VGA_SR_INDEX); sr1 = inb(VGA_SR_DATA); outb(sr1 | 1<<5, VGA_SR_DATA); - vga_put(dev->pdev, VGA_RSRC_LEGACY_IO); + vga_put(pdev, VGA_RSRC_LEGACY_IO); udelay(300); I915_WRITE(vga_reg, VGA_DISP_DISABLE); @@ -16273,15 +16493,15 @@ static bool intel_crtc_has_encoders(struct intel_crtc *crtc) return false; } -static bool intel_encoder_has_connectors(struct intel_encoder *encoder) +static struct intel_connector *intel_encoder_find_connector(struct intel_encoder *encoder) { struct drm_device *dev = encoder->base.dev; struct intel_connector *connector; for_each_connector_on_encoder(dev, &encoder->base, connector) - return true; + return connector; - return false; + return NULL; } static bool has_pch_trancoder(struct drm_i915_private *dev_priv, @@ -16386,7 +16606,6 @@ static void intel_sanitize_crtc(struct intel_crtc *crtc) static void intel_sanitize_encoder(struct intel_encoder *encoder) { struct intel_connector *connector; - struct drm_device *dev = encoder->base.dev; /* We need to check both for a crtc link (meaning that the * encoder is active and trying to read from a pipe) and the @@ -16394,7 +16613,8 @@ static void intel_sanitize_encoder(struct intel_encoder *encoder) bool has_active_crtc = encoder->base.crtc && to_intel_crtc(encoder->base.crtc)->active; - if (intel_encoder_has_connectors(encoder) && !has_active_crtc) { + connector = intel_encoder_find_connector(encoder); + if (connector && !has_active_crtc) { DRM_DEBUG_KMS("[ENCODER:%d:%s] has active connectors but no active pipe!\n", encoder->base.base.id, encoder->base.name); @@ -16403,12 +16623,14 @@ static void intel_sanitize_encoder(struct intel_encoder *encoder) * fallout from our resume register restoring. Disable * the encoder manually again. */ if (encoder->base.crtc) { + struct drm_crtc_state *crtc_state = encoder->base.crtc->state; + DRM_DEBUG_KMS("[ENCODER:%d:%s] manually disabled\n", encoder->base.base.id, encoder->base.name); - encoder->disable(encoder); + encoder->disable(encoder, to_intel_crtc_state(crtc_state), connector->base.state); if (encoder->post_disable) - encoder->post_disable(encoder); + encoder->post_disable(encoder, to_intel_crtc_state(crtc_state), connector->base.state); } encoder->base.crtc = NULL; @@ -16416,12 +16638,9 @@ static void intel_sanitize_encoder(struct intel_encoder *encoder) * a bug in one of the get_hw_state functions. Or someplace else * in our code, like the register restore mess on resume. Clamp * things to off as a safer default. */ - for_each_intel_connector(dev, connector) { - if (connector->encoder != encoder) - continue; - connector->base.dpms = DRM_MODE_DPMS_OFF; - connector->base.encoder = NULL; - } + + connector->base.dpms = DRM_MODE_DPMS_OFF; + connector->base.encoder = NULL; } /* Enabled encoders without active connectors will be fixed in * the crtc fixup. */ diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c index 364db908c191..acd0c51f74d5 100644 --- a/drivers/gpu/drm/i915/intel_dp.c +++ b/drivers/gpu/drm/i915/intel_dp.c @@ -190,6 +190,29 @@ intel_dp_max_data_rate(int max_link_clock, int max_lanes) return (max_link_clock * max_lanes * 8) / 10; } +static int +intel_dp_downstream_max_dotclock(struct intel_dp *intel_dp) +{ + struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp); + struct intel_encoder *encoder = &intel_dig_port->base; + struct drm_i915_private *dev_priv = to_i915(encoder->base.dev); + int max_dotclk = dev_priv->max_dotclk_freq; + int ds_max_dotclk; + + int type = intel_dp->downstream_ports[0] & DP_DS_PORT_TYPE_MASK; + + if (type != DP_DS_PORT_TYPE_VGA) + return max_dotclk; + + ds_max_dotclk = drm_dp_downstream_max_clock(intel_dp->dpcd, + intel_dp->downstream_ports); + + if (ds_max_dotclk != 0) + max_dotclk = min(max_dotclk, ds_max_dotclk); + + return max_dotclk; +} + static enum drm_mode_status intel_dp_mode_valid(struct drm_connector *connector, struct drm_display_mode *mode) @@ -199,7 +222,9 @@ intel_dp_mode_valid(struct drm_connector *connector, struct drm_display_mode *fixed_mode = intel_connector->panel.fixed_mode; int target_clock = mode->clock; int max_rate, mode_rate, max_lanes, max_link_clock; - int max_dotclk = to_i915(connector->dev)->max_dotclk_freq; + int max_dotclk; + + max_dotclk = intel_dp_downstream_max_dotclock(intel_dp); if (is_edp(intel_dp) && fixed_mode) { if (mode->hdisplay > fixed_mode->hdisplay) @@ -1243,7 +1268,7 @@ intel_dp_aux_fini(struct intel_dp *intel_dp) } static void -intel_dp_aux_init(struct intel_dp *intel_dp, struct intel_connector *connector) +intel_dp_aux_init(struct intel_dp *intel_dp) { struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp); enum port port = intel_dig_port->port; @@ -1419,6 +1444,44 @@ static void intel_dp_print_rates(struct intel_dp *intel_dp) DRM_DEBUG_KMS("common rates: %s\n", str); } +static void intel_dp_print_hw_revision(struct intel_dp *intel_dp) +{ + uint8_t rev; + int len; + + if ((drm_debug & DRM_UT_KMS) == 0) + return; + + if (!(intel_dp->dpcd[DP_DOWNSTREAMPORT_PRESENT] & + DP_DWN_STRM_PORT_PRESENT)) + return; + + len = drm_dp_dpcd_read(&intel_dp->aux, DP_BRANCH_HW_REV, &rev, 1); + if (len < 0) + return; + + DRM_DEBUG_KMS("sink hw revision: %d.%d\n", (rev & 0xf0) >> 4, rev & 0xf); +} + +static void intel_dp_print_sw_revision(struct intel_dp *intel_dp) +{ + uint8_t rev[2]; + int len; + + if ((drm_debug & DRM_UT_KMS) == 0) + return; + + if (!(intel_dp->dpcd[DP_DOWNSTREAMPORT_PRESENT] & + DP_DWN_STRM_PORT_PRESENT)) + return; + + len = drm_dp_dpcd_read(&intel_dp->aux, DP_BRANCH_SW_REV, &rev, 2); + if (len < 0) + return; + + DRM_DEBUG_KMS("sink sw revision: %d.%d\n", rev[0], rev[1]); +} + static int rate_to_index(int find, const int *rates) { int i = 0; @@ -1461,9 +1524,24 @@ void intel_dp_compute_rate(struct intel_dp *intel_dp, int port_clock, } } +static int intel_dp_compute_bpp(struct intel_dp *intel_dp, + struct intel_crtc_state *pipe_config) +{ + int bpp, bpc; + + bpp = pipe_config->pipe_bpp; + bpc = drm_dp_downstream_max_bpc(intel_dp->dpcd, intel_dp->downstream_ports); + + if (bpc > 0) + bpp = min(bpp, 3*bpc); + + return bpp; +} + bool intel_dp_compute_config(struct intel_encoder *encoder, - struct intel_crtc_state *pipe_config) + struct intel_crtc_state *pipe_config, + struct drm_connector_state *conn_state) { struct drm_device *dev = encoder->base.dev; struct drm_i915_private *dev_priv = to_i915(dev); @@ -1526,7 +1604,7 @@ intel_dp_compute_config(struct intel_encoder *encoder, /* Walk through all bpp values. Luckily they're all nicely spaced with 2 * bpc in between. */ - bpp = pipe_config->pipe_bpp; + bpp = intel_dp_compute_bpp(intel_dp, pipe_config); if (is_edp(intel_dp)) { /* Get bpp from vbt only for panels that dont have bpp in edid */ @@ -1640,23 +1718,28 @@ found: } void intel_dp_set_link_params(struct intel_dp *intel_dp, - const struct intel_crtc_state *pipe_config) + int link_rate, uint8_t lane_count, + bool link_mst) { - intel_dp->link_rate = pipe_config->port_clock; - intel_dp->lane_count = pipe_config->lane_count; - intel_dp->link_mst = intel_crtc_has_type(pipe_config, INTEL_OUTPUT_DP_MST); + intel_dp->link_rate = link_rate; + intel_dp->lane_count = lane_count; + intel_dp->link_mst = link_mst; } -static void intel_dp_prepare(struct intel_encoder *encoder) +static void intel_dp_prepare(struct intel_encoder *encoder, + struct intel_crtc_state *pipe_config) { struct drm_device *dev = encoder->base.dev; struct drm_i915_private *dev_priv = to_i915(dev); struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base); enum port port = dp_to_dig_port(intel_dp)->port; struct intel_crtc *crtc = to_intel_crtc(encoder->base.crtc); - const struct drm_display_mode *adjusted_mode = &crtc->config->base.adjusted_mode; + const struct drm_display_mode *adjusted_mode = &pipe_config->base.adjusted_mode; - intel_dp_set_link_params(intel_dp, crtc->config); + intel_dp_set_link_params(intel_dp, pipe_config->port_clock, + pipe_config->lane_count, + intel_crtc_has_type(pipe_config, + INTEL_OUTPUT_DP_MST)); /* * There are four kinds of DP registers: @@ -1682,7 +1765,7 @@ static void intel_dp_prepare(struct intel_encoder *encoder) /* Handle DP bits in common between all three register formats */ intel_dp->DP |= DP_VOLTAGE_0_4 | DP_PRE_EMPHASIS_0; - intel_dp->DP |= DP_PORT_WIDTH(crtc->config->lane_count); + intel_dp->DP |= DP_PORT_WIDTH(pipe_config->lane_count); /* Split out the IBX/CPU vs CPT settings */ @@ -1710,7 +1793,7 @@ static void intel_dp_prepare(struct intel_encoder *encoder) I915_WRITE(TRANS_DP_CTL(crtc->pipe), trans_dp); } else { if (!HAS_PCH_SPLIT(dev) && !IS_VALLEYVIEW(dev) && - !IS_CHERRYVIEW(dev) && crtc->config->limited_color_range) + !IS_CHERRYVIEW(dev) && pipe_config->limited_color_range) intel_dp->DP |= DP_COLOR_RANGE_16_235; if (adjusted_mode->flags & DRM_MODE_FLAG_PHSYNC) @@ -2249,10 +2332,10 @@ static void assert_edp_pll(struct drm_i915_private *dev_priv, bool state) #define assert_edp_pll_enabled(d) assert_edp_pll((d), true) #define assert_edp_pll_disabled(d) assert_edp_pll((d), false) -static void ironlake_edp_pll_on(struct intel_dp *intel_dp) +static void ironlake_edp_pll_on(struct intel_dp *intel_dp, + struct intel_crtc_state *pipe_config) { - struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp); - struct intel_crtc *crtc = to_intel_crtc(intel_dig_port->base.base.crtc); + struct intel_crtc *crtc = to_intel_crtc(pipe_config->base.crtc); struct drm_i915_private *dev_priv = to_i915(crtc->base.dev); assert_pipe_disabled(dev_priv, crtc->pipe); @@ -2260,11 +2343,11 @@ static void ironlake_edp_pll_on(struct intel_dp *intel_dp) assert_edp_pll_disabled(dev_priv); DRM_DEBUG_KMS("enabling eDP PLL for clock %d\n", - crtc->config->port_clock); + pipe_config->port_clock); intel_dp->DP &= ~DP_PLL_FREQ_MASK; - if (crtc->config->port_clock == 162000) + if (pipe_config->port_clock == 162000) intel_dp->DP |= DP_PLL_FREQ_162MHZ; else intel_dp->DP |= DP_PLL_FREQ_270MHZ; @@ -2473,16 +2556,17 @@ static void intel_dp_get_config(struct intel_encoder *encoder, } } -static void intel_disable_dp(struct intel_encoder *encoder) +static void intel_disable_dp(struct intel_encoder *encoder, + struct intel_crtc_state *old_crtc_state, + struct drm_connector_state *old_conn_state) { struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base); - struct drm_device *dev = encoder->base.dev; - struct intel_crtc *crtc = to_intel_crtc(encoder->base.crtc); + struct drm_i915_private *dev_priv = to_i915(encoder->base.dev); - if (crtc->config->has_audio) + if (old_crtc_state->has_audio) intel_audio_codec_disable(encoder); - if (HAS_PSR(dev) && !HAS_DDI(dev)) + if (HAS_PSR(dev_priv) && !HAS_DDI(dev_priv)) intel_psr_disable(intel_dp); /* Make sure the panel is off before trying to change the mode. But also @@ -2493,11 +2577,13 @@ static void intel_disable_dp(struct intel_encoder *encoder) intel_edp_panel_off(intel_dp); /* disable the port before the pipe on g4x */ - if (INTEL_INFO(dev)->gen < 5) + if (INTEL_GEN(dev_priv) < 5) intel_dp_link_down(intel_dp); } -static void ilk_post_disable_dp(struct intel_encoder *encoder) +static void ilk_post_disable_dp(struct intel_encoder *encoder, + struct intel_crtc_state *old_crtc_state, + struct drm_connector_state *old_conn_state) { struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base); enum port port = dp_to_dig_port(intel_dp)->port; @@ -2509,14 +2595,18 @@ static void ilk_post_disable_dp(struct intel_encoder *encoder) ironlake_edp_pll_off(intel_dp); } -static void vlv_post_disable_dp(struct intel_encoder *encoder) +static void vlv_post_disable_dp(struct intel_encoder *encoder, + struct intel_crtc_state *old_crtc_state, + struct drm_connector_state *old_conn_state) { struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base); intel_dp_link_down(intel_dp); } -static void chv_post_disable_dp(struct intel_encoder *encoder) +static void chv_post_disable_dp(struct intel_encoder *encoder, + struct intel_crtc_state *old_crtc_state, + struct drm_connector_state *old_conn_state) { struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base); struct drm_device *dev = encoder->base.dev; @@ -2542,6 +2632,10 @@ _intel_dp_set_link_train(struct intel_dp *intel_dp, struct drm_i915_private *dev_priv = to_i915(dev); enum port port = intel_dig_port->port; + if (dp_train_pat & DP_TRAINING_PATTERN_MASK) + DRM_DEBUG_KMS("Using DP training pattern TPS%d\n", + dp_train_pat & DP_TRAINING_PATTERN_MASK); + if (HAS_DDI(dev)) { uint32_t temp = I915_READ(DP_TP_CTL(port)); @@ -2583,7 +2677,7 @@ _intel_dp_set_link_train(struct intel_dp *intel_dp, *DP |= DP_LINK_TRAIN_PAT_2_CPT; break; case DP_TRAINING_PATTERN_3: - DRM_ERROR("DP training pattern 3 not supported\n"); + DRM_DEBUG_KMS("TPS3 not supported, using TPS2 instead\n"); *DP |= DP_LINK_TRAIN_PAT_2_CPT; break; } @@ -2608,7 +2702,7 @@ _intel_dp_set_link_train(struct intel_dp *intel_dp, if (IS_CHERRYVIEW(dev)) { *DP |= DP_LINK_TRAIN_PAT_3_CHV; } else { - DRM_ERROR("DP training pattern 3 not supported\n"); + DRM_DEBUG_KMS("TPS3 not supported, using TPS2 instead\n"); *DP |= DP_LINK_TRAIN_PAT_2; } break; @@ -2616,19 +2710,15 @@ _intel_dp_set_link_train(struct intel_dp *intel_dp, } } -static void intel_dp_enable_port(struct intel_dp *intel_dp) +static void intel_dp_enable_port(struct intel_dp *intel_dp, + struct intel_crtc_state *old_crtc_state) { struct drm_device *dev = intel_dp_to_dev(intel_dp); struct drm_i915_private *dev_priv = to_i915(dev); - struct intel_crtc *crtc = - to_intel_crtc(dp_to_dig_port(intel_dp)->base.base.crtc); /* enable with pattern 1 (as per spec) */ - _intel_dp_set_link_train(intel_dp, &intel_dp->DP, - DP_TRAINING_PATTERN_1); - I915_WRITE(intel_dp->output_reg, intel_dp->DP); - POSTING_READ(intel_dp->output_reg); + intel_dp_program_link_training_pattern(intel_dp, DP_TRAINING_PATTERN_1); /* * Magic for VLV/CHV. We _must_ first set up the register @@ -2637,14 +2727,15 @@ static void intel_dp_enable_port(struct intel_dp *intel_dp) * fail when the power sequencer is freshly used for this port. */ intel_dp->DP |= DP_PORT_EN; - if (crtc->config->has_audio) + if (old_crtc_state->has_audio) intel_dp->DP |= DP_AUDIO_OUTPUT_ENABLE; I915_WRITE(intel_dp->output_reg, intel_dp->DP); POSTING_READ(intel_dp->output_reg); } -static void intel_enable_dp(struct intel_encoder *encoder) +static void intel_enable_dp(struct intel_encoder *encoder, + struct intel_crtc_state *pipe_config) { struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base); struct drm_device *dev = encoder->base.dev; @@ -2661,7 +2752,7 @@ static void intel_enable_dp(struct intel_encoder *encoder) if (IS_VALLEYVIEW(dev) || IS_CHERRYVIEW(dev)) vlv_init_panel_power_sequencer(intel_dp); - intel_dp_enable_port(intel_dp); + intel_dp_enable_port(intel_dp, pipe_config); edp_panel_vdd_on(intel_dp); edp_panel_on(intel_dp); @@ -2673,7 +2764,7 @@ static void intel_enable_dp(struct intel_encoder *encoder) unsigned int lane_mask = 0x0; if (IS_CHERRYVIEW(dev)) - lane_mask = intel_dp_unused_lane_mask(crtc->config->lane_count); + lane_mask = intel_dp_unused_lane_mask(pipe_config->lane_count); vlv_wait_port_ready(dev_priv, dp_to_dig_port(intel_dp), lane_mask); @@ -2683,22 +2774,26 @@ static void intel_enable_dp(struct intel_encoder *encoder) intel_dp_start_link_train(intel_dp); intel_dp_stop_link_train(intel_dp); - if (crtc->config->has_audio) { + if (pipe_config->has_audio) { DRM_DEBUG_DRIVER("Enabling DP audio on pipe %c\n", pipe_name(pipe)); intel_audio_codec_enable(encoder); } } -static void g4x_enable_dp(struct intel_encoder *encoder) +static void g4x_enable_dp(struct intel_encoder *encoder, + struct intel_crtc_state *pipe_config, + struct drm_connector_state *conn_state) { struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base); - intel_enable_dp(encoder); + intel_enable_dp(encoder, pipe_config); intel_edp_backlight_on(intel_dp); } -static void vlv_enable_dp(struct intel_encoder *encoder) +static void vlv_enable_dp(struct intel_encoder *encoder, + struct intel_crtc_state *pipe_config, + struct drm_connector_state *conn_state) { struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base); @@ -2706,16 +2801,18 @@ static void vlv_enable_dp(struct intel_encoder *encoder) intel_psr_enable(intel_dp); } -static void g4x_pre_enable_dp(struct intel_encoder *encoder) +static void g4x_pre_enable_dp(struct intel_encoder *encoder, + struct intel_crtc_state *pipe_config, + struct drm_connector_state *conn_state) { struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base); enum port port = dp_to_dig_port(intel_dp)->port; - intel_dp_prepare(encoder); + intel_dp_prepare(encoder, pipe_config); /* Only ilk+ has port A */ if (port == PORT_A) - ironlake_edp_pll_on(intel_dp); + ironlake_edp_pll_on(intel_dp, pipe_config); } static void vlv_detach_power_sequencer(struct intel_dp *intel_dp) @@ -2821,38 +2918,48 @@ static void vlv_init_panel_power_sequencer(struct intel_dp *intel_dp) intel_dp_init_panel_power_sequencer_registers(dev, intel_dp); } -static void vlv_pre_enable_dp(struct intel_encoder *encoder) +static void vlv_pre_enable_dp(struct intel_encoder *encoder, + struct intel_crtc_state *pipe_config, + struct drm_connector_state *conn_state) { vlv_phy_pre_encoder_enable(encoder); - intel_enable_dp(encoder); + intel_enable_dp(encoder, pipe_config); } -static void vlv_dp_pre_pll_enable(struct intel_encoder *encoder) +static void vlv_dp_pre_pll_enable(struct intel_encoder *encoder, + struct intel_crtc_state *pipe_config, + struct drm_connector_state *conn_state) { - intel_dp_prepare(encoder); + intel_dp_prepare(encoder, pipe_config); vlv_phy_pre_pll_enable(encoder); } -static void chv_pre_enable_dp(struct intel_encoder *encoder) +static void chv_pre_enable_dp(struct intel_encoder *encoder, + struct intel_crtc_state *pipe_config, + struct drm_connector_state *conn_state) { chv_phy_pre_encoder_enable(encoder); - intel_enable_dp(encoder); + intel_enable_dp(encoder, pipe_config); /* Second common lane will stay alive on its own now */ chv_phy_release_cl2_override(encoder); } -static void chv_dp_pre_pll_enable(struct intel_encoder *encoder) +static void chv_dp_pre_pll_enable(struct intel_encoder *encoder, + struct intel_crtc_state *pipe_config, + struct drm_connector_state *conn_state) { - intel_dp_prepare(encoder); + intel_dp_prepare(encoder, pipe_config); chv_phy_pre_pll_enable(encoder); } -static void chv_dp_post_pll_disable(struct intel_encoder *encoder) +static void chv_dp_post_pll_disable(struct intel_encoder *encoder, + struct intel_crtc_state *pipe_config, + struct drm_connector_state *conn_state) { chv_phy_post_pll_disable(encoder); } @@ -4171,7 +4278,7 @@ static bool bxt_digital_port_connected(struct drm_i915_private *dev_priv, * * Return %true if @port is connected, %false otherwise. */ -bool intel_digital_port_connected(struct drm_i915_private *dev_priv, +static bool intel_digital_port_connected(struct drm_i915_private *dev_priv, struct intel_digital_port *port) { if (HAS_PCH_IBX(dev_priv)) @@ -4282,6 +4389,9 @@ intel_dp_long_pulse(struct intel_connector *intel_connector) intel_dp_probe_oui(intel_dp); + intel_dp_print_hw_revision(intel_dp); + intel_dp_print_sw_revision(intel_dp); + intel_dp_configure_mst(intel_dp); if (intel_dp->is_mst) { @@ -5022,7 +5132,8 @@ static void intel_dp_pps_init(struct drm_device *dev, /** * intel_dp_set_drrs_state - program registers for RR switch to take effect - * @dev: DRM device + * @dev_priv: i915 device + * @crtc_state: a pointer to the active intel_crtc_state * @refresh_rate: RR to be programmed * * This function gets called when refresh rate (RR) has to be changed from @@ -5032,14 +5143,14 @@ static void intel_dp_pps_init(struct drm_device *dev, * * The caller of this function needs to take a lock on dev_priv->drrs. */ -static void intel_dp_set_drrs_state(struct drm_device *dev, int refresh_rate) +static void intel_dp_set_drrs_state(struct drm_i915_private *dev_priv, + struct intel_crtc_state *crtc_state, + int refresh_rate) { - struct drm_i915_private *dev_priv = to_i915(dev); struct intel_encoder *encoder; struct intel_digital_port *dig_port = NULL; struct intel_dp *intel_dp = dev_priv->drrs.dp; - struct intel_crtc_state *config = NULL; - struct intel_crtc *intel_crtc = NULL; + struct intel_crtc *intel_crtc = to_intel_crtc(crtc_state->base.crtc); enum drrs_refresh_rate_type index = DRRS_HIGH_RR; if (refresh_rate <= 0) { @@ -5066,8 +5177,6 @@ static void intel_dp_set_drrs_state(struct drm_device *dev, int refresh_rate) return; } - config = intel_crtc->config; - if (dev_priv->drrs.type < SEAMLESS_DRRS_SUPPORT) { DRM_DEBUG_KMS("Only Seamless DRRS supported.\n"); return; @@ -5083,12 +5192,12 @@ static void intel_dp_set_drrs_state(struct drm_device *dev, int refresh_rate) return; } - if (!intel_crtc->active) { + if (!crtc_state->base.active) { DRM_DEBUG_KMS("eDP encoder disabled. CRTC not Active\n"); return; } - if (INTEL_INFO(dev)->gen >= 8 && !IS_CHERRYVIEW(dev)) { + if (INTEL_GEN(dev_priv) >= 8 && !IS_CHERRYVIEW(dev_priv)) { switch (index) { case DRRS_HIGH_RR: intel_dp_set_m_n(intel_crtc, M1_N1); @@ -5100,18 +5209,18 @@ static void intel_dp_set_drrs_state(struct drm_device *dev, int refresh_rate) default: DRM_ERROR("Unsupported refreshrate type\n"); } - } else if (INTEL_INFO(dev)->gen > 6) { - i915_reg_t reg = PIPECONF(intel_crtc->config->cpu_transcoder); + } else if (INTEL_GEN(dev_priv) > 6) { + i915_reg_t reg = PIPECONF(crtc_state->cpu_transcoder); u32 val; val = I915_READ(reg); if (index > DRRS_HIGH_RR) { - if (IS_VALLEYVIEW(dev) || IS_CHERRYVIEW(dev)) + if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv)) val |= PIPECONF_EDP_RR_MODE_SWITCH_VLV; else val |= PIPECONF_EDP_RR_MODE_SWITCH; } else { - if (IS_VALLEYVIEW(dev) || IS_CHERRYVIEW(dev)) + if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv)) val &= ~PIPECONF_EDP_RR_MODE_SWITCH_VLV; else val &= ~PIPECONF_EDP_RR_MODE_SWITCH; @@ -5127,18 +5236,17 @@ static void intel_dp_set_drrs_state(struct drm_device *dev, int refresh_rate) /** * intel_edp_drrs_enable - init drrs struct if supported * @intel_dp: DP struct + * @crtc_state: A pointer to the active crtc state. * * Initializes frontbuffer_bits and drrs.dp */ -void intel_edp_drrs_enable(struct intel_dp *intel_dp) +void intel_edp_drrs_enable(struct intel_dp *intel_dp, + struct intel_crtc_state *crtc_state) { struct drm_device *dev = intel_dp_to_dev(intel_dp); struct drm_i915_private *dev_priv = to_i915(dev); - struct intel_digital_port *dig_port = dp_to_dig_port(intel_dp); - struct drm_crtc *crtc = dig_port->base.base.crtc; - struct intel_crtc *intel_crtc = to_intel_crtc(crtc); - if (!intel_crtc->config->has_drrs) { + if (!crtc_state->has_drrs) { DRM_DEBUG_KMS("Panel doesn't support DRRS\n"); return; } @@ -5160,17 +5268,16 @@ unlock: /** * intel_edp_drrs_disable - Disable DRRS * @intel_dp: DP struct + * @old_crtc_state: Pointer to old crtc_state. * */ -void intel_edp_drrs_disable(struct intel_dp *intel_dp) +void intel_edp_drrs_disable(struct intel_dp *intel_dp, + struct intel_crtc_state *old_crtc_state) { struct drm_device *dev = intel_dp_to_dev(intel_dp); struct drm_i915_private *dev_priv = to_i915(dev); - struct intel_digital_port *dig_port = dp_to_dig_port(intel_dp); - struct drm_crtc *crtc = dig_port->base.base.crtc; - struct intel_crtc *intel_crtc = to_intel_crtc(crtc); - if (!intel_crtc->config->has_drrs) + if (!old_crtc_state->has_drrs) return; mutex_lock(&dev_priv->drrs.mutex); @@ -5180,9 +5287,8 @@ void intel_edp_drrs_disable(struct intel_dp *intel_dp) } if (dev_priv->drrs.refresh_rate_type == DRRS_LOW_RR) - intel_dp_set_drrs_state(&dev_priv->drm, - intel_dp->attached_connector->panel. - fixed_mode->vrefresh); + intel_dp_set_drrs_state(dev_priv, old_crtc_state, + intel_dp->attached_connector->panel.fixed_mode->vrefresh); dev_priv->drrs.dp = NULL; mutex_unlock(&dev_priv->drrs.mutex); @@ -5211,10 +5317,12 @@ static void intel_edp_drrs_downclock_work(struct work_struct *work) if (dev_priv->drrs.busy_frontbuffer_bits) goto unlock; - if (dev_priv->drrs.refresh_rate_type != DRRS_LOW_RR) - intel_dp_set_drrs_state(&dev_priv->drm, - intel_dp->attached_connector->panel. - downclock_mode->vrefresh); + if (dev_priv->drrs.refresh_rate_type != DRRS_LOW_RR) { + struct drm_crtc *crtc = dp_to_dig_port(intel_dp)->base.base.crtc; + + intel_dp_set_drrs_state(dev_priv, to_intel_crtc(crtc)->config, + intel_dp->attached_connector->panel.downclock_mode->vrefresh); + } unlock: mutex_unlock(&dev_priv->drrs.mutex); @@ -5255,9 +5363,8 @@ void intel_edp_drrs_invalidate(struct drm_i915_private *dev_priv, /* invalidate means busy screen hence upclock */ if (frontbuffer_bits && dev_priv->drrs.refresh_rate_type == DRRS_LOW_RR) - intel_dp_set_drrs_state(&dev_priv->drm, - dev_priv->drrs.dp->attached_connector->panel. - fixed_mode->vrefresh); + intel_dp_set_drrs_state(dev_priv, to_intel_crtc(crtc)->config, + dev_priv->drrs.dp->attached_connector->panel.fixed_mode->vrefresh); mutex_unlock(&dev_priv->drrs.mutex); } @@ -5299,9 +5406,8 @@ void intel_edp_drrs_flush(struct drm_i915_private *dev_priv, /* flush means busy screen hence upclock */ if (frontbuffer_bits && dev_priv->drrs.refresh_rate_type == DRRS_LOW_RR) - intel_dp_set_drrs_state(&dev_priv->drm, - dev_priv->drrs.dp->attached_connector->panel. - fixed_mode->vrefresh); + intel_dp_set_drrs_state(dev_priv, to_intel_crtc(crtc)->config, + dev_priv->drrs.dp->attached_connector->panel.fixed_mode->vrefresh); /* * flush also means no more activity hence schedule downclock, if all @@ -5598,7 +5704,7 @@ intel_dp_init_connector(struct intel_digital_port *intel_dig_port, connector->interlace_allowed = true; connector->doublescan_allowed = 0; - intel_dp_aux_init(intel_dp, intel_connector); + intel_dp_aux_init(intel_dp); INIT_DELAYED_WORK(&intel_dp->panel_vdd_work, edp_panel_vdd_work); diff --git a/drivers/gpu/drm/i915/intel_dp_link_training.c b/drivers/gpu/drm/i915/intel_dp_link_training.c index 60fb39cd220b..c438b02184cb 100644 --- a/drivers/gpu/drm/i915/intel_dp_link_training.c +++ b/drivers/gpu/drm/i915/intel_dp_link_training.c @@ -24,6 +24,15 @@ #include "intel_drv.h" static void +intel_dp_dump_link_status(const uint8_t link_status[DP_LINK_STATUS_SIZE]) +{ + + DRM_DEBUG_KMS("ln0_1:0x%x ln2_3:0x%x align:0x%x sink:0x%x adj_req0_1:0x%x adj_req2_3:0x%x", + link_status[0], link_status[1], link_status[2], + link_status[3], link_status[4], link_status[5]); +} + +static void intel_get_adjust_train(struct intel_dp *intel_dp, const uint8_t link_status[DP_LINK_STATUS_SIZE]) { @@ -103,13 +112,24 @@ intel_dp_update_link_train(struct intel_dp *intel_dp) return ret == intel_dp->lane_count; } +static bool intel_dp_link_max_vswing_reached(struct intel_dp *intel_dp) +{ + int lane; + + for (lane = 0; lane < intel_dp->lane_count; lane++) + if ((intel_dp->train_set[lane] & + DP_TRAIN_MAX_SWING_REACHED) == 0) + return false; + + return true; +} + /* Enable corresponding port and start training pattern 1 */ -static void +static bool intel_dp_link_training_clock_recovery(struct intel_dp *intel_dp) { - int i; uint8_t voltage; - int voltage_tries, loop_tries; + int voltage_tries, max_vswing_tries; uint8_t link_config[2]; uint8_t link_bw, rate_select; @@ -125,6 +145,7 @@ intel_dp_link_training_clock_recovery(struct intel_dp *intel_dp) if (drm_dp_enhanced_frame_cap(intel_dp->dpcd)) link_config[1] |= DP_LANE_COUNT_ENHANCED_FRAME_EN; drm_dp_dpcd_write(&intel_dp->aux, DP_LINK_BW_SET, link_config, 2); + if (intel_dp->num_sink_rates) drm_dp_dpcd_write(&intel_dp->aux, DP_LINK_RATE_SET, &rate_select, 1); @@ -140,60 +161,54 @@ intel_dp_link_training_clock_recovery(struct intel_dp *intel_dp) DP_TRAINING_PATTERN_1 | DP_LINK_SCRAMBLING_DISABLE)) { DRM_ERROR("failed to enable link training\n"); - return; + return false; } - voltage = 0xff; - voltage_tries = 0; - loop_tries = 0; + voltage_tries = 1; + max_vswing_tries = 0; for (;;) { uint8_t link_status[DP_LINK_STATUS_SIZE]; drm_dp_link_train_clock_recovery_delay(intel_dp->dpcd); + if (!intel_dp_get_link_status(intel_dp, link_status)) { DRM_ERROR("failed to get link status\n"); - break; + return false; } if (drm_dp_clock_recovery_ok(link_status, intel_dp->lane_count)) { DRM_DEBUG_KMS("clock recovery OK\n"); - break; + return true; } - /* Check to see if we've tried the max voltage */ - for (i = 0; i < intel_dp->lane_count; i++) - if ((intel_dp->train_set[i] & DP_TRAIN_MAX_SWING_REACHED) == 0) - break; - if (i == intel_dp->lane_count) { - ++loop_tries; - if (loop_tries == 5) { - DRM_ERROR("too many full retries, give up\n"); - break; - } - intel_dp_reset_link_train(intel_dp, - DP_TRAINING_PATTERN_1 | - DP_LINK_SCRAMBLING_DISABLE); - voltage_tries = 0; - continue; + if (voltage_tries == 5) { + DRM_DEBUG_KMS("Same voltage tried 5 times\n"); + return false; + } + + if (max_vswing_tries == 1) { + DRM_DEBUG_KMS("Max Voltage Swing reached\n"); + return false; } - /* Check to see if we've tried the same voltage 5 times */ - if ((intel_dp->train_set[0] & DP_TRAIN_VOLTAGE_SWING_MASK) == voltage) { - ++voltage_tries; - if (voltage_tries == 5) { - DRM_ERROR("too many voltage retries, give up\n"); - break; - } - } else - voltage_tries = 0; voltage = intel_dp->train_set[0] & DP_TRAIN_VOLTAGE_SWING_MASK; /* Update training set as requested by target */ intel_get_adjust_train(intel_dp, link_status); if (!intel_dp_update_link_train(intel_dp)) { DRM_ERROR("failed to update link training\n"); - break; + return false; } + + if ((intel_dp->train_set[0] & DP_TRAIN_VOLTAGE_SWING_MASK) == + voltage) + ++voltage_tries; + else + voltage_tries = 1; + + if (intel_dp_link_max_vswing_reached(intel_dp)) + ++max_vswing_tries; + } } @@ -229,12 +244,12 @@ static u32 intel_dp_training_pattern(struct intel_dp *intel_dp) return training_pattern; } -static void +static bool intel_dp_link_training_channel_equalization(struct intel_dp *intel_dp) { - bool channel_eq = false; - int tries, cr_tries; + int tries; u32 training_pattern; + uint8_t link_status[DP_LINK_STATUS_SIZE]; training_pattern = intel_dp_training_pattern(intel_dp); @@ -243,19 +258,11 @@ intel_dp_link_training_channel_equalization(struct intel_dp *intel_dp) training_pattern | DP_LINK_SCRAMBLING_DISABLE)) { DRM_ERROR("failed to start channel equalization\n"); - return; + return false; } - tries = 0; - cr_tries = 0; - channel_eq = false; - for (;;) { - uint8_t link_status[DP_LINK_STATUS_SIZE]; - - if (cr_tries > 5) { - DRM_ERROR("failed to train DP, aborting\n"); - break; - } + intel_dp->channel_eq_status = false; + for (tries = 0; tries < 5; tries++) { drm_dp_link_train_channel_eq_delay(intel_dp->dpcd); if (!intel_dp_get_link_status(intel_dp, link_status)) { @@ -266,44 +273,38 @@ intel_dp_link_training_channel_equalization(struct intel_dp *intel_dp) /* Make sure clock is still ok */ if (!drm_dp_clock_recovery_ok(link_status, intel_dp->lane_count)) { - intel_dp_link_training_clock_recovery(intel_dp); - intel_dp_set_link_train(intel_dp, - training_pattern | - DP_LINK_SCRAMBLING_DISABLE); - cr_tries++; - continue; + intel_dp_dump_link_status(link_status); + DRM_DEBUG_KMS("Clock recovery check failed, cannot " + "continue channel equalization\n"); + break; } if (drm_dp_channel_eq_ok(link_status, intel_dp->lane_count)) { - channel_eq = true; + intel_dp->channel_eq_status = true; + DRM_DEBUG_KMS("Channel EQ done. DP Training " + "successful\n"); break; } - /* Try 5 times, then try clock recovery if that fails */ - if (tries > 5) { - intel_dp_link_training_clock_recovery(intel_dp); - intel_dp_set_link_train(intel_dp, - training_pattern | - DP_LINK_SCRAMBLING_DISABLE); - tries = 0; - cr_tries++; - continue; - } - /* Update training set as requested by target */ intel_get_adjust_train(intel_dp, link_status); if (!intel_dp_update_link_train(intel_dp)) { DRM_ERROR("failed to update link training\n"); break; } - ++tries; + } + + /* Try 5 times, else fail and try at lower BW */ + if (tries == 5) { + intel_dp_dump_link_status(link_status); + DRM_DEBUG_KMS("Channel equalization failed 5 times\n"); } intel_dp_set_idle_link_train(intel_dp); - if (channel_eq) - DRM_DEBUG_KMS("Channel EQ done. DP Training successful\n"); + return intel_dp->channel_eq_status; + } void intel_dp_stop_link_train(struct intel_dp *intel_dp) diff --git a/drivers/gpu/drm/i915/intel_dp_mst.c b/drivers/gpu/drm/i915/intel_dp_mst.c index 629337dbca3d..54a9d7610d8f 100644 --- a/drivers/gpu/drm/i915/intel_dp_mst.c +++ b/drivers/gpu/drm/i915/intel_dp_mst.c @@ -31,18 +31,16 @@ #include <drm/drm_edid.h> static bool intel_dp_mst_compute_config(struct intel_encoder *encoder, - struct intel_crtc_state *pipe_config) + struct intel_crtc_state *pipe_config, + struct drm_connector_state *conn_state) { struct intel_dp_mst_encoder *intel_mst = enc_to_mst(&encoder->base); struct intel_digital_port *intel_dig_port = intel_mst->primary; struct intel_dp *intel_dp = &intel_dig_port->dp; struct drm_atomic_state *state; - int bpp, i; + int bpp; int lane_count, slots; const struct drm_display_mode *adjusted_mode = &pipe_config->base.adjusted_mode; - struct drm_connector *drm_connector; - struct intel_connector *connector, *found = NULL; - struct drm_connector_state *connector_state; int mst_pbn; pipe_config->dp_encoder_is_mst = true; @@ -54,7 +52,6 @@ static bool intel_dp_mst_compute_config(struct intel_encoder *encoder, */ lane_count = drm_dp_max_lane_count(intel_dp->dpcd); - pipe_config->lane_count = lane_count; pipe_config->pipe_bpp = 24; @@ -62,20 +59,6 @@ static bool intel_dp_mst_compute_config(struct intel_encoder *encoder, state = pipe_config->base.state; - for_each_connector_in_state(state, drm_connector, connector_state, i) { - connector = to_intel_connector(drm_connector); - - if (connector_state->best_encoder == &encoder->base) { - found = connector; - break; - } - } - - if (!found) { - DRM_ERROR("can't find connector\n"); - return false; - } - mst_pbn = drm_dp_calc_pbn_mode(adjusted_mode->crtc_clock, bpp); pipe_config->pbn = mst_pbn; @@ -92,16 +75,20 @@ static bool intel_dp_mst_compute_config(struct intel_encoder *encoder, } -static void intel_mst_disable_dp(struct intel_encoder *encoder) +static void intel_mst_disable_dp(struct intel_encoder *encoder, + struct intel_crtc_state *old_crtc_state, + struct drm_connector_state *old_conn_state) { struct intel_dp_mst_encoder *intel_mst = enc_to_mst(&encoder->base); struct intel_digital_port *intel_dig_port = intel_mst->primary; struct intel_dp *intel_dp = &intel_dig_port->dp; + struct intel_connector *connector = + to_intel_connector(old_conn_state->connector); int ret; DRM_DEBUG_KMS("%d\n", intel_dp->active_mst_links); - drm_dp_mst_reset_vcpi_slots(&intel_dp->mst_mgr, intel_mst->connector->port); + drm_dp_mst_reset_vcpi_slots(&intel_dp->mst_mgr, connector->port); ret = drm_dp_update_payload_part1(&intel_dp->mst_mgr); if (ret) { @@ -109,11 +96,15 @@ static void intel_mst_disable_dp(struct intel_encoder *encoder) } } -static void intel_mst_post_disable_dp(struct intel_encoder *encoder) +static void intel_mst_post_disable_dp(struct intel_encoder *encoder, + struct intel_crtc_state *old_crtc_state, + struct drm_connector_state *old_conn_state) { struct intel_dp_mst_encoder *intel_mst = enc_to_mst(&encoder->base); struct intel_digital_port *intel_dig_port = intel_mst->primary; struct intel_dp *intel_dp = &intel_dig_port->dp; + struct intel_connector *connector = + to_intel_connector(old_conn_state->connector); DRM_DEBUG_KMS("%d\n", intel_dp->active_mst_links); @@ -122,59 +113,51 @@ static void intel_mst_post_disable_dp(struct intel_encoder *encoder) /* and this can also fail */ drm_dp_update_payload_part2(&intel_dp->mst_mgr); - drm_dp_mst_deallocate_vcpi(&intel_dp->mst_mgr, intel_mst->connector->port); + drm_dp_mst_deallocate_vcpi(&intel_dp->mst_mgr, connector->port); intel_dp->active_mst_links--; intel_mst->connector = NULL; if (intel_dp->active_mst_links == 0) { - intel_dig_port->base.post_disable(&intel_dig_port->base); + intel_dig_port->base.post_disable(&intel_dig_port->base, + NULL, NULL); + intel_dp_sink_dpms(intel_dp, DRM_MODE_DPMS_OFF); } } -static void intel_mst_pre_enable_dp(struct intel_encoder *encoder) +static void intel_mst_pre_enable_dp(struct intel_encoder *encoder, + struct intel_crtc_state *pipe_config, + struct drm_connector_state *conn_state) { struct intel_dp_mst_encoder *intel_mst = enc_to_mst(&encoder->base); struct intel_digital_port *intel_dig_port = intel_mst->primary; struct intel_dp *intel_dp = &intel_dig_port->dp; - struct drm_device *dev = encoder->base.dev; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = to_i915(encoder->base.dev); enum port port = intel_dig_port->port; + struct intel_connector *connector = + to_intel_connector(conn_state->connector); int ret; uint32_t temp; - struct intel_connector *found = NULL, *connector; int slots; - struct drm_crtc *crtc = encoder->base.crtc; - struct intel_crtc *intel_crtc = to_intel_crtc(crtc); - - for_each_intel_connector(dev, connector) { - if (connector->base.state->best_encoder == &encoder->base) { - found = connector; - break; - } - } - - if (!found) { - DRM_ERROR("can't find connector\n"); - return; - } /* MST encoders are bound to a crtc, not to a connector, * force the mapping here for get_hw_state. */ - found->encoder = encoder; + connector->encoder = encoder; + intel_mst->connector = connector; DRM_DEBUG_KMS("%d\n", intel_dp->active_mst_links); - intel_mst->connector = found; - if (intel_dp->active_mst_links == 0) { - intel_ddi_clk_select(&intel_dig_port->base, intel_crtc->config); + intel_ddi_clk_select(&intel_dig_port->base, + pipe_config->shared_dpll); intel_prepare_dp_ddi_buffers(&intel_dig_port->base); - - intel_dp_set_link_params(intel_dp, intel_crtc->config); + intel_dp_set_link_params(intel_dp, + pipe_config->port_clock, + pipe_config->lane_count, + true); intel_ddi_init_dp_buf_reg(&intel_dig_port->base); @@ -185,8 +168,8 @@ static void intel_mst_pre_enable_dp(struct intel_encoder *encoder) } ret = drm_dp_mst_allocate_vcpi(&intel_dp->mst_mgr, - intel_mst->connector->port, - intel_crtc->config->pbn, &slots); + connector->port, + pipe_config->pbn, &slots); if (ret == false) { DRM_ERROR("failed to allocate vcpi\n"); return; @@ -200,13 +183,14 @@ static void intel_mst_pre_enable_dp(struct intel_encoder *encoder) ret = drm_dp_update_payload_part1(&intel_dp->mst_mgr); } -static void intel_mst_enable_dp(struct intel_encoder *encoder) +static void intel_mst_enable_dp(struct intel_encoder *encoder, + struct intel_crtc_state *pipe_config, + struct drm_connector_state *conn_state) { struct intel_dp_mst_encoder *intel_mst = enc_to_mst(&encoder->base); struct intel_digital_port *intel_dig_port = intel_mst->primary; struct intel_dp *intel_dp = &intel_dig_port->dp; - struct drm_device *dev = intel_dig_port->base.base.dev; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = to_i915(encoder->base.dev); enum port port = intel_dig_port->port; int ret; @@ -239,9 +223,8 @@ static void intel_dp_mst_enc_get_config(struct intel_encoder *encoder, { struct intel_dp_mst_encoder *intel_mst = enc_to_mst(&encoder->base); struct intel_digital_port *intel_dig_port = intel_mst->primary; - struct intel_crtc *crtc = to_intel_crtc(encoder->base.crtc); - struct drm_device *dev = encoder->base.dev; - struct drm_i915_private *dev_priv = to_i915(dev); + struct intel_crtc *crtc = to_intel_crtc(pipe_config->base.crtc); + struct drm_i915_private *dev_priv = to_i915(encoder->base.dev); enum transcoder cpu_transcoder = pipe_config->cpu_transcoder; u32 temp, flags = 0; diff --git a/drivers/gpu/drm/i915/intel_dpll_mgr.c b/drivers/gpu/drm/i915/intel_dpll_mgr.c index 5c1f2d235ffa..c26d18a574b6 100644 --- a/drivers/gpu/drm/i915/intel_dpll_mgr.c +++ b/drivers/gpu/drm/i915/intel_dpll_mgr.c @@ -24,6 +24,44 @@ #include "intel_drv.h" struct intel_shared_dpll * +skl_find_link_pll(struct drm_i915_private *dev_priv, int clock) +{ + struct intel_shared_dpll *pll = NULL; + struct intel_dpll_hw_state dpll_hw_state; + enum intel_dpll_id i; + bool found = false; + + if (!skl_ddi_dp_set_dpll_hw_state(clock, &dpll_hw_state)) + return pll; + + for (i = DPLL_ID_SKL_DPLL1; i <= DPLL_ID_SKL_DPLL3; i++) { + pll = &dev_priv->shared_dplls[i]; + + /* Only want to check enabled timings first */ + if (pll->config.crtc_mask == 0) + continue; + + if (memcmp(&dpll_hw_state, &pll->config.hw_state, + sizeof(pll->config.hw_state)) == 0) { + found = true; + break; + } + } + + /* Ok no matching timings, maybe there's a free one? */ + for (i = DPLL_ID_SKL_DPLL1; + ((found == false) && (i <= DPLL_ID_SKL_DPLL3)); i++) { + pll = &dev_priv->shared_dplls[i]; + if (pll->config.crtc_mask == 0) { + pll->config.hw_state = dpll_hw_state; + break; + } + } + + return pll; +} + +struct intel_shared_dpll * intel_get_shared_dpll_by_id(struct drm_i915_private *dev_priv, enum intel_dpll_id id) { @@ -452,26 +490,6 @@ static bool hsw_ddi_spll_get_hw_state(struct drm_i915_private *dev_priv, return val & SPLL_PLL_ENABLE; } -static uint32_t hsw_pll_to_ddi_pll_sel(struct intel_shared_dpll *pll) -{ - switch (pll->id) { - case DPLL_ID_WRPLL1: - return PORT_CLK_SEL_WRPLL1; - case DPLL_ID_WRPLL2: - return PORT_CLK_SEL_WRPLL2; - case DPLL_ID_SPLL: - return PORT_CLK_SEL_SPLL; - case DPLL_ID_LCPLL_810: - return PORT_CLK_SEL_LCPLL_810; - case DPLL_ID_LCPLL_1350: - return PORT_CLK_SEL_LCPLL_1350; - case DPLL_ID_LCPLL_2700: - return PORT_CLK_SEL_LCPLL_2700; - default: - return PORT_CLK_SEL_NONE; - } -} - #define LC_FREQ 2700 #define LC_FREQ_2K U64_C(LC_FREQ * 2000) @@ -687,11 +705,65 @@ hsw_ddi_calculate_wrpll(int clock /* in Hz */, *r2_out = best.r2; } +static struct intel_shared_dpll *hsw_ddi_hdmi_get_dpll(int clock, + struct intel_crtc *crtc, + struct intel_crtc_state *crtc_state) +{ + struct intel_shared_dpll *pll; + uint32_t val; + unsigned int p, n2, r2; + + hsw_ddi_calculate_wrpll(clock * 1000, &r2, &n2, &p); + + val = WRPLL_PLL_ENABLE | WRPLL_PLL_LCPLL | + WRPLL_DIVIDER_REFERENCE(r2) | WRPLL_DIVIDER_FEEDBACK(n2) | + WRPLL_DIVIDER_POST(p); + + crtc_state->dpll_hw_state.wrpll = val; + + pll = intel_find_shared_dpll(crtc, crtc_state, + DPLL_ID_WRPLL1, DPLL_ID_WRPLL2); + + if (!pll) + return NULL; + + return pll; +} + +struct intel_shared_dpll *hsw_ddi_dp_get_dpll(struct intel_encoder *encoder, + int clock) +{ + struct drm_i915_private *dev_priv = to_i915(encoder->base.dev); + struct intel_shared_dpll *pll; + enum intel_dpll_id pll_id; + + switch (clock / 2) { + case 81000: + pll_id = DPLL_ID_LCPLL_810; + break; + case 135000: + pll_id = DPLL_ID_LCPLL_1350; + break; + case 270000: + pll_id = DPLL_ID_LCPLL_2700; + break; + default: + DRM_DEBUG_KMS("Invalid clock for DP: %d\n", clock); + return NULL; + } + + pll = intel_get_shared_dpll_by_id(dev_priv, pll_id); + + if (!pll) + return NULL; + + return pll; +} + static struct intel_shared_dpll * hsw_get_dpll(struct intel_crtc *crtc, struct intel_crtc_state *crtc_state, struct intel_encoder *encoder) { - struct drm_i915_private *dev_priv = to_i915(crtc->base.dev); struct intel_shared_dpll *pll; int clock = crtc_state->port_clock; @@ -699,41 +771,12 @@ hsw_get_dpll(struct intel_crtc *crtc, struct intel_crtc_state *crtc_state, sizeof(crtc_state->dpll_hw_state)); if (encoder->type == INTEL_OUTPUT_HDMI) { - uint32_t val; - unsigned p, n2, r2; - - hsw_ddi_calculate_wrpll(clock * 1000, &r2, &n2, &p); - - val = WRPLL_PLL_ENABLE | WRPLL_PLL_LCPLL | - WRPLL_DIVIDER_REFERENCE(r2) | WRPLL_DIVIDER_FEEDBACK(n2) | - WRPLL_DIVIDER_POST(p); - - crtc_state->dpll_hw_state.wrpll = val; - - pll = intel_find_shared_dpll(crtc, crtc_state, - DPLL_ID_WRPLL1, DPLL_ID_WRPLL2); + pll = hsw_ddi_hdmi_get_dpll(clock, crtc, crtc_state); } else if (encoder->type == INTEL_OUTPUT_DP || encoder->type == INTEL_OUTPUT_DP_MST || encoder->type == INTEL_OUTPUT_EDP) { - enum intel_dpll_id pll_id; - - switch (clock / 2) { - case 81000: - pll_id = DPLL_ID_LCPLL_810; - break; - case 135000: - pll_id = DPLL_ID_LCPLL_1350; - break; - case 270000: - pll_id = DPLL_ID_LCPLL_2700; - break; - default: - DRM_DEBUG_KMS("Invalid clock for DP: %d\n", clock); - return NULL; - } - - pll = intel_get_shared_dpll_by_id(dev_priv, pll_id); + pll = hsw_ddi_dp_get_dpll(encoder, clock); } else if (encoder->type == INTEL_OUTPUT_ANALOG) { if (WARN_ON(crtc_state->port_clock / 2 != 135000)) @@ -751,14 +794,11 @@ hsw_get_dpll(struct intel_crtc *crtc, struct intel_crtc_state *crtc_state, if (!pll) return NULL; - crtc_state->ddi_pll_sel = hsw_pll_to_ddi_pll_sel(pll); - intel_reference_shared_dpll(pll, crtc_state); return pll; } - static const struct intel_shared_dpll_funcs hsw_ddi_wrpll_funcs = { .enable = hsw_ddi_wrpll_enable, .disable = hsw_ddi_wrpll_disable, @@ -1194,75 +1234,110 @@ skip_remaining_dividers: return true; } -static struct intel_shared_dpll * -skl_get_dpll(struct intel_crtc *crtc, struct intel_crtc_state *crtc_state, - struct intel_encoder *encoder) +static bool skl_ddi_hdmi_pll_dividers(struct intel_crtc *crtc, + struct intel_crtc_state *crtc_state, + int clock) { - struct intel_shared_dpll *pll; uint32_t ctrl1, cfgcr1, cfgcr2; - int clock = crtc_state->port_clock; + struct skl_wrpll_params wrpll_params = { 0, }; /* * See comment in intel_dpll_hw_state to understand why we always use 0 * as the DPLL id in this function. */ - ctrl1 = DPLL_CTRL1_OVERRIDE(0); - if (encoder->type == INTEL_OUTPUT_HDMI) { - struct skl_wrpll_params wrpll_params = { 0, }; + ctrl1 |= DPLL_CTRL1_HDMI_MODE(0); - ctrl1 |= DPLL_CTRL1_HDMI_MODE(0); + if (!skl_ddi_calculate_wrpll(clock * 1000, &wrpll_params)) + return false; - if (!skl_ddi_calculate_wrpll(clock * 1000, &wrpll_params)) - return NULL; + cfgcr1 = DPLL_CFGCR1_FREQ_ENABLE | + DPLL_CFGCR1_DCO_FRACTION(wrpll_params.dco_fraction) | + wrpll_params.dco_integer; + + cfgcr2 = DPLL_CFGCR2_QDIV_RATIO(wrpll_params.qdiv_ratio) | + DPLL_CFGCR2_QDIV_MODE(wrpll_params.qdiv_mode) | + DPLL_CFGCR2_KDIV(wrpll_params.kdiv) | + DPLL_CFGCR2_PDIV(wrpll_params.pdiv) | + wrpll_params.central_freq; + + memset(&crtc_state->dpll_hw_state, 0, + sizeof(crtc_state->dpll_hw_state)); + + crtc_state->dpll_hw_state.ctrl1 = ctrl1; + crtc_state->dpll_hw_state.cfgcr1 = cfgcr1; + crtc_state->dpll_hw_state.cfgcr2 = cfgcr2; + return true; +} - cfgcr1 = DPLL_CFGCR1_FREQ_ENABLE | - DPLL_CFGCR1_DCO_FRACTION(wrpll_params.dco_fraction) | - wrpll_params.dco_integer; - cfgcr2 = DPLL_CFGCR2_QDIV_RATIO(wrpll_params.qdiv_ratio) | - DPLL_CFGCR2_QDIV_MODE(wrpll_params.qdiv_mode) | - DPLL_CFGCR2_KDIV(wrpll_params.kdiv) | - DPLL_CFGCR2_PDIV(wrpll_params.pdiv) | - wrpll_params.central_freq; +bool skl_ddi_dp_set_dpll_hw_state(int clock, + struct intel_dpll_hw_state *dpll_hw_state) +{ + uint32_t ctrl1; + + /* + * See comment in intel_dpll_hw_state to understand why we always use 0 + * as the DPLL id in this function. + */ + ctrl1 = DPLL_CTRL1_OVERRIDE(0); + switch (clock / 2) { + case 81000: + ctrl1 |= DPLL_CTRL1_LINK_RATE(DPLL_CTRL1_LINK_RATE_810, 0); + break; + case 135000: + ctrl1 |= DPLL_CTRL1_LINK_RATE(DPLL_CTRL1_LINK_RATE_1350, 0); + break; + case 270000: + ctrl1 |= DPLL_CTRL1_LINK_RATE(DPLL_CTRL1_LINK_RATE_2700, 0); + break; + /* eDP 1.4 rates */ + case 162000: + ctrl1 |= DPLL_CTRL1_LINK_RATE(DPLL_CTRL1_LINK_RATE_1620, 0); + break; + case 108000: + ctrl1 |= DPLL_CTRL1_LINK_RATE(DPLL_CTRL1_LINK_RATE_1080, 0); + break; + case 216000: + ctrl1 |= DPLL_CTRL1_LINK_RATE(DPLL_CTRL1_LINK_RATE_2160, 0); + break; + } + + dpll_hw_state->ctrl1 = ctrl1; + return true; +} + +static struct intel_shared_dpll * +skl_get_dpll(struct intel_crtc *crtc, struct intel_crtc_state *crtc_state, + struct intel_encoder *encoder) +{ + struct intel_shared_dpll *pll; + int clock = crtc_state->port_clock; + bool bret; + struct intel_dpll_hw_state dpll_hw_state; + + memset(&dpll_hw_state, 0, sizeof(dpll_hw_state)); + + if (encoder->type == INTEL_OUTPUT_HDMI) { + bret = skl_ddi_hdmi_pll_dividers(crtc, crtc_state, clock); + if (!bret) { + DRM_DEBUG_KMS("Could not get HDMI pll dividers.\n"); + return NULL; + } } else if (encoder->type == INTEL_OUTPUT_DP || encoder->type == INTEL_OUTPUT_DP_MST || encoder->type == INTEL_OUTPUT_EDP) { - switch (crtc_state->port_clock / 2) { - case 81000: - ctrl1 |= DPLL_CTRL1_LINK_RATE(DPLL_CTRL1_LINK_RATE_810, 0); - break; - case 135000: - ctrl1 |= DPLL_CTRL1_LINK_RATE(DPLL_CTRL1_LINK_RATE_1350, 0); - break; - case 270000: - ctrl1 |= DPLL_CTRL1_LINK_RATE(DPLL_CTRL1_LINK_RATE_2700, 0); - break; - /* eDP 1.4 rates */ - case 162000: - ctrl1 |= DPLL_CTRL1_LINK_RATE(DPLL_CTRL1_LINK_RATE_1620, 0); - break; - case 108000: - ctrl1 |= DPLL_CTRL1_LINK_RATE(DPLL_CTRL1_LINK_RATE_1080, 0); - break; - case 216000: - ctrl1 |= DPLL_CTRL1_LINK_RATE(DPLL_CTRL1_LINK_RATE_2160, 0); - break; + bret = skl_ddi_dp_set_dpll_hw_state(clock, &dpll_hw_state); + if (!bret) { + DRM_DEBUG_KMS("Could not set DP dpll HW state.\n"); + return NULL; } - - cfgcr1 = cfgcr2 = 0; + crtc_state->dpll_hw_state = dpll_hw_state; } else { return NULL; } - memset(&crtc_state->dpll_hw_state, 0, - sizeof(crtc_state->dpll_hw_state)); - - crtc_state->dpll_hw_state.ctrl1 = ctrl1; - crtc_state->dpll_hw_state.cfgcr1 = cfgcr1; - crtc_state->dpll_hw_state.cfgcr2 = cfgcr2; - if (encoder->type == INTEL_OUTPUT_EDP) pll = intel_find_shared_dpll(crtc, crtc_state, DPLL_ID_SKL_DPLL0, @@ -1274,8 +1349,6 @@ skl_get_dpll(struct intel_crtc *crtc, struct intel_crtc_state *crtc_state, if (!pll) return NULL; - crtc_state->ddi_pll_sel = pll->id; - intel_reference_shared_dpll(pll, crtc_state); return pll; @@ -1484,6 +1557,8 @@ struct bxt_clk_div { uint32_t m2_frac; bool m2_frac_en; uint32_t n; + + int vco; }; /* pre-calculated values for DP linkrates */ @@ -1497,57 +1572,60 @@ static const struct bxt_clk_div bxt_dp_clk_val[] = { {432000, 3, 1, 32, 1677722, 1, 1} }; -static struct intel_shared_dpll * -bxt_get_dpll(struct intel_crtc *crtc, struct intel_crtc_state *crtc_state, - struct intel_encoder *encoder) +static bool +bxt_ddi_hdmi_pll_dividers(struct intel_crtc *intel_crtc, + struct intel_crtc_state *crtc_state, int clock, + struct bxt_clk_div *clk_div) { - struct drm_i915_private *dev_priv = to_i915(crtc->base.dev); - struct intel_shared_dpll *pll; - enum intel_dpll_id i; - struct intel_digital_port *intel_dig_port; - struct bxt_clk_div clk_div = {0}; - int vco = 0; - uint32_t prop_coef, int_coef, gain_ctl, targ_cnt; - uint32_t lanestagger; - int clock = crtc_state->port_clock; + struct dpll best_clock; - if (encoder->type == INTEL_OUTPUT_HDMI) { - struct dpll best_clock; + /* Calculate HDMI div */ + /* + * FIXME: tie the following calculation into + * i9xx_crtc_compute_clock + */ + if (!bxt_find_best_dpll(crtc_state, clock, &best_clock)) { + DRM_DEBUG_DRIVER("no PLL dividers found for clock %d pipe %c\n", + clock, pipe_name(intel_crtc->pipe)); + return false; + } - /* Calculate HDMI div */ - /* - * FIXME: tie the following calculation into - * i9xx_crtc_compute_clock - */ - if (!bxt_find_best_dpll(crtc_state, clock, &best_clock)) { - DRM_DEBUG_DRIVER("no PLL dividers found for clock %d pipe %c\n", - clock, pipe_name(crtc->pipe)); - return NULL; - } + clk_div->p1 = best_clock.p1; + clk_div->p2 = best_clock.p2; + WARN_ON(best_clock.m1 != 2); + clk_div->n = best_clock.n; + clk_div->m2_int = best_clock.m2 >> 22; + clk_div->m2_frac = best_clock.m2 & ((1 << 22) - 1); + clk_div->m2_frac_en = clk_div->m2_frac != 0; - clk_div.p1 = best_clock.p1; - clk_div.p2 = best_clock.p2; - WARN_ON(best_clock.m1 != 2); - clk_div.n = best_clock.n; - clk_div.m2_int = best_clock.m2 >> 22; - clk_div.m2_frac = best_clock.m2 & ((1 << 22) - 1); - clk_div.m2_frac_en = clk_div.m2_frac != 0; + clk_div->vco = best_clock.vco; - vco = best_clock.vco; - } else if (encoder->type == INTEL_OUTPUT_DP || - encoder->type == INTEL_OUTPUT_EDP) { - int i; + return true; +} - clk_div = bxt_dp_clk_val[0]; - for (i = 0; i < ARRAY_SIZE(bxt_dp_clk_val); ++i) { - if (bxt_dp_clk_val[i].clock == clock) { - clk_div = bxt_dp_clk_val[i]; - break; - } +static void bxt_ddi_dp_pll_dividers(int clock, struct bxt_clk_div *clk_div) +{ + int i; + + *clk_div = bxt_dp_clk_val[0]; + for (i = 0; i < ARRAY_SIZE(bxt_dp_clk_val); ++i) { + if (bxt_dp_clk_val[i].clock == clock) { + *clk_div = bxt_dp_clk_val[i]; + break; } - vco = clock * 10 / 2 * clk_div.p1 * clk_div.p2; } + clk_div->vco = clock * 10 / 2 * clk_div->p1 * clk_div->p2; +} + +static bool bxt_ddi_set_dpll_hw_state(int clock, + struct bxt_clk_div *clk_div, + struct intel_dpll_hw_state *dpll_hw_state) +{ + int vco = clk_div->vco; + uint32_t prop_coef, int_coef, gain_ctl, targ_cnt; + uint32_t lanestagger; + if (vco >= 6200000 && vco <= 6700000) { prop_coef = 4; int_coef = 9; @@ -1566,12 +1644,9 @@ bxt_get_dpll(struct intel_crtc *crtc, struct intel_crtc_state *crtc_state, targ_cnt = 9; } else { DRM_ERROR("Invalid VCO\n"); - return NULL; + return false; } - memset(&crtc_state->dpll_hw_state, 0, - sizeof(crtc_state->dpll_hw_state)); - if (clock > 270000) lanestagger = 0x18; else if (clock > 135000) @@ -1583,35 +1658,75 @@ bxt_get_dpll(struct intel_crtc *crtc, struct intel_crtc_state *crtc_state, else lanestagger = 0x02; - crtc_state->dpll_hw_state.ebb0 = - PORT_PLL_P1(clk_div.p1) | PORT_PLL_P2(clk_div.p2); - crtc_state->dpll_hw_state.pll0 = clk_div.m2_int; - crtc_state->dpll_hw_state.pll1 = PORT_PLL_N(clk_div.n); - crtc_state->dpll_hw_state.pll2 = clk_div.m2_frac; + dpll_hw_state->ebb0 = PORT_PLL_P1(clk_div->p1) | PORT_PLL_P2(clk_div->p2); + dpll_hw_state->pll0 = clk_div->m2_int; + dpll_hw_state->pll1 = PORT_PLL_N(clk_div->n); + dpll_hw_state->pll2 = clk_div->m2_frac; - if (clk_div.m2_frac_en) - crtc_state->dpll_hw_state.pll3 = - PORT_PLL_M2_FRAC_ENABLE; + if (clk_div->m2_frac_en) + dpll_hw_state->pll3 = PORT_PLL_M2_FRAC_ENABLE; - crtc_state->dpll_hw_state.pll6 = - prop_coef | PORT_PLL_INT_COEFF(int_coef); - crtc_state->dpll_hw_state.pll6 |= - PORT_PLL_GAIN_CTL(gain_ctl); + dpll_hw_state->pll6 = prop_coef | PORT_PLL_INT_COEFF(int_coef); + dpll_hw_state->pll6 |= PORT_PLL_GAIN_CTL(gain_ctl); - crtc_state->dpll_hw_state.pll8 = targ_cnt; + dpll_hw_state->pll8 = targ_cnt; - crtc_state->dpll_hw_state.pll9 = 5 << PORT_PLL_LOCK_THRESHOLD_SHIFT; + dpll_hw_state->pll9 = 5 << PORT_PLL_LOCK_THRESHOLD_SHIFT; - crtc_state->dpll_hw_state.pll10 = + dpll_hw_state->pll10 = PORT_PLL_DCO_AMP(PORT_PLL_DCO_AMP_DEFAULT) | PORT_PLL_DCO_AMP_OVR_EN_H; - crtc_state->dpll_hw_state.ebb4 = PORT_PLL_10BIT_CLK_ENABLE; + dpll_hw_state->ebb4 = PORT_PLL_10BIT_CLK_ENABLE; - crtc_state->dpll_hw_state.pcsdw12 = - LANESTAGGER_STRAP_OVRD | lanestagger; + dpll_hw_state->pcsdw12 = LANESTAGGER_STRAP_OVRD | lanestagger; - intel_dig_port = enc_to_dig_port(&encoder->base); + return true; +} + +bool bxt_ddi_dp_set_dpll_hw_state(int clock, + struct intel_dpll_hw_state *dpll_hw_state) +{ + struct bxt_clk_div clk_div = {0}; + + bxt_ddi_dp_pll_dividers(clock, &clk_div); + + return bxt_ddi_set_dpll_hw_state(clock, &clk_div, dpll_hw_state); +} + +static struct intel_shared_dpll * +bxt_get_dpll(struct intel_crtc *crtc, + struct intel_crtc_state *crtc_state, + struct intel_encoder *encoder) +{ + struct bxt_clk_div clk_div = {0}; + struct intel_dpll_hw_state dpll_hw_state = {0}; + struct drm_i915_private *dev_priv = to_i915(crtc->base.dev); + struct intel_digital_port *intel_dig_port; + struct intel_shared_dpll *pll; + int i, clock = crtc_state->port_clock; + + if (encoder->type == INTEL_OUTPUT_HDMI + && !bxt_ddi_hdmi_pll_dividers(crtc, crtc_state, + clock, &clk_div)) + return NULL; + + if ((encoder->type == INTEL_OUTPUT_DP || + encoder->type == INTEL_OUTPUT_EDP) && + !bxt_ddi_dp_set_dpll_hw_state(clock, &dpll_hw_state)) + return NULL; + + memset(&crtc_state->dpll_hw_state, 0, + sizeof(crtc_state->dpll_hw_state)); + + crtc_state->dpll_hw_state = dpll_hw_state; + + if (encoder->type == INTEL_OUTPUT_DP_MST) { + struct intel_dp_mst_encoder *intel_mst = enc_to_mst(&encoder->base); + + intel_dig_port = intel_mst->primary; + } else + intel_dig_port = enc_to_dig_port(&encoder->base); /* 1:1 mapping between ports and PLLs */ i = (enum intel_dpll_id) intel_dig_port->port; @@ -1622,9 +1737,6 @@ bxt_get_dpll(struct intel_crtc *crtc, struct intel_crtc_state *crtc_state, intel_reference_shared_dpll(pll, crtc_state); - /* shared DPLL id 0 is DPLL A */ - crtc_state->ddi_pll_sel = pll->id; - return pll; } diff --git a/drivers/gpu/drm/i915/intel_dpll_mgr.h b/drivers/gpu/drm/i915/intel_dpll_mgr.h index 89c5ada1a315..f4385353bc11 100644 --- a/drivers/gpu/drm/i915/intel_dpll_mgr.h +++ b/drivers/gpu/drm/i915/intel_dpll_mgr.h @@ -160,5 +160,20 @@ void intel_disable_shared_dpll(struct intel_crtc *crtc); void intel_shared_dpll_commit(struct drm_atomic_state *state); void intel_shared_dpll_init(struct drm_device *dev); +/* BXT dpll related functions */ +bool bxt_ddi_dp_set_dpll_hw_state(int clock, + struct intel_dpll_hw_state *dpll_hw_state); + + +/* SKL dpll related functions */ +bool skl_ddi_dp_set_dpll_hw_state(int clock, + struct intel_dpll_hw_state *dpll_hw_state); +struct intel_shared_dpll *skl_find_link_pll(struct drm_i915_private *dev_priv, + int clock); + + +/* HSW dpll related functions */ +struct intel_shared_dpll *hsw_ddi_dp_get_dpll(struct intel_encoder *encoder, + int clock); #endif /* _INTEL_DPLL_MGR_H_ */ diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h index 774aab342f40..8fd16adf069b 100644 --- a/drivers/gpu/drm/i915/intel_drv.h +++ b/drivers/gpu/drm/i915/intel_drv.h @@ -52,11 +52,15 @@ */ #define _wait_for(COND, US, W) ({ \ unsigned long timeout__ = jiffies + usecs_to_jiffies(US) + 1; \ - int ret__ = 0; \ - while (!(COND)) { \ - if (time_after(jiffies, timeout__)) { \ - if (!(COND)) \ - ret__ = -ETIMEDOUT; \ + int ret__; \ + for (;;) { \ + bool expired__ = time_after(jiffies, timeout__); \ + if (COND) { \ + ret__ = 0; \ + break; \ + } \ + if (expired__) { \ + ret__ = -ETIMEDOUT; \ break; \ } \ if ((W) && drm_can_sleep()) { \ @@ -205,14 +209,26 @@ struct intel_encoder { unsigned int cloneable; void (*hot_plug)(struct intel_encoder *); bool (*compute_config)(struct intel_encoder *, - struct intel_crtc_state *); - void (*pre_pll_enable)(struct intel_encoder *); - void (*pre_enable)(struct intel_encoder *); - void (*enable)(struct intel_encoder *); - void (*mode_set)(struct intel_encoder *intel_encoder); - void (*disable)(struct intel_encoder *); - void (*post_disable)(struct intel_encoder *); - void (*post_pll_disable)(struct intel_encoder *); + struct intel_crtc_state *, + struct drm_connector_state *); + void (*pre_pll_enable)(struct intel_encoder *, + struct intel_crtc_state *, + struct drm_connector_state *); + void (*pre_enable)(struct intel_encoder *, + struct intel_crtc_state *, + struct drm_connector_state *); + void (*enable)(struct intel_encoder *, + struct intel_crtc_state *, + struct drm_connector_state *); + void (*disable)(struct intel_encoder *, + struct intel_crtc_state *, + struct drm_connector_state *); + void (*post_disable)(struct intel_encoder *, + struct intel_crtc_state *, + struct drm_connector_state *); + void (*post_pll_disable)(struct intel_encoder *, + struct intel_crtc_state *, + struct drm_connector_state *); /* Read out the current hw state of this connector, returning true if * the encoder is active. If the encoder is enabled it also set the pipe * it is connected to in the pipe parameter. */ @@ -578,12 +594,6 @@ struct intel_crtc_state { /* Selected dpll when shared or NULL. */ struct intel_shared_dpll *shared_dpll; - /* - * - PORT_CLK_SEL for DDI ports on HSW/BDW. - * - enum skl_dpll on SKL - */ - uint32_t ddi_pll_sel; - /* Actual register state of the dpll, for shared dpll cross-checking. */ struct intel_dpll_hw_state dpll_hw_state; @@ -700,8 +710,8 @@ struct intel_crtc { struct intel_crtc_state *config; - /* reset counter value when the last flip was submitted */ - unsigned int reset_counter; + /* global reset count when the last flip was submitted */ + unsigned int reset_count; /* Access to these should be protected by dev_priv->irq_lock. */ bool cpu_fifo_underrun_disabled; @@ -872,6 +882,7 @@ struct intel_dp { bool link_mst; bool has_audio; bool detect_done; + bool channel_eq_status; enum hdmi_force_audio force_audio; bool limited_color_range; bool color_range_auto; @@ -1124,7 +1135,10 @@ void intel_crt_reset(struct drm_encoder *encoder); /* intel_ddi.c */ void intel_ddi_clk_select(struct intel_encoder *encoder, - const struct intel_crtc_state *pipe_config); + struct intel_shared_dpll *pll); +void intel_ddi_fdi_post_disable(struct intel_encoder *intel_encoder, + struct intel_crtc_state *old_crtc_state, + struct drm_connector_state *old_conn_state); void intel_prepare_dp_ddi_buffers(struct intel_encoder *encoder); void hsw_fdi_link_train(struct drm_crtc *crtc); void intel_ddi_init(struct drm_device *dev, enum port port); @@ -1140,7 +1154,6 @@ bool intel_ddi_pll_select(struct intel_crtc *crtc, void intel_ddi_set_pipe_settings(struct drm_crtc *crtc); void intel_ddi_prepare_link_retrain(struct intel_dp *intel_dp); bool intel_ddi_connector_get_hw_state(struct intel_connector *intel_connector); -void intel_ddi_fdi_disable(struct drm_crtc *crtc); void intel_ddi_get_config(struct intel_encoder *encoder, struct intel_crtc_state *pipe_config); struct intel_encoder * @@ -1151,7 +1164,8 @@ void intel_ddi_clock_get(struct intel_encoder *encoder, struct intel_crtc_state *pipe_config); void intel_ddi_set_vc_payload_alloc(struct drm_crtc *crtc, bool state); uint32_t ddi_signal_levels(struct intel_dp *intel_dp); - +struct intel_shared_dpll *intel_ddi_get_link_dpll(struct intel_dp *intel_dp, + int clock); unsigned int intel_fb_align_height(struct drm_device *dev, unsigned int height, uint32_t pixel_format, @@ -1171,6 +1185,8 @@ void skl_set_preferred_cdclk_vco(struct drm_i915_private *dev_priv, int vco); void intel_update_rawclk(struct drm_i915_private *dev_priv); int vlv_get_cck_clock(struct drm_i915_private *dev_priv, const char *name, u32 reg, int ref_freq); +void lpt_disable_pch_transcoder(struct drm_i915_private *dev_priv); +void lpt_disable_iclkip(struct drm_i915_private *dev_priv); extern const struct drm_plane_funcs intel_plane_funcs; void intel_init_display_hooks(struct drm_i915_private *dev_priv); unsigned int intel_fb_xy_to_linear(int x, int y, @@ -1367,7 +1383,8 @@ bool intel_dp_init(struct drm_device *dev, i915_reg_t output_reg, enum port port bool intel_dp_init_connector(struct intel_digital_port *intel_dig_port, struct intel_connector *intel_connector); void intel_dp_set_link_params(struct intel_dp *intel_dp, - const struct intel_crtc_state *pipe_config); + int link_rate, uint8_t lane_count, + bool link_mst); void intel_dp_start_link_train(struct intel_dp *intel_dp); void intel_dp_stop_link_train(struct intel_dp *intel_dp); void intel_dp_sink_dpms(struct intel_dp *intel_dp, int mode); @@ -1376,7 +1393,8 @@ void intel_dp_encoder_suspend(struct intel_encoder *intel_encoder); void intel_dp_encoder_destroy(struct drm_encoder *encoder); int intel_dp_sink_crc(struct intel_dp *intel_dp, u8 *crc); bool intel_dp_compute_config(struct intel_encoder *encoder, - struct intel_crtc_state *pipe_config); + struct intel_crtc_state *pipe_config, + struct drm_connector_state *conn_state); bool intel_dp_is_edp(struct drm_device *dev, enum port port); enum irqreturn intel_dp_hpd_pulse(struct intel_digital_port *intel_dig_port, bool long_hpd); @@ -1394,14 +1412,14 @@ void intel_dp_hot_plug(struct intel_encoder *intel_encoder); void intel_power_sequencer_reset(struct drm_i915_private *dev_priv); uint32_t intel_dp_pack_aux(const uint8_t *src, int src_bytes); void intel_plane_destroy(struct drm_plane *plane); -void intel_edp_drrs_enable(struct intel_dp *intel_dp); -void intel_edp_drrs_disable(struct intel_dp *intel_dp); +void intel_edp_drrs_enable(struct intel_dp *intel_dp, + struct intel_crtc_state *crtc_state); +void intel_edp_drrs_disable(struct intel_dp *intel_dp, + struct intel_crtc_state *crtc_state); void intel_edp_drrs_invalidate(struct drm_i915_private *dev_priv, unsigned int frontbuffer_bits); void intel_edp_drrs_flush(struct drm_i915_private *dev_priv, unsigned int frontbuffer_bits); -bool intel_digital_port_connected(struct drm_i915_private *dev_priv, - struct intel_digital_port *port); void intel_dp_program_link_training_pattern(struct intel_dp *intel_dp, @@ -1501,7 +1519,8 @@ void intel_hdmi_init_connector(struct intel_digital_port *intel_dig_port, struct intel_connector *intel_connector); struct intel_hdmi *enc_to_intel_hdmi(struct drm_encoder *encoder); bool intel_hdmi_compute_config(struct intel_encoder *encoder, - struct intel_crtc_state *pipe_config); + struct intel_crtc_state *pipe_config, + struct drm_connector_state *conn_state); void intel_dp_dual_mode_set_tmds_output(struct intel_hdmi *hdmi, bool enable); @@ -1722,6 +1741,21 @@ void ilk_wm_get_hw_state(struct drm_device *dev); void skl_wm_get_hw_state(struct drm_device *dev); void skl_ddb_get_hw_state(struct drm_i915_private *dev_priv, struct skl_ddb_allocation *ddb /* out */); +bool skl_can_enable_sagv(struct drm_atomic_state *state); +int skl_enable_sagv(struct drm_i915_private *dev_priv); +int skl_disable_sagv(struct drm_i915_private *dev_priv); +bool skl_ddb_allocation_equals(const struct skl_ddb_allocation *old, + const struct skl_ddb_allocation *new, + enum pipe pipe); +bool skl_ddb_allocation_overlaps(struct drm_atomic_state *state, + const struct skl_ddb_allocation *old, + const struct skl_ddb_allocation *new, + enum pipe pipe); +void skl_write_cursor_wm(struct intel_crtc *intel_crtc, + const struct skl_wm_values *wm); +void skl_write_plane_wm(struct intel_crtc *intel_crtc, + const struct skl_wm_values *wm, + int plane); uint32_t ilk_pipe_pixel_rate(const struct intel_crtc_state *pipe_config); bool ilk_disable_lp_wm(struct drm_device *dev); int sanitize_rc6_option(struct drm_i915_private *dev_priv, int enable_rc6); diff --git a/drivers/gpu/drm/i915/intel_dsi.c b/drivers/gpu/drm/i915/intel_dsi.c index de8e9fb51595..b2e3d3a334f7 100644 --- a/drivers/gpu/drm/i915/intel_dsi.c +++ b/drivers/gpu/drm/i915/intel_dsi.c @@ -312,7 +312,8 @@ static inline bool is_cmd_mode(struct intel_dsi *intel_dsi) } static bool intel_dsi_compute_config(struct intel_encoder *encoder, - struct intel_crtc_state *pipe_config) + struct intel_crtc_state *pipe_config, + struct drm_connector_state *conn_state) { struct drm_i915_private *dev_priv = to_i915(encoder->base.dev); struct intel_dsi *intel_dsi = container_of(encoder, struct intel_dsi, @@ -533,14 +534,15 @@ static void intel_dsi_enable(struct intel_encoder *encoder) intel_panel_enable_backlight(intel_dsi->attached_connector); } -static void intel_dsi_prepare(struct intel_encoder *intel_encoder); +static void intel_dsi_prepare(struct intel_encoder *intel_encoder, + struct intel_crtc_state *pipe_config); -static void intel_dsi_pre_enable(struct intel_encoder *encoder) +static void intel_dsi_pre_enable(struct intel_encoder *encoder, + struct intel_crtc_state *pipe_config, + struct drm_connector_state *conn_state) { - struct drm_device *dev = encoder->base.dev; - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = to_i915(encoder->base.dev); struct intel_dsi *intel_dsi = enc_to_intel_dsi(&encoder->base); - struct intel_crtc *crtc = to_intel_crtc(encoder->base.crtc); enum port port; DRM_DEBUG_KMS("\n"); @@ -550,9 +552,9 @@ static void intel_dsi_pre_enable(struct intel_encoder *encoder) * lock. It needs to be fully powered down to fix it. */ intel_disable_dsi_pll(encoder); - intel_enable_dsi_pll(encoder, crtc->config); + intel_enable_dsi_pll(encoder, pipe_config); - intel_dsi_prepare(encoder); + intel_dsi_prepare(encoder, pipe_config); /* Panel Enable over CRC PMIC */ if (intel_dsi->gpio_panel) @@ -582,7 +584,9 @@ static void intel_dsi_pre_enable(struct intel_encoder *encoder) intel_dsi_enable(encoder); } -static void intel_dsi_enable_nop(struct intel_encoder *encoder) +static void intel_dsi_enable_nop(struct intel_encoder *encoder, + struct intel_crtc_state *pipe_config, + struct drm_connector_state *conn_state) { DRM_DEBUG_KMS("\n"); @@ -592,7 +596,9 @@ static void intel_dsi_enable_nop(struct intel_encoder *encoder) */ } -static void intel_dsi_pre_disable(struct intel_encoder *encoder) +static void intel_dsi_pre_disable(struct intel_encoder *encoder, + struct intel_crtc_state *old_crtc_state, + struct drm_connector_state *old_conn_state) { struct intel_dsi *intel_dsi = enc_to_intel_dsi(&encoder->base); enum port port; @@ -694,7 +700,9 @@ static void intel_dsi_clear_device_ready(struct intel_encoder *encoder) intel_disable_dsi_pll(encoder); } -static void intel_dsi_post_disable(struct intel_encoder *encoder) +static void intel_dsi_post_disable(struct intel_encoder *encoder, + struct intel_crtc_state *pipe_config, + struct drm_connector_state *conn_state) { struct drm_i915_private *dev_priv = to_i915(encoder->base.dev); struct intel_dsi *intel_dsi = enc_to_intel_dsi(&encoder->base); @@ -819,6 +827,7 @@ static void bxt_dsi_get_pipe_config(struct intel_encoder *encoder, u16 crtc_htotal_sw, crtc_hsync_start_sw, crtc_hsync_end_sw, crtc_hblank_start_sw, crtc_hblank_end_sw; + /* FIXME: hw readout should not depend on SW state */ intel_crtc = to_intel_crtc(encoder->base.crtc); adjusted_mode_sw = &intel_crtc->config->base.adjusted_mode; @@ -1104,14 +1113,15 @@ static u32 pixel_format_to_reg(enum mipi_dsi_pixel_format fmt) } } -static void intel_dsi_prepare(struct intel_encoder *intel_encoder) +static void intel_dsi_prepare(struct intel_encoder *intel_encoder, + struct intel_crtc_state *pipe_config) { struct drm_encoder *encoder = &intel_encoder->base; struct drm_device *dev = encoder->dev; struct drm_i915_private *dev_priv = to_i915(dev); - struct intel_crtc *intel_crtc = to_intel_crtc(encoder->crtc); + struct intel_crtc *intel_crtc = to_intel_crtc(pipe_config->base.crtc); struct intel_dsi *intel_dsi = enc_to_intel_dsi(encoder); - const struct drm_display_mode *adjusted_mode = &intel_crtc->config->base.adjusted_mode; + const struct drm_display_mode *adjusted_mode = &pipe_config->base.adjusted_mode; enum port port; unsigned int bpp = mipi_dsi_pixel_format_to_bpp(intel_dsi->pixel_format); u32 val, tmp; @@ -1348,7 +1358,7 @@ static int intel_dsi_set_property(struct drm_connector *connector, intel_connector->panel.fitting_mode = val; } - crtc = intel_attached_encoder(connector)->base.crtc; + crtc = connector->state->crtc; if (crtc && crtc->state->enable) { /* * If the CRTC is enabled, the display will be changed diff --git a/drivers/gpu/drm/i915/intel_dvo.c b/drivers/gpu/drm/i915/intel_dvo.c index 47bdf9dad0d3..2e452c505e7e 100644 --- a/drivers/gpu/drm/i915/intel_dvo.c +++ b/drivers/gpu/drm/i915/intel_dvo.c @@ -174,7 +174,9 @@ static void intel_dvo_get_config(struct intel_encoder *encoder, pipe_config->base.adjusted_mode.crtc_clock = pipe_config->port_clock; } -static void intel_disable_dvo(struct intel_encoder *encoder) +static void intel_disable_dvo(struct intel_encoder *encoder, + struct intel_crtc_state *old_crtc_state, + struct drm_connector_state *old_conn_state) { struct drm_i915_private *dev_priv = to_i915(encoder->base.dev); struct intel_dvo *intel_dvo = enc_to_dvo(encoder); @@ -186,17 +188,18 @@ static void intel_disable_dvo(struct intel_encoder *encoder) I915_READ(dvo_reg); } -static void intel_enable_dvo(struct intel_encoder *encoder) +static void intel_enable_dvo(struct intel_encoder *encoder, + struct intel_crtc_state *pipe_config, + struct drm_connector_state *conn_state) { struct drm_i915_private *dev_priv = to_i915(encoder->base.dev); struct intel_dvo *intel_dvo = enc_to_dvo(encoder); - struct intel_crtc *crtc = to_intel_crtc(encoder->base.crtc); i915_reg_t dvo_reg = intel_dvo->dev.dvo_reg; u32 temp = I915_READ(dvo_reg); intel_dvo->dev.dev_ops->mode_set(&intel_dvo->dev, - &crtc->config->base.mode, - &crtc->config->base.adjusted_mode); + &pipe_config->base.mode, + &pipe_config->base.adjusted_mode); I915_WRITE(dvo_reg, temp | DVO_ENABLE); I915_READ(dvo_reg); @@ -235,7 +238,8 @@ intel_dvo_mode_valid(struct drm_connector *connector, } static bool intel_dvo_compute_config(struct intel_encoder *encoder, - struct intel_crtc_state *pipe_config) + struct intel_crtc_state *pipe_config, + struct drm_connector_state *conn_state) { struct intel_dvo *intel_dvo = enc_to_dvo(encoder); const struct drm_display_mode *fixed_mode = @@ -253,12 +257,13 @@ static bool intel_dvo_compute_config(struct intel_encoder *encoder, return true; } -static void intel_dvo_pre_enable(struct intel_encoder *encoder) +static void intel_dvo_pre_enable(struct intel_encoder *encoder, + struct intel_crtc_state *pipe_config, + struct drm_connector_state *conn_state) { - struct drm_device *dev = encoder->base.dev; - struct drm_i915_private *dev_priv = to_i915(dev); - struct intel_crtc *crtc = to_intel_crtc(encoder->base.crtc); - const struct drm_display_mode *adjusted_mode = &crtc->config->base.adjusted_mode; + struct drm_i915_private *dev_priv = to_i915(encoder->base.dev); + struct intel_crtc *crtc = to_intel_crtc(pipe_config->base.crtc); + const struct drm_display_mode *adjusted_mode = &pipe_config->base.adjusted_mode; struct intel_dvo *intel_dvo = enc_to_dvo(encoder); int pipe = crtc->pipe; u32 dvo_val; @@ -554,7 +559,6 @@ void intel_dvo_init(struct drm_device *dev) return; } - drm_encoder_cleanup(&intel_encoder->base); kfree(intel_dvo); kfree(intel_connector); } diff --git a/drivers/gpu/drm/i915/intel_engine_cs.c b/drivers/gpu/drm/i915/intel_engine_cs.c index 2e96a86105c2..e405f1080296 100644 --- a/drivers/gpu/drm/i915/intel_engine_cs.c +++ b/drivers/gpu/drm/i915/intel_engine_cs.c @@ -211,6 +211,8 @@ void intel_engine_init_hangcheck(struct intel_engine_cs *engine) { memset(&engine->hangcheck, 0, sizeof(engine->hangcheck)); clear_bit(engine->id, &engine->i915->gpu_error.missed_irq_rings); + if (intel_engine_has_waiter(engine)) + i915_queue_hangcheck(engine->i915); } static void intel_engine_init_requests(struct intel_engine_cs *engine) @@ -230,7 +232,6 @@ static void intel_engine_init_requests(struct intel_engine_cs *engine) */ void intel_engine_setup_common(struct intel_engine_cs *engine) { - INIT_LIST_HEAD(&engine->buffers); INIT_LIST_HEAD(&engine->execlist_queue); spin_lock_init(&engine->execlist_lock); @@ -306,6 +307,18 @@ int intel_engine_init_common(struct intel_engine_cs *engine) return 0; } +void intel_engine_reset_irq(struct intel_engine_cs *engine) +{ + struct drm_i915_private *dev_priv = engine->i915; + + spin_lock_irq(&dev_priv->irq_lock); + if (intel_engine_has_waiter(engine)) + engine->irq_enable(engine); + else + engine->irq_disable(engine); + spin_unlock_irq(&dev_priv->irq_lock); +} + /** * intel_engines_cleanup_common - cleans up the engine state created by * the common initiailizers. diff --git a/drivers/gpu/drm/i915/intel_fbc.c b/drivers/gpu/drm/i915/intel_fbc.c index bf8b22ad9aed..faa67624e1ed 100644 --- a/drivers/gpu/drm/i915/intel_fbc.c +++ b/drivers/gpu/drm/i915/intel_fbc.c @@ -799,10 +799,8 @@ static bool intel_fbc_can_activate(struct intel_crtc *crtc) */ if (cache->fb.tiling_mode != I915_TILING_X || cache->fb.fence_reg == I915_FENCE_REG_NONE) { - if (INTEL_GEN(dev_priv) < 5) { - fbc->no_fbc_reason = "framebuffer not tiled or fenced"; - return false; - } + fbc->no_fbc_reason = "framebuffer not tiled or fenced"; + return false; } if (INTEL_INFO(dev_priv)->gen <= 4 && !IS_G4X(dev_priv) && cache->plane.rotation != DRM_ROTATE_0) { diff --git a/drivers/gpu/drm/i915/intel_fbdev.c b/drivers/gpu/drm/i915/intel_fbdev.c index 4003e4908c09..b7098f98bb67 100644 --- a/drivers/gpu/drm/i915/intel_fbdev.c +++ b/drivers/gpu/drm/i915/intel_fbdev.c @@ -183,6 +183,7 @@ static int intelfb_create(struct drm_fb_helper *helper, struct intel_framebuffer *intel_fb = ifbdev->fb; struct drm_device *dev = helper->dev; struct drm_i915_private *dev_priv = to_i915(dev); + struct pci_dev *pdev = dev_priv->drm.pdev; struct i915_ggtt *ggtt = &dev_priv->ggtt; struct fb_info *info; struct drm_framebuffer *fb; @@ -280,7 +281,7 @@ static int intelfb_create(struct drm_fb_helper *helper, ifbdev->vma = vma; mutex_unlock(&dev->struct_mutex); - vga_switcheroo_client_fb_set(dev->pdev, info); + vga_switcheroo_client_fb_set(pdev, info); return 0; out_destroy_fbi: diff --git a/drivers/gpu/drm/i915/intel_guc.h b/drivers/gpu/drm/i915/intel_guc.h index c97326269588..b1ba86958811 100644 --- a/drivers/gpu/drm/i915/intel_guc.h +++ b/drivers/gpu/drm/i915/intel_guc.h @@ -78,9 +78,11 @@ struct i915_guc_client { uint16_t doorbell_id; uint16_t padding[3]; /* Maintain alignment */ + spinlock_t wq_lock; uint32_t wq_offset; uint32_t wq_size; uint32_t wq_tail; + uint32_t wq_rsvd; uint32_t no_wq_space; uint32_t b_fail; int retcode; @@ -157,7 +159,7 @@ extern int intel_guc_resume(struct drm_device *dev); /* i915_guc_submission.c */ int i915_guc_submission_init(struct drm_i915_private *dev_priv); int i915_guc_submission_enable(struct drm_i915_private *dev_priv); -int i915_guc_wq_check_space(struct drm_i915_gem_request *rq); +int i915_guc_wq_reserve(struct drm_i915_gem_request *rq); void i915_guc_submission_disable(struct drm_i915_private *dev_priv); void i915_guc_submission_fini(struct drm_i915_private *dev_priv); diff --git a/drivers/gpu/drm/i915/intel_guc_loader.c b/drivers/gpu/drm/i915/intel_guc_loader.c index 324812d69b70..6fd39efb7894 100644 --- a/drivers/gpu/drm/i915/intel_guc_loader.c +++ b/drivers/gpu/drm/i915/intel_guc_loader.c @@ -97,7 +97,7 @@ const char *intel_guc_fw_status_repr(enum intel_guc_fw_status status) } }; -static void direct_interrupts_to_host(struct drm_i915_private *dev_priv) +static void guc_interrupts_release(struct drm_i915_private *dev_priv) { struct intel_engine_cs *engine; int irqs; @@ -114,7 +114,7 @@ static void direct_interrupts_to_host(struct drm_i915_private *dev_priv) I915_WRITE(GUC_WD_VECS_IER, 0); } -static void direct_interrupts_to_guc(struct drm_i915_private *dev_priv) +static void guc_interrupts_capture(struct drm_i915_private *dev_priv) { struct intel_engine_cs *engine; int irqs; @@ -134,13 +134,28 @@ static void direct_interrupts_to_guc(struct drm_i915_private *dev_priv) I915_WRITE(GUC_WD_VECS_IER, ~irqs); /* - * If GuC has routed PM interrupts to itself, don't keep it. - * and keep other interrupts those are unmasked by GuC. - */ + * The REDIRECT_TO_GUC bit of the PMINTRMSK register directs all + * (unmasked) PM interrupts to the GuC. All other bits of this + * register *disable* generation of a specific interrupt. + * + * 'pm_intr_keep' indicates bits that are NOT to be set when + * writing to the PM interrupt mask register, i.e. interrupts + * that must not be disabled. + * + * If the GuC is handling these interrupts, then we must not let + * the PM code disable ANY interrupt that the GuC is expecting. + * So for each ENABLED (0) bit in this register, we must SET the + * bit in pm_intr_keep so that it's left enabled for the GuC. + * + * OTOH the REDIRECT_TO_GUC bit is initially SET in pm_intr_keep + * (so interrupts go to the DISPLAY unit at first); but here we + * need to CLEAR that bit, which will result in the register bit + * being left SET! + */ tmp = I915_READ(GEN6_PMINTRMSK); - if (tmp & GEN8_PMINTR_REDIRECT_TO_NON_DISP) { - dev_priv->rps.pm_intr_keep |= ~(tmp & ~GEN8_PMINTR_REDIRECT_TO_NON_DISP); - dev_priv->rps.pm_intr_keep &= ~GEN8_PMINTR_REDIRECT_TO_NON_DISP; + if (tmp & GEN8_PMINTR_REDIRECT_TO_GUC) { + dev_priv->rps.pm_intr_keep |= ~tmp; + dev_priv->rps.pm_intr_keep &= ~GEN8_PMINTR_REDIRECT_TO_GUC; } } @@ -152,17 +167,24 @@ static u32 get_gttype(struct drm_i915_private *dev_priv) static u32 get_core_family(struct drm_i915_private *dev_priv) { - switch (INTEL_INFO(dev_priv)->gen) { + u32 gen = INTEL_GEN(dev_priv); + + switch (gen) { case 9: return GFXCORE_FAMILY_GEN9; default: - DRM_ERROR("GUC: unsupported core family\n"); + WARN(1, "GEN%d does not support GuC operation!\n", gen); return GFXCORE_FAMILY_UNKNOWN; } } -static void set_guc_init_params(struct drm_i915_private *dev_priv) +/* + * Initialise the GuC parameter block before starting the firmware + * transfer. These parameters are read by the firmware on startup + * and cannot be changed thereafter. + */ +static void guc_params_init(struct drm_i915_private *dev_priv) { struct intel_guc *guc = &dev_priv->guc; u32 params[GUC_CTL_MAX_DWORDS]; @@ -375,11 +397,11 @@ static int guc_ucode_xfer(struct drm_i915_private *dev_priv) I915_WRITE(GEN7_MISCCPCTL, (GEN8_DOP_CLOCK_GATE_GUC_ENABLE | I915_READ(GEN7_MISCCPCTL))); - /* allows for 5us before GT can go to RC6 */ + /* allows for 5us (in 10ns units) before GT can go to RC6 */ I915_WRITE(GUC_ARAT_C6DIS, 0x1FF); } - set_guc_init_params(dev_priv); + guc_params_init(dev_priv); ret = guc_ucode_xfer_dma(dev_priv, vma); @@ -394,7 +416,7 @@ static int guc_ucode_xfer(struct drm_i915_private *dev_priv) return ret; } -static int i915_reset_guc(struct drm_i915_private *dev_priv) +static int guc_hw_reset(struct drm_i915_private *dev_priv) { int ret; u32 guc_status; @@ -447,7 +469,7 @@ int intel_guc_setup(struct drm_device *dev) goto fail; } else if (*fw_path == '\0') { /* Device has a GuC but we don't know what f/w to load? */ - DRM_INFO("No GuC firmware known for this platform\n"); + WARN(1, "No GuC firmware known for this platform!\n"); err = -ENODEV; goto fail; } @@ -461,7 +483,7 @@ int intel_guc_setup(struct drm_device *dev) goto fail; } - direct_interrupts_to_host(dev_priv); + guc_interrupts_release(dev_priv); guc_fw->guc_fw_load_status = GUC_FIRMWARE_PENDING; @@ -484,11 +506,9 @@ int intel_guc_setup(struct drm_device *dev) * Always reset the GuC just before (re)loading, so * that the state and timing are fairly predictable */ - err = i915_reset_guc(dev_priv); - if (err) { - DRM_ERROR("GuC reset failed: %d\n", err); + err = guc_hw_reset(dev_priv); + if (err) goto fail; - } err = guc_ucode_xfer(dev_priv); if (!err) @@ -511,7 +531,7 @@ int intel_guc_setup(struct drm_device *dev) err = i915_guc_submission_enable(dev_priv); if (err) goto fail; - direct_interrupts_to_guc(dev_priv); + guc_interrupts_capture(dev_priv); } return 0; @@ -520,7 +540,7 @@ fail: if (guc_fw->guc_fw_load_status == GUC_FIRMWARE_PENDING) guc_fw->guc_fw_load_status = GUC_FIRMWARE_FAIL; - direct_interrupts_to_host(dev_priv); + guc_interrupts_release(dev_priv); i915_guc_submission_disable(dev_priv); i915_guc_submission_fini(dev_priv); @@ -546,15 +566,15 @@ fail: else if (err == 0) DRM_INFO("GuC firmware load skipped\n"); else if (ret != -EIO) - DRM_INFO("GuC firmware load failed: %d\n", err); + DRM_NOTE("GuC firmware load failed: %d\n", err); else - DRM_ERROR("GuC firmware load failed: %d\n", err); + DRM_WARN("GuC firmware load failed: %d\n", err); if (i915.enable_guc_submission) { if (fw_path == NULL) DRM_INFO("GuC submission without firmware not supported\n"); if (ret == 0) - DRM_INFO("Falling back from GuC submission to execlist mode\n"); + DRM_NOTE("Falling back from GuC submission to execlist mode\n"); else DRM_ERROR("GuC init failed: %d\n", ret); } @@ -565,6 +585,7 @@ fail: static void guc_fw_fetch(struct drm_device *dev, struct intel_guc_fw *guc_fw) { + struct pci_dev *pdev = dev->pdev; struct drm_i915_gem_object *obj; const struct firmware *fw; struct guc_css_header *css; @@ -574,7 +595,7 @@ static void guc_fw_fetch(struct drm_device *dev, struct intel_guc_fw *guc_fw) DRM_DEBUG_DRIVER("before requesting firmware: GuC fw fetch status %s\n", intel_guc_fw_status_repr(guc_fw->guc_fw_fetch_status)); - err = request_firmware(&fw, guc_fw->guc_fw_path, &dev->pdev->dev); + err = request_firmware(&fw, guc_fw->guc_fw_path, &pdev->dev); if (err) goto fail; if (!fw) @@ -585,7 +606,7 @@ static void guc_fw_fetch(struct drm_device *dev, struct intel_guc_fw *guc_fw) /* Check the size of the blob before examining buffer contents */ if (fw->size < sizeof(struct guc_css_header)) { - DRM_ERROR("Firmware header is missing\n"); + DRM_NOTE("Firmware header is missing\n"); goto fail; } @@ -597,7 +618,7 @@ static void guc_fw_fetch(struct drm_device *dev, struct intel_guc_fw *guc_fw) css->key_size_dw - css->exponent_size_dw) * sizeof(u32); if (guc_fw->header_size != sizeof(struct guc_css_header)) { - DRM_ERROR("CSS header definition mismatch\n"); + DRM_NOTE("CSS header definition mismatch\n"); goto fail; } @@ -607,7 +628,7 @@ static void guc_fw_fetch(struct drm_device *dev, struct intel_guc_fw *guc_fw) /* now RSA */ if (css->key_size_dw != UOS_RSA_SCRATCH_MAX_COUNT) { - DRM_ERROR("RSA key size is bad\n"); + DRM_NOTE("RSA key size is bad\n"); goto fail; } guc_fw->rsa_offset = guc_fw->ucode_offset + guc_fw->ucode_size; @@ -616,14 +637,14 @@ static void guc_fw_fetch(struct drm_device *dev, struct intel_guc_fw *guc_fw) /* At least, it should have header, uCode and RSA. Size of all three. */ size = guc_fw->header_size + guc_fw->ucode_size + guc_fw->rsa_size; if (fw->size < size) { - DRM_ERROR("Missing firmware components\n"); + DRM_NOTE("Missing firmware components\n"); goto fail; } /* Header and uCode will be loaded to WOPCM. Size of the two. */ size = guc_fw->header_size + guc_fw->ucode_size; if (size > guc_wopcm_size(to_i915(dev))) { - DRM_ERROR("Firmware is too large to fit in WOPCM\n"); + DRM_NOTE("Firmware is too large to fit in WOPCM\n"); goto fail; } @@ -638,7 +659,7 @@ static void guc_fw_fetch(struct drm_device *dev, struct intel_guc_fw *guc_fw) if (guc_fw->guc_fw_major_found != guc_fw->guc_fw_major_wanted || guc_fw->guc_fw_minor_found < guc_fw->guc_fw_minor_wanted) { - DRM_ERROR("GuC firmware version %d.%d, required %d.%d\n", + DRM_NOTE("GuC firmware version %d.%d, required %d.%d\n", guc_fw->guc_fw_major_found, guc_fw->guc_fw_minor_found, guc_fw->guc_fw_major_wanted, guc_fw->guc_fw_minor_wanted); err = -ENOEXEC; @@ -668,10 +689,10 @@ static void guc_fw_fetch(struct drm_device *dev, struct intel_guc_fw *guc_fw) return; fail: + DRM_WARN("Failed to fetch valid GuC firmware from %s (error %d)\n", + guc_fw->guc_fw_path, err); DRM_DEBUG_DRIVER("GuC fw fetch status FAIL; err %d, fw %p, obj %p\n", err, fw, guc_fw->guc_fw_obj); - DRM_ERROR("Failed to fetch GuC firmware from %s (error %d)\n", - guc_fw->guc_fw_path, err); mutex_lock(&dev->struct_mutex); obj = guc_fw->guc_fw_obj; @@ -752,7 +773,7 @@ void intel_guc_fini(struct drm_device *dev) struct intel_guc_fw *guc_fw = &dev_priv->guc.guc_fw; mutex_lock(&dev->struct_mutex); - direct_interrupts_to_host(dev_priv); + guc_interrupts_release(dev_priv); i915_guc_submission_disable(dev_priv); i915_guc_submission_fini(dev_priv); diff --git a/drivers/gpu/drm/i915/intel_hdmi.c b/drivers/gpu/drm/i915/intel_hdmi.c index 4df9f384910c..c51073f78730 100644 --- a/drivers/gpu/drm/i915/intel_hdmi.c +++ b/drivers/gpu/drm/i915/intel_hdmi.c @@ -985,7 +985,9 @@ static void intel_enable_hdmi_audio(struct intel_encoder *encoder) intel_audio_codec_enable(encoder); } -static void g4x_enable_hdmi(struct intel_encoder *encoder) +static void g4x_enable_hdmi(struct intel_encoder *encoder, + struct intel_crtc_state *pipe_config, + struct drm_connector_state *conn_state) { struct drm_device *dev = encoder->base.dev; struct drm_i915_private *dev_priv = to_i915(dev); @@ -1006,7 +1008,9 @@ static void g4x_enable_hdmi(struct intel_encoder *encoder) intel_enable_hdmi_audio(encoder); } -static void ibx_enable_hdmi(struct intel_encoder *encoder) +static void ibx_enable_hdmi(struct intel_encoder *encoder, + struct intel_crtc_state *pipe_config, + struct drm_connector_state *conn_state) { struct drm_device *dev = encoder->base.dev; struct drm_i915_private *dev_priv = to_i915(dev); @@ -1055,7 +1059,9 @@ static void ibx_enable_hdmi(struct intel_encoder *encoder) intel_enable_hdmi_audio(encoder); } -static void cpt_enable_hdmi(struct intel_encoder *encoder) +static void cpt_enable_hdmi(struct intel_encoder *encoder, + struct intel_crtc_state *pipe_config, + struct drm_connector_state *conn_state) { struct drm_device *dev = encoder->base.dev; struct drm_i915_private *dev_priv = to_i915(dev); @@ -1108,11 +1114,15 @@ static void cpt_enable_hdmi(struct intel_encoder *encoder) intel_enable_hdmi_audio(encoder); } -static void vlv_enable_hdmi(struct intel_encoder *encoder) +static void vlv_enable_hdmi(struct intel_encoder *encoder, + struct intel_crtc_state *pipe_config, + struct drm_connector_state *conn_state) { } -static void intel_disable_hdmi(struct intel_encoder *encoder) +static void intel_disable_hdmi(struct intel_encoder *encoder, + struct intel_crtc_state *old_crtc_state, + struct drm_connector_state *old_conn_state) { struct drm_device *dev = encoder->base.dev; struct drm_i915_private *dev_priv = to_i915(dev); @@ -1164,17 +1174,21 @@ static void intel_disable_hdmi(struct intel_encoder *encoder) intel_dp_dual_mode_set_tmds_output(intel_hdmi, false); } -static void g4x_disable_hdmi(struct intel_encoder *encoder) +static void g4x_disable_hdmi(struct intel_encoder *encoder, + struct intel_crtc_state *old_crtc_state, + struct drm_connector_state *old_conn_state) { struct intel_crtc *crtc = to_intel_crtc(encoder->base.crtc); if (crtc->config->has_audio) intel_audio_codec_disable(encoder); - intel_disable_hdmi(encoder); + intel_disable_hdmi(encoder, old_crtc_state, old_conn_state); } -static void pch_disable_hdmi(struct intel_encoder *encoder) +static void pch_disable_hdmi(struct intel_encoder *encoder, + struct intel_crtc_state *old_crtc_state, + struct drm_connector_state *old_conn_state) { struct intel_crtc *crtc = to_intel_crtc(encoder->base.crtc); @@ -1182,9 +1196,11 @@ static void pch_disable_hdmi(struct intel_encoder *encoder) intel_audio_codec_disable(encoder); } -static void pch_post_disable_hdmi(struct intel_encoder *encoder) +static void pch_post_disable_hdmi(struct intel_encoder *encoder, + struct intel_crtc_state *old_crtc_state, + struct drm_connector_state *old_conn_state) { - intel_disable_hdmi(encoder); + intel_disable_hdmi(encoder, old_crtc_state, old_conn_state); } static int intel_hdmi_source_max_tmds_clock(struct drm_i915_private *dev_priv) @@ -1285,7 +1301,8 @@ static bool hdmi_12bpc_possible(struct intel_crtc_state *crtc_state) } bool intel_hdmi_compute_config(struct intel_encoder *encoder, - struct intel_crtc_state *pipe_config) + struct intel_crtc_state *pipe_config, + struct drm_connector_state *conn_state) { struct intel_hdmi *intel_hdmi = enc_to_intel_hdmi(&encoder->base); struct drm_device *dev = encoder->base.dev; @@ -1422,24 +1439,22 @@ intel_hdmi_dp_dual_mode_detect(struct drm_connector *connector, bool has_edid) } static bool -intel_hdmi_set_edid(struct drm_connector *connector, bool force) +intel_hdmi_set_edid(struct drm_connector *connector) { struct drm_i915_private *dev_priv = to_i915(connector->dev); struct intel_hdmi *intel_hdmi = intel_attached_hdmi(connector); - struct edid *edid = NULL; + struct edid *edid; bool connected = false; - if (force) { - intel_display_power_get(dev_priv, POWER_DOMAIN_GMBUS); + intel_display_power_get(dev_priv, POWER_DOMAIN_GMBUS); - edid = drm_get_edid(connector, - intel_gmbus_get_adapter(dev_priv, - intel_hdmi->ddc_bus)); + edid = drm_get_edid(connector, + intel_gmbus_get_adapter(dev_priv, + intel_hdmi->ddc_bus)); - intel_hdmi_dp_dual_mode_detect(connector, edid != NULL); + intel_hdmi_dp_dual_mode_detect(connector, edid != NULL); - intel_display_power_put(dev_priv, POWER_DOMAIN_GMBUS); - } + intel_display_power_put(dev_priv, POWER_DOMAIN_GMBUS); to_intel_connector(connector)->detect_edid = edid; if (edid && edid->input & DRM_EDID_INPUT_DIGITAL) { @@ -1465,37 +1480,16 @@ static enum drm_connector_status intel_hdmi_detect(struct drm_connector *connector, bool force) { enum drm_connector_status status; - struct intel_hdmi *intel_hdmi = intel_attached_hdmi(connector); struct drm_i915_private *dev_priv = to_i915(connector->dev); - bool live_status = false; - unsigned int try; DRM_DEBUG_KMS("[CONNECTOR:%d:%s]\n", connector->base.id, connector->name); intel_display_power_get(dev_priv, POWER_DOMAIN_GMBUS); - for (try = 0; !live_status && try < 9; try++) { - if (try) - msleep(10); - live_status = intel_digital_port_connected(dev_priv, - hdmi_to_dig_port(intel_hdmi)); - } - - if (!live_status) { - DRM_DEBUG_KMS("HDMI live status down\n"); - /* - * Live status register is not reliable on all intel platforms. - * So consider live_status only for certain platforms, for - * others, read EDID to determine presence of sink. - */ - if (INTEL_INFO(dev_priv)->gen < 7 || IS_IVYBRIDGE(dev_priv)) - live_status = true; - } - intel_hdmi_unset_edid(connector); - if (intel_hdmi_set_edid(connector, live_status)) { + if (intel_hdmi_set_edid(connector)) { struct intel_hdmi *intel_hdmi = intel_attached_hdmi(connector); hdmi_to_dig_port(intel_hdmi)->base.type = INTEL_OUTPUT_HDMI; @@ -1521,7 +1515,7 @@ intel_hdmi_force(struct drm_connector *connector) if (connector->status != connector_status_connected) return; - intel_hdmi_set_edid(connector, true); + intel_hdmi_set_edid(connector); hdmi_to_dig_port(intel_hdmi)->base.type = INTEL_OUTPUT_HDMI; } @@ -1638,7 +1632,9 @@ done: return 0; } -static void intel_hdmi_pre_enable(struct intel_encoder *encoder) +static void intel_hdmi_pre_enable(struct intel_encoder *encoder, + struct intel_crtc_state *pipe_config, + struct drm_connector_state *conn_state) { struct intel_hdmi *intel_hdmi = enc_to_intel_hdmi(&encoder->base); struct intel_crtc *intel_crtc = to_intel_crtc(encoder->base.crtc); @@ -1651,7 +1647,9 @@ static void intel_hdmi_pre_enable(struct intel_encoder *encoder) adjusted_mode); } -static void vlv_hdmi_pre_enable(struct intel_encoder *encoder) +static void vlv_hdmi_pre_enable(struct intel_encoder *encoder, + struct intel_crtc_state *pipe_config, + struct drm_connector_state *conn_state) { struct intel_digital_port *dport = enc_to_dig_port(&encoder->base); struct intel_hdmi *intel_hdmi = &dport->hdmi; @@ -1671,37 +1669,47 @@ static void vlv_hdmi_pre_enable(struct intel_encoder *encoder) intel_crtc->config->has_hdmi_sink, adjusted_mode); - g4x_enable_hdmi(encoder); + g4x_enable_hdmi(encoder, pipe_config, conn_state); vlv_wait_port_ready(dev_priv, dport, 0x0); } -static void vlv_hdmi_pre_pll_enable(struct intel_encoder *encoder) +static void vlv_hdmi_pre_pll_enable(struct intel_encoder *encoder, + struct intel_crtc_state *pipe_config, + struct drm_connector_state *conn_state) { intel_hdmi_prepare(encoder); vlv_phy_pre_pll_enable(encoder); } -static void chv_hdmi_pre_pll_enable(struct intel_encoder *encoder) +static void chv_hdmi_pre_pll_enable(struct intel_encoder *encoder, + struct intel_crtc_state *pipe_config, + struct drm_connector_state *conn_state) { intel_hdmi_prepare(encoder); chv_phy_pre_pll_enable(encoder); } -static void chv_hdmi_post_pll_disable(struct intel_encoder *encoder) +static void chv_hdmi_post_pll_disable(struct intel_encoder *encoder, + struct intel_crtc_state *old_crtc_state, + struct drm_connector_state *old_conn_state) { chv_phy_post_pll_disable(encoder); } -static void vlv_hdmi_post_disable(struct intel_encoder *encoder) +static void vlv_hdmi_post_disable(struct intel_encoder *encoder, + struct intel_crtc_state *old_crtc_state, + struct drm_connector_state *old_conn_state) { /* Reset lanes to avoid HDMI flicker (VLV w/a) */ vlv_phy_reset_lanes(encoder); } -static void chv_hdmi_post_disable(struct intel_encoder *encoder) +static void chv_hdmi_post_disable(struct intel_encoder *encoder, + struct intel_crtc_state *old_crtc_state, + struct drm_connector_state *old_conn_state) { struct drm_device *dev = encoder->base.dev; struct drm_i915_private *dev_priv = to_i915(dev); @@ -1714,7 +1722,9 @@ static void chv_hdmi_post_disable(struct intel_encoder *encoder) mutex_unlock(&dev_priv->sb_lock); } -static void chv_hdmi_pre_enable(struct intel_encoder *encoder) +static void chv_hdmi_pre_enable(struct intel_encoder *encoder, + struct intel_crtc_state *pipe_config, + struct drm_connector_state *conn_state) { struct intel_digital_port *dport = enc_to_dig_port(&encoder->base); struct intel_hdmi *intel_hdmi = &dport->hdmi; @@ -1734,7 +1744,7 @@ static void chv_hdmi_pre_enable(struct intel_encoder *encoder) intel_crtc->config->has_hdmi_sink, adjusted_mode); - g4x_enable_hdmi(encoder); + g4x_enable_hdmi(encoder, pipe_config, conn_state); vlv_wait_port_ready(dev_priv, dport, 0x0); diff --git a/drivers/gpu/drm/i915/intel_i2c.c b/drivers/gpu/drm/i915/intel_i2c.c index 1f266d7df2ec..79aab9ad6faa 100644 --- a/drivers/gpu/drm/i915/intel_i2c.c +++ b/drivers/gpu/drm/i915/intel_i2c.c @@ -255,67 +255,59 @@ intel_gpio_setup(struct intel_gmbus *bus, unsigned int pin) algo->data = bus; } -static int -gmbus_wait_hw_status(struct drm_i915_private *dev_priv, - u32 gmbus2_status, - u32 gmbus4_irq_en) +static int gmbus_wait(struct drm_i915_private *dev_priv, u32 status, u32 irq_en) { - int i; - u32 gmbus2 = 0; DEFINE_WAIT(wait); - - if (!HAS_GMBUS_IRQ(dev_priv)) - gmbus4_irq_en = 0; + u32 gmbus2; + int ret; /* Important: The hw handles only the first bit, so set only one! Since * we also need to check for NAKs besides the hw ready/idle signal, we - * need to wake up periodically and check that ourselves. */ - I915_WRITE(GMBUS4, gmbus4_irq_en); - - for (i = 0; i < msecs_to_jiffies_timeout(50); i++) { - prepare_to_wait(&dev_priv->gmbus_wait_queue, &wait, - TASK_UNINTERRUPTIBLE); + * need to wake up periodically and check that ourselves. + */ + if (!HAS_GMBUS_IRQ(dev_priv)) + irq_en = 0; - gmbus2 = I915_READ_NOTRACE(GMBUS2); - if (gmbus2 & (GMBUS_SATOER | gmbus2_status)) - break; + add_wait_queue(&dev_priv->gmbus_wait_queue, &wait); + I915_WRITE_FW(GMBUS4, irq_en); - schedule_timeout(1); - } - finish_wait(&dev_priv->gmbus_wait_queue, &wait); + status |= GMBUS_SATOER; + ret = wait_for_us((gmbus2 = I915_READ_FW(GMBUS2)) & status, 2); + if (ret) + ret = wait_for((gmbus2 = I915_READ_FW(GMBUS2)) & status, 50); - I915_WRITE(GMBUS4, 0); + I915_WRITE_FW(GMBUS4, 0); + remove_wait_queue(&dev_priv->gmbus_wait_queue, &wait); if (gmbus2 & GMBUS_SATOER) return -ENXIO; - if (gmbus2 & gmbus2_status) - return 0; - return -ETIMEDOUT; + + return ret; } static int gmbus_wait_idle(struct drm_i915_private *dev_priv) { + DEFINE_WAIT(wait); + u32 irq_enable; int ret; - if (!HAS_GMBUS_IRQ(dev_priv)) - return intel_wait_for_register(dev_priv, - GMBUS2, GMBUS_ACTIVE, 0, - 10); - /* Important: The hw handles only the first bit, so set only one! */ - I915_WRITE(GMBUS4, GMBUS_IDLE_EN); + irq_enable = 0; + if (HAS_GMBUS_IRQ(dev_priv)) + irq_enable = GMBUS_IDLE_EN; - ret = wait_event_timeout(dev_priv->gmbus_wait_queue, - (I915_READ_NOTRACE(GMBUS2) & GMBUS_ACTIVE) == 0, - msecs_to_jiffies_timeout(10)); + add_wait_queue(&dev_priv->gmbus_wait_queue, &wait); + I915_WRITE_FW(GMBUS4, irq_enable); - I915_WRITE(GMBUS4, 0); + ret = intel_wait_for_register_fw(dev_priv, + GMBUS2, GMBUS_ACTIVE, 0, + 10); - if (ret) - return 0; - else - return -ETIMEDOUT; + I915_WRITE_FW(GMBUS4, 0); + remove_wait_queue(&dev_priv->gmbus_wait_queue, &wait); + + return ret; } static int @@ -323,22 +315,21 @@ gmbus_xfer_read_chunk(struct drm_i915_private *dev_priv, unsigned short addr, u8 *buf, unsigned int len, u32 gmbus1_index) { - I915_WRITE(GMBUS1, - gmbus1_index | - GMBUS_CYCLE_WAIT | - (len << GMBUS_BYTE_COUNT_SHIFT) | - (addr << GMBUS_SLAVE_ADDR_SHIFT) | - GMBUS_SLAVE_READ | GMBUS_SW_RDY); + I915_WRITE_FW(GMBUS1, + gmbus1_index | + GMBUS_CYCLE_WAIT | + (len << GMBUS_BYTE_COUNT_SHIFT) | + (addr << GMBUS_SLAVE_ADDR_SHIFT) | + GMBUS_SLAVE_READ | GMBUS_SW_RDY); while (len) { int ret; u32 val, loop = 0; - ret = gmbus_wait_hw_status(dev_priv, GMBUS_HW_RDY, - GMBUS_HW_RDY_EN); + ret = gmbus_wait(dev_priv, GMBUS_HW_RDY, GMBUS_HW_RDY_EN); if (ret) return ret; - val = I915_READ(GMBUS3); + val = I915_READ_FW(GMBUS3); do { *buf++ = val & 0xff; val >>= 8; @@ -385,12 +376,12 @@ gmbus_xfer_write_chunk(struct drm_i915_private *dev_priv, len -= 1; } - I915_WRITE(GMBUS3, val); - I915_WRITE(GMBUS1, - GMBUS_CYCLE_WAIT | - (chunk_size << GMBUS_BYTE_COUNT_SHIFT) | - (addr << GMBUS_SLAVE_ADDR_SHIFT) | - GMBUS_SLAVE_WRITE | GMBUS_SW_RDY); + I915_WRITE_FW(GMBUS3, val); + I915_WRITE_FW(GMBUS1, + GMBUS_CYCLE_WAIT | + (chunk_size << GMBUS_BYTE_COUNT_SHIFT) | + (addr << GMBUS_SLAVE_ADDR_SHIFT) | + GMBUS_SLAVE_WRITE | GMBUS_SW_RDY); while (len) { int ret; @@ -399,10 +390,9 @@ gmbus_xfer_write_chunk(struct drm_i915_private *dev_priv, val |= *buf++ << (8 * loop); } while (--len && ++loop < 4); - I915_WRITE(GMBUS3, val); + I915_WRITE_FW(GMBUS3, val); - ret = gmbus_wait_hw_status(dev_priv, GMBUS_HW_RDY, - GMBUS_HW_RDY_EN); + ret = gmbus_wait(dev_priv, GMBUS_HW_RDY, GMBUS_HW_RDY_EN); if (ret) return ret; } @@ -460,13 +450,13 @@ gmbus_xfer_index_read(struct drm_i915_private *dev_priv, struct i2c_msg *msgs) /* GMBUS5 holds 16-bit index */ if (gmbus5) - I915_WRITE(GMBUS5, gmbus5); + I915_WRITE_FW(GMBUS5, gmbus5); ret = gmbus_xfer_read(dev_priv, &msgs[1], gmbus1_index); /* Clear GMBUS5 after each index transfer */ if (gmbus5) - I915_WRITE(GMBUS5, 0); + I915_WRITE_FW(GMBUS5, 0); return ret; } @@ -478,11 +468,15 @@ do_gmbus_xfer(struct i2c_adapter *adapter, struct i2c_msg *msgs, int num) struct intel_gmbus, adapter); struct drm_i915_private *dev_priv = bus->dev_priv; + const unsigned int fw = + intel_uncore_forcewake_for_reg(dev_priv, GMBUS0, + FW_REG_READ | FW_REG_WRITE); int i = 0, inc, try = 0; int ret = 0; + intel_uncore_forcewake_get(dev_priv, fw); retry: - I915_WRITE(GMBUS0, bus->reg0); + I915_WRITE_FW(GMBUS0, bus->reg0); for (; i < num; i += inc) { inc = 1; @@ -496,8 +490,8 @@ retry: } if (!ret) - ret = gmbus_wait_hw_status(dev_priv, GMBUS_HW_WAIT_PHASE, - GMBUS_HW_WAIT_EN); + ret = gmbus_wait(dev_priv, + GMBUS_HW_WAIT_PHASE, GMBUS_HW_WAIT_EN); if (ret == -ETIMEDOUT) goto timeout; else if (ret) @@ -508,7 +502,7 @@ retry: * a STOP on the very first cycle. To simplify the code we * unconditionally generate the STOP condition with an additional gmbus * cycle. */ - I915_WRITE(GMBUS1, GMBUS_CYCLE_STOP | GMBUS_SW_RDY); + I915_WRITE_FW(GMBUS1, GMBUS_CYCLE_STOP | GMBUS_SW_RDY); /* Mark the GMBUS interface as disabled after waiting for idle. * We will re-enable it at the start of the next xfer, @@ -519,7 +513,7 @@ retry: adapter->name); ret = -ETIMEDOUT; } - I915_WRITE(GMBUS0, 0); + I915_WRITE_FW(GMBUS0, 0); ret = ret ?: i; goto out; @@ -548,9 +542,9 @@ clear_err: * of resetting the GMBUS controller and so clearing the * BUS_ERROR raised by the slave's NAK. */ - I915_WRITE(GMBUS1, GMBUS_SW_CLR_INT); - I915_WRITE(GMBUS1, 0); - I915_WRITE(GMBUS0, 0); + I915_WRITE_FW(GMBUS1, GMBUS_SW_CLR_INT); + I915_WRITE_FW(GMBUS1, 0); + I915_WRITE_FW(GMBUS0, 0); DRM_DEBUG_KMS("GMBUS [%s] NAK for addr: %04x %c(%d)\n", adapter->name, msgs[i].addr, @@ -573,7 +567,7 @@ clear_err: timeout: DRM_DEBUG_KMS("GMBUS [%s] timed out, falling back to bit banging on pin %d\n", bus->adapter.name, bus->reg0 & 0xff); - I915_WRITE(GMBUS0, 0); + I915_WRITE_FW(GMBUS0, 0); /* * Hardware may not support GMBUS over these pins? Try GPIO bitbanging @@ -582,6 +576,7 @@ timeout: ret = -EAGAIN; out: + intel_uncore_forcewake_put(dev_priv, fw); return ret; } @@ -633,6 +628,7 @@ static const struct i2c_algorithm gmbus_algorithm = { int intel_setup_gmbus(struct drm_device *dev) { struct drm_i915_private *dev_priv = to_i915(dev); + struct pci_dev *pdev = dev_priv->drm.pdev; struct intel_gmbus *bus; unsigned int pin; int ret; @@ -663,7 +659,7 @@ int intel_setup_gmbus(struct drm_device *dev) "i915 gmbus %s", get_gmbus_pin(dev_priv, pin)->name); - bus->adapter.dev.parent = &dev->pdev->dev; + bus->adapter.dev.parent = &pdev->dev; bus->dev_priv = dev_priv; bus->adapter.algo = &gmbus_algorithm; diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c index 6b49df4316f4..251143361f31 100644 --- a/drivers/gpu/drm/i915/intel_lrc.c +++ b/drivers/gpu/drm/i915/intel_lrc.c @@ -156,6 +156,11 @@ #define GEN8_CTX_STATUS_COMPLETE (1 << 4) #define GEN8_CTX_STATUS_LITE_RESTORE (1 << 15) +#define GEN8_CTX_STATUS_COMPLETED_MASK \ + (GEN8_CTX_STATUS_ACTIVE_IDLE | \ + GEN8_CTX_STATUS_PREEMPTED | \ + GEN8_CTX_STATUS_ELEMENT_SWITCH) + #define CTX_LRI_HEADER_0 0x01 #define CTX_CONTEXT_CONTROL 0x02 #define CTX_RING_HEAD 0x04 @@ -263,12 +268,10 @@ logical_ring_init_platform_invariants(struct intel_engine_cs *engine) { struct drm_i915_private *dev_priv = engine->i915; - if (IS_GEN8(dev_priv) || IS_GEN9(dev_priv)) - engine->idle_lite_restore_wa = ~0; - - engine->disable_lite_restore_wa = (IS_SKL_REVID(dev_priv, 0, SKL_REVID_B0) || - IS_BXT_REVID(dev_priv, 0, BXT_REVID_A1)) && - (engine->id == VCS || engine->id == VCS2); + engine->disable_lite_restore_wa = + (IS_SKL_REVID(dev_priv, 0, SKL_REVID_B0) || + IS_BXT_REVID(dev_priv, 0, BXT_REVID_A1)) && + (engine->id == VCS || engine->id == VCS2); engine->ctx_desc_template = GEN8_CTX_VALID; if (IS_GEN8(dev_priv)) @@ -328,34 +331,18 @@ uint64_t intel_lr_context_descriptor(struct i915_gem_context *ctx, return ctx->engine[engine->id].lrc_desc; } -static void execlists_elsp_write(struct drm_i915_gem_request *rq0, - struct drm_i915_gem_request *rq1) +static inline void +execlists_context_status_change(struct drm_i915_gem_request *rq, + unsigned long status) { + /* + * Only used when GVT-g is enabled now. When GVT-g is disabled, + * The compiler should eliminate this function as dead-code. + */ + if (!IS_ENABLED(CONFIG_DRM_I915_GVT)) + return; - struct intel_engine_cs *engine = rq0->engine; - struct drm_i915_private *dev_priv = rq0->i915; - uint64_t desc[2]; - - if (rq1) { - desc[1] = intel_lr_context_descriptor(rq1->ctx, rq1->engine); - rq1->elsp_submitted++; - } else { - desc[1] = 0; - } - - desc[0] = intel_lr_context_descriptor(rq0->ctx, rq0->engine); - rq0->elsp_submitted++; - - /* You must always write both descriptors in the order below. */ - I915_WRITE_FW(RING_ELSP(engine), upper_32_bits(desc[1])); - I915_WRITE_FW(RING_ELSP(engine), lower_32_bits(desc[1])); - - I915_WRITE_FW(RING_ELSP(engine), upper_32_bits(desc[0])); - /* The context is automatically loaded after the following */ - I915_WRITE_FW(RING_ELSP(engine), lower_32_bits(desc[0])); - - /* ELSP is a wo register, use another nearby reg for posting */ - POSTING_READ_FW(RING_EXECLIST_STATUS_LO(engine)); + atomic_notifier_call_chain(&rq->ctx->status_notifier, status, rq); } static void @@ -367,11 +354,11 @@ execlists_update_context_pdps(struct i915_hw_ppgtt *ppgtt, u32 *reg_state) ASSIGN_CTX_PDP(ppgtt, reg_state, 0); } -static void execlists_update_context(struct drm_i915_gem_request *rq) +static u64 execlists_update_context(struct drm_i915_gem_request *rq) { - struct intel_engine_cs *engine = rq->engine; + struct intel_context *ce = &rq->ctx->engine[rq->engine->id]; struct i915_hw_ppgtt *ppgtt = rq->ctx->ppgtt; - uint32_t *reg_state = rq->ctx->engine[engine->id].lrc_reg_state; + u32 *reg_state = ce->lrc_reg_state; reg_state[CTX_RING_TAIL+1] = intel_ring_offset(rq->ring, rq->tail); @@ -382,158 +369,163 @@ static void execlists_update_context(struct drm_i915_gem_request *rq) */ if (ppgtt && !USES_FULL_48BIT_PPGTT(ppgtt->base.dev)) execlists_update_context_pdps(ppgtt, reg_state); + + return ce->lrc_desc; } -static void execlists_elsp_submit_contexts(struct drm_i915_gem_request *rq0, - struct drm_i915_gem_request *rq1) +static void execlists_submit_ports(struct intel_engine_cs *engine) { - struct drm_i915_private *dev_priv = rq0->i915; - unsigned int fw_domains = rq0->engine->fw_domains; - - execlists_update_context(rq0); + struct drm_i915_private *dev_priv = engine->i915; + struct execlist_port *port = engine->execlist_port; + u32 __iomem *elsp = + dev_priv->regs + i915_mmio_reg_offset(RING_ELSP(engine)); + u64 desc[2]; - if (rq1) - execlists_update_context(rq1); + if (!port[0].count) + execlists_context_status_change(port[0].request, + INTEL_CONTEXT_SCHEDULE_IN); + desc[0] = execlists_update_context(port[0].request); + engine->preempt_wa = port[0].count++; /* bdw only? fixed on skl? */ - spin_lock_irq(&dev_priv->uncore.lock); - intel_uncore_forcewake_get__locked(dev_priv, fw_domains); + if (port[1].request) { + GEM_BUG_ON(port[1].count); + execlists_context_status_change(port[1].request, + INTEL_CONTEXT_SCHEDULE_IN); + desc[1] = execlists_update_context(port[1].request); + port[1].count = 1; + } else { + desc[1] = 0; + } + GEM_BUG_ON(desc[0] == desc[1]); - execlists_elsp_write(rq0, rq1); + /* You must always write both descriptors in the order below. */ + writel(upper_32_bits(desc[1]), elsp); + writel(lower_32_bits(desc[1]), elsp); - intel_uncore_forcewake_put__locked(dev_priv, fw_domains); - spin_unlock_irq(&dev_priv->uncore.lock); + writel(upper_32_bits(desc[0]), elsp); + /* The context is automatically loaded after the following */ + writel(lower_32_bits(desc[0]), elsp); } -static inline void execlists_context_status_change( - struct drm_i915_gem_request *rq, - unsigned long status) +static bool ctx_single_port_submission(const struct i915_gem_context *ctx) { - /* - * Only used when GVT-g is enabled now. When GVT-g is disabled, - * The compiler should eliminate this function as dead-code. - */ - if (!IS_ENABLED(CONFIG_DRM_I915_GVT)) - return; - - atomic_notifier_call_chain(&rq->ctx->status_notifier, status, rq); + return (IS_ENABLED(CONFIG_DRM_I915_GVT) && + ctx->execlists_force_single_submission); } -static void execlists_unqueue(struct intel_engine_cs *engine) +static bool can_merge_ctx(const struct i915_gem_context *prev, + const struct i915_gem_context *next) { - struct drm_i915_gem_request *req0 = NULL, *req1 = NULL; - struct drm_i915_gem_request *cursor, *tmp; + if (prev != next) + return false; - assert_spin_locked(&engine->execlist_lock); + if (ctx_single_port_submission(prev)) + return false; - /* - * If irqs are not active generate a warning as batches that finish - * without the irqs may get lost and a GPU Hang may occur. - */ - WARN_ON(!intel_irqs_enabled(engine->i915)); - - /* Try to read in pairs */ - list_for_each_entry_safe(cursor, tmp, &engine->execlist_queue, - execlist_link) { - if (!req0) { - req0 = cursor; - } else if (req0->ctx == cursor->ctx) { - /* Same ctx: ignore first request, as second request - * will update tail past first request's workload */ - cursor->elsp_submitted = req0->elsp_submitted; - list_del(&req0->execlist_link); - i915_gem_request_put(req0); - req0 = cursor; - } else { - if (IS_ENABLED(CONFIG_DRM_I915_GVT)) { - /* - * req0 (after merged) ctx requires single - * submission, stop picking - */ - if (req0->ctx->execlists_force_single_submission) - break; - /* - * req0 ctx doesn't require single submission, - * but next req ctx requires, stop picking - */ - if (cursor->ctx->execlists_force_single_submission) - break; - } - req1 = cursor; - WARN_ON(req1->elsp_submitted); - break; - } - } + return true; +} - if (unlikely(!req0)) - return; +static void execlists_dequeue(struct intel_engine_cs *engine) +{ + struct drm_i915_gem_request *cursor, *last; + struct execlist_port *port = engine->execlist_port; + bool submit = false; + + last = port->request; + if (last) + /* WaIdleLiteRestore:bdw,skl + * Apply the wa NOOPs to prevent ring:HEAD == req:TAIL + * as we resubmit the request. See gen8_emit_request() + * for where we prepare the padding after the end of the + * request. + */ + last->tail = last->wa_tail; - execlists_context_status_change(req0, INTEL_CONTEXT_SCHEDULE_IN); + GEM_BUG_ON(port[1].request); - if (req1) - execlists_context_status_change(req1, - INTEL_CONTEXT_SCHEDULE_IN); + /* Hardware submission is through 2 ports. Conceptually each port + * has a (RING_START, RING_HEAD, RING_TAIL) tuple. RING_START is + * static for a context, and unique to each, so we only execute + * requests belonging to a single context from each ring. RING_HEAD + * is maintained by the CS in the context image, it marks the place + * where it got up to last time, and through RING_TAIL we tell the CS + * where we want to execute up to this time. + * + * In this list the requests are in order of execution. Consecutive + * requests from the same context are adjacent in the ringbuffer. We + * can combine these requests into a single RING_TAIL update: + * + * RING_HEAD...req1...req2 + * ^- RING_TAIL + * since to execute req2 the CS must first execute req1. + * + * Our goal then is to point each port to the end of a consecutive + * sequence of requests as being the most optimal (fewest wake ups + * and context switches) submission. + */ - if (req0->elsp_submitted & engine->idle_lite_restore_wa) { - /* - * WaIdleLiteRestore: make sure we never cause a lite restore - * with HEAD==TAIL. + spin_lock(&engine->execlist_lock); + list_for_each_entry(cursor, &engine->execlist_queue, execlist_link) { + /* Can we combine this request with the current port? It has to + * be the same context/ringbuffer and not have any exceptions + * (e.g. GVT saying never to combine contexts). * - * Apply the wa NOOPS to prevent ring:HEAD == req:TAIL as we - * resubmit the request. See gen8_emit_request() for where we - * prepare the padding after the end of the request. + * If we can combine the requests, we can execute both by + * updating the RING_TAIL to point to the end of the second + * request, and so we never need to tell the hardware about + * the first. */ - req0->tail += 8; - req0->tail &= req0->ring->size - 1; + if (last && !can_merge_ctx(cursor->ctx, last->ctx)) { + /* If we are on the second port and cannot combine + * this request with the last, then we are done. + */ + if (port != engine->execlist_port) + break; + + /* If GVT overrides us we only ever submit port[0], + * leaving port[1] empty. Note that we also have + * to be careful that we don't queue the same + * context (even though a different request) to + * the second port. + */ + if (ctx_single_port_submission(cursor->ctx)) + break; + + GEM_BUG_ON(last->ctx == cursor->ctx); + + i915_gem_request_assign(&port->request, last); + port++; + } + last = cursor; + submit = true; + } + if (submit) { + /* Decouple all the requests submitted from the queue */ + engine->execlist_queue.next = &cursor->execlist_link; + cursor->execlist_link.prev = &engine->execlist_queue; + + i915_gem_request_assign(&port->request, last); } + spin_unlock(&engine->execlist_lock); - execlists_elsp_submit_contexts(req0, req1); + if (submit) + execlists_submit_ports(engine); } -static unsigned int -execlists_check_remove_request(struct intel_engine_cs *engine, u32 ctx_id) +static bool execlists_elsp_idle(struct intel_engine_cs *engine) { - struct drm_i915_gem_request *head_req; - - assert_spin_locked(&engine->execlist_lock); - - head_req = list_first_entry_or_null(&engine->execlist_queue, - struct drm_i915_gem_request, - execlist_link); - - if (WARN_ON(!head_req || (head_req->ctx_hw_id != ctx_id))) - return 0; - - WARN(head_req->elsp_submitted == 0, "Never submitted head request\n"); - - if (--head_req->elsp_submitted > 0) - return 0; - - execlists_context_status_change(head_req, INTEL_CONTEXT_SCHEDULE_OUT); - - list_del(&head_req->execlist_link); - i915_gem_request_put(head_req); - - return 1; + return !engine->execlist_port[0].request; } -static u32 -get_context_status(struct intel_engine_cs *engine, unsigned int read_pointer, - u32 *context_id) +static bool execlists_elsp_ready(struct intel_engine_cs *engine) { - struct drm_i915_private *dev_priv = engine->i915; - u32 status; - - read_pointer %= GEN8_CSB_ENTRIES; - - status = I915_READ_FW(RING_CONTEXT_STATUS_BUF_LO(engine, read_pointer)); - - if (status & GEN8_CTX_STATUS_IDLE_ACTIVE) - return 0; + int port; - *context_id = I915_READ_FW(RING_CONTEXT_STATUS_BUF_HI(engine, - read_pointer)); + port = 1; /* wait for a free slot */ + if (engine->disable_lite_restore_wa || engine->preempt_wa) + port = 0; /* wait for GPU to be idle before continuing */ - return status; + return !engine->execlist_port[port].request; } /* @@ -543,103 +535,70 @@ get_context_status(struct intel_engine_cs *engine, unsigned int read_pointer, static void intel_lrc_irq_handler(unsigned long data) { struct intel_engine_cs *engine = (struct intel_engine_cs *)data; + struct execlist_port *port = engine->execlist_port; struct drm_i915_private *dev_priv = engine->i915; - u32 status_pointer; - unsigned int read_pointer, write_pointer; - u32 csb[GEN8_CSB_ENTRIES][2]; - unsigned int csb_read = 0, i; - unsigned int submit_contexts = 0; intel_uncore_forcewake_get(dev_priv, engine->fw_domains); - status_pointer = I915_READ_FW(RING_CONTEXT_STATUS_PTR(engine)); - - read_pointer = engine->next_context_status_buffer; - write_pointer = GEN8_CSB_WRITE_PTR(status_pointer); - if (read_pointer > write_pointer) - write_pointer += GEN8_CSB_ENTRIES; - - while (read_pointer < write_pointer) { - if (WARN_ON_ONCE(csb_read == GEN8_CSB_ENTRIES)) - break; - csb[csb_read][0] = get_context_status(engine, ++read_pointer, - &csb[csb_read][1]); - csb_read++; - } - - engine->next_context_status_buffer = write_pointer % GEN8_CSB_ENTRIES; - - /* Update the read pointer to the old write pointer. Manual ringbuffer - * management ftw </sarcasm> */ - I915_WRITE_FW(RING_CONTEXT_STATUS_PTR(engine), - _MASKED_FIELD(GEN8_CSB_READ_PTR_MASK, - engine->next_context_status_buffer << 8)); - - intel_uncore_forcewake_put(dev_priv, engine->fw_domains); - - spin_lock(&engine->execlist_lock); + if (!execlists_elsp_idle(engine)) { + u32 __iomem *csb_mmio = + dev_priv->regs + i915_mmio_reg_offset(RING_CONTEXT_STATUS_PTR(engine)); + u32 __iomem *buf = + dev_priv->regs + i915_mmio_reg_offset(RING_CONTEXT_STATUS_BUF_LO(engine, 0)); + unsigned int csb, head, tail; + + csb = readl(csb_mmio); + head = GEN8_CSB_READ_PTR(csb); + tail = GEN8_CSB_WRITE_PTR(csb); + if (tail < head) + tail += GEN8_CSB_ENTRIES; + while (head < tail) { + unsigned int idx = ++head % GEN8_CSB_ENTRIES; + unsigned int status = readl(buf + 2 * idx); + + if (!(status & GEN8_CTX_STATUS_COMPLETED_MASK)) + continue; + + GEM_BUG_ON(port[0].count == 0); + if (--port[0].count == 0) { + GEM_BUG_ON(status & GEN8_CTX_STATUS_PREEMPTED); + execlists_context_status_change(port[0].request, + INTEL_CONTEXT_SCHEDULE_OUT); + + i915_gem_request_put(port[0].request); + port[0] = port[1]; + memset(&port[1], 0, sizeof(port[1])); + + engine->preempt_wa = false; + } - for (i = 0; i < csb_read; i++) { - if (unlikely(csb[i][0] & GEN8_CTX_STATUS_PREEMPTED)) { - if (csb[i][0] & GEN8_CTX_STATUS_LITE_RESTORE) { - if (execlists_check_remove_request(engine, csb[i][1])) - WARN(1, "Lite Restored request removed from queue\n"); - } else - WARN(1, "Preemption without Lite Restore\n"); + GEM_BUG_ON(port[0].count == 0 && + !(status & GEN8_CTX_STATUS_ACTIVE_IDLE)); } - if (csb[i][0] & (GEN8_CTX_STATUS_ACTIVE_IDLE | - GEN8_CTX_STATUS_ELEMENT_SWITCH)) - submit_contexts += - execlists_check_remove_request(engine, csb[i][1]); - } - - if (submit_contexts) { - if (!engine->disable_lite_restore_wa || - (csb[i][0] & GEN8_CTX_STATUS_ACTIVE_IDLE)) - execlists_unqueue(engine); + writel(_MASKED_FIELD(GEN8_CSB_READ_PTR_MASK, + GEN8_CSB_WRITE_PTR(csb) << 8), + csb_mmio); } - spin_unlock(&engine->execlist_lock); + if (execlists_elsp_ready(engine)) + execlists_dequeue(engine); - if (unlikely(submit_contexts > 2)) - DRM_ERROR("More than two context complete events?\n"); + intel_uncore_forcewake_put(dev_priv, engine->fw_domains); } static void execlists_submit_request(struct drm_i915_gem_request *request) { struct intel_engine_cs *engine = request->engine; - struct drm_i915_gem_request *cursor; - int num_elements = 0; - - spin_lock_bh(&engine->execlist_lock); + unsigned long flags; - list_for_each_entry(cursor, &engine->execlist_queue, execlist_link) - if (++num_elements > 2) - break; + spin_lock_irqsave(&engine->execlist_lock, flags); - if (num_elements > 2) { - struct drm_i915_gem_request *tail_req; - - tail_req = list_last_entry(&engine->execlist_queue, - struct drm_i915_gem_request, - execlist_link); - - if (request->ctx == tail_req->ctx) { - WARN(tail_req->elsp_submitted != 0, - "More than 2 already-submitted reqs queued\n"); - list_del(&tail_req->execlist_link); - i915_gem_request_put(tail_req); - } - } - - i915_gem_request_get(request); list_add_tail(&request->execlist_link, &engine->execlist_queue); - request->ctx_hw_id = request->ctx->hw_id; - if (num_elements == 0) - execlists_unqueue(engine); + if (execlists_elsp_idle(engine)) + tasklet_hi_schedule(&engine->irq_tasklet); - spin_unlock_bh(&engine->execlist_lock); + spin_unlock_irqrestore(&engine->execlist_lock, flags); } int intel_logical_ring_alloc_request_extras(struct drm_i915_gem_request *request) @@ -668,7 +627,7 @@ int intel_logical_ring_alloc_request_extras(struct drm_i915_gem_request *request * going any further, as the i915_add_request() call * later on mustn't fail ... */ - ret = i915_guc_wq_check_space(request); + ret = i915_guc_wq_reserve(request); if (ret) return ret; } @@ -731,6 +690,7 @@ intel_logical_ring_advance(struct drm_i915_gem_request *request) intel_ring_emit(ring, MI_NOOP); intel_ring_emit(ring, MI_NOOP); intel_ring_advance(ring); + request->wa_tail = ring->tail; /* We keep the previous context alive until we retire the following * request. This ensures that any the context object is still pinned @@ -743,23 +703,6 @@ intel_logical_ring_advance(struct drm_i915_gem_request *request) return 0; } -void intel_execlists_cancel_requests(struct intel_engine_cs *engine) -{ - struct drm_i915_gem_request *req, *tmp; - LIST_HEAD(cancel_list); - - WARN_ON(!mutex_is_locked(&engine->i915->drm.struct_mutex)); - - spin_lock_bh(&engine->execlist_lock); - list_replace_init(&engine->execlist_queue, &cancel_list); - spin_unlock_bh(&engine->execlist_lock); - - list_for_each_entry_safe(req, tmp, &cancel_list, execlist_link) { - list_del(&req->execlist_link); - i915_gem_request_put(req); - } -} - static int intel_lr_context_pin(struct i915_gem_context *ctx, struct intel_engine_cs *engine) { @@ -1280,48 +1223,30 @@ static void lrc_init_hws(struct intel_engine_cs *engine) static int gen8_init_common_ring(struct intel_engine_cs *engine) { struct drm_i915_private *dev_priv = engine->i915; - unsigned int next_context_status_buffer_hw; + int ret; + + ret = intel_mocs_init_engine(engine); + if (ret) + return ret; lrc_init_hws(engine); - I915_WRITE_IMR(engine, - ~(engine->irq_enable_mask | engine->irq_keep_mask)); + intel_engine_reset_irq(engine); + I915_WRITE(RING_HWSTAM(engine->mmio_base), 0xffffffff); I915_WRITE(RING_MODE_GEN7(engine), _MASKED_BIT_DISABLE(GFX_REPLAY_MODE) | _MASKED_BIT_ENABLE(GFX_RUN_LIST_ENABLE)); - POSTING_READ(RING_MODE_GEN7(engine)); - - /* - * Instead of resetting the Context Status Buffer (CSB) read pointer to - * zero, we need to read the write pointer from hardware and use its - * value because "this register is power context save restored". - * Effectively, these states have been observed: - * - * | Suspend-to-idle (freeze) | Suspend-to-RAM (mem) | - * BDW | CSB regs not reset | CSB regs reset | - * CHT | CSB regs not reset | CSB regs not reset | - * SKL | ? | ? | - * BXT | ? | ? | - */ - next_context_status_buffer_hw = - GEN8_CSB_WRITE_PTR(I915_READ(RING_CONTEXT_STATUS_PTR(engine))); - - /* - * When the CSB registers are reset (also after power-up / gpu reset), - * CSB write pointer is set to all 1's, which is not valid, use '5' in - * this special case, so the first element read is CSB[0]. - */ - if (next_context_status_buffer_hw == GEN8_CSB_PTR_MASK) - next_context_status_buffer_hw = (GEN8_CSB_ENTRIES - 1); - engine->next_context_status_buffer = next_context_status_buffer_hw; DRM_DEBUG_DRIVER("Execlists enabled for %s\n", engine->name); intel_engine_init_hangcheck(engine); - return intel_mocs_init_engine(engine); + if (!execlists_elsp_idle(engine)) + execlists_submit_ports(engine); + + return 0; } static int gen8_init_render_ring(struct intel_engine_cs *engine) @@ -1357,6 +1282,36 @@ static int gen9_init_render_ring(struct intel_engine_cs *engine) return init_workarounds_ring(engine); } +static void reset_common_ring(struct intel_engine_cs *engine, + struct drm_i915_gem_request *request) +{ + struct drm_i915_private *dev_priv = engine->i915; + struct execlist_port *port = engine->execlist_port; + struct intel_context *ce = &request->ctx->engine[engine->id]; + + /* Move the RING_HEAD onto the breadcrumb, past the hanging batch */ + ce->lrc_reg_state[CTX_RING_HEAD+1] = request->postfix; + request->ring->head = request->postfix; + request->ring->last_retired_head = -1; + intel_ring_update_space(request->ring); + + if (i915.enable_guc_submission) + return; + + /* Catch up with any missed context-switch interrupts */ + I915_WRITE(RING_CONTEXT_STATUS_PTR(engine), _MASKED_FIELD(0xffff, 0)); + if (request->ctx != port[0].request->ctx) { + i915_gem_request_put(port[0].request); + port[0] = port[1]; + memset(&port[1], 0, sizeof(port[1])); + } + + /* CS is stopped, and we will resubmit both ports on resume */ + GEM_BUG_ON(request->ctx != port[0].request->ctx); + port[0].count = 0; + port[1].count = 0; +} + static int intel_logical_ring_emit_pdps(struct drm_i915_gem_request *req) { struct i915_hw_ppgtt *ppgtt = req->ctx->ppgtt; @@ -1702,10 +1657,6 @@ void intel_logical_ring_cleanup(struct intel_engine_cs *engine) } intel_lr_context_unpin(dev_priv->kernel_context, engine); - engine->idle_lite_restore_wa = 0; - engine->disable_lite_restore_wa = false; - engine->ctx_desc_template = 0; - lrc_destroy_wa_ctx_obj(engine); engine->i915 = NULL; } @@ -1723,6 +1674,7 @@ logical_ring_default_vfuncs(struct intel_engine_cs *engine) { /* Default vfuncs which can be overriden by each engine. */ engine->init_hw = gen8_init_common_ring; + engine->reset_hw = reset_common_ring; engine->emit_flush = gen8_emit_flush; engine->emit_request = gen8_emit_request; engine->submit_request = execlists_submit_request; @@ -1896,24 +1848,24 @@ make_rpcs(struct drm_i915_private *dev_priv) * must make an explicit request through RPCS for full * enablement. */ - if (INTEL_INFO(dev_priv)->has_slice_pg) { + if (INTEL_INFO(dev_priv)->sseu.has_slice_pg) { rpcs |= GEN8_RPCS_S_CNT_ENABLE; - rpcs |= INTEL_INFO(dev_priv)->slice_total << + rpcs |= hweight8(INTEL_INFO(dev_priv)->sseu.slice_mask) << GEN8_RPCS_S_CNT_SHIFT; rpcs |= GEN8_RPCS_ENABLE; } - if (INTEL_INFO(dev_priv)->has_subslice_pg) { + if (INTEL_INFO(dev_priv)->sseu.has_subslice_pg) { rpcs |= GEN8_RPCS_SS_CNT_ENABLE; - rpcs |= INTEL_INFO(dev_priv)->subslice_per_slice << + rpcs |= hweight8(INTEL_INFO(dev_priv)->sseu.subslice_mask) << GEN8_RPCS_SS_CNT_SHIFT; rpcs |= GEN8_RPCS_ENABLE; } - if (INTEL_INFO(dev_priv)->has_eu_pg) { - rpcs |= INTEL_INFO(dev_priv)->eu_per_subslice << + if (INTEL_INFO(dev_priv)->sseu.has_eu_pg) { + rpcs |= INTEL_INFO(dev_priv)->sseu.eu_per_subslice << GEN8_RPCS_EU_MIN_SHIFT; - rpcs |= INTEL_INFO(dev_priv)->eu_per_subslice << + rpcs |= INTEL_INFO(dev_priv)->sseu.eu_per_subslice << GEN8_RPCS_EU_MAX_SHIFT; rpcs |= GEN8_RPCS_ENABLE; } @@ -2175,9 +2127,9 @@ error_deref_obj: return ret; } -void intel_lr_context_reset(struct drm_i915_private *dev_priv, - struct i915_gem_context *ctx) +void intel_lr_context_resume(struct drm_i915_private *dev_priv) { + struct i915_gem_context *ctx = dev_priv->kernel_context; struct intel_engine_cs *engine; for_each_engine(engine, dev_priv) { diff --git a/drivers/gpu/drm/i915/intel_lrc.h b/drivers/gpu/drm/i915/intel_lrc.h index a52cf57dbd40..4fed8165f98a 100644 --- a/drivers/gpu/drm/i915/intel_lrc.h +++ b/drivers/gpu/drm/i915/intel_lrc.h @@ -87,8 +87,7 @@ void intel_lr_context_unpin(struct i915_gem_context *ctx, struct drm_i915_private; -void intel_lr_context_reset(struct drm_i915_private *dev_priv, - struct i915_gem_context *ctx); +void intel_lr_context_resume(struct drm_i915_private *dev_priv); uint64_t intel_lr_context_descriptor(struct i915_gem_context *ctx, struct intel_engine_cs *engine); @@ -97,6 +96,4 @@ int intel_sanitize_enable_execlists(struct drm_i915_private *dev_priv, int enable_execlists); void intel_execlists_enable_submission(struct drm_i915_private *dev_priv); -void intel_execlists_cancel_requests(struct intel_engine_cs *engine); - #endif /* _INTEL_LRC_H_ */ diff --git a/drivers/gpu/drm/i915/intel_lvds.c b/drivers/gpu/drm/i915/intel_lvds.c index 52d6ed6f6966..e1d47d51ea47 100644 --- a/drivers/gpu/drm/i915/intel_lvds.c +++ b/drivers/gpu/drm/i915/intel_lvds.c @@ -230,20 +230,21 @@ static void intel_lvds_pps_init_hw(struct drm_i915_private *dev_priv, I915_WRITE(PP_DIVISOR(0), val); } -static void intel_pre_enable_lvds(struct intel_encoder *encoder) +static void intel_pre_enable_lvds(struct intel_encoder *encoder, + struct intel_crtc_state *pipe_config, + struct drm_connector_state *conn_state) { struct intel_lvds_encoder *lvds_encoder = to_lvds_encoder(&encoder->base); - struct drm_device *dev = encoder->base.dev; - struct drm_i915_private *dev_priv = to_i915(dev); - struct intel_crtc *crtc = to_intel_crtc(encoder->base.crtc); - const struct drm_display_mode *adjusted_mode = &crtc->config->base.adjusted_mode; + struct drm_i915_private *dev_priv = to_i915(encoder->base.dev); + struct intel_crtc *crtc = to_intel_crtc(pipe_config->base.crtc); + const struct drm_display_mode *adjusted_mode = &pipe_config->base.adjusted_mode; int pipe = crtc->pipe; u32 temp; - if (HAS_PCH_SPLIT(dev)) { + if (HAS_PCH_SPLIT(dev_priv)) { assert_fdi_rx_pll_disabled(dev_priv, pipe); assert_shared_dpll_disabled(dev_priv, - crtc->config->shared_dpll); + pipe_config->shared_dpll); } else { assert_pll_disabled(dev_priv, pipe); } @@ -253,7 +254,7 @@ static void intel_pre_enable_lvds(struct intel_encoder *encoder) temp = lvds_encoder->init_lvds_val; temp |= LVDS_PORT_EN | LVDS_A0A2_CLKA_POWER_UP; - if (HAS_PCH_CPT(dev)) { + if (HAS_PCH_CPT(dev_priv)) { temp &= ~PORT_TRANS_SEL_MASK; temp |= PORT_TRANS_SEL_CPT(pipe); } else { @@ -266,7 +267,7 @@ static void intel_pre_enable_lvds(struct intel_encoder *encoder) /* set the corresponsding LVDS_BORDER bit */ temp &= ~LVDS_BORDER_ENABLE; - temp |= crtc->config->gmch_pfit.lvds_border_bits; + temp |= pipe_config->gmch_pfit.lvds_border_bits; /* Set the B0-B3 data pairs corresponding to whether we're going to * set the DPLLs for dual-channel mode or not. */ @@ -289,7 +290,7 @@ static void intel_pre_enable_lvds(struct intel_encoder *encoder) if (IS_GEN4(dev_priv)) { /* Bspec wording suggests that LVDS port dithering only exists * for 18bpp panels. */ - if (crtc->config->dither && crtc->config->pipe_bpp == 18) + if (pipe_config->dither && pipe_config->pipe_bpp == 18) temp |= LVDS_ENABLE_DITHER; else temp &= ~LVDS_ENABLE_DITHER; @@ -306,7 +307,9 @@ static void intel_pre_enable_lvds(struct intel_encoder *encoder) /** * Sets the power state for the panel. */ -static void intel_enable_lvds(struct intel_encoder *encoder) +static void intel_enable_lvds(struct intel_encoder *encoder, + struct intel_crtc_state *pipe_config, + struct drm_connector_state *conn_state) { struct drm_device *dev = encoder->base.dev; struct intel_lvds_encoder *lvds_encoder = to_lvds_encoder(&encoder->base); @@ -324,11 +327,12 @@ static void intel_enable_lvds(struct intel_encoder *encoder) intel_panel_enable_backlight(intel_connector); } -static void intel_disable_lvds(struct intel_encoder *encoder) +static void intel_disable_lvds(struct intel_encoder *encoder, + struct intel_crtc_state *old_crtc_state, + struct drm_connector_state *old_conn_state) { - struct drm_device *dev = encoder->base.dev; struct intel_lvds_encoder *lvds_encoder = to_lvds_encoder(&encoder->base); - struct drm_i915_private *dev_priv = to_i915(dev); + struct drm_i915_private *dev_priv = to_i915(encoder->base.dev); I915_WRITE(PP_CONTROL(0), I915_READ(PP_CONTROL(0)) & ~PANEL_POWER_ON); if (intel_wait_for_register(dev_priv, PP_STATUS(0), PP_ON, 0, 1000)) @@ -338,7 +342,10 @@ static void intel_disable_lvds(struct intel_encoder *encoder) POSTING_READ(lvds_encoder->reg); } -static void gmch_disable_lvds(struct intel_encoder *encoder) +static void gmch_disable_lvds(struct intel_encoder *encoder, + struct intel_crtc_state *old_crtc_state, + struct drm_connector_state *old_conn_state) + { struct intel_lvds_encoder *lvds_encoder = to_lvds_encoder(&encoder->base); struct intel_connector *intel_connector = @@ -346,10 +353,12 @@ static void gmch_disable_lvds(struct intel_encoder *encoder) intel_panel_disable_backlight(intel_connector); - intel_disable_lvds(encoder); + intel_disable_lvds(encoder, old_crtc_state, old_conn_state); } -static void pch_disable_lvds(struct intel_encoder *encoder) +static void pch_disable_lvds(struct intel_encoder *encoder, + struct intel_crtc_state *old_crtc_state, + struct drm_connector_state *old_conn_state) { struct intel_lvds_encoder *lvds_encoder = to_lvds_encoder(&encoder->base); struct intel_connector *intel_connector = @@ -358,9 +367,11 @@ static void pch_disable_lvds(struct intel_encoder *encoder) intel_panel_disable_backlight(intel_connector); } -static void pch_post_disable_lvds(struct intel_encoder *encoder) +static void pch_post_disable_lvds(struct intel_encoder *encoder, + struct intel_crtc_state *old_crtc_state, + struct drm_connector_state *old_conn_state) { - intel_disable_lvds(encoder); + intel_disable_lvds(encoder, old_crtc_state, old_conn_state); } static enum drm_mode_status @@ -382,7 +393,8 @@ intel_lvds_mode_valid(struct drm_connector *connector, } static bool intel_lvds_compute_config(struct intel_encoder *intel_encoder, - struct intel_crtc_state *pipe_config) + struct intel_crtc_state *pipe_config, + struct drm_connector_state *conn_state) { struct drm_device *dev = intel_encoder->base.dev; struct intel_lvds_encoder *lvds_encoder = diff --git a/drivers/gpu/drm/i915/intel_opregion.c b/drivers/gpu/drm/i915/intel_opregion.c index adca262d591a..7acbbbf97833 100644 --- a/drivers/gpu/drm/i915/intel_opregion.c +++ b/drivers/gpu/drm/i915/intel_opregion.c @@ -1047,6 +1047,23 @@ err_out: return err; } +static int intel_use_opregion_panel_type_callback(const struct dmi_system_id *id) +{ + DRM_INFO("Using panel type from OpRegion on %s\n", id->ident); + return 1; +} + +static const struct dmi_system_id intel_use_opregion_panel_type[] = { + { + .callback = intel_use_opregion_panel_type_callback, + .ident = "Conrac GmbH IX45GM2", + .matches = {DMI_MATCH(DMI_SYS_VENDOR, "Conrac GmbH"), + DMI_MATCH(DMI_PRODUCT_NAME, "IX45GM2"), + }, + }, + { } +}; + int intel_opregion_get_panel_type(struct drm_i915_private *dev_priv) { @@ -1073,6 +1090,16 @@ intel_opregion_get_panel_type(struct drm_i915_private *dev_priv) } /* + * So far we know that some machined must use it, others must not use it. + * There doesn't seem to be any way to determine which way to go, except + * via a quirk list :( + */ + if (!dmi_check_system(intel_use_opregion_panel_type)) { + DRM_DEBUG_KMS("Ignoring OpRegion panel type (%d)\n", ret - 1); + return -ENODEV; + } + + /* * FIXME On Dell XPS 13 9350 the OpRegion panel type (0) gives us * low vswing for eDP, whereas the VBT panel type (2) gives us normal * vswing instead. Low vswing results in some display flickers, so diff --git a/drivers/gpu/drm/i915/intel_panel.c b/drivers/gpu/drm/i915/intel_panel.c index 96c65d77e886..c10e9b0405e8 100644 --- a/drivers/gpu/drm/i915/intel_panel.c +++ b/drivers/gpu/drm/i915/intel_panel.c @@ -1430,10 +1430,11 @@ static int lpt_setup_backlight(struct intel_connector *connector, enum pipe unus panel->backlight.min = get_backlight_min_vbt(connector); val = lpt_get_backlight(connector); - panel->backlight.level = intel_panel_compute_brightness(connector, val); + val = intel_panel_compute_brightness(connector, val); + panel->backlight.level = clamp(val, panel->backlight.min, + panel->backlight.max); - panel->backlight.enabled = (pch_ctl1 & BLM_PCH_PWM_ENABLE) && - panel->backlight.level != 0; + panel->backlight.enabled = pch_ctl1 & BLM_PCH_PWM_ENABLE; return 0; } @@ -1459,11 +1460,13 @@ static int pch_setup_backlight(struct intel_connector *connector, enum pipe unus panel->backlight.min = get_backlight_min_vbt(connector); val = pch_get_backlight(connector); - panel->backlight.level = intel_panel_compute_brightness(connector, val); + val = intel_panel_compute_brightness(connector, val); + panel->backlight.level = clamp(val, panel->backlight.min, + panel->backlight.max); cpu_ctl2 = I915_READ(BLC_PWM_CPU_CTL2); panel->backlight.enabled = (cpu_ctl2 & BLM_PWM_ENABLE) && - (pch_ctl1 & BLM_PCH_PWM_ENABLE) && panel->backlight.level != 0; + (pch_ctl1 & BLM_PCH_PWM_ENABLE); return 0; } @@ -1498,9 +1501,11 @@ static int i9xx_setup_backlight(struct intel_connector *connector, enum pipe unu panel->backlight.min = get_backlight_min_vbt(connector); val = i9xx_get_backlight(connector); - panel->backlight.level = intel_panel_compute_brightness(connector, val); + val = intel_panel_compute_brightness(connector, val); + panel->backlight.level = clamp(val, panel->backlight.min, + panel->backlight.max); - panel->backlight.enabled = panel->backlight.level != 0; + panel->backlight.enabled = val != 0; return 0; } @@ -1530,10 +1535,11 @@ static int i965_setup_backlight(struct intel_connector *connector, enum pipe unu panel->backlight.min = get_backlight_min_vbt(connector); val = i9xx_get_backlight(connector); - panel->backlight.level = intel_panel_compute_brightness(connector, val); + val = intel_panel_compute_brightness(connector, val); + panel->backlight.level = clamp(val, panel->backlight.min, + panel->backlight.max); - panel->backlight.enabled = (ctl2 & BLM_PWM_ENABLE) && - panel->backlight.level != 0; + panel->backlight.enabled = ctl2 & BLM_PWM_ENABLE; return 0; } @@ -1562,10 +1568,11 @@ static int vlv_setup_backlight(struct intel_connector *connector, enum pipe pipe panel->backlight.min = get_backlight_min_vbt(connector); val = _vlv_get_backlight(dev_priv, pipe); - panel->backlight.level = intel_panel_compute_brightness(connector, val); + val = intel_panel_compute_brightness(connector, val); + panel->backlight.level = clamp(val, panel->backlight.min, + panel->backlight.max); - panel->backlight.enabled = (ctl2 & BLM_PWM_ENABLE) && - panel->backlight.level != 0; + panel->backlight.enabled = ctl2 & BLM_PWM_ENABLE; return 0; } @@ -1607,10 +1614,11 @@ bxt_setup_backlight(struct intel_connector *connector, enum pipe unused) return -ENODEV; val = bxt_get_backlight(connector); - panel->backlight.level = intel_panel_compute_brightness(connector, val); + val = intel_panel_compute_brightness(connector, val); + panel->backlight.level = clamp(val, panel->backlight.min, + panel->backlight.max); - panel->backlight.enabled = (pwm_ctl & BXT_BLC_PWM_ENABLE) && - panel->backlight.level != 0; + panel->backlight.enabled = pwm_ctl & BXT_BLC_PWM_ENABLE; return 0; } diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index 8a6751e14ab9..2df06b703e3d 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -2853,13 +2853,7 @@ bool ilk_disable_lp_wm(struct drm_device *dev) return _ilk_disable_lp_wm(dev_priv, WM_DIRTY_LP_ALL); } -/* - * On gen9, we need to allocate Display Data Buffer (DDB) portions to the - * different active planes. - */ - -#define SKL_DDB_SIZE 896 /* in blocks */ -#define BXT_DDB_SIZE 512 +#define SKL_SAGV_BLOCK_TIME 30 /* µs */ /* * Return the index of a plane in the SKL DDB and wm result arrays. Primary @@ -2883,6 +2877,153 @@ skl_wm_plane_id(const struct intel_plane *plane) } } +/* + * SAGV dynamically adjusts the system agent voltage and clock frequencies + * depending on power and performance requirements. The display engine access + * to system memory is blocked during the adjustment time. Because of the + * blocking time, having this enabled can cause full system hangs and/or pipe + * underruns if we don't meet all of the following requirements: + * + * - <= 1 pipe enabled + * - All planes can enable watermarks for latencies >= SAGV engine block time + * - We're not using an interlaced display configuration + */ +int +skl_enable_sagv(struct drm_i915_private *dev_priv) +{ + int ret; + + if (dev_priv->skl_sagv_status == I915_SKL_SAGV_NOT_CONTROLLED || + dev_priv->skl_sagv_status == I915_SKL_SAGV_ENABLED) + return 0; + + DRM_DEBUG_KMS("Enabling the SAGV\n"); + mutex_lock(&dev_priv->rps.hw_lock); + + ret = sandybridge_pcode_write(dev_priv, GEN9_PCODE_SAGV_CONTROL, + GEN9_SAGV_ENABLE); + + /* We don't need to wait for the SAGV when enabling */ + mutex_unlock(&dev_priv->rps.hw_lock); + + /* + * Some skl systems, pre-release machines in particular, + * don't actually have an SAGV. + */ + if (ret == -ENXIO) { + DRM_DEBUG_DRIVER("No SAGV found on system, ignoring\n"); + dev_priv->skl_sagv_status = I915_SKL_SAGV_NOT_CONTROLLED; + return 0; + } else if (ret < 0) { + DRM_ERROR("Failed to enable the SAGV\n"); + return ret; + } + + dev_priv->skl_sagv_status = I915_SKL_SAGV_ENABLED; + return 0; +} + +static int +skl_do_sagv_disable(struct drm_i915_private *dev_priv) +{ + int ret; + uint32_t temp = GEN9_SAGV_DISABLE; + + ret = sandybridge_pcode_read(dev_priv, GEN9_PCODE_SAGV_CONTROL, + &temp); + if (ret) + return ret; + else + return temp & GEN9_SAGV_IS_DISABLED; +} + +int +skl_disable_sagv(struct drm_i915_private *dev_priv) +{ + int ret, result; + + if (dev_priv->skl_sagv_status == I915_SKL_SAGV_NOT_CONTROLLED || + dev_priv->skl_sagv_status == I915_SKL_SAGV_DISABLED) + return 0; + + DRM_DEBUG_KMS("Disabling the SAGV\n"); + mutex_lock(&dev_priv->rps.hw_lock); + + /* bspec says to keep retrying for at least 1 ms */ + ret = wait_for(result = skl_do_sagv_disable(dev_priv), 1); + mutex_unlock(&dev_priv->rps.hw_lock); + + if (ret == -ETIMEDOUT) { + DRM_ERROR("Request to disable SAGV timed out\n"); + return -ETIMEDOUT; + } + + /* + * Some skl systems, pre-release machines in particular, + * don't actually have an SAGV. + */ + if (result == -ENXIO) { + DRM_DEBUG_DRIVER("No SAGV found on system, ignoring\n"); + dev_priv->skl_sagv_status = I915_SKL_SAGV_NOT_CONTROLLED; + return 0; + } else if (result < 0) { + DRM_ERROR("Failed to disable the SAGV\n"); + return result; + } + + dev_priv->skl_sagv_status = I915_SKL_SAGV_DISABLED; + return 0; +} + +bool skl_can_enable_sagv(struct drm_atomic_state *state) +{ + struct drm_device *dev = state->dev; + struct drm_i915_private *dev_priv = to_i915(dev); + struct intel_atomic_state *intel_state = to_intel_atomic_state(state); + struct drm_crtc *crtc; + enum pipe pipe; + int level, plane; + + /* + * SKL workaround: bspec recommends we disable the SAGV when we have + * more then one pipe enabled + * + * If there are no active CRTCs, no additional checks need be performed + */ + if (hweight32(intel_state->active_crtcs) == 0) + return true; + else if (hweight32(intel_state->active_crtcs) > 1) + return false; + + /* Since we're now guaranteed to only have one active CRTC... */ + pipe = ffs(intel_state->active_crtcs) - 1; + crtc = dev_priv->pipe_to_crtc_mapping[pipe]; + + if (crtc->state->mode.flags & DRM_MODE_FLAG_INTERLACE) + return false; + + for_each_plane(dev_priv, pipe, plane) { + /* Skip this plane if it's not enabled */ + if (intel_state->wm_results.plane[pipe][plane][0] == 0) + continue; + + /* Find the highest enabled wm level for this plane */ + for (level = ilk_wm_max_level(dev); + intel_state->wm_results.plane[pipe][plane][level] == 0; --level) + { } + + /* + * If any of the planes on this pipe don't enable wm levels + * that incur memory latencies higher then 30µs we can't enable + * the SAGV + */ + if (dev_priv->wm.skl_latency[level] < SKL_SAGV_BLOCK_TIME) + return false; + } + + return true; +} + static void skl_ddb_get_pipe_allocation_limits(struct drm_device *dev, const struct intel_crtc_state *cstate, @@ -2909,10 +3050,8 @@ skl_ddb_get_pipe_allocation_limits(struct drm_device *dev, else *num_active = hweight32(dev_priv->active_crtcs); - if (IS_BROXTON(dev)) - ddb_size = BXT_DDB_SIZE; - else - ddb_size = SKL_DDB_SIZE; + ddb_size = INTEL_INFO(dev_priv)->ddb_size; + WARN_ON(ddb_size == 0); ddb_size -= 4; /* 4 blocks for bypass path allocation */ @@ -3688,183 +3827,82 @@ static void skl_ddb_entry_write(struct drm_i915_private *dev_priv, I915_WRITE(reg, 0); } -static void skl_write_wm_values(struct drm_i915_private *dev_priv, - const struct skl_wm_values *new) +void skl_write_plane_wm(struct intel_crtc *intel_crtc, + const struct skl_wm_values *wm, + int plane) { - struct drm_device *dev = &dev_priv->drm; - struct intel_crtc *crtc; - - for_each_intel_crtc(dev, crtc) { - int i, level, max_level = ilk_wm_max_level(dev); - enum pipe pipe = crtc->pipe; - - if ((new->dirty_pipes & drm_crtc_mask(&crtc->base)) == 0) - continue; - if (!crtc->active) - continue; - - I915_WRITE(PIPE_WM_LINETIME(pipe), new->wm_linetime[pipe]); - - for (level = 0; level <= max_level; level++) { - for (i = 0; i < intel_num_planes(crtc); i++) - I915_WRITE(PLANE_WM(pipe, i, level), - new->plane[pipe][i][level]); - I915_WRITE(CUR_WM(pipe, level), - new->plane[pipe][PLANE_CURSOR][level]); - } - for (i = 0; i < intel_num_planes(crtc); i++) - I915_WRITE(PLANE_WM_TRANS(pipe, i), - new->plane_trans[pipe][i]); - I915_WRITE(CUR_WM_TRANS(pipe), - new->plane_trans[pipe][PLANE_CURSOR]); - - for (i = 0; i < intel_num_planes(crtc); i++) { - skl_ddb_entry_write(dev_priv, - PLANE_BUF_CFG(pipe, i), - &new->ddb.plane[pipe][i]); - skl_ddb_entry_write(dev_priv, - PLANE_NV12_BUF_CFG(pipe, i), - &new->ddb.y_plane[pipe][i]); - } + struct drm_crtc *crtc = &intel_crtc->base; + struct drm_device *dev = crtc->dev; + struct drm_i915_private *dev_priv = to_i915(dev); + int level, max_level = ilk_wm_max_level(dev); + enum pipe pipe = intel_crtc->pipe; - skl_ddb_entry_write(dev_priv, CUR_BUF_CFG(pipe), - &new->ddb.plane[pipe][PLANE_CURSOR]); + for (level = 0; level <= max_level; level++) { + I915_WRITE(PLANE_WM(pipe, plane, level), + wm->plane[pipe][plane][level]); } -} + I915_WRITE(PLANE_WM_TRANS(pipe, plane), wm->plane_trans[pipe][plane]); -/* - * When setting up a new DDB allocation arrangement, we need to correctly - * sequence the times at which the new allocations for the pipes are taken into - * account or we'll have pipes fetching from space previously allocated to - * another pipe. - * - * Roughly the sequence looks like: - * 1. re-allocate the pipe(s) with the allocation being reduced and not - * overlapping with a previous light-up pipe (another way to put it is: - * pipes with their new allocation strickly included into their old ones). - * 2. re-allocate the other pipes that get their allocation reduced - * 3. allocate the pipes having their allocation increased - * - * Steps 1. and 2. are here to take care of the following case: - * - Initially DDB looks like this: - * | B | C | - * - enable pipe A. - * - pipe B has a reduced DDB allocation that overlaps with the old pipe C - * allocation - * | A | B | C | - * - * We need to sequence the re-allocation: C, B, A (and not B, C, A). - */ + skl_ddb_entry_write(dev_priv, PLANE_BUF_CFG(pipe, plane), + &wm->ddb.plane[pipe][plane]); + skl_ddb_entry_write(dev_priv, PLANE_NV12_BUF_CFG(pipe, plane), + &wm->ddb.y_plane[pipe][plane]); +} -static void -skl_wm_flush_pipe(struct drm_i915_private *dev_priv, enum pipe pipe, int pass) +void skl_write_cursor_wm(struct intel_crtc *intel_crtc, + const struct skl_wm_values *wm) { - int plane; - - DRM_DEBUG_KMS("flush pipe %c (pass %d)\n", pipe_name(pipe), pass); + struct drm_crtc *crtc = &intel_crtc->base; + struct drm_device *dev = crtc->dev; + struct drm_i915_private *dev_priv = to_i915(dev); + int level, max_level = ilk_wm_max_level(dev); + enum pipe pipe = intel_crtc->pipe; - for_each_plane(dev_priv, pipe, plane) { - I915_WRITE(PLANE_SURF(pipe, plane), - I915_READ(PLANE_SURF(pipe, plane))); + for (level = 0; level <= max_level; level++) { + I915_WRITE(CUR_WM(pipe, level), + wm->plane[pipe][PLANE_CURSOR][level]); } - I915_WRITE(CURBASE(pipe), I915_READ(CURBASE(pipe))); + I915_WRITE(CUR_WM_TRANS(pipe), wm->plane_trans[pipe][PLANE_CURSOR]); + + skl_ddb_entry_write(dev_priv, CUR_BUF_CFG(pipe), + &wm->ddb.plane[pipe][PLANE_CURSOR]); } -static bool -skl_ddb_allocation_included(const struct skl_ddb_allocation *old, - const struct skl_ddb_allocation *new, - enum pipe pipe) +bool skl_ddb_allocation_equals(const struct skl_ddb_allocation *old, + const struct skl_ddb_allocation *new, + enum pipe pipe) { - uint16_t old_size, new_size; - - old_size = skl_ddb_entry_size(&old->pipe[pipe]); - new_size = skl_ddb_entry_size(&new->pipe[pipe]); - - return old_size != new_size && - new->pipe[pipe].start >= old->pipe[pipe].start && - new->pipe[pipe].end <= old->pipe[pipe].end; + return new->pipe[pipe].start == old->pipe[pipe].start && + new->pipe[pipe].end == old->pipe[pipe].end; } -static void skl_flush_wm_values(struct drm_i915_private *dev_priv, - struct skl_wm_values *new_values) +static inline bool skl_ddb_entries_overlap(const struct skl_ddb_entry *a, + const struct skl_ddb_entry *b) { - struct drm_device *dev = &dev_priv->drm; - struct skl_ddb_allocation *cur_ddb, *new_ddb; - bool reallocated[I915_MAX_PIPES] = {}; - struct intel_crtc *crtc; - enum pipe pipe; - - new_ddb = &new_values->ddb; - cur_ddb = &dev_priv->wm.skl_hw.ddb; - - /* - * First pass: flush the pipes with the new allocation contained into - * the old space. - * - * We'll wait for the vblank on those pipes to ensure we can safely - * re-allocate the freed space without this pipe fetching from it. - */ - for_each_intel_crtc(dev, crtc) { - if (!crtc->active) - continue; - - pipe = crtc->pipe; - - if (!skl_ddb_allocation_included(cur_ddb, new_ddb, pipe)) - continue; - - skl_wm_flush_pipe(dev_priv, pipe, 1); - intel_wait_for_vblank(dev, pipe); - - reallocated[pipe] = true; - } - + return a->start < b->end && b->start < a->end; +} - /* - * Second pass: flush the pipes that are having their allocation - * reduced, but overlapping with a previous allocation. - * - * Here as well we need to wait for the vblank to make sure the freed - * space is not used anymore. - */ - for_each_intel_crtc(dev, crtc) { - if (!crtc->active) - continue; +bool skl_ddb_allocation_overlaps(struct drm_atomic_state *state, + const struct skl_ddb_allocation *old, + const struct skl_ddb_allocation *new, + enum pipe pipe) +{ + struct drm_device *dev = state->dev; + struct intel_crtc *intel_crtc; + enum pipe otherp; - pipe = crtc->pipe; + for_each_intel_crtc(dev, intel_crtc) { + otherp = intel_crtc->pipe; - if (reallocated[pipe]) + if (otherp == pipe) continue; - if (skl_ddb_entry_size(&new_ddb->pipe[pipe]) < - skl_ddb_entry_size(&cur_ddb->pipe[pipe])) { - skl_wm_flush_pipe(dev_priv, pipe, 2); - intel_wait_for_vblank(dev, pipe); - reallocated[pipe] = true; - } + if (skl_ddb_entries_overlap(&new->pipe[pipe], + &old->pipe[otherp])) + return true; } - /* - * Third pass: flush the pipes that got more space allocated. - * - * We don't need to actively wait for the update here, next vblank - * will just get more DDB space with the correct WM values. - */ - for_each_intel_crtc(dev, crtc) { - if (!crtc->active) - continue; - - pipe = crtc->pipe; - - /* - * At this point, only the pipes more space than before are - * left to re-allocate. - */ - if (reallocated[pipe]) - continue; - - skl_wm_flush_pipe(dev_priv, pipe, 3); - } + return false; } static int skl_update_pipe_wm(struct drm_crtc_state *cstate, @@ -3964,11 +4002,33 @@ skl_compute_ddb(struct drm_atomic_state *state) ret = skl_allocate_pipe_ddb(cstate, ddb); if (ret) return ret; + + ret = drm_atomic_add_affected_planes(state, &intel_crtc->base); + if (ret) + return ret; } return 0; } +static void +skl_copy_wm_for_pipe(struct skl_wm_values *dst, + struct skl_wm_values *src, + enum pipe pipe) +{ + dst->wm_linetime[pipe] = src->wm_linetime[pipe]; + memcpy(dst->plane[pipe], src->plane[pipe], + sizeof(dst->plane[pipe])); + memcpy(dst->plane_trans[pipe], src->plane_trans[pipe], + sizeof(dst->plane_trans[pipe])); + + dst->ddb.pipe[pipe] = src->ddb.pipe[pipe]; + memcpy(dst->ddb.y_plane[pipe], src->ddb.y_plane[pipe], + sizeof(dst->ddb.y_plane[pipe])); + memcpy(dst->ddb.plane[pipe], src->ddb.plane[pipe], + sizeof(dst->ddb.plane[pipe])); +} + static int skl_compute_wm(struct drm_atomic_state *state) { @@ -4041,8 +4101,10 @@ static void skl_update_wm(struct drm_crtc *crtc) struct drm_device *dev = crtc->dev; struct drm_i915_private *dev_priv = to_i915(dev); struct skl_wm_values *results = &dev_priv->wm.skl_results; + struct skl_wm_values *hw_vals = &dev_priv->wm.skl_hw; struct intel_crtc_state *cstate = to_intel_crtc_state(crtc->state); struct skl_pipe_wm *pipe_wm = &cstate->wm.skl.optimal; + enum pipe pipe = intel_crtc->pipe; if ((results->dirty_pipes & drm_crtc_mask(crtc)) == 0) return; @@ -4051,11 +4113,22 @@ static void skl_update_wm(struct drm_crtc *crtc) mutex_lock(&dev_priv->wm.wm_mutex); - skl_write_wm_values(dev_priv, results); - skl_flush_wm_values(dev_priv, results); + /* + * If this pipe isn't active already, we're going to be enabling it + * very soon. Since it's safe to update a pipe's ddb allocation while + * the pipe's shut off, just do so here. Already active pipes will have + * their watermarks updated once we update their planes. + */ + if (crtc->state->active_changed) { + int plane; + + for (plane = 0; plane < intel_num_planes(intel_crtc); plane++) + skl_write_plane_wm(intel_crtc, results, plane); - /* store the new configuration */ - dev_priv->wm.skl_hw = *results; + skl_write_cursor_wm(intel_crtc, results); + } + + skl_copy_wm_for_pipe(hw_vals, results, pipe); mutex_unlock(&dev_priv->wm.wm_mutex); } @@ -5550,7 +5623,7 @@ static int cherryview_rps_max_freq(struct drm_i915_private *dev_priv) val = vlv_punit_read(dev_priv, FB_GFX_FMAX_AT_VMAX_FUSE); - switch (INTEL_INFO(dev_priv)->eu_total) { + switch (INTEL_INFO(dev_priv)->sseu.eu_total) { case 8: /* (2 * 4) config */ rp0 = (val >> FB_GFX_FMAX_AT_VMAX_2SS4EU_FUSE_SHIFT); @@ -6691,9 +6764,7 @@ void intel_autoenable_gt_powersave(struct drm_i915_private *dev_priv) if (IS_IRONLAKE_M(dev_priv)) { ironlake_enable_drps(dev_priv); - mutex_lock(&dev_priv->drm.struct_mutex); intel_init_emon(dev_priv); - mutex_unlock(&dev_priv->drm.struct_mutex); } else if (INTEL_INFO(dev_priv)->gen >= 6) { /* * PCU communication is slow and this doesn't need to be @@ -7659,8 +7730,54 @@ void intel_init_pm(struct drm_device *dev) } } +static inline int gen6_check_mailbox_status(struct drm_i915_private *dev_priv) +{ + uint32_t flags = + I915_READ_FW(GEN6_PCODE_MAILBOX) & GEN6_PCODE_ERROR_MASK; + + switch (flags) { + case GEN6_PCODE_SUCCESS: + return 0; + case GEN6_PCODE_UNIMPLEMENTED_CMD: + case GEN6_PCODE_ILLEGAL_CMD: + return -ENXIO; + case GEN6_PCODE_MIN_FREQ_TABLE_GT_RATIO_OUT_OF_RANGE: + case GEN7_PCODE_MIN_FREQ_TABLE_GT_RATIO_OUT_OF_RANGE: + return -EOVERFLOW; + case GEN6_PCODE_TIMEOUT: + return -ETIMEDOUT; + default: + MISSING_CASE(flags) + return 0; + } +} + +static inline int gen7_check_mailbox_status(struct drm_i915_private *dev_priv) +{ + uint32_t flags = + I915_READ_FW(GEN6_PCODE_MAILBOX) & GEN6_PCODE_ERROR_MASK; + + switch (flags) { + case GEN6_PCODE_SUCCESS: + return 0; + case GEN6_PCODE_ILLEGAL_CMD: + return -ENXIO; + case GEN7_PCODE_TIMEOUT: + return -ETIMEDOUT; + case GEN7_PCODE_ILLEGAL_DATA: + return -EINVAL; + case GEN7_PCODE_MIN_FREQ_TABLE_GT_RATIO_OUT_OF_RANGE: + return -EOVERFLOW; + default: + MISSING_CASE(flags); + return 0; + } +} + int sandybridge_pcode_read(struct drm_i915_private *dev_priv, u32 mbox, u32 *val) { + int status; + WARN_ON(!mutex_is_locked(&dev_priv->rps.hw_lock)); /* GEN6_PCODE_* are outside of the forcewake domain, we can @@ -7687,12 +7804,25 @@ int sandybridge_pcode_read(struct drm_i915_private *dev_priv, u32 mbox, u32 *val *val = I915_READ_FW(GEN6_PCODE_DATA); I915_WRITE_FW(GEN6_PCODE_DATA, 0); + if (INTEL_GEN(dev_priv) > 6) + status = gen7_check_mailbox_status(dev_priv); + else + status = gen6_check_mailbox_status(dev_priv); + + if (status) { + DRM_DEBUG_DRIVER("warning: pcode (read) mailbox access failed: %d\n", + status); + return status; + } + return 0; } int sandybridge_pcode_write(struct drm_i915_private *dev_priv, - u32 mbox, u32 val) + u32 mbox, u32 val) { + int status; + WARN_ON(!mutex_is_locked(&dev_priv->rps.hw_lock)); /* GEN6_PCODE_* are outside of the forcewake domain, we can @@ -7717,6 +7847,17 @@ int sandybridge_pcode_write(struct drm_i915_private *dev_priv, I915_WRITE_FW(GEN6_PCODE_DATA, 0); + if (INTEL_GEN(dev_priv) > 6) + status = gen7_check_mailbox_status(dev_priv); + else + status = gen6_check_mailbox_status(dev_priv); + + if (status) { + DRM_DEBUG_DRIVER("warning: pcode (write) mailbox access failed: %d\n", + status); + return status; + } + return 0; } diff --git a/drivers/gpu/drm/i915/intel_psr.c b/drivers/gpu/drm/i915/intel_psr.c index 59a21c9d2e43..108ba1e5d658 100644 --- a/drivers/gpu/drm/i915/intel_psr.c +++ b/drivers/gpu/drm/i915/intel_psr.c @@ -255,14 +255,14 @@ static void hsw_psr_enable_source(struct intel_dp *intel_dp) struct drm_i915_private *dev_priv = to_i915(dev); uint32_t max_sleep_time = 0x1f; - /* Lately it was identified that depending on panel idle frame count - * calculated at HW can be off by 1. So let's use what came - * from VBT + 1. - * There are also other cases where panel demands at least 4 - * but VBT is not being set. To cover these 2 cases lets use - * at least 5 when VBT isn't set to be on the safest side. + /* + * Let's respect VBT in case VBT asks a higher idle_frame value. + * Let's use 6 as the minimum to cover all known cases including + * the off-by-one issue that HW has in some cases. Also there are + * cases where sink should be able to train + * with the 5 or 6 idle patterns. */ - uint32_t idle_frames = dev_priv->vbt.psr.idle_frames + 1; + uint32_t idle_frames = max(6, dev_priv->vbt.psr.idle_frames); uint32_t val = EDP_PSR_ENABLE; val |= max_sleep_time << EDP_PSR_MAX_SLEEP_TIME_SHIFT; diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c index cc5bcd14b6df..7a74750076c5 100644 --- a/drivers/gpu/drm/i915/intel_ringbuffer.c +++ b/drivers/gpu/drm/i915/intel_ringbuffer.c @@ -559,10 +559,12 @@ static int init_ring_common(struct intel_engine_cs *engine) } } - if (I915_NEED_GFX_HWS(dev_priv)) - intel_ring_setup_status_page(engine); - else + if (HWS_NEEDS_PHYSICAL(dev_priv)) ring_setup_phys_status_page(engine); + else + intel_ring_setup_status_page(engine); + + intel_engine_reset_irq(engine); /* Enforce ordering by reading HEAD register back */ I915_READ_HEAD(engine); @@ -577,34 +579,33 @@ static int init_ring_common(struct intel_engine_cs *engine) if (I915_READ_HEAD(engine)) DRM_DEBUG("%s initialization failed [head=%08x], fudging\n", engine->name, I915_READ_HEAD(engine)); - I915_WRITE_HEAD(engine, 0); - (void)I915_READ_HEAD(engine); + + intel_ring_update_space(ring); + I915_WRITE_HEAD(engine, ring->head); + I915_WRITE_TAIL(engine, ring->tail); + (void)I915_READ_TAIL(engine); I915_WRITE_CTL(engine, ((ring->size - PAGE_SIZE) & RING_NR_PAGES) | RING_VALID); /* If the head is still not zero, the ring is dead */ - if (wait_for((I915_READ_CTL(engine) & RING_VALID) != 0 && - I915_READ_START(engine) == i915_ggtt_offset(ring->vma) && - (I915_READ_HEAD(engine) & HEAD_ADDR) == 0, 50)) { + if (intel_wait_for_register_fw(dev_priv, RING_CTL(engine->mmio_base), + RING_VALID, RING_VALID, + 50)) { DRM_ERROR("%s initialization failed " - "ctl %08x (valid? %d) head %08x tail %08x start %08x [expected %08x]\n", + "ctl %08x (valid? %d) head %08x [%08x] tail %08x [%08x] start %08x [expected %08x]\n", engine->name, I915_READ_CTL(engine), I915_READ_CTL(engine) & RING_VALID, - I915_READ_HEAD(engine), I915_READ_TAIL(engine), + I915_READ_HEAD(engine), ring->head, + I915_READ_TAIL(engine), ring->tail, I915_READ_START(engine), i915_ggtt_offset(ring->vma)); ret = -EIO; goto out; } - ring->last_retired_head = -1; - ring->head = I915_READ_HEAD(engine); - ring->tail = I915_READ_TAIL(engine) & TAIL_ADDR; - intel_ring_update_space(ring); - intel_engine_init_hangcheck(engine); out: @@ -613,6 +614,15 @@ out: return ret; } +static void reset_ring_common(struct intel_engine_cs *engine, + struct drm_i915_gem_request *request) +{ + struct intel_ring *ring = request->ring; + + ring->head = request->postfix; + ring->last_retired_head = -1; +} + static int intel_ring_workarounds_emit(struct drm_i915_gem_request *req) { struct intel_ring *ring = req->ring; @@ -951,7 +961,7 @@ static int skl_tune_iz_hashing(struct intel_engine_cs *engine) * Only consider slices where one, and only one, subslice has 7 * EUs */ - if (!is_power_of_2(dev_priv->info.subslice_7eu[i])) + if (!is_power_of_2(INTEL_INFO(dev_priv)->sseu.subslice_7eu[i])) continue; /* @@ -960,7 +970,7 @@ static int skl_tune_iz_hashing(struct intel_engine_cs *engine) * * -> 0 <= ss <= 3; */ - ss = ffs(dev_priv->info.subslice_7eu[i]) - 1; + ss = ffs(INTEL_INFO(dev_priv)->sseu.subslice_7eu[i]) - 1; vals[i] = 3 - ss; } @@ -2007,7 +2017,6 @@ intel_engine_create_ring(struct intel_engine_cs *engine, int size) } ring->vma = vma; - list_add(&ring->link, &engine->buffers); return ring; } @@ -2015,7 +2024,6 @@ void intel_ring_free(struct intel_ring *ring) { i915_vma_put(ring->vma); - list_del(&ring->link); kfree(ring); } @@ -2109,13 +2117,13 @@ static int intel_init_ring_buffer(struct intel_engine_cs *engine) goto error; } - if (I915_NEED_GFX_HWS(dev_priv)) { - ret = init_status_page(engine); + if (HWS_NEEDS_PHYSICAL(dev_priv)) { + WARN_ON(engine->id != RCS); + ret = init_phys_status_page(engine); if (ret) goto error; } else { - WARN_ON(engine->id != RCS); - ret = init_phys_status_page(engine); + ret = init_status_page(engine); if (ret) goto error; } @@ -2155,11 +2163,11 @@ void intel_engine_cleanup(struct intel_engine_cs *engine) if (engine->cleanup) engine->cleanup(engine); - if (I915_NEED_GFX_HWS(dev_priv)) { - cleanup_status_page(engine); - } else { + if (HWS_NEEDS_PHYSICAL(dev_priv)) { WARN_ON(engine->id != RCS); cleanup_phys_status_page(engine); + } else { + cleanup_status_page(engine); } intel_engine_cleanup_common(engine); @@ -2169,6 +2177,16 @@ void intel_engine_cleanup(struct intel_engine_cs *engine) engine->i915 = NULL; } +void intel_legacy_submission_resume(struct drm_i915_private *dev_priv) +{ + struct intel_engine_cs *engine; + + for_each_engine(engine, dev_priv) { + engine->buffer->head = engine->buffer->tail; + engine->buffer->last_retired_head = -1; + } +} + int intel_ring_alloc_request_extras(struct drm_i915_gem_request *request) { int ret; @@ -2223,13 +2241,12 @@ static int wait_for_space(struct drm_i915_gem_request *req, int bytes) if (WARN_ON(&target->ring_link == &ring->request_list)) return -ENOSPC; - ret = i915_wait_request(target, true, NULL, NO_WAITBOOST); + ret = i915_wait_request(target, + I915_WAIT_INTERRUPTIBLE | I915_WAIT_LOCKED, + NULL, NO_WAITBOOST); if (ret) return ret; - if (i915_reset_in_progress(&target->i915->gpu_error)) - return -EAGAIN; - i915_gem_request_retire_upto(target); intel_ring_update_space(ring); @@ -2655,6 +2672,7 @@ static void intel_ring_default_vfuncs(struct drm_i915_private *dev_priv, intel_ring_init_semaphores(dev_priv, engine); engine->init_hw = init_ring_common; + engine->reset_hw = reset_ring_common; engine->emit_request = i9xx_emit_request; if (i915.semaphores) diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.h b/drivers/gpu/drm/i915/intel_ringbuffer.h index 84aea549de5d..7f64d611159b 100644 --- a/drivers/gpu/drm/i915/intel_ringbuffer.h +++ b/drivers/gpu/drm/i915/intel_ringbuffer.h @@ -87,7 +87,6 @@ struct intel_ring { void *vaddr; struct intel_engine_cs *engine; - struct list_head link; struct list_head request_list; @@ -157,7 +156,6 @@ struct intel_engine_cs { u32 mmio_base; unsigned int irq_shift; struct intel_ring *buffer; - struct list_head buffers; /* Rather than have every client wait upon all user interrupts, * with the herd waking after every interrupt and each doing the @@ -211,6 +209,8 @@ struct intel_engine_cs { void (*irq_disable)(struct intel_engine_cs *engine); int (*init_hw)(struct intel_engine_cs *engine); + void (*reset_hw)(struct intel_engine_cs *engine, + struct drm_i915_gem_request *req); int (*init_context)(struct drm_i915_gem_request *req); @@ -226,7 +226,15 @@ struct intel_engine_cs { #define I915_DISPATCH_PINNED BIT(1) #define I915_DISPATCH_RS BIT(2) int (*emit_request)(struct drm_i915_gem_request *req); + + /* Pass the request to the hardware queue (e.g. directly into + * the legacy ringbuffer or to the end of an execlist). + * + * This is called from an atomic context with irqs disabled; must + * be irq safe. + */ void (*submit_request)(struct drm_i915_gem_request *req); + /* Some chipsets are not quite as coherent as advertised and need * an expensive kick to force a true read of the up-to-date seqno. * However, the up-to-date seqno is not always required and the last @@ -298,11 +306,14 @@ struct intel_engine_cs { /* Execlists */ struct tasklet_struct irq_tasklet; spinlock_t execlist_lock; /* used inside tasklet, use spin_lock_bh */ + struct execlist_port { + struct drm_i915_gem_request *request; + unsigned int count; + } execlist_port[2]; struct list_head execlist_queue; unsigned int fw_domains; - unsigned int next_context_status_buffer; - unsigned int idle_lite_restore_wa; bool disable_lite_restore_wa; + bool preempt_wa; u32 ctx_desc_template; /** @@ -441,6 +452,8 @@ void intel_ring_free(struct intel_ring *ring); void intel_engine_stop(struct intel_engine_cs *engine); void intel_engine_cleanup(struct intel_engine_cs *engine); +void intel_legacy_submission_resume(struct drm_i915_private *dev_priv); + int intel_ring_alloc_request_extras(struct drm_i915_gem_request *request); int __must_check intel_ring_begin(struct drm_i915_gem_request *req, int n); @@ -479,6 +492,7 @@ int __intel_ring_space(int head, int tail, int size); void intel_ring_update_space(struct intel_ring *ring); void intel_engine_init_seqno(struct intel_engine_cs *engine, u32 seqno); +void intel_engine_reset_irq(struct intel_engine_cs *engine); void intel_engine_setup_common(struct intel_engine_cs *engine); int intel_engine_init_common(struct intel_engine_cs *engine); @@ -486,11 +500,11 @@ int intel_engine_create_scratch(struct intel_engine_cs *engine, int size); void intel_engine_cleanup_common(struct intel_engine_cs *engine); static inline int intel_engine_idle(struct intel_engine_cs *engine, - bool interruptible) + unsigned int flags) { /* Wait upon the last request to be completed */ return i915_gem_active_wait_unlocked(&engine->last_request, - interruptible, NULL, NULL); + flags, NULL, NULL); } int intel_init_render_ring_buffer(struct intel_engine_cs *engine); diff --git a/drivers/gpu/drm/i915/intel_runtime_pm.c b/drivers/gpu/drm/i915/intel_runtime_pm.c index a1d73c2de332..6c11168facd6 100644 --- a/drivers/gpu/drm/i915/intel_runtime_pm.c +++ b/drivers/gpu/drm/i915/intel_runtime_pm.c @@ -287,6 +287,7 @@ void intel_display_set_init_power(struct drm_i915_private *dev_priv, */ static void hsw_power_well_post_enable(struct drm_i915_private *dev_priv) { + struct pci_dev *pdev = dev_priv->drm.pdev; struct drm_device *dev = &dev_priv->drm; /* @@ -299,9 +300,9 @@ static void hsw_power_well_post_enable(struct drm_i915_private *dev_priv) * sure vgacon can keep working normally without triggering interrupts * and error messages. */ - vga_get_uninterruptible(dev->pdev, VGA_RSRC_LEGACY_IO); + vga_get_uninterruptible(pdev, VGA_RSRC_LEGACY_IO); outb(inb(VGA_MSR_READ), VGA_MSR_WRITE); - vga_put(dev->pdev, VGA_RSRC_LEGACY_IO); + vga_put(pdev, VGA_RSRC_LEGACY_IO); if (IS_BROADWELL(dev)) gen8_irq_power_well_post_enable(dev_priv, @@ -318,7 +319,7 @@ static void hsw_power_well_pre_disable(struct drm_i915_private *dev_priv) static void skl_power_well_post_enable(struct drm_i915_private *dev_priv, struct i915_power_well *power_well) { - struct drm_device *dev = &dev_priv->drm; + struct pci_dev *pdev = dev_priv->drm.pdev; /* * After we re-enable the power well, if we touch VGA register 0x3d5 @@ -331,9 +332,9 @@ static void skl_power_well_post_enable(struct drm_i915_private *dev_priv, * and error messages. */ if (power_well->data == SKL_DISP_PW_2) { - vga_get_uninterruptible(dev->pdev, VGA_RSRC_LEGACY_IO); + vga_get_uninterruptible(pdev, VGA_RSRC_LEGACY_IO); outb(inb(VGA_MSR_READ), VGA_MSR_WRITE); - vga_put(dev->pdev, VGA_RSRC_LEGACY_IO); + vga_put(pdev, VGA_RSRC_LEGACY_IO); gen8_irq_power_well_post_enable(dev_priv, 1 << PIPE_C | 1 << PIPE_B); @@ -2288,7 +2289,7 @@ int intel_power_domains_init(struct drm_i915_private *dev_priv) */ void intel_power_domains_fini(struct drm_i915_private *dev_priv) { - struct device *device = &dev_priv->drm.pdev->dev; + struct device *kdev = &dev_priv->drm.pdev->dev; /* * The i915.ko module is still not prepared to be loaded when @@ -2310,7 +2311,7 @@ void intel_power_domains_fini(struct drm_i915_private *dev_priv) * the platform doesn't support runtime PM. */ if (!HAS_RUNTIME_PM(dev_priv)) - pm_runtime_put(device); + pm_runtime_put(kdev); } static void intel_power_domains_sync_hw(struct drm_i915_private *dev_priv) @@ -2651,10 +2652,10 @@ void intel_power_domains_suspend(struct drm_i915_private *dev_priv) */ void intel_runtime_pm_get(struct drm_i915_private *dev_priv) { - struct drm_device *dev = &dev_priv->drm; - struct device *device = &dev->pdev->dev; + struct pci_dev *pdev = dev_priv->drm.pdev; + struct device *kdev = &pdev->dev; - pm_runtime_get_sync(device); + pm_runtime_get_sync(kdev); atomic_inc(&dev_priv->pm.wakeref_count); assert_rpm_wakelock_held(dev_priv); @@ -2672,11 +2673,11 @@ void intel_runtime_pm_get(struct drm_i915_private *dev_priv) */ bool intel_runtime_pm_get_if_in_use(struct drm_i915_private *dev_priv) { - struct drm_device *dev = &dev_priv->drm; - struct device *device = &dev->pdev->dev; + struct pci_dev *pdev = dev_priv->drm.pdev; + struct device *kdev = &pdev->dev; if (IS_ENABLED(CONFIG_PM)) { - int ret = pm_runtime_get_if_in_use(device); + int ret = pm_runtime_get_if_in_use(kdev); /* * In cases runtime PM is disabled by the RPM core and we get @@ -2714,11 +2715,11 @@ bool intel_runtime_pm_get_if_in_use(struct drm_i915_private *dev_priv) */ void intel_runtime_pm_get_noresume(struct drm_i915_private *dev_priv) { - struct drm_device *dev = &dev_priv->drm; - struct device *device = &dev->pdev->dev; + struct pci_dev *pdev = dev_priv->drm.pdev; + struct device *kdev = &pdev->dev; assert_rpm_wakelock_held(dev_priv); - pm_runtime_get_noresume(device); + pm_runtime_get_noresume(kdev); atomic_inc(&dev_priv->pm.wakeref_count); } @@ -2733,15 +2734,15 @@ void intel_runtime_pm_get_noresume(struct drm_i915_private *dev_priv) */ void intel_runtime_pm_put(struct drm_i915_private *dev_priv) { - struct drm_device *dev = &dev_priv->drm; - struct device *device = &dev->pdev->dev; + struct pci_dev *pdev = dev_priv->drm.pdev; + struct device *kdev = &pdev->dev; assert_rpm_wakelock_held(dev_priv); if (atomic_dec_and_test(&dev_priv->pm.wakeref_count)) atomic_inc(&dev_priv->pm.atomic_seq); - pm_runtime_mark_last_busy(device); - pm_runtime_put_autosuspend(device); + pm_runtime_mark_last_busy(kdev); + pm_runtime_put_autosuspend(kdev); } /** @@ -2756,11 +2757,12 @@ void intel_runtime_pm_put(struct drm_i915_private *dev_priv) */ void intel_runtime_pm_enable(struct drm_i915_private *dev_priv) { + struct pci_dev *pdev = dev_priv->drm.pdev; struct drm_device *dev = &dev_priv->drm; - struct device *device = &dev->pdev->dev; + struct device *kdev = &pdev->dev; - pm_runtime_set_autosuspend_delay(device, 10000); /* 10s */ - pm_runtime_mark_last_busy(device); + pm_runtime_set_autosuspend_delay(kdev, 10000); /* 10s */ + pm_runtime_mark_last_busy(kdev); /* * Take a permanent reference to disable the RPM functionality and drop @@ -2769,10 +2771,10 @@ void intel_runtime_pm_enable(struct drm_i915_private *dev_priv) * platforms without RPM support. */ if (!HAS_RUNTIME_PM(dev)) { - pm_runtime_dont_use_autosuspend(device); - pm_runtime_get_sync(device); + pm_runtime_dont_use_autosuspend(kdev); + pm_runtime_get_sync(kdev); } else { - pm_runtime_use_autosuspend(device); + pm_runtime_use_autosuspend(kdev); } /* @@ -2780,6 +2782,5 @@ void intel_runtime_pm_enable(struct drm_i915_private *dev_priv) * We drop that here and will reacquire it during unloading in * intel_power_domains_fini(). */ - pm_runtime_put_autosuspend(device); + pm_runtime_put_autosuspend(kdev); } - diff --git a/drivers/gpu/drm/i915/intel_sdvo.c b/drivers/gpu/drm/i915/intel_sdvo.c index e378f35365a2..c551024d4871 100644 --- a/drivers/gpu/drm/i915/intel_sdvo.c +++ b/drivers/gpu/drm/i915/intel_sdvo.c @@ -1003,24 +1003,22 @@ static bool intel_sdvo_write_infoframe(struct intel_sdvo *intel_sdvo, } static bool intel_sdvo_set_avi_infoframe(struct intel_sdvo *intel_sdvo, - const struct drm_display_mode *adjusted_mode) + struct intel_crtc_state *pipe_config) { uint8_t sdvo_data[HDMI_INFOFRAME_SIZE(AVI)]; - struct drm_crtc *crtc = intel_sdvo->base.base.crtc; - struct intel_crtc *intel_crtc = to_intel_crtc(crtc); union hdmi_infoframe frame; int ret; ssize_t len; ret = drm_hdmi_avi_infoframe_from_display_mode(&frame.avi, - adjusted_mode); + &pipe_config->base.adjusted_mode); if (ret < 0) { DRM_ERROR("couldn't fill AVI infoframe\n"); return false; } if (intel_sdvo->rgb_quant_range_selectable) { - if (intel_crtc->config->limited_color_range) + if (pipe_config->limited_color_range) frame.avi.quantization_range = HDMI_QUANTIZATION_RANGE_LIMITED; else @@ -1125,7 +1123,8 @@ static void i9xx_adjust_sdvo_tv_clock(struct intel_crtc_state *pipe_config) } static bool intel_sdvo_compute_config(struct intel_encoder *encoder, - struct intel_crtc_state *pipe_config) + struct intel_crtc_state *pipe_config, + struct drm_connector_state *conn_state) { struct intel_sdvo *intel_sdvo = to_sdvo(encoder); struct drm_display_mode *adjusted_mode = &pipe_config->base.adjusted_mode; @@ -1192,22 +1191,21 @@ static bool intel_sdvo_compute_config(struct intel_encoder *encoder, return true; } -static void intel_sdvo_pre_enable(struct intel_encoder *intel_encoder) +static void intel_sdvo_pre_enable(struct intel_encoder *intel_encoder, + struct intel_crtc_state *crtc_state, + struct drm_connector_state *conn_state) { struct drm_device *dev = intel_encoder->base.dev; struct drm_i915_private *dev_priv = to_i915(dev); - struct intel_crtc *crtc = to_intel_crtc(intel_encoder->base.crtc); - const struct drm_display_mode *adjusted_mode = &crtc->config->base.adjusted_mode; - struct drm_display_mode *mode = &crtc->config->base.mode; + struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc); + const struct drm_display_mode *adjusted_mode = &crtc_state->base.adjusted_mode; + struct drm_display_mode *mode = &crtc_state->base.mode; struct intel_sdvo *intel_sdvo = to_sdvo(intel_encoder); u32 sdvox; struct intel_sdvo_in_out_map in_out; struct intel_sdvo_dtd input_dtd, output_dtd; int rate; - if (!mode) - return; - /* First, set the input mapping for the first input to our controlled * output. This is only correct if we're a single-input device, in * which case the first input is the output from the appropriate SDVO @@ -1240,11 +1238,11 @@ static void intel_sdvo_pre_enable(struct intel_encoder *intel_encoder) if (!intel_sdvo_set_target_input(intel_sdvo)) return; - if (crtc->config->has_hdmi_sink) { + if (crtc_state->has_hdmi_sink) { intel_sdvo_set_encode(intel_sdvo, SDVO_ENCODE_HDMI); intel_sdvo_set_colorimetry(intel_sdvo, SDVO_COLORIMETRY_RGB256); - intel_sdvo_set_avi_infoframe(intel_sdvo, adjusted_mode); + intel_sdvo_set_avi_infoframe(intel_sdvo, crtc_state); } else intel_sdvo_set_encode(intel_sdvo, SDVO_ENCODE_DVI); @@ -1260,7 +1258,7 @@ static void intel_sdvo_pre_enable(struct intel_encoder *intel_encoder) DRM_INFO("Setting input timings on %s failed\n", SDVO_NAME(intel_sdvo)); - switch (crtc->config->pixel_multiplier) { + switch (crtc_state->pixel_multiplier) { default: WARN(1, "unknown pixel multiplier specified\n"); case 1: rate = SDVO_CLOCK_RATE_MULT_1X; break; @@ -1275,7 +1273,7 @@ static void intel_sdvo_pre_enable(struct intel_encoder *intel_encoder) /* The real mode polarity is set by the SDVO commands, using * struct intel_sdvo_dtd. */ sdvox = SDVO_VSYNC_ACTIVE_HIGH | SDVO_HSYNC_ACTIVE_HIGH; - if (!HAS_PCH_SPLIT(dev) && crtc->config->limited_color_range) + if (!HAS_PCH_SPLIT(dev) && crtc_state->limited_color_range) sdvox |= HDMI_COLOR_RANGE_16_235; if (INTEL_INFO(dev)->gen < 5) sdvox |= SDVO_BORDER_ENABLE; @@ -1301,7 +1299,7 @@ static void intel_sdvo_pre_enable(struct intel_encoder *intel_encoder) } else if (IS_I945G(dev) || IS_I945GM(dev) || IS_G33(dev)) { /* done in crtc_mode_set as it lives inside the dpll register */ } else { - sdvox |= (crtc->config->pixel_multiplier - 1) + sdvox |= (crtc_state->pixel_multiplier - 1) << SDVO_PORT_MULTIPLY_SHIFT; } @@ -1434,7 +1432,9 @@ static void intel_sdvo_get_config(struct intel_encoder *encoder, pipe_config->pixel_multiplier, encoder_pixel_multiplier); } -static void intel_disable_sdvo(struct intel_encoder *encoder) +static void intel_disable_sdvo(struct intel_encoder *encoder, + struct intel_crtc_state *old_crtc_state, + struct drm_connector_state *conn_state) { struct drm_i915_private *dev_priv = to_i915(encoder->base.dev); struct intel_sdvo *intel_sdvo = to_sdvo(encoder); @@ -1477,16 +1477,22 @@ static void intel_disable_sdvo(struct intel_encoder *encoder) } } -static void pch_disable_sdvo(struct intel_encoder *encoder) +static void pch_disable_sdvo(struct intel_encoder *encoder, + struct intel_crtc_state *old_crtc_state, + struct drm_connector_state *old_conn_state) { } -static void pch_post_disable_sdvo(struct intel_encoder *encoder) +static void pch_post_disable_sdvo(struct intel_encoder *encoder, + struct intel_crtc_state *old_crtc_state, + struct drm_connector_state *old_conn_state) { - intel_disable_sdvo(encoder); + intel_disable_sdvo(encoder, old_crtc_state, old_conn_state); } -static void intel_enable_sdvo(struct intel_encoder *encoder) +static void intel_enable_sdvo(struct intel_encoder *encoder, + struct intel_crtc_state *pipe_config, + struct drm_connector_state *conn_state) { struct drm_device *dev = encoder->base.dev; struct drm_i915_private *dev_priv = to_i915(dev); @@ -2930,10 +2936,12 @@ static bool intel_sdvo_init_ddc_proxy(struct intel_sdvo *sdvo, struct drm_device *dev) { + struct pci_dev *pdev = dev->pdev; + sdvo->ddc.owner = THIS_MODULE; sdvo->ddc.class = I2C_CLASS_DDC; snprintf(sdvo->ddc.name, I2C_NAME_SIZE, "SDVO DDC proxy"); - sdvo->ddc.dev.parent = &dev->pdev->dev; + sdvo->ddc.dev.parent = &pdev->dev; sdvo->ddc.algo_data = sdvo; sdvo->ddc.algo = &intel_sdvo_ddc_proxy; diff --git a/drivers/gpu/drm/i915/intel_sprite.c b/drivers/gpu/drm/i915/intel_sprite.c index 366900dcde34..73a521fdf1bd 100644 --- a/drivers/gpu/drm/i915/intel_sprite.c +++ b/drivers/gpu/drm/i915/intel_sprite.c @@ -203,6 +203,9 @@ skl_update_plane(struct drm_plane *drm_plane, struct drm_i915_private *dev_priv = to_i915(dev); struct intel_plane *intel_plane = to_intel_plane(drm_plane); struct drm_framebuffer *fb = plane_state->base.fb; + const struct skl_wm_values *wm = &dev_priv->wm.skl_results; + struct drm_crtc *crtc = crtc_state->base.crtc; + struct intel_crtc *intel_crtc = to_intel_crtc(crtc); const int pipe = intel_plane->pipe; const int plane = intel_plane->plane + 1; u32 plane_ctl; @@ -228,6 +231,9 @@ skl_update_plane(struct drm_plane *drm_plane, plane_ctl |= skl_plane_ctl_rotation(rotation); + if (wm->dirty_pipes & drm_crtc_mask(crtc)) + skl_write_plane_wm(intel_crtc, wm, plane); + if (key->flags) { I915_WRITE(PLANE_KEYVAL(pipe, plane), key->min_value); I915_WRITE(PLANE_KEYMAX(pipe, plane), key->max_value); @@ -286,6 +292,14 @@ skl_disable_plane(struct drm_plane *dplane, struct drm_crtc *crtc) const int pipe = intel_plane->pipe; const int plane = intel_plane->plane + 1; + /* + * We only populate skl_results on watermark updates, and if the + * plane's visiblity isn't actually changing neither is its watermarks. + */ + if (!dplane->state->visible) + skl_write_plane_wm(to_intel_crtc(crtc), + &dev_priv->wm.skl_results, plane); + I915_WRITE(PLANE_CTL(pipe, plane), 0); I915_WRITE(PLANE_SURF(pipe, plane), 0); diff --git a/drivers/gpu/drm/i915/intel_tv.c b/drivers/gpu/drm/i915/intel_tv.c index 49136ad5473e..d960e4866595 100644 --- a/drivers/gpu/drm/i915/intel_tv.c +++ b/drivers/gpu/drm/i915/intel_tv.c @@ -838,7 +838,9 @@ intel_tv_get_hw_state(struct intel_encoder *encoder, enum pipe *pipe) } static void -intel_enable_tv(struct intel_encoder *encoder) +intel_enable_tv(struct intel_encoder *encoder, + struct intel_crtc_state *pipe_config, + struct drm_connector_state *conn_state) { struct drm_device *dev = encoder->base.dev; struct drm_i915_private *dev_priv = to_i915(dev); @@ -851,7 +853,9 @@ intel_enable_tv(struct intel_encoder *encoder) } static void -intel_disable_tv(struct intel_encoder *encoder) +intel_disable_tv(struct intel_encoder *encoder, + struct intel_crtc_state *old_crtc_state, + struct drm_connector_state *old_conn_state) { struct drm_device *dev = encoder->base.dev; struct drm_i915_private *dev_priv = to_i915(dev); @@ -908,7 +912,8 @@ intel_tv_get_config(struct intel_encoder *encoder, static bool intel_tv_compute_config(struct intel_encoder *encoder, - struct intel_crtc_state *pipe_config) + struct intel_crtc_state *pipe_config, + struct drm_connector_state *conn_state) { struct intel_tv *intel_tv = enc_to_tv(encoder); const struct tv_mode *tv_mode = intel_tv_mode_find(intel_tv); @@ -1010,7 +1015,9 @@ static void set_color_conversion(struct drm_i915_private *dev_priv, color_conversion->av); } -static void intel_tv_pre_enable(struct intel_encoder *encoder) +static void intel_tv_pre_enable(struct intel_encoder *encoder, + struct intel_crtc_state *pipe_config, + struct drm_connector_state *conn_state) { struct drm_device *dev = encoder->base.dev; struct drm_i915_private *dev_priv = to_i915(dev); diff --git a/drivers/gpu/drm/i915/intel_uncore.c b/drivers/gpu/drm/i915/intel_uncore.c index 43f833901b8e..a9b6c936aadd 100644 --- a/drivers/gpu/drm/i915/intel_uncore.c +++ b/drivers/gpu/drm/i915/intel_uncore.c @@ -1018,11 +1018,9 @@ gen5_write##x(struct drm_i915_private *dev_priv, i915_reg_t reg, u##x val, bool __gen5_write(8) __gen5_write(16) __gen5_write(32) -__gen5_write(64) __gen2_write(8) __gen2_write(16) __gen2_write(32) -__gen2_write(64) #undef __gen5_write #undef __gen2_write @@ -1112,23 +1110,18 @@ gen9_write##x(struct drm_i915_private *dev_priv, i915_reg_t reg, u##x val, \ __gen9_write(8) __gen9_write(16) __gen9_write(32) -__gen9_write(64) __chv_write(8) __chv_write(16) __chv_write(32) -__chv_write(64) __gen8_write(8) __gen8_write(16) __gen8_write(32) -__gen8_write(64) __hsw_write(8) __hsw_write(16) __hsw_write(32) -__hsw_write(64) __gen6_write(8) __gen6_write(16) __gen6_write(32) -__gen6_write(64) #undef __gen9_write #undef __chv_write @@ -1158,7 +1151,6 @@ static void vgpu_write##x(struct drm_i915_private *dev_priv, \ __vgpu_write(8) __vgpu_write(16) __vgpu_write(32) -__vgpu_write(64) #undef __vgpu_write #undef VGPU_WRITE_FOOTER @@ -1169,7 +1161,6 @@ do { \ dev_priv->uncore.funcs.mmio_writeb = x##_write8; \ dev_priv->uncore.funcs.mmio_writew = x##_write16; \ dev_priv->uncore.funcs.mmio_writel = x##_write32; \ - dev_priv->uncore.funcs.mmio_writeq = x##_write64; \ } while (0) #define ASSIGN_READ_MMIO_VFUNCS(x) \ @@ -1597,8 +1588,10 @@ static int gen6_reset_engines(struct drm_i915_private *dev_priv, if (engine_mask == ALL_ENGINES) { hw_mask = GEN6_GRDOM_FULL; } else { + unsigned int tmp; + hw_mask = 0; - for_each_engine_masked(engine, dev_priv, engine_mask) + for_each_engine_masked(engine, dev_priv, engine_mask, tmp) hw_mask |= hw_engine_mask[engine->id]; } @@ -1714,15 +1707,16 @@ static int gen8_reset_engines(struct drm_i915_private *dev_priv, unsigned engine_mask) { struct intel_engine_cs *engine; + unsigned int tmp; - for_each_engine_masked(engine, dev_priv, engine_mask) + for_each_engine_masked(engine, dev_priv, engine_mask, tmp) if (gen8_request_engine_reset(engine)) goto not_ready; return gen6_reset_engines(dev_priv, engine_mask); not_ready: - for_each_engine_masked(engine, dev_priv, engine_mask) + for_each_engine_masked(engine, dev_priv, engine_mask, tmp) gen8_unrequest_engine_reset(engine); return -EIO; diff --git a/drivers/gpu/drm/imx/imx-drm-core.c b/drivers/gpu/drm/imx/imx-drm-core.c index 56dfc4cd50c6..98df09c2b388 100644 --- a/drivers/gpu/drm/imx/imx-drm-core.c +++ b/drivers/gpu/drm/imx/imx-drm-core.c @@ -64,25 +64,6 @@ static void imx_drm_driver_lastclose(struct drm_device *drm) drm_fbdev_cma_restore_mode(imxdrm->fbhelper); } -static int imx_drm_driver_unload(struct drm_device *drm) -{ - struct imx_drm_device *imxdrm = drm->dev_private; - - drm_kms_helper_poll_fini(drm); - - if (imxdrm->fbhelper) - drm_fbdev_cma_fini(imxdrm->fbhelper); - - component_unbind_all(drm->dev, drm); - - drm_vblank_cleanup(drm); - drm_mode_config_cleanup(drm); - - platform_set_drvdata(drm->platformdev, NULL); - - return 0; -} - static int imx_drm_enable_vblank(struct drm_device *drm, unsigned int pipe) { struct imx_drm_device *imxdrm = drm->dev_private; @@ -146,55 +127,73 @@ static void imx_drm_output_poll_changed(struct drm_device *drm) drm_fbdev_cma_hotplug_event(imxdrm->fbhelper); } +static int imx_drm_atomic_check(struct drm_device *dev, + struct drm_atomic_state *state) +{ + int ret; + + ret = drm_atomic_helper_check_modeset(dev, state); + if (ret) + return ret; + + ret = drm_atomic_helper_check_planes(dev, state); + if (ret) + return ret; + + /* + * Check modeset again in case crtc_state->mode_changed is + * updated in plane's ->atomic_check callback. + */ + ret = drm_atomic_helper_check_modeset(dev, state); + if (ret) + return ret; + + return ret; +} + +static int imx_drm_atomic_commit(struct drm_device *dev, + struct drm_atomic_state *state, + bool nonblock) +{ + struct drm_plane_state *plane_state; + struct drm_plane *plane; + struct dma_buf *dma_buf; + int i; + + /* + * If the plane fb has an dma-buf attached, fish out the exclusive + * fence for the atomic helper to wait on. + */ + for_each_plane_in_state(state, plane, plane_state, i) { + if ((plane->state->fb != plane_state->fb) && plane_state->fb) { + dma_buf = drm_fb_cma_get_gem_obj(plane_state->fb, + 0)->base.dma_buf; + if (!dma_buf) + continue; + plane_state->fence = + reservation_object_get_excl_rcu(dma_buf->resv); + } + } + + return drm_atomic_helper_commit(dev, state, nonblock); +} + static const struct drm_mode_config_funcs imx_drm_mode_config_funcs = { .fb_create = drm_fb_cma_create, .output_poll_changed = imx_drm_output_poll_changed, - .atomic_check = drm_atomic_helper_check, - .atomic_commit = drm_atomic_helper_commit, + .atomic_check = imx_drm_atomic_check, + .atomic_commit = imx_drm_atomic_commit, }; static void imx_drm_atomic_commit_tail(struct drm_atomic_state *state) { struct drm_device *dev = state->dev; - struct drm_crtc *crtc; - struct drm_crtc_state *crtc_state; - struct drm_plane_state *plane_state; - struct drm_gem_cma_object *cma_obj; - struct fence *excl; - unsigned shared_count; - struct fence **shared; - unsigned int i, j; - int ret; - - /* Wait for fences. */ - for_each_crtc_in_state(state, crtc, crtc_state, i) { - plane_state = crtc->primary->state; - if (plane_state->fb) { - cma_obj = drm_fb_cma_get_gem_obj(plane_state->fb, 0); - if (cma_obj->base.dma_buf) { - ret = reservation_object_get_fences_rcu( - cma_obj->base.dma_buf->resv, &excl, - &shared_count, &shared); - if (unlikely(ret)) - DRM_ERROR("failed to get fences " - "for buffer\n"); - - if (excl) { - fence_wait(excl, false); - fence_put(excl); - } - for (j = 0; j < shared_count; i++) { - fence_wait(shared[j], false); - fence_put(shared[j]); - } - } - } - } drm_atomic_helper_commit_modeset_disables(dev, state); drm_atomic_helper_commit_planes(dev, state, - DRM_PLANE_COMMIT_ACTIVE_ONLY); + DRM_PLANE_COMMIT_ACTIVE_ONLY | + DRM_PLANE_COMMIT_NO_DISABLE_AFTER_MODESET); drm_atomic_helper_commit_modeset_enables(dev, state); @@ -210,111 +209,6 @@ static struct drm_mode_config_helper_funcs imx_drm_mode_config_helpers = { }; /* - * Main DRM initialisation. This binds, initialises and registers - * with DRM the subcomponents of the driver. - */ -static int imx_drm_driver_load(struct drm_device *drm, unsigned long flags) -{ - struct imx_drm_device *imxdrm; - struct drm_connector *connector; - int ret; - - imxdrm = devm_kzalloc(drm->dev, sizeof(*imxdrm), GFP_KERNEL); - if (!imxdrm) - return -ENOMEM; - - imxdrm->drm = drm; - - drm->dev_private = imxdrm; - - /* - * enable drm irq mode. - * - with irq_enabled = true, we can use the vblank feature. - * - * P.S. note that we wouldn't use drm irq handler but - * just specific driver own one instead because - * drm framework supports only one irq handler and - * drivers can well take care of their interrupts - */ - drm->irq_enabled = true; - - /* - * set max width and height as default value(4096x4096). - * this value would be used to check framebuffer size limitation - * at drm_mode_addfb(). - */ - drm->mode_config.min_width = 64; - drm->mode_config.min_height = 64; - drm->mode_config.max_width = 4096; - drm->mode_config.max_height = 4096; - drm->mode_config.funcs = &imx_drm_mode_config_funcs; - drm->mode_config.helper_private = &imx_drm_mode_config_helpers; - - drm_mode_config_init(drm); - - ret = drm_vblank_init(drm, MAX_CRTC); - if (ret) - goto err_kms; - - platform_set_drvdata(drm->platformdev, drm); - - /* Now try and bind all our sub-components */ - ret = component_bind_all(drm->dev, drm); - if (ret) - goto err_vblank; - - /* - * All components are now added, we can publish the connector sysfs - * entries to userspace. This will generate hotplug events and so - * userspace will expect to be able to access DRM at this point. - */ - list_for_each_entry(connector, &drm->mode_config.connector_list, head) { - ret = drm_connector_register(connector); - if (ret) { - dev_err(drm->dev, - "[CONNECTOR:%d:%s] drm_connector_register failed: %d\n", - connector->base.id, - connector->name, ret); - goto err_unbind; - } - } - - drm_mode_config_reset(drm); - - /* - * All components are now initialised, so setup the fb helper. - * The fb helper takes copies of key hardware information, so the - * crtcs/connectors/encoders must not change after this point. - */ -#if IS_ENABLED(CONFIG_DRM_FBDEV_EMULATION) - if (legacyfb_depth != 16 && legacyfb_depth != 32) { - dev_warn(drm->dev, "Invalid legacyfb_depth. Defaulting to 16bpp\n"); - legacyfb_depth = 16; - } - imxdrm->fbhelper = drm_fbdev_cma_init(drm, legacyfb_depth, - drm->mode_config.num_crtc, MAX_CRTC); - if (IS_ERR(imxdrm->fbhelper)) { - ret = PTR_ERR(imxdrm->fbhelper); - imxdrm->fbhelper = NULL; - goto err_unbind; - } -#endif - - drm_kms_helper_poll_init(drm); - - return 0; - -err_unbind: - component_unbind_all(drm->dev, drm); -err_vblank: - drm_vblank_cleanup(drm); -err_kms: - drm_mode_config_cleanup(drm); - - return ret; -} - -/* * imx_drm_add_crtc - add a new crtc */ int imx_drm_add_crtc(struct drm_device *drm, struct drm_crtc *crtc, @@ -406,8 +300,6 @@ static const struct drm_ioctl_desc imx_drm_ioctls[] = { static struct drm_driver imx_drm_driver = { .driver_features = DRIVER_MODESET | DRIVER_GEM | DRIVER_PRIME | DRIVER_ATOMIC, - .load = imx_drm_driver_load, - .unload = imx_drm_driver_unload, .lastclose = imx_drm_driver_lastclose, .gem_free_object_unlocked = drm_gem_cma_free_object, .gem_vm_ops = &drm_gem_cma_vm_ops, @@ -460,12 +352,122 @@ static int compare_of(struct device *dev, void *data) static int imx_drm_bind(struct device *dev) { - return drm_platform_init(&imx_drm_driver, to_platform_device(dev)); + struct drm_device *drm; + struct imx_drm_device *imxdrm; + int ret; + + drm = drm_dev_alloc(&imx_drm_driver, dev); + if (!drm) + return -ENOMEM; + + imxdrm = devm_kzalloc(dev, sizeof(*imxdrm), GFP_KERNEL); + if (!imxdrm) { + ret = -ENOMEM; + goto err_unref; + } + + imxdrm->drm = drm; + drm->dev_private = imxdrm; + + /* + * enable drm irq mode. + * - with irq_enabled = true, we can use the vblank feature. + * + * P.S. note that we wouldn't use drm irq handler but + * just specific driver own one instead because + * drm framework supports only one irq handler and + * drivers can well take care of their interrupts + */ + drm->irq_enabled = true; + + /* + * set max width and height as default value(4096x4096). + * this value would be used to check framebuffer size limitation + * at drm_mode_addfb(). + */ + drm->mode_config.min_width = 64; + drm->mode_config.min_height = 64; + drm->mode_config.max_width = 4096; + drm->mode_config.max_height = 4096; + drm->mode_config.funcs = &imx_drm_mode_config_funcs; + drm->mode_config.helper_private = &imx_drm_mode_config_helpers; + + drm_mode_config_init(drm); + + ret = drm_vblank_init(drm, MAX_CRTC); + if (ret) + goto err_kms; + + dev_set_drvdata(dev, drm); + + /* Now try and bind all our sub-components */ + ret = component_bind_all(dev, drm); + if (ret) + goto err_vblank; + + drm_mode_config_reset(drm); + + /* + * All components are now initialised, so setup the fb helper. + * The fb helper takes copies of key hardware information, so the + * crtcs/connectors/encoders must not change after this point. + */ +#if IS_ENABLED(CONFIG_DRM_FBDEV_EMULATION) + if (legacyfb_depth != 16 && legacyfb_depth != 32) { + dev_warn(dev, "Invalid legacyfb_depth. Defaulting to 16bpp\n"); + legacyfb_depth = 16; + } + imxdrm->fbhelper = drm_fbdev_cma_init(drm, legacyfb_depth, + drm->mode_config.num_crtc, MAX_CRTC); + if (IS_ERR(imxdrm->fbhelper)) { + ret = PTR_ERR(imxdrm->fbhelper); + imxdrm->fbhelper = NULL; + goto err_unbind; + } +#endif + + drm_kms_helper_poll_init(drm); + + ret = drm_dev_register(drm, 0); + if (ret) + goto err_fbhelper; + + return 0; + +err_fbhelper: + drm_kms_helper_poll_fini(drm); + if (imxdrm->fbhelper) + drm_fbdev_cma_fini(imxdrm->fbhelper); +err_unbind: + component_unbind_all(drm->dev, drm); +err_vblank: + drm_vblank_cleanup(drm); +err_kms: + drm_mode_config_cleanup(drm); +err_unref: + drm_dev_unref(drm); + + return ret; } static void imx_drm_unbind(struct device *dev) { - drm_put_dev(dev_get_drvdata(dev)); + struct drm_device *drm = dev_get_drvdata(dev); + struct imx_drm_device *imxdrm = drm->dev_private; + + drm_dev_unregister(drm); + + drm_kms_helper_poll_fini(drm); + + if (imxdrm->fbhelper) + drm_fbdev_cma_fini(imxdrm->fbhelper); + + drm_mode_config_cleanup(drm); + + component_unbind_all(drm->dev, drm); + dev_set_drvdata(dev, NULL); + + drm_dev_unref(drm); } static const struct component_master_ops imx_drm_ops = { diff --git a/drivers/gpu/drm/imx/imx-ldb.c b/drivers/gpu/drm/imx/imx-ldb.c index 4eed3a6addad..3ce391c239b0 100644 --- a/drivers/gpu/drm/imx/imx-ldb.c +++ b/drivers/gpu/drm/imx/imx-ldb.c @@ -757,11 +757,10 @@ static void imx_ldb_unbind(struct device *dev, struct device *master, for (i = 0; i < 2; i++) { struct imx_ldb_channel *channel = &imx_ldb->channel[i]; - if (!channel->connector.funcs) - continue; - - channel->connector.funcs->destroy(&channel->connector); - channel->encoder.funcs->destroy(&channel->encoder); + if (channel->bridge) + drm_bridge_detach(channel->bridge); + if (channel->panel) + drm_panel_detach(channel->panel); kfree(channel->edid); i2c_put_adapter(channel->ddc); diff --git a/drivers/gpu/drm/imx/imx-tve.c b/drivers/gpu/drm/imx/imx-tve.c index 5e875944ffa2..8fc088843e55 100644 --- a/drivers/gpu/drm/imx/imx-tve.c +++ b/drivers/gpu/drm/imx/imx-tve.c @@ -685,9 +685,6 @@ static void imx_tve_unbind(struct device *dev, struct device *master, { struct imx_tve *tve = dev_get_drvdata(dev); - tve->connector.funcs->destroy(&tve->connector); - tve->encoder.funcs->destroy(&tve->encoder); - if (!IS_ERR(tve->dac_reg)) regulator_disable(tve->dac_reg); } diff --git a/drivers/gpu/drm/imx/ipuv3-crtc.c b/drivers/gpu/drm/imx/ipuv3-crtc.c index 6e1dc902522c..9df29f1cb16a 100644 --- a/drivers/gpu/drm/imx/ipuv3-crtc.c +++ b/drivers/gpu/drm/imx/ipuv3-crtc.c @@ -60,7 +60,8 @@ static void ipu_crtc_enable(struct drm_crtc *crtc) ipu_di_enable(ipu_crtc->di); } -static void ipu_crtc_disable(struct drm_crtc *crtc) +static void ipu_crtc_atomic_disable(struct drm_crtc *crtc, + struct drm_crtc_state *old_crtc_state) { struct ipu_crtc *ipu_crtc = to_ipu_crtc(crtc); struct ipu_soc *ipu = dev_get_drvdata(ipu_crtc->dev->parent); @@ -75,6 +76,9 @@ static void ipu_crtc_disable(struct drm_crtc *crtc) crtc->state->event = NULL; } spin_unlock_irq(&crtc->dev->event_lock); + + /* always disable planes on the CRTC */ + drm_atomic_helper_disable_planes_on_crtc(old_crtc_state, true); } static void imx_drm_crtc_reset(struct drm_crtc *crtc) @@ -120,9 +124,14 @@ static void imx_drm_crtc_destroy_state(struct drm_crtc *crtc, kfree(to_imx_crtc_state(state)); } +static void imx_drm_crtc_destroy(struct drm_crtc *crtc) +{ + imx_drm_remove_crtc(to_ipu_crtc(crtc)->imx_crtc); +} + static const struct drm_crtc_funcs ipu_crtc_funcs = { .set_config = drm_atomic_helper_set_config, - .destroy = drm_crtc_cleanup, + .destroy = imx_drm_crtc_destroy, .page_flip = drm_atomic_helper_page_flip, .reset = imx_drm_crtc_reset, .atomic_duplicate_state = imx_drm_crtc_duplicate_state, @@ -241,7 +250,7 @@ static const struct drm_crtc_helper_funcs ipu_helper_funcs = { .mode_set_nofb = ipu_crtc_mode_set_nofb, .atomic_check = ipu_crtc_atomic_check, .atomic_begin = ipu_crtc_atomic_begin, - .disable = ipu_crtc_disable, + .atomic_disable = ipu_crtc_atomic_disable, .enable = ipu_crtc_enable, }; @@ -409,8 +418,6 @@ static void ipu_drm_unbind(struct device *dev, struct device *master, { struct ipu_crtc *ipu_crtc = dev_get_drvdata(dev); - imx_drm_remove_crtc(ipu_crtc->imx_crtc); - ipu_put_resources(ipu_crtc); if (ipu_crtc->plane[1]) ipu_plane_put_resources(ipu_crtc->plane[1]); diff --git a/drivers/gpu/drm/imx/ipuv3-plane.c b/drivers/gpu/drm/imx/ipuv3-plane.c index 4ad67d015ec7..ce22d0a0ddc8 100644 --- a/drivers/gpu/drm/imx/ipuv3-plane.c +++ b/drivers/gpu/drm/imx/ipuv3-plane.c @@ -213,8 +213,12 @@ static void ipu_plane_enable(struct ipu_plane *ipu_plane) ipu_dp_enable_channel(ipu_plane->dp); } -static void ipu_plane_disable(struct ipu_plane *ipu_plane) +static int ipu_disable_plane(struct drm_plane *plane) { + struct ipu_plane *ipu_plane = to_ipu_plane(plane); + + DRM_DEBUG_KMS("[%d] %s\n", __LINE__, __func__); + ipu_idmac_wait_busy(ipu_plane->ipu_ch, 50); if (ipu_plane->dp) @@ -223,15 +227,6 @@ static void ipu_plane_disable(struct ipu_plane *ipu_plane) ipu_dmfc_disable_channel(ipu_plane->dmfc); if (ipu_plane->dp) ipu_dp_disable(ipu_plane->ipu); -} - -static int ipu_disable_plane(struct drm_plane *plane) -{ - struct ipu_plane *ipu_plane = to_ipu_plane(plane); - - DRM_DEBUG_KMS("[%d] %s\n", __LINE__, __func__); - - ipu_plane_disable(ipu_plane); return 0; } @@ -242,7 +237,6 @@ static void ipu_plane_destroy(struct drm_plane *plane) DRM_DEBUG_KMS("[%d] %s\n", __LINE__, __func__); - ipu_disable_plane(plane); drm_plane_cleanup(plane); kfree(ipu_plane); } @@ -319,13 +313,16 @@ static int ipu_plane_atomic_check(struct drm_plane *plane, return -EINVAL; /* - * since we cannot touch active IDMAC channels, we do not support - * resizing the enabled plane or changing its format + * We support resizing active plane or changing its format by + * forcing CRTC mode change in plane's ->atomic_check callback + * and disabling all affected active planes in CRTC's ->atomic_disable + * callback. The planes will be reenabled in plane's ->atomic_update + * callback. */ if (old_fb && (state->src_w != old_state->src_w || state->src_h != old_state->src_h || fb->pixel_format != old_fb->pixel_format)) - return -EINVAL; + crtc_state->mode_changed = true; eba = drm_plane_state_to_eba(state); @@ -336,7 +333,7 @@ static int ipu_plane_atomic_check(struct drm_plane *plane, return -EINVAL; if (old_fb && fb->pitches[0] != old_fb->pitches[0]) - return -EINVAL; + crtc_state->mode_changed = true; switch (fb->pixel_format) { case DRM_FORMAT_YUV420: @@ -372,7 +369,7 @@ static int ipu_plane_atomic_check(struct drm_plane *plane, return -EINVAL; if (old_fb && old_fb->pitches[1] != fb->pitches[1]) - return -EINVAL; + crtc_state->mode_changed = true; } return 0; @@ -392,8 +389,12 @@ static void ipu_plane_atomic_update(struct drm_plane *plane, enum ipu_color_space ics; if (old_state->fb) { - ipu_plane_atomic_set_base(ipu_plane, old_state); - return; + struct drm_crtc_state *crtc_state = state->crtc->state; + + if (!drm_atomic_crtc_needs_modeset(crtc_state)) { + ipu_plane_atomic_set_base(ipu_plane, old_state); + return; + } } switch (ipu_plane->dp_flow) { diff --git a/drivers/gpu/drm/imx/parallel-display.c b/drivers/gpu/drm/imx/parallel-display.c index 74b0ac06fdab..d796ada2a47a 100644 --- a/drivers/gpu/drm/imx/parallel-display.c +++ b/drivers/gpu/drm/imx/parallel-display.c @@ -293,8 +293,10 @@ static void imx_pd_unbind(struct device *dev, struct device *master, { struct imx_parallel_display *imxpd = dev_get_drvdata(dev); - imxpd->encoder.funcs->destroy(&imxpd->encoder); - imxpd->connector.funcs->destroy(&imxpd->connector); + if (imxpd->bridge) + drm_bridge_detach(imxpd->bridge); + if (imxpd->panel) + drm_panel_detach(imxpd->panel); kfree(imxpd->edid); } diff --git a/drivers/gpu/drm/qxl/qxl_object.c b/drivers/gpu/drm/qxl/qxl_object.c index 5e1d7899dd72..fa5440dc9a19 100644 --- a/drivers/gpu/drm/qxl/qxl_object.c +++ b/drivers/gpu/drm/qxl/qxl_object.c @@ -61,7 +61,7 @@ void qxl_ttm_placement_from_domain(struct qxl_bo *qbo, u32 domain, bool pinned) if (domain == QXL_GEM_DOMAIN_VRAM) qbo->placements[c++].flags = TTM_PL_FLAG_CACHED | TTM_PL_FLAG_VRAM | pflag; if (domain == QXL_GEM_DOMAIN_SURFACE) - qbo->placements[c++].flags = TTM_PL_FLAG_CACHED | TTM_PL_FLAG_PRIV0 | pflag; + qbo->placements[c++].flags = TTM_PL_FLAG_CACHED | TTM_PL_FLAG_PRIV | pflag; if (domain == QXL_GEM_DOMAIN_CPU) qbo->placements[c++].flags = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM | pflag; if (!c) @@ -151,7 +151,7 @@ void *qxl_bo_kmap_atomic_page(struct qxl_device *qdev, if (bo->tbo.mem.mem_type == TTM_PL_VRAM) map = qdev->vram_mapping; - else if (bo->tbo.mem.mem_type == TTM_PL_PRIV0) + else if (bo->tbo.mem.mem_type == TTM_PL_PRIV) map = qdev->surface_mapping; else goto fallback; @@ -191,7 +191,7 @@ void qxl_bo_kunmap_atomic_page(struct qxl_device *qdev, if (bo->tbo.mem.mem_type == TTM_PL_VRAM) map = qdev->vram_mapping; - else if (bo->tbo.mem.mem_type == TTM_PL_PRIV0) + else if (bo->tbo.mem.mem_type == TTM_PL_PRIV) map = qdev->surface_mapping; else goto fallback; @@ -311,7 +311,7 @@ int qxl_bo_check_id(struct qxl_device *qdev, struct qxl_bo *bo) int qxl_surf_evict(struct qxl_device *qdev) { - return ttm_bo_evict_mm(&qdev->mman.bdev, TTM_PL_PRIV0); + return ttm_bo_evict_mm(&qdev->mman.bdev, TTM_PL_PRIV); } int qxl_vram_evict(struct qxl_device *qdev) diff --git a/drivers/gpu/drm/qxl/qxl_ttm.c b/drivers/gpu/drm/qxl/qxl_ttm.c index b2fe2f8e6ae9..e26c82db948b 100644 --- a/drivers/gpu/drm/qxl/qxl_ttm.c +++ b/drivers/gpu/drm/qxl/qxl_ttm.c @@ -168,7 +168,7 @@ static int qxl_init_mem_type(struct ttm_bo_device *bdev, uint32_t type, man->default_caching = TTM_PL_FLAG_CACHED; break; case TTM_PL_VRAM: - case TTM_PL_PRIV0: + case TTM_PL_PRIV: /* "On-card" video ram */ man->func = &ttm_bo_manager_func; man->gpu_offset = 0; @@ -236,7 +236,7 @@ static int qxl_ttm_io_mem_reserve(struct ttm_bo_device *bdev, mem->bus.base = qdev->vram_base; mem->bus.offset = mem->start << PAGE_SHIFT; break; - case TTM_PL_PRIV0: + case TTM_PL_PRIV: mem->bus.is_iomem = true; mem->bus.base = qdev->surfaceram_base; mem->bus.offset = mem->start << PAGE_SHIFT; @@ -377,7 +377,7 @@ static void qxl_bo_move_notify(struct ttm_buffer_object *bo, qbo = to_qxl_bo(bo); qdev = qbo->gem_base.dev->dev_private; - if (bo->mem.mem_type == TTM_PL_PRIV0 && qbo->surface_id) + if (bo->mem.mem_type == TTM_PL_PRIV && qbo->surface_id) qxl_surface_evict(qdev, qbo, new_mem ? true : false); } @@ -423,7 +423,7 @@ int qxl_ttm_init(struct qxl_device *qdev) DRM_ERROR("Failed initializing VRAM heap.\n"); return r; } - r = ttm_bo_init_mm(&qdev->mman.bdev, TTM_PL_PRIV0, + r = ttm_bo_init_mm(&qdev->mman.bdev, TTM_PL_PRIV, qdev->surfaceram_size / PAGE_SIZE); if (r) { DRM_ERROR("Failed initializing Surfaces heap.\n"); @@ -446,7 +446,7 @@ int qxl_ttm_init(struct qxl_device *qdev) void qxl_ttm_fini(struct qxl_device *qdev) { ttm_bo_clean_mm(&qdev->mman.bdev, TTM_PL_VRAM); - ttm_bo_clean_mm(&qdev->mman.bdev, TTM_PL_PRIV0); + ttm_bo_clean_mm(&qdev->mman.bdev, TTM_PL_PRIV); ttm_bo_device_release(&qdev->mman.bdev); qxl_ttm_global_fini(qdev); DRM_INFO("qxl: ttm finalized\n"); @@ -490,7 +490,7 @@ static int qxl_ttm_debugfs_init(struct qxl_device *qdev) if (i == 0) qxl_mem_types_list[i].data = qdev->mman.bdev.man[TTM_PL_VRAM].priv; else - qxl_mem_types_list[i].data = qdev->mman.bdev.man[TTM_PL_PRIV0].priv; + qxl_mem_types_list[i].data = qdev->mman.bdev.man[TTM_PL_PRIV].priv; } return qxl_debugfs_add_files(qdev, qxl_mem_types_list, i); diff --git a/drivers/gpu/drm/radeon/radeon_device.c b/drivers/gpu/drm/radeon/radeon_device.c index a00dd2f74527..b423c0159581 100644 --- a/drivers/gpu/drm/radeon/radeon_device.c +++ b/drivers/gpu/drm/radeon/radeon_device.c @@ -639,7 +639,7 @@ void radeon_gtt_location(struct radeon_device *rdev, struct radeon_mc *mc) * Used at driver startup. * Returns true if virtual or false if not. */ -static bool radeon_device_is_virtual(void) +bool radeon_device_is_virtual(void) { #ifdef CONFIG_X86 return boot_cpu_has(X86_FEATURE_HYPERVISOR); @@ -1594,7 +1594,8 @@ int radeon_suspend_kms(struct drm_device *dev, bool suspend, rdev = dev->dev_private; - if (dev->switch_power_state == DRM_SWITCH_POWER_OFF) + if (dev->switch_power_state == DRM_SWITCH_POWER_OFF || + dev->switch_power_state == DRM_SWITCH_POWER_DYNAMIC_OFF) return 0; drm_kms_helper_poll_disable(dev); @@ -1689,7 +1690,8 @@ int radeon_resume_kms(struct drm_device *dev, bool resume, bool fbcon) struct drm_crtc *crtc; int r; - if (dev->switch_power_state == DRM_SWITCH_POWER_OFF) + if (dev->switch_power_state == DRM_SWITCH_POWER_OFF || + dev->switch_power_state == DRM_SWITCH_POWER_DYNAMIC_OFF) return 0; if (fbcon) { @@ -1956,14 +1958,3 @@ static void radeon_debugfs_remove_files(struct radeon_device *rdev) } #endif } - -#if defined(CONFIG_DEBUG_FS) -int radeon_debugfs_init(struct drm_minor *minor) -{ - return 0; -} - -void radeon_debugfs_cleanup(struct drm_minor *minor) -{ -} -#endif diff --git a/drivers/gpu/drm/radeon/radeon_drv.c b/drivers/gpu/drm/radeon/radeon_drv.c index 07e44931f1f1..78367ba8bb7d 100644 --- a/drivers/gpu/drm/radeon/radeon_drv.c +++ b/drivers/gpu/drm/radeon/radeon_drv.c @@ -156,11 +156,6 @@ void radeon_gem_prime_vunmap(struct drm_gem_object *obj, void *vaddr); extern long radeon_kms_compat_ioctl(struct file *filp, unsigned int cmd, unsigned long arg); -#if defined(CONFIG_DEBUG_FS) -int radeon_debugfs_init(struct drm_minor *minor); -void radeon_debugfs_cleanup(struct drm_minor *minor); -#endif - /* atpx handler */ #if defined(CONFIG_VGA_SWITCHEROO) void radeon_register_atpx_handler(void); @@ -311,6 +306,8 @@ MODULE_DEVICE_TABLE(pci, pciidlist); static struct drm_driver kms_driver; +bool radeon_device_is_virtual(void); + static int radeon_kick_out_firmware_fb(struct pci_dev *pdev) { struct apertures_struct *ap; @@ -364,6 +361,16 @@ radeon_pci_remove(struct pci_dev *pdev) drm_put_dev(dev); } +static void +radeon_pci_shutdown(struct pci_dev *pdev) +{ + /* if we are running in a VM, make sure the device + * torn down properly on reboot/shutdown + */ + if (radeon_device_is_virtual()) + radeon_pci_remove(pdev); +} + static int radeon_pmops_suspend(struct device *dev) { struct pci_dev *pdev = to_pci_dev(dev); @@ -375,6 +382,14 @@ static int radeon_pmops_resume(struct device *dev) { struct pci_dev *pdev = to_pci_dev(dev); struct drm_device *drm_dev = pci_get_drvdata(pdev); + + /* GPU comes up enabled by the bios on resume */ + if (radeon_is_px(drm_dev)) { + pm_runtime_disable(dev); + pm_runtime_set_active(dev); + pm_runtime_enable(dev); + } + return radeon_resume_kms(drm_dev, true, true); } @@ -531,10 +546,6 @@ static struct drm_driver kms_driver = { .disable_vblank = radeon_disable_vblank_kms, .get_vblank_timestamp = radeon_get_vblank_timestamp_kms, .get_scanout_position = radeon_get_crtc_scanoutpos, -#if defined(CONFIG_DEBUG_FS) - .debugfs_init = radeon_debugfs_init, - .debugfs_cleanup = radeon_debugfs_cleanup, -#endif .irq_preinstall = radeon_driver_irq_preinstall_kms, .irq_postinstall = radeon_driver_irq_postinstall_kms, .irq_uninstall = radeon_driver_irq_uninstall_kms, @@ -576,6 +587,7 @@ static struct pci_driver radeon_kms_pci_driver = { .id_table = pciidlist, .probe = radeon_pci_probe, .remove = radeon_pci_remove, + .shutdown = radeon_pci_shutdown, .driver.pm = &radeon_pm_ops, }; diff --git a/drivers/gpu/drm/radeon/radeon_fb.c b/drivers/gpu/drm/radeon/radeon_fb.c index 568e036d547e..0daad446d2c7 100644 --- a/drivers/gpu/drm/radeon/radeon_fb.c +++ b/drivers/gpu/drm/radeon/radeon_fb.c @@ -25,6 +25,7 @@ */ #include <linux/module.h> #include <linux/slab.h> +#include <linux/pm_runtime.h> #include <drm/drmP.h> #include <drm/drm_crtc.h> @@ -46,8 +47,35 @@ struct radeon_fbdev { struct radeon_device *rdev; }; +static int +radeonfb_open(struct fb_info *info, int user) +{ + struct radeon_fbdev *rfbdev = info->par; + struct radeon_device *rdev = rfbdev->rdev; + int ret = pm_runtime_get_sync(rdev->ddev->dev); + if (ret < 0 && ret != -EACCES) { + pm_runtime_mark_last_busy(rdev->ddev->dev); + pm_runtime_put_autosuspend(rdev->ddev->dev); + return ret; + } + return 0; +} + +static int +radeonfb_release(struct fb_info *info, int user) +{ + struct radeon_fbdev *rfbdev = info->par; + struct radeon_device *rdev = rfbdev->rdev; + + pm_runtime_mark_last_busy(rdev->ddev->dev); + pm_runtime_put_autosuspend(rdev->ddev->dev); + return 0; +} + static struct fb_ops radeonfb_ops = { .owner = THIS_MODULE, + .fb_open = radeonfb_open, + .fb_release = radeonfb_release, .fb_check_var = drm_fb_helper_check_var, .fb_set_par = drm_fb_helper_set_par, .fb_fillrect = drm_fb_helper_cfb_fillrect, diff --git a/drivers/gpu/drm/radeon/radeon_kms.c b/drivers/gpu/drm/radeon/radeon_kms.c index 835563c1f0ed..4388ddeec8d2 100644 --- a/drivers/gpu/drm/radeon/radeon_kms.c +++ b/drivers/gpu/drm/radeon/radeon_kms.c @@ -641,11 +641,11 @@ int radeon_driver_open_kms(struct drm_device *dev, struct drm_file *file_priv) if (rdev->family >= CHIP_CAYMAN) { struct radeon_fpriv *fpriv; struct radeon_vm *vm; - int r; fpriv = kzalloc(sizeof(*fpriv), GFP_KERNEL); if (unlikely(!fpriv)) { - return -ENOMEM; + r = -ENOMEM; + goto out_suspend; } if (rdev->accel_working) { @@ -653,14 +653,14 @@ int radeon_driver_open_kms(struct drm_device *dev, struct drm_file *file_priv) r = radeon_vm_init(rdev, vm); if (r) { kfree(fpriv); - return r; + goto out_suspend; } r = radeon_bo_reserve(rdev->ring_tmp_bo.bo, false); if (r) { radeon_vm_fini(rdev, vm); kfree(fpriv); - return r; + goto out_suspend; } /* map the ib pool buffer read only into @@ -674,15 +674,16 @@ int radeon_driver_open_kms(struct drm_device *dev, struct drm_file *file_priv) if (r) { radeon_vm_fini(rdev, vm); kfree(fpriv); - return r; + goto out_suspend; } } file_priv->driver_priv = fpriv; } +out_suspend: pm_runtime_mark_last_busy(dev->dev); pm_runtime_put_autosuspend(dev->dev); - return 0; + return r; } /** @@ -717,6 +718,8 @@ void radeon_driver_postclose_kms(struct drm_device *dev, kfree(fpriv); file_priv->driver_priv = NULL; } + pm_runtime_mark_last_busy(dev->dev); + pm_runtime_put_autosuspend(dev->dev); } /** @@ -733,6 +736,8 @@ void radeon_driver_preclose_kms(struct drm_device *dev, { struct radeon_device *rdev = dev->dev_private; + pm_runtime_get_sync(dev->dev); + mutex_lock(&rdev->gem.mutex); if (rdev->hyperz_filp == file_priv) rdev->hyperz_filp = NULL; diff --git a/drivers/gpu/drm/ttm/ttm_bo.c b/drivers/gpu/drm/ttm/ttm_bo.c index c2a30bdc8a01..fc6217dfe401 100644 --- a/drivers/gpu/drm/ttm/ttm_bo.c +++ b/drivers/gpu/drm/ttm/ttm_bo.c @@ -57,14 +57,14 @@ static struct attribute ttm_bo_count = { static inline int ttm_mem_type_from_place(const struct ttm_place *place, uint32_t *mem_type) { - int i; + int pos; - for (i = 0; i <= TTM_PL_PRIV5; i++) - if (place->flags & (1 << i)) { - *mem_type = i; - return 0; - } - return -EINVAL; + pos = ffs(place->flags & TTM_PL_MASK_MEM); + if (unlikely(!pos)) + return -EINVAL; + + *mem_type = pos - 1; + return 0; } static void ttm_mem_type_debug(struct ttm_bo_device *bdev, int mem_type) diff --git a/drivers/gpu/drm/ttm/ttm_page_alloc_dma.c b/drivers/gpu/drm/ttm/ttm_page_alloc_dma.c index bef9f6feb635..cec4b4baa179 100644 --- a/drivers/gpu/drm/ttm/ttm_page_alloc_dma.c +++ b/drivers/gpu/drm/ttm/ttm_page_alloc_dma.c @@ -858,7 +858,6 @@ static int ttm_dma_pool_get_pages(struct dma_pool *pool, if (count) { d_page = list_first_entry(&pool->free_list, struct dma_page, page_list); ttm->pages[index] = d_page->p; - ttm_dma->cpu_address[index] = d_page->vaddr; ttm_dma->dma_address[index] = d_page->dma; list_move_tail(&d_page->page_list, &ttm_dma->pages_list); r = 0; @@ -989,7 +988,6 @@ void ttm_dma_unpopulate(struct ttm_dma_tt *ttm_dma, struct device *dev) INIT_LIST_HEAD(&ttm_dma->pages_list); for (i = 0; i < ttm->num_pages; i++) { ttm->pages[i] = NULL; - ttm_dma->cpu_address[i] = 0; ttm_dma->dma_address[i] = 0; } diff --git a/drivers/gpu/drm/ttm/ttm_tt.c b/drivers/gpu/drm/ttm/ttm_tt.c index bc5aa573f466..aee3c00f836e 100644 --- a/drivers/gpu/drm/ttm/ttm_tt.c +++ b/drivers/gpu/drm/ttm/ttm_tt.c @@ -57,10 +57,8 @@ static void ttm_dma_tt_alloc_page_directory(struct ttm_dma_tt *ttm) { ttm->ttm.pages = drm_calloc_large(ttm->ttm.num_pages, sizeof(*ttm->ttm.pages) + - sizeof(*ttm->dma_address) + - sizeof(*ttm->cpu_address)); - ttm->cpu_address = (void *) (ttm->ttm.pages + ttm->ttm.num_pages); - ttm->dma_address = (void *) (ttm->cpu_address + ttm->ttm.num_pages); + sizeof(*ttm->dma_address)); + ttm->dma_address = (void *) (ttm->ttm.pages + ttm->ttm.num_pages); } #ifdef CONFIG_X86 @@ -244,7 +242,6 @@ void ttm_dma_tt_fini(struct ttm_dma_tt *ttm_dma) drm_free_large(ttm->pages); ttm->pages = NULL; - ttm_dma->cpu_address = NULL; ttm_dma->dma_address = NULL; } EXPORT_SYMBOL(ttm_dma_tt_fini); diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h index 74304b03f9d4..070d750af16d 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h @@ -67,10 +67,10 @@ VMWGFX_NUM_GB_SURFACE +\ VMWGFX_NUM_GB_SCREEN_TARGET) -#define VMW_PL_GMR TTM_PL_PRIV0 -#define VMW_PL_FLAG_GMR TTM_PL_FLAG_PRIV0 -#define VMW_PL_MOB TTM_PL_PRIV1 -#define VMW_PL_FLAG_MOB TTM_PL_FLAG_PRIV1 +#define VMW_PL_GMR (TTM_PL_PRIV + 0) +#define VMW_PL_FLAG_GMR (TTM_PL_FLAG_PRIV << 0) +#define VMW_PL_MOB (TTM_PL_PRIV + 1) +#define VMW_PL_FLAG_MOB (TTM_PL_FLAG_PRIV << 1) #define VMW_RES_CONTEXT ttm_driver_type0 #define VMW_RES_SURFACE ttm_driver_type1 diff --git a/drivers/gpu/ipu-v3/Makefile b/drivers/gpu/ipu-v3/Makefile index 107ec236a4a6..5f961416c4ee 100644 --- a/drivers/gpu/ipu-v3/Makefile +++ b/drivers/gpu/ipu-v3/Makefile @@ -1,4 +1,5 @@ obj-$(CONFIG_IMX_IPUV3_CORE) += imx-ipu-v3.o imx-ipu-v3-objs := ipu-common.o ipu-cpmem.o ipu-csi.o ipu-dc.o ipu-di.o \ - ipu-dp.o ipu-dmfc.o ipu-ic.o ipu-smfc.o + ipu-dp.o ipu-dmfc.o ipu-ic.o ipu-image-convert.o \ + ipu-smfc.o ipu-vdi.o diff --git a/drivers/gpu/ipu-v3/ipu-common.c b/drivers/gpu/ipu-v3/ipu-common.c index d230988ddb8f..b9539f7c5e9a 100644 --- a/drivers/gpu/ipu-v3/ipu-common.c +++ b/drivers/gpu/ipu-v3/ipu-common.c @@ -730,6 +730,137 @@ void ipu_set_ic_src_mux(struct ipu_soc *ipu, int csi_id, bool vdi) } EXPORT_SYMBOL_GPL(ipu_set_ic_src_mux); + +/* Frame Synchronization Unit Channel Linking */ + +struct fsu_link_reg_info { + int chno; + u32 reg; + u32 mask; + u32 val; +}; + +struct fsu_link_info { + struct fsu_link_reg_info src; + struct fsu_link_reg_info sink; +}; + +static const struct fsu_link_info fsu_link_info[] = { + { + .src = { IPUV3_CHANNEL_IC_PRP_ENC_MEM, IPU_FS_PROC_FLOW2, + FS_PRP_ENC_DEST_SEL_MASK, FS_PRP_ENC_DEST_SEL_IRT_ENC }, + .sink = { IPUV3_CHANNEL_MEM_ROT_ENC, IPU_FS_PROC_FLOW1, + FS_PRPENC_ROT_SRC_SEL_MASK, FS_PRPENC_ROT_SRC_SEL_ENC }, + }, { + .src = { IPUV3_CHANNEL_IC_PRP_VF_MEM, IPU_FS_PROC_FLOW2, + FS_PRPVF_DEST_SEL_MASK, FS_PRPVF_DEST_SEL_IRT_VF }, + .sink = { IPUV3_CHANNEL_MEM_ROT_VF, IPU_FS_PROC_FLOW1, + FS_PRPVF_ROT_SRC_SEL_MASK, FS_PRPVF_ROT_SRC_SEL_VF }, + }, { + .src = { IPUV3_CHANNEL_IC_PP_MEM, IPU_FS_PROC_FLOW2, + FS_PP_DEST_SEL_MASK, FS_PP_DEST_SEL_IRT_PP }, + .sink = { IPUV3_CHANNEL_MEM_ROT_PP, IPU_FS_PROC_FLOW1, + FS_PP_ROT_SRC_SEL_MASK, FS_PP_ROT_SRC_SEL_PP }, + }, { + .src = { IPUV3_CHANNEL_CSI_DIRECT, 0 }, + .sink = { IPUV3_CHANNEL_CSI_VDI_PREV, IPU_FS_PROC_FLOW1, + FS_VDI_SRC_SEL_MASK, FS_VDI_SRC_SEL_CSI_DIRECT }, + }, +}; + +static const struct fsu_link_info *find_fsu_link_info(int src, int sink) +{ + int i; + + for (i = 0; i < ARRAY_SIZE(fsu_link_info); i++) { + if (src == fsu_link_info[i].src.chno && + sink == fsu_link_info[i].sink.chno) + return &fsu_link_info[i]; + } + + return NULL; +} + +/* + * Links a source channel to a sink channel in the FSU. + */ +int ipu_fsu_link(struct ipu_soc *ipu, int src_ch, int sink_ch) +{ + const struct fsu_link_info *link; + u32 src_reg, sink_reg; + unsigned long flags; + + link = find_fsu_link_info(src_ch, sink_ch); + if (!link) + return -EINVAL; + + spin_lock_irqsave(&ipu->lock, flags); + + if (link->src.mask) { + src_reg = ipu_cm_read(ipu, link->src.reg); + src_reg &= ~link->src.mask; + src_reg |= link->src.val; + ipu_cm_write(ipu, src_reg, link->src.reg); + } + + if (link->sink.mask) { + sink_reg = ipu_cm_read(ipu, link->sink.reg); + sink_reg &= ~link->sink.mask; + sink_reg |= link->sink.val; + ipu_cm_write(ipu, sink_reg, link->sink.reg); + } + + spin_unlock_irqrestore(&ipu->lock, flags); + return 0; +} +EXPORT_SYMBOL_GPL(ipu_fsu_link); + +/* + * Unlinks source and sink channels in the FSU. + */ +int ipu_fsu_unlink(struct ipu_soc *ipu, int src_ch, int sink_ch) +{ + const struct fsu_link_info *link; + u32 src_reg, sink_reg; + unsigned long flags; + + link = find_fsu_link_info(src_ch, sink_ch); + if (!link) + return -EINVAL; + + spin_lock_irqsave(&ipu->lock, flags); + + if (link->src.mask) { + src_reg = ipu_cm_read(ipu, link->src.reg); + src_reg &= ~link->src.mask; + ipu_cm_write(ipu, src_reg, link->src.reg); + } + + if (link->sink.mask) { + sink_reg = ipu_cm_read(ipu, link->sink.reg); + sink_reg &= ~link->sink.mask; + ipu_cm_write(ipu, sink_reg, link->sink.reg); + } + + spin_unlock_irqrestore(&ipu->lock, flags); + return 0; +} +EXPORT_SYMBOL_GPL(ipu_fsu_unlink); + +/* Link IDMAC channels in the FSU */ +int ipu_idmac_link(struct ipuv3_channel *src, struct ipuv3_channel *sink) +{ + return ipu_fsu_link(src->ipu, src->num, sink->num); +} +EXPORT_SYMBOL_GPL(ipu_idmac_link); + +/* Unlink IDMAC channels in the FSU */ +int ipu_idmac_unlink(struct ipuv3_channel *src, struct ipuv3_channel *sink) +{ + return ipu_fsu_unlink(src->ipu, src->num, sink->num); +} +EXPORT_SYMBOL_GPL(ipu_idmac_unlink); + struct ipu_devtype { const char *name; unsigned long cm_ofs; @@ -839,6 +970,20 @@ static int ipu_submodules_init(struct ipu_soc *ipu, goto err_ic; } + ret = ipu_vdi_init(ipu, dev, ipu_base + devtype->vdi_ofs, + IPU_CONF_VDI_EN | IPU_CONF_ISP_EN | + IPU_CONF_IC_INPUT); + if (ret) { + unit = "vdi"; + goto err_vdi; + } + + ret = ipu_image_convert_init(ipu, dev); + if (ret) { + unit = "image_convert"; + goto err_image_convert; + } + ret = ipu_di_init(ipu, dev, 0, ipu_base + devtype->disp0_ofs, IPU_CONF_DI0_EN, ipu_clk); if (ret) { @@ -893,6 +1038,10 @@ err_dc: err_di_1: ipu_di_exit(ipu, 0); err_di_0: + ipu_image_convert_exit(ipu); +err_image_convert: + ipu_vdi_exit(ipu); +err_vdi: ipu_ic_exit(ipu); err_ic: ipu_csi_exit(ipu, 1); @@ -977,6 +1126,8 @@ static void ipu_submodules_exit(struct ipu_soc *ipu) ipu_dc_exit(ipu); ipu_di_exit(ipu, 1); ipu_di_exit(ipu, 0); + ipu_image_convert_exit(ipu); + ipu_vdi_exit(ipu); ipu_ic_exit(ipu); ipu_csi_exit(ipu, 1); ipu_csi_exit(ipu, 0); @@ -1213,8 +1364,6 @@ EXPORT_SYMBOL_GPL(ipu_dump); static int ipu_probe(struct platform_device *pdev) { - const struct of_device_id *of_id = - of_match_device(imx_ipu_dt_ids, &pdev->dev); struct device_node *np = pdev->dev.of_node; struct ipu_soc *ipu; struct resource *res; @@ -1222,7 +1371,9 @@ static int ipu_probe(struct platform_device *pdev) int i, ret, irq_sync, irq_err; const struct ipu_devtype *devtype; - devtype = of_id->data; + devtype = of_device_get_match_data(&pdev->dev); + if (!devtype) + return -EINVAL; irq_sync = platform_get_irq(pdev, 0); irq_err = platform_get_irq(pdev, 1); diff --git a/drivers/gpu/ipu-v3/ipu-dmfc.c b/drivers/gpu/ipu-v3/ipu-dmfc.c index 42705bb5aaa3..a40f211f382f 100644 --- a/drivers/gpu/ipu-v3/ipu-dmfc.c +++ b/drivers/gpu/ipu-v3/ipu-dmfc.c @@ -123,20 +123,6 @@ int ipu_dmfc_enable_channel(struct dmfc_channel *dmfc) } EXPORT_SYMBOL_GPL(ipu_dmfc_enable_channel); -static void ipu_dmfc_wait_fifos(struct ipu_dmfc_priv *priv) -{ - unsigned long timeout = jiffies + msecs_to_jiffies(1000); - - while ((readl(priv->base + DMFC_STAT) & 0x02fff000) != 0x02fff000) { - if (time_after(jiffies, timeout)) { - dev_warn(priv->dev, - "Timeout waiting for DMFC FIFOs to clear\n"); - break; - } - cpu_relax(); - } -} - void ipu_dmfc_disable_channel(struct dmfc_channel *dmfc) { struct ipu_dmfc_priv *priv = dmfc->priv; @@ -145,10 +131,8 @@ void ipu_dmfc_disable_channel(struct dmfc_channel *dmfc) priv->use_count--; - if (!priv->use_count) { - ipu_dmfc_wait_fifos(priv); + if (!priv->use_count) ipu_module_disable(priv->ipu, IPU_CONF_DMFC_EN); - } if (priv->use_count < 0) priv->use_count = 0; diff --git a/drivers/gpu/ipu-v3/ipu-ic.c b/drivers/gpu/ipu-v3/ipu-ic.c index 1a37afcd85bd..321eb983c2f5 100644 --- a/drivers/gpu/ipu-v3/ipu-ic.c +++ b/drivers/gpu/ipu-v3/ipu-ic.c @@ -619,7 +619,7 @@ int ipu_ic_task_idma_init(struct ipu_ic *ic, struct ipuv3_channel *channel, ipu_ic_write(ic, ic_idmac_2, IC_IDMAC_2); ipu_ic_write(ic, ic_idmac_3, IC_IDMAC_3); - if (rot >= IPU_ROTATE_90_RIGHT) + if (ipu_rot_mode_is_irt(rot)) ic->rotation = true; unlock: diff --git a/drivers/gpu/ipu-v3/ipu-image-convert.c b/drivers/gpu/ipu-v3/ipu-image-convert.c new file mode 100644 index 000000000000..2ba7d437a2af --- /dev/null +++ b/drivers/gpu/ipu-v3/ipu-image-convert.c @@ -0,0 +1,1709 @@ +/* + * Copyright (C) 2012-2016 Mentor Graphics Inc. + * + * Queued image conversion support, with tiling and rotation. + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License as published by the + * Free Software Foundation; either version 2 of the License, or (at your + * option) any later version. + * + * This program is distributed in the hope that it will be useful, but + * WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY + * or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License + * for more details. + */ + +#include <linux/interrupt.h> +#include <linux/dma-mapping.h> +#include <video/imx-ipu-image-convert.h> +#include "ipu-prv.h" + +/* + * The IC Resizer has a restriction that the output frame from the + * resizer must be 1024 or less in both width (pixels) and height + * (lines). + * + * The image converter attempts to split up a conversion when + * the desired output (converted) frame resolution exceeds the + * IC resizer limit of 1024 in either dimension. + * + * If either dimension of the output frame exceeds the limit, the + * dimension is split into 1, 2, or 4 equal stripes, for a maximum + * of 4*4 or 16 tiles. A conversion is then carried out for each + * tile (but taking care to pass the full frame stride length to + * the DMA channel's parameter memory!). IDMA double-buffering is used + * to convert each tile back-to-back when possible (see note below + * when double_buffering boolean is set). + * + * Note that the input frame must be split up into the same number + * of tiles as the output frame. + * + * FIXME: at this point there is no attempt to deal with visible seams + * at the tile boundaries when upscaling. The seams are caused by a reset + * of the bilinear upscale interpolation when starting a new tile. The + * seams are barely visible for small upscale factors, but become + * increasingly visible as the upscale factor gets larger, since more + * interpolated pixels get thrown out at the tile boundaries. A possilble + * fix might be to overlap tiles of different sizes, but this must be done + * while also maintaining the IDMAC dma buffer address alignment and 8x8 IRT + * alignment restrictions of each tile. + */ + +#define MAX_STRIPES_W 4 +#define MAX_STRIPES_H 4 +#define MAX_TILES (MAX_STRIPES_W * MAX_STRIPES_H) + +#define MIN_W 16 +#define MIN_H 8 +#define MAX_W 4096 +#define MAX_H 4096 + +enum ipu_image_convert_type { + IMAGE_CONVERT_IN = 0, + IMAGE_CONVERT_OUT, +}; + +struct ipu_image_convert_dma_buf { + void *virt; + dma_addr_t phys; + unsigned long len; +}; + +struct ipu_image_convert_dma_chan { + int in; + int out; + int rot_in; + int rot_out; + int vdi_in_p; + int vdi_in; + int vdi_in_n; +}; + +/* dimensions of one tile */ +struct ipu_image_tile { + u32 width; + u32 height; + /* size and strides are in bytes */ + u32 size; + u32 stride; + u32 rot_stride; + /* start Y or packed offset of this tile */ + u32 offset; + /* offset from start to tile in U plane, for planar formats */ + u32 u_off; + /* offset from start to tile in V plane, for planar formats */ + u32 v_off; +}; + +struct ipu_image_convert_image { + struct ipu_image base; + enum ipu_image_convert_type type; + + const struct ipu_image_pixfmt *fmt; + unsigned int stride; + + /* # of rows (horizontal stripes) if dest height is > 1024 */ + unsigned int num_rows; + /* # of columns (vertical stripes) if dest width is > 1024 */ + unsigned int num_cols; + + struct ipu_image_tile tile[MAX_TILES]; +}; + +struct ipu_image_pixfmt { + u32 fourcc; /* V4L2 fourcc */ + int bpp; /* total bpp */ + int uv_width_dec; /* decimation in width for U/V planes */ + int uv_height_dec; /* decimation in height for U/V planes */ + bool planar; /* planar format */ + bool uv_swapped; /* U and V planes are swapped */ + bool uv_packed; /* partial planar (U and V in same plane) */ +}; + +struct ipu_image_convert_ctx; +struct ipu_image_convert_chan; +struct ipu_image_convert_priv; + +struct ipu_image_convert_ctx { + struct ipu_image_convert_chan *chan; + + ipu_image_convert_cb_t complete; + void *complete_context; + + /* Source/destination image data and rotation mode */ + struct ipu_image_convert_image in; + struct ipu_image_convert_image out; + enum ipu_rotate_mode rot_mode; + + /* intermediate buffer for rotation */ + struct ipu_image_convert_dma_buf rot_intermediate[2]; + + /* current buffer number for double buffering */ + int cur_buf_num; + + bool aborting; + struct completion aborted; + + /* can we use double-buffering for this conversion operation? */ + bool double_buffering; + /* num_rows * num_cols */ + unsigned int num_tiles; + /* next tile to process */ + unsigned int next_tile; + /* where to place converted tile in dest image */ + unsigned int out_tile_map[MAX_TILES]; + + struct list_head list; +}; + +struct ipu_image_convert_chan { + struct ipu_image_convert_priv *priv; + + enum ipu_ic_task ic_task; + const struct ipu_image_convert_dma_chan *dma_ch; + + struct ipu_ic *ic; + struct ipuv3_channel *in_chan; + struct ipuv3_channel *out_chan; + struct ipuv3_channel *rotation_in_chan; + struct ipuv3_channel *rotation_out_chan; + + /* the IPU end-of-frame irqs */ + int out_eof_irq; + int rot_out_eof_irq; + + spinlock_t irqlock; + + /* list of convert contexts */ + struct list_head ctx_list; + /* queue of conversion runs */ + struct list_head pending_q; + /* queue of completed runs */ + struct list_head done_q; + + /* the current conversion run */ + struct ipu_image_convert_run *current_run; +}; + +struct ipu_image_convert_priv { + struct ipu_image_convert_chan chan[IC_NUM_TASKS]; + struct ipu_soc *ipu; +}; + +static const struct ipu_image_convert_dma_chan +image_convert_dma_chan[IC_NUM_TASKS] = { + [IC_TASK_VIEWFINDER] = { + .in = IPUV3_CHANNEL_MEM_IC_PRP_VF, + .out = IPUV3_CHANNEL_IC_PRP_VF_MEM, + .rot_in = IPUV3_CHANNEL_MEM_ROT_VF, + .rot_out = IPUV3_CHANNEL_ROT_VF_MEM, + .vdi_in_p = IPUV3_CHANNEL_MEM_VDI_PREV, + .vdi_in = IPUV3_CHANNEL_MEM_VDI_CUR, + .vdi_in_n = IPUV3_CHANNEL_MEM_VDI_NEXT, + }, + [IC_TASK_POST_PROCESSOR] = { + .in = IPUV3_CHANNEL_MEM_IC_PP, + .out = IPUV3_CHANNEL_IC_PP_MEM, + .rot_in = IPUV3_CHANNEL_MEM_ROT_PP, + .rot_out = IPUV3_CHANNEL_ROT_PP_MEM, + }, +}; + +static const struct ipu_image_pixfmt image_convert_formats[] = { + { + .fourcc = V4L2_PIX_FMT_RGB565, + .bpp = 16, + }, { + .fourcc = V4L2_PIX_FMT_RGB24, + .bpp = 24, + }, { + .fourcc = V4L2_PIX_FMT_BGR24, + .bpp = 24, + }, { + .fourcc = V4L2_PIX_FMT_RGB32, + .bpp = 32, + }, { + .fourcc = V4L2_PIX_FMT_BGR32, + .bpp = 32, + }, { + .fourcc = V4L2_PIX_FMT_YUYV, + .bpp = 16, + .uv_width_dec = 2, + .uv_height_dec = 1, + }, { + .fourcc = V4L2_PIX_FMT_UYVY, + .bpp = 16, + .uv_width_dec = 2, + .uv_height_dec = 1, + }, { + .fourcc = V4L2_PIX_FMT_YUV420, + .bpp = 12, + .planar = true, + .uv_width_dec = 2, + .uv_height_dec = 2, + }, { + .fourcc = V4L2_PIX_FMT_YVU420, + .bpp = 12, + .planar = true, + .uv_width_dec = 2, + .uv_height_dec = 2, + .uv_swapped = true, + }, { + .fourcc = V4L2_PIX_FMT_NV12, + .bpp = 12, + .planar = true, + .uv_width_dec = 2, + .uv_height_dec = 2, + .uv_packed = true, + }, { + .fourcc = V4L2_PIX_FMT_YUV422P, + .bpp = 16, + .planar = true, + .uv_width_dec = 2, + .uv_height_dec = 1, + }, { + .fourcc = V4L2_PIX_FMT_NV16, + .bpp = 16, + .planar = true, + .uv_width_dec = 2, + .uv_height_dec = 1, + .uv_packed = true, + }, +}; + +static const struct ipu_image_pixfmt *get_format(u32 fourcc) +{ + const struct ipu_image_pixfmt *ret = NULL; + unsigned int i; + + for (i = 0; i < ARRAY_SIZE(image_convert_formats); i++) { + if (image_convert_formats[i].fourcc == fourcc) { + ret = &image_convert_formats[i]; + break; + } + } + + return ret; +} + +static void dump_format(struct ipu_image_convert_ctx *ctx, + struct ipu_image_convert_image *ic_image) +{ + struct ipu_image_convert_chan *chan = ctx->chan; + struct ipu_image_convert_priv *priv = chan->priv; + + dev_dbg(priv->ipu->dev, + "task %u: ctx %p: %s format: %dx%d (%dx%d tiles of size %dx%d), %c%c%c%c\n", + chan->ic_task, ctx, + ic_image->type == IMAGE_CONVERT_OUT ? "Output" : "Input", + ic_image->base.pix.width, ic_image->base.pix.height, + ic_image->num_cols, ic_image->num_rows, + ic_image->tile[0].width, ic_image->tile[0].height, + ic_image->fmt->fourcc & 0xff, + (ic_image->fmt->fourcc >> 8) & 0xff, + (ic_image->fmt->fourcc >> 16) & 0xff, + (ic_image->fmt->fourcc >> 24) & 0xff); +} + +int ipu_image_convert_enum_format(int index, u32 *fourcc) +{ + const struct ipu_image_pixfmt *fmt; + + if (index >= (int)ARRAY_SIZE(image_convert_formats)) + return -EINVAL; + + /* Format found */ + fmt = &image_convert_formats[index]; + *fourcc = fmt->fourcc; + return 0; +} +EXPORT_SYMBOL_GPL(ipu_image_convert_enum_format); + +static void free_dma_buf(struct ipu_image_convert_priv *priv, + struct ipu_image_convert_dma_buf *buf) +{ + if (buf->virt) + dma_free_coherent(priv->ipu->dev, + buf->len, buf->virt, buf->phys); + buf->virt = NULL; + buf->phys = 0; +} + +static int alloc_dma_buf(struct ipu_image_convert_priv *priv, + struct ipu_image_convert_dma_buf *buf, + int size) +{ + buf->len = PAGE_ALIGN(size); + buf->virt = dma_alloc_coherent(priv->ipu->dev, buf->len, &buf->phys, + GFP_DMA | GFP_KERNEL); + if (!buf->virt) { + dev_err(priv->ipu->dev, "failed to alloc dma buffer\n"); + return -ENOMEM; + } + + return 0; +} + +static inline int num_stripes(int dim) +{ + if (dim <= 1024) + return 1; + else if (dim <= 2048) + return 2; + else + return 4; +} + +static void calc_tile_dimensions(struct ipu_image_convert_ctx *ctx, + struct ipu_image_convert_image *image) +{ + int i; + + for (i = 0; i < ctx->num_tiles; i++) { + struct ipu_image_tile *tile = &image->tile[i]; + + tile->height = image->base.pix.height / image->num_rows; + tile->width = image->base.pix.width / image->num_cols; + tile->size = ((tile->height * image->fmt->bpp) >> 3) * + tile->width; + + if (image->fmt->planar) { + tile->stride = tile->width; + tile->rot_stride = tile->height; + } else { + tile->stride = + (image->fmt->bpp * tile->width) >> 3; + tile->rot_stride = + (image->fmt->bpp * tile->height) >> 3; + } + } +} + +/* + * Use the rotation transformation to find the tile coordinates + * (row, col) of a tile in the destination frame that corresponds + * to the given tile coordinates of a source frame. The destination + * coordinate is then converted to a tile index. + */ +static int transform_tile_index(struct ipu_image_convert_ctx *ctx, + int src_row, int src_col) +{ + struct ipu_image_convert_chan *chan = ctx->chan; + struct ipu_image_convert_priv *priv = chan->priv; + struct ipu_image_convert_image *s_image = &ctx->in; + struct ipu_image_convert_image *d_image = &ctx->out; + int dst_row, dst_col; + + /* with no rotation it's a 1:1 mapping */ + if (ctx->rot_mode == IPU_ROTATE_NONE) + return src_row * s_image->num_cols + src_col; + + /* + * before doing the transform, first we have to translate + * source row,col for an origin in the center of s_image + */ + src_row = src_row * 2 - (s_image->num_rows - 1); + src_col = src_col * 2 - (s_image->num_cols - 1); + + /* do the rotation transform */ + if (ctx->rot_mode & IPU_ROT_BIT_90) { + dst_col = -src_row; + dst_row = src_col; + } else { + dst_col = src_col; + dst_row = src_row; + } + + /* apply flip */ + if (ctx->rot_mode & IPU_ROT_BIT_HFLIP) + dst_col = -dst_col; + if (ctx->rot_mode & IPU_ROT_BIT_VFLIP) + dst_row = -dst_row; + + dev_dbg(priv->ipu->dev, "task %u: ctx %p: [%d,%d] --> [%d,%d]\n", + chan->ic_task, ctx, src_col, src_row, dst_col, dst_row); + + /* + * finally translate dest row,col using an origin in upper + * left of d_image + */ + dst_row += d_image->num_rows - 1; + dst_col += d_image->num_cols - 1; + dst_row /= 2; + dst_col /= 2; + + return dst_row * d_image->num_cols + dst_col; +} + +/* + * Fill the out_tile_map[] with transformed destination tile indeces. + */ +static void calc_out_tile_map(struct ipu_image_convert_ctx *ctx) +{ + struct ipu_image_convert_image *s_image = &ctx->in; + unsigned int row, col, tile = 0; + + for (row = 0; row < s_image->num_rows; row++) { + for (col = 0; col < s_image->num_cols; col++) { + ctx->out_tile_map[tile] = + transform_tile_index(ctx, row, col); + tile++; + } + } +} + +static void calc_tile_offsets_planar(struct ipu_image_convert_ctx *ctx, + struct ipu_image_convert_image *image) +{ + struct ipu_image_convert_chan *chan = ctx->chan; + struct ipu_image_convert_priv *priv = chan->priv; + const struct ipu_image_pixfmt *fmt = image->fmt; + unsigned int row, col, tile = 0; + u32 H, w, h, y_stride, uv_stride; + u32 uv_row_off, uv_col_off, uv_off, u_off, v_off, tmp; + u32 y_row_off, y_col_off, y_off; + u32 y_size, uv_size; + + /* setup some convenience vars */ + H = image->base.pix.height; + + y_stride = image->stride; + uv_stride = y_stride / fmt->uv_width_dec; + if (fmt->uv_packed) + uv_stride *= 2; + + y_size = H * y_stride; + uv_size = y_size / (fmt->uv_width_dec * fmt->uv_height_dec); + + for (row = 0; row < image->num_rows; row++) { + w = image->tile[tile].width; + h = image->tile[tile].height; + y_row_off = row * h * y_stride; + uv_row_off = (row * h * uv_stride) / fmt->uv_height_dec; + + for (col = 0; col < image->num_cols; col++) { + y_col_off = col * w; + uv_col_off = y_col_off / fmt->uv_width_dec; + if (fmt->uv_packed) + uv_col_off *= 2; + + y_off = y_row_off + y_col_off; + uv_off = uv_row_off + uv_col_off; + + u_off = y_size - y_off + uv_off; + v_off = (fmt->uv_packed) ? 0 : u_off + uv_size; + if (fmt->uv_swapped) { + tmp = u_off; + u_off = v_off; + v_off = tmp; + } + + image->tile[tile].offset = y_off; + image->tile[tile].u_off = u_off; + image->tile[tile++].v_off = v_off; + + dev_dbg(priv->ipu->dev, + "task %u: ctx %p: %s@[%d,%d]: y_off %08x, u_off %08x, v_off %08x\n", + chan->ic_task, ctx, + image->type == IMAGE_CONVERT_IN ? + "Input" : "Output", row, col, + y_off, u_off, v_off); + } + } +} + +static void calc_tile_offsets_packed(struct ipu_image_convert_ctx *ctx, + struct ipu_image_convert_image *image) +{ + struct ipu_image_convert_chan *chan = ctx->chan; + struct ipu_image_convert_priv *priv = chan->priv; + const struct ipu_image_pixfmt *fmt = image->fmt; + unsigned int row, col, tile = 0; + u32 w, h, bpp, stride; + u32 row_off, col_off; + + /* setup some convenience vars */ + stride = image->stride; + bpp = fmt->bpp; + + for (row = 0; row < image->num_rows; row++) { + w = image->tile[tile].width; + h = image->tile[tile].height; + row_off = row * h * stride; + + for (col = 0; col < image->num_cols; col++) { + col_off = (col * w * bpp) >> 3; + + image->tile[tile].offset = row_off + col_off; + image->tile[tile].u_off = 0; + image->tile[tile++].v_off = 0; + + dev_dbg(priv->ipu->dev, + "task %u: ctx %p: %s@[%d,%d]: phys %08x\n", + chan->ic_task, ctx, + image->type == IMAGE_CONVERT_IN ? + "Input" : "Output", row, col, + row_off + col_off); + } + } +} + +static void calc_tile_offsets(struct ipu_image_convert_ctx *ctx, + struct ipu_image_convert_image *image) +{ + if (image->fmt->planar) + calc_tile_offsets_planar(ctx, image); + else + calc_tile_offsets_packed(ctx, image); +} + +/* + * return the number of runs in given queue (pending_q or done_q) + * for this context. hold irqlock when calling. + */ +static int get_run_count(struct ipu_image_convert_ctx *ctx, + struct list_head *q) +{ + struct ipu_image_convert_run *run; + int count = 0; + + lockdep_assert_held(&ctx->chan->irqlock); + + list_for_each_entry(run, q, list) { + if (run->ctx == ctx) + count++; + } + + return count; +} + +static void convert_stop(struct ipu_image_convert_run *run) +{ + struct ipu_image_convert_ctx *ctx = run->ctx; + struct ipu_image_convert_chan *chan = ctx->chan; + struct ipu_image_convert_priv *priv = chan->priv; + + dev_dbg(priv->ipu->dev, "%s: task %u: stopping ctx %p run %p\n", + __func__, chan->ic_task, ctx, run); + + /* disable IC tasks and the channels */ + ipu_ic_task_disable(chan->ic); + ipu_idmac_disable_channel(chan->in_chan); + ipu_idmac_disable_channel(chan->out_chan); + + if (ipu_rot_mode_is_irt(ctx->rot_mode)) { + ipu_idmac_disable_channel(chan->rotation_in_chan); + ipu_idmac_disable_channel(chan->rotation_out_chan); + ipu_idmac_unlink(chan->out_chan, chan->rotation_in_chan); + } + + ipu_ic_disable(chan->ic); +} + +static void init_idmac_channel(struct ipu_image_convert_ctx *ctx, + struct ipuv3_channel *channel, + struct ipu_image_convert_image *image, + enum ipu_rotate_mode rot_mode, + bool rot_swap_width_height) +{ + struct ipu_image_convert_chan *chan = ctx->chan; + unsigned int burst_size; + u32 width, height, stride; + dma_addr_t addr0, addr1 = 0; + struct ipu_image tile_image; + unsigned int tile_idx[2]; + + if (image->type == IMAGE_CONVERT_OUT) { + tile_idx[0] = ctx->out_tile_map[0]; + tile_idx[1] = ctx->out_tile_map[1]; + } else { + tile_idx[0] = 0; + tile_idx[1] = 1; + } + + if (rot_swap_width_height) { + width = image->tile[0].height; + height = image->tile[0].width; + stride = image->tile[0].rot_stride; + addr0 = ctx->rot_intermediate[0].phys; + if (ctx->double_buffering) + addr1 = ctx->rot_intermediate[1].phys; + } else { + width = image->tile[0].width; + height = image->tile[0].height; + stride = image->stride; + addr0 = image->base.phys0 + + image->tile[tile_idx[0]].offset; + if (ctx->double_buffering) + addr1 = image->base.phys0 + + image->tile[tile_idx[1]].offset; + } + + ipu_cpmem_zero(channel); + + memset(&tile_image, 0, sizeof(tile_image)); + tile_image.pix.width = tile_image.rect.width = width; + tile_image.pix.height = tile_image.rect.height = height; + tile_image.pix.bytesperline = stride; + tile_image.pix.pixelformat = image->fmt->fourcc; + tile_image.phys0 = addr0; + tile_image.phys1 = addr1; + ipu_cpmem_set_image(channel, &tile_image); + + if (image->fmt->planar && !rot_swap_width_height) + ipu_cpmem_set_uv_offset(channel, + image->tile[tile_idx[0]].u_off, + image->tile[tile_idx[0]].v_off); + + if (rot_mode) + ipu_cpmem_set_rotation(channel, rot_mode); + + if (channel == chan->rotation_in_chan || + channel == chan->rotation_out_chan) { + burst_size = 8; + ipu_cpmem_set_block_mode(channel); + } else + burst_size = (width % 16) ? 8 : 16; + + ipu_cpmem_set_burstsize(channel, burst_size); + + ipu_ic_task_idma_init(chan->ic, channel, width, height, + burst_size, rot_mode); + + ipu_cpmem_set_axi_id(channel, 1); + + ipu_idmac_set_double_buffer(channel, ctx->double_buffering); +} + +static int convert_start(struct ipu_image_convert_run *run) +{ + struct ipu_image_convert_ctx *ctx = run->ctx; + struct ipu_image_convert_chan *chan = ctx->chan; + struct ipu_image_convert_priv *priv = chan->priv; + struct ipu_image_convert_image *s_image = &ctx->in; + struct ipu_image_convert_image *d_image = &ctx->out; + enum ipu_color_space src_cs, dest_cs; + unsigned int dest_width, dest_height; + int ret; + + dev_dbg(priv->ipu->dev, "%s: task %u: starting ctx %p run %p\n", + __func__, chan->ic_task, ctx, run); + + src_cs = ipu_pixelformat_to_colorspace(s_image->fmt->fourcc); + dest_cs = ipu_pixelformat_to_colorspace(d_image->fmt->fourcc); + + if (ipu_rot_mode_is_irt(ctx->rot_mode)) { + /* swap width/height for resizer */ + dest_width = d_image->tile[0].height; + dest_height = d_image->tile[0].width; + } else { + dest_width = d_image->tile[0].width; + dest_height = d_image->tile[0].height; + } + + /* setup the IC resizer and CSC */ + ret = ipu_ic_task_init(chan->ic, + s_image->tile[0].width, + s_image->tile[0].height, + dest_width, + dest_height, + src_cs, dest_cs); + if (ret) { + dev_err(priv->ipu->dev, "ipu_ic_task_init failed, %d\n", ret); + return ret; + } + + /* init the source MEM-->IC PP IDMAC channel */ + init_idmac_channel(ctx, chan->in_chan, s_image, + IPU_ROTATE_NONE, false); + + if (ipu_rot_mode_is_irt(ctx->rot_mode)) { + /* init the IC PP-->MEM IDMAC channel */ + init_idmac_channel(ctx, chan->out_chan, d_image, + IPU_ROTATE_NONE, true); + + /* init the MEM-->IC PP ROT IDMAC channel */ + init_idmac_channel(ctx, chan->rotation_in_chan, d_image, + ctx->rot_mode, true); + + /* init the destination IC PP ROT-->MEM IDMAC channel */ + init_idmac_channel(ctx, chan->rotation_out_chan, d_image, + IPU_ROTATE_NONE, false); + + /* now link IC PP-->MEM to MEM-->IC PP ROT */ + ipu_idmac_link(chan->out_chan, chan->rotation_in_chan); + } else { + /* init the destination IC PP-->MEM IDMAC channel */ + init_idmac_channel(ctx, chan->out_chan, d_image, + ctx->rot_mode, false); + } + + /* enable the IC */ + ipu_ic_enable(chan->ic); + + /* set buffers ready */ + ipu_idmac_select_buffer(chan->in_chan, 0); + ipu_idmac_select_buffer(chan->out_chan, 0); + if (ipu_rot_mode_is_irt(ctx->rot_mode)) + ipu_idmac_select_buffer(chan->rotation_out_chan, 0); + if (ctx->double_buffering) { + ipu_idmac_select_buffer(chan->in_chan, 1); + ipu_idmac_select_buffer(chan->out_chan, 1); + if (ipu_rot_mode_is_irt(ctx->rot_mode)) + ipu_idmac_select_buffer(chan->rotation_out_chan, 1); + } + + /* enable the channels! */ + ipu_idmac_enable_channel(chan->in_chan); + ipu_idmac_enable_channel(chan->out_chan); + if (ipu_rot_mode_is_irt(ctx->rot_mode)) { + ipu_idmac_enable_channel(chan->rotation_in_chan); + ipu_idmac_enable_channel(chan->rotation_out_chan); + } + + ipu_ic_task_enable(chan->ic); + + ipu_cpmem_dump(chan->in_chan); + ipu_cpmem_dump(chan->out_chan); + if (ipu_rot_mode_is_irt(ctx->rot_mode)) { + ipu_cpmem_dump(chan->rotation_in_chan); + ipu_cpmem_dump(chan->rotation_out_chan); + } + + ipu_dump(priv->ipu); + + return 0; +} + +/* hold irqlock when calling */ +static int do_run(struct ipu_image_convert_run *run) +{ + struct ipu_image_convert_ctx *ctx = run->ctx; + struct ipu_image_convert_chan *chan = ctx->chan; + + lockdep_assert_held(&chan->irqlock); + + ctx->in.base.phys0 = run->in_phys; + ctx->out.base.phys0 = run->out_phys; + + ctx->cur_buf_num = 0; + ctx->next_tile = 1; + + /* remove run from pending_q and set as current */ + list_del(&run->list); + chan->current_run = run; + + return convert_start(run); +} + +/* hold irqlock when calling */ +static void run_next(struct ipu_image_convert_chan *chan) +{ + struct ipu_image_convert_priv *priv = chan->priv; + struct ipu_image_convert_run *run, *tmp; + int ret; + + lockdep_assert_held(&chan->irqlock); + + list_for_each_entry_safe(run, tmp, &chan->pending_q, list) { + /* skip contexts that are aborting */ + if (run->ctx->aborting) { + dev_dbg(priv->ipu->dev, + "%s: task %u: skipping aborting ctx %p run %p\n", + __func__, chan->ic_task, run->ctx, run); + continue; + } + + ret = do_run(run); + if (!ret) + break; + + /* + * something went wrong with start, add the run + * to done q and continue to the next run in the + * pending q. + */ + run->status = ret; + list_add_tail(&run->list, &chan->done_q); + chan->current_run = NULL; + } +} + +static void empty_done_q(struct ipu_image_convert_chan *chan) +{ + struct ipu_image_convert_priv *priv = chan->priv; + struct ipu_image_convert_run *run; + unsigned long flags; + + spin_lock_irqsave(&chan->irqlock, flags); + + while (!list_empty(&chan->done_q)) { + run = list_entry(chan->done_q.next, + struct ipu_image_convert_run, + list); + + list_del(&run->list); + + dev_dbg(priv->ipu->dev, + "%s: task %u: completing ctx %p run %p with %d\n", + __func__, chan->ic_task, run->ctx, run, run->status); + + /* call the completion callback and free the run */ + spin_unlock_irqrestore(&chan->irqlock, flags); + run->ctx->complete(run, run->ctx->complete_context); + spin_lock_irqsave(&chan->irqlock, flags); + } + + spin_unlock_irqrestore(&chan->irqlock, flags); +} + +/* + * the bottom half thread clears out the done_q, calling the + * completion handler for each. + */ +static irqreturn_t do_bh(int irq, void *dev_id) +{ + struct ipu_image_convert_chan *chan = dev_id; + struct ipu_image_convert_priv *priv = chan->priv; + struct ipu_image_convert_ctx *ctx; + unsigned long flags; + + dev_dbg(priv->ipu->dev, "%s: task %u: enter\n", __func__, + chan->ic_task); + + empty_done_q(chan); + + spin_lock_irqsave(&chan->irqlock, flags); + + /* + * the done_q is cleared out, signal any contexts + * that are aborting that abort can complete. + */ + list_for_each_entry(ctx, &chan->ctx_list, list) { + if (ctx->aborting) { + dev_dbg(priv->ipu->dev, + "%s: task %u: signaling abort for ctx %p\n", + __func__, chan->ic_task, ctx); + complete(&ctx->aborted); + } + } + + spin_unlock_irqrestore(&chan->irqlock, flags); + + dev_dbg(priv->ipu->dev, "%s: task %u: exit\n", __func__, + chan->ic_task); + + return IRQ_HANDLED; +} + +/* hold irqlock when calling */ +static irqreturn_t do_irq(struct ipu_image_convert_run *run) +{ + struct ipu_image_convert_ctx *ctx = run->ctx; + struct ipu_image_convert_chan *chan = ctx->chan; + struct ipu_image_tile *src_tile, *dst_tile; + struct ipu_image_convert_image *s_image = &ctx->in; + struct ipu_image_convert_image *d_image = &ctx->out; + struct ipuv3_channel *outch; + unsigned int dst_idx; + + lockdep_assert_held(&chan->irqlock); + + outch = ipu_rot_mode_is_irt(ctx->rot_mode) ? + chan->rotation_out_chan : chan->out_chan; + + /* + * It is difficult to stop the channel DMA before the channels + * enter the paused state. Without double-buffering the channels + * are always in a paused state when the EOF irq occurs, so it + * is safe to stop the channels now. For double-buffering we + * just ignore the abort until the operation completes, when it + * is safe to shut down. + */ + if (ctx->aborting && !ctx->double_buffering) { + convert_stop(run); + run->status = -EIO; + goto done; + } + + if (ctx->next_tile == ctx->num_tiles) { + /* + * the conversion is complete + */ + convert_stop(run); + run->status = 0; + goto done; + } + + /* + * not done, place the next tile buffers. + */ + if (!ctx->double_buffering) { + + src_tile = &s_image->tile[ctx->next_tile]; + dst_idx = ctx->out_tile_map[ctx->next_tile]; + dst_tile = &d_image->tile[dst_idx]; + + ipu_cpmem_set_buffer(chan->in_chan, 0, + s_image->base.phys0 + src_tile->offset); + ipu_cpmem_set_buffer(outch, 0, + d_image->base.phys0 + dst_tile->offset); + if (s_image->fmt->planar) + ipu_cpmem_set_uv_offset(chan->in_chan, + src_tile->u_off, + src_tile->v_off); + if (d_image->fmt->planar) + ipu_cpmem_set_uv_offset(outch, + dst_tile->u_off, + dst_tile->v_off); + + ipu_idmac_select_buffer(chan->in_chan, 0); + ipu_idmac_select_buffer(outch, 0); + + } else if (ctx->next_tile < ctx->num_tiles - 1) { + + src_tile = &s_image->tile[ctx->next_tile + 1]; + dst_idx = ctx->out_tile_map[ctx->next_tile + 1]; + dst_tile = &d_image->tile[dst_idx]; + + ipu_cpmem_set_buffer(chan->in_chan, ctx->cur_buf_num, + s_image->base.phys0 + src_tile->offset); + ipu_cpmem_set_buffer(outch, ctx->cur_buf_num, + d_image->base.phys0 + dst_tile->offset); + + ipu_idmac_select_buffer(chan->in_chan, ctx->cur_buf_num); + ipu_idmac_select_buffer(outch, ctx->cur_buf_num); + + ctx->cur_buf_num ^= 1; + } + + ctx->next_tile++; + return IRQ_HANDLED; +done: + list_add_tail(&run->list, &chan->done_q); + chan->current_run = NULL; + run_next(chan); + return IRQ_WAKE_THREAD; +} + +static irqreturn_t norotate_irq(int irq, void *data) +{ + struct ipu_image_convert_chan *chan = data; + struct ipu_image_convert_ctx *ctx; + struct ipu_image_convert_run *run; + unsigned long flags; + irqreturn_t ret; + + spin_lock_irqsave(&chan->irqlock, flags); + + /* get current run and its context */ + run = chan->current_run; + if (!run) { + ret = IRQ_NONE; + goto out; + } + + ctx = run->ctx; + + if (ipu_rot_mode_is_irt(ctx->rot_mode)) { + /* this is a rotation operation, just ignore */ + spin_unlock_irqrestore(&chan->irqlock, flags); + return IRQ_HANDLED; + } + + ret = do_irq(run); +out: + spin_unlock_irqrestore(&chan->irqlock, flags); + return ret; +} + +static irqreturn_t rotate_irq(int irq, void *data) +{ + struct ipu_image_convert_chan *chan = data; + struct ipu_image_convert_priv *priv = chan->priv; + struct ipu_image_convert_ctx *ctx; + struct ipu_image_convert_run *run; + unsigned long flags; + irqreturn_t ret; + + spin_lock_irqsave(&chan->irqlock, flags); + + /* get current run and its context */ + run = chan->current_run; + if (!run) { + ret = IRQ_NONE; + goto out; + } + + ctx = run->ctx; + + if (!ipu_rot_mode_is_irt(ctx->rot_mode)) { + /* this was NOT a rotation operation, shouldn't happen */ + dev_err(priv->ipu->dev, "Unexpected rotation interrupt\n"); + spin_unlock_irqrestore(&chan->irqlock, flags); + return IRQ_HANDLED; + } + + ret = do_irq(run); +out: + spin_unlock_irqrestore(&chan->irqlock, flags); + return ret; +} + +/* + * try to force the completion of runs for this ctx. Called when + * abort wait times out in ipu_image_convert_abort(). + */ +static void force_abort(struct ipu_image_convert_ctx *ctx) +{ + struct ipu_image_convert_chan *chan = ctx->chan; + struct ipu_image_convert_run *run; + unsigned long flags; + + spin_lock_irqsave(&chan->irqlock, flags); + + run = chan->current_run; + if (run && run->ctx == ctx) { + convert_stop(run); + run->status = -EIO; + list_add_tail(&run->list, &chan->done_q); + chan->current_run = NULL; + run_next(chan); + } + + spin_unlock_irqrestore(&chan->irqlock, flags); + + empty_done_q(chan); +} + +static void release_ipu_resources(struct ipu_image_convert_chan *chan) +{ + if (chan->out_eof_irq >= 0) + free_irq(chan->out_eof_irq, chan); + if (chan->rot_out_eof_irq >= 0) + free_irq(chan->rot_out_eof_irq, chan); + + if (!IS_ERR_OR_NULL(chan->in_chan)) + ipu_idmac_put(chan->in_chan); + if (!IS_ERR_OR_NULL(chan->out_chan)) + ipu_idmac_put(chan->out_chan); + if (!IS_ERR_OR_NULL(chan->rotation_in_chan)) + ipu_idmac_put(chan->rotation_in_chan); + if (!IS_ERR_OR_NULL(chan->rotation_out_chan)) + ipu_idmac_put(chan->rotation_out_chan); + if (!IS_ERR_OR_NULL(chan->ic)) + ipu_ic_put(chan->ic); + + chan->in_chan = chan->out_chan = chan->rotation_in_chan = + chan->rotation_out_chan = NULL; + chan->out_eof_irq = chan->rot_out_eof_irq = -1; +} + +static int get_ipu_resources(struct ipu_image_convert_chan *chan) +{ + const struct ipu_image_convert_dma_chan *dma = chan->dma_ch; + struct ipu_image_convert_priv *priv = chan->priv; + int ret; + + /* get IC */ + chan->ic = ipu_ic_get(priv->ipu, chan->ic_task); + if (IS_ERR(chan->ic)) { + dev_err(priv->ipu->dev, "could not acquire IC\n"); + ret = PTR_ERR(chan->ic); + goto err; + } + + /* get IDMAC channels */ + chan->in_chan = ipu_idmac_get(priv->ipu, dma->in); + chan->out_chan = ipu_idmac_get(priv->ipu, dma->out); + if (IS_ERR(chan->in_chan) || IS_ERR(chan->out_chan)) { + dev_err(priv->ipu->dev, "could not acquire idmac channels\n"); + ret = -EBUSY; + goto err; + } + + chan->rotation_in_chan = ipu_idmac_get(priv->ipu, dma->rot_in); + chan->rotation_out_chan = ipu_idmac_get(priv->ipu, dma->rot_out); + if (IS_ERR(chan->rotation_in_chan) || IS_ERR(chan->rotation_out_chan)) { + dev_err(priv->ipu->dev, + "could not acquire idmac rotation channels\n"); + ret = -EBUSY; + goto err; + } + + /* acquire the EOF interrupts */ + chan->out_eof_irq = ipu_idmac_channel_irq(priv->ipu, + chan->out_chan, + IPU_IRQ_EOF); + + ret = request_threaded_irq(chan->out_eof_irq, norotate_irq, do_bh, + 0, "ipu-ic", chan); + if (ret < 0) { + dev_err(priv->ipu->dev, "could not acquire irq %d\n", + chan->out_eof_irq); + chan->out_eof_irq = -1; + goto err; + } + + chan->rot_out_eof_irq = ipu_idmac_channel_irq(priv->ipu, + chan->rotation_out_chan, + IPU_IRQ_EOF); + + ret = request_threaded_irq(chan->rot_out_eof_irq, rotate_irq, do_bh, + 0, "ipu-ic", chan); + if (ret < 0) { + dev_err(priv->ipu->dev, "could not acquire irq %d\n", + chan->rot_out_eof_irq); + chan->rot_out_eof_irq = -1; + goto err; + } + + return 0; +err: + release_ipu_resources(chan); + return ret; +} + +static int fill_image(struct ipu_image_convert_ctx *ctx, + struct ipu_image_convert_image *ic_image, + struct ipu_image *image, + enum ipu_image_convert_type type) +{ + struct ipu_image_convert_priv *priv = ctx->chan->priv; + + ic_image->base = *image; + ic_image->type = type; + + ic_image->fmt = get_format(image->pix.pixelformat); + if (!ic_image->fmt) { + dev_err(priv->ipu->dev, "pixelformat not supported for %s\n", + type == IMAGE_CONVERT_OUT ? "Output" : "Input"); + return -EINVAL; + } + + if (ic_image->fmt->planar) + ic_image->stride = ic_image->base.pix.width; + else + ic_image->stride = ic_image->base.pix.bytesperline; + + calc_tile_dimensions(ctx, ic_image); + calc_tile_offsets(ctx, ic_image); + + return 0; +} + +/* borrowed from drivers/media/v4l2-core/v4l2-common.c */ +static unsigned int clamp_align(unsigned int x, unsigned int min, + unsigned int max, unsigned int align) +{ + /* Bits that must be zero to be aligned */ + unsigned int mask = ~((1 << align) - 1); + + /* Clamp to aligned min and max */ + x = clamp(x, (min + ~mask) & mask, max & mask); + + /* Round to nearest aligned value */ + if (align) + x = (x + (1 << (align - 1))) & mask; + + return x; +} + +/* + * We have to adjust the tile width such that the tile physaddrs and + * U and V plane offsets are multiples of 8 bytes as required by + * the IPU DMA Controller. For the planar formats, this corresponds + * to a pixel alignment of 16 (but use a more formal equation since + * the variables are available). For all the packed formats, 8 is + * good enough. + */ +static inline u32 tile_width_align(const struct ipu_image_pixfmt *fmt) +{ + return fmt->planar ? 8 * fmt->uv_width_dec : 8; +} + +/* + * For tile height alignment, we have to ensure that the output tile + * heights are multiples of 8 lines if the IRT is required by the + * given rotation mode (the IRT performs rotations on 8x8 blocks + * at a time). If the IRT is not used, or for input image tiles, + * 2 lines are good enough. + */ +static inline u32 tile_height_align(enum ipu_image_convert_type type, + enum ipu_rotate_mode rot_mode) +{ + return (type == IMAGE_CONVERT_OUT && + ipu_rot_mode_is_irt(rot_mode)) ? 8 : 2; +} + +/* Adjusts input/output images to IPU restrictions */ +void ipu_image_convert_adjust(struct ipu_image *in, struct ipu_image *out, + enum ipu_rotate_mode rot_mode) +{ + const struct ipu_image_pixfmt *infmt, *outfmt; + unsigned int num_in_rows, num_in_cols; + unsigned int num_out_rows, num_out_cols; + u32 w_align, h_align; + + infmt = get_format(in->pix.pixelformat); + outfmt = get_format(out->pix.pixelformat); + + /* set some default pixel formats if needed */ + if (!infmt) { + in->pix.pixelformat = V4L2_PIX_FMT_RGB24; + infmt = get_format(V4L2_PIX_FMT_RGB24); + } + if (!outfmt) { + out->pix.pixelformat = V4L2_PIX_FMT_RGB24; + outfmt = get_format(V4L2_PIX_FMT_RGB24); + } + + /* image converter does not handle fields */ + in->pix.field = out->pix.field = V4L2_FIELD_NONE; + + /* resizer cannot downsize more than 4:1 */ + if (ipu_rot_mode_is_irt(rot_mode)) { + out->pix.height = max_t(__u32, out->pix.height, + in->pix.width / 4); + out->pix.width = max_t(__u32, out->pix.width, + in->pix.height / 4); + } else { + out->pix.width = max_t(__u32, out->pix.width, + in->pix.width / 4); + out->pix.height = max_t(__u32, out->pix.height, + in->pix.height / 4); + } + + /* get tiling rows/cols from output format */ + num_out_rows = num_stripes(out->pix.height); + num_out_cols = num_stripes(out->pix.width); + if (ipu_rot_mode_is_irt(rot_mode)) { + num_in_rows = num_out_cols; + num_in_cols = num_out_rows; + } else { + num_in_rows = num_out_rows; + num_in_cols = num_out_cols; + } + + /* align input width/height */ + w_align = ilog2(tile_width_align(infmt) * num_in_cols); + h_align = ilog2(tile_height_align(IMAGE_CONVERT_IN, rot_mode) * + num_in_rows); + in->pix.width = clamp_align(in->pix.width, MIN_W, MAX_W, w_align); + in->pix.height = clamp_align(in->pix.height, MIN_H, MAX_H, h_align); + + /* align output width/height */ + w_align = ilog2(tile_width_align(outfmt) * num_out_cols); + h_align = ilog2(tile_height_align(IMAGE_CONVERT_OUT, rot_mode) * + num_out_rows); + out->pix.width = clamp_align(out->pix.width, MIN_W, MAX_W, w_align); + out->pix.height = clamp_align(out->pix.height, MIN_H, MAX_H, h_align); + + /* set input/output strides and image sizes */ + in->pix.bytesperline = (in->pix.width * infmt->bpp) >> 3; + in->pix.sizeimage = in->pix.height * in->pix.bytesperline; + out->pix.bytesperline = (out->pix.width * outfmt->bpp) >> 3; + out->pix.sizeimage = out->pix.height * out->pix.bytesperline; +} +EXPORT_SYMBOL_GPL(ipu_image_convert_adjust); + +/* + * this is used by ipu_image_convert_prepare() to verify set input and + * output images are valid before starting the conversion. Clients can + * also call it before calling ipu_image_convert_prepare(). + */ +int ipu_image_convert_verify(struct ipu_image *in, struct ipu_image *out, + enum ipu_rotate_mode rot_mode) +{ + struct ipu_image testin, testout; + + testin = *in; + testout = *out; + + ipu_image_convert_adjust(&testin, &testout, rot_mode); + + if (testin.pix.width != in->pix.width || + testin.pix.height != in->pix.height || + testout.pix.width != out->pix.width || + testout.pix.height != out->pix.height) + return -EINVAL; + + return 0; +} +EXPORT_SYMBOL_GPL(ipu_image_convert_verify); + +/* + * Call ipu_image_convert_prepare() to prepare for the conversion of + * given images and rotation mode. Returns a new conversion context. + */ +struct ipu_image_convert_ctx * +ipu_image_convert_prepare(struct ipu_soc *ipu, enum ipu_ic_task ic_task, + struct ipu_image *in, struct ipu_image *out, + enum ipu_rotate_mode rot_mode, + ipu_image_convert_cb_t complete, + void *complete_context) +{ + struct ipu_image_convert_priv *priv = ipu->image_convert_priv; + struct ipu_image_convert_image *s_image, *d_image; + struct ipu_image_convert_chan *chan; + struct ipu_image_convert_ctx *ctx; + unsigned long flags; + bool get_res; + int ret; + + if (!in || !out || !complete || + (ic_task != IC_TASK_VIEWFINDER && + ic_task != IC_TASK_POST_PROCESSOR)) + return ERR_PTR(-EINVAL); + + /* verify the in/out images before continuing */ + ret = ipu_image_convert_verify(in, out, rot_mode); + if (ret) { + dev_err(priv->ipu->dev, "%s: in/out formats invalid\n", + __func__); + return ERR_PTR(ret); + } + + chan = &priv->chan[ic_task]; + + ctx = kzalloc(sizeof(*ctx), GFP_KERNEL); + if (!ctx) + return ERR_PTR(-ENOMEM); + + dev_dbg(priv->ipu->dev, "%s: task %u: ctx %p\n", __func__, + chan->ic_task, ctx); + + ctx->chan = chan; + init_completion(&ctx->aborted); + + s_image = &ctx->in; + d_image = &ctx->out; + + /* set tiling and rotation */ + d_image->num_rows = num_stripes(out->pix.height); + d_image->num_cols = num_stripes(out->pix.width); + if (ipu_rot_mode_is_irt(rot_mode)) { + s_image->num_rows = d_image->num_cols; + s_image->num_cols = d_image->num_rows; + } else { + s_image->num_rows = d_image->num_rows; + s_image->num_cols = d_image->num_cols; + } + + ctx->num_tiles = d_image->num_cols * d_image->num_rows; + ctx->rot_mode = rot_mode; + + ret = fill_image(ctx, s_image, in, IMAGE_CONVERT_IN); + if (ret) + goto out_free; + ret = fill_image(ctx, d_image, out, IMAGE_CONVERT_OUT); + if (ret) + goto out_free; + + calc_out_tile_map(ctx); + + dump_format(ctx, s_image); + dump_format(ctx, d_image); + + ctx->complete = complete; + ctx->complete_context = complete_context; + + /* + * Can we use double-buffering for this operation? If there is + * only one tile (the whole image can be converted in a single + * operation) there's no point in using double-buffering. Also, + * the IPU's IDMAC channels allow only a single U and V plane + * offset shared between both buffers, but these offsets change + * for every tile, and therefore would have to be updated for + * each buffer which is not possible. So double-buffering is + * impossible when either the source or destination images are + * a planar format (YUV420, YUV422P, etc.). + */ + ctx->double_buffering = (ctx->num_tiles > 1 && + !s_image->fmt->planar && + !d_image->fmt->planar); + + if (ipu_rot_mode_is_irt(ctx->rot_mode)) { + ret = alloc_dma_buf(priv, &ctx->rot_intermediate[0], + d_image->tile[0].size); + if (ret) + goto out_free; + if (ctx->double_buffering) { + ret = alloc_dma_buf(priv, + &ctx->rot_intermediate[1], + d_image->tile[0].size); + if (ret) + goto out_free_dmabuf0; + } + } + + spin_lock_irqsave(&chan->irqlock, flags); + + get_res = list_empty(&chan->ctx_list); + + list_add_tail(&ctx->list, &chan->ctx_list); + + spin_unlock_irqrestore(&chan->irqlock, flags); + + if (get_res) { + ret = get_ipu_resources(chan); + if (ret) + goto out_free_dmabuf1; + } + + return ctx; + +out_free_dmabuf1: + free_dma_buf(priv, &ctx->rot_intermediate[1]); + spin_lock_irqsave(&chan->irqlock, flags); + list_del(&ctx->list); + spin_unlock_irqrestore(&chan->irqlock, flags); +out_free_dmabuf0: + free_dma_buf(priv, &ctx->rot_intermediate[0]); +out_free: + kfree(ctx); + return ERR_PTR(ret); +} +EXPORT_SYMBOL_GPL(ipu_image_convert_prepare); + +/* + * Carry out a single image conversion run. Only the physaddr's of the input + * and output image buffers are needed. The conversion context must have + * been created previously with ipu_image_convert_prepare(). + */ +int ipu_image_convert_queue(struct ipu_image_convert_run *run) +{ + struct ipu_image_convert_chan *chan; + struct ipu_image_convert_priv *priv; + struct ipu_image_convert_ctx *ctx; + unsigned long flags; + int ret = 0; + + if (!run || !run->ctx || !run->in_phys || !run->out_phys) + return -EINVAL; + + ctx = run->ctx; + chan = ctx->chan; + priv = chan->priv; + + dev_dbg(priv->ipu->dev, "%s: task %u: ctx %p run %p\n", __func__, + chan->ic_task, ctx, run); + + INIT_LIST_HEAD(&run->list); + + spin_lock_irqsave(&chan->irqlock, flags); + + if (ctx->aborting) { + ret = -EIO; + goto unlock; + } + + list_add_tail(&run->list, &chan->pending_q); + + if (!chan->current_run) { + ret = do_run(run); + if (ret) + chan->current_run = NULL; + } +unlock: + spin_unlock_irqrestore(&chan->irqlock, flags); + return ret; +} +EXPORT_SYMBOL_GPL(ipu_image_convert_queue); + +/* Abort any active or pending conversions for this context */ +void ipu_image_convert_abort(struct ipu_image_convert_ctx *ctx) +{ + struct ipu_image_convert_chan *chan = ctx->chan; + struct ipu_image_convert_priv *priv = chan->priv; + struct ipu_image_convert_run *run, *active_run, *tmp; + unsigned long flags; + int run_count, ret; + bool need_abort; + + reinit_completion(&ctx->aborted); + + spin_lock_irqsave(&chan->irqlock, flags); + + /* move all remaining pending runs in this context to done_q */ + list_for_each_entry_safe(run, tmp, &chan->pending_q, list) { + if (run->ctx != ctx) + continue; + run->status = -EIO; + list_move_tail(&run->list, &chan->done_q); + } + + run_count = get_run_count(ctx, &chan->done_q); + active_run = (chan->current_run && chan->current_run->ctx == ctx) ? + chan->current_run : NULL; + + need_abort = (run_count || active_run); + + ctx->aborting = need_abort; + + spin_unlock_irqrestore(&chan->irqlock, flags); + + if (!need_abort) { + dev_dbg(priv->ipu->dev, + "%s: task %u: no abort needed for ctx %p\n", + __func__, chan->ic_task, ctx); + return; + } + + dev_dbg(priv->ipu->dev, + "%s: task %u: wait for completion: %d runs, active run %p\n", + __func__, chan->ic_task, run_count, active_run); + + ret = wait_for_completion_timeout(&ctx->aborted, + msecs_to_jiffies(10000)); + if (ret == 0) { + dev_warn(priv->ipu->dev, "%s: timeout\n", __func__); + force_abort(ctx); + } + + ctx->aborting = false; +} +EXPORT_SYMBOL_GPL(ipu_image_convert_abort); + +/* Unprepare image conversion context */ +void ipu_image_convert_unprepare(struct ipu_image_convert_ctx *ctx) +{ + struct ipu_image_convert_chan *chan = ctx->chan; + struct ipu_image_convert_priv *priv = chan->priv; + unsigned long flags; + bool put_res; + + /* make sure no runs are hanging around */ + ipu_image_convert_abort(ctx); + + dev_dbg(priv->ipu->dev, "%s: task %u: removing ctx %p\n", __func__, + chan->ic_task, ctx); + + spin_lock_irqsave(&chan->irqlock, flags); + + list_del(&ctx->list); + + put_res = list_empty(&chan->ctx_list); + + spin_unlock_irqrestore(&chan->irqlock, flags); + + if (put_res) + release_ipu_resources(chan); + + free_dma_buf(priv, &ctx->rot_intermediate[1]); + free_dma_buf(priv, &ctx->rot_intermediate[0]); + + kfree(ctx); +} +EXPORT_SYMBOL_GPL(ipu_image_convert_unprepare); + +/* + * "Canned" asynchronous single image conversion. Allocates and returns + * a new conversion run. On successful return the caller must free the + * run and call ipu_image_convert_unprepare() after conversion completes. + */ +struct ipu_image_convert_run * +ipu_image_convert(struct ipu_soc *ipu, enum ipu_ic_task ic_task, + struct ipu_image *in, struct ipu_image *out, + enum ipu_rotate_mode rot_mode, + ipu_image_convert_cb_t complete, + void *complete_context) +{ + struct ipu_image_convert_ctx *ctx; + struct ipu_image_convert_run *run; + int ret; + + ctx = ipu_image_convert_prepare(ipu, ic_task, in, out, rot_mode, + complete, complete_context); + if (IS_ERR(ctx)) + return ERR_PTR(PTR_ERR(ctx)); + + run = kzalloc(sizeof(*run), GFP_KERNEL); + if (!run) { + ipu_image_convert_unprepare(ctx); + return ERR_PTR(-ENOMEM); + } + + run->ctx = ctx; + run->in_phys = in->phys0; + run->out_phys = out->phys0; + + ret = ipu_image_convert_queue(run); + if (ret) { + ipu_image_convert_unprepare(ctx); + kfree(run); + return ERR_PTR(ret); + } + + return run; +} +EXPORT_SYMBOL_GPL(ipu_image_convert); + +/* "Canned" synchronous single image conversion */ +static void image_convert_sync_complete(struct ipu_image_convert_run *run, + void *data) +{ + struct completion *comp = data; + + complete(comp); +} + +int ipu_image_convert_sync(struct ipu_soc *ipu, enum ipu_ic_task ic_task, + struct ipu_image *in, struct ipu_image *out, + enum ipu_rotate_mode rot_mode) +{ + struct ipu_image_convert_run *run; + struct completion comp; + int ret; + + init_completion(&comp); + + run = ipu_image_convert(ipu, ic_task, in, out, rot_mode, + image_convert_sync_complete, &comp); + if (IS_ERR(run)) + return PTR_ERR(run); + + ret = wait_for_completion_timeout(&comp, msecs_to_jiffies(10000)); + ret = (ret == 0) ? -ETIMEDOUT : 0; + + ipu_image_convert_unprepare(run->ctx); + kfree(run); + + return ret; +} +EXPORT_SYMBOL_GPL(ipu_image_convert_sync); + +int ipu_image_convert_init(struct ipu_soc *ipu, struct device *dev) +{ + struct ipu_image_convert_priv *priv; + int i; + + priv = devm_kzalloc(dev, sizeof(*priv), GFP_KERNEL); + if (!priv) + return -ENOMEM; + + ipu->image_convert_priv = priv; + priv->ipu = ipu; + + for (i = 0; i < IC_NUM_TASKS; i++) { + struct ipu_image_convert_chan *chan = &priv->chan[i]; + + chan->ic_task = i; + chan->priv = priv; + chan->dma_ch = &image_convert_dma_chan[i]; + chan->out_eof_irq = -1; + chan->rot_out_eof_irq = -1; + + spin_lock_init(&chan->irqlock); + INIT_LIST_HEAD(&chan->ctx_list); + INIT_LIST_HEAD(&chan->pending_q); + INIT_LIST_HEAD(&chan->done_q); + } + + return 0; +} + +void ipu_image_convert_exit(struct ipu_soc *ipu) +{ +} diff --git a/drivers/gpu/ipu-v3/ipu-prv.h b/drivers/gpu/ipu-v3/ipu-prv.h index fd47f8f555cd..22e47b68b14a 100644 --- a/drivers/gpu/ipu-v3/ipu-prv.h +++ b/drivers/gpu/ipu-v3/ipu-prv.h @@ -75,6 +75,33 @@ struct ipu_soc; #define IPU_INT_CTRL(n) IPU_CM_REG(0x003C + 4 * (n)) #define IPU_INT_STAT(n) IPU_CM_REG(0x0200 + 4 * (n)) +/* FS_PROC_FLOW1 */ +#define FS_PRPENC_ROT_SRC_SEL_MASK (0xf << 0) +#define FS_PRPENC_ROT_SRC_SEL_ENC (0x7 << 0) +#define FS_PRPVF_ROT_SRC_SEL_MASK (0xf << 8) +#define FS_PRPVF_ROT_SRC_SEL_VF (0x8 << 8) +#define FS_PP_SRC_SEL_MASK (0xf << 12) +#define FS_PP_ROT_SRC_SEL_MASK (0xf << 16) +#define FS_PP_ROT_SRC_SEL_PP (0x5 << 16) +#define FS_VDI1_SRC_SEL_MASK (0x3 << 20) +#define FS_VDI3_SRC_SEL_MASK (0x3 << 20) +#define FS_PRP_SRC_SEL_MASK (0xf << 24) +#define FS_VDI_SRC_SEL_MASK (0x3 << 28) +#define FS_VDI_SRC_SEL_CSI_DIRECT (0x1 << 28) +#define FS_VDI_SRC_SEL_VDOA (0x2 << 28) + +/* FS_PROC_FLOW2 */ +#define FS_PRP_ENC_DEST_SEL_MASK (0xf << 0) +#define FS_PRP_ENC_DEST_SEL_IRT_ENC (0x1 << 0) +#define FS_PRPVF_DEST_SEL_MASK (0xf << 4) +#define FS_PRPVF_DEST_SEL_IRT_VF (0x1 << 4) +#define FS_PRPVF_ROT_DEST_SEL_MASK (0xf << 8) +#define FS_PP_DEST_SEL_MASK (0xf << 12) +#define FS_PP_DEST_SEL_IRT_PP (0x3 << 12) +#define FS_PP_ROT_DEST_SEL_MASK (0xf << 16) +#define FS_PRPENC_ROT_DEST_SEL_MASK (0xf << 20) +#define FS_PRP_DEST_SEL_MASK (0xf << 24) + #define IPU_DI0_COUNTER_RELEASE (1 << 24) #define IPU_DI1_COUNTER_RELEASE (1 << 25) @@ -138,6 +165,8 @@ struct ipu_dc_priv; struct ipu_dmfc_priv; struct ipu_di; struct ipu_ic_priv; +struct ipu_vdi; +struct ipu_image_convert_priv; struct ipu_smfc_priv; struct ipu_devtype; @@ -170,6 +199,8 @@ struct ipu_soc { struct ipu_di *di_priv[2]; struct ipu_csi *csi_priv[2]; struct ipu_ic_priv *ic_priv; + struct ipu_vdi *vdi_priv; + struct ipu_image_convert_priv *image_convert_priv; struct ipu_smfc_priv *smfc_priv; }; @@ -200,6 +231,13 @@ int ipu_ic_init(struct ipu_soc *ipu, struct device *dev, unsigned long base, unsigned long tpmem_base); void ipu_ic_exit(struct ipu_soc *ipu); +int ipu_vdi_init(struct ipu_soc *ipu, struct device *dev, + unsigned long base, u32 module); +void ipu_vdi_exit(struct ipu_soc *ipu); + +int ipu_image_convert_init(struct ipu_soc *ipu, struct device *dev); +void ipu_image_convert_exit(struct ipu_soc *ipu); + int ipu_di_init(struct ipu_soc *ipu, struct device *dev, int id, unsigned long base, u32 module, struct clk *ipu_clk); void ipu_di_exit(struct ipu_soc *ipu, int id); diff --git a/drivers/gpu/ipu-v3/ipu-vdi.c b/drivers/gpu/ipu-v3/ipu-vdi.c new file mode 100644 index 000000000000..f27bf5a12ebc --- /dev/null +++ b/drivers/gpu/ipu-v3/ipu-vdi.c @@ -0,0 +1,243 @@ +/* + * Copyright (C) 2012-2016 Mentor Graphics Inc. + * Copyright (C) 2005-2009 Freescale Semiconductor, Inc. + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License as published by the + * Free Software Foundation; either version 2 of the License, or (at your + * option) any later version. + * + * This program is distributed in the hope that it will be useful, but + * WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY + * or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License + * for more details. + */ +#include <linux/io.h> +#include "ipu-prv.h" + +struct ipu_vdi { + void __iomem *base; + u32 module; + spinlock_t lock; + int use_count; + struct ipu_soc *ipu; +}; + + +/* VDI Register Offsets */ +#define VDI_FSIZE 0x0000 +#define VDI_C 0x0004 + +/* VDI Register Fields */ +#define VDI_C_CH_420 (0 << 1) +#define VDI_C_CH_422 (1 << 1) +#define VDI_C_MOT_SEL_MASK (0x3 << 2) +#define VDI_C_MOT_SEL_FULL (2 << 2) +#define VDI_C_MOT_SEL_LOW (1 << 2) +#define VDI_C_MOT_SEL_MED (0 << 2) +#define VDI_C_BURST_SIZE1_4 (3 << 4) +#define VDI_C_BURST_SIZE2_4 (3 << 8) +#define VDI_C_BURST_SIZE3_4 (3 << 12) +#define VDI_C_BURST_SIZE_MASK 0xF +#define VDI_C_BURST_SIZE1_OFFSET 4 +#define VDI_C_BURST_SIZE2_OFFSET 8 +#define VDI_C_BURST_SIZE3_OFFSET 12 +#define VDI_C_VWM1_SET_1 (0 << 16) +#define VDI_C_VWM1_SET_2 (1 << 16) +#define VDI_C_VWM1_CLR_2 (1 << 19) +#define VDI_C_VWM3_SET_1 (0 << 22) +#define VDI_C_VWM3_SET_2 (1 << 22) +#define VDI_C_VWM3_CLR_2 (1 << 25) +#define VDI_C_TOP_FIELD_MAN_1 (1 << 30) +#define VDI_C_TOP_FIELD_AUTO_1 (1 << 31) + +static inline u32 ipu_vdi_read(struct ipu_vdi *vdi, unsigned int offset) +{ + return readl(vdi->base + offset); +} + +static inline void ipu_vdi_write(struct ipu_vdi *vdi, u32 value, + unsigned int offset) +{ + writel(value, vdi->base + offset); +} + +void ipu_vdi_set_field_order(struct ipu_vdi *vdi, v4l2_std_id std, u32 field) +{ + bool top_field_0 = false; + unsigned long flags; + u32 reg; + + switch (field) { + case V4L2_FIELD_INTERLACED_TB: + case V4L2_FIELD_SEQ_TB: + case V4L2_FIELD_TOP: + top_field_0 = true; + break; + case V4L2_FIELD_INTERLACED_BT: + case V4L2_FIELD_SEQ_BT: + case V4L2_FIELD_BOTTOM: + top_field_0 = false; + break; + default: + top_field_0 = (std & V4L2_STD_525_60) ? true : false; + break; + } + + spin_lock_irqsave(&vdi->lock, flags); + + reg = ipu_vdi_read(vdi, VDI_C); + if (top_field_0) + reg &= ~VDI_C_TOP_FIELD_MAN_1; + else + reg |= VDI_C_TOP_FIELD_MAN_1; + ipu_vdi_write(vdi, reg, VDI_C); + + spin_unlock_irqrestore(&vdi->lock, flags); +} +EXPORT_SYMBOL_GPL(ipu_vdi_set_field_order); + +void ipu_vdi_set_motion(struct ipu_vdi *vdi, enum ipu_motion_sel motion_sel) +{ + unsigned long flags; + u32 reg; + + spin_lock_irqsave(&vdi->lock, flags); + + reg = ipu_vdi_read(vdi, VDI_C); + + reg &= ~VDI_C_MOT_SEL_MASK; + + switch (motion_sel) { + case MED_MOTION: + reg |= VDI_C_MOT_SEL_MED; + break; + case HIGH_MOTION: + reg |= VDI_C_MOT_SEL_FULL; + break; + default: + reg |= VDI_C_MOT_SEL_LOW; + break; + } + + ipu_vdi_write(vdi, reg, VDI_C); + + spin_unlock_irqrestore(&vdi->lock, flags); +} +EXPORT_SYMBOL_GPL(ipu_vdi_set_motion); + +void ipu_vdi_setup(struct ipu_vdi *vdi, u32 code, int xres, int yres) +{ + unsigned long flags; + u32 pixel_fmt, reg; + + spin_lock_irqsave(&vdi->lock, flags); + + reg = ((yres - 1) << 16) | (xres - 1); + ipu_vdi_write(vdi, reg, VDI_FSIZE); + + /* + * Full motion, only vertical filter is used. + * Burst size is 4 accesses + */ + if (code == MEDIA_BUS_FMT_UYVY8_2X8 || + code == MEDIA_BUS_FMT_UYVY8_1X16 || + code == MEDIA_BUS_FMT_YUYV8_2X8 || + code == MEDIA_BUS_FMT_YUYV8_1X16) + pixel_fmt = VDI_C_CH_422; + else + pixel_fmt = VDI_C_CH_420; + + reg = ipu_vdi_read(vdi, VDI_C); + reg |= pixel_fmt; + reg |= VDI_C_BURST_SIZE2_4; + reg |= VDI_C_BURST_SIZE1_4 | VDI_C_VWM1_CLR_2; + reg |= VDI_C_BURST_SIZE3_4 | VDI_C_VWM3_CLR_2; + ipu_vdi_write(vdi, reg, VDI_C); + + spin_unlock_irqrestore(&vdi->lock, flags); +} +EXPORT_SYMBOL_GPL(ipu_vdi_setup); + +void ipu_vdi_unsetup(struct ipu_vdi *vdi) +{ + unsigned long flags; + + spin_lock_irqsave(&vdi->lock, flags); + ipu_vdi_write(vdi, 0, VDI_FSIZE); + ipu_vdi_write(vdi, 0, VDI_C); + spin_unlock_irqrestore(&vdi->lock, flags); +} +EXPORT_SYMBOL_GPL(ipu_vdi_unsetup); + +int ipu_vdi_enable(struct ipu_vdi *vdi) +{ + unsigned long flags; + + spin_lock_irqsave(&vdi->lock, flags); + + if (!vdi->use_count) + ipu_module_enable(vdi->ipu, vdi->module); + + vdi->use_count++; + + spin_unlock_irqrestore(&vdi->lock, flags); + + return 0; +} +EXPORT_SYMBOL_GPL(ipu_vdi_enable); + +int ipu_vdi_disable(struct ipu_vdi *vdi) +{ + unsigned long flags; + + spin_lock_irqsave(&vdi->lock, flags); + + if (vdi->use_count) { + if (!--vdi->use_count) + ipu_module_disable(vdi->ipu, vdi->module); + } + + spin_unlock_irqrestore(&vdi->lock, flags); + + return 0; +} +EXPORT_SYMBOL_GPL(ipu_vdi_disable); + +struct ipu_vdi *ipu_vdi_get(struct ipu_soc *ipu) +{ + return ipu->vdi_priv; +} +EXPORT_SYMBOL_GPL(ipu_vdi_get); + +void ipu_vdi_put(struct ipu_vdi *vdi) +{ +} +EXPORT_SYMBOL_GPL(ipu_vdi_put); + +int ipu_vdi_init(struct ipu_soc *ipu, struct device *dev, + unsigned long base, u32 module) +{ + struct ipu_vdi *vdi; + + vdi = devm_kzalloc(dev, sizeof(*vdi), GFP_KERNEL); + if (!vdi) + return -ENOMEM; + + ipu->vdi_priv = vdi; + + spin_lock_init(&vdi->lock); + vdi->module = module; + vdi->base = devm_ioremap(dev, base, PAGE_SIZE); + if (!vdi->base) + return -ENOMEM; + + dev_dbg(dev, "VDI base: 0x%08lx remapped to %p\n", base, vdi->base); + vdi->ipu = ipu; + + return 0; +} + +void ipu_vdi_exit(struct ipu_soc *ipu) +{ +} |