diff options
Diffstat (limited to 'drivers/gpu')
66 files changed, 549 insertions, 275 deletions
diff --git a/drivers/gpu/drm/drm_ioctl.c b/drivers/gpu/drm/drm_ioctl.c index dffc836144cc..f4dc9b7a3831 100644 --- a/drivers/gpu/drm/drm_ioctl.c +++ b/drivers/gpu/drm/drm_ioctl.c @@ -296,6 +296,18 @@ int drm_getcap(struct drm_device *dev, void *data, struct drm_file *file_priv) case DRM_CAP_ASYNC_PAGE_FLIP: req->value = dev->mode_config.async_page_flip; break; + case DRM_CAP_CURSOR_WIDTH: + if (dev->mode_config.cursor_width) + req->value = dev->mode_config.cursor_width; + else + req->value = 64; + break; + case DRM_CAP_CURSOR_HEIGHT: + if (dev->mode_config.cursor_height) + req->value = dev->mode_config.cursor_height; + else + req->value = 64; + break; default: return -EINVAL; } diff --git a/drivers/gpu/drm/exynos/Kconfig b/drivers/gpu/drm/exynos/Kconfig index f227f544aa36..6e1a1a20cf6b 100644 --- a/drivers/gpu/drm/exynos/Kconfig +++ b/drivers/gpu/drm/exynos/Kconfig @@ -51,7 +51,7 @@ config DRM_EXYNOS_G2D config DRM_EXYNOS_IPP bool "Exynos DRM IPP" - depends on DRM_EXYNOS && !ARCH_MULTIPLATFORM + depends on DRM_EXYNOS help Choose this option if you want to use IPP feature for DRM. @@ -69,6 +69,6 @@ config DRM_EXYNOS_ROTATOR config DRM_EXYNOS_GSC bool "Exynos DRM GSC" - depends on DRM_EXYNOS_IPP && ARCH_EXYNOS5 + depends on DRM_EXYNOS_IPP && ARCH_EXYNOS5 && !ARCH_MULTIPLATFORM help Choose this option if you want to use Exynos GSC for DRM. diff --git a/drivers/gpu/drm/exynos/exynos_drm_drv.c b/drivers/gpu/drm/exynos/exynos_drm_drv.c index 9d096a0c5f8d..215131ab1dd2 100644 --- a/drivers/gpu/drm/exynos/exynos_drm_drv.c +++ b/drivers/gpu/drm/exynos/exynos_drm_drv.c @@ -171,22 +171,24 @@ static int exynos_drm_open(struct drm_device *dev, struct drm_file *file) file->driver_priv = file_priv; ret = exynos_drm_subdrv_open(dev, file); - if (ret) { - kfree(file_priv); - file->driver_priv = NULL; - } + if (ret) + goto out; anon_filp = anon_inode_getfile("exynos_gem", &exynos_drm_gem_fops, NULL, 0); if (IS_ERR(anon_filp)) { - kfree(file_priv); - return PTR_ERR(anon_filp); + ret = PTR_ERR(anon_filp); + goto out; } anon_filp->f_mode = FMODE_READ | FMODE_WRITE; file_priv->anon_filp = anon_filp; return ret; +out: + kfree(file_priv); + file->driver_priv = NULL; + return ret; } static void exynos_drm_preclose(struct drm_device *dev, diff --git a/drivers/gpu/drm/exynos/exynos_drm_g2d.c b/drivers/gpu/drm/exynos/exynos_drm_g2d.c index 380aec28840b..6c1885eedfdf 100644 --- a/drivers/gpu/drm/exynos/exynos_drm_g2d.c +++ b/drivers/gpu/drm/exynos/exynos_drm_g2d.c @@ -607,7 +607,7 @@ static enum g2d_reg_type g2d_get_reg_type(int reg_offset) reg_type = REG_TYPE_NONE; DRM_ERROR("Unknown register offset![%d]\n", reg_offset); break; - }; + } return reg_type; } diff --git a/drivers/gpu/drm/exynos/exynos_drm_ipp.c b/drivers/gpu/drm/exynos/exynos_drm_ipp.c index d519a4e5fe40..09312b877470 100644 --- a/drivers/gpu/drm/exynos/exynos_drm_ipp.c +++ b/drivers/gpu/drm/exynos/exynos_drm_ipp.c @@ -16,7 +16,6 @@ #include <linux/types.h> #include <linux/clk.h> #include <linux/pm_runtime.h> -#include <plat/map-base.h> #include <drm/drmP.h> #include <drm/exynos_drm.h> @@ -826,7 +825,7 @@ static void ipp_put_event(struct drm_exynos_ipp_cmd_node *c_node, DRM_DEBUG_KMS("count[%d]e[0x%x]\n", count++, (int)e); /* - * quf == NULL condition means all event deletion. + * qbuf == NULL condition means all event deletion. * stop operations want to delete all event list. * another case delete only same buf id. */ diff --git a/drivers/gpu/drm/exynos/exynos_hdmi.c b/drivers/gpu/drm/exynos/exynos_hdmi.c index a0e10aeb0e67..c021ddc1ffb4 100644 --- a/drivers/gpu/drm/exynos/exynos_hdmi.c +++ b/drivers/gpu/drm/exynos/exynos_hdmi.c @@ -34,6 +34,7 @@ #include <linux/io.h> #include <linux/of.h> #include <linux/of_gpio.h> +#include <linux/hdmi.h> #include <drm/exynos_drm.h> @@ -59,19 +60,6 @@ #define HDMI_AUI_VERSION 0x01 #define HDMI_AUI_LENGTH 0x0A -/* HDMI infoframe to configure HDMI out packet header, AUI and AVI */ -enum HDMI_PACKET_TYPE { - /* refer to Table 5-8 Packet Type in HDMI specification v1.4a */ - /* InfoFrame packet type */ - HDMI_PACKET_TYPE_INFOFRAME = 0x80, - /* Vendor-Specific InfoFrame */ - HDMI_PACKET_TYPE_VSI = HDMI_PACKET_TYPE_INFOFRAME + 1, - /* Auxiliary Video information InfoFrame */ - HDMI_PACKET_TYPE_AVI = HDMI_PACKET_TYPE_INFOFRAME + 2, - /* Audio information InfoFrame */ - HDMI_PACKET_TYPE_AUI = HDMI_PACKET_TYPE_INFOFRAME + 4 -}; - enum hdmi_type { HDMI_TYPE13, HDMI_TYPE14, @@ -379,12 +367,6 @@ static const struct hdmiphy_config hdmiphy_v14_configs[] = { }, }; -struct hdmi_infoframe { - enum HDMI_PACKET_TYPE type; - u8 ver; - u8 len; -}; - static inline u32 hdmi_reg_read(struct hdmi_context *hdata, u32 reg_id) { return readl(hdata->regs + reg_id); @@ -682,7 +664,7 @@ static u8 hdmi_chksum(struct hdmi_context *hdata, } static void hdmi_reg_infoframe(struct hdmi_context *hdata, - struct hdmi_infoframe *infoframe) + union hdmi_infoframe *infoframe) { u32 hdr_sum; u8 chksum; @@ -700,13 +682,15 @@ static void hdmi_reg_infoframe(struct hdmi_context *hdata, return; } - switch (infoframe->type) { - case HDMI_PACKET_TYPE_AVI: + switch (infoframe->any.type) { + case HDMI_INFOFRAME_TYPE_AVI: hdmi_reg_writeb(hdata, HDMI_AVI_CON, HDMI_AVI_CON_EVERY_VSYNC); - hdmi_reg_writeb(hdata, HDMI_AVI_HEADER0, infoframe->type); - hdmi_reg_writeb(hdata, HDMI_AVI_HEADER1, infoframe->ver); - hdmi_reg_writeb(hdata, HDMI_AVI_HEADER2, infoframe->len); - hdr_sum = infoframe->type + infoframe->ver + infoframe->len; + hdmi_reg_writeb(hdata, HDMI_AVI_HEADER0, infoframe->any.type); + hdmi_reg_writeb(hdata, HDMI_AVI_HEADER1, + infoframe->any.version); + hdmi_reg_writeb(hdata, HDMI_AVI_HEADER2, infoframe->any.length); + hdr_sum = infoframe->any.type + infoframe->any.version + + infoframe->any.length; /* Output format zero hardcoded ,RGB YBCR selection */ hdmi_reg_writeb(hdata, HDMI_AVI_BYTE(1), 0 << 5 | @@ -722,18 +706,20 @@ static void hdmi_reg_infoframe(struct hdmi_context *hdata, hdmi_reg_writeb(hdata, HDMI_AVI_BYTE(4), vic); chksum = hdmi_chksum(hdata, HDMI_AVI_BYTE(1), - infoframe->len, hdr_sum); + infoframe->any.length, hdr_sum); DRM_DEBUG_KMS("AVI checksum = 0x%x\n", chksum); hdmi_reg_writeb(hdata, HDMI_AVI_CHECK_SUM, chksum); break; - case HDMI_PACKET_TYPE_AUI: + case HDMI_INFOFRAME_TYPE_AUDIO: hdmi_reg_writeb(hdata, HDMI_AUI_CON, 0x02); - hdmi_reg_writeb(hdata, HDMI_AUI_HEADER0, infoframe->type); - hdmi_reg_writeb(hdata, HDMI_AUI_HEADER1, infoframe->ver); - hdmi_reg_writeb(hdata, HDMI_AUI_HEADER2, infoframe->len); - hdr_sum = infoframe->type + infoframe->ver + infoframe->len; + hdmi_reg_writeb(hdata, HDMI_AUI_HEADER0, infoframe->any.type); + hdmi_reg_writeb(hdata, HDMI_AUI_HEADER1, + infoframe->any.version); + hdmi_reg_writeb(hdata, HDMI_AUI_HEADER2, infoframe->any.length); + hdr_sum = infoframe->any.type + infoframe->any.version + + infoframe->any.length; chksum = hdmi_chksum(hdata, HDMI_AUI_BYTE(1), - infoframe->len, hdr_sum); + infoframe->any.length, hdr_sum); DRM_DEBUG_KMS("AUI checksum = 0x%x\n", chksum); hdmi_reg_writeb(hdata, HDMI_AUI_CHECK_SUM, chksum); break; @@ -985,7 +971,7 @@ static void hdmi_conf_reset(struct hdmi_context *hdata) static void hdmi_conf_init(struct hdmi_context *hdata) { - struct hdmi_infoframe infoframe; + union hdmi_infoframe infoframe; /* disable HPD interrupts from HDMI IP block, use GPIO instead */ hdmi_reg_writemask(hdata, HDMI_INTC_CON, 0, HDMI_INTC_EN_GLOBAL | @@ -1021,14 +1007,14 @@ static void hdmi_conf_init(struct hdmi_context *hdata) hdmi_reg_writeb(hdata, HDMI_V13_AUI_CON, 0x02); hdmi_reg_writeb(hdata, HDMI_V13_ACR_CON, 0x04); } else { - infoframe.type = HDMI_PACKET_TYPE_AVI; - infoframe.ver = HDMI_AVI_VERSION; - infoframe.len = HDMI_AVI_LENGTH; + infoframe.any.type = HDMI_INFOFRAME_TYPE_AVI; + infoframe.any.version = HDMI_AVI_VERSION; + infoframe.any.length = HDMI_AVI_LENGTH; hdmi_reg_infoframe(hdata, &infoframe); - infoframe.type = HDMI_PACKET_TYPE_AUI; - infoframe.ver = HDMI_AUI_VERSION; - infoframe.len = HDMI_AUI_LENGTH; + infoframe.any.type = HDMI_INFOFRAME_TYPE_AUDIO; + infoframe.any.version = HDMI_AUI_VERSION; + infoframe.any.length = HDMI_AUI_LENGTH; hdmi_reg_infoframe(hdata, &infoframe); /* enable AVI packet every vsync, fixes purple line problem */ diff --git a/drivers/gpu/drm/i2c/tda998x_drv.c b/drivers/gpu/drm/i2c/tda998x_drv.c index 400b0c4a10fb..faa77f543a07 100644 --- a/drivers/gpu/drm/i2c/tda998x_drv.c +++ b/drivers/gpu/drm/i2c/tda998x_drv.c @@ -208,7 +208,7 @@ struct tda998x_priv { # define PLL_SERIAL_1_SRL_IZ(x) (((x) & 3) << 1) # define PLL_SERIAL_1_SRL_MAN_IZ (1 << 6) #define REG_PLL_SERIAL_2 REG(0x02, 0x01) /* read/write */ -# define PLL_SERIAL_2_SRL_NOSC(x) (((x) & 3) << 0) +# define PLL_SERIAL_2_SRL_NOSC(x) ((x) << 0) # define PLL_SERIAL_2_SRL_PR(x) (((x) & 0xf) << 4) #define REG_PLL_SERIAL_3 REG(0x02, 0x02) /* read/write */ # define PLL_SERIAL_3_SRL_CCIR (1 << 0) @@ -528,10 +528,10 @@ tda998x_write_aif(struct drm_encoder *encoder, struct tda998x_encoder_params *p) { uint8_t buf[PB(5) + 1]; + memset(buf, 0, sizeof(buf)); buf[HB(0)] = 0x84; buf[HB(1)] = 0x01; buf[HB(2)] = 10; - buf[PB(0)] = 0; buf[PB(1)] = p->audio_frame[1] & 0x07; /* CC */ buf[PB(2)] = p->audio_frame[2] & 0x1c; /* SF */ buf[PB(4)] = p->audio_frame[4]; @@ -824,6 +824,11 @@ tda998x_encoder_mode_set(struct drm_encoder *encoder, } div = 148500 / mode->clock; + if (div != 0) { + div--; + if (div > 3) + div = 3; + } /* mute the audio FIFO: */ reg_set(encoder, REG_AIP_CNTRL_0, AIP_CNTRL_0_RST_FIFO); @@ -913,7 +918,7 @@ tda998x_encoder_mode_set(struct drm_encoder *encoder, if (priv->rev == TDA19988) { /* let incoming pixels fill the active space (if any) */ - reg_write(encoder, REG_ENABLE_SPACE, 0x01); + reg_write(encoder, REG_ENABLE_SPACE, 0x00); } /* must be last register set: */ @@ -1094,6 +1099,8 @@ tda998x_encoder_destroy(struct drm_encoder *encoder) { struct tda998x_priv *priv = to_tda998x_priv(encoder); drm_i2c_encoder_destroy(encoder); + if (priv->cec) + i2c_unregister_device(priv->cec); kfree(priv); } @@ -1142,8 +1149,12 @@ tda998x_encoder_init(struct i2c_client *client, priv->vip_cntrl_1 = VIP_CNTRL_1_SWAP_C(0) | VIP_CNTRL_1_SWAP_D(1); priv->vip_cntrl_2 = VIP_CNTRL_2_SWAP_E(4) | VIP_CNTRL_2_SWAP_F(5); - priv->current_page = 0; + priv->current_page = 0xff; priv->cec = i2c_new_dummy(client->adapter, 0x34); + if (!priv->cec) { + kfree(priv); + return -ENODEV; + } priv->dpms = DRM_MODE_DPMS_OFF; encoder_slave->slave_priv = priv; diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h index 4a2bf8e3f739..df77e20e3c3d 100644 --- a/drivers/gpu/drm/i915/i915_drv.h +++ b/drivers/gpu/drm/i915/i915_drv.h @@ -1831,6 +1831,14 @@ struct drm_i915_file_private { /* Early gen2 have a totally busted CS tlb and require pinned batches. */ #define HAS_BROKEN_CS_TLB(dev) (IS_I830(dev) || IS_845G(dev)) +/* + * dp aux and gmbus irq on gen4 seems to be able to generate legacy interrupts + * even when in MSI mode. This results in spurious interrupt warnings if the + * legacy irq no. is shared with another device. The kernel then disables that + * 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) /* 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. diff --git a/drivers/gpu/drm/i915/i915_gpu_error.c b/drivers/gpu/drm/i915/i915_gpu_error.c index d7fd2fd2f0a5..990cf8f43efd 100644 --- a/drivers/gpu/drm/i915/i915_gpu_error.c +++ b/drivers/gpu/drm/i915/i915_gpu_error.c @@ -146,7 +146,10 @@ static void i915_error_vprintf(struct drm_i915_error_state_buf *e, va_list tmp; va_copy(tmp, args); - if (!__i915_error_seek(e, vsnprintf(NULL, 0, f, tmp))) + len = vsnprintf(NULL, 0, f, tmp); + va_end(tmp); + + if (!__i915_error_seek(e, len)) return; } diff --git a/drivers/gpu/drm/i915/i915_irq.c b/drivers/gpu/drm/i915/i915_irq.c index 17d8fcb1b6f7..9fec71175571 100644 --- a/drivers/gpu/drm/i915/i915_irq.c +++ b/drivers/gpu/drm/i915/i915_irq.c @@ -567,8 +567,7 @@ static u32 i915_get_vblank_counter(struct drm_device *dev, int pipe) vbl_start = mode->crtc_vblank_start * mode->crtc_htotal; } else { - enum transcoder cpu_transcoder = - intel_pipe_to_cpu_transcoder(dev_priv, pipe); + enum transcoder cpu_transcoder = (enum transcoder) pipe; u32 htotal; htotal = ((I915_READ(HTOTAL(cpu_transcoder)) >> 16) & 0x1fff) + 1; diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index 9fa24347963a..4c1672809493 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -8586,6 +8586,20 @@ static int intel_gen7_queue_flip(struct drm_device *dev, if (ring->id == RCS) len += 6; + /* + * BSpec MI_DISPLAY_FLIP for IVB: + * "The full packet must be contained within the same cache line." + * + * Currently the LRI+SRM+MI_DISPLAY_FLIP all fit within the same + * cacheline, if we ever start emitting more commands before + * the MI_DISPLAY_FLIP we may need to first emit everything else, + * then do the cacheline alignment, and finally emit the + * MI_DISPLAY_FLIP. + */ + ret = intel_ring_cacheline_align(ring); + if (ret) + goto err_unpin; + ret = intel_ring_begin(ring, len); if (ret) goto err_unpin; diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c index 5ede4e8e290d..57552eb386b0 100644 --- a/drivers/gpu/drm/i915/intel_dp.c +++ b/drivers/gpu/drm/i915/intel_dp.c @@ -404,7 +404,7 @@ intel_dp_aux_ch(struct intel_dp *intel_dp, int i, ret, recv_bytes; uint32_t status; int try, precharge, clock = 0; - bool has_aux_irq = true; + bool has_aux_irq = HAS_AUX_IRQ(dev); uint32_t timeout; /* dp aux is extremely sensitive to irq latency, hence request the @@ -537,6 +537,7 @@ intel_dp_aux_native_write(struct intel_dp *intel_dp, uint8_t msg[20]; int msg_bytes; uint8_t ack; + int retry; if (WARN_ON(send_bytes > 16)) return -E2BIG; @@ -548,19 +549,21 @@ intel_dp_aux_native_write(struct intel_dp *intel_dp, msg[3] = send_bytes - 1; memcpy(&msg[4], send, send_bytes); msg_bytes = send_bytes + 4; - for (;;) { + for (retry = 0; retry < 7; retry++) { ret = intel_dp_aux_ch(intel_dp, msg, msg_bytes, &ack, 1); if (ret < 0) return ret; ack >>= 4; if ((ack & DP_AUX_NATIVE_REPLY_MASK) == DP_AUX_NATIVE_REPLY_ACK) - break; + return send_bytes; else if ((ack & DP_AUX_NATIVE_REPLY_MASK) == DP_AUX_NATIVE_REPLY_DEFER) - udelay(100); + usleep_range(400, 500); else return -EIO; } - return send_bytes; + + DRM_ERROR("too many retries, giving up\n"); + return -EIO; } /* Write a single byte to the aux channel in native mode */ @@ -582,6 +585,7 @@ intel_dp_aux_native_read(struct intel_dp *intel_dp, int reply_bytes; uint8_t ack; int ret; + int retry; if (WARN_ON(recv_bytes > 19)) return -E2BIG; @@ -595,7 +599,7 @@ intel_dp_aux_native_read(struct intel_dp *intel_dp, msg_bytes = 4; reply_bytes = recv_bytes + 1; - for (;;) { + for (retry = 0; retry < 7; retry++) { ret = intel_dp_aux_ch(intel_dp, msg, msg_bytes, reply, reply_bytes); if (ret == 0) @@ -608,10 +612,13 @@ intel_dp_aux_native_read(struct intel_dp *intel_dp, return ret - 1; } else if ((ack & DP_AUX_NATIVE_REPLY_MASK) == DP_AUX_NATIVE_REPLY_DEFER) - udelay(100); + usleep_range(400, 500); else return -EIO; } + + DRM_ERROR("too many retries, giving up\n"); + return -EIO; } static int @@ -1869,10 +1876,12 @@ static void vlv_pre_enable_dp(struct intel_encoder *encoder) mutex_unlock(&dev_priv->dpio_lock); - /* init power sequencer on this pipe and port */ - intel_dp_init_panel_power_sequencer(dev, intel_dp, &power_seq); - intel_dp_init_panel_power_sequencer_registers(dev, intel_dp, - &power_seq); + if (is_edp(intel_dp)) { + /* init power sequencer on this pipe and port */ + intel_dp_init_panel_power_sequencer(dev, intel_dp, &power_seq); + intel_dp_init_panel_power_sequencer_registers(dev, intel_dp, + &power_seq); + } intel_enable_dp(encoder); diff --git a/drivers/gpu/drm/i915/intel_i2c.c b/drivers/gpu/drm/i915/intel_i2c.c index b1dc33f47899..d33b61d0dd33 100644 --- a/drivers/gpu/drm/i915/intel_i2c.c +++ b/drivers/gpu/drm/i915/intel_i2c.c @@ -258,13 +258,6 @@ intel_gpio_setup(struct intel_gmbus *bus, u32 pin) algo->data = bus; } -/* - * gmbus on gen4 seems to be able to generate legacy interrupts even when in MSI - * mode. This results in spurious interrupt warnings if the legacy irq no. is - * shared with another device. The kernel then disables that interrupt source - * and so prevents the other device from working properly. - */ -#define HAS_GMBUS_IRQ(dev) (INTEL_INFO(dev)->gen >= 5) static int gmbus_wait_hw_status(struct drm_i915_private *dev_priv, u32 gmbus2_status, diff --git a/drivers/gpu/drm/i915/intel_opregion.c b/drivers/gpu/drm/i915/intel_opregion.c index 4e960ec7419f..acde2945eb8a 100644 --- a/drivers/gpu/drm/i915/intel_opregion.c +++ b/drivers/gpu/drm/i915/intel_opregion.c @@ -226,6 +226,8 @@ struct opregion_asle { #define ACPI_DIGITAL_OUTPUT (3<<8) #define ACPI_LVDS_OUTPUT (4<<8) +#define MAX_DSLP 1500 + #ifdef CONFIG_ACPI static int swsci(struct drm_device *dev, u32 function, u32 parm, u32 *parm_out) { @@ -260,10 +262,11 @@ static int swsci(struct drm_device *dev, u32 function, u32 parm, u32 *parm_out) /* The spec says 2ms should be the default, but it's too small * for some machines. */ dslp = 50; - } else if (dslp > 500) { + } else if (dslp > MAX_DSLP) { /* Hey bios, trust must be earned. */ - WARN_ONCE(1, "excessive driver sleep timeout (DSPL) %u\n", dslp); - dslp = 500; + DRM_INFO_ONCE("ACPI BIOS requests an excessive sleep of %u ms, " + "using %u ms instead\n", dslp, MAX_DSLP); + dslp = MAX_DSLP; } /* The spec tells us to do this, but we are the only user... */ diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c index b7f1742caf87..31b36c5ac894 100644 --- a/drivers/gpu/drm/i915/intel_ringbuffer.c +++ b/drivers/gpu/drm/i915/intel_ringbuffer.c @@ -1653,6 +1653,27 @@ int intel_ring_begin(struct intel_ring_buffer *ring, return 0; } +/* Align the ring tail to a cacheline boundary */ +int intel_ring_cacheline_align(struct intel_ring_buffer *ring) +{ + int num_dwords = (64 - (ring->tail & 63)) / sizeof(uint32_t); + int ret; + + if (num_dwords == 0) + return 0; + + ret = intel_ring_begin(ring, num_dwords); + if (ret) + return ret; + + while (num_dwords--) + intel_ring_emit(ring, MI_NOOP); + + intel_ring_advance(ring); + + return 0; +} + void intel_ring_init_seqno(struct intel_ring_buffer *ring, u32 seqno) { struct drm_i915_private *dev_priv = ring->dev->dev_private; diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.h b/drivers/gpu/drm/i915/intel_ringbuffer.h index 71a73f4fe252..0b243ce33714 100644 --- a/drivers/gpu/drm/i915/intel_ringbuffer.h +++ b/drivers/gpu/drm/i915/intel_ringbuffer.h @@ -233,6 +233,7 @@ intel_write_status_page(struct intel_ring_buffer *ring, void intel_cleanup_ring_buffer(struct intel_ring_buffer *ring); int __must_check intel_ring_begin(struct intel_ring_buffer *ring, int n); +int __must_check intel_ring_cacheline_align(struct intel_ring_buffer *ring); static inline void intel_ring_emit(struct intel_ring_buffer *ring, u32 data) { diff --git a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_crtc.c b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_crtc.c index 1964f4f0d452..84c5b13b33c9 100644 --- a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_crtc.c +++ b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_crtc.c @@ -39,6 +39,7 @@ struct mdp4_crtc { spinlock_t lock; bool stale; uint32_t width, height; + uint32_t x, y; /* next cursor to scan-out: */ uint32_t next_iova; @@ -57,9 +58,16 @@ struct mdp4_crtc { #define PENDING_FLIP 0x2 atomic_t pending; - /* the fb that we currently hold a scanout ref to: */ + /* the fb that we logically (from PoV of KMS API) hold a ref + * to. Which we may not yet be scanning out (we may still + * be scanning out previous in case of page_flip while waiting + * for gpu rendering to complete: + */ struct drm_framebuffer *fb; + /* the fb that we currently hold a scanout ref to: */ + struct drm_framebuffer *scanout_fb; + /* for unref'ing framebuffers after scanout completes: */ struct drm_flip_work unref_fb_work; @@ -77,24 +85,73 @@ static struct mdp4_kms *get_kms(struct drm_crtc *crtc) return to_mdp4_kms(to_mdp_kms(priv->kms)); } -static void update_fb(struct drm_crtc *crtc, bool async, - struct drm_framebuffer *new_fb) +static void request_pending(struct drm_crtc *crtc, uint32_t pending) { struct mdp4_crtc *mdp4_crtc = to_mdp4_crtc(crtc); - struct drm_framebuffer *old_fb = mdp4_crtc->fb; - if (old_fb) - drm_flip_work_queue(&mdp4_crtc->unref_fb_work, old_fb); + atomic_or(pending, &mdp4_crtc->pending); + mdp_irq_register(&get_kms(crtc)->base, &mdp4_crtc->vblank); +} + +static void crtc_flush(struct drm_crtc *crtc) +{ + struct mdp4_crtc *mdp4_crtc = to_mdp4_crtc(crtc); + struct mdp4_kms *mdp4_kms = get_kms(crtc); + uint32_t i, flush = 0; + + for (i = 0; i < ARRAY_SIZE(mdp4_crtc->planes); i++) { + struct drm_plane *plane = mdp4_crtc->planes[i]; + if (plane) { + enum mdp4_pipe pipe_id = mdp4_plane_pipe(plane); + flush |= pipe2flush(pipe_id); + } + } + flush |= ovlp2flush(mdp4_crtc->ovlp); + + DBG("%s: flush=%08x", mdp4_crtc->name, flush); + + mdp4_write(mdp4_kms, REG_MDP4_OVERLAY_FLUSH, flush); +} + +static void update_fb(struct drm_crtc *crtc, struct drm_framebuffer *new_fb) +{ + struct mdp4_crtc *mdp4_crtc = to_mdp4_crtc(crtc); + struct drm_framebuffer *old_fb = mdp4_crtc->fb; /* grab reference to incoming scanout fb: */ drm_framebuffer_reference(new_fb); mdp4_crtc->base.fb = new_fb; mdp4_crtc->fb = new_fb; - if (!async) { - /* enable vblank to pick up the old_fb */ - mdp_irq_register(&get_kms(crtc)->base, &mdp4_crtc->vblank); - } + if (old_fb) + drm_flip_work_queue(&mdp4_crtc->unref_fb_work, old_fb); +} + +/* unlike update_fb(), take a ref to the new scanout fb *before* updating + * plane, then call this. Needed to ensure we don't unref the buffer that + * is actually still being scanned out. + * + * Note that this whole thing goes away with atomic.. since we can defer + * calling into driver until rendering is done. + */ +static void update_scanout(struct drm_crtc *crtc, struct drm_framebuffer *fb) +{ + struct mdp4_crtc *mdp4_crtc = to_mdp4_crtc(crtc); + + /* flush updates, to make sure hw is updated to new scanout fb, + * so that we can safely queue unref to current fb (ie. next + * vblank we know hw is done w/ previous scanout_fb). + */ + crtc_flush(crtc); + + if (mdp4_crtc->scanout_fb) + drm_flip_work_queue(&mdp4_crtc->unref_fb_work, + mdp4_crtc->scanout_fb); + + mdp4_crtc->scanout_fb = fb; + + /* enable vblank to complete flip: */ + request_pending(crtc, PENDING_FLIP); } /* if file!=NULL, this is preclose potential cancel-flip path */ @@ -120,34 +177,6 @@ static void complete_flip(struct drm_crtc *crtc, struct drm_file *file) spin_unlock_irqrestore(&dev->event_lock, flags); } -static void crtc_flush(struct drm_crtc *crtc) -{ - struct mdp4_crtc *mdp4_crtc = to_mdp4_crtc(crtc); - struct mdp4_kms *mdp4_kms = get_kms(crtc); - uint32_t i, flush = 0; - - for (i = 0; i < ARRAY_SIZE(mdp4_crtc->planes); i++) { - struct drm_plane *plane = mdp4_crtc->planes[i]; - if (plane) { - enum mdp4_pipe pipe_id = mdp4_plane_pipe(plane); - flush |= pipe2flush(pipe_id); - } - } - flush |= ovlp2flush(mdp4_crtc->ovlp); - - DBG("%s: flush=%08x", mdp4_crtc->name, flush); - - mdp4_write(mdp4_kms, REG_MDP4_OVERLAY_FLUSH, flush); -} - -static void request_pending(struct drm_crtc *crtc, uint32_t pending) -{ - struct mdp4_crtc *mdp4_crtc = to_mdp4_crtc(crtc); - - atomic_or(pending, &mdp4_crtc->pending); - mdp_irq_register(&get_kms(crtc)->base, &mdp4_crtc->vblank); -} - static void pageflip_cb(struct msm_fence_cb *cb) { struct mdp4_crtc *mdp4_crtc = @@ -158,11 +187,9 @@ static void pageflip_cb(struct msm_fence_cb *cb) if (!fb) return; + drm_framebuffer_reference(fb); mdp4_plane_set_scanout(mdp4_crtc->plane, fb); - crtc_flush(crtc); - - /* enable vblank to complete flip: */ - request_pending(crtc, PENDING_FLIP); + update_scanout(crtc, fb); } static void unref_fb_worker(struct drm_flip_work *work, void *val) @@ -320,6 +347,20 @@ static int mdp4_crtc_mode_set(struct drm_crtc *crtc, mode->vsync_end, mode->vtotal, mode->type, mode->flags); + /* grab extra ref for update_scanout() */ + drm_framebuffer_reference(crtc->fb); + + ret = mdp4_plane_mode_set(mdp4_crtc->plane, crtc, crtc->fb, + 0, 0, mode->hdisplay, mode->vdisplay, + x << 16, y << 16, + mode->hdisplay << 16, mode->vdisplay << 16); + if (ret) { + drm_framebuffer_unreference(crtc->fb); + dev_err(crtc->dev->dev, "%s: failed to set mode on plane: %d\n", + mdp4_crtc->name, ret); + return ret; + } + mdp4_write(mdp4_kms, REG_MDP4_DMA_SRC_SIZE(dma), MDP4_DMA_SRC_SIZE_WIDTH(mode->hdisplay) | MDP4_DMA_SRC_SIZE_HEIGHT(mode->vdisplay)); @@ -341,24 +382,15 @@ static int mdp4_crtc_mode_set(struct drm_crtc *crtc, mdp4_write(mdp4_kms, REG_MDP4_OVLP_CFG(ovlp), 1); - update_fb(crtc, false, crtc->fb); - - ret = mdp4_plane_mode_set(mdp4_crtc->plane, crtc, crtc->fb, - 0, 0, mode->hdisplay, mode->vdisplay, - x << 16, y << 16, - mode->hdisplay << 16, mode->vdisplay << 16); - if (ret) { - dev_err(crtc->dev->dev, "%s: failed to set mode on plane: %d\n", - mdp4_crtc->name, ret); - return ret; - } - if (dma == DMA_E) { mdp4_write(mdp4_kms, REG_MDP4_DMA_E_QUANT(0), 0x00ff0000); mdp4_write(mdp4_kms, REG_MDP4_DMA_E_QUANT(1), 0x00ff0000); mdp4_write(mdp4_kms, REG_MDP4_DMA_E_QUANT(2), 0x00ff0000); } + update_fb(crtc, crtc->fb); + update_scanout(crtc, crtc->fb); + return 0; } @@ -385,13 +417,24 @@ static int mdp4_crtc_mode_set_base(struct drm_crtc *crtc, int x, int y, struct mdp4_crtc *mdp4_crtc = to_mdp4_crtc(crtc); struct drm_plane *plane = mdp4_crtc->plane; struct drm_display_mode *mode = &crtc->mode; + int ret; - update_fb(crtc, false, crtc->fb); + /* grab extra ref for update_scanout() */ + drm_framebuffer_reference(crtc->fb); - return mdp4_plane_mode_set(plane, crtc, crtc->fb, + ret = mdp4_plane_mode_set(plane, crtc, crtc->fb, 0, 0, mode->hdisplay, mode->vdisplay, x << 16, y << 16, mode->hdisplay << 16, mode->vdisplay << 16); + if (ret) { + drm_framebuffer_unreference(crtc->fb); + return ret; + } + + update_fb(crtc, crtc->fb); + update_scanout(crtc, crtc->fb); + + return 0; } static void mdp4_crtc_load_lut(struct drm_crtc *crtc) @@ -419,7 +462,7 @@ static int mdp4_crtc_page_flip(struct drm_crtc *crtc, mdp4_crtc->event = event; spin_unlock_irqrestore(&dev->event_lock, flags); - update_fb(crtc, true, new_fb); + update_fb(crtc, new_fb); return msm_gem_queue_inactive_cb(obj, &mdp4_crtc->pageflip_cb); } @@ -442,12 +485,12 @@ static int mdp4_crtc_set_property(struct drm_crtc *crtc, static void update_cursor(struct drm_crtc *crtc) { struct mdp4_crtc *mdp4_crtc = to_mdp4_crtc(crtc); + struct mdp4_kms *mdp4_kms = get_kms(crtc); enum mdp4_dma dma = mdp4_crtc->dma; unsigned long flags; spin_lock_irqsave(&mdp4_crtc->cursor.lock, flags); if (mdp4_crtc->cursor.stale) { - struct mdp4_kms *mdp4_kms = get_kms(crtc); struct drm_gem_object *next_bo = mdp4_crtc->cursor.next_bo; struct drm_gem_object *prev_bo = mdp4_crtc->cursor.scanout_bo; uint32_t iova = mdp4_crtc->cursor.next_iova; @@ -479,6 +522,11 @@ static void update_cursor(struct drm_crtc *crtc) mdp4_crtc->cursor.scanout_bo = next_bo; mdp4_crtc->cursor.stale = false; } + + mdp4_write(mdp4_kms, REG_MDP4_DMA_CURSOR_POS(dma), + MDP4_DMA_CURSOR_POS_X(mdp4_crtc->cursor.x) | + MDP4_DMA_CURSOR_POS_Y(mdp4_crtc->cursor.y)); + spin_unlock_irqrestore(&mdp4_crtc->cursor.lock, flags); } @@ -530,6 +578,7 @@ static int mdp4_crtc_cursor_set(struct drm_crtc *crtc, drm_gem_object_unreference_unlocked(old_bo); } + crtc_flush(crtc); request_pending(crtc, PENDING_CURSOR); return 0; @@ -542,12 +591,15 @@ fail: static int mdp4_crtc_cursor_move(struct drm_crtc *crtc, int x, int y) { struct mdp4_crtc *mdp4_crtc = to_mdp4_crtc(crtc); - struct mdp4_kms *mdp4_kms = get_kms(crtc); - enum mdp4_dma dma = mdp4_crtc->dma; + unsigned long flags; - mdp4_write(mdp4_kms, REG_MDP4_DMA_CURSOR_POS(dma), - MDP4_DMA_CURSOR_POS_X(x) | - MDP4_DMA_CURSOR_POS_Y(y)); + spin_lock_irqsave(&mdp4_crtc->cursor.lock, flags); + mdp4_crtc->cursor.x = x; + mdp4_crtc->cursor.y = y; + spin_unlock_irqrestore(&mdp4_crtc->cursor.lock, flags); + + crtc_flush(crtc); + request_pending(crtc, PENDING_CURSOR); return 0; } @@ -713,6 +765,7 @@ struct drm_crtc *mdp4_crtc_init(struct drm_device *dev, crtc = &mdp4_crtc->base; mdp4_crtc->plane = plane; + mdp4_crtc->id = id; mdp4_crtc->ovlp = ovlp_id; mdp4_crtc->dma = dma_id; diff --git a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_plane.c b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_plane.c index 2406027200ec..1e893dd13859 100644 --- a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_plane.c +++ b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_plane.c @@ -170,8 +170,8 @@ int mdp4_plane_mode_set(struct drm_plane *plane, MDP4_PIPE_DST_SIZE_HEIGHT(crtc_h)); mdp4_write(mdp4_kms, REG_MDP4_PIPE_DST_XY(pipe), - MDP4_PIPE_SRC_XY_X(crtc_x) | - MDP4_PIPE_SRC_XY_Y(crtc_y)); + MDP4_PIPE_DST_XY_X(crtc_x) | + MDP4_PIPE_DST_XY_Y(crtc_y)); mdp4_plane_set_scanout(plane, fb); diff --git a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_crtc.c b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_crtc.c index 71a3b2345eb3..f2794021f086 100644 --- a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_crtc.c +++ b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_crtc.c @@ -296,6 +296,7 @@ static int mdp5_crtc_mode_set(struct drm_crtc *crtc, x << 16, y << 16, mode->hdisplay << 16, mode->vdisplay << 16); if (ret) { + drm_framebuffer_unreference(crtc->fb); dev_err(crtc->dev->dev, "%s: failed to set mode on plane: %d\n", mdp5_crtc->name, ret); return ret; @@ -343,11 +344,15 @@ static int mdp5_crtc_mode_set_base(struct drm_crtc *crtc, int x, int y, 0, 0, mode->hdisplay, mode->vdisplay, x << 16, y << 16, mode->hdisplay << 16, mode->vdisplay << 16); + if (ret) { + drm_framebuffer_unreference(crtc->fb); + return ret; + } update_fb(crtc, crtc->fb); update_scanout(crtc, crtc->fb); - return ret; + return 0; } static void mdp5_crtc_load_lut(struct drm_crtc *crtc) diff --git a/drivers/gpu/drm/msm/msm_gem.c b/drivers/gpu/drm/msm/msm_gem.c index d8d60c969ac7..3da8264d3039 100644 --- a/drivers/gpu/drm/msm/msm_gem.c +++ b/drivers/gpu/drm/msm/msm_gem.c @@ -644,7 +644,7 @@ struct drm_gem_object *msm_gem_new(struct drm_device *dev, fail: if (obj) - drm_gem_object_unreference_unlocked(obj); + drm_gem_object_unreference(obj); return ERR_PTR(ret); } diff --git a/drivers/gpu/drm/msm/msm_gem_submit.c b/drivers/gpu/drm/msm/msm_gem_submit.c index 5281d4bc37f7..5423e914e491 100644 --- a/drivers/gpu/drm/msm/msm_gem_submit.c +++ b/drivers/gpu/drm/msm/msm_gem_submit.c @@ -163,7 +163,7 @@ retry: /* if locking succeeded, pin bo: */ - ret = msm_gem_get_iova(&msm_obj->base, + ret = msm_gem_get_iova_locked(&msm_obj->base, submit->gpu->id, &iova); /* this would break the logic in the fail path.. there is no @@ -247,7 +247,7 @@ static int submit_reloc(struct msm_gem_submit *submit, struct msm_gem_object *ob /* For now, just map the entire thing. Eventually we probably * to do it page-by-page, w/ kmap() if not vmap()d.. */ - ptr = msm_gem_vaddr(&obj->base); + ptr = msm_gem_vaddr_locked(&obj->base); if (IS_ERR(ptr)) { ret = PTR_ERR(ptr); @@ -307,14 +307,12 @@ static void submit_cleanup(struct msm_gem_submit *submit, bool fail) { unsigned i; - mutex_lock(&submit->dev->struct_mutex); for (i = 0; i < submit->nr_bos; i++) { struct msm_gem_object *msm_obj = submit->bos[i].obj; submit_unlock_unpin_bo(submit, i); list_del_init(&msm_obj->submit_entry); drm_gem_object_unreference(&msm_obj->base); } - mutex_unlock(&submit->dev->struct_mutex); ww_acquire_fini(&submit->ticket); kfree(submit); @@ -342,6 +340,8 @@ int msm_ioctl_gem_submit(struct drm_device *dev, void *data, if (args->nr_cmds > MAX_CMDS) return -EINVAL; + mutex_lock(&dev->struct_mutex); + submit = submit_create(dev, gpu, args->nr_bos); if (!submit) { ret = -ENOMEM; @@ -410,5 +410,6 @@ int msm_ioctl_gem_submit(struct drm_device *dev, void *data, out: if (submit) submit_cleanup(submit, !!ret); + mutex_unlock(&dev->struct_mutex); return ret; } diff --git a/drivers/gpu/drm/msm/msm_gpu.c b/drivers/gpu/drm/msm/msm_gpu.c index 4ebce8be489d..0cfe3f426ee4 100644 --- a/drivers/gpu/drm/msm/msm_gpu.c +++ b/drivers/gpu/drm/msm/msm_gpu.c @@ -298,8 +298,6 @@ int msm_gpu_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit, struct msm_drm_private *priv = dev->dev_private; int i, ret; - mutex_lock(&dev->struct_mutex); - submit->fence = ++priv->next_fence; gpu->submitted_fence = submit->fence; @@ -331,7 +329,6 @@ int msm_gpu_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit, msm_gem_move_to_active(&msm_obj->base, gpu, true, submit->fence); } hangcheck_timer_reset(gpu); - mutex_unlock(&dev->struct_mutex); return ret; } diff --git a/drivers/gpu/drm/nouveau/Makefile b/drivers/gpu/drm/nouveau/Makefile index e88145ba1bf5..d310c195bdfe 100644 --- a/drivers/gpu/drm/nouveau/Makefile +++ b/drivers/gpu/drm/nouveau/Makefile @@ -141,6 +141,7 @@ nouveau-y += core/subdev/mc/base.o nouveau-y += core/subdev/mc/nv04.o nouveau-y += core/subdev/mc/nv40.o nouveau-y += core/subdev/mc/nv44.o +nouveau-y += core/subdev/mc/nv4c.o nouveau-y += core/subdev/mc/nv50.o nouveau-y += core/subdev/mc/nv94.o nouveau-y += core/subdev/mc/nv98.o diff --git a/drivers/gpu/drm/nouveau/core/engine/device/nv40.c b/drivers/gpu/drm/nouveau/core/engine/device/nv40.c index 1b653dd74a70..08b88591ed60 100644 --- a/drivers/gpu/drm/nouveau/core/engine/device/nv40.c +++ b/drivers/gpu/drm/nouveau/core/engine/device/nv40.c @@ -311,7 +311,7 @@ nv40_identify(struct nouveau_device *device) device->oclass[NVDEV_SUBDEV_CLOCK ] = &nv40_clock_oclass; device->oclass[NVDEV_SUBDEV_THERM ] = &nv40_therm_oclass; device->oclass[NVDEV_SUBDEV_DEVINIT] = nv1a_devinit_oclass; - device->oclass[NVDEV_SUBDEV_MC ] = nv44_mc_oclass; + device->oclass[NVDEV_SUBDEV_MC ] = nv4c_mc_oclass; device->oclass[NVDEV_SUBDEV_BUS ] = nv31_bus_oclass; device->oclass[NVDEV_SUBDEV_TIMER ] = &nv04_timer_oclass; device->oclass[NVDEV_SUBDEV_FB ] = nv46_fb_oclass; @@ -334,7 +334,7 @@ nv40_identify(struct nouveau_device *device) device->oclass[NVDEV_SUBDEV_CLOCK ] = &nv40_clock_oclass; device->oclass[NVDEV_SUBDEV_THERM ] = &nv40_therm_oclass; device->oclass[NVDEV_SUBDEV_DEVINIT] = nv1a_devinit_oclass; - device->oclass[NVDEV_SUBDEV_MC ] = nv44_mc_oclass; + device->oclass[NVDEV_SUBDEV_MC ] = nv4c_mc_oclass; device->oclass[NVDEV_SUBDEV_BUS ] = nv31_bus_oclass; device->oclass[NVDEV_SUBDEV_TIMER ] = &nv04_timer_oclass; device->oclass[NVDEV_SUBDEV_FB ] = nv4e_fb_oclass; @@ -357,7 +357,7 @@ nv40_identify(struct nouveau_device *device) device->oclass[NVDEV_SUBDEV_CLOCK ] = &nv40_clock_oclass; device->oclass[NVDEV_SUBDEV_THERM ] = &nv40_therm_oclass; device->oclass[NVDEV_SUBDEV_DEVINIT] = nv1a_devinit_oclass; - device->oclass[NVDEV_SUBDEV_MC ] = nv44_mc_oclass; + device->oclass[NVDEV_SUBDEV_MC ] = nv4c_mc_oclass; device->oclass[NVDEV_SUBDEV_BUS ] = nv31_bus_oclass; device->oclass[NVDEV_SUBDEV_TIMER ] = &nv04_timer_oclass; device->oclass[NVDEV_SUBDEV_FB ] = nv46_fb_oclass; @@ -380,7 +380,7 @@ nv40_identify(struct nouveau_device *device) device->oclass[NVDEV_SUBDEV_CLOCK ] = &nv40_clock_oclass; device->oclass[NVDEV_SUBDEV_THERM ] = &nv40_therm_oclass; device->oclass[NVDEV_SUBDEV_DEVINIT] = nv1a_devinit_oclass; - device->oclass[NVDEV_SUBDEV_MC ] = nv44_mc_oclass; + device->oclass[NVDEV_SUBDEV_MC ] = nv4c_mc_oclass; device->oclass[NVDEV_SUBDEV_BUS ] = nv31_bus_oclass; device->oclass[NVDEV_SUBDEV_TIMER ] = &nv04_timer_oclass; device->oclass[NVDEV_SUBDEV_FB ] = nv46_fb_oclass; @@ -403,7 +403,7 @@ nv40_identify(struct nouveau_device *device) device->oclass[NVDEV_SUBDEV_CLOCK ] = &nv40_clock_oclass; device->oclass[NVDEV_SUBDEV_THERM ] = &nv40_therm_oclass; device->oclass[NVDEV_SUBDEV_DEVINIT] = nv1a_devinit_oclass; - device->oclass[NVDEV_SUBDEV_MC ] = nv44_mc_oclass; + device->oclass[NVDEV_SUBDEV_MC ] = nv4c_mc_oclass; device->oclass[NVDEV_SUBDEV_BUS ] = nv31_bus_oclass; device->oclass[NVDEV_SUBDEV_TIMER ] = &nv04_timer_oclass; device->oclass[NVDEV_SUBDEV_FB ] = nv46_fb_oclass; diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/nv50.c b/drivers/gpu/drm/nouveau/core/engine/disp/nv50.c index 940eaa5d8b9a..9ad722e4e087 100644 --- a/drivers/gpu/drm/nouveau/core/engine/disp/nv50.c +++ b/drivers/gpu/drm/nouveau/core/engine/disp/nv50.c @@ -1142,7 +1142,7 @@ nv50_disp_intr_unk20_2(struct nv50_disp_priv *priv, int head) if (conf != ~0) { if (outp.location == 0 && outp.type == DCB_OUTPUT_DP) { u32 soff = (ffs(outp.or) - 1) * 0x08; - u32 ctrl = nv_rd32(priv, 0x610798 + soff); + u32 ctrl = nv_rd32(priv, 0x610794 + soff); u32 datarate; switch ((ctrl & 0x000f0000) >> 16) { diff --git a/drivers/gpu/drm/nouveau/core/engine/fifo/nve0.c b/drivers/gpu/drm/nouveau/core/engine/fifo/nve0.c index 9a850fe19515..54c1b5b471cd 100644 --- a/drivers/gpu/drm/nouveau/core/engine/fifo/nve0.c +++ b/drivers/gpu/drm/nouveau/core/engine/fifo/nve0.c @@ -112,7 +112,7 @@ nve0_fifo_runlist_update(struct nve0_fifo_priv *priv, u32 engine) nv_wr32(priv, 0x002270, cur->addr >> 12); nv_wr32(priv, 0x002274, (engine << 20) | (p >> 3)); - if (!nv_wait(priv, 0x002284 + (engine * 4), 0x00100000, 0x00000000)) + if (!nv_wait(priv, 0x002284 + (engine * 8), 0x00100000, 0x00000000)) nv_error(priv, "runlist %d update timeout\n", engine); mutex_unlock(&nv_subdev(priv)->mutex); } diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/nv50.c b/drivers/gpu/drm/nouveau/core/engine/graph/nv50.c index 30ed19c52e05..7a367c402978 100644 --- a/drivers/gpu/drm/nouveau/core/engine/graph/nv50.c +++ b/drivers/gpu/drm/nouveau/core/engine/graph/nv50.c @@ -539,7 +539,7 @@ nv50_priv_tp_trap(struct nv50_graph_priv *priv, int type, u32 ustatus_old, ustatus &= ~0x04030000; } if (ustatus && display) { - nv_error("%s - TP%d:", name, i); + nv_error(priv, "%s - TP%d:", name, i); nouveau_bitfield_print(nv50_mpc_traps, ustatus); pr_cont("\n"); ustatus = 0; diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/mc.h b/drivers/gpu/drm/nouveau/core/include/subdev/mc.h index adc88b73d911..3c6738edd127 100644 --- a/drivers/gpu/drm/nouveau/core/include/subdev/mc.h +++ b/drivers/gpu/drm/nouveau/core/include/subdev/mc.h @@ -47,6 +47,7 @@ struct nouveau_mc_oclass { extern struct nouveau_oclass *nv04_mc_oclass; extern struct nouveau_oclass *nv40_mc_oclass; extern struct nouveau_oclass *nv44_mc_oclass; +extern struct nouveau_oclass *nv4c_mc_oclass; extern struct nouveau_oclass *nv50_mc_oclass; extern struct nouveau_oclass *nv94_mc_oclass; extern struct nouveau_oclass *nv98_mc_oclass; diff --git a/drivers/gpu/drm/nouveau/core/subdev/bios/base.c b/drivers/gpu/drm/nouveau/core/subdev/bios/base.c index aa0fbbec7f08..ef0c9c4a8cc3 100644 --- a/drivers/gpu/drm/nouveau/core/subdev/bios/base.c +++ b/drivers/gpu/drm/nouveau/core/subdev/bios/base.c @@ -130,6 +130,10 @@ nouveau_bios_shadow_prom(struct nouveau_bios *bios) u16 pcir; int i; + /* there is no prom on nv4x IGP's */ + if (device->card_type == NV_40 && device->chipset >= 0x4c) + return; + /* enable access to rom */ if (device->card_type >= NV_50) pcireg = 0x088050; diff --git a/drivers/gpu/drm/nouveau/core/subdev/fb/nv1a.c b/drivers/gpu/drm/nouveau/core/subdev/fb/nv1a.c index 9159a5ccee93..265d1253624a 100644 --- a/drivers/gpu/drm/nouveau/core/subdev/fb/nv1a.c +++ b/drivers/gpu/drm/nouveau/core/subdev/fb/nv1a.c @@ -36,7 +36,7 @@ nv1a_fb_oclass = &(struct nv04_fb_impl) { .fini = _nouveau_fb_fini, }, .base.memtype = nv04_fb_memtype_valid, - .base.ram = &nv10_ram_oclass, + .base.ram = &nv1a_ram_oclass, .tile.regions = 8, .tile.init = nv10_fb_tile_init, .tile.fini = nv10_fb_tile_fini, diff --git a/drivers/gpu/drm/nouveau/core/subdev/mc/nv04.h b/drivers/gpu/drm/nouveau/core/subdev/mc/nv04.h index b0d5c31606c1..81a408e7d034 100644 --- a/drivers/gpu/drm/nouveau/core/subdev/mc/nv04.h +++ b/drivers/gpu/drm/nouveau/core/subdev/mc/nv04.h @@ -14,6 +14,7 @@ int nv04_mc_ctor(struct nouveau_object *, struct nouveau_object *, extern const struct nouveau_mc_intr nv04_mc_intr[]; int nv04_mc_init(struct nouveau_object *); void nv40_mc_msi_rearm(struct nouveau_mc *); +int nv44_mc_init(struct nouveau_object *object); int nv50_mc_init(struct nouveau_object *); extern const struct nouveau_mc_intr nv50_mc_intr[]; extern const struct nouveau_mc_intr nvc0_mc_intr[]; diff --git a/drivers/gpu/drm/nouveau/core/subdev/mc/nv44.c b/drivers/gpu/drm/nouveau/core/subdev/mc/nv44.c index 3bfee5c6c4f2..cc4d0d2d886e 100644 --- a/drivers/gpu/drm/nouveau/core/subdev/mc/nv44.c +++ b/drivers/gpu/drm/nouveau/core/subdev/mc/nv44.c @@ -24,7 +24,7 @@ #include "nv04.h" -static int +int nv44_mc_init(struct nouveau_object *object) { struct nv04_mc_priv *priv = (void *)object; diff --git a/drivers/gpu/drm/nouveau/core/subdev/mc/nv4c.c b/drivers/gpu/drm/nouveau/core/subdev/mc/nv4c.c new file mode 100644 index 000000000000..a75c35ccf25c --- /dev/null +++ b/drivers/gpu/drm/nouveau/core/subdev/mc/nv4c.c @@ -0,0 +1,45 @@ +/* + * Copyright 2014 Ilia Mirkin + * + * 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: Ilia Mirkin + */ + +#include "nv04.h" + +static void +nv4c_mc_msi_rearm(struct nouveau_mc *pmc) +{ + struct nv04_mc_priv *priv = (void *)pmc; + nv_wr08(priv, 0x088050, 0xff); +} + +struct nouveau_oclass * +nv4c_mc_oclass = &(struct nouveau_mc_oclass) { + .base.handle = NV_SUBDEV(MC, 0x4c), + .base.ofuncs = &(struct nouveau_ofuncs) { + .ctor = nv04_mc_ctor, + .dtor = _nouveau_mc_dtor, + .init = nv44_mc_init, + .fini = _nouveau_mc_fini, + }, + .intr = nv04_mc_intr, + .msi_rearm = nv4c_mc_msi_rearm, +}.base; diff --git a/drivers/gpu/drm/nouveau/nouveau_bo.c b/drivers/gpu/drm/nouveau/nouveau_bo.c index 488686d490c0..4aed1714b9ab 100644 --- a/drivers/gpu/drm/nouveau/nouveau_bo.c +++ b/drivers/gpu/drm/nouveau/nouveau_bo.c @@ -1249,7 +1249,7 @@ nouveau_ttm_io_mem_reserve(struct ttm_bo_device *bdev, struct ttm_mem_reg *mem) mem->bus.is_iomem = !dev->agp->cant_use_aperture; } #endif - if (!node->memtype) + if (nv_device(drm->device)->card_type < NV_50 || !node->memtype) /* untiled */ break; /* fallthrough, tiled memory */ diff --git a/drivers/gpu/drm/nouveau/nouveau_drm.c b/drivers/gpu/drm/nouveau/nouveau_drm.c index 78c8e7146d56..89c484d8ac26 100644 --- a/drivers/gpu/drm/nouveau/nouveau_drm.c +++ b/drivers/gpu/drm/nouveau/nouveau_drm.c @@ -376,6 +376,8 @@ nouveau_drm_load(struct drm_device *dev, unsigned long flags) if (ret) goto fail_device; + dev->irq_enabled = true; + /* workaround an odd issue on nvc1 by disabling the device's * nosnoop capability. hopefully won't cause issues until a * better fix is found - assuming there is one... @@ -475,6 +477,7 @@ nouveau_drm_remove(struct pci_dev *pdev) struct nouveau_drm *drm = nouveau_drm(dev); struct nouveau_object *device; + dev->irq_enabled = false; device = drm->client.base.device; drm_put_dev(dev); diff --git a/drivers/gpu/drm/nouveau/nouveau_vga.c b/drivers/gpu/drm/nouveau/nouveau_vga.c index 81638d7f2eff..471347edc27e 100644 --- a/drivers/gpu/drm/nouveau/nouveau_vga.c +++ b/drivers/gpu/drm/nouveau/nouveau_vga.c @@ -14,7 +14,9 @@ nouveau_vga_set_decode(void *priv, bool state) { struct nouveau_device *device = nouveau_dev(priv); - if (device->chipset >= 0x40) + if (device->card_type == NV_40 && device->chipset >= 0x4c) + nv_wr32(device, 0x088060, state); + else if (device->chipset >= 0x40) nv_wr32(device, 0x088054, state); else nv_wr32(device, 0x001854, state); diff --git a/drivers/gpu/drm/radeon/atombios_crtc.c b/drivers/gpu/drm/radeon/atombios_crtc.c index a9338c85630f..0d19f4f94d5a 100644 --- a/drivers/gpu/drm/radeon/atombios_crtc.c +++ b/drivers/gpu/drm/radeon/atombios_crtc.c @@ -559,7 +559,7 @@ static u32 atombios_adjust_pll(struct drm_crtc *crtc, u32 adjusted_clock = mode->clock; int encoder_mode = atombios_get_encoder_mode(encoder); u32 dp_clock = mode->clock; - int bpc = radeon_get_monitor_bpc(connector); + int bpc = radeon_crtc->bpc; bool is_duallink = radeon_dig_monitor_is_duallink(encoder, mode->clock); /* reset the pll flags */ @@ -1176,7 +1176,7 @@ static int dce4_crtc_do_set_base(struct drm_crtc *crtc, evergreen_tiling_fields(tiling_flags, &bankw, &bankh, &mtaspect, &tile_split); /* Set NUM_BANKS. */ - if (rdev->family >= CHIP_BONAIRE) { + if (rdev->family >= CHIP_TAHITI) { unsigned tileb, index, num_banks, tile_split_bytes; /* Calculate the macrotile mode index. */ @@ -1194,13 +1194,14 @@ static int dce4_crtc_do_set_base(struct drm_crtc *crtc, return -EINVAL; } - num_banks = (rdev->config.cik.macrotile_mode_array[index] >> 6) & 0x3; + if (rdev->family >= CHIP_BONAIRE) + num_banks = (rdev->config.cik.macrotile_mode_array[index] >> 6) & 0x3; + else + num_banks = (rdev->config.si.tile_mode_array[index] >> 20) & 0x3; fb_format |= EVERGREEN_GRPH_NUM_BANKS(num_banks); } else { - /* SI and older. */ - if (rdev->family >= CHIP_TAHITI) - tmp = rdev->config.si.tile_config; - else if (rdev->family >= CHIP_CAYMAN) + /* NI and older. */ + if (rdev->family >= CHIP_CAYMAN) tmp = rdev->config.cayman.tile_config; else tmp = rdev->config.evergreen.tile_config; diff --git a/drivers/gpu/drm/radeon/atombios_encoders.c b/drivers/gpu/drm/radeon/atombios_encoders.c index a42d61571f49..2cec2ab02f80 100644 --- a/drivers/gpu/drm/radeon/atombios_encoders.c +++ b/drivers/gpu/drm/radeon/atombios_encoders.c @@ -464,11 +464,12 @@ atombios_tv_setup(struct drm_encoder *encoder, int action) static u8 radeon_atom_get_bpc(struct drm_encoder *encoder) { - struct drm_connector *connector = radeon_get_connector_for_encoder(encoder); int bpc = 8; - if (connector) - bpc = radeon_get_monitor_bpc(connector); + if (encoder->crtc) { + struct radeon_crtc *radeon_crtc = to_radeon_crtc(encoder->crtc); + bpc = radeon_crtc->bpc; + } switch (bpc) { case 0: diff --git a/drivers/gpu/drm/radeon/btc_dpm.c b/drivers/gpu/drm/radeon/btc_dpm.c index 0fbd36f3d4e9..ea103ccdf4bd 100644 --- a/drivers/gpu/drm/radeon/btc_dpm.c +++ b/drivers/gpu/drm/radeon/btc_dpm.c @@ -29,6 +29,7 @@ #include "cypress_dpm.h" #include "btc_dpm.h" #include "atom.h" +#include <linux/seq_file.h> #define MC_CG_ARB_FREQ_F0 0x0a #define MC_CG_ARB_FREQ_F1 0x0b @@ -2756,6 +2757,37 @@ void btc_dpm_fini(struct radeon_device *rdev) r600_free_extended_power_table(rdev); } +void btc_dpm_debugfs_print_current_performance_level(struct radeon_device *rdev, + struct seq_file *m) +{ + struct evergreen_power_info *eg_pi = evergreen_get_pi(rdev); + struct radeon_ps *rps = &eg_pi->current_rps; + struct rv7xx_ps *ps = rv770_get_ps(rps); + struct rv7xx_pl *pl; + u32 current_index = + (RREG32(TARGET_AND_CURRENT_PROFILE_INDEX) & CURRENT_PROFILE_INDEX_MASK) >> + CURRENT_PROFILE_INDEX_SHIFT; + + if (current_index > 2) { + seq_printf(m, "invalid dpm profile %d\n", current_index); + } else { + if (current_index == 0) + pl = &ps->low; + else if (current_index == 1) + pl = &ps->medium; + else /* current_index == 2 */ + pl = &ps->high; + seq_printf(m, "uvd vclk: %d dclk: %d\n", rps->vclk, rps->dclk); + if (rdev->family >= CHIP_CEDAR) { + seq_printf(m, "power level %d sclk: %u mclk: %u vddc: %u vddci: %u\n", + current_index, pl->sclk, pl->mclk, pl->vddc, pl->vddci); + } else { + seq_printf(m, "power level %d sclk: %u mclk: %u vddc: %u\n", + current_index, pl->sclk, pl->mclk, pl->vddc); + } + } +} + u32 btc_dpm_get_sclk(struct radeon_device *rdev, bool low) { struct evergreen_power_info *eg_pi = evergreen_get_pi(rdev); diff --git a/drivers/gpu/drm/radeon/btcd.h b/drivers/gpu/drm/radeon/btcd.h index 29e32de7e025..9c65be2d55a9 100644 --- a/drivers/gpu/drm/radeon/btcd.h +++ b/drivers/gpu/drm/radeon/btcd.h @@ -44,6 +44,10 @@ # define DYN_SPREAD_SPECTRUM_EN (1 << 23) # define AC_DC_SW (1 << 24) +#define TARGET_AND_CURRENT_PROFILE_INDEX 0x66c +# define CURRENT_PROFILE_INDEX_MASK (0xf << 4) +# define CURRENT_PROFILE_INDEX_SHIFT 4 + #define CG_BIF_REQ_AND_RSP 0x7f4 #define CG_CLIENT_REQ(x) ((x) << 0) #define CG_CLIENT_REQ_MASK (0xff << 0) diff --git a/drivers/gpu/drm/radeon/evergreen.c b/drivers/gpu/drm/radeon/evergreen.c index f2b9e21ce4da..5623e7542d99 100644 --- a/drivers/gpu/drm/radeon/evergreen.c +++ b/drivers/gpu/drm/radeon/evergreen.c @@ -1680,7 +1680,7 @@ bool evergreen_hpd_sense(struct radeon_device *rdev, enum radeon_hpd_id hpd) case RADEON_HPD_6: if (RREG32(DC_HPD6_INT_STATUS) & DC_HPDx_SENSE) connected = true; - break; + break; default: break; } diff --git a/drivers/gpu/drm/radeon/kv_dpm.c b/drivers/gpu/drm/radeon/kv_dpm.c index b6e01d5d2cce..351db361239d 100644 --- a/drivers/gpu/drm/radeon/kv_dpm.c +++ b/drivers/gpu/drm/radeon/kv_dpm.c @@ -1223,7 +1223,7 @@ int kv_dpm_enable(struct radeon_device *rdev) int kv_dpm_late_enable(struct radeon_device *rdev) { - int ret; + int ret = 0; if (rdev->irq.installed && r600_is_internal_thermal_sensor(rdev->pm.int_thermal_type)) { diff --git a/drivers/gpu/drm/radeon/ni_dpm.c b/drivers/gpu/drm/radeon/ni_dpm.c index c351226ecb31..ca814276b075 100644 --- a/drivers/gpu/drm/radeon/ni_dpm.c +++ b/drivers/gpu/drm/radeon/ni_dpm.c @@ -2588,7 +2588,7 @@ static int ni_populate_sq_ramping_values(struct radeon_device *rdev, if (NISLANDS_DPM2_SQ_RAMP_STI_SIZE > (STI_SIZE_MASK >> STI_SIZE_SHIFT)) enable_sq_ramping = false; - if (NISLANDS_DPM2_SQ_RAMP_LTI_RATIO <= (LTI_RATIO_MASK >> LTI_RATIO_SHIFT)) + if (NISLANDS_DPM2_SQ_RAMP_LTI_RATIO > (LTI_RATIO_MASK >> LTI_RATIO_SHIFT)) enable_sq_ramping = false; for (i = 0; i < state->performance_level_count; i++) { @@ -3945,7 +3945,6 @@ static void ni_parse_pplib_clock_info(struct radeon_device *rdev, struct rv7xx_power_info *pi = rv770_get_pi(rdev); struct evergreen_power_info *eg_pi = evergreen_get_pi(rdev); struct ni_ps *ps = ni_get_ps(rps); - u16 vddc; struct rv7xx_pl *pl = &ps->performance_levels[index]; ps->performance_level_count = index + 1; @@ -3961,8 +3960,8 @@ static void ni_parse_pplib_clock_info(struct radeon_device *rdev, /* patch up vddc if necessary */ if (pl->vddc == 0xff01) { - if (radeon_atom_get_max_vddc(rdev, 0, 0, &vddc) == 0) - pl->vddc = vddc; + if (pi->max_vddc) + pl->vddc = pi->max_vddc; } if (rps->class & ATOM_PPLIB_CLASSIFICATION_ACPI) { @@ -4322,7 +4321,8 @@ void ni_dpm_print_power_state(struct radeon_device *rdev, void ni_dpm_debugfs_print_current_performance_level(struct radeon_device *rdev, struct seq_file *m) { - struct radeon_ps *rps = rdev->pm.dpm.current_ps; + struct evergreen_power_info *eg_pi = evergreen_get_pi(rdev); + struct radeon_ps *rps = &eg_pi->current_rps; struct ni_ps *ps = ni_get_ps(rps); struct rv7xx_pl *pl; u32 current_index = diff --git a/drivers/gpu/drm/radeon/r600.c b/drivers/gpu/drm/radeon/r600.c index 56140b4e5bb2..cdbc4171fe73 100644 --- a/drivers/gpu/drm/radeon/r600.c +++ b/drivers/gpu/drm/radeon/r600.c @@ -3991,6 +3991,10 @@ restart_ih: break; } break; + case 124: /* UVD */ + DRM_DEBUG("IH: UVD int: 0x%08x\n", src_data); + radeon_fence_process(rdev, R600_RING_TYPE_UVD_INDEX); + break; case 176: /* CP_INT in ring buffer */ case 177: /* CP_INT in IB1 */ case 178: /* CP_INT in IB2 */ diff --git a/drivers/gpu/drm/radeon/radeon.h b/drivers/gpu/drm/radeon/radeon.h index 4a8ac1cd6b4c..024db37b1832 100644 --- a/drivers/gpu/drm/radeon/radeon.h +++ b/drivers/gpu/drm/radeon/radeon.h @@ -135,6 +135,9 @@ extern int radeon_hard_reset; /* R600+ */ #define R600_RING_TYPE_UVD_INDEX 5 +/* number of hw syncs before falling back on blocking */ +#define RADEON_NUM_SYNCS 4 + /* hardcode those limit for now */ #define RADEON_VA_IB_OFFSET (1 << 20) #define RADEON_VA_RESERVED_SIZE (8 << 20) @@ -554,7 +557,6 @@ int radeon_mode_dumb_mmap(struct drm_file *filp, /* * Semaphores. */ -/* everything here is constant */ struct radeon_semaphore { struct radeon_sa_bo *sa_bo; signed waiters; diff --git a/drivers/gpu/drm/radeon/radeon_asic.c b/drivers/gpu/drm/radeon/radeon_asic.c index f74db43346fd..dda02bfc10a4 100644 --- a/drivers/gpu/drm/radeon/radeon_asic.c +++ b/drivers/gpu/drm/radeon/radeon_asic.c @@ -1555,7 +1555,7 @@ static struct radeon_asic btc_asic = { .get_sclk = &btc_dpm_get_sclk, .get_mclk = &btc_dpm_get_mclk, .print_power_state = &rv770_dpm_print_power_state, - .debugfs_print_current_performance_level = &rv770_dpm_debugfs_print_current_performance_level, + .debugfs_print_current_performance_level = &btc_dpm_debugfs_print_current_performance_level, .force_performance_level = &rv770_dpm_force_performance_level, .vblank_too_short = &btc_dpm_vblank_too_short, }, diff --git a/drivers/gpu/drm/radeon/radeon_asic.h b/drivers/gpu/drm/radeon/radeon_asic.h index b3bc433eed4c..ae637cfda783 100644 --- a/drivers/gpu/drm/radeon/radeon_asic.h +++ b/drivers/gpu/drm/radeon/radeon_asic.h @@ -551,6 +551,8 @@ void btc_dpm_fini(struct radeon_device *rdev); u32 btc_dpm_get_sclk(struct radeon_device *rdev, bool low); u32 btc_dpm_get_mclk(struct radeon_device *rdev, bool low); bool btc_dpm_vblank_too_short(struct radeon_device *rdev); +void btc_dpm_debugfs_print_current_performance_level(struct radeon_device *rdev, + struct seq_file *m); int sumo_dpm_init(struct radeon_device *rdev); int sumo_dpm_enable(struct radeon_device *rdev); int sumo_dpm_late_enable(struct radeon_device *rdev); diff --git a/drivers/gpu/drm/radeon/radeon_display.c b/drivers/gpu/drm/radeon/radeon_display.c index d680608f6f5b..fbd8b930f2be 100644 --- a/drivers/gpu/drm/radeon/radeon_display.c +++ b/drivers/gpu/drm/radeon/radeon_display.c @@ -571,6 +571,8 @@ static void radeon_crtc_init(struct drm_device *dev, int index) radeon_crtc->max_cursor_width = CURSOR_WIDTH; radeon_crtc->max_cursor_height = CURSOR_HEIGHT; } + dev->mode_config.cursor_width = radeon_crtc->max_cursor_width; + dev->mode_config.cursor_height = radeon_crtc->max_cursor_height; #if 0 radeon_crtc->mode_set.crtc = &radeon_crtc->base; diff --git a/drivers/gpu/drm/radeon/radeon_ring.c b/drivers/gpu/drm/radeon/radeon_ring.c index 1b783f0e6d3a..15e44a7281ab 100644 --- a/drivers/gpu/drm/radeon/radeon_ring.c +++ b/drivers/gpu/drm/radeon/radeon_ring.c @@ -139,7 +139,7 @@ int radeon_ib_schedule(struct radeon_device *rdev, struct radeon_ib *ib, } /* 64 dwords should be enough for fence too */ - r = radeon_ring_lock(rdev, ring, 64 + RADEON_NUM_RINGS * 8); + r = radeon_ring_lock(rdev, ring, 64 + RADEON_NUM_SYNCS * 8); if (r) { dev_err(rdev->dev, "scheduling IB failed (%d).\n", r); return r; diff --git a/drivers/gpu/drm/radeon/radeon_semaphore.c b/drivers/gpu/drm/radeon/radeon_semaphore.c index 2b42aa1914f2..9006b32d5eed 100644 --- a/drivers/gpu/drm/radeon/radeon_semaphore.c +++ b/drivers/gpu/drm/radeon/radeon_semaphore.c @@ -34,14 +34,15 @@ int radeon_semaphore_create(struct radeon_device *rdev, struct radeon_semaphore **semaphore) { + uint32_t *cpu_addr; int i, r; *semaphore = kmalloc(sizeof(struct radeon_semaphore), GFP_KERNEL); if (*semaphore == NULL) { return -ENOMEM; } - r = radeon_sa_bo_new(rdev, &rdev->ring_tmp_bo, - &(*semaphore)->sa_bo, 8, 8, true); + r = radeon_sa_bo_new(rdev, &rdev->ring_tmp_bo, &(*semaphore)->sa_bo, + 8 * RADEON_NUM_SYNCS, 8, true); if (r) { kfree(*semaphore); *semaphore = NULL; @@ -49,7 +50,10 @@ int radeon_semaphore_create(struct radeon_device *rdev, } (*semaphore)->waiters = 0; (*semaphore)->gpu_addr = radeon_sa_bo_gpu_addr((*semaphore)->sa_bo); - *((uint64_t*)radeon_sa_bo_cpu_addr((*semaphore)->sa_bo)) = 0; + + cpu_addr = radeon_sa_bo_cpu_addr((*semaphore)->sa_bo); + for (i = 0; i < RADEON_NUM_SYNCS; ++i) + cpu_addr[i] = 0; for (i = 0; i < RADEON_NUM_RINGS; ++i) (*semaphore)->sync_to[i] = NULL; @@ -125,6 +129,7 @@ int radeon_semaphore_sync_rings(struct radeon_device *rdev, struct radeon_semaphore *semaphore, int ring) { + unsigned count = 0; int i, r; for (i = 0; i < RADEON_NUM_RINGS; ++i) { @@ -140,6 +145,12 @@ int radeon_semaphore_sync_rings(struct radeon_device *rdev, return -EINVAL; } + if (++count > RADEON_NUM_SYNCS) { + /* not enough room, wait manually */ + radeon_fence_wait_locked(fence); + continue; + } + /* allocate enough space for sync command */ r = radeon_ring_alloc(rdev, &rdev->ring[i], 16); if (r) { @@ -164,6 +175,8 @@ int radeon_semaphore_sync_rings(struct radeon_device *rdev, radeon_ring_commit(rdev, &rdev->ring[i]); radeon_fence_note_sync(fence, ring); + + semaphore->gpu_addr += 8; } return 0; diff --git a/drivers/gpu/drm/radeon/rv770_dpm.c b/drivers/gpu/drm/radeon/rv770_dpm.c index 80c595aba359..b5f63f5e22a3 100644 --- a/drivers/gpu/drm/radeon/rv770_dpm.c +++ b/drivers/gpu/drm/radeon/rv770_dpm.c @@ -2174,7 +2174,6 @@ static void rv7xx_parse_pplib_clock_info(struct radeon_device *rdev, struct evergreen_power_info *eg_pi = evergreen_get_pi(rdev); struct rv7xx_ps *ps = rv770_get_ps(rps); u32 sclk, mclk; - u16 vddc; struct rv7xx_pl *pl; switch (index) { @@ -2214,8 +2213,8 @@ static void rv7xx_parse_pplib_clock_info(struct radeon_device *rdev, /* patch up vddc if necessary */ if (pl->vddc == 0xff01) { - if (radeon_atom_get_max_vddc(rdev, 0, 0, &vddc) == 0) - pl->vddc = vddc; + if (pi->max_vddc) + pl->vddc = pi->max_vddc; } if (rps->class & ATOM_PPLIB_CLASSIFICATION_ACPI) { @@ -2527,14 +2526,7 @@ u32 rv770_dpm_get_mclk(struct radeon_device *rdev, bool low) bool rv770_dpm_vblank_too_short(struct radeon_device *rdev) { u32 vblank_time = r600_dpm_get_vblank_time(rdev); - u32 switch_limit = 300; - - /* quirks */ - /* ASUS K70AF */ - if ((rdev->pdev->device == 0x9553) && - (rdev->pdev->subsystem_vendor == 0x1043) && - (rdev->pdev->subsystem_device == 0x1c42)) - switch_limit = 200; + u32 switch_limit = 200; /* 300 */ /* RV770 */ /* mclk switching doesn't seem to work reliably on desktop RV770s */ diff --git a/drivers/gpu/drm/radeon/si.c b/drivers/gpu/drm/radeon/si.c index 09ec4f6c53bb..83578324e5d1 100644 --- a/drivers/gpu/drm/radeon/si.c +++ b/drivers/gpu/drm/radeon/si.c @@ -6338,6 +6338,10 @@ restart_ih: break; } break; + case 124: /* UVD */ + DRM_DEBUG("IH: UVD int: 0x%08x\n", src_data); + radeon_fence_process(rdev, R600_RING_TYPE_UVD_INDEX); + break; case 146: case 147: addr = RREG32(VM_CONTEXT1_PROTECTION_FAULT_ADDR); diff --git a/drivers/gpu/drm/radeon/si_dpm.c b/drivers/gpu/drm/radeon/si_dpm.c index 0471501338fb..0a2f5b4bca43 100644 --- a/drivers/gpu/drm/radeon/si_dpm.c +++ b/drivers/gpu/drm/radeon/si_dpm.c @@ -2395,7 +2395,7 @@ static int si_populate_sq_ramping_values(struct radeon_device *rdev, 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)) + 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++) { @@ -6472,7 +6472,8 @@ void si_dpm_fini(struct radeon_device *rdev) void si_dpm_debugfs_print_current_performance_level(struct radeon_device *rdev, struct seq_file *m) { - struct radeon_ps *rps = rdev->pm.dpm.current_ps; + struct evergreen_power_info *eg_pi = evergreen_get_pi(rdev); + struct radeon_ps *rps = &eg_pi->current_rps; struct ni_ps *ps = ni_get_ps(rps); struct rv7xx_pl *pl; u32 current_index = diff --git a/drivers/gpu/drm/radeon/sumo_dpm.c b/drivers/gpu/drm/radeon/sumo_dpm.c index f121efe12dc5..8b47b3cd0357 100644 --- a/drivers/gpu/drm/radeon/sumo_dpm.c +++ b/drivers/gpu/drm/radeon/sumo_dpm.c @@ -1807,7 +1807,7 @@ void sumo_dpm_debugfs_print_current_performance_level(struct radeon_device *rdev struct seq_file *m) { struct sumo_power_info *pi = sumo_get_pi(rdev); - struct radeon_ps *rps = rdev->pm.dpm.current_ps; + struct radeon_ps *rps = &pi->current_rps; struct sumo_ps *ps = sumo_get_ps(rps); struct sumo_pl *pl; u32 current_index = diff --git a/drivers/gpu/drm/radeon/trinity_dpm.c b/drivers/gpu/drm/radeon/trinity_dpm.c index 2d447192d6f7..2da0e17eb960 100644 --- a/drivers/gpu/drm/radeon/trinity_dpm.c +++ b/drivers/gpu/drm/radeon/trinity_dpm.c @@ -1926,7 +1926,8 @@ void trinity_dpm_print_power_state(struct radeon_device *rdev, void trinity_dpm_debugfs_print_current_performance_level(struct radeon_device *rdev, struct seq_file *m) { - struct radeon_ps *rps = rdev->pm.dpm.current_ps; + struct trinity_power_info *pi = trinity_get_pi(rdev); + struct radeon_ps *rps = &pi->current_rps; struct trinity_ps *ps = trinity_get_ps(rps); struct trinity_pl *pl; u32 current_index = diff --git a/drivers/gpu/drm/radeon/uvd_v2_2.c b/drivers/gpu/drm/radeon/uvd_v2_2.c index 824550db3fed..d1771004cb52 100644 --- a/drivers/gpu/drm/radeon/uvd_v2_2.c +++ b/drivers/gpu/drm/radeon/uvd_v2_2.c @@ -57,7 +57,6 @@ void uvd_v2_2_fence_emit(struct radeon_device *rdev, radeon_ring_write(ring, 0); radeon_ring_write(ring, PACKET0(UVD_GPCOM_VCPU_CMD, 0)); radeon_ring_write(ring, 2); - return; } /** diff --git a/drivers/gpu/drm/ttm/ttm_agp_backend.c b/drivers/gpu/drm/ttm/ttm_agp_backend.c index 3302f99e7497..764be36397fd 100644 --- a/drivers/gpu/drm/ttm/ttm_agp_backend.c +++ b/drivers/gpu/drm/ttm/ttm_agp_backend.c @@ -126,6 +126,7 @@ struct ttm_tt *ttm_agp_tt_create(struct ttm_bo_device *bdev, agp_be->ttm.func = &ttm_agp_func; if (ttm_tt_init(&agp_be->ttm, bdev, size, page_flags, dummy_read_page)) { + kfree(agp_be); return NULL; } diff --git a/drivers/gpu/drm/vmwgfx/svga3d_reg.h b/drivers/gpu/drm/vmwgfx/svga3d_reg.h index b645647b7776..bb594c11605e 100644 --- a/drivers/gpu/drm/vmwgfx/svga3d_reg.h +++ b/drivers/gpu/drm/vmwgfx/svga3d_reg.h @@ -1223,9 +1223,19 @@ typedef enum { #define SVGA_3D_CMD_INVALIDATE_GB_IMAGE_PARTIAL 1129 #define SVGA_3D_CMD_SET_GB_SHADERCONSTS_INLINE 1130 - +#define SVGA_3D_CMD_GB_SCREEN_DMA 1131 +#define SVGA_3D_CMD_BIND_GB_SURFACE_WITH_PITCH 1132 +#define SVGA_3D_CMD_GB_MOB_FENCE 1133 +#define SVGA_3D_CMD_DEFINE_GB_SURFACE_V2 1134 #define SVGA_3D_CMD_DEFINE_GB_MOB64 1135 #define SVGA_3D_CMD_REDEFINE_GB_MOB64 1136 +#define SVGA_3D_CMD_NOP_ERROR 1137 + +#define SVGA_3D_CMD_RESERVED1 1138 +#define SVGA_3D_CMD_RESERVED2 1139 +#define SVGA_3D_CMD_RESERVED3 1140 +#define SVGA_3D_CMD_RESERVED4 1141 +#define SVGA_3D_CMD_RESERVED5 1142 #define SVGA_3D_CMD_MAX 1142 #define SVGA_3D_CMD_FUTURE_MAX 3000 @@ -1973,8 +1983,7 @@ struct { uint32 sizeInBytes; uint32 validSizeInBytes; SVGAMobFormat ptDepth; -} -__attribute__((__packed__)) +} __packed SVGA3dCmdSetOTableBase; /* SVGA_3D_CMD_SET_OTABLE_BASE */ typedef @@ -1984,15 +1993,13 @@ struct { uint32 sizeInBytes; uint32 validSizeInBytes; SVGAMobFormat ptDepth; -} -__attribute__((__packed__)) +} __packed SVGA3dCmdSetOTableBase64; /* SVGA_3D_CMD_SET_OTABLE_BASE64 */ typedef struct { SVGAOTableType type; -} -__attribute__((__packed__)) +} __packed SVGA3dCmdReadbackOTable; /* SVGA_3D_CMD_READBACK_OTABLE */ /* @@ -2005,8 +2012,7 @@ struct SVGA3dCmdDefineGBMob { SVGAMobFormat ptDepth; PPN base; uint32 sizeInBytes; -} -__attribute__((__packed__)) +} __packed SVGA3dCmdDefineGBMob; /* SVGA_3D_CMD_DEFINE_GB_MOB */ @@ -2017,8 +2023,7 @@ SVGA3dCmdDefineGBMob; /* SVGA_3D_CMD_DEFINE_GB_MOB */ typedef struct SVGA3dCmdDestroyGBMob { SVGAMobId mobid; -} -__attribute__((__packed__)) +} __packed SVGA3dCmdDestroyGBMob; /* SVGA_3D_CMD_DESTROY_GB_MOB */ /* @@ -2031,8 +2036,7 @@ struct SVGA3dCmdRedefineGBMob { SVGAMobFormat ptDepth; PPN base; uint32 sizeInBytes; -} -__attribute__((__packed__)) +} __packed SVGA3dCmdRedefineGBMob; /* SVGA_3D_CMD_REDEFINE_GB_MOB */ /* @@ -2045,8 +2049,7 @@ struct SVGA3dCmdDefineGBMob64 { SVGAMobFormat ptDepth; PPN64 base; uint32 sizeInBytes; -} -__attribute__((__packed__)) +} __packed SVGA3dCmdDefineGBMob64; /* SVGA_3D_CMD_DEFINE_GB_MOB64 */ /* @@ -2059,8 +2062,7 @@ struct SVGA3dCmdRedefineGBMob64 { SVGAMobFormat ptDepth; PPN64 base; uint32 sizeInBytes; -} -__attribute__((__packed__)) +} __packed SVGA3dCmdRedefineGBMob64; /* SVGA_3D_CMD_REDEFINE_GB_MOB64 */ /* @@ -2070,8 +2072,7 @@ SVGA3dCmdRedefineGBMob64; /* SVGA_3D_CMD_REDEFINE_GB_MOB64 */ typedef struct SVGA3dCmdUpdateGBMobMapping { SVGAMobId mobid; -} -__attribute__((__packed__)) +} __packed SVGA3dCmdUpdateGBMobMapping; /* SVGA_3D_CMD_UPDATE_GB_MOB_MAPPING */ /* @@ -2087,7 +2088,8 @@ struct SVGA3dCmdDefineGBSurface { uint32 multisampleCount; SVGA3dTextureFilter autogenFilter; SVGA3dSize size; -} SVGA3dCmdDefineGBSurface; /* SVGA_3D_CMD_DEFINE_GB_SURFACE */ +} __packed +SVGA3dCmdDefineGBSurface; /* SVGA_3D_CMD_DEFINE_GB_SURFACE */ /* * Destroy a guest-backed surface. @@ -2096,7 +2098,8 @@ struct SVGA3dCmdDefineGBSurface { typedef struct SVGA3dCmdDestroyGBSurface { uint32 sid; -} SVGA3dCmdDestroyGBSurface; /* SVGA_3D_CMD_DESTROY_GB_SURFACE */ +} __packed +SVGA3dCmdDestroyGBSurface; /* SVGA_3D_CMD_DESTROY_GB_SURFACE */ /* * Bind a guest-backed surface to an object. @@ -2106,7 +2109,8 @@ typedef struct SVGA3dCmdBindGBSurface { uint32 sid; SVGAMobId mobid; -} SVGA3dCmdBindGBSurface; /* SVGA_3D_CMD_BIND_GB_SURFACE */ +} __packed +SVGA3dCmdBindGBSurface; /* SVGA_3D_CMD_BIND_GB_SURFACE */ /* * Conditionally bind a mob to a guest backed surface if testMobid @@ -2123,7 +2127,7 @@ struct{ SVGAMobId testMobid; SVGAMobId mobid; uint32 flags; -} +} __packed SVGA3dCmdCondBindGBSurface; /* SVGA_3D_CMD_COND_BIND_GB_SURFACE */ /* @@ -2135,7 +2139,8 @@ typedef struct SVGA3dCmdUpdateGBImage { SVGA3dSurfaceImageId image; SVGA3dBox box; -} SVGA3dCmdUpdateGBImage; /* SVGA_3D_CMD_UPDATE_GB_IMAGE */ +} __packed +SVGA3dCmdUpdateGBImage; /* SVGA_3D_CMD_UPDATE_GB_IMAGE */ /* * Update an entire guest-backed surface. @@ -2145,7 +2150,8 @@ struct SVGA3dCmdUpdateGBImage { typedef struct SVGA3dCmdUpdateGBSurface { uint32 sid; -} SVGA3dCmdUpdateGBSurface; /* SVGA_3D_CMD_UPDATE_GB_SURFACE */ +} __packed +SVGA3dCmdUpdateGBSurface; /* SVGA_3D_CMD_UPDATE_GB_SURFACE */ /* * Readback an image in a guest-backed surface. @@ -2155,7 +2161,8 @@ struct SVGA3dCmdUpdateGBSurface { typedef struct SVGA3dCmdReadbackGBImage { SVGA3dSurfaceImageId image; -} SVGA3dCmdReadbackGBImage; /* SVGA_3D_CMD_READBACK_GB_IMAGE*/ +} __packed +SVGA3dCmdReadbackGBImage; /* SVGA_3D_CMD_READBACK_GB_IMAGE*/ /* * Readback an entire guest-backed surface. @@ -2165,7 +2172,8 @@ struct SVGA3dCmdReadbackGBImage { typedef struct SVGA3dCmdReadbackGBSurface { uint32 sid; -} SVGA3dCmdReadbackGBSurface; /* SVGA_3D_CMD_READBACK_GB_SURFACE */ +} __packed +SVGA3dCmdReadbackGBSurface; /* SVGA_3D_CMD_READBACK_GB_SURFACE */ /* * Readback a sub rect of an image in a guest-backed surface. After @@ -2179,7 +2187,7 @@ struct SVGA3dCmdReadbackGBImagePartial { SVGA3dSurfaceImageId image; SVGA3dBox box; uint32 invertBox; -} +} __packed SVGA3dCmdReadbackGBImagePartial; /* SVGA_3D_CMD_READBACK_GB_IMAGE_PARTIAL */ /* @@ -2190,7 +2198,8 @@ SVGA3dCmdReadbackGBImagePartial; /* SVGA_3D_CMD_READBACK_GB_IMAGE_PARTIAL */ typedef struct SVGA3dCmdInvalidateGBImage { SVGA3dSurfaceImageId image; -} SVGA3dCmdInvalidateGBImage; /* SVGA_3D_CMD_INVALIDATE_GB_IMAGE */ +} __packed +SVGA3dCmdInvalidateGBImage; /* SVGA_3D_CMD_INVALIDATE_GB_IMAGE */ /* * Invalidate an entire guest-backed surface. @@ -2200,7 +2209,8 @@ struct SVGA3dCmdInvalidateGBImage { typedef struct SVGA3dCmdInvalidateGBSurface { uint32 sid; -} SVGA3dCmdInvalidateGBSurface; /* SVGA_3D_CMD_INVALIDATE_GB_SURFACE */ +} __packed +SVGA3dCmdInvalidateGBSurface; /* SVGA_3D_CMD_INVALIDATE_GB_SURFACE */ /* * Invalidate a sub rect of an image in a guest-backed surface. After @@ -2214,7 +2224,7 @@ struct SVGA3dCmdInvalidateGBImagePartial { SVGA3dSurfaceImageId image; SVGA3dBox box; uint32 invertBox; -} +} __packed SVGA3dCmdInvalidateGBImagePartial; /* SVGA_3D_CMD_INVALIDATE_GB_IMAGE_PARTIAL */ /* @@ -2224,7 +2234,8 @@ SVGA3dCmdInvalidateGBImagePartial; /* SVGA_3D_CMD_INVALIDATE_GB_IMAGE_PARTIAL */ typedef struct SVGA3dCmdDefineGBContext { uint32 cid; -} SVGA3dCmdDefineGBContext; /* SVGA_3D_CMD_DEFINE_GB_CONTEXT */ +} __packed +SVGA3dCmdDefineGBContext; /* SVGA_3D_CMD_DEFINE_GB_CONTEXT */ /* * Destroy a guest-backed context. @@ -2233,7 +2244,8 @@ struct SVGA3dCmdDefineGBContext { typedef struct SVGA3dCmdDestroyGBContext { uint32 cid; -} SVGA3dCmdDestroyGBContext; /* SVGA_3D_CMD_DESTROY_GB_CONTEXT */ +} __packed +SVGA3dCmdDestroyGBContext; /* SVGA_3D_CMD_DESTROY_GB_CONTEXT */ /* * Bind a guest-backed context. @@ -2252,7 +2264,8 @@ struct SVGA3dCmdBindGBContext { uint32 cid; SVGAMobId mobid; uint32 validContents; -} SVGA3dCmdBindGBContext; /* SVGA_3D_CMD_BIND_GB_CONTEXT */ +} __packed +SVGA3dCmdBindGBContext; /* SVGA_3D_CMD_BIND_GB_CONTEXT */ /* * Readback a guest-backed context. @@ -2262,7 +2275,8 @@ struct SVGA3dCmdBindGBContext { typedef struct SVGA3dCmdReadbackGBContext { uint32 cid; -} SVGA3dCmdReadbackGBContext; /* SVGA_3D_CMD_READBACK_GB_CONTEXT */ +} __packed +SVGA3dCmdReadbackGBContext; /* SVGA_3D_CMD_READBACK_GB_CONTEXT */ /* * Invalidate a guest-backed context. @@ -2270,7 +2284,8 @@ struct SVGA3dCmdReadbackGBContext { typedef struct SVGA3dCmdInvalidateGBContext { uint32 cid; -} SVGA3dCmdInvalidateGBContext; /* SVGA_3D_CMD_INVALIDATE_GB_CONTEXT */ +} __packed +SVGA3dCmdInvalidateGBContext; /* SVGA_3D_CMD_INVALIDATE_GB_CONTEXT */ /* * Define a guest-backed shader. @@ -2281,7 +2296,8 @@ struct SVGA3dCmdDefineGBShader { uint32 shid; SVGA3dShaderType type; uint32 sizeInBytes; -} SVGA3dCmdDefineGBShader; /* SVGA_3D_CMD_DEFINE_GB_SHADER */ +} __packed +SVGA3dCmdDefineGBShader; /* SVGA_3D_CMD_DEFINE_GB_SHADER */ /* * Bind a guest-backed shader. @@ -2291,7 +2307,8 @@ typedef struct SVGA3dCmdBindGBShader { uint32 shid; SVGAMobId mobid; uint32 offsetInBytes; -} SVGA3dCmdBindGBShader; /* SVGA_3D_CMD_BIND_GB_SHADER */ +} __packed +SVGA3dCmdBindGBShader; /* SVGA_3D_CMD_BIND_GB_SHADER */ /* * Destroy a guest-backed shader. @@ -2299,7 +2316,8 @@ typedef struct SVGA3dCmdBindGBShader { typedef struct SVGA3dCmdDestroyGBShader { uint32 shid; -} SVGA3dCmdDestroyGBShader; /* SVGA_3D_CMD_DESTROY_GB_SHADER */ +} __packed +SVGA3dCmdDestroyGBShader; /* SVGA_3D_CMD_DESTROY_GB_SHADER */ typedef struct { @@ -2314,14 +2332,16 @@ struct { * Note that FLOAT and INT constants are 4-dwords in length, while * BOOL constants are 1-dword in length. */ -} SVGA3dCmdSetGBShaderConstInline; +} __packed +SVGA3dCmdSetGBShaderConstInline; /* SVGA_3D_CMD_SET_GB_SHADERCONSTS_INLINE */ typedef struct { uint32 cid; SVGA3dQueryType type; -} SVGA3dCmdBeginGBQuery; /* SVGA_3D_CMD_BEGIN_GB_QUERY */ +} __packed +SVGA3dCmdBeginGBQuery; /* SVGA_3D_CMD_BEGIN_GB_QUERY */ typedef struct { @@ -2329,7 +2349,8 @@ struct { SVGA3dQueryType type; SVGAMobId mobid; uint32 offset; -} SVGA3dCmdEndGBQuery; /* SVGA_3D_CMD_END_GB_QUERY */ +} __packed +SVGA3dCmdEndGBQuery; /* SVGA_3D_CMD_END_GB_QUERY */ /* @@ -2346,21 +2367,22 @@ struct { SVGA3dQueryType type; SVGAMobId mobid; uint32 offset; -} SVGA3dCmdWaitForGBQuery; /* SVGA_3D_CMD_WAIT_FOR_GB_QUERY */ +} __packed +SVGA3dCmdWaitForGBQuery; /* SVGA_3D_CMD_WAIT_FOR_GB_QUERY */ typedef struct { SVGAMobId mobid; uint32 fbOffset; uint32 initalized; -} +} __packed SVGA3dCmdEnableGart; /* SVGA_3D_CMD_ENABLE_GART */ typedef struct { SVGAMobId mobid; uint32 gartOffset; -} +} __packed SVGA3dCmdMapMobIntoGart; /* SVGA_3D_CMD_MAP_MOB_INTO_GART */ @@ -2368,7 +2390,7 @@ typedef struct { uint32 gartOffset; uint32 numPages; -} +} __packed SVGA3dCmdUnmapGartRange; /* SVGA_3D_CMD_UNMAP_GART_RANGE */ @@ -2385,27 +2407,27 @@ struct { int32 xRoot; int32 yRoot; uint32 flags; -} +} __packed SVGA3dCmdDefineGBScreenTarget; /* SVGA_3D_CMD_DEFINE_GB_SCREENTARGET */ typedef struct { uint32 stid; -} +} __packed SVGA3dCmdDestroyGBScreenTarget; /* SVGA_3D_CMD_DESTROY_GB_SCREENTARGET */ typedef struct { uint32 stid; SVGA3dSurfaceImageId image; -} +} __packed SVGA3dCmdBindGBScreenTarget; /* SVGA_3D_CMD_BIND_GB_SCREENTARGET */ typedef struct { uint32 stid; SVGA3dBox box; -} +} __packed SVGA3dCmdUpdateGBScreenTarget; /* SVGA_3D_CMD_UPDATE_GB_SCREENTARGET */ /* diff --git a/drivers/gpu/drm/vmwgfx/svga3d_surfacedefs.h b/drivers/gpu/drm/vmwgfx/svga3d_surfacedefs.h index 8369c3ba10fe..ef3385096145 100644 --- a/drivers/gpu/drm/vmwgfx/svga3d_surfacedefs.h +++ b/drivers/gpu/drm/vmwgfx/svga3d_surfacedefs.h @@ -38,8 +38,11 @@ #define DIV_ROUND_UP(x, y) (((x) + (y) - 1) / (y)) #define max_t(type, x, y) ((x) > (y) ? (x) : (y)) +#define min_t(type, x, y) ((x) < (y) ? (x) : (y)) #define surf_size_struct SVGA3dSize #define u32 uint32 +#define u64 uint64_t +#define U32_MAX ((u32)~0U) #endif /* __KERNEL__ */ @@ -704,8 +707,8 @@ static const struct svga3d_surface_desc svga3d_surface_descs[] = { static inline u32 clamped_umul32(u32 a, u32 b) { - uint64_t tmp = (uint64_t) a*b; - return (tmp > (uint64_t) ((u32) -1)) ? (u32) -1 : tmp; + u64 tmp = (u64) a*b; + return (tmp > (u64) U32_MAX) ? U32_MAX : tmp; } static inline const struct svga3d_surface_desc * @@ -834,7 +837,7 @@ svga3dsurface_get_serialized_size(SVGA3dSurfaceFormat format, bool cubemap) { const struct svga3d_surface_desc *desc = svga3dsurface_get_desc(format); - u32 total_size = 0; + u64 total_size = 0; u32 mip; for (mip = 0; mip < num_mip_levels; mip++) { @@ -847,7 +850,7 @@ svga3dsurface_get_serialized_size(SVGA3dSurfaceFormat format, if (cubemap) total_size *= SVGA3D_MAX_SURFACE_FACES; - return total_size; + return (u32) min_t(u64, total_size, (u64) U32_MAX); } diff --git a/drivers/gpu/drm/vmwgfx/svga_reg.h b/drivers/gpu/drm/vmwgfx/svga_reg.h index 71defa4d2d75..11323dd5196f 100644 --- a/drivers/gpu/drm/vmwgfx/svga_reg.h +++ b/drivers/gpu/drm/vmwgfx/svga_reg.h @@ -169,10 +169,17 @@ enum { SVGA_REG_TRACES = 45, /* Enable trace-based updates even when FIFO is on */ SVGA_REG_GMRS_MAX_PAGES = 46, /* Maximum number of 4KB pages for all GMRs */ SVGA_REG_MEMORY_SIZE = 47, /* Total dedicated device memory excluding FIFO */ + SVGA_REG_COMMAND_LOW = 48, /* Lower 32 bits and submits commands */ + SVGA_REG_COMMAND_HIGH = 49, /* Upper 32 bits of command buffer PA */ SVGA_REG_MAX_PRIMARY_BOUNDING_BOX_MEM = 50, /* Max primary memory */ SVGA_REG_SUGGESTED_GBOBJECT_MEM_SIZE_KB = 51, /* Suggested limit on mob mem */ SVGA_REG_DEV_CAP = 52, /* Write dev cap index, read value */ - SVGA_REG_TOP = 53, /* Must be 1 more than the last register */ + SVGA_REG_CMD_PREPEND_LOW = 53, + SVGA_REG_CMD_PREPEND_HIGH = 54, + SVGA_REG_SCREENTARGET_MAX_WIDTH = 55, + SVGA_REG_SCREENTARGET_MAX_HEIGHT = 56, + SVGA_REG_MOB_MAX_SIZE = 57, + SVGA_REG_TOP = 58, /* Must be 1 more than the last register */ SVGA_PALETTE_BASE = 1024, /* Base of SVGA color map */ /* Next 768 (== 256*3) registers exist for colormap */ diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_context.c b/drivers/gpu/drm/vmwgfx/vmwgfx_context.c index 9426c53fb483..1e80152674b5 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_context.c +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_context.c @@ -551,8 +551,7 @@ static int vmw_context_scrub_shader(struct vmw_ctx_bindinfo *bi, bool rebind) cmd->header.size = sizeof(cmd->body); cmd->body.cid = bi->ctx->id; cmd->body.type = bi->i1.shader_type; - cmd->body.shid = - cpu_to_le32((rebind) ? bi->res->id : SVGA3D_INVALID_ID); + cmd->body.shid = ((rebind) ? bi->res->id : SVGA3D_INVALID_ID); vmw_fifo_commit(dev_priv, sizeof(*cmd)); return 0; @@ -585,8 +584,7 @@ static int vmw_context_scrub_render_target(struct vmw_ctx_bindinfo *bi, cmd->header.size = sizeof(cmd->body); cmd->body.cid = bi->ctx->id; cmd->body.type = bi->i1.rt_type; - cmd->body.target.sid = - cpu_to_le32((rebind) ? bi->res->id : SVGA3D_INVALID_ID); + cmd->body.target.sid = ((rebind) ? bi->res->id : SVGA3D_INVALID_ID); cmd->body.target.face = 0; cmd->body.target.mipmap = 0; vmw_fifo_commit(dev_priv, sizeof(*cmd)); @@ -628,8 +626,7 @@ static int vmw_context_scrub_texture(struct vmw_ctx_bindinfo *bi, cmd->body.c.cid = bi->ctx->id; cmd->body.s1.stage = bi->i1.texture_stage; cmd->body.s1.name = SVGA3D_TS_BIND_TEXTURE; - cmd->body.s1.value = - cpu_to_le32((rebind) ? bi->res->id : SVGA3D_INVALID_ID); + cmd->body.s1.value = ((rebind) ? bi->res->id : SVGA3D_INVALID_ID); vmw_fifo_commit(dev_priv, sizeof(*cmd)); return 0; diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.c b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.c index 3bdc0adc656d..0083cbf99edf 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.c +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.c @@ -667,6 +667,7 @@ static int vmw_driver_load(struct drm_device *dev, unsigned long chipset) dev_priv->memory_size = 512*1024*1024; } dev_priv->max_mob_pages = 0; + dev_priv->max_mob_size = 0; if (dev_priv->capabilities & SVGA_CAP_GBOBJECTS) { uint64_t mem_size = vmw_read(dev_priv, @@ -676,6 +677,8 @@ static int vmw_driver_load(struct drm_device *dev, unsigned long chipset) dev_priv->prim_bb_mem = vmw_read(dev_priv, SVGA_REG_MAX_PRIMARY_BOUNDING_BOX_MEM); + dev_priv->max_mob_size = + vmw_read(dev_priv, SVGA_REG_MOB_MAX_SIZE); } else dev_priv->prim_bb_mem = dev_priv->vram_size; diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h index ecaa302a6154..9e4be1725985 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h @@ -386,6 +386,7 @@ struct vmw_private { uint32_t max_gmr_ids; uint32_t max_gmr_pages; uint32_t max_mob_pages; + uint32_t max_mob_size; uint32_t memory_size; bool has_gmr; bool has_mob; diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_execbuf.c b/drivers/gpu/drm/vmwgfx/vmwgfx_execbuf.c index 269b85cc875a..efb575a7996c 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_execbuf.c +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_execbuf.c @@ -602,7 +602,7 @@ static int vmw_cmd_cid_check(struct vmw_private *dev_priv, { struct vmw_cid_cmd { SVGA3dCmdHeader header; - __le32 cid; + uint32_t cid; } *cmd; cmd = container_of(header, struct vmw_cid_cmd, header); @@ -1835,7 +1835,7 @@ static int vmw_cmd_check_not_3d(struct vmw_private *dev_priv, return 0; } -static const struct vmw_cmd_entry const vmw_cmd_entries[SVGA_3D_CMD_MAX] = { +static const struct vmw_cmd_entry vmw_cmd_entries[SVGA_3D_CMD_MAX] = { VMW_CMD_DEF(SVGA_3D_CMD_SURFACE_DEFINE, &vmw_cmd_invalid, false, false, false), VMW_CMD_DEF(SVGA_3D_CMD_SURFACE_DESTROY, &vmw_cmd_invalid, @@ -2032,6 +2032,9 @@ static int vmw_cmd_check(struct vmw_private *dev_priv, goto out_invalid; entry = &vmw_cmd_entries[cmd_id]; + if (unlikely(!entry->func)) + goto out_invalid; + if (unlikely(!entry->user_allow && !sw_context->kernel)) goto out_privileged; @@ -2469,7 +2472,7 @@ int vmw_execbuf_process(struct drm_file *file_priv, if (dev_priv->has_mob) { ret = vmw_rebind_contexts(sw_context); if (unlikely(ret != 0)) - goto out_err; + goto out_unlock_binding; } cmd = vmw_fifo_reserve(dev_priv, command_size); diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_ioctl.c b/drivers/gpu/drm/vmwgfx/vmwgfx_ioctl.c index f9881f9e62bd..47b70949bf3a 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_ioctl.c +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_ioctl.c @@ -102,6 +102,9 @@ int vmw_getparam_ioctl(struct drm_device *dev, void *data, vmw_fp->gb_aware = true; param->value = dev_priv->max_mob_pages * PAGE_SIZE; break; + case DRM_VMW_PARAM_MAX_MOB_SIZE: + param->value = dev_priv->max_mob_size; + break; default: DRM_ERROR("Illegal vmwgfx get param request: %d\n", param->param); diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_shader.c b/drivers/gpu/drm/vmwgfx/vmwgfx_shader.c index 217d941b8176..ee3856578a12 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_shader.c +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_shader.c @@ -371,13 +371,13 @@ int vmw_shader_destroy_ioctl(struct drm_device *dev, void *data, TTM_REF_USAGE); } -int vmw_shader_alloc(struct vmw_private *dev_priv, - struct vmw_dma_buffer *buffer, - size_t shader_size, - size_t offset, - SVGA3dShaderType shader_type, - struct ttm_object_file *tfile, - u32 *handle) +static int vmw_shader_alloc(struct vmw_private *dev_priv, + struct vmw_dma_buffer *buffer, + size_t shader_size, + size_t offset, + SVGA3dShaderType shader_type, + struct ttm_object_file *tfile, + u32 *handle) { struct vmw_user_shader *ushader; struct vmw_resource *res, *tmp; @@ -779,6 +779,8 @@ vmw_compat_shader_man_create(struct vmw_private *dev_priv) int ret; man = kzalloc(sizeof(*man), GFP_KERNEL); + if (man == NULL) + return ERR_PTR(-ENOMEM); man->dev_priv = dev_priv; INIT_LIST_HEAD(&man->list); |