diff options
Diffstat (limited to 'drivers/dma')
64 files changed, 2747 insertions, 271 deletions
diff --git a/drivers/dma/Kconfig b/drivers/dma/Kconfig index cc0a62c34861..d9ec1e69e428 100644 --- a/drivers/dma/Kconfig +++ b/drivers/dma/Kconfig @@ -369,6 +369,15 @@ config K3_DMA Support the DMA engine for Hisilicon K3 platform devices. +config LOONGSON1_APB_DMA + tristate "Loongson1 APB DMA support" + depends on MACH_LOONGSON32 || COMPILE_TEST + select DMA_ENGINE + select DMA_VIRTUAL_CHANNELS + help + This selects support for the APB DMA controller in Loongson1 SoCs, + which is required by Loongson1 NAND and audio support. + config LPC18XX_DMAMUX bool "NXP LPC18xx/43xx DMA MUX for PL080" depends on ARCH_LPC18XX || COMPILE_TEST @@ -378,6 +387,15 @@ config LPC18XX_DMAMUX Enable support for DMA on NXP LPC18xx/43xx platforms with PL080 and multiplexed DMA request lines. +config LPC32XX_DMAMUX + bool "NXP LPC32xx DMA MUX for PL080" + depends on ARCH_LPC32XX || COMPILE_TEST + depends on OF && AMBA_PL08X + select MFD_SYSCON + help + Support for PL080 multiplexed DMA request lines on + LPC32XX platrofm. + config LS2X_APB_DMA tristate "Loongson LS2X APB DMA support" depends on LOONGARCH || COMPILE_TEST @@ -716,6 +734,8 @@ config XILINX_ZYNQMP_DPDMA display driver. # driver files +source "drivers/dma/amd/Kconfig" + source "drivers/dma/bestcomm/Kconfig" source "drivers/dma/mediatek/Kconfig" diff --git a/drivers/dma/Makefile b/drivers/dma/Makefile index 374ea98faf43..ad6a03c052ec 100644 --- a/drivers/dma/Makefile +++ b/drivers/dma/Makefile @@ -49,7 +49,9 @@ obj-$(CONFIG_INTEL_IDMA64) += idma64.o obj-$(CONFIG_INTEL_IOATDMA) += ioat/ obj-y += idxd/ obj-$(CONFIG_K3_DMA) += k3dma.o +obj-$(CONFIG_LOONGSON1_APB_DMA) += loongson1-apb-dma.o obj-$(CONFIG_LPC18XX_DMAMUX) += lpc18xx-dmamux.o +obj-$(CONFIG_LPC32XX_DMAMUX) += lpc32xx-dmamux.o obj-$(CONFIG_LS2X_APB_DMA) += ls2x-apb-dma.o obj-$(CONFIG_MILBEAUT_HDMAC) += milbeaut-hdmac.o obj-$(CONFIG_MILBEAUT_XDMAC) += milbeaut-xdmac.o @@ -83,6 +85,7 @@ obj-$(CONFIG_ST_FDMA) += st_fdma.o obj-$(CONFIG_FSL_DPAA2_QDMA) += fsl-dpaa2-qdma/ obj-$(CONFIG_INTEL_LDMA) += lgm/ +obj-y += amd/ obj-y += mediatek/ obj-y += qcom/ obj-y += stm32/ diff --git a/drivers/dma/acpi-dma.c b/drivers/dma/acpi-dma.c index 5906eae26e2a..a58a1600dd65 100644 --- a/drivers/dma/acpi-dma.c +++ b/drivers/dma/acpi-dma.c @@ -112,7 +112,7 @@ static int acpi_dma_parse_resource_group(const struct acpi_csrt_group *grp, } /** - * acpi_dma_parse_csrt - parse CSRT to exctract additional DMA resources + * acpi_dma_parse_csrt - parse CSRT to extract additional DMA resources * @adev: ACPI device to match with * @adma: struct acpi_dma of the given DMA controller * @@ -305,7 +305,7 @@ EXPORT_SYMBOL_GPL(devm_acpi_dma_controller_free); * found. * * Return: - * 0, if no information is avaiable, -1 on mismatch, and 1 otherwise. + * 0, if no information is available, -1 on mismatch, and 1 otherwise. */ static int acpi_dma_update_dma_spec(struct acpi_dma *adma, struct acpi_dma_spec *dma_spec) diff --git a/drivers/dma/altera-msgdma.c b/drivers/dma/altera-msgdma.c index 0968176f323d..e6a6566b309e 100644 --- a/drivers/dma/altera-msgdma.c +++ b/drivers/dma/altera-msgdma.c @@ -153,7 +153,7 @@ struct msgdma_extended_desc { /** * struct msgdma_sw_desc - implements a sw descriptor * @async_tx: support for the async_tx api - * @hw_desc: assosiated HW descriptor + * @hw_desc: associated HW descriptor * @node: node to move from the free list to the tx list * @tx_list: transmit list node */ @@ -511,7 +511,7 @@ static void msgdma_copy_one(struct msgdma_device *mdev, * of the DMA controller. The descriptor will get flushed to the * FIFO, once the last word (control word) is written. Since we * are not 100% sure that memcpy() writes all word in the "correct" - * oder (address from low to high) on all architectures, we make + * order (address from low to high) on all architectures, we make * sure this control word is written last by single coding it and * adding some write-barriers here. */ diff --git a/drivers/dma/amba-pl08x.c b/drivers/dma/amba-pl08x.c index 73a5cfb4da8a..38cdbca59485 100644 --- a/drivers/dma/amba-pl08x.c +++ b/drivers/dma/amba-pl08x.c @@ -2,7 +2,7 @@ /* * Copyright (c) 2006 ARM Ltd. * Copyright (c) 2010 ST-Ericsson SA - * Copyirght (c) 2017 Linaro Ltd. + * Copyright (c) 2017 Linaro Ltd. * * Author: Peter Pearse <peter.pearse@arm.com> * Author: Linus Walleij <linus.walleij@linaro.org> diff --git a/drivers/dma/amd/Kconfig b/drivers/dma/amd/Kconfig new file mode 100644 index 000000000000..7d1f51d69675 --- /dev/null +++ b/drivers/dma/amd/Kconfig @@ -0,0 +1,14 @@ +# SPDX-License-Identifier: GPL-2.0-only + +config AMD_QDMA + tristate "AMD Queue-based DMA" + depends on HAS_IOMEM + select DMA_ENGINE + select DMA_VIRTUAL_CHANNELS + select REGMAP_MMIO + help + Enable support for the AMD Queue-based DMA subsystem. The primary + mechanism to transfer data using the QDMA is for the QDMA engine to + operate on instructions (descriptors) provided by the host operating + system. Using the descriptors, the QDMA can move data in either the + Host to Card (H2C) direction or the Card to Host (C2H) direction. diff --git a/drivers/dma/amd/Makefile b/drivers/dma/amd/Makefile new file mode 100644 index 000000000000..37212be9364f --- /dev/null +++ b/drivers/dma/amd/Makefile @@ -0,0 +1,3 @@ +# SPDX-License-Identifier: GPL-2.0 + +obj-$(CONFIG_AMD_QDMA) += qdma/ diff --git a/drivers/dma/amd/qdma/Makefile b/drivers/dma/amd/qdma/Makefile new file mode 100644 index 000000000000..011268fef377 --- /dev/null +++ b/drivers/dma/amd/qdma/Makefile @@ -0,0 +1,5 @@ +# SPDX-License-Identifier: GPL-2.0 + +obj-$(CONFIG_AMD_QDMA) += amd-qdma.o + +amd-qdma-$(CONFIG_AMD_QDMA) := qdma.o qdma-comm-regs.o diff --git a/drivers/dma/amd/qdma/qdma-comm-regs.c b/drivers/dma/amd/qdma/qdma-comm-regs.c new file mode 100644 index 000000000000..9162f9d367cc --- /dev/null +++ b/drivers/dma/amd/qdma/qdma-comm-regs.c @@ -0,0 +1,64 @@ +// SPDX-License-Identifier: GPL-2.0-or-later +/* + * Copyright (C) 2023-2024, Advanced Micro Devices, Inc. + */ + +#ifndef __QDMA_REGS_DEF_H +#define __QDMA_REGS_DEF_H + +#include "qdma.h" + +const struct qdma_reg qdma_regos_default[QDMA_REGO_MAX] = { + [QDMA_REGO_CTXT_DATA] = QDMA_REGO(0x804, 8), + [QDMA_REGO_CTXT_CMD] = QDMA_REGO(0x844, 1), + [QDMA_REGO_CTXT_MASK] = QDMA_REGO(0x824, 8), + [QDMA_REGO_MM_H2C_CTRL] = QDMA_REGO(0x1004, 1), + [QDMA_REGO_MM_C2H_CTRL] = QDMA_REGO(0x1204, 1), + [QDMA_REGO_QUEUE_COUNT] = QDMA_REGO(0x120, 1), + [QDMA_REGO_RING_SIZE] = QDMA_REGO(0x204, 1), + [QDMA_REGO_H2C_PIDX] = QDMA_REGO(0x18004, 1), + [QDMA_REGO_C2H_PIDX] = QDMA_REGO(0x18008, 1), + [QDMA_REGO_INTR_CIDX] = QDMA_REGO(0x18000, 1), + [QDMA_REGO_FUNC_ID] = QDMA_REGO(0x12c, 1), + [QDMA_REGO_ERR_INT] = QDMA_REGO(0xb04, 1), + [QDMA_REGO_ERR_STAT] = QDMA_REGO(0x248, 1), +}; + +const struct qdma_reg_field qdma_regfs_default[QDMA_REGF_MAX] = { + /* QDMA_REGO_CTXT_DATA fields */ + [QDMA_REGF_IRQ_ENABLE] = QDMA_REGF(53, 53), + [QDMA_REGF_WBK_ENABLE] = QDMA_REGF(52, 52), + [QDMA_REGF_WBI_CHECK] = QDMA_REGF(34, 34), + [QDMA_REGF_IRQ_ARM] = QDMA_REGF(16, 16), + [QDMA_REGF_IRQ_VEC] = QDMA_REGF(138, 128), + [QDMA_REGF_IRQ_AGG] = QDMA_REGF(139, 139), + [QDMA_REGF_WBI_INTVL_ENABLE] = QDMA_REGF(35, 35), + [QDMA_REGF_MRKR_DISABLE] = QDMA_REGF(62, 62), + [QDMA_REGF_QUEUE_ENABLE] = QDMA_REGF(32, 32), + [QDMA_REGF_QUEUE_MODE] = QDMA_REGF(63, 63), + [QDMA_REGF_DESC_BASE] = QDMA_REGF(127, 64), + [QDMA_REGF_DESC_SIZE] = QDMA_REGF(49, 48), + [QDMA_REGF_RING_ID] = QDMA_REGF(47, 44), + [QDMA_REGF_QUEUE_BASE] = QDMA_REGF(11, 0), + [QDMA_REGF_QUEUE_MAX] = QDMA_REGF(44, 32), + [QDMA_REGF_FUNCTION_ID] = QDMA_REGF(24, 17), + [QDMA_REGF_INTR_AGG_BASE] = QDMA_REGF(66, 15), + [QDMA_REGF_INTR_VECTOR] = QDMA_REGF(11, 1), + [QDMA_REGF_INTR_SIZE] = QDMA_REGF(69, 67), + [QDMA_REGF_INTR_VALID] = QDMA_REGF(0, 0), + [QDMA_REGF_INTR_COLOR] = QDMA_REGF(14, 14), + [QDMA_REGF_INTR_FUNCTION_ID] = QDMA_REGF(125, 114), + /* QDMA_REGO_CTXT_CMD fields */ + [QDMA_REGF_CMD_INDX] = QDMA_REGF(19, 7), + [QDMA_REGF_CMD_CMD] = QDMA_REGF(6, 5), + [QDMA_REGF_CMD_TYPE] = QDMA_REGF(4, 1), + [QDMA_REGF_CMD_BUSY] = QDMA_REGF(0, 0), + /* QDMA_REGO_QUEUE_COUNT fields */ + [QDMA_REGF_QUEUE_COUNT] = QDMA_REGF(11, 0), + /* QDMA_REGO_ERR_INT fields */ + [QDMA_REGF_ERR_INT_FUNC] = QDMA_REGF(11, 0), + [QDMA_REGF_ERR_INT_VEC] = QDMA_REGF(22, 12), + [QDMA_REGF_ERR_INT_ARM] = QDMA_REGF(24, 24), +}; + +#endif /* __QDMA_REGS_DEF_H */ diff --git a/drivers/dma/amd/qdma/qdma.c b/drivers/dma/amd/qdma/qdma.c new file mode 100644 index 000000000000..b0a1f3ad851b --- /dev/null +++ b/drivers/dma/amd/qdma/qdma.c @@ -0,0 +1,1143 @@ +// SPDX-License-Identifier: GPL-2.0-or-later +/* + * DMA driver for AMD Queue-based DMA Subsystem + * + * Copyright (C) 2023-2024, Advanced Micro Devices, Inc. + */ +#include <linux/bitfield.h> +#include <linux/bitops.h> +#include <linux/dmaengine.h> +#include <linux/module.h> +#include <linux/mod_devicetable.h> +#include <linux/dma-map-ops.h> +#include <linux/platform_device.h> +#include <linux/platform_data/amd_qdma.h> +#include <linux/regmap.h> + +#include "qdma.h" + +#define CHAN_STR(q) (((q)->dir == DMA_MEM_TO_DEV) ? "H2C" : "C2H") +#define QDMA_REG_OFF(d, r) ((d)->roffs[r].off) + +/* MMIO regmap config for all QDMA registers */ +static const struct regmap_config qdma_regmap_config = { + .reg_bits = 32, + .val_bits = 32, + .reg_stride = 4, +}; + +static inline struct qdma_queue *to_qdma_queue(struct dma_chan *chan) +{ + return container_of(chan, struct qdma_queue, vchan.chan); +} + +static inline struct qdma_mm_vdesc *to_qdma_vdesc(struct virt_dma_desc *vdesc) +{ + return container_of(vdesc, struct qdma_mm_vdesc, vdesc); +} + +static inline u32 qdma_get_intr_ring_idx(struct qdma_device *qdev) +{ + u32 idx; + + idx = qdev->qintr_rings[qdev->qintr_ring_idx++].ridx; + qdev->qintr_ring_idx %= qdev->qintr_ring_num; + + return idx; +} + +static u64 qdma_get_field(const struct qdma_device *qdev, const u32 *data, + enum qdma_reg_fields field) +{ + const struct qdma_reg_field *f = &qdev->rfields[field]; + u16 low_pos, hi_pos, low_bit, hi_bit; + u64 value = 0, mask; + + low_pos = f->lsb / BITS_PER_TYPE(*data); + hi_pos = f->msb / BITS_PER_TYPE(*data); + + if (low_pos == hi_pos) { + low_bit = f->lsb % BITS_PER_TYPE(*data); + hi_bit = f->msb % BITS_PER_TYPE(*data); + mask = GENMASK(hi_bit, low_bit); + value = (data[low_pos] & mask) >> low_bit; + } else if (hi_pos == low_pos + 1) { + low_bit = f->lsb % BITS_PER_TYPE(*data); + hi_bit = low_bit + (f->msb - f->lsb); + value = ((u64)data[hi_pos] << BITS_PER_TYPE(*data)) | + data[low_pos]; + mask = GENMASK_ULL(hi_bit, low_bit); + value = (value & mask) >> low_bit; + } else { + hi_bit = f->msb % BITS_PER_TYPE(*data); + mask = GENMASK(hi_bit, 0); + value = data[hi_pos] & mask; + low_bit = f->msb - f->lsb - hi_bit; + value <<= low_bit; + low_bit -= 32; + value |= (u64)data[hi_pos - 1] << low_bit; + mask = GENMASK(31, 32 - low_bit); + value |= (data[hi_pos - 2] & mask) >> low_bit; + } + + return value; +} + +static void qdma_set_field(const struct qdma_device *qdev, u32 *data, + enum qdma_reg_fields field, u64 value) +{ + const struct qdma_reg_field *f = &qdev->rfields[field]; + u16 low_pos, hi_pos, low_bit; + + low_pos = f->lsb / BITS_PER_TYPE(*data); + hi_pos = f->msb / BITS_PER_TYPE(*data); + low_bit = f->lsb % BITS_PER_TYPE(*data); + + data[low_pos++] |= value << low_bit; + if (low_pos <= hi_pos) + data[low_pos++] |= (u32)(value >> (32 - low_bit)); + if (low_pos <= hi_pos) + data[low_pos] |= (u32)(value >> (64 - low_bit)); +} + +static inline int qdma_reg_write(const struct qdma_device *qdev, + const u32 *data, enum qdma_regs reg) +{ + const struct qdma_reg *r = &qdev->roffs[reg]; + int ret; + + if (r->count > 1) + ret = regmap_bulk_write(qdev->regmap, r->off, data, r->count); + else + ret = regmap_write(qdev->regmap, r->off, *data); + + return ret; +} + +static inline int qdma_reg_read(const struct qdma_device *qdev, u32 *data, + enum qdma_regs reg) +{ + const struct qdma_reg *r = &qdev->roffs[reg]; + int ret; + + if (r->count > 1) + ret = regmap_bulk_read(qdev->regmap, r->off, data, r->count); + else + ret = regmap_read(qdev->regmap, r->off, data); + + return ret; +} + +static int qdma_context_cmd_execute(const struct qdma_device *qdev, + enum qdma_ctxt_type type, + enum qdma_ctxt_cmd cmd, u16 index) +{ + u32 value = 0; + int ret; + + qdma_set_field(qdev, &value, QDMA_REGF_CMD_INDX, index); + qdma_set_field(qdev, &value, QDMA_REGF_CMD_CMD, cmd); + qdma_set_field(qdev, &value, QDMA_REGF_CMD_TYPE, type); + + ret = qdma_reg_write(qdev, &value, QDMA_REGO_CTXT_CMD); + if (ret) + return ret; + + ret = regmap_read_poll_timeout(qdev->regmap, + QDMA_REG_OFF(qdev, QDMA_REGO_CTXT_CMD), + value, + !qdma_get_field(qdev, &value, + QDMA_REGF_CMD_BUSY), + QDMA_POLL_INTRVL_US, + QDMA_POLL_TIMEOUT_US); + if (ret) { + qdma_err(qdev, "Context command execution timed out"); + return ret; + } + + return 0; +} + +static int qdma_context_write_data(const struct qdma_device *qdev, + const u32 *data) +{ + u32 mask[QDMA_CTXT_REGMAP_LEN]; + int ret; + + memset(mask, ~0, sizeof(mask)); + + ret = qdma_reg_write(qdev, mask, QDMA_REGO_CTXT_MASK); + if (ret) + return ret; + + ret = qdma_reg_write(qdev, data, QDMA_REGO_CTXT_DATA); + if (ret) + return ret; + + return 0; +} + +static void qdma_prep_sw_desc_context(const struct qdma_device *qdev, + const struct qdma_ctxt_sw_desc *ctxt, + u32 *data) +{ + memset(data, 0, QDMA_CTXT_REGMAP_LEN * sizeof(*data)); + qdma_set_field(qdev, data, QDMA_REGF_DESC_BASE, ctxt->desc_base); + qdma_set_field(qdev, data, QDMA_REGF_IRQ_VEC, ctxt->vec); + qdma_set_field(qdev, data, QDMA_REGF_FUNCTION_ID, qdev->fid); + + qdma_set_field(qdev, data, QDMA_REGF_DESC_SIZE, QDMA_DESC_SIZE_32B); + qdma_set_field(qdev, data, QDMA_REGF_RING_ID, QDMA_DEFAULT_RING_ID); + qdma_set_field(qdev, data, QDMA_REGF_QUEUE_MODE, QDMA_QUEUE_OP_MM); + qdma_set_field(qdev, data, QDMA_REGF_IRQ_ENABLE, 1); + qdma_set_field(qdev, data, QDMA_REGF_WBK_ENABLE, 1); + qdma_set_field(qdev, data, QDMA_REGF_WBI_CHECK, 1); + qdma_set_field(qdev, data, QDMA_REGF_IRQ_ARM, 1); + qdma_set_field(qdev, data, QDMA_REGF_IRQ_AGG, 1); + qdma_set_field(qdev, data, QDMA_REGF_WBI_INTVL_ENABLE, 1); + qdma_set_field(qdev, data, QDMA_REGF_QUEUE_ENABLE, 1); + qdma_set_field(qdev, data, QDMA_REGF_MRKR_DISABLE, 1); +} + +static void qdma_prep_intr_context(const struct qdma_device *qdev, + const struct qdma_ctxt_intr *ctxt, + u32 *data) +{ + memset(data, 0, QDMA_CTXT_REGMAP_LEN * sizeof(*data)); + qdma_set_field(qdev, data, QDMA_REGF_INTR_AGG_BASE, ctxt->agg_base); + qdma_set_field(qdev, data, QDMA_REGF_INTR_VECTOR, ctxt->vec); + qdma_set_field(qdev, data, QDMA_REGF_INTR_SIZE, ctxt->size); + qdma_set_field(qdev, data, QDMA_REGF_INTR_VALID, ctxt->valid); + qdma_set_field(qdev, data, QDMA_REGF_INTR_COLOR, ctxt->color); + qdma_set_field(qdev, data, QDMA_REGF_INTR_FUNCTION_ID, qdev->fid); +} + +static void qdma_prep_fmap_context(const struct qdma_device *qdev, + const struct qdma_ctxt_fmap *ctxt, + u32 *data) +{ + memset(data, 0, QDMA_CTXT_REGMAP_LEN * sizeof(*data)); + qdma_set_field(qdev, data, QDMA_REGF_QUEUE_BASE, ctxt->qbase); + qdma_set_field(qdev, data, QDMA_REGF_QUEUE_MAX, ctxt->qmax); +} + +/* + * Program the indirect context register space + * + * Once the queue is enabled, context is dynamically updated by hardware. Any + * modification of the context through this API when the queue is enabled can + * result in unexpected behavior. Reading the context when the queue is enabled + * is not recommended as it can result in reduced performance. + */ +static int qdma_prog_context(struct qdma_device *qdev, enum qdma_ctxt_type type, + enum qdma_ctxt_cmd cmd, u16 index, u32 *ctxt) +{ + int ret; + + mutex_lock(&qdev->ctxt_lock); + if (cmd == QDMA_CTXT_WRITE) { + ret = qdma_context_write_data(qdev, ctxt); + if (ret) + goto failed; + } + + ret = qdma_context_cmd_execute(qdev, type, cmd, index); + if (ret) + goto failed; + + if (cmd == QDMA_CTXT_READ) { + ret = qdma_reg_read(qdev, ctxt, QDMA_REGO_CTXT_DATA); + if (ret) + goto failed; + } + +failed: + mutex_unlock(&qdev->ctxt_lock); + + return ret; +} + +static int qdma_check_queue_status(struct qdma_device *qdev, + enum dma_transfer_direction dir, u16 qid) +{ + u32 status, data[QDMA_CTXT_REGMAP_LEN] = {0}; + enum qdma_ctxt_type type; + int ret; + + if (dir == DMA_MEM_TO_DEV) + type = QDMA_CTXT_DESC_SW_H2C; + else + type = QDMA_CTXT_DESC_SW_C2H; + ret = qdma_prog_context(qdev, type, QDMA_CTXT_READ, qid, data); + if (ret) + return ret; + + status = qdma_get_field(qdev, data, QDMA_REGF_QUEUE_ENABLE); + if (status) { + qdma_err(qdev, "queue %d already in use", qid); + return -EBUSY; + } + + return 0; +} + +static int qdma_clear_queue_context(const struct qdma_queue *queue) +{ + enum qdma_ctxt_type h2c_types[] = { QDMA_CTXT_DESC_SW_H2C, + QDMA_CTXT_DESC_HW_H2C, + QDMA_CTXT_DESC_CR_H2C, + QDMA_CTXT_PFTCH, }; + enum qdma_ctxt_type c2h_types[] = { QDMA_CTXT_DESC_SW_C2H, + QDMA_CTXT_DESC_HW_C2H, + QDMA_CTXT_DESC_CR_C2H, + QDMA_CTXT_PFTCH, }; + struct qdma_device *qdev = queue->qdev; + enum qdma_ctxt_type *type; + int ret, num, i; + + if (queue->dir == DMA_MEM_TO_DEV) { + type = h2c_types; + num = ARRAY_SIZE(h2c_types); + } else { + type = c2h_types; + num = ARRAY_SIZE(c2h_types); + } + for (i = 0; i < num; i++) { + ret = qdma_prog_context(qdev, type[i], QDMA_CTXT_CLEAR, + queue->qid, NULL); + if (ret) { + qdma_err(qdev, "Failed to clear ctxt %d", type[i]); + return ret; + } + } + + return 0; +} + +static int qdma_setup_fmap_context(struct qdma_device *qdev) +{ + u32 ctxt[QDMA_CTXT_REGMAP_LEN]; + struct qdma_ctxt_fmap fmap; + int ret; + + ret = qdma_prog_context(qdev, QDMA_CTXT_FMAP, QDMA_CTXT_CLEAR, + qdev->fid, NULL); + if (ret) { + qdma_err(qdev, "Failed clearing context"); + return ret; + } + + fmap.qbase = 0; + fmap.qmax = qdev->chan_num * 2; + qdma_prep_fmap_context(qdev, &fmap, ctxt); + ret = qdma_prog_context(qdev, QDMA_CTXT_FMAP, QDMA_CTXT_WRITE, + qdev->fid, ctxt); + if (ret) + qdma_err(qdev, "Failed setup fmap, ret %d", ret); + + return ret; +} + +static int qdma_setup_queue_context(struct qdma_device *qdev, + const struct qdma_ctxt_sw_desc *sw_desc, + enum dma_transfer_direction dir, u16 qid) +{ + u32 ctxt[QDMA_CTXT_REGMAP_LEN]; + enum qdma_ctxt_type type; + int ret; + + if (dir == DMA_MEM_TO_DEV) + type = QDMA_CTXT_DESC_SW_H2C; + else + type = QDMA_CTXT_DESC_SW_C2H; + + qdma_prep_sw_desc_context(qdev, sw_desc, ctxt); + /* Setup SW descriptor context */ + ret = qdma_prog_context(qdev, type, QDMA_CTXT_WRITE, qid, ctxt); + if (ret) + qdma_err(qdev, "Failed setup SW desc ctxt for queue: %d", qid); + + return ret; +} + +/* + * Enable or disable memory-mapped DMA engines + * 1: enable, 0: disable + */ +static int qdma_sgdma_control(struct qdma_device *qdev, u32 ctrl) +{ + int ret; + + ret = qdma_reg_write(qdev, &ctrl, QDMA_REGO_MM_H2C_CTRL); + ret |= qdma_reg_write(qdev, &ctrl, QDMA_REGO_MM_C2H_CTRL); + + return ret; +} + +static int qdma_get_hw_info(struct qdma_device *qdev) +{ + struct qdma_platdata *pdata = dev_get_platdata(&qdev->pdev->dev); + u32 value = 0; + int ret; + + ret = qdma_reg_read(qdev, &value, QDMA_REGO_QUEUE_COUNT); + if (ret) + return ret; + + value = qdma_get_field(qdev, &value, QDMA_REGF_QUEUE_COUNT) + 1; + if (pdata->max_mm_channels * 2 > value) { + qdma_err(qdev, "not enough hw queues %d", value); + return -EINVAL; + } + qdev->chan_num = pdata->max_mm_channels; + + ret = qdma_reg_read(qdev, &qdev->fid, QDMA_REGO_FUNC_ID); + if (ret) + return ret; + + qdma_info(qdev, "max channel %d, function id %d", + qdev->chan_num, qdev->fid); + + return 0; +} + +static inline int qdma_update_pidx(const struct qdma_queue *queue, u16 pidx) +{ + struct qdma_device *qdev = queue->qdev; + + return regmap_write(qdev->regmap, queue->pidx_reg, + pidx | QDMA_QUEUE_ARM_BIT); +} + +static inline int qdma_update_cidx(const struct qdma_queue *queue, + u16 ridx, u16 cidx) +{ + struct qdma_device *qdev = queue->qdev; + + return regmap_write(qdev->regmap, queue->cidx_reg, + ((u32)ridx << 16) | cidx); +} + +/** + * qdma_free_vdesc - Free descriptor + * @vdesc: Virtual DMA descriptor + */ +static void qdma_free_vdesc(struct virt_dma_desc *vdesc) +{ + struct qdma_mm_vdesc *vd = to_qdma_vdesc(vdesc); + + kfree(vd); +} + +static int qdma_alloc_queues(struct qdma_device *qdev, + enum dma_transfer_direction dir) +{ + struct qdma_queue *q, **queues; + u32 i, pidx_base; + int ret; + + if (dir == DMA_MEM_TO_DEV) { + queues = &qdev->h2c_queues; + pidx_base = QDMA_REG_OFF(qdev, QDMA_REGO_H2C_PIDX); + } else { + queues = &qdev->c2h_queues; + pidx_base = QDMA_REG_OFF(qdev, QDMA_REGO_C2H_PIDX); + } + + *queues = devm_kcalloc(&qdev->pdev->dev, qdev->chan_num, sizeof(*q), + GFP_KERNEL); + if (!*queues) + return -ENOMEM; + + for (i = 0; i < qdev->chan_num; i++) { + ret = qdma_check_queue_status(qdev, dir, i); + if (ret) + return ret; + + q = &(*queues)[i]; + q->ring_size = QDMA_DEFAULT_RING_SIZE; + q->idx_mask = q->ring_size - 2; + q->qdev = qdev; + q->dir = dir; + q->qid = i; + q->pidx_reg = pidx_base + i * QDMA_DMAP_REG_STRIDE; + q->cidx_reg = QDMA_REG_OFF(qdev, QDMA_REGO_INTR_CIDX) + + i * QDMA_DMAP_REG_STRIDE; + q->vchan.desc_free = qdma_free_vdesc; + vchan_init(&q->vchan, &qdev->dma_dev); + } + + return 0; +} + +static int qdma_device_verify(struct qdma_device *qdev) +{ + u32 value; + int ret; + + ret = regmap_read(qdev->regmap, QDMA_IDENTIFIER_REGOFF, &value); + if (ret) + return ret; + + value = FIELD_GET(QDMA_IDENTIFIER_MASK, value); + if (value != QDMA_IDENTIFIER) { + qdma_err(qdev, "Invalid identifier"); + return -ENODEV; + } + qdev->rfields = qdma_regfs_default; + qdev->roffs = qdma_regos_default; + + return 0; +} + +static int qdma_device_setup(struct qdma_device *qdev) +{ + struct device *dev = &qdev->pdev->dev; + u32 ring_sz = QDMA_DEFAULT_RING_SIZE; + int ret = 0; + + while (dev && get_dma_ops(dev)) + dev = dev->parent; + if (!dev) { + qdma_err(qdev, "dma device not found"); + return -EINVAL; + } + set_dma_ops(&qdev->pdev->dev, get_dma_ops(dev)); + + ret = qdma_setup_fmap_context(qdev); + if (ret) { + qdma_err(qdev, "Failed setup fmap context"); + return ret; + } + + /* Setup global ring buffer size at QDMA_DEFAULT_RING_ID index */ + ret = qdma_reg_write(qdev, &ring_sz, QDMA_REGO_RING_SIZE); + if (ret) { + qdma_err(qdev, "Failed to setup ring %d of size %ld", + QDMA_DEFAULT_RING_ID, QDMA_DEFAULT_RING_SIZE); + return ret; + } + + /* Enable memory-mapped DMA engine in both directions */ + ret = qdma_sgdma_control(qdev, 1); + if (ret) { + qdma_err(qdev, "Failed to SGDMA with error %d", ret); + return ret; + } + + ret = qdma_alloc_queues(qdev, DMA_MEM_TO_DEV); + if (ret) { + qdma_err(qdev, "Failed to alloc H2C queues, ret %d", ret); + return ret; + } + + ret = qdma_alloc_queues(qdev, DMA_DEV_TO_MEM); + if (ret) { + qdma_err(qdev, "Failed to alloc C2H queues, ret %d", ret); + return ret; + } + + return 0; +} + +/** + * qdma_free_queue_resources() - Free queue resources + * @chan: DMA channel + */ +static void qdma_free_queue_resources(struct dma_chan *chan) +{ + struct qdma_queue *queue = to_qdma_queue(chan); + struct qdma_device *qdev = queue->qdev; + struct device *dev = qdev->dma_dev.dev; + + qdma_clear_queue_context(queue); + vchan_free_chan_resources(&queue->vchan); + dma_free_coherent(dev, queue->ring_size * QDMA_MM_DESC_SIZE, + queue->desc_base, queue->dma_desc_base); +} + +/** + * qdma_alloc_queue_resources() - Allocate queue resources + * @chan: DMA channel + */ +static int qdma_alloc_queue_resources(struct dma_chan *chan) +{ + struct qdma_queue *queue = to_qdma_queue(chan); + struct qdma_device *qdev = queue->qdev; + struct qdma_ctxt_sw_desc desc; + size_t size; + int ret; + + ret = qdma_clear_queue_context(queue); + if (ret) + return ret; + + size = queue->ring_size * QDMA_MM_DESC_SIZE; + queue->desc_base = dma_alloc_coherent(qdev->dma_dev.dev, size, + &queue->dma_desc_base, + GFP_KERNEL); + if (!queue->desc_base) { + qdma_err(qdev, "Failed to allocate descriptor ring"); + return -ENOMEM; + } + + /* Setup SW descriptor queue context for DMA memory map */ + desc.vec = qdma_get_intr_ring_idx(qdev); + desc.desc_base = queue->dma_desc_base; + ret = qdma_setup_queue_context(qdev, &desc, queue->dir, queue->qid); + if (ret) { + qdma_err(qdev, "Failed to setup SW desc ctxt for %s", + chan->name); + dma_free_coherent(qdev->dma_dev.dev, size, queue->desc_base, + queue->dma_desc_base); + return ret; + } + + queue->pidx = 0; + queue->cidx = 0; + + return 0; +} + +static bool qdma_filter_fn(struct dma_chan *chan, void *param) +{ + struct qdma_queue *queue = to_qdma_queue(chan); + struct qdma_queue_info *info = param; + + return info->dir == queue->dir; +} + +static int qdma_xfer_start(struct qdma_queue *queue) +{ + struct qdma_device *qdev = queue->qdev; + int ret; + + if (!vchan_next_desc(&queue->vchan)) + return 0; + + qdma_dbg(qdev, "Tnx kickoff with P: %d for %s%d", + queue->issued_vdesc->pidx, CHAN_STR(queue), queue->qid); + + ret = qdma_update_pidx(queue, queue->issued_vdesc->pidx); + if (ret) { + qdma_err(qdev, "Failed to update PIDX to %d for %s queue: %d", + queue->pidx, CHAN_STR(queue), queue->qid); + } + + return ret; +} + +static void qdma_issue_pending(struct dma_chan *chan) +{ + struct qdma_queue *queue = to_qdma_queue(chan); + unsigned long flags; + + spin_lock_irqsave(&queue->vchan.lock, flags); + if (vchan_issue_pending(&queue->vchan)) { + if (queue->submitted_vdesc) { + queue->issued_vdesc = queue->submitted_vdesc; + queue->submitted_vdesc = NULL; + } + qdma_xfer_start(queue); + } + + spin_unlock_irqrestore(&queue->vchan.lock, flags); +} + +static struct qdma_mm_desc *qdma_get_desc(struct qdma_queue *q) +{ + struct qdma_mm_desc *desc; + + if (((q->pidx + 1) & q->idx_mask) == q->cidx) + return NULL; + + desc = q->desc_base + q->pidx; + q->pidx = (q->pidx + 1) & q->idx_mask; + + return desc; +} + +static int qdma_hw_enqueue(struct qdma_queue *q, struct qdma_mm_vdesc *vdesc) +{ + struct qdma_mm_desc *desc; + struct scatterlist *sg; + u64 addr, *src, *dst; + u32 rest, len; + int ret = 0; + u32 i; + + if (!vdesc->sg_len) + return 0; + + if (q->dir == DMA_MEM_TO_DEV) { + dst = &vdesc->dev_addr; + src = &addr; + } else { + dst = &addr; + src = &vdesc->dev_addr; + } + + for_each_sg(vdesc->sgl, sg, vdesc->sg_len, i) { + addr = sg_dma_address(sg) + vdesc->sg_off; + rest = sg_dma_len(sg) - vdesc->sg_off; + while (rest) { + len = min_t(u32, rest, QDMA_MM_DESC_MAX_LEN); + desc = qdma_get_desc(q); + if (!desc) { + ret = -EBUSY; + goto out; + } + + desc->src_addr = cpu_to_le64(*src); + desc->dst_addr = cpu_to_le64(*dst); + desc->len = cpu_to_le32(len); + + vdesc->dev_addr += len; + vdesc->sg_off += len; + vdesc->pending_descs++; + addr += len; + rest -= len; + } + vdesc->sg_off = 0; + } +out: + vdesc->sg_len -= i; + vdesc->pidx = q->pidx; + return ret; +} + +static void qdma_fill_pending_vdesc(struct qdma_queue *q) +{ + struct virt_dma_chan *vc = &q->vchan; + struct qdma_mm_vdesc *vdesc = NULL; + struct virt_dma_desc *vd; + int ret; + + if (!list_empty(&vc->desc_issued)) { + vd = &q->issued_vdesc->vdesc; + list_for_each_entry_from(vd, &vc->desc_issued, node) { + vdesc = to_qdma_vdesc(vd); + ret = qdma_hw_enqueue(q, vdesc); + if (ret) { + q->issued_vdesc = vdesc; + return; + } + } + q->issued_vdesc = vdesc; + } + + if (list_empty(&vc->desc_submitted)) + return; + + if (q->submitted_vdesc) + vd = &q->submitted_vdesc->vdesc; + else + vd = list_first_entry(&vc->desc_submitted, typeof(*vd), node); + + list_for_each_entry_from(vd, &vc->desc_submitted, node) { + vdesc = to_qdma_vdesc(vd); + ret = qdma_hw_enqueue(q, vdesc); + if (ret) + break; + } + q->submitted_vdesc = vdesc; +} + +static dma_cookie_t qdma_tx_submit(struct dma_async_tx_descriptor *tx) +{ + struct virt_dma_chan *vc = to_virt_chan(tx->chan); + struct qdma_queue *q = to_qdma_queue(&vc->chan); + struct virt_dma_desc *vd; + unsigned long flags; + dma_cookie_t cookie; + + vd = container_of(tx, struct virt_dma_desc, tx); + spin_lock_irqsave(&vc->lock, flags); + cookie = dma_cookie_assign(tx); + + list_move_tail(&vd->node, &vc->desc_submitted); + qdma_fill_pending_vdesc(q); + spin_unlock_irqrestore(&vc->lock, flags); + + return cookie; +} + +static struct dma_async_tx_descriptor * +qdma_prep_device_sg(struct dma_chan *chan, struct scatterlist *sgl, + unsigned int sg_len, enum dma_transfer_direction dir, + unsigned long flags, void *context) +{ + struct qdma_queue *q = to_qdma_queue(chan); + struct dma_async_tx_descriptor *tx; + struct qdma_mm_vdesc *vdesc; + + vdesc = kzalloc(sizeof(*vdesc), GFP_NOWAIT); + if (!vdesc) + return NULL; + vdesc->sgl = sgl; + vdesc->sg_len = sg_len; + if (dir == DMA_MEM_TO_DEV) + vdesc->dev_addr = q->cfg.dst_addr; + else + vdesc->dev_addr = q->cfg.src_addr; + + tx = vchan_tx_prep(&q->vchan, &vdesc->vdesc, flags); + tx->tx_submit = qdma_tx_submit; + + return tx; +} + +static int qdma_device_config(struct dma_chan *chan, + struct dma_slave_config *cfg) +{ + struct qdma_queue *q = to_qdma_queue(chan); + + memcpy(&q->cfg, cfg, sizeof(*cfg)); + + return 0; +} + +static int qdma_arm_err_intr(const struct qdma_device *qdev) +{ + u32 value = 0; + + qdma_set_field(qdev, &value, QDMA_REGF_ERR_INT_FUNC, qdev->fid); + qdma_set_field(qdev, &value, QDMA_REGF_ERR_INT_VEC, qdev->err_irq_idx); + qdma_set_field(qdev, &value, QDMA_REGF_ERR_INT_ARM, 1); + + return qdma_reg_write(qdev, &value, QDMA_REGO_ERR_INT); +} + +static irqreturn_t qdma_error_isr(int irq, void *data) +{ + struct qdma_device *qdev = data; + u32 err_stat = 0; + int ret; + + ret = qdma_reg_read(qdev, &err_stat, QDMA_REGO_ERR_STAT); + if (ret) { + qdma_err(qdev, "read error state failed, ret %d", ret); + goto out; + } + + qdma_err(qdev, "global error %d", err_stat); + ret = qdma_reg_write(qdev, &err_stat, QDMA_REGO_ERR_STAT); + if (ret) + qdma_err(qdev, "clear error state failed, ret %d", ret); + +out: + qdma_arm_err_intr(qdev); + return IRQ_HANDLED; +} + +static irqreturn_t qdma_queue_isr(int irq, void *data) +{ + struct qdma_intr_ring *intr = data; + struct qdma_queue *q = NULL; + struct qdma_device *qdev; + u32 index, comp_desc; + u64 intr_ent; + u8 color; + int ret; + u16 qid; + + qdev = intr->qdev; + index = intr->cidx; + while (1) { + struct virt_dma_desc *vd; + struct qdma_mm_vdesc *vdesc; + unsigned long flags; + u32 cidx; + + intr_ent = le64_to_cpu(intr->base[index]); + color = FIELD_GET(QDMA_INTR_MASK_COLOR, intr_ent); + if (color != intr->color) + break; + + qid = FIELD_GET(QDMA_INTR_MASK_QID, intr_ent); + if (FIELD_GET(QDMA_INTR_MASK_TYPE, intr_ent)) + q = qdev->c2h_queues; + else + q = qdev->h2c_queues; + q += qid; + + cidx = FIELD_GET(QDMA_INTR_MASK_CIDX, intr_ent); + + spin_lock_irqsave(&q->vchan.lock, flags); + comp_desc = (cidx - q->cidx) & q->idx_mask; + + vd = vchan_next_desc(&q->vchan); + if (!vd) + goto skip; + + vdesc = to_qdma_vdesc(vd); + while (comp_desc > vdesc->pending_descs) { + list_del(&vd->node); + vchan_cookie_complete(vd); + comp_desc -= vdesc->pending_descs; + vd = vchan_next_desc(&q->vchan); + vdesc = to_qdma_vdesc(vd); + } + vdesc->pending_descs -= comp_desc; + if (!vdesc->pending_descs && QDMA_VDESC_QUEUED(vdesc)) { + list_del(&vd->node); + vchan_cookie_complete(vd); + } + q->cidx = cidx; + + qdma_fill_pending_vdesc(q); + qdma_xfer_start(q); + +skip: + spin_unlock_irqrestore(&q->vchan.lock, flags); + + /* + * Wrap the index value and flip the expected color value if + * interrupt aggregation PIDX has wrapped around. + */ + index++; + index &= QDMA_INTR_RING_IDX_MASK; + if (!index) + intr->color = !intr->color; + } + + /* + * Update the software interrupt aggregation ring CIDX if a valid entry + * was found. + */ + if (q) { + qdma_dbg(qdev, "update intr ring%d %d", intr->ridx, index); + + /* + * Record the last read index of status descriptor from the + * interrupt aggregation ring. + */ + intr->cidx = index; + + ret = qdma_update_cidx(q, intr->ridx, index); + if (ret) { + qdma_err(qdev, "Failed to update IRQ CIDX"); + return IRQ_NONE; + } + } + + return IRQ_HANDLED; +} + +static int qdma_init_error_irq(struct qdma_device *qdev) +{ + struct device *dev = &qdev->pdev->dev; + int ret; + u32 vec; + + vec = qdev->queue_irq_start - 1; + + ret = devm_request_threaded_irq(dev, vec, NULL, qdma_error_isr, + IRQF_ONESHOT, "amd-qdma-error", qdev); + if (ret) { + qdma_err(qdev, "Failed to request error IRQ vector: %d", vec); + return ret; + } + + ret = qdma_arm_err_intr(qdev); + if (ret) + qdma_err(qdev, "Failed to arm err interrupt, ret %d", ret); + + return ret; +} + +static int qdmam_alloc_qintr_rings(struct qdma_device *qdev) +{ + u32 ctxt[QDMA_CTXT_REGMAP_LEN]; + struct device *dev = &qdev->pdev->dev; + struct qdma_intr_ring *ring; + struct qdma_ctxt_intr intr_ctxt; + u32 vector; + int ret, i; + + qdev->qintr_ring_num = qdev->queue_irq_num; + qdev->qintr_rings = devm_kcalloc(dev, qdev->qintr_ring_num, + sizeof(*qdev->qintr_rings), + GFP_KERNEL); + if (!qdev->qintr_rings) + return -ENOMEM; + + vector = qdev->queue_irq_start; + for (i = 0; i < qdev->qintr_ring_num; i++, vector++) { + ring = &qdev->qintr_rings[i]; + ring->qdev = qdev; + ring->msix_id = qdev->err_irq_idx + i + 1; + ring->ridx = i; + ring->color = 1; + ring->base = dmam_alloc_coherent(dev, QDMA_INTR_RING_SIZE, + &ring->dev_base, GFP_KERNEL); + if (!ring->base) { + qdma_err(qdev, "Failed to alloc intr ring %d", i); + return -ENOMEM; + } + intr_ctxt.agg_base = QDMA_INTR_RING_BASE(ring->dev_base); + intr_ctxt.size = (QDMA_INTR_RING_SIZE - 1) / 4096; + intr_ctxt.vec = ring->msix_id; + intr_ctxt.valid = true; + intr_ctxt.color = true; + ret = qdma_prog_context(qdev, QDMA_CTXT_INTR_COAL, + QDMA_CTXT_CLEAR, ring->ridx, NULL); + if (ret) { + qdma_err(qdev, "Failed clear intr ctx, ret %d", ret); + return ret; + } + + qdma_prep_intr_context(qdev, &intr_ctxt, ctxt); + ret = qdma_prog_context(qdev, QDMA_CTXT_INTR_COAL, + QDMA_CTXT_WRITE, ring->ridx, ctxt); + if (ret) { + qdma_err(qdev, "Failed setup intr ctx, ret %d", ret); + return ret; + } + + ret = devm_request_threaded_irq(dev, vector, NULL, + qdma_queue_isr, IRQF_ONESHOT, + "amd-qdma-queue", ring); + if (ret) { + qdma_err(qdev, "Failed to request irq %d", vector); + return ret; + } + } + + return 0; +} + +static int qdma_intr_init(struct qdma_device *qdev) +{ + int ret; + + ret = qdma_init_error_irq(qdev); + if (ret) { + qdma_err(qdev, "Failed to init error IRQs, ret %d", ret); + return ret; + } + + ret = qdmam_alloc_qintr_rings(qdev); + if (ret) { + qdma_err(qdev, "Failed to init queue IRQs, ret %d", ret); + return ret; + } + + return 0; +} + +static void amd_qdma_remove(struct platform_device *pdev) +{ + struct qdma_device *qdev = platform_get_drvdata(pdev); + + qdma_sgdma_control(qdev, 0); + dma_async_device_unregister(&qdev->dma_dev); + + mutex_destroy(&qdev->ctxt_lock); +} + +static int amd_qdma_probe(struct platform_device *pdev) +{ + struct qdma_platdata *pdata = dev_get_platdata(&pdev->dev); + struct qdma_device *qdev; + struct resource *res; + void __iomem *regs; + int ret; + + qdev = devm_kzalloc(&pdev->dev, sizeof(*qdev), GFP_KERNEL); + if (!qdev) + return -ENOMEM; + + platform_set_drvdata(pdev, qdev); + qdev->pdev = pdev; + mutex_init(&qdev->ctxt_lock); + + res = platform_get_resource(pdev, IORESOURCE_IRQ, 0); + if (!res) { + qdma_err(qdev, "Failed to get IRQ resource"); + ret = -ENODEV; + goto failed; + } + qdev->err_irq_idx = pdata->irq_index; + qdev->queue_irq_start = res->start + 1; + qdev->queue_irq_num = resource_size(res) - 1; + + regs = devm_platform_get_and_ioremap_resource(pdev, 0, NULL); + if (IS_ERR(regs)) { + ret = PTR_ERR(regs); + qdma_err(qdev, "Failed to map IO resource, err %d", ret); + goto failed; + } + + qdev->regmap = devm_regmap_init_mmio(&pdev->dev, regs, + &qdma_regmap_config); + if (IS_ERR(qdev->regmap)) { + ret = PTR_ERR(qdev->regmap); + qdma_err(qdev, "Regmap init failed, err %d", ret); + goto failed; + } + + ret = qdma_device_verify(qdev); + if (ret) + goto failed; + + ret = qdma_get_hw_info(qdev); + if (ret) + goto failed; + + INIT_LIST_HEAD(&qdev->dma_dev.channels); + + ret = qdma_device_setup(qdev); + if (ret) + goto failed; + + ret = qdma_intr_init(qdev); + if (ret) { + qdma_err(qdev, "Failed to initialize IRQs %d", ret); + goto failed_disable_engine; + } + + dma_cap_set(DMA_SLAVE, qdev->dma_dev.cap_mask); + dma_cap_set(DMA_PRIVATE, qdev->dma_dev.cap_mask); + + qdev->dma_dev.dev = &pdev->dev; + qdev->dma_dev.filter.map = pdata->device_map; + qdev->dma_dev.filter.mapcnt = qdev->chan_num * 2; + qdev->dma_dev.filter.fn = qdma_filter_fn; + qdev->dma_dev.device_alloc_chan_resources = qdma_alloc_queue_resources; + qdev->dma_dev.device_free_chan_resources = qdma_free_queue_resources; + qdev->dma_dev.device_prep_slave_sg = qdma_prep_device_sg; + qdev->dma_dev.device_config = qdma_device_config; + qdev->dma_dev.device_issue_pending = qdma_issue_pending; + qdev->dma_dev.device_tx_status = dma_cookie_status; + qdev->dma_dev.directions = BIT(DMA_DEV_TO_MEM) | BIT(DMA_MEM_TO_DEV); + + ret = dma_async_device_register(&qdev->dma_dev); + if (ret) { + qdma_err(qdev, "Failed to register AMD QDMA: %d", ret); + goto failed_disable_engine; + } + + return 0; + +failed_disable_engine: + qdma_sgdma_control(qdev, 0); +failed: + mutex_destroy(&qdev->ctxt_lock); + qdma_err(qdev, "Failed to probe AMD QDMA driver"); + return ret; +} + +static struct platform_driver amd_qdma_driver = { + .driver = { + .name = "amd-qdma", + }, + .probe = amd_qdma_probe, + .remove_new = amd_qdma_remove, +}; + +module_platform_driver(amd_qdma_driver); + +MODULE_DESCRIPTION("AMD QDMA driver"); +MODULE_AUTHOR("XRT Team <runtimeca39d@amd.com>"); +MODULE_LICENSE("GPL"); diff --git a/drivers/dma/amd/qdma/qdma.h b/drivers/dma/amd/qdma/qdma.h new file mode 100644 index 000000000000..94089f1f0c11 --- /dev/null +++ b/drivers/dma/amd/qdma/qdma.h @@ -0,0 +1,266 @@ +/* SPDX-License-Identifier: GPL-2.0-or-later */ +/* + * DMA header for AMD Queue-based DMA Subsystem + * + * Copyright (C) 2023-2024, Advanced Micro Devices, Inc. + */ + +#ifndef __QDMA_H +#define __QDMA_H + +#include <linux/bitfield.h> +#include <linux/dmaengine.h> +#include <linux/kernel.h> +#include <linux/platform_device.h> +#include <linux/regmap.h> + +#include "../../virt-dma.h" + +#define DISABLE 0 +#define ENABLE 1 + +#define QDMA_MIN_IRQ 3 +#define QDMA_INTR_NAME_MAX_LEN 30 +#define QDMA_INTR_PREFIX "amd-qdma" + +#define QDMA_IDENTIFIER 0x1FD3 +#define QDMA_DEFAULT_RING_SIZE (BIT(10) + 1) +#define QDMA_DEFAULT_RING_ID 0 +#define QDMA_POLL_INTRVL_US 10 /* 10us */ +#define QDMA_POLL_TIMEOUT_US (500 * 1000) /* 500ms */ +#define QDMA_DMAP_REG_STRIDE 16 +#define QDMA_CTXT_REGMAP_LEN 8 /* 8 regs */ +#define QDMA_MM_DESC_SIZE 32 /* Bytes */ +#define QDMA_MM_DESC_LEN_BITS 28 +#define QDMA_MM_DESC_MAX_LEN (BIT(QDMA_MM_DESC_LEN_BITS) - 1) +#define QDMA_MIN_DMA_ALLOC_SIZE 4096 +#define QDMA_INTR_RING_SIZE BIT(13) +#define QDMA_INTR_RING_IDX_MASK GENMASK(9, 0) +#define QDMA_INTR_RING_BASE(_addr) ((_addr) >> 12) + +#define QDMA_IDENTIFIER_REGOFF 0x0 +#define QDMA_IDENTIFIER_MASK GENMASK(31, 16) +#define QDMA_QUEUE_ARM_BIT BIT(16) + +#define qdma_err(qdev, fmt, args...) \ + dev_err(&(qdev)->pdev->dev, fmt, ##args) + +#define qdma_dbg(qdev, fmt, args...) \ + dev_dbg(&(qdev)->pdev->dev, fmt, ##args) + +#define qdma_info(qdev, fmt, args...) \ + dev_info(&(qdev)->pdev->dev, fmt, ##args) + +enum qdma_reg_fields { + QDMA_REGF_IRQ_ENABLE, + QDMA_REGF_WBK_ENABLE, + QDMA_REGF_WBI_CHECK, + QDMA_REGF_IRQ_ARM, + QDMA_REGF_IRQ_VEC, + QDMA_REGF_IRQ_AGG, + QDMA_REGF_WBI_INTVL_ENABLE, + QDMA_REGF_MRKR_DISABLE, + QDMA_REGF_QUEUE_ENABLE, + QDMA_REGF_QUEUE_MODE, + QDMA_REGF_DESC_BASE, + QDMA_REGF_DESC_SIZE, + QDMA_REGF_RING_ID, + QDMA_REGF_CMD_INDX, + QDMA_REGF_CMD_CMD, + QDMA_REGF_CMD_TYPE, + QDMA_REGF_CMD_BUSY, + QDMA_REGF_QUEUE_COUNT, + QDMA_REGF_QUEUE_MAX, + QDMA_REGF_QUEUE_BASE, + QDMA_REGF_FUNCTION_ID, + QDMA_REGF_INTR_AGG_BASE, + QDMA_REGF_INTR_VECTOR, + QDMA_REGF_INTR_SIZE, + QDMA_REGF_INTR_VALID, + QDMA_REGF_INTR_COLOR, + QDMA_REGF_INTR_FUNCTION_ID, + QDMA_REGF_ERR_INT_FUNC, + QDMA_REGF_ERR_INT_VEC, + QDMA_REGF_ERR_INT_ARM, + QDMA_REGF_MAX +}; + +enum qdma_regs { + QDMA_REGO_CTXT_DATA, + QDMA_REGO_CTXT_CMD, + QDMA_REGO_CTXT_MASK, + QDMA_REGO_MM_H2C_CTRL, + QDMA_REGO_MM_C2H_CTRL, + QDMA_REGO_QUEUE_COUNT, + QDMA_REGO_RING_SIZE, + QDMA_REGO_H2C_PIDX, + QDMA_REGO_C2H_PIDX, + QDMA_REGO_INTR_CIDX, + QDMA_REGO_FUNC_ID, + QDMA_REGO_ERR_INT, + QDMA_REGO_ERR_STAT, + QDMA_REGO_MAX +}; + +struct qdma_reg_field { + u16 lsb; /* Least significant bit of field */ + u16 msb; /* Most significant bit of field */ +}; + +struct qdma_reg { + u32 off; + u32 count; +}; + +#define QDMA_REGF(_msb, _lsb) { \ + .lsb = (_lsb), \ + .msb = (_msb), \ +} + +#define QDMA_REGO(_off, _count) { \ + .off = (_off), \ + .count = (_count), \ +} + +enum qdma_desc_size { + QDMA_DESC_SIZE_8B, + QDMA_DESC_SIZE_16B, + QDMA_DESC_SIZE_32B, + QDMA_DESC_SIZE_64B, +}; + +enum qdma_queue_op_mode { + QDMA_QUEUE_OP_STREAM, + QDMA_QUEUE_OP_MM, +}; + +enum qdma_ctxt_type { + QDMA_CTXT_DESC_SW_C2H, + QDMA_CTXT_DESC_SW_H2C, + QDMA_CTXT_DESC_HW_C2H, + QDMA_CTXT_DESC_HW_H2C, + QDMA_CTXT_DESC_CR_C2H, + QDMA_CTXT_DESC_CR_H2C, + QDMA_CTXT_WRB, + QDMA_CTXT_PFTCH, + QDMA_CTXT_INTR_COAL, + QDMA_CTXT_RSVD, + QDMA_CTXT_HOST_PROFILE, + QDMA_CTXT_TIMER, + QDMA_CTXT_FMAP, + QDMA_CTXT_FNC_STS, +}; + +enum qdma_ctxt_cmd { + QDMA_CTXT_CLEAR, + QDMA_CTXT_WRITE, + QDMA_CTXT_READ, + QDMA_CTXT_INVALIDATE, + QDMA_CTXT_MAX +}; + +struct qdma_ctxt_sw_desc { + u64 desc_base; + u16 vec; +}; + +struct qdma_ctxt_intr { + u64 agg_base; + u16 vec; + u32 size; + bool valid; + bool color; +}; + +struct qdma_ctxt_fmap { + u16 qbase; + u16 qmax; +}; + +struct qdma_device; + +struct qdma_mm_desc { + __le64 src_addr; + __le32 len; + __le32 reserved1; + __le64 dst_addr; + __le64 reserved2; +} __packed; + +struct qdma_mm_vdesc { + struct virt_dma_desc vdesc; + struct qdma_queue *queue; + struct scatterlist *sgl; + u64 sg_off; + u32 sg_len; + u64 dev_addr; + u32 pidx; + u32 pending_descs; + struct dma_slave_config cfg; +}; + +#define QDMA_VDESC_QUEUED(vdesc) (!(vdesc)->sg_len) + +struct qdma_queue { + struct qdma_device *qdev; + struct virt_dma_chan vchan; + enum dma_transfer_direction dir; + struct dma_slave_config cfg; + struct qdma_mm_desc *desc_base; + struct qdma_mm_vdesc *submitted_vdesc; + struct qdma_mm_vdesc *issued_vdesc; + dma_addr_t dma_desc_base; + u32 pidx_reg; + u32 cidx_reg; + u32 ring_size; + u32 idx_mask; + u16 qid; + u32 pidx; + u32 cidx; +}; + +struct qdma_intr_ring { + struct qdma_device *qdev; + __le64 *base; + dma_addr_t dev_base; + char msix_name[QDMA_INTR_NAME_MAX_LEN]; + u32 msix_vector; + u16 msix_id; + u32 ring_size; + u16 ridx; + u16 cidx; + u8 color; +}; + +#define QDMA_INTR_MASK_PIDX GENMASK_ULL(15, 0) +#define QDMA_INTR_MASK_CIDX GENMASK_ULL(31, 16) +#define QDMA_INTR_MASK_DESC_COLOR GENMASK_ULL(32, 32) +#define QDMA_INTR_MASK_STATE GENMASK_ULL(34, 33) +#define QDMA_INTR_MASK_ERROR GENMASK_ULL(36, 35) +#define QDMA_INTR_MASK_TYPE GENMASK_ULL(38, 38) +#define QDMA_INTR_MASK_QID GENMASK_ULL(62, 39) +#define QDMA_INTR_MASK_COLOR GENMASK_ULL(63, 63) + +struct qdma_device { + struct platform_device *pdev; + struct dma_device dma_dev; + struct regmap *regmap; + struct mutex ctxt_lock; /* protect ctxt registers */ + const struct qdma_reg_field *rfields; + const struct qdma_reg *roffs; + struct qdma_queue *h2c_queues; + struct qdma_queue *c2h_queues; + struct qdma_intr_ring *qintr_rings; + u32 qintr_ring_num; + u32 qintr_ring_idx; + u32 chan_num; + u32 queue_irq_start; + u32 queue_irq_num; + u32 err_irq_idx; + u32 fid; +}; + +extern const struct qdma_reg qdma_regos_default[QDMA_REGO_MAX]; +extern const struct qdma_reg_field qdma_regfs_default[QDMA_REGF_MAX]; + +#endif /* __QDMA_H */ diff --git a/drivers/dma/at_hdmac.c b/drivers/dma/at_hdmac.c index 40052d1bd0b5..baebddc740b0 100644 --- a/drivers/dma/at_hdmac.c +++ b/drivers/dma/at_hdmac.c @@ -339,7 +339,7 @@ static inline u8 convert_buswidth(enum dma_slave_buswidth addr_width) * @regs: memory mapped register base * @clk: dma controller clock * @save_imr: interrupt mask register that is saved on suspend/resume cycle - * @all_chan_mask: all channels availlable in a mask + * @all_chan_mask: all channels available in a mask * @lli_pool: hw lli table * @memset_pool: hw memset pool * @chan: channels table to store at_dma_chan structures @@ -668,7 +668,7 @@ static inline u32 atc_calc_bytes_left(u32 current_len, u32 ctrla) * CTRLA is read in turn, next the DSCR is read a second time. If the two * consecutive read values of the DSCR are the same then we assume both refers * to the very same LLI as well as the CTRLA value read inbetween does. For - * cyclic tranfers, the assumption is that a full loop is "not so fast". If the + * cyclic transfers, the assumption is that a full loop is "not so fast". If the * two DSCR values are different, we read again the CTRLA then the DSCR till two * consecutive read values from DSCR are equal or till the maximum trials is * reach. This algorithm is very unlikely not to find a stable value for DSCR. @@ -700,7 +700,7 @@ static int atc_get_llis_residue(struct at_dma_chan *atchan, break; /* - * DSCR has changed inside the DMA controller, so the previouly + * DSCR has changed inside the DMA controller, so the previously * read value of CTRLA may refer to an already processed * descriptor hence could be outdated. We need to update ctrla * to match the current descriptor. diff --git a/drivers/dma/bcm-sba-raid.c b/drivers/dma/bcm-sba-raid.c index fbaacb4c19b2..cfa6e1167a1f 100644 --- a/drivers/dma/bcm-sba-raid.c +++ b/drivers/dma/bcm-sba-raid.c @@ -15,7 +15,7 @@ * number of hardware rings over one or more SBA hardware devices. By * design, the internal buffer size of SBA hardware device is limited * but all offload operations supported by SBA can be broken down into - * multiple small size requests and executed parallely on multiple SBA + * multiple small size requests and executed parallelly on multiple SBA * hardware devices for achieving high through-put. * * The Broadcom SBA RAID driver does not require any register programming @@ -135,7 +135,7 @@ struct sba_device { u32 max_xor_srcs; u32 max_resp_pool_size; u32 max_cmds_pool_size; - /* Maibox client and Mailbox channels */ + /* Mailbox client and Mailbox channels */ struct mbox_client client; struct mbox_chan *mchan; struct device *mbox_dev; diff --git a/drivers/dma/bcm2835-dma.c b/drivers/dma/bcm2835-dma.c index 9d74fe97452e..e1b92b4d7b05 100644 --- a/drivers/dma/bcm2835-dma.c +++ b/drivers/dma/bcm2835-dma.c @@ -369,7 +369,7 @@ static struct bcm2835_desc *bcm2835_dma_create_cb_chain( /* the last frame requires extra flags */ d->cb_list[d->frames - 1].cb->info |= finalextrainfo; - /* detect a size missmatch */ + /* detect a size mismatch */ if (buf_len && (d->size != buf_len)) goto error_cb; diff --git a/drivers/dma/dmaengine.c b/drivers/dma/dmaengine.c index c380a4dda77a..c1357d7f3dc6 100644 --- a/drivers/dma/dmaengine.c +++ b/drivers/dma/dmaengine.c @@ -1070,7 +1070,7 @@ static int __dma_async_device_channel_register(struct dma_device *device, if (!name) dev_set_name(&chan->dev->device, "dma%dchan%d", device->dev_id, chan->chan_id); else - dev_set_name(&chan->dev->device, name); + dev_set_name(&chan->dev->device, "%s", name); rc = device_register(&chan->dev->device); if (rc) goto err_out_ida; diff --git a/drivers/dma/dmatest.c b/drivers/dma/dmatest.c index 1f201a542b37..91b2fbc0b864 100644 --- a/drivers/dma/dmatest.c +++ b/drivers/dma/dmatest.c @@ -500,7 +500,7 @@ static unsigned long long dmatest_persec(s64 runtime, unsigned int val) per_sec *= val; per_sec = INT_TO_FIXPT(per_sec); - do_div(per_sec, runtime); + do_div(per_sec, (u32)runtime); return per_sec; } diff --git a/drivers/dma/dw-edma/dw-hdma-v0-core.c b/drivers/dma/dw-edma/dw-hdma-v0-core.c index 10e8f0715114..e3f8db4fe909 100644 --- a/drivers/dma/dw-edma/dw-hdma-v0-core.c +++ b/drivers/dma/dw-edma/dw-hdma-v0-core.c @@ -17,8 +17,8 @@ enum dw_hdma_control { DW_HDMA_V0_CB = BIT(0), DW_HDMA_V0_TCB = BIT(1), DW_HDMA_V0_LLP = BIT(2), - DW_HDMA_V0_LIE = BIT(3), - DW_HDMA_V0_RIE = BIT(4), + DW_HDMA_V0_LWIE = BIT(3), + DW_HDMA_V0_RWIE = BIT(4), DW_HDMA_V0_CCS = BIT(8), DW_HDMA_V0_LLE = BIT(9), }; @@ -195,25 +195,14 @@ static void dw_hdma_v0_write_ll_link(struct dw_edma_chunk *chunk, static void dw_hdma_v0_core_write_chunk(struct dw_edma_chunk *chunk) { struct dw_edma_burst *child; - struct dw_edma_chan *chan = chunk->chan; u32 control = 0, i = 0; - int j; if (chunk->cb) control = DW_HDMA_V0_CB; - j = chunk->bursts_alloc; - list_for_each_entry(child, &chunk->burst->list, list) { - j--; - if (!j) { - control |= DW_HDMA_V0_LIE; - if (!(chan->dw->chip->flags & DW_EDMA_CHIP_LOCAL)) - control |= DW_HDMA_V0_RIE; - } - + list_for_each_entry(child, &chunk->burst->list, list) dw_hdma_v0_write_ll_data(chunk, i++, control, child->sz, child->sar, child->dar); - } control = DW_HDMA_V0_LLP | DW_HDMA_V0_TCB; if (!chunk->cb) @@ -247,10 +236,11 @@ static void dw_hdma_v0_core_start(struct dw_edma_chunk *chunk, bool first) if (first) { /* Enable engine */ SET_CH_32(dw, chan->dir, chan->id, ch_en, BIT(0)); - /* Interrupt enable&unmask - done, abort */ - tmp = GET_CH_32(dw, chan->dir, chan->id, int_setup) | - HDMA_V0_STOP_INT_MASK | HDMA_V0_ABORT_INT_MASK | - HDMA_V0_LOCAL_STOP_INT_EN | HDMA_V0_LOCAL_ABORT_INT_EN; + /* Interrupt unmask - stop, abort */ + tmp = GET_CH_32(dw, chan->dir, chan->id, int_setup); + tmp &= ~(HDMA_V0_STOP_INT_MASK | HDMA_V0_ABORT_INT_MASK); + /* Interrupt enable - stop, abort */ + tmp |= HDMA_V0_LOCAL_STOP_INT_EN | HDMA_V0_LOCAL_ABORT_INT_EN; if (!(dw->chip->flags & DW_EDMA_CHIP_LOCAL)) tmp |= HDMA_V0_REMOTE_STOP_INT_EN | HDMA_V0_REMOTE_ABORT_INT_EN; SET_CH_32(dw, chan->dir, chan->id, int_setup, tmp); diff --git a/drivers/dma/dw/core.c b/drivers/dma/dw/core.c index 5f7d690e3dba..dd75f97a33b3 100644 --- a/drivers/dma/dw/core.c +++ b/drivers/dma/dw/core.c @@ -16,6 +16,7 @@ #include <linux/init.h> #include <linux/interrupt.h> #include <linux/io.h> +#include <linux/log2.h> #include <linux/mm.h> #include <linux/module.h> #include <linux/slab.h> @@ -621,12 +622,10 @@ dwc_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl, struct dw_desc *prev; struct dw_desc *first; u32 ctllo, ctlhi; - u8 m_master = dwc->dws.m_master; - u8 lms = DWC_LLP_LMS(m_master); + u8 lms = DWC_LLP_LMS(dwc->dws.m_master); dma_addr_t reg; unsigned int reg_width; unsigned int mem_width; - unsigned int data_width = dw->pdata->data_width[m_master]; unsigned int i; struct scatterlist *sg; size_t total_len = 0; @@ -660,7 +659,7 @@ dwc_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl, mem = sg_dma_address(sg); len = sg_dma_len(sg); - mem_width = __ffs(data_width | mem | len); + mem_width = __ffs(sconfig->src_addr_width | mem | len); slave_sg_todev_fill_desc: desc = dwc_desc_get(dwc); @@ -720,7 +719,7 @@ slave_sg_fromdev_fill_desc: lli_write(desc, sar, reg); lli_write(desc, dar, mem); lli_write(desc, ctlhi, ctlhi); - mem_width = __ffs(data_width | mem); + mem_width = __ffs(sconfig->dst_addr_width | mem); lli_write(desc, ctllo, ctllo | DWC_CTLL_DST_WIDTH(mem_width)); desc->len = dlen; @@ -780,20 +779,108 @@ bool dw_dma_filter(struct dma_chan *chan, void *param) } EXPORT_SYMBOL_GPL(dw_dma_filter); -static int dwc_config(struct dma_chan *chan, struct dma_slave_config *sconfig) +static int dwc_verify_maxburst(struct dma_chan *chan) { struct dw_dma_chan *dwc = to_dw_dma_chan(chan); - struct dw_dma *dw = to_dw_dma(chan->device); - memcpy(&dwc->dma_sconfig, sconfig, sizeof(*sconfig)); + dwc->dma_sconfig.src_maxburst = + clamp(dwc->dma_sconfig.src_maxburst, 1U, dwc->max_burst); + dwc->dma_sconfig.dst_maxburst = + clamp(dwc->dma_sconfig.dst_maxburst, 1U, dwc->max_burst); dwc->dma_sconfig.src_maxburst = - clamp(dwc->dma_sconfig.src_maxburst, 0U, dwc->max_burst); + rounddown_pow_of_two(dwc->dma_sconfig.src_maxburst); dwc->dma_sconfig.dst_maxburst = - clamp(dwc->dma_sconfig.dst_maxburst, 0U, dwc->max_burst); + rounddown_pow_of_two(dwc->dma_sconfig.dst_maxburst); - dw->encode_maxburst(dwc, &dwc->dma_sconfig.src_maxburst); - dw->encode_maxburst(dwc, &dwc->dma_sconfig.dst_maxburst); + return 0; +} + +static int dwc_verify_p_buswidth(struct dma_chan *chan) +{ + struct dw_dma_chan *dwc = to_dw_dma_chan(chan); + struct dw_dma *dw = to_dw_dma(chan->device); + u32 reg_width, max_width; + + if (dwc->dma_sconfig.direction == DMA_MEM_TO_DEV) + reg_width = dwc->dma_sconfig.dst_addr_width; + else if (dwc->dma_sconfig.direction == DMA_DEV_TO_MEM) + reg_width = dwc->dma_sconfig.src_addr_width; + else /* DMA_MEM_TO_MEM */ + return 0; + + max_width = dw->pdata->data_width[dwc->dws.p_master]; + + /* Fall-back to 1-byte transfer width if undefined */ + if (reg_width == DMA_SLAVE_BUSWIDTH_UNDEFINED) + reg_width = DMA_SLAVE_BUSWIDTH_1_BYTE; + else if (!is_power_of_2(reg_width) || reg_width > max_width) + return -EINVAL; + else /* bus width is valid */ + return 0; + + /* Update undefined addr width value */ + if (dwc->dma_sconfig.direction == DMA_MEM_TO_DEV) + dwc->dma_sconfig.dst_addr_width = reg_width; + else /* DMA_DEV_TO_MEM */ + dwc->dma_sconfig.src_addr_width = reg_width; + + return 0; +} + +static int dwc_verify_m_buswidth(struct dma_chan *chan) +{ + struct dw_dma_chan *dwc = to_dw_dma_chan(chan); + struct dw_dma *dw = to_dw_dma(chan->device); + u32 reg_width, reg_burst, mem_width; + + mem_width = dw->pdata->data_width[dwc->dws.m_master]; + + /* + * It's possible to have a data portion locked in the DMA FIFO in case + * of the channel suspension. Subsequent channel disabling will cause + * that data silent loss. In order to prevent that maintain the src and + * dst transfer widths coherency by means of the relation: + * (CTLx.SRC_TR_WIDTH * CTLx.SRC_MSIZE >= CTLx.DST_TR_WIDTH) + * Look for the details in the commit message that brings this change. + * + * Note the DMA configs utilized in the calculations below must have + * been verified to have correct values by this method call. + */ + if (dwc->dma_sconfig.direction == DMA_MEM_TO_DEV) { + reg_width = dwc->dma_sconfig.dst_addr_width; + if (mem_width < reg_width) + return -EINVAL; + + dwc->dma_sconfig.src_addr_width = mem_width; + } else if (dwc->dma_sconfig.direction == DMA_DEV_TO_MEM) { + reg_width = dwc->dma_sconfig.src_addr_width; + reg_burst = dwc->dma_sconfig.src_maxburst; + + dwc->dma_sconfig.dst_addr_width = min(mem_width, reg_width * reg_burst); + } + + return 0; +} + +static int dwc_config(struct dma_chan *chan, struct dma_slave_config *sconfig) +{ + struct dw_dma_chan *dwc = to_dw_dma_chan(chan); + int ret; + + memcpy(&dwc->dma_sconfig, sconfig, sizeof(*sconfig)); + + ret = dwc_verify_maxburst(chan); + if (ret) + return ret; + + ret = dwc_verify_p_buswidth(chan); + if (ret) + return ret; + + ret = dwc_verify_m_buswidth(chan); + if (ret) + return ret; return 0; } @@ -1068,7 +1155,7 @@ int do_dma_probe(struct dw_dma_chip *chip) bool autocfg = false; unsigned int dw_params; unsigned int i; - int err; + int ret; dw->pdata = devm_kzalloc(chip->dev, sizeof(*dw->pdata), GFP_KERNEL); if (!dw->pdata) @@ -1084,7 +1171,7 @@ int do_dma_probe(struct dw_dma_chip *chip) autocfg = dw_params >> DW_PARAMS_EN & 1; if (!autocfg) { - err = -EINVAL; + ret = -EINVAL; goto err_pdata; } @@ -1104,7 +1191,7 @@ int do_dma_probe(struct dw_dma_chip *chip) pdata->chan_allocation_order = CHAN_ALLOCATION_ASCENDING; pdata->chan_priority = CHAN_PRIORITY_ASCENDING; } else if (chip->pdata->nr_channels > DW_DMA_MAX_NR_CHANNELS) { - err = -EINVAL; + ret = -EINVAL; goto err_pdata; } else { memcpy(dw->pdata, chip->pdata, sizeof(*dw->pdata)); @@ -1116,7 +1203,7 @@ int do_dma_probe(struct dw_dma_chip *chip) dw->chan = devm_kcalloc(chip->dev, pdata->nr_channels, sizeof(*dw->chan), GFP_KERNEL); if (!dw->chan) { - err = -ENOMEM; + ret = -ENOMEM; goto err_pdata; } @@ -1134,15 +1221,15 @@ int do_dma_probe(struct dw_dma_chip *chip) sizeof(struct dw_desc), 4, 0); if (!dw->desc_pool) { dev_err(chip->dev, "No memory for descriptors dma pool\n"); - err = -ENOMEM; + ret = -ENOMEM; goto err_pdata; } tasklet_setup(&dw->tasklet, dw_dma_tasklet); - err = request_irq(chip->irq, dw_dma_interrupt, IRQF_SHARED, + ret = request_irq(chip->irq, dw_dma_interrupt, IRQF_SHARED, dw->name, dw); - if (err) + if (ret) goto err_pdata; INIT_LIST_HEAD(&dw->dma.channels); @@ -1254,8 +1341,8 @@ int do_dma_probe(struct dw_dma_chip *chip) */ dma_set_max_seg_size(dw->dma.dev, dw->chan[0].block_size); - err = dma_async_device_register(&dw->dma); - if (err) + ret = dma_async_device_register(&dw->dma); + if (ret) goto err_dma_register; dev_info(chip->dev, "DesignWare DMA Controller, %d channels\n", @@ -1269,7 +1356,7 @@ err_dma_register: free_irq(chip->irq, dw); err_pdata: pm_runtime_put_sync_suspend(chip->dev); - return err; + return ret; } int do_dma_remove(struct dw_dma_chip *chip) diff --git a/drivers/dma/dw/dw.c b/drivers/dma/dw/dw.c index a4862263ff14..6766142884b6 100644 --- a/drivers/dma/dw/dw.c +++ b/drivers/dma/dw/dw.c @@ -64,30 +64,39 @@ static size_t dw_dma_block2bytes(struct dw_dma_chan *dwc, u32 block, u32 width) return DWC_CTLH_BLOCK_TS(block) << width; } +static inline u8 dw_dma_encode_maxburst(u32 maxburst) +{ + /* + * Fix burst size according to dw_dmac. We need to convert them as: + * 1 -> 0, 4 -> 1, 8 -> 2, 16 -> 3. + */ + return maxburst > 1 ? fls(maxburst) - 2 : 0; +} + static u32 dw_dma_prepare_ctllo(struct dw_dma_chan *dwc) { struct dma_slave_config *sconfig = &dwc->dma_sconfig; - u8 smsize = (dwc->direction == DMA_DEV_TO_MEM) ? sconfig->src_maxburst : 0; - u8 dmsize = (dwc->direction == DMA_MEM_TO_DEV) ? sconfig->dst_maxburst : 0; - u8 p_master = dwc->dws.p_master; - u8 m_master = dwc->dws.m_master; - u8 dms = (dwc->direction == DMA_MEM_TO_DEV) ? p_master : m_master; - u8 sms = (dwc->direction == DMA_DEV_TO_MEM) ? p_master : m_master; + u8 smsize = 0, dmsize = 0; + u8 sms, dms; + + if (dwc->direction == DMA_MEM_TO_DEV) { + sms = dwc->dws.m_master; + dms = dwc->dws.p_master; + dmsize = dw_dma_encode_maxburst(sconfig->dst_maxburst); + } else if (dwc->direction == DMA_DEV_TO_MEM) { + sms = dwc->dws.p_master; + dms = dwc->dws.m_master; + smsize = dw_dma_encode_maxburst(sconfig->src_maxburst); + } else /* DMA_MEM_TO_MEM */ { + sms = dwc->dws.m_master; + dms = dwc->dws.m_master; + } return DWC_CTLL_LLP_D_EN | DWC_CTLL_LLP_S_EN | DWC_CTLL_DST_MSIZE(dmsize) | DWC_CTLL_SRC_MSIZE(smsize) | DWC_CTLL_DMS(dms) | DWC_CTLL_SMS(sms); } -static void dw_dma_encode_maxburst(struct dw_dma_chan *dwc, u32 *maxburst) -{ - /* - * Fix burst size according to dw_dmac. We need to convert them as: - * 1 -> 0, 4 -> 1, 8 -> 2, 16 -> 3. - */ - *maxburst = *maxburst > 1 ? fls(*maxburst) - 2 : 0; -} - static void dw_dma_set_device_name(struct dw_dma *dw, int id) { snprintf(dw->name, sizeof(dw->name), "dw:dmac%d", id); @@ -116,7 +125,6 @@ int dw_dma_probe(struct dw_dma_chip *chip) dw->suspend_chan = dw_dma_suspend_chan; dw->resume_chan = dw_dma_resume_chan; dw->prepare_ctllo = dw_dma_prepare_ctllo; - dw->encode_maxburst = dw_dma_encode_maxburst; dw->bytes2block = dw_dma_bytes2block; dw->block2bytes = dw_dma_block2bytes; diff --git a/drivers/dma/dw/idma32.c b/drivers/dma/dw/idma32.c index 58f4078d83fe..dac617c183e6 100644 --- a/drivers/dma/dw/idma32.c +++ b/drivers/dma/dw/idma32.c @@ -199,21 +199,25 @@ static size_t idma32_block2bytes(struct dw_dma_chan *dwc, u32 block, u32 width) return IDMA32C_CTLH_BLOCK_TS(block); } +static inline u8 idma32_encode_maxburst(u32 maxburst) +{ + return maxburst > 1 ? fls(maxburst) - 1 : 0; +} + static u32 idma32_prepare_ctllo(struct dw_dma_chan *dwc) { struct dma_slave_config *sconfig = &dwc->dma_sconfig; - u8 smsize = (dwc->direction == DMA_DEV_TO_MEM) ? sconfig->src_maxburst : 0; - u8 dmsize = (dwc->direction == DMA_MEM_TO_DEV) ? sconfig->dst_maxburst : 0; + u8 smsize = 0, dmsize = 0; + + if (dwc->direction == DMA_MEM_TO_DEV) + dmsize = idma32_encode_maxburst(sconfig->dst_maxburst); + else if (dwc->direction == DMA_DEV_TO_MEM) + smsize = idma32_encode_maxburst(sconfig->src_maxburst); return DWC_CTLL_LLP_D_EN | DWC_CTLL_LLP_S_EN | DWC_CTLL_DST_MSIZE(dmsize) | DWC_CTLL_SRC_MSIZE(smsize); } -static void idma32_encode_maxburst(struct dw_dma_chan *dwc, u32 *maxburst) -{ - *maxburst = *maxburst > 1 ? fls(*maxburst) - 1 : 0; -} - static void idma32_set_device_name(struct dw_dma *dw, int id) { snprintf(dw->name, sizeof(dw->name), "idma32:dmac%d", id); @@ -270,7 +274,6 @@ int idma32_dma_probe(struct dw_dma_chip *chip) dw->suspend_chan = idma32_suspend_chan; dw->resume_chan = idma32_resume_chan; dw->prepare_ctllo = idma32_prepare_ctllo; - dw->encode_maxburst = idma32_encode_maxburst; dw->bytes2block = idma32_bytes2block; dw->block2bytes = idma32_block2bytes; diff --git a/drivers/dma/dw/platform.c b/drivers/dma/dw/platform.c index 7d9d4c951724..47c58ad468cb 100644 --- a/drivers/dma/dw/platform.c +++ b/drivers/dma/dw/platform.c @@ -29,7 +29,7 @@ static int dw_probe(struct platform_device *pdev) struct dw_dma_chip_pdata *data; struct dw_dma_chip *chip; struct device *dev = &pdev->dev; - int err; + int ret; match = device_get_match_data(dev); if (!match) @@ -51,9 +51,9 @@ static int dw_probe(struct platform_device *pdev) if (IS_ERR(chip->regs)) return PTR_ERR(chip->regs); - err = dma_coerce_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(32)); - if (err) - return err; + ret = dma_coerce_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(32)); + if (ret) + return ret; if (!data->pdata) data->pdata = dev_get_platdata(dev); @@ -69,14 +69,14 @@ static int dw_probe(struct platform_device *pdev) chip->clk = devm_clk_get_optional(chip->dev, "hclk"); if (IS_ERR(chip->clk)) return PTR_ERR(chip->clk); - err = clk_prepare_enable(chip->clk); - if (err) - return err; + ret = clk_prepare_enable(chip->clk); + if (ret) + return ret; pm_runtime_enable(&pdev->dev); - err = data->probe(chip); - if (err) + ret = data->probe(chip); + if (ret) goto err_dw_dma_probe; platform_set_drvdata(pdev, data); @@ -90,7 +90,7 @@ static int dw_probe(struct platform_device *pdev) err_dw_dma_probe: pm_runtime_disable(&pdev->dev); clk_disable_unprepare(chip->clk); - return err; + return ret; } static void dw_remove(struct platform_device *pdev) diff --git a/drivers/dma/dw/regs.h b/drivers/dma/dw/regs.h index 76654bd13c1a..5969d9cc8d7a 100644 --- a/drivers/dma/dw/regs.h +++ b/drivers/dma/dw/regs.h @@ -327,7 +327,6 @@ struct dw_dma { void (*suspend_chan)(struct dw_dma_chan *dwc, bool drain); void (*resume_chan)(struct dw_dma_chan *dwc, bool drain); u32 (*prepare_ctllo)(struct dw_dma_chan *dwc); - void (*encode_maxburst)(struct dw_dma_chan *dwc, u32 *maxburst); u32 (*bytes2block)(struct dw_dma_chan *dwc, size_t bytes, unsigned int width, size_t *len); size_t (*block2bytes)(struct dw_dma_chan *dwc, u32 block, u32 width); diff --git a/drivers/dma/ep93xx_dma.c b/drivers/dma/ep93xx_dma.c index 8514b9ab8517..995427afe077 100644 --- a/drivers/dma/ep93xx_dma.c +++ b/drivers/dma/ep93xx_dma.c @@ -898,7 +898,7 @@ static dma_cookie_t ep93xx_dma_tx_submit(struct dma_async_tx_descriptor *tx) desc = container_of(tx, struct ep93xx_dma_desc, txd); /* - * If nothing is currently prosessed, we push this descriptor + * If nothing is currently processed, we push this descriptor * directly to the hardware. Otherwise we put the descriptor * to the pending queue. */ @@ -1076,7 +1076,7 @@ fail: * @chan: channel * @sgl: list of buffers to transfer * @sg_len: number of entries in @sgl - * @dir: direction of tha DMA transfer + * @dir: direction of the DMA transfer * @flags: flags for the descriptor * @context: operation context (ignored) * diff --git a/drivers/dma/fsl-dpaa2-qdma/dpaa2-qdma.h b/drivers/dma/fsl-dpaa2-qdma/dpaa2-qdma.h index 2c80077cb7c0..36c284a3d184 100644 --- a/drivers/dma/fsl-dpaa2-qdma/dpaa2-qdma.h +++ b/drivers/dma/fsl-dpaa2-qdma/dpaa2-qdma.h @@ -12,8 +12,8 @@ struct dpaa2_qdma_sd_d { u32 rsv:32; union { struct { - u32 ssd:12; /* souce stride distance */ - u32 sss:12; /* souce stride size */ + u32 ssd:12; /* source stride distance */ + u32 sss:12; /* source stride size */ u32 rsv1:8; } sdf; struct { @@ -48,7 +48,7 @@ struct dpaa2_qdma_sd_d { #define QDMA_SER_DISABLE (8) /* no notification */ #define QDMA_SER_CTX BIT(8) /* notification by FQD_CTX[fqid] */ #define QDMA_SER_DEST (2 << 8) /* notification by destination desc */ -#define QDMA_SER_BOTH (3 << 8) /* soruce and dest notification */ +#define QDMA_SER_BOTH (3 << 8) /* source and dest notification */ #define QDMA_FD_SPF_ENALBE BIT(30) /* source prefetch enable */ #define QMAN_FD_VA_ENABLE BIT(14) /* Address used is virtual address */ diff --git a/drivers/dma/fsl-edma-main.c b/drivers/dma/fsl-edma-main.c index c66185c5a199..f9f1eda79254 100644 --- a/drivers/dma/fsl-edma-main.c +++ b/drivers/dma/fsl-edma-main.c @@ -100,6 +100,22 @@ static irqreturn_t fsl_edma_irq_handler(int irq, void *dev_id) return fsl_edma_err_handler(irq, dev_id); } +static bool fsl_edma_srcid_in_use(struct fsl_edma_engine *fsl_edma, u32 srcid) +{ + struct fsl_edma_chan *fsl_chan; + int i; + + for (i = 0; i < fsl_edma->n_chans; i++) { + fsl_chan = &fsl_edma->chans[i]; + + if (fsl_chan->srcid && srcid == fsl_chan->srcid) { + dev_err(&fsl_chan->pdev->dev, "The srcid is in use, can't use!"); + return true; + } + } + return false; +} + static struct dma_chan *fsl_edma_xlate(struct of_phandle_args *dma_spec, struct of_dma *ofdma) { @@ -117,6 +133,10 @@ static struct dma_chan *fsl_edma_xlate(struct of_phandle_args *dma_spec, list_for_each_entry_safe(chan, _chan, &fsl_edma->dma_dev.channels, device_node) { if (chan->client_count) continue; + + if (fsl_edma_srcid_in_use(fsl_edma, dma_spec->args[1])) + return NULL; + if ((chan->chan_id / chans_per_mux) == dma_spec->args[0]) { chan = dma_get_slave_channel(chan); if (chan) { @@ -153,7 +173,7 @@ static struct dma_chan *fsl_edma3_xlate(struct of_phandle_args *dma_spec, b_chmux = !!(fsl_edma->drvdata->flags & FSL_EDMA_DRV_HAS_CHMUX); - mutex_lock(&fsl_edma->fsl_edma_mutex); + guard(mutex)(&fsl_edma->fsl_edma_mutex); list_for_each_entry_safe(chan, _chan, &fsl_edma->dma_dev.channels, device_node) { @@ -161,6 +181,8 @@ static struct dma_chan *fsl_edma3_xlate(struct of_phandle_args *dma_spec, continue; fsl_chan = to_fsl_edma_chan(chan); + if (fsl_edma_srcid_in_use(fsl_edma, dma_spec->args[0])) + return NULL; i = fsl_chan - fsl_edma->chans; fsl_chan->priority = dma_spec->args[1]; @@ -177,18 +199,15 @@ static struct dma_chan *fsl_edma3_xlate(struct of_phandle_args *dma_spec, if (!b_chmux && i == dma_spec->args[0]) { chan = dma_get_slave_channel(chan); chan->device->privatecnt++; - mutex_unlock(&fsl_edma->fsl_edma_mutex); return chan; } else if (b_chmux && !fsl_chan->srcid) { /* if controller support channel mux, choose a free channel */ chan = dma_get_slave_channel(chan); chan->device->privatecnt++; fsl_chan->srcid = dma_spec->args[0]; - mutex_unlock(&fsl_edma->fsl_edma_mutex); return chan; } } - mutex_unlock(&fsl_edma->fsl_edma_mutex); return NULL; } diff --git a/drivers/dma/hisi_dma.c b/drivers/dma/hisi_dma.c index 4c47bff81064..25a4134be36b 100644 --- a/drivers/dma/hisi_dma.c +++ b/drivers/dma/hisi_dma.c @@ -677,7 +677,7 @@ static void hisi_dma_init_hw_qp(struct hisi_dma_dev *hdma_dev, u32 index) writel_relaxed(tmp, addr); /* - * 0 - dma should process FLR whith CPU. + * 0 - dma should process FLR with CPU. * 1 - dma not process FLR, only cpu process FLR. */ addr = q_base + HISI_DMA_HIP09_DMA_FLR_DISABLE + diff --git a/drivers/dma/idma64.c b/drivers/dma/idma64.c index e3505e56784b..3c648308a54a 100644 --- a/drivers/dma/idma64.c +++ b/drivers/dma/idma64.c @@ -290,7 +290,7 @@ static void idma64_desc_fill(struct idma64_chan *idma64c, desc->length += hw->len; } while (i); - /* Trigger an interrupt after the last block is transfered */ + /* Trigger an interrupt after the last block is transferred */ lli->ctllo |= IDMA64C_CTLL_INT_EN; /* Disable LLP transfer in the last block */ @@ -364,7 +364,7 @@ static size_t idma64_active_desc_size(struct idma64_chan *idma64c) if (!i) return bytes; - /* The current chunk is not fully transfered yet */ + /* The current chunk is not fully transferred yet */ bytes += desc->hw[--i].len; return bytes - IDMA64C_CTLH_BLOCK_TS(ctlhi); @@ -598,9 +598,7 @@ static int idma64_probe(struct idma64_chip *chip) idma64->dma.dev = chip->sysdev; - ret = dma_set_max_seg_size(idma64->dma.dev, IDMA64C_CTLH_BLOCK_TS_MASK); - if (ret) - return ret; + dma_set_max_seg_size(idma64->dma.dev, IDMA64C_CTLH_BLOCK_TS_MASK); ret = dma_async_device_register(&idma64->dma); if (ret) diff --git a/drivers/dma/idxd/idxd.h b/drivers/dma/idxd/idxd.h index 868b724a3b75..d84e21daa991 100644 --- a/drivers/dma/idxd/idxd.h +++ b/drivers/dma/idxd/idxd.h @@ -124,7 +124,6 @@ struct idxd_pmu { struct pmu pmu; char name[IDXD_NAME_SIZE]; - int cpu; int n_counters; int counter_width; @@ -135,8 +134,6 @@ struct idxd_pmu { unsigned long supported_filters; int n_filters; - - struct hlist_node cpuhp_node; }; #define IDXD_MAX_PRIORITY 0xf @@ -803,14 +800,10 @@ void idxd_user_counter_increment(struct idxd_wq *wq, u32 pasid, int index); int perfmon_pmu_init(struct idxd_device *idxd); void perfmon_pmu_remove(struct idxd_device *idxd); void perfmon_counter_overflow(struct idxd_device *idxd); -void perfmon_init(void); -void perfmon_exit(void); #else static inline int perfmon_pmu_init(struct idxd_device *idxd) { return 0; } static inline void perfmon_pmu_remove(struct idxd_device *idxd) {} static inline void perfmon_counter_overflow(struct idxd_device *idxd) {} -static inline void perfmon_init(void) {} -static inline void perfmon_exit(void) {} #endif /* debugfs */ diff --git a/drivers/dma/idxd/init.c b/drivers/dma/idxd/init.c index 21f6905b554d..234c1c658ec7 100644 --- a/drivers/dma/idxd/init.c +++ b/drivers/dma/idxd/init.c @@ -69,9 +69,15 @@ static struct idxd_driver_data idxd_driver_data[] = { static struct pci_device_id idxd_pci_tbl[] = { /* DSA ver 1.0 platforms */ { PCI_DEVICE_DATA(INTEL, DSA_SPR0, &idxd_driver_data[IDXD_TYPE_DSA]) }, + /* DSA on GNR-D platforms */ + { PCI_DEVICE_DATA(INTEL, DSA_GNRD, &idxd_driver_data[IDXD_TYPE_DSA]) }, + /* DSA on DMR platforms */ + { PCI_DEVICE_DATA(INTEL, DSA_DMR, &idxd_driver_data[IDXD_TYPE_DSA]) }, /* IAX ver 1.0 platforms */ { PCI_DEVICE_DATA(INTEL, IAX_SPR0, &idxd_driver_data[IDXD_TYPE_IAX]) }, + /* IAA on DMR platforms */ + { PCI_DEVICE_DATA(INTEL, IAA_DMR, &idxd_driver_data[IDXD_TYPE_IAX]) }, { 0, } }; MODULE_DEVICE_TABLE(pci, idxd_pci_tbl); @@ -878,8 +884,6 @@ static int __init idxd_init_module(void) else support_enqcmd = true; - perfmon_init(); - err = idxd_driver_register(&idxd_drv); if (err < 0) goto err_idxd_driver_register; @@ -928,7 +932,6 @@ static void __exit idxd_exit_module(void) idxd_driver_unregister(&idxd_drv); pci_unregister_driver(&idxd_pci_driver); idxd_cdev_remove(); - perfmon_exit(); idxd_remove_debugfs(); } module_exit(idxd_exit_module); diff --git a/drivers/dma/idxd/perfmon.c b/drivers/dma/idxd/perfmon.c index 5e94247e1ea7..4b6af2f15d8a 100644 --- a/drivers/dma/idxd/perfmon.c +++ b/drivers/dma/idxd/perfmon.c @@ -6,29 +6,6 @@ #include "idxd.h" #include "perfmon.h" -static ssize_t cpumask_show(struct device *dev, struct device_attribute *attr, - char *buf); - -static cpumask_t perfmon_dsa_cpu_mask; -static bool cpuhp_set_up; -static enum cpuhp_state cpuhp_slot; - -/* - * perf userspace reads this attribute to determine which cpus to open - * counters on. It's connected to perfmon_dsa_cpu_mask, which is - * maintained by the cpu hotplug handlers. - */ -static DEVICE_ATTR_RO(cpumask); - -static struct attribute *perfmon_cpumask_attrs[] = { - &dev_attr_cpumask.attr, - NULL, -}; - -static struct attribute_group cpumask_attr_group = { - .attrs = perfmon_cpumask_attrs, -}; - /* * These attributes specify the bits in the config word that the perf * syscall uses to pass the event ids and categories to perfmon. @@ -67,16 +44,9 @@ static struct attribute_group perfmon_format_attr_group = { static const struct attribute_group *perfmon_attr_groups[] = { &perfmon_format_attr_group, - &cpumask_attr_group, NULL, }; -static ssize_t cpumask_show(struct device *dev, struct device_attribute *attr, - char *buf) -{ - return cpumap_print_to_pagebuf(true, buf, &perfmon_dsa_cpu_mask); -} - static bool is_idxd_event(struct idxd_pmu *idxd_pmu, struct perf_event *event) { return &idxd_pmu->pmu == event->pmu; @@ -217,7 +187,6 @@ static int perfmon_pmu_event_init(struct perf_event *event) return -EINVAL; event->hw.event_base = ioread64(PERFMON_TABLE_OFFSET(idxd)); - event->cpu = idxd->idxd_pmu->cpu; event->hw.config = event->attr.config; if (event->group_leader != event) @@ -480,14 +449,15 @@ static void idxd_pmu_init(struct idxd_pmu *idxd_pmu) idxd_pmu->pmu.attr_groups = perfmon_attr_groups; idxd_pmu->pmu.task_ctx_nr = perf_invalid_context; idxd_pmu->pmu.event_init = perfmon_pmu_event_init; - idxd_pmu->pmu.pmu_enable = perfmon_pmu_enable, - idxd_pmu->pmu.pmu_disable = perfmon_pmu_disable, + idxd_pmu->pmu.pmu_enable = perfmon_pmu_enable; + idxd_pmu->pmu.pmu_disable = perfmon_pmu_disable; idxd_pmu->pmu.add = perfmon_pmu_event_add; idxd_pmu->pmu.del = perfmon_pmu_event_del; idxd_pmu->pmu.start = perfmon_pmu_event_start; idxd_pmu->pmu.stop = perfmon_pmu_event_stop; idxd_pmu->pmu.read = perfmon_pmu_event_update; idxd_pmu->pmu.capabilities = PERF_PMU_CAP_NO_EXCLUDE; + idxd_pmu->pmu.scope = PERF_PMU_SCOPE_SYS_WIDE; idxd_pmu->pmu.module = THIS_MODULE; } @@ -496,47 +466,11 @@ void perfmon_pmu_remove(struct idxd_device *idxd) if (!idxd->idxd_pmu) return; - cpuhp_state_remove_instance(cpuhp_slot, &idxd->idxd_pmu->cpuhp_node); perf_pmu_unregister(&idxd->idxd_pmu->pmu); kfree(idxd->idxd_pmu); idxd->idxd_pmu = NULL; } -static int perf_event_cpu_online(unsigned int cpu, struct hlist_node *node) -{ - struct idxd_pmu *idxd_pmu; - - idxd_pmu = hlist_entry_safe(node, typeof(*idxd_pmu), cpuhp_node); - - /* select the first online CPU as the designated reader */ - if (cpumask_empty(&perfmon_dsa_cpu_mask)) { - cpumask_set_cpu(cpu, &perfmon_dsa_cpu_mask); - idxd_pmu->cpu = cpu; - } - - return 0; -} - -static int perf_event_cpu_offline(unsigned int cpu, struct hlist_node *node) -{ - struct idxd_pmu *idxd_pmu; - unsigned int target; - - idxd_pmu = hlist_entry_safe(node, typeof(*idxd_pmu), cpuhp_node); - - if (!cpumask_test_and_clear_cpu(cpu, &perfmon_dsa_cpu_mask)) - return 0; - - target = cpumask_any_but(cpu_online_mask, cpu); - /* migrate events if there is a valid target */ - if (target < nr_cpu_ids) { - cpumask_set_cpu(target, &perfmon_dsa_cpu_mask); - perf_pmu_migrate_context(&idxd_pmu->pmu, cpu, target); - } - - return 0; -} - int perfmon_pmu_init(struct idxd_device *idxd) { union idxd_perfcap perfcap; @@ -544,12 +478,6 @@ int perfmon_pmu_init(struct idxd_device *idxd) int rc = -ENODEV; /* - * perfmon module initialization failed, nothing to do - */ - if (!cpuhp_set_up) - return -ENODEV; - - /* * If perfmon_offset or num_counters is 0, it means perfmon is * not supported on this hardware. */ @@ -624,11 +552,6 @@ int perfmon_pmu_init(struct idxd_device *idxd) if (rc) goto free; - rc = cpuhp_state_add_instance(cpuhp_slot, &idxd_pmu->cpuhp_node); - if (rc) { - perf_pmu_unregister(&idxd->idxd_pmu->pmu); - goto free; - } out: return rc; free: @@ -637,22 +560,3 @@ free: goto out; } - -void __init perfmon_init(void) -{ - int rc = cpuhp_setup_state_multi(CPUHP_AP_ONLINE_DYN, - "driver/dma/idxd/perf:online", - perf_event_cpu_online, - perf_event_cpu_offline); - if (WARN_ON(rc < 0)) - return; - - cpuhp_slot = rc; - cpuhp_set_up = true; -} - -void __exit perfmon_exit(void) -{ - if (cpuhp_set_up) - cpuhp_remove_multi_state(cpuhp_slot); -} diff --git a/drivers/dma/idxd/submit.c b/drivers/dma/idxd/submit.c index 817a564413b0..94eca25ae9b9 100644 --- a/drivers/dma/idxd/submit.c +++ b/drivers/dma/idxd/submit.c @@ -134,7 +134,7 @@ static void llist_abort_desc(struct idxd_wq *wq, struct idxd_irq_entry *ie, * completing the descriptor will return desc to allocator and * the desc can be acquired by a different process and the * desc->list can be modified. Delete desc from list so the - * list trasversing does not get corrupted by the other process. + * list traversing does not get corrupted by the other process. */ list_for_each_entry_safe(d, t, &flist, list) { list_del_init(&d->list); diff --git a/drivers/dma/imx-dma.c b/drivers/dma/imx-dma.c index ebf7c115d553..e913f0db99da 100644 --- a/drivers/dma/imx-dma.c +++ b/drivers/dma/imx-dma.c @@ -167,7 +167,6 @@ struct imxdma_channel { enum imx_dma_type { IMX1_DMA, - IMX21_DMA, IMX27_DMA, }; @@ -195,8 +194,6 @@ static const struct of_device_id imx_dma_of_dev_id[] = { { .compatible = "fsl,imx1-dma", .data = (const void *)IMX1_DMA, }, { - .compatible = "fsl,imx21-dma", .data = (const void *)IMX21_DMA, - }, { .compatible = "fsl,imx27-dma", .data = (const void *)IMX27_DMA, }, { /* sentinel */ diff --git a/drivers/dma/ioat/init.c b/drivers/dma/ioat/init.c index 7b502b60b38b..cc9ddd6c325b 100644 --- a/drivers/dma/ioat/init.c +++ b/drivers/dma/ioat/init.c @@ -905,7 +905,7 @@ static int ioat_xor_val_self_test(struct ioatdma_device *ioat_dma) op = IOAT_OP_XOR_VAL; - /* validate the sources with the destintation page */ + /* validate the sources with the destination page */ for (i = 0; i < IOAT_NUM_SRC_TEST; i++) xor_val_srcs[i] = xor_srcs[i]; xor_val_srcs[i] = dest; diff --git a/drivers/dma/lgm/lgm-dma.c b/drivers/dma/lgm/lgm-dma.c index 4117c7b67e9c..8173c3f1075a 100644 --- a/drivers/dma/lgm/lgm-dma.c +++ b/drivers/dma/lgm/lgm-dma.c @@ -107,7 +107,7 @@ * If header mode is set in DMA descriptor, * If bit 30 is disabled, HDR_LEN must be configured according to channel * requirement. - * If bit 30 is enabled(checksum with heade mode), HDR_LEN has no need to + * If bit 30 is enabled(checksum with header mode), HDR_LEN has no need to * be configured. It will enable check sum for switch * If header mode is not set in DMA descriptor, * This register setting doesn't matter diff --git a/drivers/dma/loongson1-apb-dma.c b/drivers/dma/loongson1-apb-dma.c new file mode 100644 index 000000000000..255fe7eca212 --- /dev/null +++ b/drivers/dma/loongson1-apb-dma.c @@ -0,0 +1,660 @@ +// SPDX-License-Identifier: GPL-2.0-or-later +/* + * Driver for Loongson-1 APB DMA Controller + * + * Copyright (C) 2015-2024 Keguang Zhang <keguang.zhang@gmail.com> + */ + +#include <linux/dmapool.h> +#include <linux/dma-mapping.h> +#include <linux/init.h> +#include <linux/interrupt.h> +#include <linux/iopoll.h> +#include <linux/module.h> +#include <linux/of.h> +#include <linux/of_dma.h> +#include <linux/platform_device.h> +#include <linux/slab.h> + +#include "dmaengine.h" +#include "virt-dma.h" + +/* Loongson-1 DMA Control Register */ +#define LS1X_DMA_CTRL 0x0 + +/* DMA Control Register Bits */ +#define LS1X_DMA_STOP BIT(4) +#define LS1X_DMA_START BIT(3) +#define LS1X_DMA_ASK_VALID BIT(2) + +/* DMA Next Field Bits */ +#define LS1X_DMA_NEXT_VALID BIT(0) + +/* DMA Command Field Bits */ +#define LS1X_DMA_RAM2DEV BIT(12) +#define LS1X_DMA_INT BIT(1) +#define LS1X_DMA_INT_MASK BIT(0) + +#define LS1X_DMA_LLI_ALIGNMENT 64 +#define LS1X_DMA_LLI_ADDR_MASK GENMASK(31, __ffs(LS1X_DMA_LLI_ALIGNMENT)) +#define LS1X_DMA_MAX_CHANNELS 3 + +enum ls1x_dmadesc_offsets { + LS1X_DMADESC_NEXT = 0, + LS1X_DMADESC_SADDR, + LS1X_DMADESC_DADDR, + LS1X_DMADESC_LENGTH, + LS1X_DMADESC_STRIDE, + LS1X_DMADESC_CYCLES, + LS1X_DMADESC_CMD, + LS1X_DMADESC_SIZE +}; + +struct ls1x_dma_lli { + unsigned int hw[LS1X_DMADESC_SIZE]; + dma_addr_t phys; + struct list_head node; +} __aligned(LS1X_DMA_LLI_ALIGNMENT); + +struct ls1x_dma_desc { + struct virt_dma_desc vd; + struct list_head lli_list; +}; + +struct ls1x_dma_chan { + struct virt_dma_chan vc; + struct dma_pool *lli_pool; + phys_addr_t src_addr; + phys_addr_t dst_addr; + enum dma_slave_buswidth src_addr_width; + enum dma_slave_buswidth dst_addr_width; + unsigned int bus_width; + void __iomem *reg_base; + int irq; + bool is_cyclic; + struct ls1x_dma_lli *curr_lli; +}; + +struct ls1x_dma { + struct dma_device ddev; + unsigned int nr_chans; + struct ls1x_dma_chan chan[]; +}; + +static irqreturn_t ls1x_dma_irq_handler(int irq, void *data); + +#define to_ls1x_dma_chan(dchan) \ + container_of(dchan, struct ls1x_dma_chan, vc.chan) + +#define to_ls1x_dma_desc(d) \ + container_of(d, struct ls1x_dma_desc, vd) + +static inline struct device *chan2dev(struct dma_chan *chan) +{ + return &chan->dev->device; +} + +static inline int ls1x_dma_query(struct ls1x_dma_chan *chan, + dma_addr_t *lli_phys) +{ + struct dma_chan *dchan = &chan->vc.chan; + int val, ret; + + val = *lli_phys & LS1X_DMA_LLI_ADDR_MASK; + val |= LS1X_DMA_ASK_VALID; + val |= dchan->chan_id; + writel(val, chan->reg_base + LS1X_DMA_CTRL); + ret = readl_poll_timeout_atomic(chan->reg_base + LS1X_DMA_CTRL, val, + !(val & LS1X_DMA_ASK_VALID), 0, 3000); + if (ret) + dev_err(chan2dev(dchan), "failed to query DMA\n"); + + return ret; +} + +static inline int ls1x_dma_start(struct ls1x_dma_chan *chan, + dma_addr_t *lli_phys) +{ + struct dma_chan *dchan = &chan->vc.chan; + struct device *dev = chan2dev(dchan); + int val, ret; + + val = *lli_phys & LS1X_DMA_LLI_ADDR_MASK; + val |= LS1X_DMA_START; + val |= dchan->chan_id; + writel(val, chan->reg_base + LS1X_DMA_CTRL); + ret = readl_poll_timeout(chan->reg_base + LS1X_DMA_CTRL, val, + !(val & LS1X_DMA_START), 0, 1000); + if (!ret) + dev_dbg(dev, "start DMA with lli_phys=%pad\n", lli_phys); + else + dev_err(dev, "failed to start DMA\n"); + + return ret; +} + +static inline void ls1x_dma_stop(struct ls1x_dma_chan *chan) +{ + int val = readl(chan->reg_base + LS1X_DMA_CTRL); + + writel(val | LS1X_DMA_STOP, chan->reg_base + LS1X_DMA_CTRL); +} + +static void ls1x_dma_free_chan_resources(struct dma_chan *dchan) +{ + struct ls1x_dma_chan *chan = to_ls1x_dma_chan(dchan); + struct device *dev = chan2dev(dchan); + + dma_free_coherent(dev, sizeof(struct ls1x_dma_lli), + chan->curr_lli, chan->curr_lli->phys); + dma_pool_destroy(chan->lli_pool); + chan->lli_pool = NULL; + devm_free_irq(dev, chan->irq, chan); + vchan_free_chan_resources(&chan->vc); +} + +static int ls1x_dma_alloc_chan_resources(struct dma_chan *dchan) +{ + struct ls1x_dma_chan *chan = to_ls1x_dma_chan(dchan); + struct device *dev = chan2dev(dchan); + dma_addr_t phys; + int ret; + + ret = devm_request_irq(dev, chan->irq, ls1x_dma_irq_handler, + IRQF_SHARED, dma_chan_name(dchan), chan); + if (ret) { + dev_err(dev, "failed to request IRQ %d\n", chan->irq); + return ret; + } + + chan->lli_pool = dma_pool_create(dma_chan_name(dchan), dev, + sizeof(struct ls1x_dma_lli), + __alignof__(struct ls1x_dma_lli), 0); + if (!chan->lli_pool) + return -ENOMEM; + + /* allocate memory for querying the current lli */ + dma_set_coherent_mask(dev, DMA_BIT_MASK(32)); + chan->curr_lli = dma_alloc_coherent(dev, sizeof(struct ls1x_dma_lli), + &phys, GFP_KERNEL); + if (!chan->curr_lli) { + dma_pool_destroy(chan->lli_pool); + return -ENOMEM; + } + chan->curr_lli->phys = phys; + + return 0; +} + +static void ls1x_dma_free_desc(struct virt_dma_desc *vd) +{ + struct ls1x_dma_desc *desc = to_ls1x_dma_desc(vd); + struct ls1x_dma_chan *chan = to_ls1x_dma_chan(vd->tx.chan); + struct ls1x_dma_lli *lli, *_lli; + + list_for_each_entry_safe(lli, _lli, &desc->lli_list, node) { + list_del(&lli->node); + dma_pool_free(chan->lli_pool, lli, lli->phys); + } + + kfree(desc); +} + +static struct ls1x_dma_desc *ls1x_dma_alloc_desc(void) +{ + struct ls1x_dma_desc *desc; + + desc = kzalloc(sizeof(*desc), GFP_NOWAIT); + if (!desc) + return NULL; + + INIT_LIST_HEAD(&desc->lli_list); + + return desc; +} + +static int ls1x_dma_prep_lli(struct dma_chan *dchan, struct ls1x_dma_desc *desc, + struct scatterlist *sgl, unsigned int sg_len, + enum dma_transfer_direction dir, bool is_cyclic) +{ + struct ls1x_dma_chan *chan = to_ls1x_dma_chan(dchan); + struct ls1x_dma_lli *lli, *prev = NULL, *first = NULL; + struct device *dev = chan2dev(dchan); + struct list_head *pos = NULL; + struct scatterlist *sg; + unsigned int dev_addr, cmd, i; + + switch (dir) { + case DMA_MEM_TO_DEV: + dev_addr = chan->dst_addr; + chan->bus_width = chan->dst_addr_width; + cmd = LS1X_DMA_RAM2DEV | LS1X_DMA_INT; + break; + case DMA_DEV_TO_MEM: + dev_addr = chan->src_addr; + chan->bus_width = chan->src_addr_width; + cmd = LS1X_DMA_INT; + break; + default: + dev_err(dev, "unsupported DMA direction: %s\n", + dmaengine_get_direction_text(dir)); + return -EINVAL; + } + + for_each_sg(sgl, sg, sg_len, i) { + dma_addr_t buf_addr = sg_dma_address(sg); + size_t buf_len = sg_dma_len(sg); + dma_addr_t phys; + + if (!is_dma_copy_aligned(dchan->device, buf_addr, 0, buf_len)) { + dev_err(dev, "buffer is not aligned\n"); + return -EINVAL; + } + + /* allocate HW descriptors */ + lli = dma_pool_zalloc(chan->lli_pool, GFP_NOWAIT, &phys); + if (!lli) { + dev_err(dev, "failed to alloc lli %u\n", i); + return -ENOMEM; + } + + /* setup HW descriptors */ + lli->phys = phys; + lli->hw[LS1X_DMADESC_SADDR] = buf_addr; + lli->hw[LS1X_DMADESC_DADDR] = dev_addr; + lli->hw[LS1X_DMADESC_LENGTH] = buf_len / chan->bus_width; + lli->hw[LS1X_DMADESC_STRIDE] = 0; + lli->hw[LS1X_DMADESC_CYCLES] = 1; + lli->hw[LS1X_DMADESC_CMD] = cmd; + + if (prev) + prev->hw[LS1X_DMADESC_NEXT] = + lli->phys | LS1X_DMA_NEXT_VALID; + prev = lli; + + if (!first) + first = lli; + + list_add_tail(&lli->node, &desc->lli_list); + } + + if (is_cyclic) { + lli->hw[LS1X_DMADESC_NEXT] = first->phys | LS1X_DMA_NEXT_VALID; + chan->is_cyclic = is_cyclic; + } + + list_for_each(pos, &desc->lli_list) { + lli = list_entry(pos, struct ls1x_dma_lli, node); + print_hex_dump_debug("LLI: ", DUMP_PREFIX_OFFSET, 16, 4, + lli, sizeof(*lli), false); + } + + return 0; +} + +static struct dma_async_tx_descriptor * +ls1x_dma_prep_slave_sg(struct dma_chan *dchan, struct scatterlist *sgl, + unsigned int sg_len, enum dma_transfer_direction dir, + unsigned long flags, void *context) +{ + struct ls1x_dma_desc *desc; + + dev_dbg(chan2dev(dchan), "sg_len=%u flags=0x%lx dir=%s\n", + sg_len, flags, dmaengine_get_direction_text(dir)); + + desc = ls1x_dma_alloc_desc(); + if (!desc) + return NULL; + + if (ls1x_dma_prep_lli(dchan, desc, sgl, sg_len, dir, false)) { + ls1x_dma_free_desc(&desc->vd); + return NULL; + } + + return vchan_tx_prep(to_virt_chan(dchan), &desc->vd, flags); +} + +static struct dma_async_tx_descriptor * +ls1x_dma_prep_dma_cyclic(struct dma_chan *dchan, dma_addr_t buf_addr, + size_t buf_len, size_t period_len, + enum dma_transfer_direction dir, unsigned long flags) +{ + struct ls1x_dma_desc *desc; + struct scatterlist *sgl; + unsigned int sg_len; + unsigned int i; + int ret; + + dev_dbg(chan2dev(dchan), + "buf_len=%zu period_len=%zu flags=0x%lx dir=%s\n", + buf_len, period_len, flags, dmaengine_get_direction_text(dir)); + + desc = ls1x_dma_alloc_desc(); + if (!desc) + return NULL; + + /* allocate the scatterlist */ + sg_len = buf_len / period_len; + sgl = kmalloc_array(sg_len, sizeof(*sgl), GFP_NOWAIT); + if (!sgl) + return NULL; + + sg_init_table(sgl, sg_len); + for (i = 0; i < sg_len; ++i) { + sg_set_page(&sgl[i], pfn_to_page(PFN_DOWN(buf_addr)), + period_len, offset_in_page(buf_addr)); + sg_dma_address(&sgl[i]) = buf_addr; + sg_dma_len(&sgl[i]) = period_len; + buf_addr += period_len; + } + + ret = ls1x_dma_prep_lli(dchan, desc, sgl, sg_len, dir, true); + kfree(sgl); + if (ret) { + ls1x_dma_free_desc(&desc->vd); + return NULL; + } + + return vchan_tx_prep(to_virt_chan(dchan), &desc->vd, flags); +} + +static int ls1x_dma_slave_config(struct dma_chan *dchan, + struct dma_slave_config *config) +{ + struct ls1x_dma_chan *chan = to_ls1x_dma_chan(dchan); + + chan->src_addr = config->src_addr; + chan->src_addr_width = config->src_addr_width; + chan->dst_addr = config->dst_addr; + chan->dst_addr_width = config->dst_addr_width; + + return 0; +} + +static int ls1x_dma_pause(struct dma_chan *dchan) +{ + struct ls1x_dma_chan *chan = to_ls1x_dma_chan(dchan); + int ret; + + guard(spinlock_irqsave)(&chan->vc.lock); + /* save the current lli */ + ret = ls1x_dma_query(chan, &chan->curr_lli->phys); + if (!ret) + ls1x_dma_stop(chan); + + return ret; +} + +static int ls1x_dma_resume(struct dma_chan *dchan) +{ + struct ls1x_dma_chan *chan = to_ls1x_dma_chan(dchan); + + guard(spinlock_irqsave)(&chan->vc.lock); + + return ls1x_dma_start(chan, &chan->curr_lli->phys); +} + +static int ls1x_dma_terminate_all(struct dma_chan *dchan) +{ + struct ls1x_dma_chan *chan = to_ls1x_dma_chan(dchan); + struct virt_dma_desc *vd; + LIST_HEAD(head); + + ls1x_dma_stop(chan); + + scoped_guard(spinlock_irqsave, &chan->vc.lock) { + vd = vchan_next_desc(&chan->vc); + if (vd) + vchan_terminate_vdesc(vd); + + vchan_get_all_descriptors(&chan->vc, &head); + } + + vchan_dma_desc_free_list(&chan->vc, &head); + + return 0; +} + +static void ls1x_dma_synchronize(struct dma_chan *dchan) +{ + vchan_synchronize(to_virt_chan(dchan)); +} + +static enum dma_status ls1x_dma_tx_status(struct dma_chan *dchan, + dma_cookie_t cookie, + struct dma_tx_state *state) +{ + struct ls1x_dma_chan *chan = to_ls1x_dma_chan(dchan); + struct virt_dma_desc *vd; + enum dma_status status; + size_t bytes = 0; + + status = dma_cookie_status(dchan, cookie, state); + if (status == DMA_COMPLETE) + return status; + + scoped_guard(spinlock_irqsave, &chan->vc.lock) { + vd = vchan_find_desc(&chan->vc, cookie); + if (vd) { + struct ls1x_dma_desc *desc = to_ls1x_dma_desc(vd); + struct ls1x_dma_lli *lli; + dma_addr_t next_phys; + + /* get the current lli */ + if (ls1x_dma_query(chan, &chan->curr_lli->phys)) + return status; + + /* locate the current lli */ + next_phys = chan->curr_lli->hw[LS1X_DMADESC_NEXT]; + list_for_each_entry(lli, &desc->lli_list, node) + if (lli->hw[LS1X_DMADESC_NEXT] == next_phys) + break; + + dev_dbg(chan2dev(dchan), "current lli_phys=%pad", + &lli->phys); + + /* count the residues */ + list_for_each_entry_from(lli, &desc->lli_list, node) + bytes += lli->hw[LS1X_DMADESC_LENGTH] * + chan->bus_width; + } + } + + dma_set_residue(state, bytes); + + return status; +} + +static void ls1x_dma_issue_pending(struct dma_chan *dchan) +{ + struct ls1x_dma_chan *chan = to_ls1x_dma_chan(dchan); + + guard(spinlock_irqsave)(&chan->vc.lock); + + if (vchan_issue_pending(&chan->vc)) { + struct virt_dma_desc *vd = vchan_next_desc(&chan->vc); + + if (vd) { + struct ls1x_dma_desc *desc = to_ls1x_dma_desc(vd); + struct ls1x_dma_lli *lli; + + lli = list_first_entry(&desc->lli_list, + struct ls1x_dma_lli, node); + ls1x_dma_start(chan, &lli->phys); + } + } +} + +static irqreturn_t ls1x_dma_irq_handler(int irq, void *data) +{ + struct ls1x_dma_chan *chan = data; + struct dma_chan *dchan = &chan->vc.chan; + struct device *dev = chan2dev(dchan); + struct virt_dma_desc *vd; + + scoped_guard(spinlock, &chan->vc.lock) { + vd = vchan_next_desc(&chan->vc); + if (!vd) { + dev_warn(dev, + "IRQ %d with no active desc on channel %d\n", + irq, dchan->chan_id); + return IRQ_NONE; + } + + if (chan->is_cyclic) { + vchan_cyclic_callback(vd); + } else { + list_del(&vd->node); + vchan_cookie_complete(vd); + } + } + + dev_dbg(dev, "DMA IRQ %d on channel %d\n", irq, dchan->chan_id); + + return IRQ_HANDLED; +} + +static int ls1x_dma_chan_probe(struct platform_device *pdev, + struct ls1x_dma *dma) +{ + void __iomem *reg_base; + int id; + + reg_base = devm_platform_ioremap_resource(pdev, 0); + if (IS_ERR(reg_base)) + return PTR_ERR(reg_base); + + for (id = 0; id < dma->nr_chans; id++) { + struct ls1x_dma_chan *chan = &dma->chan[id]; + char pdev_irqname[16]; + + snprintf(pdev_irqname, sizeof(pdev_irqname), "ch%d", id); + chan->irq = platform_get_irq_byname(pdev, pdev_irqname); + if (chan->irq < 0) + return dev_err_probe(&pdev->dev, chan->irq, + "failed to get IRQ for ch%d\n", + id); + + chan->reg_base = reg_base; + chan->vc.desc_free = ls1x_dma_free_desc; + vchan_init(&chan->vc, &dma->ddev); + } + + return 0; +} + +static void ls1x_dma_chan_remove(struct ls1x_dma *dma) +{ + int id; + + for (id = 0; id < dma->nr_chans; id++) { + struct ls1x_dma_chan *chan = &dma->chan[id]; + + if (chan->vc.chan.device == &dma->ddev) { + list_del(&chan->vc.chan.device_node); + tasklet_kill(&chan->vc.task); + } + } +} + +static int ls1x_dma_probe(struct platform_device *pdev) +{ + struct device *dev = &pdev->dev; + struct dma_device *ddev; + struct ls1x_dma *dma; + int ret; + + ret = platform_irq_count(pdev); + if (ret <= 0 || ret > LS1X_DMA_MAX_CHANNELS) + return dev_err_probe(dev, -EINVAL, + "Invalid number of IRQ channels: %d\n", + ret); + + dma = devm_kzalloc(dev, struct_size(dma, chan, ret), GFP_KERNEL); + if (!dma) + return -ENOMEM; + dma->nr_chans = ret; + + /* initialize DMA device */ + ddev = &dma->ddev; + ddev->dev = dev; + ddev->copy_align = DMAENGINE_ALIGN_4_BYTES; + ddev->src_addr_widths = BIT(DMA_SLAVE_BUSWIDTH_1_BYTE) | + BIT(DMA_SLAVE_BUSWIDTH_2_BYTES) | + BIT(DMA_SLAVE_BUSWIDTH_4_BYTES); + ddev->dst_addr_widths = BIT(DMA_SLAVE_BUSWIDTH_1_BYTE) | + BIT(DMA_SLAVE_BUSWIDTH_2_BYTES) | + BIT(DMA_SLAVE_BUSWIDTH_4_BYTES); + ddev->directions = BIT(DMA_DEV_TO_MEM) | BIT(DMA_MEM_TO_DEV); + ddev->residue_granularity = DMA_RESIDUE_GRANULARITY_SEGMENT; + ddev->device_alloc_chan_resources = ls1x_dma_alloc_chan_resources; + ddev->device_free_chan_resources = ls1x_dma_free_chan_resources; + ddev->device_prep_slave_sg = ls1x_dma_prep_slave_sg; + ddev->device_prep_dma_cyclic = ls1x_dma_prep_dma_cyclic; + ddev->device_config = ls1x_dma_slave_config; + ddev->device_pause = ls1x_dma_pause; + ddev->device_resume = ls1x_dma_resume; + ddev->device_terminate_all = ls1x_dma_terminate_all; + ddev->device_synchronize = ls1x_dma_synchronize; + ddev->device_tx_status = ls1x_dma_tx_status; + ddev->device_issue_pending = ls1x_dma_issue_pending; + dma_cap_set(DMA_SLAVE, ddev->cap_mask); + INIT_LIST_HEAD(&ddev->channels); + + /* initialize DMA channels */ + ret = ls1x_dma_chan_probe(pdev, dma); + if (ret) + goto err; + + ret = dmaenginem_async_device_register(ddev); + if (ret) { + dev_err(dev, "failed to register DMA device\n"); + goto err; + } + + ret = of_dma_controller_register(dev->of_node, of_dma_xlate_by_chan_id, + ddev); + if (ret) { + dev_err(dev, "failed to register DMA controller\n"); + goto err; + } + + platform_set_drvdata(pdev, dma); + dev_info(dev, "Loongson1 DMA driver registered\n"); + + return 0; + +err: + ls1x_dma_chan_remove(dma); + + return ret; +} + +static void ls1x_dma_remove(struct platform_device *pdev) +{ + struct ls1x_dma *dma = platform_get_drvdata(pdev); + + of_dma_controller_free(pdev->dev.of_node); + ls1x_dma_chan_remove(dma); +} + +static const struct of_device_id ls1x_dma_match[] = { + { .compatible = "loongson,ls1b-apbdma" }, + { /* sentinel */ } +}; +MODULE_DEVICE_TABLE(of, ls1x_dma_match); + +static struct platform_driver ls1x_dma_driver = { + .probe = ls1x_dma_probe, + .remove = ls1x_dma_remove, + .driver = { + .name = KBUILD_MODNAME, + .of_match_table = ls1x_dma_match, + }, +}; + +module_platform_driver(ls1x_dma_driver); + +MODULE_AUTHOR("Keguang Zhang <keguang.zhang@gmail.com>"); +MODULE_DESCRIPTION("Loongson-1 APB DMA Controller driver"); +MODULE_LICENSE("GPL"); diff --git a/drivers/dma/lpc32xx-dmamux.c b/drivers/dma/lpc32xx-dmamux.c new file mode 100644 index 000000000000..351d7e23e615 --- /dev/null +++ b/drivers/dma/lpc32xx-dmamux.c @@ -0,0 +1,195 @@ +// SPDX-License-Identifier: GPL-2.0-only +// +// Copyright 2024 Timesys Corporation <piotr.wojtaszczyk@timesys.com> +// +// Based on TI DMA Crossbar driver by: +// Copyright (C) 2015 Texas Instruments Incorporated - http://www.ti.com +// Author: Peter Ujfalusi <peter.ujfalusi@ti.com> + +#include <linux/err.h> +#include <linux/init.h> +#include <linux/mfd/syscon.h> +#include <linux/of.h> +#include <linux/of_dma.h> +#include <linux/of_platform.h> +#include <linux/platform_device.h> +#include <linux/regmap.h> +#include <linux/spinlock.h> + +#define LPC32XX_SSP_CLK_CTRL 0x78 +#define LPC32XX_I2S_CLK_CTRL 0x7c + +struct lpc32xx_dmamux { + int signal; + char *name_sel0; + char *name_sel1; + int muxval; + int muxreg; + int bit; + bool busy; +}; + +struct lpc32xx_dmamux_data { + struct dma_router dmarouter; + struct regmap *reg; + spinlock_t lock; /* protects busy status flag */ +}; + +/* From LPC32x0 User manual "3.2.1 DMA request signals" */ +static struct lpc32xx_dmamux lpc32xx_muxes[] = { + { + .signal = 3, + .name_sel0 = "spi2-rx-tx", + .name_sel1 = "ssp1-rx", + .muxreg = LPC32XX_SSP_CLK_CTRL, + .bit = 5, + }, + { + .signal = 10, + .name_sel0 = "uart7-rx", + .name_sel1 = "i2s1-dma1", + .muxreg = LPC32XX_I2S_CLK_CTRL, + .bit = 4, + }, + { + .signal = 11, + .name_sel0 = "spi1-rx-tx", + .name_sel1 = "ssp1-tx", + .muxreg = LPC32XX_SSP_CLK_CTRL, + .bit = 4, + }, + { + .signal = 14, + .name_sel0 = "none", + .name_sel1 = "ssp0-rx", + .muxreg = LPC32XX_SSP_CLK_CTRL, + .bit = 3, + }, + { + .signal = 15, + .name_sel0 = "none", + .name_sel1 = "ssp0-tx", + .muxreg = LPC32XX_SSP_CLK_CTRL, + .bit = 2, + }, +}; + +static void lpc32xx_dmamux_release(struct device *dev, void *route_data) +{ + struct lpc32xx_dmamux_data *dmamux = dev_get_drvdata(dev); + struct lpc32xx_dmamux *mux = route_data; + + dev_dbg(dev, "releasing dma request signal %d routed to %s\n", + mux->signal, mux->muxval ? mux->name_sel1 : mux->name_sel1); + + guard(spinlock)(&dmamux->lock); + + mux->busy = false; +} + +static void *lpc32xx_dmamux_reserve(struct of_phandle_args *dma_spec, + struct of_dma *ofdma) +{ + struct platform_device *pdev = of_find_device_by_node(ofdma->of_node); + struct device *dev = &pdev->dev; + struct lpc32xx_dmamux_data *dmamux = platform_get_drvdata(pdev); + unsigned long flags; + struct lpc32xx_dmamux *mux = NULL; + int i; + + if (dma_spec->args_count != 3) { + dev_err(&pdev->dev, "invalid number of dma mux args\n"); + return ERR_PTR(-EINVAL); + } + + for (i = 0; i < ARRAY_SIZE(lpc32xx_muxes); i++) { + if (lpc32xx_muxes[i].signal == dma_spec->args[0]) { + mux = &lpc32xx_muxes[i]; + break; + } + } + if (!mux) { + dev_err(&pdev->dev, "invalid mux request number: %d\n", + dma_spec->args[0]); + return ERR_PTR(-EINVAL); + } + + if (dma_spec->args[2] > 1) { + dev_err(&pdev->dev, "invalid dma mux value: %d\n", + dma_spec->args[1]); + return ERR_PTR(-EINVAL); + } + + /* The of_node_put() will be done in the core for the node */ + dma_spec->np = of_parse_phandle(ofdma->of_node, "dma-masters", 0); + if (!dma_spec->np) { + dev_err(&pdev->dev, "can't get dma master\n"); + return ERR_PTR(-EINVAL); + } + + spin_lock_irqsave(&dmamux->lock, flags); + if (mux->busy) { + spin_unlock_irqrestore(&dmamux->lock, flags); + dev_err(dev, "dma request signal %d busy, routed to %s\n", + mux->signal, mux->muxval ? mux->name_sel1 : mux->name_sel1); + of_node_put(dma_spec->np); + return ERR_PTR(-EBUSY); + } + + mux->busy = true; + mux->muxval = dma_spec->args[2] ? BIT(mux->bit) : 0; + + regmap_update_bits(dmamux->reg, mux->muxreg, BIT(mux->bit), mux->muxval); + spin_unlock_irqrestore(&dmamux->lock, flags); + + dma_spec->args[2] = 0; + dma_spec->args_count = 2; + + dev_dbg(dev, "dma request signal %d routed to %s\n", + mux->signal, mux->muxval ? mux->name_sel1 : mux->name_sel1); + + return mux; +} + +static int lpc32xx_dmamux_probe(struct platform_device *pdev) +{ + struct device_node *np = pdev->dev.of_node; + struct lpc32xx_dmamux_data *dmamux; + + dmamux = devm_kzalloc(&pdev->dev, sizeof(*dmamux), GFP_KERNEL); + if (!dmamux) + return -ENOMEM; + + dmamux->reg = syscon_node_to_regmap(np->parent); + if (IS_ERR(dmamux->reg)) { + dev_err(&pdev->dev, "syscon lookup failed\n"); + return PTR_ERR(dmamux->reg); + } + + spin_lock_init(&dmamux->lock); + platform_set_drvdata(pdev, dmamux); + dmamux->dmarouter.dev = &pdev->dev; + dmamux->dmarouter.route_free = lpc32xx_dmamux_release; + + return of_dma_router_register(np, lpc32xx_dmamux_reserve, + &dmamux->dmarouter); +} + +static const struct of_device_id lpc32xx_dmamux_match[] = { + { .compatible = "nxp,lpc3220-dmamux" }, + {}, +}; + +static struct platform_driver lpc32xx_dmamux_driver = { + .probe = lpc32xx_dmamux_probe, + .driver = { + .name = "lpc32xx-dmamux", + .of_match_table = lpc32xx_dmamux_match, + }, +}; + +static int __init lpc32xx_dmamux_init(void) +{ + return platform_driver_register(&lpc32xx_dmamux_driver); +} +arch_initcall(lpc32xx_dmamux_init); diff --git a/drivers/dma/ls2x-apb-dma.c b/drivers/dma/ls2x-apb-dma.c index a49913f3ed3f..9652e8666722 100644 --- a/drivers/dma/ls2x-apb-dma.c +++ b/drivers/dma/ls2x-apb-dma.c @@ -33,11 +33,11 @@ #define LDMA_STOP BIT(4) /* DMA stop operation */ #define LDMA_CONFIG_MASK GENMASK(4, 0) /* DMA controller config bits mask */ -/* Bitfields in ndesc_addr field of HW decriptor */ +/* Bitfields in ndesc_addr field of HW descriptor */ #define LDMA_DESC_EN BIT(0) /*1: The next descriptor is valid */ #define LDMA_DESC_ADDR_LOW GENMASK(31, 1) -/* Bitfields in cmd field of HW decriptor */ +/* Bitfields in cmd field of HW descriptor */ #define LDMA_INT BIT(1) /* Enable DMA interrupts */ #define LDMA_DATA_DIRECTION BIT(12) /* 1: write to device, 0: read from device */ diff --git a/drivers/dma/mediatek/mtk-cqdma.c b/drivers/dma/mediatek/mtk-cqdma.c index 529100c5b9f5..b69eabf12a24 100644 --- a/drivers/dma/mediatek/mtk-cqdma.c +++ b/drivers/dma/mediatek/mtk-cqdma.c @@ -518,7 +518,7 @@ mtk_cqdma_prep_dma_memcpy(struct dma_chan *c, dma_addr_t dest, /* setup dma channel */ cvd[i]->ch = c; - /* setup sourece, destination, and length */ + /* setup source, destination, and length */ tlen = (len > MTK_CQDMA_MAX_LEN) ? MTK_CQDMA_MAX_LEN : len; cvd[i]->len = tlen; cvd[i]->src = src; @@ -617,7 +617,7 @@ static int mtk_cqdma_alloc_chan_resources(struct dma_chan *c) u32 i, min_refcnt = U32_MAX, refcnt; unsigned long flags; - /* allocate PC with the minimun refcount */ + /* allocate PC with the minimum refcount */ for (i = 0; i < cqdma->dma_channels; ++i) { refcnt = refcount_read(&cqdma->pc[i]->refcnt); if (refcnt < min_refcnt) { diff --git a/drivers/dma/mediatek/mtk-hsdma.c b/drivers/dma/mediatek/mtk-hsdma.c index 36ff11e909ea..58c7961ab9ad 100644 --- a/drivers/dma/mediatek/mtk-hsdma.c +++ b/drivers/dma/mediatek/mtk-hsdma.c @@ -226,7 +226,7 @@ struct mtk_hsdma_soc { * @pc_refcnt: Track how many VCs are using the PC * @lock: Lock protect agaisting multiple VCs access PC * @soc: The pointer to area holding differences among - * vaious platform + * various platform */ struct mtk_hsdma_device { struct dma_device ddev; diff --git a/drivers/dma/mv_xor.c b/drivers/dma/mv_xor.c index bcd3b623ac6c..43efce77bb57 100644 --- a/drivers/dma/mv_xor.c +++ b/drivers/dma/mv_xor.c @@ -414,7 +414,7 @@ mv_xor_tx_submit(struct dma_async_tx_descriptor *tx) if (!mv_chan_is_busy(mv_chan)) { u32 current_desc = mv_chan_get_current_desc(mv_chan); /* - * and the curren desc is the end of the chain before + * and the current desc is the end of the chain before * the append, then we need to start the channel */ if (current_desc == old_chain_tail->async_tx.phys) @@ -1074,7 +1074,7 @@ mv_xor_channel_add(struct mv_xor_device *xordev, if (!mv_chan->dma_desc_pool_virt) return ERR_PTR(-ENOMEM); - /* discover transaction capabilites from the platform data */ + /* discover transaction capabilities from the platform data */ dma_dev->cap_mask = cap_mask; INIT_LIST_HEAD(&dma_dev->channels); diff --git a/drivers/dma/mv_xor.h b/drivers/dma/mv_xor.h index d86086b05b0e..c87cefd38a07 100644 --- a/drivers/dma/mv_xor.h +++ b/drivers/dma/mv_xor.h @@ -99,7 +99,7 @@ struct mv_xor_device { * @common: common dmaengine channel object members * @slots_allocated: records the actual size of the descriptor slot pool * @irq_tasklet: bottom half where mv_xor_slot_cleanup runs - * @op_in_desc: new mode of driver, each op is writen to descriptor. + * @op_in_desc: new mode of driver, each op is written to descriptor. */ struct mv_xor_chan { int pending; diff --git a/drivers/dma/mv_xor_v2.c b/drivers/dma/mv_xor_v2.c index 97ebc791a30b..c8c67f4d982c 100644 --- a/drivers/dma/mv_xor_v2.c +++ b/drivers/dma/mv_xor_v2.c @@ -175,7 +175,7 @@ struct mv_xor_v2_device { * struct mv_xor_v2_sw_desc - implements a xor SW descriptor * @idx: descriptor index * @async_tx: support for the async_tx api - * @hw_desc: assosiated HW descriptor + * @hw_desc: associated HW descriptor * @free_list: node of the free SW descriprots list */ struct mv_xor_v2_sw_desc { diff --git a/drivers/dma/nbpfaxi.c b/drivers/dma/nbpfaxi.c index c08916339aa7..3b011a91d48e 100644 --- a/drivers/dma/nbpfaxi.c +++ b/drivers/dma/nbpfaxi.c @@ -897,7 +897,7 @@ static int nbpf_config(struct dma_chan *dchan, /* * We could check config->slave_id to match chan->terminal here, * but with DT they would be coming from the same source, so - * such a check would be superflous + * such a check would be superfluous */ chan->slave_dst_addr = config->dst_addr; diff --git a/drivers/dma/of-dma.c b/drivers/dma/of-dma.c index e588fff9f21d..423442e55d36 100644 --- a/drivers/dma/of-dma.c +++ b/drivers/dma/of-dma.c @@ -26,7 +26,7 @@ static DEFINE_MUTEX(of_dma_lock); * * Finds a DMA controller with matching device node and number for dma cells * in a list of registered DMA controllers. If a match is found a valid pointer - * to the DMA data stored is retuned. A NULL pointer is returned if no match is + * to the DMA data stored is returned. A NULL pointer is returned if no match is * found. */ static struct of_dma *of_dma_find_controller(const struct of_phandle_args *dma_spec) @@ -342,7 +342,7 @@ EXPORT_SYMBOL_GPL(of_dma_simple_xlate); * * This function can be used as the of xlate callback for DMA driver which wants * to match the channel based on the channel id. When using this xlate function - * the #dma-cells propety of the DMA controller dt node needs to be set to 1. + * the #dma-cells property of the DMA controller dt node needs to be set to 1. * The data parameter of of_dma_controller_register must be a pointer to the * dma_device struct the function should match upon. * diff --git a/drivers/dma/owl-dma.c b/drivers/dma/owl-dma.c index e001f4f7aa64..aa436f9e3571 100644 --- a/drivers/dma/owl-dma.c +++ b/drivers/dma/owl-dma.c @@ -1156,7 +1156,7 @@ static int owl_dma_probe(struct platform_device *pdev) } /* - * Eventhough the DMA controller is capable of generating 4 + * Even though the DMA controller is capable of generating 4 * IRQ's for DMA priority feature, we only use 1 IRQ for * simplification. */ diff --git a/drivers/dma/pl330.c b/drivers/dma/pl330.c index 60c4de8dac1d..82a9fe88ad54 100644 --- a/drivers/dma/pl330.c +++ b/drivers/dma/pl330.c @@ -3163,10 +3163,7 @@ pl330_probe(struct amba_device *adev, const struct amba_id *id) * This is the limit for transfers with a buswidth of 1, larger * buswidths will have larger limits. */ - ret = dma_set_max_seg_size(&adev->dev, 1900800); - if (ret) - dev_err(&adev->dev, "unable to set the seg size\n"); - + dma_set_max_seg_size(&adev->dev, 1900800); init_pl330_debugfs(pl330); dev_info(&adev->dev, diff --git a/drivers/dma/ppc4xx/adma.c b/drivers/dma/ppc4xx/adma.c index bbb60a970dab..7b78759ac734 100644 --- a/drivers/dma/ppc4xx/adma.c +++ b/drivers/dma/ppc4xx/adma.c @@ -9,7 +9,7 @@ */ /* - * This driver supports the asynchrounous DMA copy and RAID engines available + * This driver supports the asynchronous DMA copy and RAID engines available * on the AMCC PPC440SPe Processors. * Based on the Intel Xscale(R) family of I/O Processors (IOP 32x, 33x, 134x) * ADMA driver written by D.Williams. diff --git a/drivers/dma/ppc4xx/dma.h b/drivers/dma/ppc4xx/dma.h index 1ff4be23db0f..b5725481bfa6 100644 --- a/drivers/dma/ppc4xx/dma.h +++ b/drivers/dma/ppc4xx/dma.h @@ -14,7 +14,7 @@ /* Number of elements in the array with statical CDBs */ #define MAX_STAT_DMA_CDBS 16 -/* Number of DMA engines available on the contoller */ +/* Number of DMA engines available on the controller */ #define DMA_ENGINES_NUM 2 /* Maximum h/w supported number of destinations */ diff --git a/drivers/dma/ptdma/ptdma.h b/drivers/dma/ptdma/ptdma.h index 21b4bf895200..39bc37268235 100644 --- a/drivers/dma/ptdma/ptdma.h +++ b/drivers/dma/ptdma/ptdma.h @@ -192,7 +192,7 @@ struct pt_cmd_queue { /* Queue dma pool */ struct dma_pool *dma_pool; - /* Queue base address (not neccessarily aligned)*/ + /* Queue base address (not necessarily aligned)*/ struct ptdma_desc *qbase; /* Aligned queue start address (per requirement) */ diff --git a/drivers/dma/qcom/bam_dma.c b/drivers/dma/qcom/bam_dma.c index 5e7d332731e0..d43a881e43b9 100644 --- a/drivers/dma/qcom/bam_dma.c +++ b/drivers/dma/qcom/bam_dma.c @@ -440,7 +440,7 @@ static void bam_reset(struct bam_device *bdev) val |= BAM_EN; writel_relaxed(val, bam_addr(bdev, 0, BAM_CTRL)); - /* set descriptor threshhold, start with 4 bytes */ + /* set descriptor threshold, start with 4 bytes */ writel_relaxed(DEFAULT_CNT_THRSHLD, bam_addr(bdev, 0, BAM_DESC_CNT_TRSHLD)); @@ -667,7 +667,7 @@ static struct dma_async_tx_descriptor *bam_prep_slave_sg(struct dma_chan *chan, for_each_sg(sgl, sg, sg_len, i) num_alloc += DIV_ROUND_UP(sg_dma_len(sg), BAM_FIFO_SIZE); - /* allocate enough room to accomodate the number of entries */ + /* allocate enough room to accommodate the number of entries */ async_desc = kzalloc(struct_size(async_desc, desc, num_alloc), GFP_NOWAIT); @@ -1325,11 +1325,7 @@ static int bam_dma_probe(struct platform_device *pdev) /* set max dma segment size */ bdev->common.dev = bdev->dev; - ret = dma_set_max_seg_size(bdev->common.dev, BAM_FIFO_SIZE); - if (ret) { - dev_err(bdev->dev, "cannot set maximum segment size\n"); - goto err_bam_channel_exit; - } + dma_set_max_seg_size(bdev->common.dev, BAM_FIFO_SIZE); platform_set_drvdata(pdev, bdev); diff --git a/drivers/dma/qcom/gpi.c b/drivers/dma/qcom/gpi.c index e6ebd688d746..52a7c8f2498f 100644 --- a/drivers/dma/qcom/gpi.c +++ b/drivers/dma/qcom/gpi.c @@ -1856,7 +1856,7 @@ static void gpi_issue_pending(struct dma_chan *chan) read_lock_irqsave(&gpii->pm_lock, pm_lock_flags); - /* move all submitted discriptors to issued list */ + /* move all submitted descriptors to issued list */ spin_lock_irqsave(&gchan->vc.lock, flags); if (vchan_issue_pending(&gchan->vc)) vd = list_last_entry(&gchan->vc.desc_issued, diff --git a/drivers/dma/qcom/qcom_adm.c b/drivers/dma/qcom/qcom_adm.c index 53f4273b657c..c1db398adc84 100644 --- a/drivers/dma/qcom/qcom_adm.c +++ b/drivers/dma/qcom/qcom_adm.c @@ -650,7 +650,7 @@ static enum dma_status adm_tx_status(struct dma_chan *chan, dma_cookie_t cookie, /* * residue is either the full length if it is in the issued list, or 0 * if it is in progress. We have no reliable way of determining - * anything inbetween + * anything in between */ dma_set_residue(txstate, residue); diff --git a/drivers/dma/sh/rcar-dmac.c b/drivers/dma/sh/rcar-dmac.c index 40482cb73d79..1094a2f82164 100644 --- a/drivers/dma/sh/rcar-dmac.c +++ b/drivers/dma/sh/rcar-dmac.c @@ -1868,9 +1868,7 @@ static int rcar_dmac_probe(struct platform_device *pdev) dmac->dev = &pdev->dev; platform_set_drvdata(pdev, dmac); - ret = dma_set_max_seg_size(dmac->dev, RCAR_DMATCR_MASK); - if (ret) - return ret; + dma_set_max_seg_size(dmac->dev, RCAR_DMATCR_MASK); ret = dma_set_mask_and_coherent(dmac->dev, DMA_BIT_MASK(40)); if (ret) diff --git a/drivers/dma/sh/shdmac.c b/drivers/dma/sh/shdmac.c index 7cc9eb2217e8..8ead0a1fd237 100644 --- a/drivers/dma/sh/shdmac.c +++ b/drivers/dma/sh/shdmac.c @@ -318,7 +318,7 @@ static void sh_dmae_setup_xfer(struct shdma_chan *schan, } /* - * Find a slave channel configuration from the contoller list by either a slave + * Find a slave channel configuration from the controller list by either a slave * ID in the non-DT case, or by a MID/RID value in the DT case */ static const struct sh_dmae_slave_config *dmae_find_slave( diff --git a/drivers/dma/ste_dma40.c b/drivers/dma/ste_dma40.c index 2c489299148e..d52e1685aed5 100644 --- a/drivers/dma/ste_dma40.c +++ b/drivers/dma/ste_dma40.c @@ -3632,11 +3632,7 @@ static int __init d40_probe(struct platform_device *pdev) if (ret) goto destroy_cache; - ret = dma_set_max_seg_size(base->dev, STEDMA40_MAX_SEG_SIZE); - if (ret) { - d40_err(dev, "Failed to set dma max seg size\n"); - goto destroy_cache; - } + dma_set_max_seg_size(base->dev, STEDMA40_MAX_SEG_SIZE); d40_hw_init(base); diff --git a/drivers/dma/ste_dma40.h b/drivers/dma/ste_dma40.h index c697bfe16a01..a90c786acc1f 100644 --- a/drivers/dma/ste_dma40.h +++ b/drivers/dma/ste_dma40.h @@ -4,7 +4,7 @@ #define STE_DMA40_H /* - * Maxium size for a single dma descriptor + * Maximum size for a single dma descriptor * Size is limited to 16 bits. * Size is in the units of addr-widths (1,2,4,8 bytes) * Larger transfers will be split up to multiple linked desc diff --git a/drivers/dma/ste_dma40_ll.h b/drivers/dma/ste_dma40_ll.h index c504e855eb02..2e30e9a94a1e 100644 --- a/drivers/dma/ste_dma40_ll.h +++ b/drivers/dma/ste_dma40_ll.h @@ -369,7 +369,7 @@ struct d40_phy_lli_bidir { * @lcsp02: Either maps to register lcsp0 if src or lcsp2 if dst. * @lcsp13: Either maps to register lcsp1 if src or lcsp3 if dst. * - * This struct must be 8 bytes aligned since it will be accessed directy by + * This struct must be 8 bytes aligned since it will be accessed directly by * the DMA. Never add any none hw mapped registers to this struct. */ diff --git a/drivers/dma/stm32/stm32-dma3.c b/drivers/dma/stm32/stm32-dma3.c index 4087e0263a48..0be6e944df6f 100644 --- a/drivers/dma/stm32/stm32-dma3.c +++ b/drivers/dma/stm32/stm32-dma3.c @@ -403,6 +403,7 @@ static struct stm32_dma3_swdesc *stm32_dma3_chan_desc_alloc(struct stm32_dma3_ch swdesc = kzalloc(struct_size(swdesc, lli, count), GFP_NOWAIT); if (!swdesc) return NULL; + swdesc->lli_size = count; for (i = 0; i < count; i++) { swdesc->lli[i].hwdesc = dma_pool_zalloc(chan->lli_pool, GFP_NOWAIT, @@ -410,7 +411,6 @@ static struct stm32_dma3_swdesc *stm32_dma3_chan_desc_alloc(struct stm32_dma3_ch if (!swdesc->lli[i].hwdesc) goto err_pool_free; } - swdesc->lli_size = count; swdesc->ccr = 0; /* Set LL base address */ diff --git a/drivers/dma/tegra20-apb-dma.c b/drivers/dma/tegra20-apb-dma.c index ac69778827f2..7d1acda2d72b 100644 --- a/drivers/dma/tegra20-apb-dma.c +++ b/drivers/dma/tegra20-apb-dma.c @@ -463,7 +463,7 @@ static void tegra_dma_configure_for_next(struct tegra_dma_channel *tdc, /* * If interrupt is pending then do nothing as the ISR will handle - * the programing for new request. + * the programming for new request. */ if (status & TEGRA_APBDMA_STATUS_ISE_EOC) { dev_err(tdc2dev(tdc), diff --git a/drivers/dma/ti/k3-udma.h b/drivers/dma/ti/k3-udma.h index d349c6d482ae..9062a237cd16 100644 --- a/drivers/dma/ti/k3-udma.h +++ b/drivers/dma/ti/k3-udma.h @@ -131,7 +131,6 @@ int xudma_navss_psil_unpair(struct udma_dev *ud, u32 src_thread, struct udma_dev *of_xudma_dev_get(struct device_node *np, const char *property); struct device *xudma_get_device(struct udma_dev *ud); struct k3_ringacc *xudma_get_ringacc(struct udma_dev *ud); -void xudma_dev_put(struct udma_dev *ud); u32 xudma_dev_get_psil_base(struct udma_dev *ud); struct udma_tisci_rm *xudma_dev_get_tisci_rm(struct udma_dev *ud); diff --git a/drivers/dma/ti/omap-dma.c b/drivers/dma/ti/omap-dma.c index 7e6c04afbe89..6ab9bfbdc480 100644 --- a/drivers/dma/ti/omap-dma.c +++ b/drivers/dma/ti/omap-dma.c @@ -1186,10 +1186,10 @@ static struct dma_async_tx_descriptor *omap_dma_prep_dma_cyclic( d->dev_addr = dev_addr; d->fi = burst; d->es = es; + d->sglen = 1; d->sg[0].addr = buf_addr; d->sg[0].en = period_len / es_bytes[es]; d->sg[0].fn = buf_len / period_len; - d->sglen = 1; d->ccr = c->ccr; if (dir == DMA_DEV_TO_MEM) @@ -1258,10 +1258,10 @@ static struct dma_async_tx_descriptor *omap_dma_prep_dma_memcpy( d->dev_addr = src; d->fi = 0; d->es = data_type; + d->sglen = 1; d->sg[0].en = len / BIT(data_type); d->sg[0].fn = 1; d->sg[0].addr = dest; - d->sglen = 1; d->ccr = c->ccr; d->ccr |= CCR_DST_AMODE_POSTINC | CCR_SRC_AMODE_POSTINC; @@ -1309,6 +1309,7 @@ static struct dma_async_tx_descriptor *omap_dma_prep_dma_interleaved( if (data_type > CSDP_DATA_TYPE_32) data_type = CSDP_DATA_TYPE_32; + d->sglen = 1; sg = &d->sg[0]; d->dir = DMA_MEM_TO_MEM; d->dev_addr = xt->src_start; @@ -1316,7 +1317,6 @@ static struct dma_async_tx_descriptor *omap_dma_prep_dma_interleaved( sg->en = xt->sgl[0].size / BIT(data_type); sg->fn = xt->numf; sg->addr = xt->dst_start; - d->sglen = 1; d->ccr = c->ccr; src_icg = dmaengine_get_src_icg(xt, &xt->sgl[0]); diff --git a/drivers/dma/xgene-dma.c b/drivers/dma/xgene-dma.c index fd4397adeb79..275848a9c450 100644 --- a/drivers/dma/xgene-dma.c +++ b/drivers/dma/xgene-dma.c @@ -1742,7 +1742,7 @@ static int xgene_dma_probe(struct platform_device *pdev) /* Initialize DMA channels software state */ xgene_dma_init_channels(pdma); - /* Configue DMA rings */ + /* Configure DMA rings */ ret = xgene_dma_init_rings(pdma); if (ret) goto err_clk_enable; diff --git a/drivers/dma/xilinx/xilinx_dpdma.c b/drivers/dma/xilinx/xilinx_dpdma.c index 36bd4825d389..be87764af9e8 100644 --- a/drivers/dma/xilinx/xilinx_dpdma.c +++ b/drivers/dma/xilinx/xilinx_dpdma.c @@ -149,7 +149,7 @@ struct xilinx_dpdma_chan; * @addr_ext: upper 16 bit of 48 bit address (next_desc and src_addr) * @next_desc: next descriptor 32 bit address * @src_addr: payload source address (1st page, 32 LSB) - * @addr_ext_23: payload source address (3nd and 3rd pages, 16 LSBs) + * @addr_ext_23: payload source address (2nd and 3rd pages, 16 LSBs) * @addr_ext_45: payload source address (4th and 5th pages, 16 LSBs) * @src_addr2: payload source address (2nd page, 32 LSB) * @src_addr3: payload source address (3rd page, 32 LSB) @@ -210,7 +210,7 @@ struct xilinx_dpdma_tx_desc { * @vchan: virtual DMA channel * @reg: register base address * @id: channel ID - * @wait_to_stop: queue to wait for outstanding transacitons before stopping + * @wait_to_stop: queue to wait for outstanding transactions before stopping * @running: true if the channel is running * @first_frame: flag for the first frame of stream * @video_group: flag if multi-channel operation is needed for video channels @@ -671,6 +671,84 @@ static void xilinx_dpdma_chan_free_tx_desc(struct virt_dma_desc *vdesc) } /** + * xilinx_dpdma_chan_prep_cyclic - Prepare a cyclic dma descriptor + * @chan: DPDMA channel + * @buf_addr: buffer address + * @buf_len: buffer length + * @period_len: number of periods + * @flags: tx flags argument passed in to prepare function + * + * Prepare a tx descriptor incudling internal software/hardware descriptors + * for the given cyclic transaction. + * + * Return: A dma async tx descriptor on success, or NULL. + */ +static struct dma_async_tx_descriptor * +xilinx_dpdma_chan_prep_cyclic(struct xilinx_dpdma_chan *chan, + dma_addr_t buf_addr, size_t buf_len, + size_t period_len, unsigned long flags) +{ + struct xilinx_dpdma_tx_desc *tx_desc; + struct xilinx_dpdma_sw_desc *sw_desc, *last = NULL; + unsigned int periods = buf_len / period_len; + unsigned int i; + + tx_desc = xilinx_dpdma_chan_alloc_tx_desc(chan); + if (!tx_desc) + return NULL; + + for (i = 0; i < periods; i++) { + struct xilinx_dpdma_hw_desc *hw_desc; + + if (!IS_ALIGNED(buf_addr, XILINX_DPDMA_ALIGN_BYTES)) { + dev_err(chan->xdev->dev, + "buffer should be aligned at %d B\n", + XILINX_DPDMA_ALIGN_BYTES); + goto error; + } + + sw_desc = xilinx_dpdma_chan_alloc_sw_desc(chan); + if (!sw_desc) + goto error; + + xilinx_dpdma_sw_desc_set_dma_addrs(chan->xdev, sw_desc, last, + &buf_addr, 1); + hw_desc = &sw_desc->hw; + hw_desc->xfer_size = period_len; + hw_desc->hsize_stride = + FIELD_PREP(XILINX_DPDMA_DESC_HSIZE_STRIDE_HSIZE_MASK, + period_len) | + FIELD_PREP(XILINX_DPDMA_DESC_HSIZE_STRIDE_STRIDE_MASK, + period_len); + hw_desc->control = XILINX_DPDMA_DESC_CONTROL_PREEMBLE | + XILINX_DPDMA_DESC_CONTROL_IGNORE_DONE | + XILINX_DPDMA_DESC_CONTROL_COMPLETE_INTR; + + list_add_tail(&sw_desc->node, &tx_desc->descriptors); + + buf_addr += period_len; + last = sw_desc; + } + + sw_desc = list_first_entry(&tx_desc->descriptors, + struct xilinx_dpdma_sw_desc, node); + last->hw.next_desc = lower_32_bits(sw_desc->dma_addr); + if (chan->xdev->ext_addr) + last->hw.addr_ext |= + FIELD_PREP(XILINX_DPDMA_DESC_ADDR_EXT_NEXT_ADDR_MASK, + upper_32_bits(sw_desc->dma_addr)); + + last->hw.control |= XILINX_DPDMA_DESC_CONTROL_LAST_OF_FRAME; + + return vchan_tx_prep(&chan->vchan, &tx_desc->vdesc, flags); + +error: + xilinx_dpdma_chan_free_tx_desc(&tx_desc->vdesc); + + return NULL; +} + +/** * xilinx_dpdma_chan_prep_interleaved_dma - Prepare an interleaved dma * descriptor * @chan: DPDMA channel @@ -1189,6 +1267,23 @@ out_unlock: /* ----------------------------------------------------------------------------- * DMA Engine Operations */ +static struct dma_async_tx_descriptor * +xilinx_dpdma_prep_dma_cyclic(struct dma_chan *dchan, dma_addr_t buf_addr, + size_t buf_len, size_t period_len, + enum dma_transfer_direction direction, + unsigned long flags) +{ + struct xilinx_dpdma_chan *chan = to_xilinx_chan(dchan); + + if (direction != DMA_MEM_TO_DEV) + return NULL; + + if (buf_len % period_len) + return NULL; + + return xilinx_dpdma_chan_prep_cyclic(chan, buf_addr, buf_len, + period_len, flags); +} static struct dma_async_tx_descriptor * xilinx_dpdma_prep_interleaved_dma(struct dma_chan *dchan, @@ -1672,6 +1767,7 @@ static int xilinx_dpdma_probe(struct platform_device *pdev) dma_cap_set(DMA_SLAVE, ddev->cap_mask); dma_cap_set(DMA_PRIVATE, ddev->cap_mask); + dma_cap_set(DMA_CYCLIC, ddev->cap_mask); dma_cap_set(DMA_INTERLEAVE, ddev->cap_mask); dma_cap_set(DMA_REPEAT, ddev->cap_mask); dma_cap_set(DMA_LOAD_EOT, ddev->cap_mask); @@ -1679,6 +1775,7 @@ static int xilinx_dpdma_probe(struct platform_device *pdev) ddev->device_alloc_chan_resources = xilinx_dpdma_alloc_chan_resources; ddev->device_free_chan_resources = xilinx_dpdma_free_chan_resources; + ddev->device_prep_dma_cyclic = xilinx_dpdma_prep_dma_cyclic; ddev->device_prep_interleaved_dma = xilinx_dpdma_prep_interleaved_dma; /* TODO: Can we achieve better granularity ? */ ddev->device_tx_status = dma_cookie_status; diff --git a/drivers/dma/xilinx/zynqmp_dma.c b/drivers/dma/xilinx/zynqmp_dma.c index f31631bef961..9ae46f1198fe 100644 --- a/drivers/dma/xilinx/zynqmp_dma.c +++ b/drivers/dma/xilinx/zynqmp_dma.c @@ -22,10 +22,10 @@ #include "../dmaengine.h" /* Register Offsets */ -#define ZYNQMP_DMA_ISR 0x100 -#define ZYNQMP_DMA_IMR 0x104 -#define ZYNQMP_DMA_IER 0x108 -#define ZYNQMP_DMA_IDS 0x10C +#define ZYNQMP_DMA_ISR (chan->irq_offset + 0x100) +#define ZYNQMP_DMA_IMR (chan->irq_offset + 0x104) +#define ZYNQMP_DMA_IER (chan->irq_offset + 0x108) +#define ZYNQMP_DMA_IDS (chan->irq_offset + 0x10c) #define ZYNQMP_DMA_CTRL0 0x110 #define ZYNQMP_DMA_CTRL1 0x114 #define ZYNQMP_DMA_DATA_ATTR 0x120 @@ -145,6 +145,9 @@ #define tx_to_desc(tx) container_of(tx, struct zynqmp_dma_desc_sw, \ async_tx) +/* IRQ Register offset for Versal Gen 2 */ +#define IRQ_REG_OFFSET 0x308 + /** * struct zynqmp_dma_desc_ll - Hw linked list descriptor * @addr: Buffer address @@ -211,6 +214,7 @@ struct zynqmp_dma_desc_sw { * @bus_width: Bus width * @src_burst_len: Source burst length * @dst_burst_len: Dest burst length + * @irq_offset: Irq register offset */ struct zynqmp_dma_chan { struct zynqmp_dma_device *zdev; @@ -235,6 +239,7 @@ struct zynqmp_dma_chan { u32 bus_width; u32 src_burst_len; u32 dst_burst_len; + u32 irq_offset; }; /** @@ -253,6 +258,14 @@ struct zynqmp_dma_device { struct clk *clk_apb; }; +struct zynqmp_dma_config { + u32 offset; +}; + +static const struct zynqmp_dma_config versal2_dma_config = { + .offset = IRQ_REG_OFFSET, +}; + static inline void zynqmp_dma_writeq(struct zynqmp_dma_chan *chan, u32 reg, u64 value) { @@ -892,6 +905,7 @@ static int zynqmp_dma_chan_probe(struct zynqmp_dma_device *zdev, { struct zynqmp_dma_chan *chan; struct device_node *node = pdev->dev.of_node; + const struct zynqmp_dma_config *match_data; int err; chan = devm_kzalloc(zdev->dev, sizeof(*chan), GFP_KERNEL); @@ -919,6 +933,10 @@ static int zynqmp_dma_chan_probe(struct zynqmp_dma_device *zdev, return -EINVAL; } + match_data = of_device_get_match_data(&pdev->dev); + if (match_data) + chan->irq_offset = match_data->offset; + chan->is_dmacoherent = of_property_read_bool(node, "dma-coherent"); zdev->chan = chan; tasklet_setup(&chan->tasklet, zynqmp_dma_do_tasklet); @@ -1161,6 +1179,7 @@ static void zynqmp_dma_remove(struct platform_device *pdev) } static const struct of_device_id zynqmp_dma_of_match[] = { + { .compatible = "amd,versal2-dma-1.0", .data = &versal2_dma_config }, { .compatible = "xlnx,zynqmp-dma-1.0", }, {} }; |