diff options
Diffstat (limited to 'drivers/gpu')
372 files changed, 9039 insertions, 3649 deletions
diff --git a/drivers/gpu/drm/Kconfig b/drivers/gpu/drm/Kconfig index 78d7fc0ebb57..83cb2a88c204 100644 --- a/drivers/gpu/drm/Kconfig +++ b/drivers/gpu/drm/Kconfig @@ -246,6 +246,8 @@ source "drivers/gpu/drm/fsl-dcu/Kconfig" source "drivers/gpu/drm/tegra/Kconfig" +source "drivers/gpu/drm/stm/Kconfig" + source "drivers/gpu/drm/panel/Kconfig" source "drivers/gpu/drm/bridge/Kconfig" @@ -274,6 +276,8 @@ source "drivers/gpu/drm/meson/Kconfig" source "drivers/gpu/drm/tinydrm/Kconfig" +source "drivers/gpu/drm/pl111/Kconfig" + # Keep legacy drivers last menuconfig DRM_LEGACY diff --git a/drivers/gpu/drm/Makefile b/drivers/gpu/drm/Makefile index 59f0f9b696eb..c156fecfb362 100644 --- a/drivers/gpu/drm/Makefile +++ b/drivers/gpu/drm/Makefile @@ -82,6 +82,7 @@ obj-$(CONFIG_DRM_BOCHS) += bochs/ obj-$(CONFIG_DRM_VIRTIO_GPU) += virtio/ obj-$(CONFIG_DRM_MSM) += msm/ obj-$(CONFIG_DRM_TEGRA) += tegra/ +obj-$(CONFIG_DRM_STM) += stm/ obj-$(CONFIG_DRM_STI) += sti/ obj-$(CONFIG_DRM_IMX) += imx/ obj-$(CONFIG_DRM_MEDIATEK) += mediatek/ @@ -96,3 +97,4 @@ obj-y += hisilicon/ obj-$(CONFIG_DRM_ZTE) += zte/ obj-$(CONFIG_DRM_MXSFB) += mxsfb/ obj-$(CONFIG_DRM_TINYDRM) += tinydrm/ +obj-$(CONFIG_DRM_PL111) += pl111/ diff --git a/drivers/gpu/drm/amd/amdgpu/Makefile b/drivers/gpu/drm/amd/amdgpu/Makefile index 660786aba7d2..20bde726419e 100644 --- a/drivers/gpu/drm/amd/amdgpu/Makefile +++ b/drivers/gpu/drm/amd/amdgpu/Makefile @@ -4,7 +4,7 @@ FULL_AMD_PATH=$(src)/.. -ccflags-y := -Iinclude/drm -I$(FULL_AMD_PATH)/include/asic_reg \ +ccflags-y := -I$(FULL_AMD_PATH)/include/asic_reg \ -I$(FULL_AMD_PATH)/include \ -I$(FULL_AMD_PATH)/amdgpu \ -I$(FULL_AMD_PATH)/scheduler \ diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h index 6a8129949333..77ff68f9932b 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h @@ -36,11 +36,11 @@ #include <linux/hashtable.h> #include <linux/dma-fence.h> -#include <ttm/ttm_bo_api.h> -#include <ttm/ttm_bo_driver.h> -#include <ttm/ttm_placement.h> -#include <ttm/ttm_module.h> -#include <ttm/ttm_execbuf_util.h> +#include <drm/ttm/ttm_bo_api.h> +#include <drm/ttm/ttm_bo_driver.h> +#include <drm/ttm/ttm_placement.h> +#include <drm/ttm/ttm_module.h> +#include <drm/ttm/ttm_execbuf_util.h> #include <drm/drmP.h> #include <drm/drm_gem.h> @@ -110,6 +110,7 @@ extern int amdgpu_pos_buf_per_se; extern int amdgpu_cntl_sb_buf_per_se; extern int amdgpu_param_buf_per_se; +#define AMDGPU_DEFAULT_GTT_SIZE_MB 3072ULL /* 3GB by default */ #define AMDGPU_WAIT_IDLE_TIMEOUT_IN_MS 3000 #define AMDGPU_MAX_USEC_TIMEOUT 100000 /* 100 ms */ #define AMDGPU_FENCE_JIFFIES_TIMEOUT (HZ / 2) @@ -966,6 +967,8 @@ struct amdgpu_gfx_config { unsigned mc_arb_ramcfg; unsigned gb_addr_config; unsigned num_rbs; + unsigned gs_vgt_table_depth; + unsigned gs_prim_buffer_depth; uint32_t tile_mode_array[32]; uint32_t macrotile_mode_array[16]; @@ -980,6 +983,7 @@ struct amdgpu_gfx_config { struct amdgpu_cu_info { uint32_t number; /* total active CU number */ uint32_t ao_cu_mask; + uint32_t wave_front_size; uint32_t bitmap[4][4]; }; @@ -1000,10 +1004,10 @@ struct amdgpu_ngg_buf { }; enum { - PRIM = 0, - POS, - CNTL, - PARAM, + NGG_PRIM = 0, + NGG_POS, + NGG_CNTL, + NGG_PARAM, NGG_BUF_MAX }; @@ -1125,6 +1129,7 @@ struct amdgpu_job { void *owner; uint64_t fence_ctx; /* the fence_context this job uses */ bool vm_needs_flush; + bool need_pipeline_sync; unsigned vm_id; uint64_t vm_pd_addr; uint32_t gds_base, gds_size; @@ -1704,9 +1709,6 @@ void amdgpu_mm_wdoorbell64(struct amdgpu_device *adev, u32 index, u64 v); #define WREG32_FIELD_OFFSET(reg, offset, field, val) \ WREG32(mm##reg + offset, (RREG32(mm##reg + offset) & ~REG_FIELD_MASK(reg, field)) | (val) << REG_FIELD_SHIFT(reg, field)) -#define WREG32_FIELD15(ip, idx, reg, field, val) \ - WREG32(SOC15_REG_OFFSET(ip, idx, mm##reg), (RREG32(SOC15_REG_OFFSET(ip, idx, mm##reg)) & ~REG_FIELD_MASK(reg, field)) | (val) << REG_FIELD_SHIFT(reg, field)) - /* * BIOS helpers. */ @@ -1910,10 +1912,6 @@ int amdgpu_device_resume(struct drm_device *dev, bool resume, bool fbcon); u32 amdgpu_get_vblank_counter_kms(struct drm_device *dev, unsigned int pipe); int amdgpu_enable_vblank_kms(struct drm_device *dev, unsigned int pipe); void amdgpu_disable_vblank_kms(struct drm_device *dev, unsigned int pipe); -int amdgpu_get_vblank_timestamp_kms(struct drm_device *dev, unsigned int pipe, - int *max_error, - struct timeval *vblank_time, - unsigned flags); long amdgpu_kms_compat_ioctl(struct file *filp, unsigned int cmd, unsigned long arg); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_atombios.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_atombios.c index ad4329922f79..1cf78f4dd339 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_atombios.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_atombios.c @@ -1727,6 +1727,12 @@ void amdgpu_atombios_scratch_regs_restore(struct amdgpu_device *adev) { int i; + /* + * VBIOS will check ASIC_INIT_COMPLETE bit to decide if + * execute ASIC_Init posting via driver + */ + adev->bios_scratch[7] &= ~ATOM_S7_ASIC_INIT_COMPLETE_MASK; + for (i = 0; i < AMDGPU_BIOS_NUM_SCRATCH; i++) WREG32(mmBIOS_SCRATCH_0 + i, adev->bios_scratch[i]); } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.c index 4b9abd68e04f..4bdda56fccee 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.c @@ -26,6 +26,7 @@ #include "atomfirmware.h" #include "amdgpu_atomfirmware.h" #include "atom.h" +#include "atombios.h" #define get_index_into_master_table(master_table, table_name) (offsetof(struct master_table, table_name) / sizeof(uint16_t)) @@ -77,10 +78,29 @@ void amdgpu_atomfirmware_scratch_regs_restore(struct amdgpu_device *adev) { int i; + /* + * VBIOS will check ASIC_INIT_COMPLETE bit to decide if + * execute ASIC_Init posting via driver + */ + adev->bios_scratch[7] &= ~ATOM_S7_ASIC_INIT_COMPLETE_MASK; + for (i = 0; i < AMDGPU_BIOS_NUM_SCRATCH; i++) WREG32(adev->bios_scratch_reg_offset + i, adev->bios_scratch[i]); } +void amdgpu_atomfirmware_scratch_regs_engine_hung(struct amdgpu_device *adev, + bool hung) +{ + u32 tmp = RREG32(adev->bios_scratch_reg_offset + 3); + + if (hung) + tmp |= ATOM_S3_ASIC_GUI_ENGINE_HUNG; + else + tmp &= ~ATOM_S3_ASIC_GUI_ENGINE_HUNG; + + WREG32(adev->bios_scratch_reg_offset + 3, tmp); +} + int amdgpu_atomfirmware_allocate_fb_scratch(struct amdgpu_device *adev) { struct atom_context *ctx = adev->mode_info.atom_context; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.h index d0c4dcd7fa96..a2c3ebe22c71 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.h @@ -28,6 +28,8 @@ bool amdgpu_atomfirmware_gpu_supports_virtualization(struct amdgpu_device *adev) void amdgpu_atomfirmware_scratch_regs_init(struct amdgpu_device *adev); void amdgpu_atomfirmware_scratch_regs_save(struct amdgpu_device *adev); void amdgpu_atomfirmware_scratch_regs_restore(struct amdgpu_device *adev); +void amdgpu_atomfirmware_scratch_regs_engine_hung(struct amdgpu_device *adev, + bool hung); int amdgpu_atomfirmware_allocate_fb_scratch(struct amdgpu_device *adev); #endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_benchmark.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_benchmark.c index cc97eee93226..1beae5b930d0 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_benchmark.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_benchmark.c @@ -117,8 +117,13 @@ static void amdgpu_benchmark_move(struct amdgpu_device *adev, unsigned size, } out_cleanup: + /* Check error value now. The value can be overwritten when clean up.*/ + if (r) { + DRM_ERROR("Error while benchmarking BO move.\n"); + } + if (sobj) { - r = amdgpu_bo_reserve(sobj, false); + r = amdgpu_bo_reserve(sobj, true); if (likely(r == 0)) { amdgpu_bo_unpin(sobj); amdgpu_bo_unreserve(sobj); @@ -126,17 +131,13 @@ out_cleanup: amdgpu_bo_unref(&sobj); } if (dobj) { - r = amdgpu_bo_reserve(dobj, false); + r = amdgpu_bo_reserve(dobj, true); if (likely(r == 0)) { amdgpu_bo_unpin(dobj); amdgpu_bo_unreserve(dobj); } amdgpu_bo_unref(&dobj); } - - if (r) { - DRM_ERROR("Error while benchmarking BO move.\n"); - } } void amdgpu_benchmark(struct amdgpu_device *adev, int test_number) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_bo_list.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_bo_list.c index a6649874e6ce..9f0247cdda5e 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_bo_list.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_bo_list.c @@ -96,7 +96,7 @@ static int amdgpu_bo_list_set(struct amdgpu_device *adev, int r; unsigned long total_size = 0; - array = drm_malloc_ab(num_entries, sizeof(struct amdgpu_bo_list_entry)); + array = kvmalloc_array(num_entries, sizeof(struct amdgpu_bo_list_entry), GFP_KERNEL); if (!array) return -ENOMEM; memset(array, 0, num_entries * sizeof(struct amdgpu_bo_list_entry)); @@ -148,7 +148,7 @@ static int amdgpu_bo_list_set(struct amdgpu_device *adev, for (i = 0; i < list->num_entries; ++i) amdgpu_bo_unref(&list->array[i].robj); - drm_free_large(list->array); + kvfree(list->array); list->gds_obj = gds_obj; list->gws_obj = gws_obj; @@ -163,7 +163,7 @@ static int amdgpu_bo_list_set(struct amdgpu_device *adev, error_free: while (i--) amdgpu_bo_unref(&array[i].robj); - drm_free_large(array); + kvfree(array); return r; } @@ -224,7 +224,7 @@ void amdgpu_bo_list_free(struct amdgpu_bo_list *list) amdgpu_bo_unref(&list->array[i].robj); mutex_destroy(&list->lock); - drm_free_large(list->array); + kvfree(list->array); kfree(list); } @@ -244,8 +244,8 @@ int amdgpu_bo_list_ioctl(struct drm_device *dev, void *data, int r; - info = drm_malloc_ab(args->in.bo_number, - sizeof(struct drm_amdgpu_bo_list_entry)); + info = kvmalloc_array(args->in.bo_number, + sizeof(struct drm_amdgpu_bo_list_entry), GFP_KERNEL); if (!info) return -ENOMEM; @@ -311,11 +311,11 @@ int amdgpu_bo_list_ioctl(struct drm_device *dev, void *data, memset(args, 0, sizeof(*args)); args->out.list_handle = handle; - drm_free_large(info); + kvfree(info); return 0; error_free: - drm_free_large(info); + kvfree(info); return r; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c index 1c7e6c28f93a..c6dba1eaefbd 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c @@ -42,82 +42,6 @@ struct amdgpu_cgs_device { struct amdgpu_device *adev = \ ((struct amdgpu_cgs_device *)cgs_device)->adev -static int amdgpu_cgs_gpu_mem_info(struct cgs_device *cgs_device, enum cgs_gpu_mem_type type, - uint64_t *mc_start, uint64_t *mc_size, - uint64_t *mem_size) -{ - CGS_FUNC_ADEV; - switch(type) { - case CGS_GPU_MEM_TYPE__VISIBLE_CONTIG_FB: - case CGS_GPU_MEM_TYPE__VISIBLE_FB: - *mc_start = 0; - *mc_size = adev->mc.visible_vram_size; - *mem_size = adev->mc.visible_vram_size - adev->vram_pin_size; - break; - case CGS_GPU_MEM_TYPE__INVISIBLE_CONTIG_FB: - case CGS_GPU_MEM_TYPE__INVISIBLE_FB: - *mc_start = adev->mc.visible_vram_size; - *mc_size = adev->mc.real_vram_size - adev->mc.visible_vram_size; - *mem_size = *mc_size; - break; - case CGS_GPU_MEM_TYPE__GART_CACHEABLE: - case CGS_GPU_MEM_TYPE__GART_WRITECOMBINE: - *mc_start = adev->mc.gtt_start; - *mc_size = adev->mc.gtt_size; - *mem_size = adev->mc.gtt_size - adev->gart_pin_size; - break; - default: - return -EINVAL; - } - - return 0; -} - -static int amdgpu_cgs_gmap_kmem(struct cgs_device *cgs_device, void *kmem, - uint64_t size, - uint64_t min_offset, uint64_t max_offset, - cgs_handle_t *kmem_handle, uint64_t *mcaddr) -{ - CGS_FUNC_ADEV; - int ret; - struct amdgpu_bo *bo; - struct page *kmem_page = vmalloc_to_page(kmem); - int npages = ALIGN(size, PAGE_SIZE) >> PAGE_SHIFT; - - struct sg_table *sg = drm_prime_pages_to_sg(&kmem_page, npages); - ret = amdgpu_bo_create(adev, size, PAGE_SIZE, false, - AMDGPU_GEM_DOMAIN_GTT, 0, sg, NULL, &bo); - if (ret) - return ret; - ret = amdgpu_bo_reserve(bo, false); - if (unlikely(ret != 0)) - return ret; - - /* pin buffer into GTT */ - ret = amdgpu_bo_pin_restricted(bo, AMDGPU_GEM_DOMAIN_GTT, - min_offset, max_offset, mcaddr); - amdgpu_bo_unreserve(bo); - - *kmem_handle = (cgs_handle_t)bo; - return ret; -} - -static int amdgpu_cgs_gunmap_kmem(struct cgs_device *cgs_device, cgs_handle_t kmem_handle) -{ - struct amdgpu_bo *obj = (struct amdgpu_bo *)kmem_handle; - - if (obj) { - int r = amdgpu_bo_reserve(obj, false); - if (likely(r == 0)) { - amdgpu_bo_unpin(obj); - amdgpu_bo_unreserve(obj); - } - amdgpu_bo_unref(&obj); - - } - return 0; -} - static int amdgpu_cgs_alloc_gpu_mem(struct cgs_device *cgs_device, enum cgs_gpu_mem_type type, uint64_t size, uint64_t align, @@ -215,7 +139,7 @@ static int amdgpu_cgs_free_gpu_mem(struct cgs_device *cgs_device, cgs_handle_t h struct amdgpu_bo *obj = (struct amdgpu_bo *)handle; if (obj) { - int r = amdgpu_bo_reserve(obj, false); + int r = amdgpu_bo_reserve(obj, true); if (likely(r == 0)) { amdgpu_bo_kunmap(obj); amdgpu_bo_unpin(obj); @@ -239,7 +163,7 @@ static int amdgpu_cgs_gmap_gpu_mem(struct cgs_device *cgs_device, cgs_handle_t h min_offset = obj->placements[0].fpfn << PAGE_SHIFT; max_offset = obj->placements[0].lpfn << PAGE_SHIFT; - r = amdgpu_bo_reserve(obj, false); + r = amdgpu_bo_reserve(obj, true); if (unlikely(r != 0)) return r; r = amdgpu_bo_pin_restricted(obj, obj->prefered_domains, @@ -252,7 +176,7 @@ static int amdgpu_cgs_gunmap_gpu_mem(struct cgs_device *cgs_device, cgs_handle_t { int r; struct amdgpu_bo *obj = (struct amdgpu_bo *)handle; - r = amdgpu_bo_reserve(obj, false); + r = amdgpu_bo_reserve(obj, true); if (unlikely(r != 0)) return r; r = amdgpu_bo_unpin(obj); @@ -265,7 +189,7 @@ static int amdgpu_cgs_kmap_gpu_mem(struct cgs_device *cgs_device, cgs_handle_t h { int r; struct amdgpu_bo *obj = (struct amdgpu_bo *)handle; - r = amdgpu_bo_reserve(obj, false); + r = amdgpu_bo_reserve(obj, true); if (unlikely(r != 0)) return r; r = amdgpu_bo_kmap(obj, map); @@ -277,7 +201,7 @@ static int amdgpu_cgs_kunmap_gpu_mem(struct cgs_device *cgs_device, cgs_handle_t { int r; struct amdgpu_bo *obj = (struct amdgpu_bo *)handle; - r = amdgpu_bo_reserve(obj, false); + r = amdgpu_bo_reserve(obj, true); if (unlikely(r != 0)) return r; amdgpu_bo_kunmap(obj); @@ -349,62 +273,6 @@ static void amdgpu_cgs_write_ind_register(struct cgs_device *cgs_device, WARN(1, "Invalid indirect register space"); } -static uint8_t amdgpu_cgs_read_pci_config_byte(struct cgs_device *cgs_device, unsigned addr) -{ - CGS_FUNC_ADEV; - uint8_t val; - int ret = pci_read_config_byte(adev->pdev, addr, &val); - if (WARN(ret, "pci_read_config_byte error")) - return 0; - return val; -} - -static uint16_t amdgpu_cgs_read_pci_config_word(struct cgs_device *cgs_device, unsigned addr) -{ - CGS_FUNC_ADEV; - uint16_t val; - int ret = pci_read_config_word(adev->pdev, addr, &val); - if (WARN(ret, "pci_read_config_word error")) - return 0; - return val; -} - -static uint32_t amdgpu_cgs_read_pci_config_dword(struct cgs_device *cgs_device, - unsigned addr) -{ - CGS_FUNC_ADEV; - uint32_t val; - int ret = pci_read_config_dword(adev->pdev, addr, &val); - if (WARN(ret, "pci_read_config_dword error")) - return 0; - return val; -} - -static void amdgpu_cgs_write_pci_config_byte(struct cgs_device *cgs_device, unsigned addr, - uint8_t value) -{ - CGS_FUNC_ADEV; - int ret = pci_write_config_byte(adev->pdev, addr, value); - WARN(ret, "pci_write_config_byte error"); -} - -static void amdgpu_cgs_write_pci_config_word(struct cgs_device *cgs_device, unsigned addr, - uint16_t value) -{ - CGS_FUNC_ADEV; - int ret = pci_write_config_word(adev->pdev, addr, value); - WARN(ret, "pci_write_config_word error"); -} - -static void amdgpu_cgs_write_pci_config_dword(struct cgs_device *cgs_device, unsigned addr, - uint32_t value) -{ - CGS_FUNC_ADEV; - int ret = pci_write_config_dword(adev->pdev, addr, value); - WARN(ret, "pci_write_config_dword error"); -} - - static int amdgpu_cgs_get_pci_resource(struct cgs_device *cgs_device, enum cgs_resource_type resource_type, uint64_t size, @@ -477,56 +345,6 @@ static int amdgpu_cgs_atom_exec_cmd_table(struct cgs_device *cgs_device, unsigne adev->mode_info.atom_context, table, args); } -static int amdgpu_cgs_create_pm_request(struct cgs_device *cgs_device, cgs_handle_t *request) -{ - /* TODO */ - return 0; -} - -static int amdgpu_cgs_destroy_pm_request(struct cgs_device *cgs_device, cgs_handle_t request) -{ - /* TODO */ - return 0; -} - -static int amdgpu_cgs_set_pm_request(struct cgs_device *cgs_device, cgs_handle_t request, - int active) -{ - /* TODO */ - return 0; -} - -static int amdgpu_cgs_pm_request_clock(struct cgs_device *cgs_device, cgs_handle_t request, - enum cgs_clock clock, unsigned freq) -{ - /* TODO */ - return 0; -} - -static int amdgpu_cgs_pm_request_engine(struct cgs_device *cgs_device, cgs_handle_t request, - enum cgs_engine engine, int powered) -{ - /* TODO */ - return 0; -} - - - -static int amdgpu_cgs_pm_query_clock_limits(struct cgs_device *cgs_device, - enum cgs_clock clock, - struct cgs_clock_limits *limits) -{ - /* TODO */ - return 0; -} - -static int amdgpu_cgs_set_camera_voltages(struct cgs_device *cgs_device, uint32_t mask, - const uint32_t *voltages) -{ - DRM_ERROR("not implemented"); - return -EPERM; -} - struct cgs_irq_params { unsigned src_id; cgs_irq_source_set_func_t set; @@ -1269,9 +1087,6 @@ static int amdgpu_cgs_call_acpi_method(struct cgs_device *cgs_device, } static const struct cgs_ops amdgpu_cgs_ops = { - .gpu_mem_info = amdgpu_cgs_gpu_mem_info, - .gmap_kmem = amdgpu_cgs_gmap_kmem, - .gunmap_kmem = amdgpu_cgs_gunmap_kmem, .alloc_gpu_mem = amdgpu_cgs_alloc_gpu_mem, .free_gpu_mem = amdgpu_cgs_free_gpu_mem, .gmap_gpu_mem = amdgpu_cgs_gmap_gpu_mem, @@ -1282,23 +1097,10 @@ static const struct cgs_ops amdgpu_cgs_ops = { .write_register = amdgpu_cgs_write_register, .read_ind_register = amdgpu_cgs_read_ind_register, .write_ind_register = amdgpu_cgs_write_ind_register, - .read_pci_config_byte = amdgpu_cgs_read_pci_config_byte, - .read_pci_config_word = amdgpu_cgs_read_pci_config_word, - .read_pci_config_dword = amdgpu_cgs_read_pci_config_dword, - .write_pci_config_byte = amdgpu_cgs_write_pci_config_byte, - .write_pci_config_word = amdgpu_cgs_write_pci_config_word, - .write_pci_config_dword = amdgpu_cgs_write_pci_config_dword, .get_pci_resource = amdgpu_cgs_get_pci_resource, .atom_get_data_table = amdgpu_cgs_atom_get_data_table, .atom_get_cmd_table_revs = amdgpu_cgs_atom_get_cmd_table_revs, .atom_exec_cmd_table = amdgpu_cgs_atom_exec_cmd_table, - .create_pm_request = amdgpu_cgs_create_pm_request, - .destroy_pm_request = amdgpu_cgs_destroy_pm_request, - .set_pm_request = amdgpu_cgs_set_pm_request, - .pm_request_clock = amdgpu_cgs_pm_request_clock, - .pm_request_engine = amdgpu_cgs_pm_request_engine, - .pm_query_clock_limits = amdgpu_cgs_pm_query_clock_limits, - .set_camera_voltages = amdgpu_cgs_set_camera_voltages, .get_firmware_info = amdgpu_cgs_get_firmware_info, .rel_firmware = amdgpu_cgs_rel_firmware, .set_powergating_state = amdgpu_cgs_set_powergating_state, diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c index ec71b9320561..5b3e0f63a115 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c @@ -194,7 +194,7 @@ int amdgpu_cs_parser_init(struct amdgpu_cs_parser *p, void *data) size = p->chunks[i].length_dw; cdata = (void __user *)(uintptr_t)user_chunk.chunk_data; - p->chunks[i].kdata = drm_malloc_ab(size, sizeof(uint32_t)); + p->chunks[i].kdata = kvmalloc_array(size, sizeof(uint32_t), GFP_KERNEL); if (p->chunks[i].kdata == NULL) { ret = -ENOMEM; i--; @@ -247,7 +247,7 @@ free_all_kdata: i = p->nchunks - 1; free_partial_kdata: for (; i >= 0; i--) - drm_free_large(p->chunks[i].kdata); + kvfree(p->chunks[i].kdata); kfree(p->chunks); p->chunks = NULL; p->nchunks = 0; @@ -505,7 +505,7 @@ static int amdgpu_cs_list_validate(struct amdgpu_cs_parser *p, return r; if (binding_userptr) { - drm_free_large(lobj->user_pages); + kvfree(lobj->user_pages); lobj->user_pages = NULL; } } @@ -571,7 +571,7 @@ static int amdgpu_cs_parser_bos(struct amdgpu_cs_parser *p, release_pages(e->user_pages, e->robj->tbo.ttm->num_pages, false); - drm_free_large(e->user_pages); + kvfree(e->user_pages); e->user_pages = NULL; } @@ -601,8 +601,9 @@ static int amdgpu_cs_parser_bos(struct amdgpu_cs_parser *p, list_for_each_entry(e, &need_pages, tv.head) { struct ttm_tt *ttm = e->robj->tbo.ttm; - e->user_pages = drm_calloc_large(ttm->num_pages, - sizeof(struct page*)); + e->user_pages = kvmalloc_array(ttm->num_pages, + sizeof(struct page*), + GFP_KERNEL | __GFP_ZERO); if (!e->user_pages) { r = -ENOMEM; DRM_ERROR("calloc failure in %s\n", __func__); @@ -612,7 +613,7 @@ static int amdgpu_cs_parser_bos(struct amdgpu_cs_parser *p, r = amdgpu_ttm_tt_get_user_pages(ttm, e->user_pages); if (r) { DRM_ERROR("amdgpu_ttm_tt_get_user_pages failed.\n"); - drm_free_large(e->user_pages); + kvfree(e->user_pages); e->user_pages = NULL; goto error_free_pages; } @@ -708,7 +709,7 @@ error_free_pages: release_pages(e->user_pages, e->robj->tbo.ttm->num_pages, false); - drm_free_large(e->user_pages); + kvfree(e->user_pages); } } @@ -761,7 +762,7 @@ static void amdgpu_cs_parser_fini(struct amdgpu_cs_parser *parser, int error, bo amdgpu_bo_list_put(parser->bo_list); for (i = 0; i < parser->nchunks; i++) - drm_free_large(parser->chunks[i].kdata); + kvfree(parser->chunks[i].kdata); kfree(parser->chunks); if (parser->job) amdgpu_job_free(parser->job); @@ -1074,6 +1075,7 @@ static int amdgpu_cs_submit(struct amdgpu_cs_parser *p, cs->out.handle = amdgpu_ctx_add_fence(p->ctx, ring, p->fence); job->uf_sequence = cs->out.handle; amdgpu_job_free_resources(job); + amdgpu_cs_parser_fini(p, 0, true); trace_amdgpu_cs_ioctl(job); amd_sched_entity_push_job(&job->base); @@ -1129,7 +1131,10 @@ int amdgpu_cs_ioctl(struct drm_device *dev, void *data, struct drm_file *filp) goto out; r = amdgpu_cs_submit(&parser, cs); + if (r) + goto out; + return 0; out: amdgpu_cs_parser_fini(&parser, r, reserved_buffers); return r; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c index cf0500671353..90d1ac8a80f8 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c @@ -273,6 +273,9 @@ struct dma_fence *amdgpu_ctx_get_fence(struct amdgpu_ctx *ctx, spin_lock(&ctx->ring_lock); + if (seq == ~0ull) + seq = ctx->rings[ring->idx].sequence - 1; + if (seq >= cring->sequence) { spin_unlock(&ctx->ring_lock); return ERR_PTR(-EINVAL); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c index 483660742f75..43ca16b6eee2 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c @@ -53,7 +53,6 @@ #include "bif/bif_4_1_d.h" #include <linux/pci.h> #include <linux/firmware.h> -#include "amdgpu_pm.h" static int amdgpu_debugfs_regs_init(struct amdgpu_device *adev); static void amdgpu_debugfs_regs_cleanup(struct amdgpu_device *adev); @@ -350,7 +349,7 @@ static void amdgpu_vram_scratch_fini(struct amdgpu_device *adev) if (adev->vram_scratch.robj == NULL) { return; } - r = amdgpu_bo_reserve(adev->vram_scratch.robj, false); + r = amdgpu_bo_reserve(adev->vram_scratch.robj, true); if (likely(r == 0)) { amdgpu_bo_kunmap(adev->vram_scratch.robj); amdgpu_bo_unpin(adev->vram_scratch.robj); @@ -422,12 +421,11 @@ static int amdgpu_doorbell_init(struct amdgpu_device *adev) if (adev->doorbell.num_doorbells == 0) return -EINVAL; - adev->doorbell.ptr = ioremap(adev->doorbell.base, adev->doorbell.num_doorbells * sizeof(u32)); - if (adev->doorbell.ptr == NULL) { + adev->doorbell.ptr = ioremap(adev->doorbell.base, + adev->doorbell.num_doorbells * + sizeof(u32)); + if (adev->doorbell.ptr == NULL) return -ENOMEM; - } - DRM_INFO("doorbell mmio base: 0x%08X\n", (uint32_t)adev->doorbell.base); - DRM_INFO("doorbell mmio size: %u\n", (unsigned)adev->doorbell.size); return 0; } @@ -1584,9 +1582,6 @@ static int amdgpu_late_init(struct amdgpu_device *adev) } } - amdgpu_dpm_enable_uvd(adev, false); - amdgpu_dpm_enable_vce(adev, false); - return 0; } @@ -1854,7 +1849,6 @@ int amdgpu_device_init(struct amdgpu_device *adev, /* mutex initialization are all done here so we * can recall function without having locking issues */ - mutex_init(&adev->vm_manager.lock); atomic_set(&adev->irq.ih.lock, 0); mutex_init(&adev->firmware.mutex); mutex_init(&adev->pm.mutex); @@ -2071,7 +2065,8 @@ void amdgpu_device_fini(struct amdgpu_device *adev) DRM_INFO("amdgpu: finishing device.\n"); adev->shutdown = true; - drm_crtc_force_disable_all(adev->ddev); + if (adev->mode_info.mode_config_initialized) + drm_crtc_force_disable_all(adev->ddev); /* evict vram memory */ amdgpu_bo_evict_vram(adev); amdgpu_ib_pool_fini(adev); @@ -2146,7 +2141,7 @@ int amdgpu_device_suspend(struct drm_device *dev, bool suspend, bool fbcon) if (amdgpu_crtc->cursor_bo) { struct amdgpu_bo *aobj = gem_to_amdgpu_bo(amdgpu_crtc->cursor_bo); - r = amdgpu_bo_reserve(aobj, false); + r = amdgpu_bo_reserve(aobj, true); if (r == 0) { amdgpu_bo_unpin(aobj); amdgpu_bo_unreserve(aobj); @@ -2159,7 +2154,7 @@ int amdgpu_device_suspend(struct drm_device *dev, bool suspend, bool fbcon) robj = gem_to_amdgpu_bo(rfb->obj); /* don't unpin kernel fb objects */ if (!amdgpu_fbdev_robj_is_fb(adev, robj)) { - r = amdgpu_bo_reserve(robj, false); + r = amdgpu_bo_reserve(robj, true); if (r == 0) { amdgpu_bo_unpin(robj); amdgpu_bo_unreserve(robj); @@ -2216,7 +2211,7 @@ int amdgpu_device_resume(struct drm_device *dev, bool resume, bool fbcon) struct drm_connector *connector; struct amdgpu_device *adev = dev->dev_private; struct drm_crtc *crtc; - int r; + int r = 0; if (dev->switch_power_state == DRM_SWITCH_POWER_OFF) return 0; @@ -2228,11 +2223,8 @@ int amdgpu_device_resume(struct drm_device *dev, bool resume, bool fbcon) pci_set_power_state(dev->pdev, PCI_D0); pci_restore_state(dev->pdev); r = pci_enable_device(dev->pdev); - if (r) { - if (fbcon) - console_unlock(); - return r; - } + if (r) + goto unlock; } if (adev->is_atom_fw) amdgpu_atomfirmware_scratch_regs_restore(adev); @@ -2249,7 +2241,7 @@ int amdgpu_device_resume(struct drm_device *dev, bool resume, bool fbcon) r = amdgpu_resume(adev); if (r) { DRM_ERROR("amdgpu_resume failed (%d).\n", r); - return r; + goto unlock; } amdgpu_fence_driver_resume(adev); @@ -2260,11 +2252,8 @@ int amdgpu_device_resume(struct drm_device *dev, bool resume, bool fbcon) } r = amdgpu_late_init(adev); - if (r) { - if (fbcon) - console_unlock(); - return r; - } + if (r) + goto unlock; /* pin cursors */ list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) { @@ -2272,7 +2261,7 @@ int amdgpu_device_resume(struct drm_device *dev, bool resume, bool fbcon) if (amdgpu_crtc->cursor_bo) { struct amdgpu_bo *aobj = gem_to_amdgpu_bo(amdgpu_crtc->cursor_bo); - r = amdgpu_bo_reserve(aobj, false); + r = amdgpu_bo_reserve(aobj, true); if (r == 0) { r = amdgpu_bo_pin(aobj, AMDGPU_GEM_DOMAIN_VRAM, @@ -2314,12 +2303,14 @@ int amdgpu_device_resume(struct drm_device *dev, bool resume, bool fbcon) dev->dev->power.disable_depth--; #endif - if (fbcon) { + if (fbcon) amdgpu_fbdev_set_suspend(adev, 0); + +unlock: + if (fbcon) console_unlock(); - } - return 0; + return r; } static bool amdgpu_check_soft_reset(struct amdgpu_device *adev) @@ -2430,25 +2421,37 @@ static int amdgpu_recover_vram_from_shadow(struct amdgpu_device *adev, uint32_t domain; int r; - if (!bo->shadow) - return 0; + if (!bo->shadow) + return 0; + + r = amdgpu_bo_reserve(bo, true); + if (r) + return r; + domain = amdgpu_mem_type_to_domain(bo->tbo.mem.mem_type); + /* if bo has been evicted, then no need to recover */ + if (domain == AMDGPU_GEM_DOMAIN_VRAM) { + r = amdgpu_bo_validate(bo->shadow); + if (r) { + DRM_ERROR("bo validate failed!\n"); + goto err; + } - r = amdgpu_bo_reserve(bo, false); - if (r) - return r; - domain = amdgpu_mem_type_to_domain(bo->tbo.mem.mem_type); - /* if bo has been evicted, then no need to recover */ - if (domain == AMDGPU_GEM_DOMAIN_VRAM) { - r = amdgpu_bo_restore_from_shadow(adev, ring, bo, + r = amdgpu_ttm_bind(&bo->shadow->tbo, &bo->shadow->tbo.mem); + if (r) { + DRM_ERROR("%p bind failed\n", bo->shadow); + goto err; + } + + r = amdgpu_bo_restore_from_shadow(adev, ring, bo, NULL, fence, true); - if (r) { - DRM_ERROR("recover page table failed!\n"); - goto err; - } - } + if (r) { + DRM_ERROR("recover page table failed!\n"); + goto err; + } + } err: - amdgpu_bo_unreserve(bo); - return r; + amdgpu_bo_unreserve(bo); + return r; } /** @@ -2520,6 +2523,7 @@ int amdgpu_sriov_gpu_reset(struct amdgpu_device *adev, bool voluntary) ring = adev->mman.buffer_funcs_ring; mutex_lock(&adev->shadow_list_lock); list_for_each_entry_safe(bo, tmp, &adev->shadow_list, shadow_list) { + next = NULL; amdgpu_recover_vram_from_shadow(adev, ring, bo, &next); if (fence) { r = dma_fence_wait(fence, false); @@ -2593,7 +2597,7 @@ int amdgpu_gpu_reset(struct amdgpu_device *adev) for (i = 0; i < AMDGPU_MAX_RINGS; ++i) { struct amdgpu_ring *ring = adev->rings[i]; - if (!ring) + if (!ring || !ring->sched.thread) continue; kthread_park(ring->sched.thread); amd_sched_hw_job_reset(&ring->sched); @@ -2666,6 +2670,7 @@ retry: DRM_INFO("recover vram bo from shadow\n"); mutex_lock(&adev->shadow_list_lock); list_for_each_entry_safe(bo, tmp, &adev->shadow_list, shadow_list) { + next = NULL; amdgpu_recover_vram_from_shadow(adev, ring, bo, &next); if (fence) { r = dma_fence_wait(fence, false); @@ -2688,7 +2693,8 @@ retry: } for (i = 0; i < AMDGPU_MAX_RINGS; ++i) { struct amdgpu_ring *ring = adev->rings[i]; - if (!ring) + + if (!ring || !ring->sched.thread) continue; amd_sched_job_recovery(&ring->sched); @@ -2697,7 +2703,7 @@ retry: } else { dev_err(adev->dev, "asic resume failed (%d).\n", r); for (i = 0; i < AMDGPU_MAX_RINGS; ++i) { - if (adev->rings[i]) { + if (adev->rings[i] && adev->rings[i]->sched.thread) { kthread_unpark(adev->rings[i]->sched.thread); } } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c index 96926a221bd5..cdf2ab20166a 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c @@ -123,7 +123,7 @@ static void amdgpu_unpin_work_func(struct work_struct *__work) int r; /* unpin of the old buffer */ - r = amdgpu_bo_reserve(work->old_abo, false); + r = amdgpu_bo_reserve(work->old_abo, true); if (likely(r == 0)) { r = amdgpu_bo_unpin(work->old_abo); if (unlikely(r != 0)) { @@ -138,52 +138,11 @@ static void amdgpu_unpin_work_func(struct work_struct *__work) kfree(work); } - -static void amdgpu_flip_work_cleanup(struct amdgpu_flip_work *work) -{ - int i; - - amdgpu_bo_unref(&work->old_abo); - dma_fence_put(work->excl); - for (i = 0; i < work->shared_count; ++i) - dma_fence_put(work->shared[i]); - kfree(work->shared); - kfree(work); -} - -static void amdgpu_flip_cleanup_unreserve(struct amdgpu_flip_work *work, - struct amdgpu_bo *new_abo) -{ - amdgpu_bo_unreserve(new_abo); - amdgpu_flip_work_cleanup(work); -} - -static void amdgpu_flip_cleanup_unpin(struct amdgpu_flip_work *work, - struct amdgpu_bo *new_abo) -{ - if (unlikely(amdgpu_bo_unpin(new_abo) != 0)) - DRM_ERROR("failed to unpin new abo in error path\n"); - amdgpu_flip_cleanup_unreserve(work, new_abo); -} - -void amdgpu_crtc_cleanup_flip_ctx(struct amdgpu_flip_work *work, - struct amdgpu_bo *new_abo) -{ - if (unlikely(amdgpu_bo_reserve(new_abo, false) != 0)) { - DRM_ERROR("failed to reserve new abo in error path\n"); - amdgpu_flip_work_cleanup(work); - return; - } - amdgpu_flip_cleanup_unpin(work, new_abo); -} - -int amdgpu_crtc_prepare_flip(struct drm_crtc *crtc, - struct drm_framebuffer *fb, - struct drm_pending_vblank_event *event, - uint32_t page_flip_flags, - uint32_t target, - struct amdgpu_flip_work **work_p, - struct amdgpu_bo **new_abo_p) +int amdgpu_crtc_page_flip_target(struct drm_crtc *crtc, + struct drm_framebuffer *fb, + struct drm_pending_vblank_event *event, + uint32_t page_flip_flags, uint32_t target, + struct drm_modeset_acquire_ctx *ctx) { struct drm_device *dev = crtc->dev; struct amdgpu_device *adev = dev->dev_private; @@ -196,7 +155,7 @@ int amdgpu_crtc_prepare_flip(struct drm_crtc *crtc, unsigned long flags; u64 tiling_flags; u64 base; - int r; + int i, r; work = kzalloc(sizeof *work, GFP_KERNEL); if (work == NULL) @@ -257,80 +216,41 @@ int amdgpu_crtc_prepare_flip(struct drm_crtc *crtc, spin_unlock_irqrestore(&crtc->dev->event_lock, flags); r = -EBUSY; goto pflip_cleanup; - } - spin_unlock_irqrestore(&crtc->dev->event_lock, flags); - - *work_p = work; - *new_abo_p = new_abo; - - return 0; - -pflip_cleanup: - amdgpu_crtc_cleanup_flip_ctx(work, new_abo); - return r; - -unpin: - amdgpu_flip_cleanup_unpin(work, new_abo); - return r; - -unreserve: - amdgpu_flip_cleanup_unreserve(work, new_abo); - return r; -cleanup: - amdgpu_flip_work_cleanup(work); - return r; - -} - -void amdgpu_crtc_submit_flip(struct drm_crtc *crtc, - struct drm_framebuffer *fb, - struct amdgpu_flip_work *work, - struct amdgpu_bo *new_abo) -{ - unsigned long flags; - struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc); - - spin_lock_irqsave(&crtc->dev->event_lock, flags); amdgpu_crtc->pflip_status = AMDGPU_FLIP_PENDING; amdgpu_crtc->pflip_works = work; + + DRM_DEBUG_DRIVER("crtc:%d[%p], pflip_stat:AMDGPU_FLIP_PENDING, work: %p,\n", + amdgpu_crtc->crtc_id, amdgpu_crtc, work); /* update crtc fb */ crtc->primary->fb = fb; spin_unlock_irqrestore(&crtc->dev->event_lock, flags); - - DRM_DEBUG_DRIVER( - "crtc:%d[%p], pflip_stat:AMDGPU_FLIP_PENDING, work: %p,\n", - amdgpu_crtc->crtc_id, amdgpu_crtc, work); - amdgpu_flip_work_func(&work->flip_work.work); -} - -int amdgpu_crtc_page_flip_target(struct drm_crtc *crtc, - struct drm_framebuffer *fb, - struct drm_pending_vblank_event *event, - uint32_t page_flip_flags, - uint32_t target, - struct drm_modeset_acquire_ctx *ctx) -{ - struct amdgpu_bo *new_abo; - struct amdgpu_flip_work *work; - int r; + return 0; - r = amdgpu_crtc_prepare_flip(crtc, - fb, - event, - page_flip_flags, - target, - &work, - &new_abo); - if (r) - return r; +pflip_cleanup: + if (unlikely(amdgpu_bo_reserve(new_abo, false) != 0)) { + DRM_ERROR("failed to reserve new abo in error path\n"); + goto cleanup; + } +unpin: + if (unlikely(amdgpu_bo_unpin(new_abo) != 0)) { + DRM_ERROR("failed to unpin new abo in error path\n"); + } +unreserve: + amdgpu_bo_unreserve(new_abo); - amdgpu_crtc_submit_flip(crtc, fb, work, new_abo); +cleanup: + amdgpu_bo_unref(&work->old_abo); + dma_fence_put(work->excl); + for (i = 0; i < work->shared_count; ++i) + dma_fence_put(work->shared[i]); + kfree(work->shared); + kfree(work); - return 0; + return r; } int amdgpu_crtc_set_config(struct drm_mode_set *set, diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.c index 38e9b0d3659a..1cb52fd19060 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.c @@ -22,7 +22,7 @@ * Authors: Alex Deucher */ -#include "drmP.h" +#include <drm/drmP.h> #include "amdgpu.h" #include "amdgpu_atombios.h" #include "amdgpu_i2c.h" diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c index 4e0f7d2d87f1..31eddd85eb40 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c @@ -39,7 +39,7 @@ #include <linux/module.h> #include <linux/pm_runtime.h> #include <linux/vga_switcheroo.h> -#include "drm_crtc_helper.h" +#include <drm/drm_crtc_helper.h> #include "amdgpu.h" #include "amdgpu_irq.h" @@ -63,9 +63,11 @@ * - 3.11.0 - Add support for sensor query info (clocks, temp, etc). * - 3.12.0 - Add query for double offchip LDS buffers * - 3.13.0 - Add PRT support + * - 3.14.0 - Fix race in amdgpu_ctx_get_fence() and note new functionality + * - 3.15.0 - Export more gpu info for gfx9 */ #define KMS_DRIVER_MAJOR 3 -#define KMS_DRIVER_MINOR 13 +#define KMS_DRIVER_MINOR 15 #define KMS_DRIVER_PATCHLEVEL 0 int amdgpu_vram_limit = 0; @@ -453,7 +455,9 @@ static const struct pci_device_id pciidlist[] = { {0x1002, 0x6861, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA10|AMD_EXP_HW_SUPPORT}, {0x1002, 0x6862, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA10|AMD_EXP_HW_SUPPORT}, {0x1002, 0x6863, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA10|AMD_EXP_HW_SUPPORT}, + {0x1002, 0x6864, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA10|AMD_EXP_HW_SUPPORT}, {0x1002, 0x6867, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA10|AMD_EXP_HW_SUPPORT}, + {0x1002, 0x6868, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA10|AMD_EXP_HW_SUPPORT}, {0x1002, 0x686c, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA10|AMD_EXP_HW_SUPPORT}, {0x1002, 0x687f, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA10|AMD_EXP_HW_SUPPORT}, {0, 0, 0} @@ -711,6 +715,16 @@ static const struct file_operations amdgpu_driver_kms_fops = { #endif }; +static bool +amdgpu_get_crtc_scanout_position(struct drm_device *dev, unsigned int pipe, + bool in_vblank_irq, int *vpos, int *hpos, + ktime_t *stime, ktime_t *etime, + const struct drm_display_mode *mode) +{ + return amdgpu_get_crtc_scanoutpos(dev, pipe, 0, vpos, hpos, + stime, etime, mode); +} + static struct drm_driver kms_driver = { .driver_features = DRIVER_USE_AGP | @@ -725,8 +739,8 @@ static struct drm_driver kms_driver = { .get_vblank_counter = amdgpu_get_vblank_counter_kms, .enable_vblank = amdgpu_enable_vblank_kms, .disable_vblank = amdgpu_disable_vblank_kms, - .get_vblank_timestamp = amdgpu_get_vblank_timestamp_kms, - .get_scanout_position = amdgpu_get_crtc_scanoutpos, + .get_vblank_timestamp = drm_calc_vbltimestamp_from_scanoutpos, + .get_scanout_position = amdgpu_get_crtc_scanout_position, #if defined(CONFIG_DEBUG_FS) .debugfs_init = amdgpu_debugfs_init, #endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c index a48142d930c6..236d9950221b 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c @@ -112,7 +112,7 @@ static void amdgpufb_destroy_pinned_object(struct drm_gem_object *gobj) struct amdgpu_bo *abo = gem_to_amdgpu_bo(gobj); int ret; - ret = amdgpu_bo_reserve(abo, false); + ret = amdgpu_bo_reserve(abo, true); if (likely(ret == 0)) { amdgpu_bo_kunmap(abo); amdgpu_bo_unpin(abo); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.c index 6d691abe889c..902e6015abca 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.c @@ -27,6 +27,9 @@ */ #include <drm/drmP.h> #include <drm/amdgpu_drm.h> +#ifdef CONFIG_X86 +#include <asm/set_memory.h> +#endif #include "amdgpu.h" /* @@ -183,7 +186,7 @@ void amdgpu_gart_table_vram_unpin(struct amdgpu_device *adev) if (adev->gart.robj == NULL) { return; } - r = amdgpu_bo_reserve(adev->gart.robj, false); + r = amdgpu_bo_reserve(adev->gart.robj, true); if (likely(r == 0)) { amdgpu_bo_kunmap(adev->gart.robj); amdgpu_bo_unpin(adev->gart.robj); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c index 03a9c5cad222..94cb91cf93eb 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c @@ -139,6 +139,35 @@ int amdgpu_gem_object_open(struct drm_gem_object *obj, return 0; } +static int amdgpu_gem_vm_check(void *param, struct amdgpu_bo *bo) +{ + /* if anything is swapped out don't swap it in here, + just abort and wait for the next CS */ + if (!amdgpu_bo_gpu_accessible(bo)) + return -ERESTARTSYS; + + if (bo->shadow && !amdgpu_bo_gpu_accessible(bo->shadow)) + return -ERESTARTSYS; + + return 0; +} + +static bool amdgpu_gem_vm_ready(struct amdgpu_device *adev, + struct amdgpu_vm *vm, + struct list_head *list) +{ + struct ttm_validate_buffer *entry; + + list_for_each_entry(entry, list, head) { + struct amdgpu_bo *bo = + container_of(entry->bo, struct amdgpu_bo, tbo); + if (amdgpu_gem_vm_check(NULL, bo)) + return false; + } + + return !amdgpu_vm_validate_pt_bos(adev, vm, amdgpu_gem_vm_check, NULL); +} + void amdgpu_gem_object_close(struct drm_gem_object *obj, struct drm_file *file_priv) { @@ -148,15 +177,13 @@ void amdgpu_gem_object_close(struct drm_gem_object *obj, struct amdgpu_vm *vm = &fpriv->vm; struct amdgpu_bo_list_entry vm_pd; - struct list_head list, duplicates; + struct list_head list; struct ttm_validate_buffer tv; struct ww_acquire_ctx ticket; struct amdgpu_bo_va *bo_va; - struct dma_fence *fence = NULL; int r; INIT_LIST_HEAD(&list); - INIT_LIST_HEAD(&duplicates); tv.bo = &bo->tbo; tv.shared = true; @@ -164,16 +191,18 @@ void amdgpu_gem_object_close(struct drm_gem_object *obj, amdgpu_vm_get_pd_bo(vm, &list, &vm_pd); - r = ttm_eu_reserve_buffers(&ticket, &list, false, &duplicates); + r = ttm_eu_reserve_buffers(&ticket, &list, false, NULL); if (r) { dev_err(adev->dev, "leaking bo va because " "we fail to reserve bo (%d)\n", r); return; } bo_va = amdgpu_vm_bo_find(vm, bo); - if (bo_va) { - if (--bo_va->ref_count == 0) { - amdgpu_vm_bo_rmv(adev, bo_va); + if (bo_va && --bo_va->ref_count == 0) { + amdgpu_vm_bo_rmv(adev, bo_va); + + if (amdgpu_gem_vm_ready(adev, vm, &list)) { + struct dma_fence *fence = NULL; r = amdgpu_vm_clear_freed(adev, vm, &fence); if (unlikely(r)) { @@ -502,19 +531,6 @@ out: return r; } -static int amdgpu_gem_va_check(void *param, struct amdgpu_bo *bo) -{ - /* if anything is swapped out don't swap it in here, - just abort and wait for the next CS */ - if (!amdgpu_bo_gpu_accessible(bo)) - return -ERESTARTSYS; - - if (bo->shadow && !amdgpu_bo_gpu_accessible(bo->shadow)) - return -ERESTARTSYS; - - return 0; -} - /** * amdgpu_gem_va_update_vm -update the bo_va in its VM * @@ -533,19 +549,9 @@ static void amdgpu_gem_va_update_vm(struct amdgpu_device *adev, struct list_head *list, uint32_t operation) { - struct ttm_validate_buffer *entry; int r = -ERESTARTSYS; - list_for_each_entry(entry, list, head) { - struct amdgpu_bo *bo = - container_of(entry->bo, struct amdgpu_bo, tbo); - if (amdgpu_gem_va_check(NULL, bo)) - goto error; - } - - r = amdgpu_vm_validate_pt_bos(adev, vm, amdgpu_gem_va_check, - NULL); - if (r) + if (!amdgpu_gem_vm_ready(adev, vm, list)) goto error; r = amdgpu_vm_update_directories(adev, vm); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gtt_mgr.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gtt_mgr.c index 0335c2f331e9..f7d22c44034d 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gtt_mgr.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gtt_mgr.c @@ -134,6 +134,15 @@ int amdgpu_gtt_mgr_alloc(struct ttm_mem_type_manager *man, return r; } +void amdgpu_gtt_mgr_print(struct seq_file *m, struct ttm_mem_type_manager *man) +{ + struct amdgpu_device *adev = amdgpu_ttm_adev(man->bdev); + struct amdgpu_gtt_mgr *mgr = man->priv; + + seq_printf(m, "man size:%llu pages, gtt available:%llu pages, usage:%lluMB\n", + man->size, mgr->available, (u64)atomic64_read(&adev->gtt_usage) >> 20); + +} /** * amdgpu_gtt_mgr_new - allocate a new node * diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c index aab857d89d03..6e4ae0d983c2 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c @@ -160,6 +160,8 @@ int amdgpu_ib_schedule(struct amdgpu_ring *ring, unsigned num_ibs, dev_err(adev->dev, "scheduling IB failed (%d).\n", r); return r; } + if (ring->funcs->emit_pipeline_sync && job && job->need_pipeline_sync) + amdgpu_ring_emit_pipeline_sync(ring); if (vm) { r = amdgpu_vm_flush(ring, job); @@ -217,7 +219,8 @@ int amdgpu_ib_schedule(struct amdgpu_ring *ring, unsigned num_ibs, if (r) { dev_err(adev->dev, "failed to emit fence (%d)\n", r); if (job && job->vm_id) - amdgpu_vm_reset_id(adev, job->vm_id); + amdgpu_vm_reset_id(adev, ring->funcs->vmhub, + job->vm_id); amdgpu_ring_undo(ring); return r; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c index 86a12424c162..7570f2439a11 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c @@ -57,6 +57,7 @@ int amdgpu_job_alloc(struct amdgpu_device *adev, unsigned num_ibs, (*job)->vm = vm; (*job)->ibs = (void *)&(*job)[1]; (*job)->num_ibs = num_ibs; + (*job)->need_pipeline_sync = false; amdgpu_sync_create(&(*job)->sync); @@ -139,7 +140,7 @@ static struct dma_fence *amdgpu_job_dependency(struct amd_sched_job *sched_job) struct dma_fence *fence = amdgpu_sync_get_fence(&job->sync); - if (fence == NULL && vm && !job->vm_id) { + while (fence == NULL && vm && !job->vm_id) { struct amdgpu_ring *ring = job->ring; int r; @@ -152,6 +153,9 @@ static struct dma_fence *amdgpu_job_dependency(struct amd_sched_job *sched_job) fence = amdgpu_sync_get_fence(&job->sync); } + if (amd_sched_dependency_optimized(fence, sched_job->s_entity)) + job->need_pipeline_sync = true; + return fence; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c index 832be632478f..dca4be970d13 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c @@ -545,11 +545,22 @@ static int amdgpu_info_ioctl(struct drm_device *dev, void *data, struct drm_file adev->gfx.config.double_offchip_lds_buf; if (amdgpu_ngg) { - dev_info.prim_buf_gpu_addr = adev->gfx.ngg.buf[PRIM].gpu_addr; - dev_info.pos_buf_gpu_addr = adev->gfx.ngg.buf[POS].gpu_addr; - dev_info.cntl_sb_buf_gpu_addr = adev->gfx.ngg.buf[CNTL].gpu_addr; - dev_info.param_buf_gpu_addr = adev->gfx.ngg.buf[PARAM].gpu_addr; + dev_info.prim_buf_gpu_addr = adev->gfx.ngg.buf[NGG_PRIM].gpu_addr; + dev_info.prim_buf_size = adev->gfx.ngg.buf[NGG_PRIM].size; + dev_info.pos_buf_gpu_addr = adev->gfx.ngg.buf[NGG_POS].gpu_addr; + dev_info.pos_buf_size = adev->gfx.ngg.buf[NGG_POS].size; + dev_info.cntl_sb_buf_gpu_addr = adev->gfx.ngg.buf[NGG_CNTL].gpu_addr; + dev_info.cntl_sb_buf_size = adev->gfx.ngg.buf[NGG_CNTL].size; + dev_info.param_buf_gpu_addr = adev->gfx.ngg.buf[NGG_PARAM].gpu_addr; + dev_info.param_buf_size = adev->gfx.ngg.buf[NGG_PARAM].size; } + dev_info.wave_front_size = adev->gfx.cu_info.wave_front_size; + dev_info.num_shader_visible_vgprs = adev->gfx.config.max_gprs; + dev_info.num_cu_per_sh = adev->gfx.config.max_cu_per_sh; + dev_info.num_tcc_blocks = adev->gfx.config.max_texture_channel_caches; + dev_info.gs_vgt_table_depth = adev->gfx.config.gs_vgt_table_depth; + dev_info.gs_prim_buffer_depth = adev->gfx.config.gs_prim_buffer_depth; + dev_info.max_gs_waves_per_vgt = adev->gfx.config.max_gs_threads; return copy_to_user(out, &dev_info, min((size_t)size, sizeof(dev_info))) ? -EFAULT : 0; @@ -810,7 +821,7 @@ void amdgpu_driver_postclose_kms(struct drm_device *dev, if (amdgpu_sriov_vf(adev)) { /* TODO: how to handle reserve failure */ - BUG_ON(amdgpu_bo_reserve(adev->virt.csa_obj, false)); + BUG_ON(amdgpu_bo_reserve(adev->virt.csa_obj, true)); amdgpu_vm_bo_rmv(adev, fpriv->vm.csa_bo_va); fpriv->vm.csa_bo_va = NULL; amdgpu_bo_unreserve(adev->virt.csa_obj); @@ -934,47 +945,6 @@ void amdgpu_disable_vblank_kms(struct drm_device *dev, unsigned int pipe) amdgpu_irq_put(adev, &adev->crtc_irq, idx); } -/** - * amdgpu_get_vblank_timestamp_kms - get vblank timestamp - * - * @dev: drm dev pointer - * @crtc: crtc to get the timestamp for - * @max_error: max error - * @vblank_time: time value - * @flags: flags passed to the driver - * - * Gets the timestamp on the requested crtc based on the - * scanout position. (all asics). - * Returns postive status flags on success, negative error on failure. - */ -int amdgpu_get_vblank_timestamp_kms(struct drm_device *dev, unsigned int pipe, - int *max_error, - struct timeval *vblank_time, - unsigned flags) -{ - struct drm_crtc *crtc; - struct amdgpu_device *adev = dev->dev_private; - - if (pipe >= dev->num_crtcs) { - DRM_ERROR("Invalid crtc %u\n", pipe); - return -EINVAL; - } - - /* Get associated drm_crtc: */ - crtc = &adev->mode_info.crtcs[pipe]->base; - if (!crtc) { - /* This can occur on driver load if some component fails to - * initialize completely and driver is unloaded */ - DRM_ERROR("Uninitialized crtc %d\n", pipe); - return -EINVAL; - } - - /* Helper routine in DRM core does all the work: */ - return drm_calc_vbltimestamp_from_scanoutpos(dev, pipe, max_error, - vblank_time, flags, - &crtc->hwmode); -} - const struct drm_ioctl_desc amdgpu_ioctls_kms[] = { DRM_IOCTL_DEF_DRV(AMDGPU_GEM_CREATE, amdgpu_gem_create_ioctl, DRM_AUTH|DRM_RENDER_ALLOW), DRM_IOCTL_DEF_DRV(AMDGPU_CTX, amdgpu_ctx_ioctl, DRM_AUTH|DRM_RENDER_ALLOW), diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h index db8f8dda209c..43a9d3aec6c4 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h @@ -534,6 +534,9 @@ struct amdgpu_framebuffer { ((em) == ATOM_ENCODER_MODE_DP_MST)) /* Driver internal use only flags of amdgpu_get_crtc_scanoutpos() */ +#define DRM_SCANOUTPOS_VALID (1 << 0) +#define DRM_SCANOUTPOS_IN_VBLANK (1 << 1) +#define DRM_SCANOUTPOS_ACCURATE (1 << 2) #define USE_REAL_VBLANKSTART (1 << 30) #define GET_DISTANCE_TO_VBLANKSTART (1 << 31) @@ -597,21 +600,6 @@ int amdgpu_crtc_page_flip_target(struct drm_crtc *crtc, struct drm_pending_vblank_event *event, uint32_t page_flip_flags, uint32_t target, struct drm_modeset_acquire_ctx *ctx); -void amdgpu_crtc_cleanup_flip_ctx(struct amdgpu_flip_work *work, - struct amdgpu_bo *new_abo); -int amdgpu_crtc_prepare_flip(struct drm_crtc *crtc, - struct drm_framebuffer *fb, - struct drm_pending_vblank_event *event, - uint32_t page_flip_flags, - uint32_t target, - struct amdgpu_flip_work **work, - struct amdgpu_bo **new_abo); - -void amdgpu_crtc_submit_flip(struct drm_crtc *crtc, - struct drm_framebuffer *fb, - struct amdgpu_flip_work *work, - struct amdgpu_bo *new_abo); - extern const struct drm_mode_config_funcs amdgpu_mode_funcs; #endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c index cb89fff863c0..365883d7948d 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c @@ -295,7 +295,7 @@ void amdgpu_bo_free_kernel(struct amdgpu_bo **bo, u64 *gpu_addr, if (*bo == NULL) return; - if (likely(amdgpu_bo_reserve(*bo, false) == 0)) { + if (likely(amdgpu_bo_reserve(*bo, true) == 0)) { if (cpu_addr) amdgpu_bo_kunmap(*bo); @@ -543,6 +543,27 @@ err: return r; } +int amdgpu_bo_validate(struct amdgpu_bo *bo) +{ + uint32_t domain; + int r; + + if (bo->pin_count) + return 0; + + domain = bo->prefered_domains; + +retry: + amdgpu_ttm_placement_from_domain(bo, domain); + r = ttm_bo_validate(&bo->tbo, &bo->placement, false, false); + if (unlikely(r == -ENOMEM) && domain != bo->allowed_domains) { + domain = bo->allowed_domains; + goto retry; + } + + return r; +} + int amdgpu_bo_restore_from_shadow(struct amdgpu_device *adev, struct amdgpu_ring *ring, struct amdgpu_bo *bo, diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h index 15a723adca76..382485115b06 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h @@ -175,6 +175,7 @@ int amdgpu_bo_backup_to_shadow(struct amdgpu_device *adev, struct amdgpu_bo *bo, struct reservation_object *resv, struct dma_fence **fence, bool direct); +int amdgpu_bo_validate(struct amdgpu_bo *bo); int amdgpu_bo_restore_from_shadow(struct amdgpu_device *adev, struct amdgpu_ring *ring, struct amdgpu_bo *bo, diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c index 990fde2cf4fd..7df503aedb69 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c @@ -867,8 +867,7 @@ static ssize_t amdgpu_hwmon_get_pwm1_enable(struct device *dev, pwm_mode = amdgpu_dpm_get_fan_control_mode(adev); - /* never 0 (full-speed), fuse or smc-controlled always */ - return sprintf(buf, "%i\n", pwm_mode == FDO_PWM_MODE_STATIC ? 1 : 2); + return sprintf(buf, "%i\n", pwm_mode); } static ssize_t amdgpu_hwmon_set_pwm1_enable(struct device *dev, @@ -887,14 +886,7 @@ static ssize_t amdgpu_hwmon_set_pwm1_enable(struct device *dev, if (err) return err; - switch (value) { - case 1: /* manual, percent-based */ - amdgpu_dpm_set_fan_control_mode(adev, FDO_PWM_MODE_STATIC); - break; - default: /* disable */ - amdgpu_dpm_set_fan_control_mode(adev, 0); - break; - } + amdgpu_dpm_set_fan_control_mode(adev, value); return count; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_prime.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_prime.c index 3826d5aea0a6..6bdc866570ab 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_prime.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_prime.c @@ -113,7 +113,7 @@ void amdgpu_gem_prime_unpin(struct drm_gem_object *obj) struct amdgpu_bo *bo = gem_to_amdgpu_bo(obj); int ret = 0; - ret = amdgpu_bo_reserve(bo, false); + ret = amdgpu_bo_reserve(bo, true); if (unlikely(ret != 0)) return; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c index ed6e5799016e..596e3957bdd9 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c @@ -24,7 +24,7 @@ */ #include <linux/firmware.h> -#include "drmP.h" +#include <drm/drmP.h> #include "amdgpu.h" #include "amdgpu_psp.h" #include "amdgpu_ucode.h" @@ -55,6 +55,8 @@ static int psp_sw_init(void *handle) psp->bootloader_load_sos = psp_v3_1_bootloader_load_sos; psp->prep_cmd_buf = psp_v3_1_prep_cmd_buf; psp->ring_init = psp_v3_1_ring_init; + psp->ring_create = psp_v3_1_ring_create; + psp->ring_destroy = psp_v3_1_ring_destroy; psp->cmd_submit = psp_v3_1_cmd_submit; psp->compare_sram_data = psp_v3_1_compare_sram_data; psp->smu_reload_quirk = psp_v3_1_smu_reload_quirk; @@ -152,11 +154,6 @@ static void psp_prep_tmr_cmd_buf(struct psp_gfx_cmd_resp *cmd, static int psp_tmr_init(struct psp_context *psp) { int ret; - struct psp_gfx_cmd_resp *cmd; - - cmd = kzalloc(sizeof(struct psp_gfx_cmd_resp), GFP_KERNEL); - if (!cmd) - return -ENOMEM; /* * Allocate 3M memory aligned to 1M from Frame Buffer (local @@ -168,22 +165,30 @@ static int psp_tmr_init(struct psp_context *psp) ret = amdgpu_bo_create_kernel(psp->adev, 0x300000, 0x100000, AMDGPU_GEM_DOMAIN_VRAM, &psp->tmr_bo, &psp->tmr_mc_addr, &psp->tmr_buf); - if (ret) - goto failed; + + return ret; +} + +static int psp_tmr_load(struct psp_context *psp) +{ + int ret; + struct psp_gfx_cmd_resp *cmd; + + cmd = kzalloc(sizeof(struct psp_gfx_cmd_resp), GFP_KERNEL); + if (!cmd) + return -ENOMEM; psp_prep_tmr_cmd_buf(cmd, psp->tmr_mc_addr, 0x300000); ret = psp_cmd_submit_buf(psp, NULL, cmd, psp->fence_buf_mc_addr, 1); if (ret) - goto failed_mem; + goto failed; kfree(cmd); return 0; -failed_mem: - amdgpu_bo_free_kernel(&psp->tmr_bo, &psp->tmr_mc_addr, &psp->tmr_buf); failed: kfree(cmd); return ret; @@ -203,104 +208,78 @@ static void psp_prep_asd_cmd_buf(struct psp_gfx_cmd_resp *cmd, cmd->cmd.cmd_load_ta.cmd_buf_len = shared_size; } -static int psp_asd_load(struct psp_context *psp) +static int psp_asd_init(struct psp_context *psp) { int ret; - struct amdgpu_bo *asd_bo, *asd_shared_bo; - uint64_t asd_mc_addr, asd_shared_mc_addr; - void *asd_buf, *asd_shared_buf; - struct psp_gfx_cmd_resp *cmd; - - cmd = kzalloc(sizeof(struct psp_gfx_cmd_resp), GFP_KERNEL); - if (!cmd) - return -ENOMEM; /* * Allocate 16k memory aligned to 4k from Frame Buffer (local * physical) for shared ASD <-> Driver */ - ret = amdgpu_bo_create_kernel(psp->adev, PSP_ASD_SHARED_MEM_SIZE, PAGE_SIZE, - AMDGPU_GEM_DOMAIN_VRAM, - &asd_shared_bo, &asd_shared_mc_addr, &asd_buf); - if (ret) - goto failed; + ret = amdgpu_bo_create_kernel(psp->adev, PSP_ASD_SHARED_MEM_SIZE, + PAGE_SIZE, AMDGPU_GEM_DOMAIN_VRAM, + &psp->asd_shared_bo, + &psp->asd_shared_mc_addr, + &psp->asd_shared_buf); - /* - * Allocate 256k memory aligned to 4k from Frame Buffer (local - * physical) for ASD firmware - */ - ret = amdgpu_bo_create_kernel(psp->adev, PSP_ASD_BIN_SIZE, PAGE_SIZE, - AMDGPU_GEM_DOMAIN_VRAM, - &asd_bo, &asd_mc_addr, &asd_buf); - if (ret) - goto failed_mem; + return ret; +} + +static int psp_asd_load(struct psp_context *psp) +{ + int ret; + struct psp_gfx_cmd_resp *cmd; + + cmd = kzalloc(sizeof(struct psp_gfx_cmd_resp), GFP_KERNEL); + if (!cmd) + return -ENOMEM; - memcpy(asd_buf, psp->asd_start_addr, psp->asd_ucode_size); + memset(psp->fw_pri_buf, 0, PSP_1_MEG); + memcpy(psp->fw_pri_buf, psp->asd_start_addr, psp->asd_ucode_size); - psp_prep_asd_cmd_buf(cmd, asd_mc_addr, asd_shared_mc_addr, + psp_prep_asd_cmd_buf(cmd, psp->fw_pri_mc_addr, psp->asd_shared_mc_addr, psp->asd_ucode_size, PSP_ASD_SHARED_MEM_SIZE); ret = psp_cmd_submit_buf(psp, NULL, cmd, psp->fence_buf_mc_addr, 2); - if (ret) - goto failed_mem1; - amdgpu_bo_free_kernel(&asd_bo, &asd_mc_addr, &asd_buf); - amdgpu_bo_free_kernel(&asd_shared_bo, &asd_shared_mc_addr, &asd_shared_buf); kfree(cmd); - return 0; - -failed_mem1: - amdgpu_bo_free_kernel(&asd_bo, &asd_mc_addr, &asd_buf); -failed_mem: - amdgpu_bo_free_kernel(&asd_shared_bo, &asd_shared_mc_addr, &asd_shared_buf); -failed: - kfree(cmd); return ret; } -static int psp_load_fw(struct amdgpu_device *adev) +static int psp_hw_start(struct psp_context *psp) { int ret; - struct psp_gfx_cmd_resp *cmd; - int i; - struct amdgpu_firmware_info *ucode; - struct psp_context *psp = &adev->psp; - - cmd = kzalloc(sizeof(struct psp_gfx_cmd_resp), GFP_KERNEL); - if (!cmd) - return -ENOMEM; ret = psp_bootloader_load_sysdrv(psp); if (ret) - goto failed; + return ret; ret = psp_bootloader_load_sos(psp); if (ret) - goto failed; - - ret = psp_ring_init(psp, PSP_RING_TYPE__KM); - if (ret) - goto failed; + return ret; - ret = amdgpu_bo_create_kernel(adev, PSP_FENCE_BUFFER_SIZE, PAGE_SIZE, - AMDGPU_GEM_DOMAIN_VRAM, - &psp->fence_buf_bo, - &psp->fence_buf_mc_addr, - &psp->fence_buf); + ret = psp_ring_create(psp, PSP_RING_TYPE__KM); if (ret) - goto failed; - - memset(psp->fence_buf, 0, PSP_FENCE_BUFFER_SIZE); + return ret; - ret = psp_tmr_init(psp); + ret = psp_tmr_load(psp); if (ret) - goto failed_mem; + return ret; ret = psp_asd_load(psp); if (ret) - goto failed_mem; + return ret; + + return 0; +} + +static int psp_np_fw_load(struct psp_context *psp) +{ + int i, ret; + struct amdgpu_firmware_info *ucode; + struct amdgpu_device* adev = psp->adev; for (i = 0; i < adev->firmware.max_ucodes; i++) { ucode = &adev->firmware.ucode[i]; @@ -310,15 +289,21 @@ static int psp_load_fw(struct amdgpu_device *adev) if (ucode->ucode_id == AMDGPU_UCODE_ID_SMC && psp_smu_reload_quirk(psp)) continue; + if (amdgpu_sriov_vf(adev) && + (ucode->ucode_id == AMDGPU_UCODE_ID_SDMA0 + || ucode->ucode_id == AMDGPU_UCODE_ID_SDMA1 + || ucode->ucode_id == AMDGPU_UCODE_ID_RLC_G)) + /*skip ucode loading in SRIOV VF */ + continue; - ret = psp_prep_cmd_buf(ucode, cmd); + ret = psp_prep_cmd_buf(ucode, psp->cmd); if (ret) - goto failed_mem; + return ret; - ret = psp_cmd_submit_buf(psp, ucode, cmd, + ret = psp_cmd_submit_buf(psp, ucode, psp->cmd, psp->fence_buf_mc_addr, i + 3); if (ret) - goto failed_mem; + return ret; #if 0 /* check if firmware loaded sucessfully */ @@ -327,8 +312,59 @@ static int psp_load_fw(struct amdgpu_device *adev) #endif } - amdgpu_bo_free_kernel(&psp->fence_buf_bo, - &psp->fence_buf_mc_addr, &psp->fence_buf); + return 0; +} + +static int psp_load_fw(struct amdgpu_device *adev) +{ + int ret; + struct psp_context *psp = &adev->psp; + struct psp_gfx_cmd_resp *cmd; + + cmd = kzalloc(sizeof(struct psp_gfx_cmd_resp), GFP_KERNEL); + if (!cmd) + return -ENOMEM; + + psp->cmd = cmd; + + ret = amdgpu_bo_create_kernel(adev, PSP_1_MEG, PSP_1_MEG, + AMDGPU_GEM_DOMAIN_GTT, + &psp->fw_pri_bo, + &psp->fw_pri_mc_addr, + &psp->fw_pri_buf); + if (ret) + goto failed; + + ret = amdgpu_bo_create_kernel(adev, PSP_FENCE_BUFFER_SIZE, PAGE_SIZE, + AMDGPU_GEM_DOMAIN_VRAM, + &psp->fence_buf_bo, + &psp->fence_buf_mc_addr, + &psp->fence_buf); + if (ret) + goto failed_mem1; + + memset(psp->fence_buf, 0, PSP_FENCE_BUFFER_SIZE); + + ret = psp_ring_init(psp, PSP_RING_TYPE__KM); + if (ret) + goto failed_mem1; + + ret = psp_tmr_init(psp); + if (ret) + goto failed_mem; + + ret = psp_asd_init(psp); + if (ret) + goto failed_mem; + + ret = psp_hw_start(psp); + if (ret) + goto failed_mem; + + ret = psp_np_fw_load(psp); + if (ret) + goto failed_mem; + kfree(cmd); return 0; @@ -336,6 +372,9 @@ static int psp_load_fw(struct amdgpu_device *adev) failed_mem: amdgpu_bo_free_kernel(&psp->fence_buf_bo, &psp->fence_buf_mc_addr, &psp->fence_buf); +failed_mem1: + amdgpu_bo_free_kernel(&psp->fw_pri_bo, + &psp->fw_pri_mc_addr, &psp->fw_pri_buf); failed: kfree(cmd); return ret; @@ -379,12 +418,24 @@ static int psp_hw_fini(void *handle) struct amdgpu_device *adev = (struct amdgpu_device *)handle; struct psp_context *psp = &adev->psp; - if (adev->firmware.load_type == AMDGPU_FW_LOAD_PSP) - amdgpu_ucode_fini_bo(adev); + if (adev->firmware.load_type != AMDGPU_FW_LOAD_PSP) + return 0; + + amdgpu_ucode_fini_bo(adev); + + psp_ring_destroy(psp, PSP_RING_TYPE__KM); if (psp->tmr_buf) amdgpu_bo_free_kernel(&psp->tmr_bo, &psp->tmr_mc_addr, &psp->tmr_buf); + if (psp->fw_pri_buf) + amdgpu_bo_free_kernel(&psp->fw_pri_bo, + &psp->fw_pri_mc_addr, &psp->fw_pri_buf); + + if (psp->fence_buf_bo) + amdgpu_bo_free_kernel(&psp->fence_buf_bo, + &psp->fence_buf_mc_addr, &psp->fence_buf); + return 0; } @@ -397,18 +448,30 @@ static int psp_resume(void *handle) { int ret; struct amdgpu_device *adev = (struct amdgpu_device *)handle; + struct psp_context *psp = &adev->psp; if (adev->firmware.load_type != AMDGPU_FW_LOAD_PSP) return 0; + DRM_INFO("PSP is resuming...\n"); + mutex_lock(&adev->firmware.mutex); - ret = psp_load_fw(adev); + ret = psp_hw_start(psp); if (ret) - DRM_ERROR("PSP resume failed\n"); + goto failed; + + ret = psp_np_fw_load(psp); + if (ret) + goto failed; mutex_unlock(&adev->firmware.mutex); + return 0; + +failed: + DRM_ERROR("PSP resume failed\n"); + mutex_unlock(&adev->firmware.mutex); return ret; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.h index e9f35e025b59..0301e4e0b297 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.h @@ -30,8 +30,8 @@ #define PSP_FENCE_BUFFER_SIZE 0x1000 #define PSP_CMD_BUFFER_SIZE 0x1000 -#define PSP_ASD_BIN_SIZE 0x40000 #define PSP_ASD_SHARED_MEM_SIZE 0x4000 +#define PSP_1_MEG 0x100000 enum psp_ring_type { @@ -57,6 +57,7 @@ struct psp_context { struct amdgpu_device *adev; struct psp_ring km_ring; + struct psp_gfx_cmd_resp *cmd; int (*init_microcode)(struct psp_context *psp); int (*bootloader_load_sysdrv)(struct psp_context *psp); @@ -64,6 +65,9 @@ struct psp_context int (*prep_cmd_buf)(struct amdgpu_firmware_info *ucode, struct psp_gfx_cmd_resp *cmd); int (*ring_init)(struct psp_context *psp, enum psp_ring_type ring_type); + int (*ring_create)(struct psp_context *psp, enum psp_ring_type ring_type); + int (*ring_destroy)(struct psp_context *psp, + enum psp_ring_type ring_type); int (*cmd_submit)(struct psp_context *psp, struct amdgpu_firmware_info *ucode, uint64_t cmd_buf_mc_addr, uint64_t fence_mc_addr, int index); bool (*compare_sram_data)(struct psp_context *psp, @@ -71,6 +75,11 @@ struct psp_context enum AMDGPU_UCODE_ID ucode_type); bool (*smu_reload_quirk)(struct psp_context *psp); + /* fence buffer */ + struct amdgpu_bo *fw_pri_bo; + uint64_t fw_pri_mc_addr; + void *fw_pri_buf; + /* sos firmware */ const struct firmware *sos_fw; uint32_t sos_fw_version; @@ -85,12 +94,15 @@ struct psp_context uint64_t tmr_mc_addr; void *tmr_buf; - /* asd firmware */ + /* asd firmware and buffer */ const struct firmware *asd_fw; uint32_t asd_fw_version; uint32_t asd_feature_version; uint32_t asd_ucode_size; uint8_t *asd_start_addr; + struct amdgpu_bo *asd_shared_bo; + uint64_t asd_shared_mc_addr; + void *asd_shared_buf; /* fence buffer */ struct amdgpu_bo *fence_buf_bo; @@ -105,6 +117,8 @@ struct amdgpu_psp_funcs { #define psp_prep_cmd_buf(ucode, type) (psp)->prep_cmd_buf((ucode), (type)) #define psp_ring_init(psp, type) (psp)->ring_init((psp), (type)) +#define psp_ring_create(psp, type) (psp)->ring_create((psp), (type)) +#define psp_ring_destroy(psp, type) ((psp)->ring_destroy((psp), (type))) #define psp_cmd_submit(psp, ucode, cmd_mc, fence_mc, index) \ (psp)->cmd_submit((psp), (ucode), (cmd_mc), (fence_mc), (index)) #define psp_compare_sram_data(psp, ucode, type) \ diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h index 63e56398ca9a..944443c5b90a 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h @@ -99,6 +99,7 @@ struct amdgpu_ring_funcs { uint32_t align_mask; u32 nop; bool support_64bit_ptrs; + unsigned vmhub; /* ring read/write ptr handling */ u64 (*get_rptr)(struct amdgpu_ring *ring); @@ -178,6 +179,7 @@ struct amdgpu_ring { unsigned cond_exe_offs; u64 cond_exe_gpu_addr; volatile u32 *cond_exe_cpu_addr; + unsigned vm_inv_eng; #if defined(CONFIG_DEBUG_FS) struct dentry *ent; #endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_sa.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_sa.c index de9f919ae336..5ca75a456ad2 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_sa.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_sa.c @@ -130,7 +130,7 @@ int amdgpu_sa_bo_manager_suspend(struct amdgpu_device *adev, return -EINVAL; } - r = amdgpu_bo_reserve(sa_manager->bo, false); + r = amdgpu_bo_reserve(sa_manager->bo, true); if (!r) { amdgpu_bo_kunmap(sa_manager->bo); amdgpu_bo_unpin(sa_manager->bo); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h index ee9d0f346d75..8601904e670a 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h @@ -190,26 +190,29 @@ TRACE_EVENT(amdgpu_sched_run_job, TRACE_EVENT(amdgpu_vm_grab_id, - TP_PROTO(struct amdgpu_vm *vm, int ring, struct amdgpu_job *job), + TP_PROTO(struct amdgpu_vm *vm, struct amdgpu_ring *ring, + struct amdgpu_job *job), TP_ARGS(vm, ring, job), TP_STRUCT__entry( __field(struct amdgpu_vm *, vm) __field(u32, ring) - __field(u32, vmid) + __field(u32, vm_id) + __field(u32, vm_hub) __field(u64, pd_addr) __field(u32, needs_flush) ), TP_fast_assign( __entry->vm = vm; - __entry->ring = ring; - __entry->vmid = job->vm_id; + __entry->ring = ring->idx; + __entry->vm_id = job->vm_id; + __entry->vm_hub = ring->funcs->vmhub, __entry->pd_addr = job->vm_pd_addr; __entry->needs_flush = job->vm_needs_flush; ), - TP_printk("vm=%p, ring=%u, id=%u, pd_addr=%010Lx needs_flush=%u", - __entry->vm, __entry->ring, __entry->vmid, - __entry->pd_addr, __entry->needs_flush) + TP_printk("vm=%p, ring=%u, id=%u, hub=%u, pd_addr=%010Lx needs_flush=%u", + __entry->vm, __entry->ring, __entry->vm_id, + __entry->vm_hub, __entry->pd_addr, __entry->needs_flush) ); TRACE_EVENT(amdgpu_vm_bo_map, @@ -331,21 +334,25 @@ TRACE_EVENT(amdgpu_vm_copy_ptes, ); TRACE_EVENT(amdgpu_vm_flush, - TP_PROTO(uint64_t pd_addr, unsigned ring, unsigned id), - TP_ARGS(pd_addr, ring, id), + TP_PROTO(struct amdgpu_ring *ring, unsigned vm_id, + uint64_t pd_addr), + TP_ARGS(ring, vm_id, pd_addr), TP_STRUCT__entry( - __field(u64, pd_addr) __field(u32, ring) - __field(u32, id) + __field(u32, vm_id) + __field(u32, vm_hub) + __field(u64, pd_addr) ), TP_fast_assign( + __entry->ring = ring->idx; + __entry->vm_id = vm_id; + __entry->vm_hub = ring->funcs->vmhub; __entry->pd_addr = pd_addr; - __entry->ring = ring; - __entry->id = id; ), - TP_printk("ring=%u, id=%u, pd_addr=%010Lx", - __entry->ring, __entry->id, __entry->pd_addr) + TP_printk("ring=%u, id=%u, hub=%u, pd_addr=%010Lx", + __entry->ring, __entry->vm_id, + __entry->vm_hub,__entry->pd_addr) ); TRACE_EVENT(amdgpu_bo_list_set, diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c index 35d53a0d9ba6..b5fa003c1341 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c @@ -29,11 +29,11 @@ * Thomas Hellstrom <thomas-at-tungstengraphics-dot-com> * Dave Airlie */ -#include <ttm/ttm_bo_api.h> -#include <ttm/ttm_bo_driver.h> -#include <ttm/ttm_placement.h> -#include <ttm/ttm_module.h> -#include <ttm/ttm_page_alloc.h> +#include <drm/ttm/ttm_bo_api.h> +#include <drm/ttm/ttm_bo_driver.h> +#include <drm/ttm/ttm_placement.h> +#include <drm/ttm/ttm_module.h> +#include <drm/ttm/ttm_page_alloc.h> #include <drm/drmP.h> #include <drm/amdgpu_drm.h> #include <linux/seq_file.h> @@ -203,7 +203,9 @@ static void amdgpu_evict_flags(struct ttm_buffer_object *bo, abo = container_of(bo, struct amdgpu_bo, tbo); switch (bo->mem.mem_type) { case TTM_PL_VRAM: - if (adev->mman.buffer_funcs_ring->ready == false) { + if (adev->mman.buffer_funcs && + adev->mman.buffer_funcs_ring && + adev->mman.buffer_funcs_ring->ready == false) { amdgpu_ttm_placement_from_domain(abo, AMDGPU_GEM_DOMAIN_CPU); } else { amdgpu_ttm_placement_from_domain(abo, AMDGPU_GEM_DOMAIN_GTT); @@ -763,7 +765,7 @@ int amdgpu_ttm_recover_gart(struct amdgpu_device *adev) { struct amdgpu_ttm_tt *gtt, *tmp; struct ttm_mem_reg bo_mem; - uint32_t flags; + uint64_t flags; int r; bo_mem.mem_type = TTM_PL_TT; @@ -1038,11 +1040,17 @@ uint64_t amdgpu_ttm_tt_pte_flags(struct amdgpu_device *adev, struct ttm_tt *ttm, static bool amdgpu_ttm_bo_eviction_valuable(struct ttm_buffer_object *bo, const struct ttm_place *place) { - if (bo->mem.mem_type == TTM_PL_VRAM && - bo->mem.start == AMDGPU_BO_INVALID_OFFSET) { - unsigned long num_pages = bo->mem.num_pages; - struct drm_mm_node *node = bo->mem.mm_node; + unsigned long num_pages = bo->mem.num_pages; + struct drm_mm_node *node = bo->mem.mm_node; + + if (bo->mem.start != AMDGPU_BO_INVALID_OFFSET) + return ttm_bo_eviction_valuable(bo, place); + + switch (bo->mem.mem_type) { + case TTM_PL_TT: + return true; + case TTM_PL_VRAM: /* Check each drm MM node individually */ while (num_pages) { if (place->fpfn < (node->start + node->size) && @@ -1052,8 +1060,10 @@ static bool amdgpu_ttm_bo_eviction_valuable(struct ttm_buffer_object *bo, num_pages -= node->size; ++node; } + break; - return false; + default: + break; } return ttm_bo_eviction_valuable(bo, place); @@ -1188,7 +1198,7 @@ void amdgpu_ttm_fini(struct amdgpu_device *adev) return; amdgpu_ttm_debugfs_fini(adev); if (adev->stollen_vga_memory) { - r = amdgpu_bo_reserve(adev->stollen_vga_memory, false); + r = amdgpu_bo_reserve(adev->stollen_vga_memory, true); if (r == 0) { amdgpu_bo_unpin(adev->stollen_vga_memory); amdgpu_bo_unreserve(adev->stollen_vga_memory); @@ -1401,6 +1411,8 @@ error_free: #if defined(CONFIG_DEBUG_FS) +extern void amdgpu_gtt_mgr_print(struct seq_file *m, struct ttm_mem_type_manager + *man); static int amdgpu_mm_dump_table(struct seq_file *m, void *data) { struct drm_info_node *node = (struct drm_info_node *)m->private; @@ -1414,11 +1426,17 @@ static int amdgpu_mm_dump_table(struct seq_file *m, void *data) spin_lock(&glob->lru_lock); drm_mm_print(mm, &p); spin_unlock(&glob->lru_lock); - if (ttm_pl == TTM_PL_VRAM) + switch (ttm_pl) { + case TTM_PL_VRAM: seq_printf(m, "man size:%llu pages, ram usage:%lluMB, vis usage:%lluMB\n", adev->mman.bdev.man[ttm_pl].size, (u64)atomic64_read(&adev->vram_usage) >> 20, (u64)atomic64_read(&adev->vram_vis_usage) >> 20); + break; + case TTM_PL_TT: + amdgpu_gtt_mgr_print(m, &adev->mman.bdev.man[TTM_PL_TT]); + break; + } return 0; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c index a1891c93cdbf..dfd1c98efa7c 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c @@ -382,10 +382,14 @@ int amdgpu_ucode_init_bo(struct amdgpu_device *adev) * if SMU loaded firmware, it needn't add SMC, UVD, and VCE * ucode info here */ - if (adev->firmware.load_type != AMDGPU_FW_LOAD_PSP) - adev->firmware.max_ucodes = AMDGPU_UCODE_ID_MAXIMUM - 4; - else + if (adev->firmware.load_type != AMDGPU_FW_LOAD_PSP) { + if (amdgpu_sriov_vf(adev)) + adev->firmware.max_ucodes = AMDGPU_UCODE_ID_MAXIMUM - 3; + else + adev->firmware.max_ucodes = AMDGPU_UCODE_ID_MAXIMUM - 4; + } else { adev->firmware.max_ucodes = AMDGPU_UCODE_ID_MAXIMUM; + } for (i = 0; i < adev->firmware.max_ucodes; i++) { ucode = &adev->firmware.ucode[i]; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c index c853400805d1..735c38d7db0d 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c @@ -955,11 +955,11 @@ int amdgpu_vce_ring_test_ring(struct amdgpu_ring *ring) struct amdgpu_device *adev = ring->adev; uint32_t rptr = amdgpu_ring_get_rptr(ring); unsigned i; - int r; + int r, timeout = adev->usec_timeout; - /* TODO: remove it if VCE can work for sriov */ + /* workaround VCE ring test slow issue for sriov*/ if (amdgpu_sriov_vf(adev)) - return 0; + timeout *= 10; r = amdgpu_ring_alloc(ring, 16); if (r) { @@ -970,13 +970,13 @@ int amdgpu_vce_ring_test_ring(struct amdgpu_ring *ring) amdgpu_ring_write(ring, VCE_CMD_END); amdgpu_ring_commit(ring); - for (i = 0; i < adev->usec_timeout; i++) { + for (i = 0; i < timeout; i++) { if (amdgpu_ring_get_rptr(ring) != rptr) break; DRM_UDELAY(1); } - if (i < adev->usec_timeout) { + if (i < timeout) { DRM_INFO("ring test on %d succeeded in %d usecs\n", ring->idx, i); } else { @@ -999,10 +999,6 @@ int amdgpu_vce_ring_test_ib(struct amdgpu_ring *ring, long timeout) struct dma_fence *fence = NULL; long r; - /* TODO: remove it if VCE can work for sriov */ - if (amdgpu_sriov_vf(ring->adev)) - return 0; - /* skip vce ring1/2 ib test for now, since it's not reliable */ if (ring != &ring->adev->vce.ring[0]) return 0; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c index ba8b8ae6234f..6bf5cea294f2 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c @@ -225,3 +225,49 @@ int amdgpu_virt_reset_gpu(struct amdgpu_device *adev) return 0; } + +/** + * amdgpu_virt_alloc_mm_table() - alloc memory for mm table + * @amdgpu: amdgpu device. + * MM table is used by UVD and VCE for its initialization + * Return: Zero if allocate success. + */ +int amdgpu_virt_alloc_mm_table(struct amdgpu_device *adev) +{ + int r; + + if (!amdgpu_sriov_vf(adev) || adev->virt.mm_table.gpu_addr) + return 0; + + r = amdgpu_bo_create_kernel(adev, PAGE_SIZE, PAGE_SIZE, + AMDGPU_GEM_DOMAIN_VRAM, + &adev->virt.mm_table.bo, + &adev->virt.mm_table.gpu_addr, + (void *)&adev->virt.mm_table.cpu_addr); + if (r) { + DRM_ERROR("failed to alloc mm table and error = %d.\n", r); + return r; + } + + memset((void *)adev->virt.mm_table.cpu_addr, 0, PAGE_SIZE); + DRM_INFO("MM table gpu addr = 0x%llx, cpu addr = %p.\n", + adev->virt.mm_table.gpu_addr, + adev->virt.mm_table.cpu_addr); + return 0; +} + +/** + * amdgpu_virt_free_mm_table() - free mm table memory + * @amdgpu: amdgpu device. + * Free MM table memory + */ +void amdgpu_virt_free_mm_table(struct amdgpu_device *adev) +{ + if (!amdgpu_sriov_vf(adev) || !adev->virt.mm_table.gpu_addr) + return; + + amdgpu_bo_free_kernel(&adev->virt.mm_table.bo, + &adev->virt.mm_table.gpu_addr, + (void *)&adev->virt.mm_table.cpu_addr); + adev->virt.mm_table.gpu_addr = 0; +} diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.h index 1ee0a190b33b..a8ed162cc0bc 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.h @@ -98,5 +98,7 @@ int amdgpu_virt_request_full_gpu(struct amdgpu_device *adev, bool init); int amdgpu_virt_release_full_gpu(struct amdgpu_device *adev, bool init); int amdgpu_virt_reset_gpu(struct amdgpu_device *adev); int amdgpu_sriov_gpu_reset(struct amdgpu_device *adev, bool voluntary); +int amdgpu_virt_alloc_mm_table(struct amdgpu_device *adev); +void amdgpu_virt_free_mm_table(struct amdgpu_device *adev); #endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c index 7ed5302b511a..749a6cde7985 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c @@ -279,8 +279,9 @@ static int amdgpu_vm_alloc_levels(struct amdgpu_device *adev, if (!parent->entries) { unsigned num_entries = amdgpu_vm_num_entries(adev, level); - parent->entries = drm_calloc_large(num_entries, - sizeof(struct amdgpu_vm_pt)); + parent->entries = kvmalloc_array(num_entries, + sizeof(struct amdgpu_vm_pt), + GFP_KERNEL | __GFP_ZERO); if (!parent->entries) return -ENOMEM; memset(parent->entries, 0 , sizeof(struct amdgpu_vm_pt)); @@ -406,6 +407,8 @@ int amdgpu_vm_grab_id(struct amdgpu_vm *vm, struct amdgpu_ring *ring, struct amdgpu_job *job) { struct amdgpu_device *adev = ring->adev; + unsigned vmhub = ring->funcs->vmhub; + struct amdgpu_vm_id_manager *id_mgr = &adev->vm_manager.id_mgr[vmhub]; uint64_t fence_context = adev->fence_context + ring->idx; struct dma_fence *updates = sync->last_vm_update; struct amdgpu_vm_id *id, *idle; @@ -413,16 +416,15 @@ int amdgpu_vm_grab_id(struct amdgpu_vm *vm, struct amdgpu_ring *ring, unsigned i; int r = 0; - fences = kmalloc_array(sizeof(void *), adev->vm_manager.num_ids, - GFP_KERNEL); + fences = kmalloc_array(sizeof(void *), id_mgr->num_ids, GFP_KERNEL); if (!fences) return -ENOMEM; - mutex_lock(&adev->vm_manager.lock); + mutex_lock(&id_mgr->lock); /* Check if we have an idle VMID */ i = 0; - list_for_each_entry(idle, &adev->vm_manager.ids_lru, list) { + list_for_each_entry(idle, &id_mgr->ids_lru, list) { fences[i] = amdgpu_sync_peek_fence(&idle->active, ring); if (!fences[i]) break; @@ -430,7 +432,7 @@ int amdgpu_vm_grab_id(struct amdgpu_vm *vm, struct amdgpu_ring *ring, } /* If we can't find a idle VMID to use, wait till one becomes available */ - if (&idle->list == &adev->vm_manager.ids_lru) { + if (&idle->list == &id_mgr->ids_lru) { u64 fence_context = adev->vm_manager.fence_context + ring->idx; unsigned seqno = ++adev->vm_manager.seqno[ring->idx]; struct dma_fence_array *array; @@ -455,25 +457,19 @@ int amdgpu_vm_grab_id(struct amdgpu_vm *vm, struct amdgpu_ring *ring, if (r) goto error; - mutex_unlock(&adev->vm_manager.lock); + mutex_unlock(&id_mgr->lock); return 0; } kfree(fences); - job->vm_needs_flush = true; + job->vm_needs_flush = false; /* Check if we can use a VMID already assigned to this VM */ - i = ring->idx; - do { + list_for_each_entry_reverse(id, &id_mgr->ids_lru, list) { struct dma_fence *flushed; - - id = vm->ids[i++]; - if (i == AMDGPU_MAX_RINGS) - i = 0; + bool needs_flush = false; /* Check all the prerequisites to using this VMID */ - if (!id) - continue; if (amdgpu_vm_had_gpu_reset(adev, id)) continue; @@ -483,16 +479,17 @@ int amdgpu_vm_grab_id(struct amdgpu_vm *vm, struct amdgpu_ring *ring, if (job->vm_pd_addr != id->pd_gpu_addr) continue; - if (!id->last_flush) - continue; - - if (id->last_flush->context != fence_context && - !dma_fence_is_signaled(id->last_flush)) - continue; + if (!id->last_flush || + (id->last_flush->context != fence_context && + !dma_fence_is_signaled(id->last_flush))) + needs_flush = true; flushed = id->flushed_updates; - if (updates && - (!flushed || dma_fence_is_later(updates, flushed))) + if (updates && (!flushed || dma_fence_is_later(updates, flushed))) + needs_flush = true; + + /* Concurrent flushes are only possible starting with Vega10 */ + if (adev->asic_type < CHIP_VEGA10 && needs_flush) continue; /* Good we can use this VMID. Remember this submission as @@ -502,17 +499,17 @@ int amdgpu_vm_grab_id(struct amdgpu_vm *vm, struct amdgpu_ring *ring, if (r) goto error; - list_move_tail(&id->list, &adev->vm_manager.ids_lru); - vm->ids[ring->idx] = id; - - job->vm_id = id - adev->vm_manager.ids; - job->vm_needs_flush = false; - trace_amdgpu_vm_grab_id(vm, ring->idx, job); + if (updates && (!flushed || dma_fence_is_later(updates, flushed))) { + dma_fence_put(id->flushed_updates); + id->flushed_updates = dma_fence_get(updates); + } - mutex_unlock(&adev->vm_manager.lock); - return 0; + if (needs_flush) + goto needs_flush; + else + goto no_flush_needed; - } while (i != ring->idx); + }; /* Still no ID to use? Then use the idle one found earlier */ id = idle; @@ -522,23 +519,25 @@ int amdgpu_vm_grab_id(struct amdgpu_vm *vm, struct amdgpu_ring *ring, if (r) goto error; - dma_fence_put(id->last_flush); - id->last_flush = NULL; - + id->pd_gpu_addr = job->vm_pd_addr; dma_fence_put(id->flushed_updates); id->flushed_updates = dma_fence_get(updates); - - id->pd_gpu_addr = job->vm_pd_addr; id->current_gpu_reset_count = atomic_read(&adev->gpu_reset_counter); - list_move_tail(&id->list, &adev->vm_manager.ids_lru); atomic64_set(&id->owner, vm->client_id); - vm->ids[ring->idx] = id; - job->vm_id = id - adev->vm_manager.ids; - trace_amdgpu_vm_grab_id(vm, ring->idx, job); +needs_flush: + job->vm_needs_flush = true; + dma_fence_put(id->last_flush); + id->last_flush = NULL; + +no_flush_needed: + list_move_tail(&id->list, &id_mgr->ids_lru); + + job->vm_id = id - id_mgr->ids; + trace_amdgpu_vm_grab_id(vm, ring, job); error: - mutex_unlock(&adev->vm_manager.lock); + mutex_unlock(&id_mgr->lock); return r; } @@ -590,7 +589,9 @@ static u64 amdgpu_vm_adjust_mc_addr(struct amdgpu_device *adev, u64 mc_addr) int amdgpu_vm_flush(struct amdgpu_ring *ring, struct amdgpu_job *job) { struct amdgpu_device *adev = ring->adev; - struct amdgpu_vm_id *id = &adev->vm_manager.ids[job->vm_id]; + unsigned vmhub = ring->funcs->vmhub; + struct amdgpu_vm_id_manager *id_mgr = &adev->vm_manager.id_mgr[vmhub]; + struct amdgpu_vm_id *id = &id_mgr->ids[job->vm_id]; bool gds_switch_needed = ring->funcs->emit_gds_switch && ( id->gds_base != job->gds_base || id->gds_size != job->gds_size || @@ -614,24 +615,24 @@ int amdgpu_vm_flush(struct amdgpu_ring *ring, struct amdgpu_job *job) if (ring->funcs->init_cond_exec) patch_offset = amdgpu_ring_init_cond_exec(ring); - if (ring->funcs->emit_pipeline_sync) + if (ring->funcs->emit_pipeline_sync && !job->need_pipeline_sync) amdgpu_ring_emit_pipeline_sync(ring); if (ring->funcs->emit_vm_flush && vm_flush_needed) { u64 pd_addr = amdgpu_vm_adjust_mc_addr(adev, job->vm_pd_addr); struct dma_fence *fence; - trace_amdgpu_vm_flush(pd_addr, ring->idx, job->vm_id); + trace_amdgpu_vm_flush(ring, job->vm_id, pd_addr); amdgpu_ring_emit_vm_flush(ring, job->vm_id, pd_addr); r = amdgpu_fence_emit(ring, &fence); if (r) return r; - mutex_lock(&adev->vm_manager.lock); + mutex_lock(&id_mgr->lock); dma_fence_put(id->last_flush); id->last_flush = fence; - mutex_unlock(&adev->vm_manager.lock); + mutex_unlock(&id_mgr->lock); } if (gds_switch_needed) { @@ -666,9 +667,11 @@ int amdgpu_vm_flush(struct amdgpu_ring *ring, struct amdgpu_job *job) * * Reset saved GDW, GWS and OA to force switch on next flush. */ -void amdgpu_vm_reset_id(struct amdgpu_device *adev, unsigned vm_id) +void amdgpu_vm_reset_id(struct amdgpu_device *adev, unsigned vmhub, + unsigned vmid) { - struct amdgpu_vm_id *id = &adev->vm_manager.ids[vm_id]; + struct amdgpu_vm_id_manager *id_mgr = &adev->vm_manager.id_mgr[vmhub]; + struct amdgpu_vm_id *id = &id_mgr->ids[vmid]; id->gds_base = 0; id->gds_size = 0; @@ -1336,6 +1339,12 @@ static int amdgpu_vm_bo_split_mapping(struct amdgpu_device *adev, flags &= ~AMDGPU_PTE_MTYPE_MASK; flags |= (mapping->flags & AMDGPU_PTE_MTYPE_MASK); + if ((mapping->flags & AMDGPU_PTE_PRT) && + (adev->asic_type >= CHIP_VEGA10)) { + flags |= AMDGPU_PTE_PRT; + flags &= ~AMDGPU_PTE_VALID; + } + trace_amdgpu_vm_bo_update(mapping); pfn = mapping->offset >> PAGE_SHIFT; @@ -1629,8 +1638,9 @@ int amdgpu_vm_clear_freed(struct amdgpu_device *adev, struct amdgpu_bo_va_mapping, list); list_del(&mapping->list); - r = amdgpu_vm_bo_split_mapping(adev, NULL, 0, NULL, vm, mapping, - 0, 0, &f); + r = amdgpu_vm_bo_update_mapping(adev, NULL, 0, NULL, vm, + mapping->start, mapping->last, + 0, 0, &f); amdgpu_vm_free_mapping(adev, vm, mapping, f); if (r) { dma_fence_put(f); @@ -2117,10 +2127,8 @@ int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm) unsigned ring_instance; struct amdgpu_ring *ring; struct amd_sched_rq *rq; - int i, r; + int r; - for (i = 0; i < AMDGPU_MAX_RINGS; ++i) - vm->ids[i] = NULL; vm->va = RB_ROOT; vm->client_id = atomic64_inc_return(&adev->vm_manager.client_counter); spin_lock_init(&vm->status_lock); @@ -2191,7 +2199,7 @@ static void amdgpu_vm_free_levels(struct amdgpu_vm_pt *level) for (i = 0; i <= level->last_entry_used; i++) amdgpu_vm_free_levels(&level->entries[i]); - drm_free_large(level->entries); + kvfree(level->entries); } /** @@ -2241,16 +2249,21 @@ void amdgpu_vm_fini(struct amdgpu_device *adev, struct amdgpu_vm *vm) */ void amdgpu_vm_manager_init(struct amdgpu_device *adev) { - unsigned i; + unsigned i, j; + + for (i = 0; i < AMDGPU_MAX_VMHUBS; ++i) { + struct amdgpu_vm_id_manager *id_mgr = + &adev->vm_manager.id_mgr[i]; - INIT_LIST_HEAD(&adev->vm_manager.ids_lru); + mutex_init(&id_mgr->lock); + INIT_LIST_HEAD(&id_mgr->ids_lru); - /* skip over VMID 0, since it is the system VM */ - for (i = 1; i < adev->vm_manager.num_ids; ++i) { - amdgpu_vm_reset_id(adev, i); - amdgpu_sync_create(&adev->vm_manager.ids[i].active); - list_add_tail(&adev->vm_manager.ids[i].list, - &adev->vm_manager.ids_lru); + /* skip over VMID 0, since it is the system VM */ + for (j = 1; j < id_mgr->num_ids; ++j) { + amdgpu_vm_reset_id(adev, i, j); + amdgpu_sync_create(&id_mgr->ids[i].active); + list_add_tail(&id_mgr->ids[j].list, &id_mgr->ids_lru); + } } adev->vm_manager.fence_context = @@ -2258,6 +2271,7 @@ void amdgpu_vm_manager_init(struct amdgpu_device *adev) for (i = 0; i < AMDGPU_MAX_RINGS; ++i) adev->vm_manager.seqno[i] = 0; + atomic_set(&adev->vm_manager.vm_pte_next_ring, 0); atomic64_set(&adev->vm_manager.client_counter, 0); spin_lock_init(&adev->vm_manager.prt_lock); @@ -2273,13 +2287,19 @@ void amdgpu_vm_manager_init(struct amdgpu_device *adev) */ void amdgpu_vm_manager_fini(struct amdgpu_device *adev) { - unsigned i; + unsigned i, j; - for (i = 0; i < AMDGPU_NUM_VM; ++i) { - struct amdgpu_vm_id *id = &adev->vm_manager.ids[i]; + for (i = 0; i < AMDGPU_MAX_VMHUBS; ++i) { + struct amdgpu_vm_id_manager *id_mgr = + &adev->vm_manager.id_mgr[i]; - amdgpu_sync_free(&adev->vm_manager.ids[i].active); - dma_fence_put(id->flushed_updates); - dma_fence_put(id->last_flush); + mutex_destroy(&id_mgr->lock); + for (j = 0; j < AMDGPU_NUM_VM; ++j) { + struct amdgpu_vm_id *id = &id_mgr->ids[j]; + + amdgpu_sync_free(&id->active); + dma_fence_put(id->flushed_updates); + dma_fence_put(id->last_flush); + } } } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h index d9e57290dc71..d97e28b4bdc4 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h @@ -65,7 +65,8 @@ struct amdgpu_bo_list_entry; #define AMDGPU_PTE_FRAG(x) ((x & 0x1fULL) << 7) -#define AMDGPU_PTE_PRT (1ULL << 63) +/* TILED for VEGA10, reserved for older ASICs */ +#define AMDGPU_PTE_PRT (1ULL << 51) /* VEGA10 only */ #define AMDGPU_PTE_MTYPE(a) ((uint64_t)a << 57) @@ -114,9 +115,6 @@ struct amdgpu_vm { struct dma_fence *last_dir_update; uint64_t last_eviction_counter; - /* for id and flush management per ring */ - struct amdgpu_vm_id *ids[AMDGPU_MAX_RINGS]; - /* protecting freed */ spinlock_t freed_lock; @@ -149,12 +147,16 @@ struct amdgpu_vm_id { uint32_t oa_size; }; +struct amdgpu_vm_id_manager { + struct mutex lock; + unsigned num_ids; + struct list_head ids_lru; + struct amdgpu_vm_id ids[AMDGPU_NUM_VM]; +}; + struct amdgpu_vm_manager { /* Handling of VMIDs */ - struct mutex lock; - unsigned num_ids; - struct list_head ids_lru; - struct amdgpu_vm_id ids[AMDGPU_NUM_VM]; + struct amdgpu_vm_id_manager id_mgr[AMDGPU_MAX_VMHUBS]; /* Handling of VM fences */ u64 fence_context; @@ -200,7 +202,8 @@ int amdgpu_vm_grab_id(struct amdgpu_vm *vm, struct amdgpu_ring *ring, struct amdgpu_sync *sync, struct dma_fence *fence, struct amdgpu_job *job); int amdgpu_vm_flush(struct amdgpu_ring *ring, struct amdgpu_job *job); -void amdgpu_vm_reset_id(struct amdgpu_device *adev, unsigned vm_id); +void amdgpu_vm_reset_id(struct amdgpu_device *adev, unsigned vmhub, + unsigned vmid); int amdgpu_vm_update_directories(struct amdgpu_device *adev, struct amdgpu_vm *vm); int amdgpu_vm_clear_freed(struct amdgpu_device *adev, diff --git a/drivers/gpu/drm/amd/amdgpu/ci_dpm.c b/drivers/gpu/drm/amd/amdgpu/ci_dpm.c index 11ccda83d767..00e56a28b593 100644 --- a/drivers/gpu/drm/amd/amdgpu/ci_dpm.c +++ b/drivers/gpu/drm/amd/amdgpu/ci_dpm.c @@ -22,7 +22,7 @@ */ #include <linux/firmware.h> -#include "drmP.h" +#include <drm/drmP.h> #include "amdgpu.h" #include "amdgpu_pm.h" #include "amdgpu_ucode.h" @@ -1267,30 +1267,33 @@ static int ci_dpm_set_fan_speed_percent(struct amdgpu_device *adev, static void ci_dpm_set_fan_control_mode(struct amdgpu_device *adev, u32 mode) { - if (mode) { - /* stop auto-manage */ + switch (mode) { + case AMD_FAN_CTRL_NONE: if (adev->pm.dpm.fan.ucode_fan_control) ci_fan_ctrl_stop_smc_fan_control(adev); - ci_fan_ctrl_set_static_mode(adev, mode); - } else { - /* restart auto-manage */ + ci_dpm_set_fan_speed_percent(adev, 100); + break; + case AMD_FAN_CTRL_MANUAL: + if (adev->pm.dpm.fan.ucode_fan_control) + ci_fan_ctrl_stop_smc_fan_control(adev); + break; + case AMD_FAN_CTRL_AUTO: if (adev->pm.dpm.fan.ucode_fan_control) ci_thermal_start_smc_fan_control(adev); - else - ci_fan_ctrl_set_default_mode(adev); + break; + default: + break; } } static u32 ci_dpm_get_fan_control_mode(struct amdgpu_device *adev) { struct ci_power_info *pi = ci_get_pi(adev); - u32 tmp; if (pi->fan_is_controlled_by_smc) - return 0; - - tmp = RREG32_SMC(ixCG_FDO_CTRL2) & CG_FDO_CTRL2__FDO_PWM_MODE_MASK; - return (tmp >> CG_FDO_CTRL2__FDO_PWM_MODE__SHIFT); + return AMD_FAN_CTRL_AUTO; + else + return AMD_FAN_CTRL_MANUAL; } #if 0 @@ -3036,6 +3039,7 @@ static int ci_populate_single_memory_level(struct amdgpu_device *adev, memory_clock, &memory_level->MinVddcPhases); + memory_level->EnabledForActivity = 1; memory_level->EnabledForThrottle = 1; memory_level->UpH = 0; memory_level->DownH = 100; @@ -3468,8 +3472,6 @@ static int ci_populate_all_memory_levels(struct amdgpu_device *adev) return ret; } - pi->smc_state_table.MemoryLevel[0].EnabledForActivity = 1; - if ((dpm_table->mclk_table.count >= 2) && ((adev->pdev->device == 0x67B0) || (adev->pdev->device == 0x67B1))) { pi->smc_state_table.MemoryLevel[1].MinVddc = diff --git a/drivers/gpu/drm/amd/amdgpu/ci_smc.c b/drivers/gpu/drm/amd/amdgpu/ci_smc.c index 7eb9069db8e3..b8ba51e045b5 100644 --- a/drivers/gpu/drm/amd/amdgpu/ci_smc.c +++ b/drivers/gpu/drm/amd/amdgpu/ci_smc.c @@ -23,7 +23,7 @@ */ #include <linux/firmware.h> -#include "drmP.h" +#include <drm/drmP.h> #include "amdgpu.h" #include "cikd.h" #include "ppsmc.h" diff --git a/drivers/gpu/drm/amd/amdgpu/cik.c b/drivers/gpu/drm/amd/amdgpu/cik.c index 9d33e5641419..6b2034533f68 100644 --- a/drivers/gpu/drm/amd/amdgpu/cik.c +++ b/drivers/gpu/drm/amd/amdgpu/cik.c @@ -24,7 +24,7 @@ #include <linux/firmware.h> #include <linux/slab.h> #include <linux/module.h> -#include "drmP.h" +#include <drm/drmP.h> #include "amdgpu.h" #include "amdgpu_atombios.h" #include "amdgpu_ih.h" diff --git a/drivers/gpu/drm/amd/amdgpu/cik_ih.c b/drivers/gpu/drm/amd/amdgpu/cik_ih.c index c57c3f18af01..b8918432c572 100644 --- a/drivers/gpu/drm/amd/amdgpu/cik_ih.c +++ b/drivers/gpu/drm/amd/amdgpu/cik_ih.c @@ -20,7 +20,7 @@ * OTHER DEALINGS IN THE SOFTWARE. * */ -#include "drmP.h" +#include <drm/drmP.h> #include "amdgpu.h" #include "amdgpu_ih.h" #include "cikd.h" diff --git a/drivers/gpu/drm/amd/amdgpu/cz_ih.c b/drivers/gpu/drm/amd/amdgpu/cz_ih.c index a5f294ebff5c..0c1209cdd1cb 100644 --- a/drivers/gpu/drm/amd/amdgpu/cz_ih.c +++ b/drivers/gpu/drm/amd/amdgpu/cz_ih.c @@ -20,7 +20,7 @@ * OTHER DEALINGS IN THE SOFTWARE. * */ -#include "drmP.h" +#include <drm/drmP.h> #include "amdgpu.h" #include "amdgpu_ih.h" #include "vid.h" diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c index ba98d35340a3..3c62c45f43a1 100644 --- a/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c +++ b/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c @@ -20,7 +20,7 @@ * OTHER DEALINGS IN THE SOFTWARE. * */ -#include "drmP.h" +#include <drm/drmP.h> #include "amdgpu.h" #include "amdgpu_pm.h" #include "amdgpu_i2c.h" @@ -2230,7 +2230,7 @@ static int dce_v10_0_crtc_do_set_base(struct drm_crtc *crtc, if (!atomic && fb && fb != crtc->primary->fb) { amdgpu_fb = to_amdgpu_framebuffer(fb); abo = gem_to_amdgpu_bo(amdgpu_fb->obj); - r = amdgpu_bo_reserve(abo, false); + r = amdgpu_bo_reserve(abo, true); if (unlikely(r != 0)) return r; amdgpu_bo_unpin(abo); @@ -2589,7 +2589,7 @@ static int dce_v10_0_crtc_cursor_set2(struct drm_crtc *crtc, unpin: if (amdgpu_crtc->cursor_bo) { struct amdgpu_bo *aobj = gem_to_amdgpu_bo(amdgpu_crtc->cursor_bo); - ret = amdgpu_bo_reserve(aobj, false); + ret = amdgpu_bo_reserve(aobj, true); if (likely(ret == 0)) { amdgpu_bo_unpin(aobj); amdgpu_bo_unreserve(aobj); @@ -2720,7 +2720,7 @@ static void dce_v10_0_crtc_disable(struct drm_crtc *crtc) amdgpu_fb = to_amdgpu_framebuffer(crtc->primary->fb); abo = gem_to_amdgpu_bo(amdgpu_fb->obj); - r = amdgpu_bo_reserve(abo, false); + r = amdgpu_bo_reserve(abo, true); if (unlikely(r)) DRM_ERROR("failed to reserve abo before unpin\n"); else { diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c index e59bc42df18c..c8ed0facddcd 100644 --- a/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c +++ b/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c @@ -20,7 +20,7 @@ * OTHER DEALINGS IN THE SOFTWARE. * */ -#include "drmP.h" +#include <drm/drmP.h> #include "amdgpu.h" #include "amdgpu_pm.h" #include "amdgpu_i2c.h" @@ -2214,7 +2214,7 @@ static int dce_v11_0_crtc_do_set_base(struct drm_crtc *crtc, if (!atomic && fb && fb != crtc->primary->fb) { amdgpu_fb = to_amdgpu_framebuffer(fb); abo = gem_to_amdgpu_bo(amdgpu_fb->obj); - r = amdgpu_bo_reserve(abo, false); + r = amdgpu_bo_reserve(abo, true); if (unlikely(r != 0)) return r; amdgpu_bo_unpin(abo); @@ -2609,7 +2609,7 @@ static int dce_v11_0_crtc_cursor_set2(struct drm_crtc *crtc, unpin: if (amdgpu_crtc->cursor_bo) { struct amdgpu_bo *aobj = gem_to_amdgpu_bo(amdgpu_crtc->cursor_bo); - ret = amdgpu_bo_reserve(aobj, false); + ret = amdgpu_bo_reserve(aobj, true); if (likely(ret == 0)) { amdgpu_bo_unpin(aobj); amdgpu_bo_unreserve(aobj); @@ -2740,7 +2740,7 @@ static void dce_v11_0_crtc_disable(struct drm_crtc *crtc) amdgpu_fb = to_amdgpu_framebuffer(crtc->primary->fb); abo = gem_to_amdgpu_bo(amdgpu_fb->obj); - r = amdgpu_bo_reserve(abo, false); + r = amdgpu_bo_reserve(abo, true); if (unlikely(r)) DRM_ERROR("failed to reserve abo before unpin\n"); else { diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c index 307269bda4fa..3f3a25493327 100644 --- a/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c +++ b/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c @@ -20,7 +20,7 @@ * OTHER DEALINGS IN THE SOFTWARE. * */ -#include "drmP.h" +#include <drm/drmP.h> #include "amdgpu.h" #include "amdgpu_pm.h" #include "amdgpu_i2c.h" @@ -979,7 +979,7 @@ static void dce_v6_0_program_watermarks(struct amdgpu_device *adev, u32 priority_a_mark = 0, priority_b_mark = 0; u32 priority_a_cnt = PRIORITY_OFF; u32 priority_b_cnt = PRIORITY_OFF; - u32 tmp, arb_control3; + u32 tmp, arb_control3, lb_vblank_lead_lines = 0; fixed20_12 a, b, c; if (amdgpu_crtc->base.enabled && num_heads && mode) { @@ -1091,6 +1091,8 @@ static void dce_v6_0_program_watermarks(struct amdgpu_device *adev, c.full = dfixed_div(c, a); priority_b_mark = dfixed_trunc(c); priority_b_cnt |= priority_b_mark & PRIORITY_MARK_MASK; + + lb_vblank_lead_lines = DIV_ROUND_UP(lb_size, mode->crtc_hdisplay); } /* select wm A */ @@ -1120,6 +1122,9 @@ static void dce_v6_0_program_watermarks(struct amdgpu_device *adev, /* save values for DPM */ amdgpu_crtc->line_time = line_time; amdgpu_crtc->wm_high = latency_watermark_a; + + /* Save number of lines the linebuffer leads before the scanout */ + amdgpu_crtc->lb_vblank_lead_lines = lb_vblank_lead_lines; } /* watermark setup */ @@ -1640,7 +1645,7 @@ static int dce_v6_0_crtc_do_set_base(struct drm_crtc *crtc, if (!atomic && fb && fb != crtc->primary->fb) { amdgpu_fb = to_amdgpu_framebuffer(fb); abo = gem_to_amdgpu_bo(amdgpu_fb->obj); - r = amdgpu_bo_reserve(abo, false); + r = amdgpu_bo_reserve(abo, true); if (unlikely(r != 0)) return r; amdgpu_bo_unpin(abo); @@ -1957,7 +1962,7 @@ static int dce_v6_0_crtc_cursor_set2(struct drm_crtc *crtc, unpin: if (amdgpu_crtc->cursor_bo) { struct amdgpu_bo *aobj = gem_to_amdgpu_bo(amdgpu_crtc->cursor_bo); - ret = amdgpu_bo_reserve(aobj, false); + ret = amdgpu_bo_reserve(aobj, true); if (likely(ret == 0)) { amdgpu_bo_unpin(aobj); amdgpu_bo_unreserve(aobj); @@ -2083,7 +2088,7 @@ static void dce_v6_0_crtc_disable(struct drm_crtc *crtc) amdgpu_fb = to_amdgpu_framebuffer(crtc->primary->fb); abo = gem_to_amdgpu_bo(amdgpu_fb->obj); - r = amdgpu_bo_reserve(abo, false); + r = amdgpu_bo_reserve(abo, true); if (unlikely(r)) DRM_ERROR("failed to reserve abo before unpin\n"); else { diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c index 6df7a28e8aac..3e90c19b9c7f 100644 --- a/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c +++ b/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c @@ -20,7 +20,7 @@ * OTHER DEALINGS IN THE SOFTWARE. * */ -#include "drmP.h" +#include <drm/drmP.h> #include "amdgpu.h" #include "amdgpu_pm.h" #include "amdgpu_i2c.h" @@ -2089,7 +2089,7 @@ static int dce_v8_0_crtc_do_set_base(struct drm_crtc *crtc, if (!atomic && fb && fb != crtc->primary->fb) { amdgpu_fb = to_amdgpu_framebuffer(fb); abo = gem_to_amdgpu_bo(amdgpu_fb->obj); - r = amdgpu_bo_reserve(abo, false); + r = amdgpu_bo_reserve(abo, true); if (unlikely(r != 0)) return r; amdgpu_bo_unpin(abo); @@ -2440,7 +2440,7 @@ static int dce_v8_0_crtc_cursor_set2(struct drm_crtc *crtc, unpin: if (amdgpu_crtc->cursor_bo) { struct amdgpu_bo *aobj = gem_to_amdgpu_bo(amdgpu_crtc->cursor_bo); - ret = amdgpu_bo_reserve(aobj, false); + ret = amdgpu_bo_reserve(aobj, true); if (likely(ret == 0)) { amdgpu_bo_unpin(aobj); amdgpu_bo_unreserve(aobj); @@ -2571,7 +2571,7 @@ static void dce_v8_0_crtc_disable(struct drm_crtc *crtc) amdgpu_fb = to_amdgpu_framebuffer(crtc->primary->fb); abo = gem_to_amdgpu_bo(amdgpu_fb->obj); - r = amdgpu_bo_reserve(abo, false); + r = amdgpu_bo_reserve(abo, true); if (unlikely(r)) DRM_ERROR("failed to reserve abo before unpin\n"); else { diff --git a/drivers/gpu/drm/amd/amdgpu/dce_virtual.c b/drivers/gpu/drm/amd/amdgpu/dce_virtual.c index 81a24b6b4846..90bb08309a53 100644 --- a/drivers/gpu/drm/amd/amdgpu/dce_virtual.c +++ b/drivers/gpu/drm/amd/amdgpu/dce_virtual.c @@ -20,7 +20,7 @@ * OTHER DEALINGS IN THE SOFTWARE. * */ -#include "drmP.h" +#include <drm/drmP.h> #include "amdgpu.h" #include "amdgpu_pm.h" #include "amdgpu_i2c.h" @@ -248,7 +248,7 @@ static void dce_virtual_crtc_disable(struct drm_crtc *crtc) amdgpu_fb = to_amdgpu_framebuffer(crtc->primary->fb); abo = gem_to_amdgpu_bo(amdgpu_fb->obj); - r = amdgpu_bo_reserve(abo, false); + r = amdgpu_bo_reserve(abo, true); if (unlikely(r)) DRM_ERROR("failed to reserve abo before unpin\n"); else { diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c index 4c4874fdf59f..a125f9d44577 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c @@ -1579,7 +1579,7 @@ static void gfx_v6_0_setup_spi(struct amdgpu_device *adev) static void gfx_v6_0_config_init(struct amdgpu_device *adev) { - adev->gfx.config.double_offchip_lds_buf = 1; + adev->gfx.config.double_offchip_lds_buf = 0; } static void gfx_v6_0_gpu_init(struct amdgpu_device *adev) @@ -2437,7 +2437,7 @@ static void gfx_v6_0_rlc_fini(struct amdgpu_device *adev) int r; if (adev->gfx.rlc.save_restore_obj) { - r = amdgpu_bo_reserve(adev->gfx.rlc.save_restore_obj, false); + r = amdgpu_bo_reserve(adev->gfx.rlc.save_restore_obj, true); if (unlikely(r != 0)) dev_warn(adev->dev, "(%d) reserve RLC sr bo failed\n", r); amdgpu_bo_unpin(adev->gfx.rlc.save_restore_obj); @@ -2448,7 +2448,7 @@ static void gfx_v6_0_rlc_fini(struct amdgpu_device *adev) } if (adev->gfx.rlc.clear_state_obj) { - r = amdgpu_bo_reserve(adev->gfx.rlc.clear_state_obj, false); + r = amdgpu_bo_reserve(adev->gfx.rlc.clear_state_obj, true); if (unlikely(r != 0)) dev_warn(adev->dev, "(%d) reserve RLC c bo failed\n", r); amdgpu_bo_unpin(adev->gfx.rlc.clear_state_obj); @@ -2459,7 +2459,7 @@ static void gfx_v6_0_rlc_fini(struct amdgpu_device *adev) } if (adev->gfx.rlc.cp_table_obj) { - r = amdgpu_bo_reserve(adev->gfx.rlc.cp_table_obj, false); + r = amdgpu_bo_reserve(adev->gfx.rlc.cp_table_obj, true); if (unlikely(r != 0)) dev_warn(adev->dev, "(%d) reserve RLC cp table bo failed\n", r); amdgpu_bo_unpin(adev->gfx.rlc.cp_table_obj); @@ -3292,7 +3292,7 @@ static int gfx_v6_0_sw_init(void *handle) ring->me = 1; ring->pipe = i; ring->queue = i; - sprintf(ring->name, "comp %d.%d.%d", ring->me, ring->pipe, ring->queue); + sprintf(ring->name, "comp_%d.%d.%d", ring->me, ring->pipe, ring->queue); irq_type = AMDGPU_CP_IRQ_COMPUTE_MEC1_PIPE0_EOP + ring->pipe; r = amdgpu_ring_init(adev, ring, 1024, &adev->gfx.eop_irq, irq_type); diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c index 8a8bc2fe6f2e..f7414cabd4ff 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c @@ -21,7 +21,7 @@ * */ #include <linux/firmware.h> -#include "drmP.h" +#include <drm/drmP.h> #include "amdgpu.h" #include "amdgpu_ih.h" #include "amdgpu_gfx.h" @@ -1935,7 +1935,7 @@ static void gfx_v7_0_gpu_init(struct amdgpu_device *adev) INDEX_STRIDE, 3); mutex_lock(&adev->srbm_mutex); - for (i = 0; i < adev->vm_manager.num_ids; i++) { + for (i = 0; i < adev->vm_manager.id_mgr[0].num_ids; i++) { if (i == 0) sh_mem_base = 0; else @@ -2792,7 +2792,7 @@ static void gfx_v7_0_cp_compute_fini(struct amdgpu_device *adev) struct amdgpu_ring *ring = &adev->gfx.compute_ring[i]; if (ring->mqd_obj) { - r = amdgpu_bo_reserve(ring->mqd_obj, false); + r = amdgpu_bo_reserve(ring->mqd_obj, true); if (unlikely(r != 0)) dev_warn(adev->dev, "(%d) reserve MQD bo failed\n", r); @@ -2810,7 +2810,7 @@ static void gfx_v7_0_mec_fini(struct amdgpu_device *adev) int r; if (adev->gfx.mec.hpd_eop_obj) { - r = amdgpu_bo_reserve(adev->gfx.mec.hpd_eop_obj, false); + r = amdgpu_bo_reserve(adev->gfx.mec.hpd_eop_obj, true); if (unlikely(r != 0)) dev_warn(adev->dev, "(%d) reserve HPD EOP bo failed\n", r); amdgpu_bo_unpin(adev->gfx.mec.hpd_eop_obj); @@ -3359,7 +3359,7 @@ static void gfx_v7_0_rlc_fini(struct amdgpu_device *adev) /* save restore block */ if (adev->gfx.rlc.save_restore_obj) { - r = amdgpu_bo_reserve(adev->gfx.rlc.save_restore_obj, false); + r = amdgpu_bo_reserve(adev->gfx.rlc.save_restore_obj, true); if (unlikely(r != 0)) dev_warn(adev->dev, "(%d) reserve RLC sr bo failed\n", r); amdgpu_bo_unpin(adev->gfx.rlc.save_restore_obj); @@ -3371,7 +3371,7 @@ static void gfx_v7_0_rlc_fini(struct amdgpu_device *adev) /* clear state block */ if (adev->gfx.rlc.clear_state_obj) { - r = amdgpu_bo_reserve(adev->gfx.rlc.clear_state_obj, false); + r = amdgpu_bo_reserve(adev->gfx.rlc.clear_state_obj, true); if (unlikely(r != 0)) dev_warn(adev->dev, "(%d) reserve RLC c bo failed\n", r); amdgpu_bo_unpin(adev->gfx.rlc.clear_state_obj); @@ -3383,7 +3383,7 @@ static void gfx_v7_0_rlc_fini(struct amdgpu_device *adev) /* clear state block */ if (adev->gfx.rlc.cp_table_obj) { - r = amdgpu_bo_reserve(adev->gfx.rlc.cp_table_obj, false); + r = amdgpu_bo_reserve(adev->gfx.rlc.cp_table_obj, true); if (unlikely(r != 0)) dev_warn(adev->dev, "(%d) reserve RLC cp table bo failed\n", r); amdgpu_bo_unpin(adev->gfx.rlc.cp_table_obj); diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c index dad8a4cd1b37..404d12785853 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c @@ -21,7 +21,7 @@ * */ #include <linux/firmware.h> -#include "drmP.h" +#include <drm/drmP.h> #include "amdgpu.h" #include "amdgpu_gfx.h" #include "vi.h" @@ -1239,7 +1239,7 @@ static void gfx_v8_0_rlc_fini(struct amdgpu_device *adev) /* clear state block */ if (adev->gfx.rlc.clear_state_obj) { - r = amdgpu_bo_reserve(adev->gfx.rlc.clear_state_obj, false); + r = amdgpu_bo_reserve(adev->gfx.rlc.clear_state_obj, true); if (unlikely(r != 0)) dev_warn(adev->dev, "(%d) reserve RLC cbs bo failed\n", r); amdgpu_bo_unpin(adev->gfx.rlc.clear_state_obj); @@ -1250,7 +1250,7 @@ static void gfx_v8_0_rlc_fini(struct amdgpu_device *adev) /* jump table block */ if (adev->gfx.rlc.cp_table_obj) { - r = amdgpu_bo_reserve(adev->gfx.rlc.cp_table_obj, false); + r = amdgpu_bo_reserve(adev->gfx.rlc.cp_table_obj, true); if (unlikely(r != 0)) dev_warn(adev->dev, "(%d) reserve RLC cp table bo failed\n", r); amdgpu_bo_unpin(adev->gfx.rlc.cp_table_obj); @@ -1363,7 +1363,7 @@ static void gfx_v8_0_mec_fini(struct amdgpu_device *adev) int r; if (adev->gfx.mec.hpd_eop_obj) { - r = amdgpu_bo_reserve(adev->gfx.mec.hpd_eop_obj, false); + r = amdgpu_bo_reserve(adev->gfx.mec.hpd_eop_obj, true); if (unlikely(r != 0)) dev_warn(adev->dev, "(%d) reserve HPD EOP bo failed\n", r); amdgpu_bo_unpin(adev->gfx.mec.hpd_eop_obj); @@ -1490,7 +1490,7 @@ static int gfx_v8_0_kiq_init(struct amdgpu_device *adev) memset(hpd, 0, MEC_HPD_SIZE); - r = amdgpu_bo_reserve(kiq->eop_obj, false); + r = amdgpu_bo_reserve(kiq->eop_obj, true); if (unlikely(r != 0)) dev_warn(adev->dev, "(%d) reserve kiq eop bo failed\n", r); amdgpu_bo_kunmap(kiq->eop_obj); @@ -1932,6 +1932,7 @@ static int gfx_v8_0_gpu_early_init(struct amdgpu_device *adev) case 0xca: case 0xce: case 0x88: + case 0xe6: /* B6 */ adev->gfx.config.max_cu_per_sh = 6; break; @@ -1964,17 +1965,28 @@ static int gfx_v8_0_gpu_early_init(struct amdgpu_device *adev) adev->gfx.config.max_backends_per_se = 1; switch (adev->pdev->revision) { + case 0x80: + case 0x81: case 0xc0: case 0xc1: case 0xc2: case 0xc4: case 0xc8: case 0xc9: + case 0xd6: + case 0xda: + case 0xe9: + case 0xea: adev->gfx.config.max_cu_per_sh = 3; break; + case 0x83: case 0xd0: case 0xd1: case 0xd2: + case 0xd4: + case 0xdb: + case 0xe1: + case 0xe2: default: adev->gfx.config.max_cu_per_sh = 2; break; @@ -3890,7 +3902,7 @@ static void gfx_v8_0_gpu_init(struct amdgpu_device *adev) sh_static_mem_cfg = REG_SET_FIELD(sh_static_mem_cfg, SH_STATIC_MEM_CONFIG, INDEX_STRIDE, 3); mutex_lock(&adev->srbm_mutex); - for (i = 0; i < adev->vm_manager.num_ids; i++) { + for (i = 0; i < adev->vm_manager.id_mgr[0].num_ids; i++) { vi_srbm_select(adev, 0, 0, 0, i); /* CP and shaders */ if (i == 0) { diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c index a447b70841c9..125b11950071 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c @@ -21,7 +21,7 @@ * */ #include <linux/firmware.h> -#include "drmP.h" +#include <drm/drmP.h> #include "amdgpu.h" #include "amdgpu_gfx.h" #include "soc15.h" @@ -39,7 +39,6 @@ #define GFX9_NUM_GFX_RINGS 1 #define GFX9_NUM_COMPUTE_RINGS 8 -#define GFX9_NUM_SE 4 #define RLCG_UCODE_LOADING_START_ADDRESS 0x2000 MODULE_FIRMWARE("amdgpu/vega10_ce.bin"); @@ -453,7 +452,7 @@ static void gfx_v9_0_mec_fini(struct amdgpu_device *adev) int r; if (adev->gfx.mec.hpd_eop_obj) { - r = amdgpu_bo_reserve(adev->gfx.mec.hpd_eop_obj, false); + r = amdgpu_bo_reserve(adev->gfx.mec.hpd_eop_obj, true); if (unlikely(r != 0)) dev_warn(adev->dev, "(%d) reserve HPD EOP bo failed\n", r); amdgpu_bo_unpin(adev->gfx.mec.hpd_eop_obj); @@ -463,7 +462,7 @@ static void gfx_v9_0_mec_fini(struct amdgpu_device *adev) adev->gfx.mec.hpd_eop_obj = NULL; } if (adev->gfx.mec.mec_fw_obj) { - r = amdgpu_bo_reserve(adev->gfx.mec.mec_fw_obj, false); + r = amdgpu_bo_reserve(adev->gfx.mec.mec_fw_obj, true); if (unlikely(r != 0)) dev_warn(adev->dev, "(%d) reserve mec firmware bo failed\n", r); amdgpu_bo_unpin(adev->gfx.mec.mec_fw_obj); @@ -599,7 +598,7 @@ static int gfx_v9_0_kiq_init(struct amdgpu_device *adev) memset(hpd, 0, MEC_HPD_SIZE); - r = amdgpu_bo_reserve(kiq->eop_obj, false); + r = amdgpu_bo_reserve(kiq->eop_obj, true); if (unlikely(r != 0)) dev_warn(adev->dev, "(%d) reserve kiq eop bo failed\n", r); amdgpu_bo_kunmap(kiq->eop_obj); @@ -631,7 +630,6 @@ static int gfx_v9_0_kiq_init_ring(struct amdgpu_device *adev, ring->pipe = 1; } - irq->data = ring; ring->queue = 0; ring->eop_gpu_addr = kiq->eop_gpu_addr; sprintf(ring->name, "kiq %d.%d.%d", ring->me, ring->pipe, ring->queue); @@ -647,7 +645,6 @@ static void gfx_v9_0_kiq_free_ring(struct amdgpu_ring *ring, { amdgpu_wb_free(ring->adev, ring->adev->virt.reg_val_offs); amdgpu_ring_fini(ring); - irq->data = NULL; } /* create MQD for each compute queue */ @@ -705,19 +702,19 @@ static void gfx_v9_0_compute_mqd_sw_fini(struct amdgpu_device *adev) static uint32_t wave_read_ind(struct amdgpu_device *adev, uint32_t simd, uint32_t wave, uint32_t address) { - WREG32(SOC15_REG_OFFSET(GC, 0, mmSQ_IND_INDEX), + WREG32_SOC15(GC, 0, mmSQ_IND_INDEX, (wave << SQ_IND_INDEX__WAVE_ID__SHIFT) | (simd << SQ_IND_INDEX__SIMD_ID__SHIFT) | (address << SQ_IND_INDEX__INDEX__SHIFT) | (SQ_IND_INDEX__FORCE_READ_MASK)); - return RREG32(SOC15_REG_OFFSET(GC, 0, mmSQ_IND_DATA)); + return RREG32_SOC15(GC, 0, mmSQ_IND_DATA); } static void wave_read_regs(struct amdgpu_device *adev, uint32_t simd, uint32_t wave, uint32_t thread, uint32_t regno, uint32_t num, uint32_t *out) { - WREG32(SOC15_REG_OFFSET(GC, 0, mmSQ_IND_INDEX), + WREG32_SOC15(GC, 0, mmSQ_IND_INDEX, (wave << SQ_IND_INDEX__WAVE_ID__SHIFT) | (simd << SQ_IND_INDEX__SIMD_ID__SHIFT) | (regno << SQ_IND_INDEX__INDEX__SHIFT) | @@ -725,7 +722,7 @@ static void wave_read_regs(struct amdgpu_device *adev, uint32_t simd, (SQ_IND_INDEX__FORCE_READ_MASK) | (SQ_IND_INDEX__AUTO_INCR_MASK)); while (num--) - *(out++) = RREG32(SOC15_REG_OFFSET(GC, 0, mmSQ_IND_DATA)); + *(out++) = RREG32_SOC15(GC, 0, mmSQ_IND_DATA); } static void gfx_v9_0_read_wave_data(struct amdgpu_device *adev, uint32_t simd, uint32_t wave, uint32_t *dst, int *no_fields) @@ -774,7 +771,6 @@ static void gfx_v9_0_gpu_early_init(struct amdgpu_device *adev) switch (adev->asic_type) { case CHIP_VEGA10: adev->gfx.config.max_shader_engines = 4; - adev->gfx.config.max_tile_pipes = 8; //?? adev->gfx.config.max_cu_per_sh = 16; adev->gfx.config.max_sh_per_se = 1; adev->gfx.config.max_backends_per_se = 4; @@ -787,6 +783,8 @@ static void gfx_v9_0_gpu_early_init(struct amdgpu_device *adev) adev->gfx.config.sc_prim_fifo_size_backend = 0x100; adev->gfx.config.sc_hiz_tile_fifo_size = 0x30; adev->gfx.config.sc_earlyz_tile_fifo_size = 0x4C0; + adev->gfx.config.gs_vgt_table_depth = 32; + adev->gfx.config.gs_prim_buffer_depth = 1792; gb_addr_config = VEGA10_GB_ADDR_CONFIG_GOLDEN; break; default: @@ -801,6 +799,10 @@ static void gfx_v9_0_gpu_early_init(struct amdgpu_device *adev) adev->gfx.config.gb_addr_config, GB_ADDR_CONFIG, NUM_PIPES); + + adev->gfx.config.max_tile_pipes = + adev->gfx.config.gb_addr_config_fields.num_pipes; + adev->gfx.config.gb_addr_config_fields.num_banks = 1 << REG_GET_FIELD( adev->gfx.config.gb_addr_config, @@ -841,7 +843,7 @@ static int gfx_v9_0_ngg_create_buf(struct amdgpu_device *adev, } size_se = size_se ? size_se : default_size_se; - ngg_buf->size = size_se * GFX9_NUM_SE; + ngg_buf->size = size_se * adev->gfx.config.max_shader_engines; r = amdgpu_bo_create_kernel(adev, ngg_buf->size, PAGE_SIZE, AMDGPU_GEM_DOMAIN_VRAM, &ngg_buf->bo, @@ -888,7 +890,7 @@ static int gfx_v9_0_ngg_init(struct amdgpu_device *adev) adev->gfx.ngg.gds_reserve_addr += adev->gds.mem.gfx_partition_size; /* Primitive Buffer */ - r = gfx_v9_0_ngg_create_buf(adev, &adev->gfx.ngg.buf[PRIM], + r = gfx_v9_0_ngg_create_buf(adev, &adev->gfx.ngg.buf[NGG_PRIM], amdgpu_prim_buf_per_se, 64 * 1024); if (r) { @@ -897,7 +899,7 @@ static int gfx_v9_0_ngg_init(struct amdgpu_device *adev) } /* Position Buffer */ - r = gfx_v9_0_ngg_create_buf(adev, &adev->gfx.ngg.buf[POS], + r = gfx_v9_0_ngg_create_buf(adev, &adev->gfx.ngg.buf[NGG_POS], amdgpu_pos_buf_per_se, 256 * 1024); if (r) { @@ -906,7 +908,7 @@ static int gfx_v9_0_ngg_init(struct amdgpu_device *adev) } /* Control Sideband */ - r = gfx_v9_0_ngg_create_buf(adev, &adev->gfx.ngg.buf[CNTL], + r = gfx_v9_0_ngg_create_buf(adev, &adev->gfx.ngg.buf[NGG_CNTL], amdgpu_cntl_sb_buf_per_se, 256); if (r) { @@ -918,7 +920,7 @@ static int gfx_v9_0_ngg_init(struct amdgpu_device *adev) if (amdgpu_param_buf_per_se <= 0) goto out; - r = gfx_v9_0_ngg_create_buf(adev, &adev->gfx.ngg.buf[PARAM], + r = gfx_v9_0_ngg_create_buf(adev, &adev->gfx.ngg.buf[NGG_PARAM], amdgpu_param_buf_per_se, 512 * 1024); if (r) { @@ -947,47 +949,47 @@ static int gfx_v9_0_ngg_en(struct amdgpu_device *adev) /* Program buffer size */ data = 0; - size = adev->gfx.ngg.buf[PRIM].size / 256; + size = adev->gfx.ngg.buf[NGG_PRIM].size / 256; data = REG_SET_FIELD(data, WD_BUF_RESOURCE_1, INDEX_BUF_SIZE, size); - size = adev->gfx.ngg.buf[POS].size / 256; + size = adev->gfx.ngg.buf[NGG_POS].size / 256; data = REG_SET_FIELD(data, WD_BUF_RESOURCE_1, POS_BUF_SIZE, size); - WREG32(SOC15_REG_OFFSET(GC, 0, mmWD_BUF_RESOURCE_1), data); + WREG32_SOC15(GC, 0, mmWD_BUF_RESOURCE_1, data); data = 0; - size = adev->gfx.ngg.buf[CNTL].size / 256; + size = adev->gfx.ngg.buf[NGG_CNTL].size / 256; data = REG_SET_FIELD(data, WD_BUF_RESOURCE_2, CNTL_SB_BUF_SIZE, size); - size = adev->gfx.ngg.buf[PARAM].size / 1024; + size = adev->gfx.ngg.buf[NGG_PARAM].size / 1024; data = REG_SET_FIELD(data, WD_BUF_RESOURCE_2, PARAM_BUF_SIZE, size); - WREG32(SOC15_REG_OFFSET(GC, 0, mmWD_BUF_RESOURCE_2), data); + WREG32_SOC15(GC, 0, mmWD_BUF_RESOURCE_2, data); /* Program buffer base address */ - base = lower_32_bits(adev->gfx.ngg.buf[PRIM].gpu_addr); + base = lower_32_bits(adev->gfx.ngg.buf[NGG_PRIM].gpu_addr); data = REG_SET_FIELD(0, WD_INDEX_BUF_BASE, BASE, base); - WREG32(SOC15_REG_OFFSET(GC, 0, mmWD_INDEX_BUF_BASE), data); + WREG32_SOC15(GC, 0, mmWD_INDEX_BUF_BASE, data); - base = upper_32_bits(adev->gfx.ngg.buf[PRIM].gpu_addr); + base = upper_32_bits(adev->gfx.ngg.buf[NGG_PRIM].gpu_addr); data = REG_SET_FIELD(0, WD_INDEX_BUF_BASE_HI, BASE_HI, base); - WREG32(SOC15_REG_OFFSET(GC, 0, mmWD_INDEX_BUF_BASE_HI), data); + WREG32_SOC15(GC, 0, mmWD_INDEX_BUF_BASE_HI, data); - base = lower_32_bits(adev->gfx.ngg.buf[POS].gpu_addr); + base = lower_32_bits(adev->gfx.ngg.buf[NGG_POS].gpu_addr); data = REG_SET_FIELD(0, WD_POS_BUF_BASE, BASE, base); - WREG32(SOC15_REG_OFFSET(GC, 0, mmWD_POS_BUF_BASE), data); + WREG32_SOC15(GC, 0, mmWD_POS_BUF_BASE, data); - base = upper_32_bits(adev->gfx.ngg.buf[POS].gpu_addr); + base = upper_32_bits(adev->gfx.ngg.buf[NGG_POS].gpu_addr); data = REG_SET_FIELD(0, WD_POS_BUF_BASE_HI, BASE_HI, base); - WREG32(SOC15_REG_OFFSET(GC, 0, mmWD_POS_BUF_BASE_HI), data); + WREG32_SOC15(GC, 0, mmWD_POS_BUF_BASE_HI, data); - base = lower_32_bits(adev->gfx.ngg.buf[CNTL].gpu_addr); + base = lower_32_bits(adev->gfx.ngg.buf[NGG_CNTL].gpu_addr); data = REG_SET_FIELD(0, WD_CNTL_SB_BUF_BASE, BASE, base); - WREG32(SOC15_REG_OFFSET(GC, 0, mmWD_CNTL_SB_BUF_BASE), data); + WREG32_SOC15(GC, 0, mmWD_CNTL_SB_BUF_BASE, data); - base = upper_32_bits(adev->gfx.ngg.buf[CNTL].gpu_addr); + base = upper_32_bits(adev->gfx.ngg.buf[NGG_CNTL].gpu_addr); data = REG_SET_FIELD(0, WD_CNTL_SB_BUF_BASE_HI, BASE_HI, base); - WREG32(SOC15_REG_OFFSET(GC, 0, mmWD_CNTL_SB_BUF_BASE_HI), data); + WREG32_SOC15(GC, 0, mmWD_CNTL_SB_BUF_BASE_HI, data); /* Clear GDS reserved memory */ r = amdgpu_ring_alloc(ring, 17); @@ -1096,7 +1098,7 @@ static int gfx_v9_0_sw_init(void *handle) ring->pipe = i / 8; ring->queue = i % 8; ring->eop_gpu_addr = adev->gfx.mec.hpd_eop_gpu_addr + (i * MEC_HPD_SIZE); - sprintf(ring->name, "comp %d.%d.%d", ring->me, ring->pipe, ring->queue); + sprintf(ring->name, "comp_%d.%d.%d", ring->me, ring->pipe, ring->queue); irq_type = AMDGPU_CP_IRQ_COMPUTE_MEC1_PIPE0_EOP + ring->pipe; /* type-2 packets are deprecated on MEC, use type-3 instead */ r = amdgpu_ring_init(adev, ring, 1024, @@ -1203,7 +1205,7 @@ static void gfx_v9_0_select_se_sh(struct amdgpu_device *adev, u32 se_num, u32 sh data = REG_SET_FIELD(data, GRBM_GFX_INDEX, SH_INDEX, sh_num); data = REG_SET_FIELD(data, GRBM_GFX_INDEX, SE_INDEX, se_num); } - WREG32( SOC15_REG_OFFSET(GC, 0, mmGRBM_GFX_INDEX), data); + WREG32_SOC15(GC, 0, mmGRBM_GFX_INDEX, data); } static u32 gfx_v9_0_create_bitmask(u32 bit_width) @@ -1215,8 +1217,8 @@ static u32 gfx_v9_0_get_rb_active_bitmap(struct amdgpu_device *adev) { u32 data, mask; - data = RREG32(SOC15_REG_OFFSET(GC, 0, mmCC_RB_BACKEND_DISABLE)); - data |= RREG32(SOC15_REG_OFFSET(GC, 0, mmGC_USER_RB_BACKEND_DISABLE)); + data = RREG32_SOC15(GC, 0, mmCC_RB_BACKEND_DISABLE); + data |= RREG32_SOC15(GC, 0, mmGC_USER_RB_BACKEND_DISABLE); data &= CC_RB_BACKEND_DISABLE__BACKEND_DISABLE_MASK; data >>= GC_USER_RB_BACKEND_DISABLE__BACKEND_DISABLE__SHIFT; @@ -1276,8 +1278,8 @@ static void gfx_v9_0_init_compute_vmid(struct amdgpu_device *adev) for (i = FIRST_COMPUTE_VMID; i < LAST_COMPUTE_VMID; i++) { soc15_grbm_select(adev, 0, 0, 0, i); /* CP and shaders */ - WREG32(SOC15_REG_OFFSET(GC, 0, mmSH_MEM_CONFIG), sh_mem_config); - WREG32(SOC15_REG_OFFSET(GC, 0, mmSH_MEM_BASES), sh_mem_bases); + WREG32_SOC15(GC, 0, mmSH_MEM_CONFIG, sh_mem_config); + WREG32_SOC15(GC, 0, mmSH_MEM_BASES, sh_mem_bases); } soc15_grbm_select(adev, 0, 0, 0, 0); mutex_unlock(&adev->srbm_mutex); @@ -1304,8 +1306,8 @@ static void gfx_v9_0_gpu_init(struct amdgpu_device *adev) tmp = 0; tmp = REG_SET_FIELD(tmp, SH_MEM_CONFIG, ALIGNMENT_MODE, SH_MEM_ALIGNMENT_MODE_UNALIGNED); - WREG32(SOC15_REG_OFFSET(GC, 0, mmSH_MEM_CONFIG), tmp); - WREG32(SOC15_REG_OFFSET(GC, 0, mmSH_MEM_BASES), 0); + WREG32_SOC15(GC, 0, mmSH_MEM_CONFIG, tmp); + WREG32_SOC15(GC, 0, mmSH_MEM_BASES, 0); } soc15_grbm_select(adev, 0, 0, 0, 0); @@ -1320,7 +1322,7 @@ static void gfx_v9_0_gpu_init(struct amdgpu_device *adev) */ gfx_v9_0_select_se_sh(adev, 0xffffffff, 0xffffffff, 0xffffffff); - WREG32(SOC15_REG_OFFSET(GC, 0, mmPA_SC_FIFO_SIZE), + WREG32_SOC15(GC, 0, mmPA_SC_FIFO_SIZE, (adev->gfx.config.sc_prim_fifo_size_frontend << PA_SC_FIFO_SIZE__SC_FRONTEND_PRIM_FIFO_SIZE__SHIFT) | (adev->gfx.config.sc_prim_fifo_size_backend << @@ -1343,7 +1345,7 @@ static void gfx_v9_0_wait_for_rlc_serdes(struct amdgpu_device *adev) for (j = 0; j < adev->gfx.config.max_sh_per_se; j++) { gfx_v9_0_select_se_sh(adev, i, j, 0xffffffff); for (k = 0; k < adev->usec_timeout; k++) { - if (RREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_SERDES_CU_MASTER_BUSY)) == 0) + if (RREG32_SOC15(GC, 0, mmRLC_SERDES_CU_MASTER_BUSY) == 0) break; udelay(1); } @@ -1357,7 +1359,7 @@ static void gfx_v9_0_wait_for_rlc_serdes(struct amdgpu_device *adev) RLC_SERDES_NONCU_MASTER_BUSY__TC0_MASTER_BUSY_MASK | RLC_SERDES_NONCU_MASTER_BUSY__TC1_MASTER_BUSY_MASK; for (k = 0; k < adev->usec_timeout; k++) { - if ((RREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_SERDES_NONCU_MASTER_BUSY)) & mask) == 0) + if ((RREG32_SOC15(GC, 0, mmRLC_SERDES_NONCU_MASTER_BUSY) & mask) == 0) break; udelay(1); } @@ -1366,7 +1368,7 @@ static void gfx_v9_0_wait_for_rlc_serdes(struct amdgpu_device *adev) static void gfx_v9_0_enable_gui_idle_interrupt(struct amdgpu_device *adev, bool enable) { - u32 tmp = RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_INT_CNTL_RING0)); + u32 tmp = RREG32_SOC15(GC, 0, mmCP_INT_CNTL_RING0); if (enable) return; @@ -1376,15 +1378,15 @@ static void gfx_v9_0_enable_gui_idle_interrupt(struct amdgpu_device *adev, tmp = REG_SET_FIELD(tmp, CP_INT_CNTL_RING0, CMP_BUSY_INT_ENABLE, enable ? 1 : 0); tmp = REG_SET_FIELD(tmp, CP_INT_CNTL_RING0, GFX_IDLE_INT_ENABLE, enable ? 1 : 0); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_INT_CNTL_RING0), tmp); + WREG32_SOC15(GC, 0, mmCP_INT_CNTL_RING0, tmp); } void gfx_v9_0_rlc_stop(struct amdgpu_device *adev) { - u32 tmp = RREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_CNTL)); + u32 tmp = RREG32_SOC15(GC, 0, mmRLC_CNTL); tmp = REG_SET_FIELD(tmp, RLC_CNTL, RLC_ENABLE_F32, 0); - WREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_CNTL), tmp); + WREG32_SOC15(GC, 0, mmRLC_CNTL, tmp); gfx_v9_0_enable_gui_idle_interrupt(adev, false); @@ -1415,17 +1417,17 @@ static void gfx_v9_0_rlc_start(struct amdgpu_device *adev) #ifdef AMDGPU_RLC_DEBUG_RETRY /* RLC_GPM_GENERAL_6 : RLC Ucode version */ - rlc_ucode_ver = RREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_GPM_GENERAL_6)); + rlc_ucode_ver = RREG32_SOC15(GC, 0, mmRLC_GPM_GENERAL_6); if(rlc_ucode_ver == 0x108) { DRM_INFO("Using rlc debug ucode. mmRLC_GPM_GENERAL_6 ==0x08%x / fw_ver == %i \n", rlc_ucode_ver, adev->gfx.rlc_fw_version); /* RLC_GPM_TIMER_INT_3 : Timer interval in RefCLK cycles, * default is 0x9C4 to create a 100us interval */ - WREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_GPM_TIMER_INT_3), 0x9C4); + WREG32_SOC15(GC, 0, mmRLC_GPM_TIMER_INT_3, 0x9C4); /* RLC_GPM_GENERAL_12 : Minimum gap between wptr and rptr * to disable the page fault retry interrupts, default is * 0x100 (256) */ - WREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_GPM_GENERAL_12), 0x100); + WREG32_SOC15(GC, 0, mmRLC_GPM_GENERAL_12, 0x100); } #endif } @@ -1446,11 +1448,11 @@ static int gfx_v9_0_rlc_load_microcode(struct amdgpu_device *adev) le32_to_cpu(hdr->header.ucode_array_offset_bytes)); fw_size = le32_to_cpu(hdr->header.ucode_size_bytes) / 4; - WREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_GPM_UCODE_ADDR), + WREG32_SOC15(GC, 0, mmRLC_GPM_UCODE_ADDR, RLCG_UCODE_LOADING_START_ADDRESS); for (i = 0; i < fw_size; i++) - WREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_GPM_UCODE_DATA), le32_to_cpup(fw_data++)); - WREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_GPM_UCODE_ADDR), adev->gfx.rlc_fw_version); + WREG32_SOC15(GC, 0, mmRLC_GPM_UCODE_DATA, le32_to_cpup(fw_data++)); + WREG32_SOC15(GC, 0, mmRLC_GPM_UCODE_ADDR, adev->gfx.rlc_fw_version); return 0; } @@ -1465,10 +1467,10 @@ static int gfx_v9_0_rlc_resume(struct amdgpu_device *adev) gfx_v9_0_rlc_stop(adev); /* disable CG */ - WREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_CGCG_CGLS_CTRL), 0); + WREG32_SOC15(GC, 0, mmRLC_CGCG_CGLS_CTRL, 0); /* disable PG */ - WREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_PG_CNTL), 0); + WREG32_SOC15(GC, 0, mmRLC_PG_CNTL, 0); gfx_v9_0_rlc_reset(adev); @@ -1487,7 +1489,7 @@ static int gfx_v9_0_rlc_resume(struct amdgpu_device *adev) static void gfx_v9_0_cp_gfx_enable(struct amdgpu_device *adev, bool enable) { int i; - u32 tmp = RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_ME_CNTL)); + u32 tmp = RREG32_SOC15(GC, 0, mmCP_ME_CNTL); tmp = REG_SET_FIELD(tmp, CP_ME_CNTL, ME_HALT, enable ? 0 : 1); tmp = REG_SET_FIELD(tmp, CP_ME_CNTL, PFP_HALT, enable ? 0 : 1); @@ -1496,7 +1498,7 @@ static void gfx_v9_0_cp_gfx_enable(struct amdgpu_device *adev, bool enable) for (i = 0; i < adev->gfx.num_gfx_rings; i++) adev->gfx.gfx_ring[i].ready = false; } - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_ME_CNTL), tmp); + WREG32_SOC15(GC, 0, mmCP_ME_CNTL, tmp); udelay(50); } @@ -1529,30 +1531,30 @@ static int gfx_v9_0_cp_gfx_load_microcode(struct amdgpu_device *adev) (adev->gfx.pfp_fw->data + le32_to_cpu(pfp_hdr->header.ucode_array_offset_bytes)); fw_size = le32_to_cpu(pfp_hdr->header.ucode_size_bytes) / 4; - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_PFP_UCODE_ADDR), 0); + WREG32_SOC15(GC, 0, mmCP_PFP_UCODE_ADDR, 0); for (i = 0; i < fw_size; i++) - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_PFP_UCODE_DATA), le32_to_cpup(fw_data++)); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_PFP_UCODE_ADDR), adev->gfx.pfp_fw_version); + WREG32_SOC15(GC, 0, mmCP_PFP_UCODE_DATA, le32_to_cpup(fw_data++)); + WREG32_SOC15(GC, 0, mmCP_PFP_UCODE_ADDR, adev->gfx.pfp_fw_version); /* CE */ fw_data = (const __le32 *) (adev->gfx.ce_fw->data + le32_to_cpu(ce_hdr->header.ucode_array_offset_bytes)); fw_size = le32_to_cpu(ce_hdr->header.ucode_size_bytes) / 4; - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_CE_UCODE_ADDR), 0); + WREG32_SOC15(GC, 0, mmCP_CE_UCODE_ADDR, 0); for (i = 0; i < fw_size; i++) - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_CE_UCODE_DATA), le32_to_cpup(fw_data++)); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_CE_UCODE_ADDR), adev->gfx.ce_fw_version); + WREG32_SOC15(GC, 0, mmCP_CE_UCODE_DATA, le32_to_cpup(fw_data++)); + WREG32_SOC15(GC, 0, mmCP_CE_UCODE_ADDR, adev->gfx.ce_fw_version); /* ME */ fw_data = (const __le32 *) (adev->gfx.me_fw->data + le32_to_cpu(me_hdr->header.ucode_array_offset_bytes)); fw_size = le32_to_cpu(me_hdr->header.ucode_size_bytes) / 4; - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_ME_RAM_WADDR), 0); + WREG32_SOC15(GC, 0, mmCP_ME_RAM_WADDR, 0); for (i = 0; i < fw_size; i++) - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_ME_RAM_DATA), le32_to_cpup(fw_data++)); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_ME_RAM_WADDR), adev->gfx.me_fw_version); + WREG32_SOC15(GC, 0, mmCP_ME_RAM_DATA, le32_to_cpup(fw_data++)); + WREG32_SOC15(GC, 0, mmCP_ME_RAM_WADDR, adev->gfx.me_fw_version); return 0; } @@ -1594,8 +1596,8 @@ static int gfx_v9_0_cp_gfx_start(struct amdgpu_device *adev) int r, i; /* init the CP */ - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_MAX_CONTEXT), adev->gfx.config.max_hw_contexts - 1); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_DEVICE_ID), 1); + WREG32_SOC15(GC, 0, mmCP_MAX_CONTEXT, adev->gfx.config.max_hw_contexts - 1); + WREG32_SOC15(GC, 0, mmCP_DEVICE_ID, 1); gfx_v9_0_cp_gfx_enable(adev, true); @@ -1650,10 +1652,10 @@ static int gfx_v9_0_cp_gfx_resume(struct amdgpu_device *adev) u64 rb_addr, rptr_addr, wptr_gpu_addr; /* Set the write pointer delay */ - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_RB_WPTR_DELAY), 0); + WREG32_SOC15(GC, 0, mmCP_RB_WPTR_DELAY, 0); /* set the RB to use vmid 0 */ - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_RB_VMID), 0); + WREG32_SOC15(GC, 0, mmCP_RB_VMID, 0); /* Set ring buffer size */ ring = &adev->gfx.gfx_ring[0]; @@ -1663,30 +1665,30 @@ static int gfx_v9_0_cp_gfx_resume(struct amdgpu_device *adev) #ifdef __BIG_ENDIAN tmp = REG_SET_FIELD(tmp, CP_RB0_CNTL, BUF_SWAP, 1); #endif - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_RB0_CNTL), tmp); + WREG32_SOC15(GC, 0, mmCP_RB0_CNTL, tmp); /* Initialize the ring buffer's write pointers */ ring->wptr = 0; - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_RB0_WPTR), lower_32_bits(ring->wptr)); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_RB0_WPTR_HI), upper_32_bits(ring->wptr)); + WREG32_SOC15(GC, 0, mmCP_RB0_WPTR, lower_32_bits(ring->wptr)); + WREG32_SOC15(GC, 0, mmCP_RB0_WPTR_HI, upper_32_bits(ring->wptr)); /* set the wb address wether it's enabled or not */ rptr_addr = adev->wb.gpu_addr + (ring->rptr_offs * 4); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_RB0_RPTR_ADDR), lower_32_bits(rptr_addr)); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_RB0_RPTR_ADDR_HI), upper_32_bits(rptr_addr) & CP_RB_RPTR_ADDR_HI__RB_RPTR_ADDR_HI_MASK); + WREG32_SOC15(GC, 0, mmCP_RB0_RPTR_ADDR, lower_32_bits(rptr_addr)); + WREG32_SOC15(GC, 0, mmCP_RB0_RPTR_ADDR_HI, upper_32_bits(rptr_addr) & CP_RB_RPTR_ADDR_HI__RB_RPTR_ADDR_HI_MASK); wptr_gpu_addr = adev->wb.gpu_addr + (ring->wptr_offs * 4); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_RB_WPTR_POLL_ADDR_LO), lower_32_bits(wptr_gpu_addr)); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_RB_WPTR_POLL_ADDR_HI), upper_32_bits(wptr_gpu_addr)); + WREG32_SOC15(GC, 0, mmCP_RB_WPTR_POLL_ADDR_LO, lower_32_bits(wptr_gpu_addr)); + WREG32_SOC15(GC, 0, mmCP_RB_WPTR_POLL_ADDR_HI, upper_32_bits(wptr_gpu_addr)); mdelay(1); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_RB0_CNTL), tmp); + WREG32_SOC15(GC, 0, mmCP_RB0_CNTL, tmp); rb_addr = ring->gpu_addr >> 8; - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_RB0_BASE), rb_addr); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_RB0_BASE_HI), upper_32_bits(rb_addr)); + WREG32_SOC15(GC, 0, mmCP_RB0_BASE, rb_addr); + WREG32_SOC15(GC, 0, mmCP_RB0_BASE_HI, upper_32_bits(rb_addr)); - tmp = RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_RB_DOORBELL_CONTROL)); + tmp = RREG32_SOC15(GC, 0, mmCP_RB_DOORBELL_CONTROL); if (ring->use_doorbell) { tmp = REG_SET_FIELD(tmp, CP_RB_DOORBELL_CONTROL, DOORBELL_OFFSET, ring->doorbell_index); @@ -1695,13 +1697,13 @@ static int gfx_v9_0_cp_gfx_resume(struct amdgpu_device *adev) } else { tmp = REG_SET_FIELD(tmp, CP_RB_DOORBELL_CONTROL, DOORBELL_EN, 0); } - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_RB_DOORBELL_CONTROL), tmp); + WREG32_SOC15(GC, 0, mmCP_RB_DOORBELL_CONTROL, tmp); tmp = REG_SET_FIELD(0, CP_RB_DOORBELL_RANGE_LOWER, DOORBELL_RANGE_LOWER, ring->doorbell_index); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_RB_DOORBELL_RANGE_LOWER), tmp); + WREG32_SOC15(GC, 0, mmCP_RB_DOORBELL_RANGE_LOWER, tmp); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_RB_DOORBELL_RANGE_UPPER), + WREG32_SOC15(GC, 0, mmCP_RB_DOORBELL_RANGE_UPPER, CP_RB_DOORBELL_RANGE_UPPER__DOORBELL_RANGE_UPPER_MASK); @@ -1717,9 +1719,9 @@ static void gfx_v9_0_cp_compute_enable(struct amdgpu_device *adev, bool enable) int i; if (enable) { - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_MEC_CNTL), 0); + WREG32_SOC15(GC, 0, mmCP_MEC_CNTL, 0); } else { - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_MEC_CNTL), + WREG32_SOC15(GC, 0, mmCP_MEC_CNTL, (CP_MEC_CNTL__MEC_ME1_HALT_MASK | CP_MEC_CNTL__MEC_ME2_HALT_MASK)); for (i = 0; i < adev->gfx.num_compute_rings; i++) adev->gfx.compute_ring[i].ready = false; @@ -1756,21 +1758,21 @@ static int gfx_v9_0_cp_compute_load_microcode(struct amdgpu_device *adev) tmp = 0; tmp = REG_SET_FIELD(tmp, CP_CPC_IC_BASE_CNTL, VMID, 0); tmp = REG_SET_FIELD(tmp, CP_CPC_IC_BASE_CNTL, CACHE_POLICY, 0); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_CPC_IC_BASE_CNTL), tmp); + WREG32_SOC15(GC, 0, mmCP_CPC_IC_BASE_CNTL, tmp); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_CPC_IC_BASE_LO), + WREG32_SOC15(GC, 0, mmCP_CPC_IC_BASE_LO, adev->gfx.mec.mec_fw_gpu_addr & 0xFFFFF000); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_CPC_IC_BASE_HI), + WREG32_SOC15(GC, 0, mmCP_CPC_IC_BASE_HI, upper_32_bits(adev->gfx.mec.mec_fw_gpu_addr)); /* MEC1 */ - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_MEC_ME1_UCODE_ADDR), + WREG32_SOC15(GC, 0, mmCP_MEC_ME1_UCODE_ADDR, mec_hdr->jt_offset); for (i = 0; i < mec_hdr->jt_size; i++) - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_MEC_ME1_UCODE_DATA), + WREG32_SOC15(GC, 0, mmCP_MEC_ME1_UCODE_DATA, le32_to_cpup(fw_data + mec_hdr->jt_offset + i)); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_MEC_ME1_UCODE_ADDR), + WREG32_SOC15(GC, 0, mmCP_MEC_ME1_UCODE_ADDR, adev->gfx.mec_fw_version); /* Todo : Loading MEC2 firmware is only necessary if MEC2 should run different microcode than MEC1. */ @@ -1785,7 +1787,7 @@ static void gfx_v9_0_cp_compute_fini(struct amdgpu_device *adev) struct amdgpu_ring *ring = &adev->gfx.compute_ring[i]; if (ring->mqd_obj) { - r = amdgpu_bo_reserve(ring->mqd_obj, false); + r = amdgpu_bo_reserve(ring->mqd_obj, true); if (unlikely(r != 0)) dev_warn(adev->dev, "(%d) reserve MQD bo failed\n", r); @@ -1823,12 +1825,12 @@ static void gfx_v9_0_kiq_setting(struct amdgpu_ring *ring) struct amdgpu_device *adev = ring->adev; /* tell RLC which is KIQ queue */ - tmp = RREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_CP_SCHEDULERS)); + tmp = RREG32_SOC15(GC, 0, mmRLC_CP_SCHEDULERS); tmp &= 0xffffff00; tmp |= (ring->me << 5) | (ring->pipe << 3) | (ring->queue); - WREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_CP_SCHEDULERS), tmp); + WREG32_SOC15(GC, 0, mmRLC_CP_SCHEDULERS, tmp); tmp |= 0x80; - WREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_CP_SCHEDULERS), tmp); + WREG32_SOC15(GC, 0, mmRLC_CP_SCHEDULERS, tmp); } static void gfx_v9_0_kiq_enable(struct amdgpu_ring *ring) @@ -1898,14 +1900,14 @@ static int gfx_v9_0_mqd_init(struct amdgpu_ring *ring) mqd->cp_hqd_eop_base_addr_hi = upper_32_bits(eop_base_addr); /* set the EOP size, register value is 2^(EOP_SIZE+1) dwords */ - tmp = RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_EOP_CONTROL)); + tmp = RREG32_SOC15(GC, 0, mmCP_HQD_EOP_CONTROL); tmp = REG_SET_FIELD(tmp, CP_HQD_EOP_CONTROL, EOP_SIZE, (order_base_2(MEC_HPD_SIZE / 4) - 1)); mqd->cp_hqd_eop_control = tmp; /* enable doorbell? */ - tmp = RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_DOORBELL_CONTROL)); + tmp = RREG32_SOC15(GC, 0, mmCP_HQD_PQ_DOORBELL_CONTROL); if (ring->use_doorbell) { tmp = REG_SET_FIELD(tmp, CP_HQD_PQ_DOORBELL_CONTROL, @@ -1935,7 +1937,7 @@ static int gfx_v9_0_mqd_init(struct amdgpu_ring *ring) mqd->cp_mqd_base_addr_hi = upper_32_bits(ring->mqd_gpu_addr); /* set MQD vmid to 0 */ - tmp = RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_MQD_CONTROL)); + tmp = RREG32_SOC15(GC, 0, mmCP_MQD_CONTROL); tmp = REG_SET_FIELD(tmp, CP_MQD_CONTROL, VMID, 0); mqd->cp_mqd_control = tmp; @@ -1945,7 +1947,7 @@ static int gfx_v9_0_mqd_init(struct amdgpu_ring *ring) mqd->cp_hqd_pq_base_hi = upper_32_bits(hqd_gpu_addr); /* set up the HQD, this is similar to CP_RB0_CNTL */ - tmp = RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_CONTROL)); + tmp = RREG32_SOC15(GC, 0, mmCP_HQD_PQ_CONTROL); tmp = REG_SET_FIELD(tmp, CP_HQD_PQ_CONTROL, QUEUE_SIZE, (order_base_2(ring->ring_size / 4) - 1)); tmp = REG_SET_FIELD(tmp, CP_HQD_PQ_CONTROL, RPTR_BLOCK_SIZE, @@ -1973,7 +1975,7 @@ static int gfx_v9_0_mqd_init(struct amdgpu_ring *ring) tmp = 0; /* enable the doorbell if requested */ if (ring->use_doorbell) { - tmp = RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_DOORBELL_CONTROL)); + tmp = RREG32_SOC15(GC, 0, mmCP_HQD_PQ_DOORBELL_CONTROL); tmp = REG_SET_FIELD(tmp, CP_HQD_PQ_DOORBELL_CONTROL, DOORBELL_OFFSET, ring->doorbell_index); @@ -1989,15 +1991,20 @@ static int gfx_v9_0_mqd_init(struct amdgpu_ring *ring) /* reset read and write pointers, similar to CP_RB0_WPTR/_RPTR */ ring->wptr = 0; - mqd->cp_hqd_pq_rptr = RREG32(mmCP_HQD_PQ_RPTR); + mqd->cp_hqd_pq_rptr = RREG32_SOC15(GC, 0, mmCP_HQD_PQ_RPTR); /* set the vmid for the queue */ mqd->cp_hqd_vmid = 0; - tmp = RREG32(mmCP_HQD_PERSISTENT_STATE); + tmp = RREG32_SOC15(GC, 0, mmCP_HQD_PERSISTENT_STATE); tmp = REG_SET_FIELD(tmp, CP_HQD_PERSISTENT_STATE, PRELOAD_SIZE, 0x53); mqd->cp_hqd_persistent_state = tmp; + /* set MIN_IB_AVAIL_SIZE */ + tmp = RREG32_SOC15(GC, 0, mmCP_HQD_IB_CONTROL); + tmp = REG_SET_FIELD(tmp, CP_HQD_IB_CONTROL, MIN_IB_AVAIL_SIZE, 3); + mqd->cp_hqd_ib_control = tmp; + /* activate the queue */ mqd->cp_hqd_active = 1; @@ -2013,94 +2020,94 @@ static int gfx_v9_0_kiq_init_register(struct amdgpu_ring *ring) /* disable wptr polling */ WREG32_FIELD15(GC, 0, CP_PQ_WPTR_POLL_CNTL, EN, 0); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_EOP_BASE_ADDR), + WREG32_SOC15(GC, 0, mmCP_HQD_EOP_BASE_ADDR, mqd->cp_hqd_eop_base_addr_lo); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_EOP_BASE_ADDR_HI), + WREG32_SOC15(GC, 0, mmCP_HQD_EOP_BASE_ADDR_HI, mqd->cp_hqd_eop_base_addr_hi); /* set the EOP size, register value is 2^(EOP_SIZE+1) dwords */ - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_EOP_CONTROL), + WREG32_SOC15(GC, 0, mmCP_HQD_EOP_CONTROL, mqd->cp_hqd_eop_control); /* enable doorbell? */ - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_DOORBELL_CONTROL), + WREG32_SOC15(GC, 0, mmCP_HQD_PQ_DOORBELL_CONTROL, mqd->cp_hqd_pq_doorbell_control); /* disable the queue if it's active */ - if (RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_ACTIVE)) & 1) { - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_DEQUEUE_REQUEST), 1); + if (RREG32_SOC15(GC, 0, mmCP_HQD_ACTIVE) & 1) { + WREG32_SOC15(GC, 0, mmCP_HQD_DEQUEUE_REQUEST, 1); for (j = 0; j < adev->usec_timeout; j++) { - if (!(RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_ACTIVE)) & 1)) + if (!(RREG32_SOC15(GC, 0, mmCP_HQD_ACTIVE) & 1)) break; udelay(1); } - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_DEQUEUE_REQUEST), + WREG32_SOC15(GC, 0, mmCP_HQD_DEQUEUE_REQUEST, mqd->cp_hqd_dequeue_request); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_RPTR), + WREG32_SOC15(GC, 0, mmCP_HQD_PQ_RPTR, mqd->cp_hqd_pq_rptr); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_WPTR_LO), + WREG32_SOC15(GC, 0, mmCP_HQD_PQ_WPTR_LO, mqd->cp_hqd_pq_wptr_lo); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_WPTR_HI), + WREG32_SOC15(GC, 0, mmCP_HQD_PQ_WPTR_HI, mqd->cp_hqd_pq_wptr_hi); } /* set the pointer to the MQD */ - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_MQD_BASE_ADDR), + WREG32_SOC15(GC, 0, mmCP_MQD_BASE_ADDR, mqd->cp_mqd_base_addr_lo); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_MQD_BASE_ADDR_HI), + WREG32_SOC15(GC, 0, mmCP_MQD_BASE_ADDR_HI, mqd->cp_mqd_base_addr_hi); /* set MQD vmid to 0 */ - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_MQD_CONTROL), + WREG32_SOC15(GC, 0, mmCP_MQD_CONTROL, mqd->cp_mqd_control); /* set the pointer to the HQD, this is similar CP_RB0_BASE/_HI */ - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_BASE), + WREG32_SOC15(GC, 0, mmCP_HQD_PQ_BASE, mqd->cp_hqd_pq_base_lo); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_BASE_HI), + WREG32_SOC15(GC, 0, mmCP_HQD_PQ_BASE_HI, mqd->cp_hqd_pq_base_hi); /* set up the HQD, this is similar to CP_RB0_CNTL */ - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_CONTROL), + WREG32_SOC15(GC, 0, mmCP_HQD_PQ_CONTROL, mqd->cp_hqd_pq_control); /* set the wb address whether it's enabled or not */ - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_RPTR_REPORT_ADDR), + WREG32_SOC15(GC, 0, mmCP_HQD_PQ_RPTR_REPORT_ADDR, mqd->cp_hqd_pq_rptr_report_addr_lo); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_RPTR_REPORT_ADDR_HI), + WREG32_SOC15(GC, 0, mmCP_HQD_PQ_RPTR_REPORT_ADDR_HI, mqd->cp_hqd_pq_rptr_report_addr_hi); /* only used if CP_PQ_WPTR_POLL_CNTL.CP_PQ_WPTR_POLL_CNTL__EN_MASK=1 */ - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_WPTR_POLL_ADDR), + WREG32_SOC15(GC, 0, mmCP_HQD_PQ_WPTR_POLL_ADDR, mqd->cp_hqd_pq_wptr_poll_addr_lo); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_WPTR_POLL_ADDR_HI), + WREG32_SOC15(GC, 0, mmCP_HQD_PQ_WPTR_POLL_ADDR_HI, mqd->cp_hqd_pq_wptr_poll_addr_hi); /* enable the doorbell if requested */ if (ring->use_doorbell) { - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_MEC_DOORBELL_RANGE_LOWER), + WREG32_SOC15(GC, 0, mmCP_MEC_DOORBELL_RANGE_LOWER, (AMDGPU_DOORBELL64_KIQ *2) << 2); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_MEC_DOORBELL_RANGE_UPPER), + WREG32_SOC15(GC, 0, mmCP_MEC_DOORBELL_RANGE_UPPER, (AMDGPU_DOORBELL64_USERQUEUE_END * 2) << 2); } - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_DOORBELL_CONTROL), + WREG32_SOC15(GC, 0, mmCP_HQD_PQ_DOORBELL_CONTROL, mqd->cp_hqd_pq_doorbell_control); /* reset read and write pointers, similar to CP_RB0_WPTR/_RPTR */ - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_WPTR_LO), + WREG32_SOC15(GC, 0, mmCP_HQD_PQ_WPTR_LO, mqd->cp_hqd_pq_wptr_lo); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_WPTR_HI), + WREG32_SOC15(GC, 0, mmCP_HQD_PQ_WPTR_HI, mqd->cp_hqd_pq_wptr_hi); /* set the vmid for the queue */ - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_VMID), mqd->cp_hqd_vmid); + WREG32_SOC15(GC, 0, mmCP_HQD_VMID, mqd->cp_hqd_vmid); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PERSISTENT_STATE), + WREG32_SOC15(GC, 0, mmCP_HQD_PERSISTENT_STATE, mqd->cp_hqd_persistent_state); /* activate the queue */ - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_ACTIVE), + WREG32_SOC15(GC, 0, mmCP_HQD_ACTIVE, mqd->cp_hqd_active); if (ring->use_doorbell) @@ -2323,7 +2330,7 @@ static bool gfx_v9_0_is_idle(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; - if (REG_GET_FIELD(RREG32(SOC15_REG_OFFSET(GC, 0, mmGRBM_STATUS)), + if (REG_GET_FIELD(RREG32_SOC15(GC, 0, mmGRBM_STATUS), GRBM_STATUS, GUI_ACTIVE)) return false; else @@ -2338,7 +2345,7 @@ static int gfx_v9_0_wait_for_idle(void *handle) for (i = 0; i < adev->usec_timeout; i++) { /* read MC_STATUS */ - tmp = RREG32(SOC15_REG_OFFSET(GC, 0, mmGRBM_STATUS)) & + tmp = RREG32_SOC15(GC, 0, mmGRBM_STATUS) & GRBM_STATUS__GUI_ACTIVE_MASK; if (!REG_GET_FIELD(tmp, GRBM_STATUS, GUI_ACTIVE)) @@ -2355,7 +2362,7 @@ static int gfx_v9_0_soft_reset(void *handle) struct amdgpu_device *adev = (struct amdgpu_device *)handle; /* GRBM_STATUS */ - tmp = RREG32(SOC15_REG_OFFSET(GC, 0, mmGRBM_STATUS)); + tmp = RREG32_SOC15(GC, 0, mmGRBM_STATUS); if (tmp & (GRBM_STATUS__PA_BUSY_MASK | GRBM_STATUS__SC_BUSY_MASK | GRBM_STATUS__BCI_BUSY_MASK | GRBM_STATUS__SX_BUSY_MASK | GRBM_STATUS__TA_BUSY_MASK | GRBM_STATUS__VGT_BUSY_MASK | @@ -2374,7 +2381,7 @@ static int gfx_v9_0_soft_reset(void *handle) } /* GRBM_STATUS2 */ - tmp = RREG32(SOC15_REG_OFFSET(GC, 0, mmGRBM_STATUS2)); + tmp = RREG32_SOC15(GC, 0, mmGRBM_STATUS2); if (REG_GET_FIELD(tmp, GRBM_STATUS2, RLC_BUSY)) grbm_soft_reset = REG_SET_FIELD(grbm_soft_reset, GRBM_SOFT_RESET, SOFT_RESET_RLC, 1); @@ -2391,17 +2398,17 @@ static int gfx_v9_0_soft_reset(void *handle) gfx_v9_0_cp_compute_enable(adev, false); if (grbm_soft_reset) { - tmp = RREG32(SOC15_REG_OFFSET(GC, 0, mmGRBM_SOFT_RESET)); + tmp = RREG32_SOC15(GC, 0, mmGRBM_SOFT_RESET); tmp |= grbm_soft_reset; dev_info(adev->dev, "GRBM_SOFT_RESET=0x%08X\n", tmp); - WREG32(SOC15_REG_OFFSET(GC, 0, mmGRBM_SOFT_RESET), tmp); - tmp = RREG32(SOC15_REG_OFFSET(GC, 0, mmGRBM_SOFT_RESET)); + WREG32_SOC15(GC, 0, mmGRBM_SOFT_RESET, tmp); + tmp = RREG32_SOC15(GC, 0, mmGRBM_SOFT_RESET); udelay(50); tmp &= ~grbm_soft_reset; - WREG32(SOC15_REG_OFFSET(GC, 0, mmGRBM_SOFT_RESET), tmp); - tmp = RREG32(SOC15_REG_OFFSET(GC, 0, mmGRBM_SOFT_RESET)); + WREG32_SOC15(GC, 0, mmGRBM_SOFT_RESET, tmp); + tmp = RREG32_SOC15(GC, 0, mmGRBM_SOFT_RESET); } /* Wait a little for things to settle down */ @@ -2415,9 +2422,9 @@ static uint64_t gfx_v9_0_get_gpu_clock_counter(struct amdgpu_device *adev) uint64_t clock; mutex_lock(&adev->gfx.gpu_clock_mutex); - WREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_CAPTURE_GPU_CLOCK_COUNT), 1); - clock = (uint64_t)RREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_GPU_CLOCK_COUNT_LSB)) | - ((uint64_t)RREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_GPU_CLOCK_COUNT_MSB)) << 32ULL); + WREG32_SOC15(GC, 0, mmRLC_CAPTURE_GPU_CLOCK_COUNT, 1); + clock = (uint64_t)RREG32_SOC15(GC, 0, mmRLC_GPU_CLOCK_COUNT_LSB) | + ((uint64_t)RREG32_SOC15(GC, 0, mmRLC_GPU_CLOCK_COUNT_MSB) << 32ULL); mutex_unlock(&adev->gfx.gpu_clock_mutex); return clock; } @@ -2497,7 +2504,7 @@ static void gfx_v9_0_enter_rlc_safe_mode(struct amdgpu_device *adev) return; /* if RLC is not enabled, do nothing */ - rlc_setting = RREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_CNTL)); + rlc_setting = RREG32_SOC15(GC, 0, mmRLC_CNTL); if (!(rlc_setting & RLC_CNTL__RLC_ENABLE_F32_MASK)) return; @@ -2506,7 +2513,7 @@ static void gfx_v9_0_enter_rlc_safe_mode(struct amdgpu_device *adev) AMD_CG_SUPPORT_GFX_3D_CGCG)) { data = RLC_SAFE_MODE__CMD_MASK; data |= (1 << RLC_SAFE_MODE__MESSAGE__SHIFT); - WREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_SAFE_MODE), data); + WREG32_SOC15(GC, 0, mmRLC_SAFE_MODE, data); /* wait for RLC_SAFE_MODE */ for (i = 0; i < adev->usec_timeout; i++) { @@ -2526,7 +2533,7 @@ static void gfx_v9_0_exit_rlc_safe_mode(struct amdgpu_device *adev) return; /* if RLC is not enabled, do nothing */ - rlc_setting = RREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_CNTL)); + rlc_setting = RREG32_SOC15(GC, 0, mmRLC_CNTL); if (!(rlc_setting & RLC_CNTL__RLC_ENABLE_F32_MASK)) return; @@ -2537,7 +2544,7 @@ static void gfx_v9_0_exit_rlc_safe_mode(struct amdgpu_device *adev) * mode. */ data = RLC_SAFE_MODE__CMD_MASK; - WREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_SAFE_MODE), data); + WREG32_SOC15(GC, 0, mmRLC_SAFE_MODE, data); adev->gfx.rlc.in_safe_mode = false; } } @@ -2550,7 +2557,7 @@ static void gfx_v9_0_update_medium_grain_clock_gating(struct amdgpu_device *adev /* It is disabled by HW by default */ if (enable && (adev->cg_flags & AMD_CG_SUPPORT_GFX_MGCG)) { /* 1 - RLC_CGTT_MGCG_OVERRIDE */ - def = data = RREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_CGTT_MGCG_OVERRIDE)); + def = data = RREG32_SOC15(GC, 0, mmRLC_CGTT_MGCG_OVERRIDE); data &= ~(RLC_CGTT_MGCG_OVERRIDE__CPF_CGTT_SCLK_OVERRIDE_MASK | RLC_CGTT_MGCG_OVERRIDE__GRBM_CGTT_SCLK_OVERRIDE_MASK | RLC_CGTT_MGCG_OVERRIDE__GFXIP_MGCG_OVERRIDE_MASK | @@ -2560,48 +2567,48 @@ static void gfx_v9_0_update_medium_grain_clock_gating(struct amdgpu_device *adev data |= RLC_CGTT_MGCG_OVERRIDE__RLC_CGTT_SCLK_OVERRIDE_MASK; if (def != data) - WREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_CGTT_MGCG_OVERRIDE), data); + WREG32_SOC15(GC, 0, mmRLC_CGTT_MGCG_OVERRIDE, data); /* MGLS is a global flag to control all MGLS in GFX */ if (adev->cg_flags & AMD_CG_SUPPORT_GFX_MGLS) { /* 2 - RLC memory Light sleep */ if (adev->cg_flags & AMD_CG_SUPPORT_GFX_RLC_LS) { - def = data = RREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_MEM_SLP_CNTL)); + def = data = RREG32_SOC15(GC, 0, mmRLC_MEM_SLP_CNTL); data |= RLC_MEM_SLP_CNTL__RLC_MEM_LS_EN_MASK; if (def != data) - WREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_MEM_SLP_CNTL), data); + WREG32_SOC15(GC, 0, mmRLC_MEM_SLP_CNTL, data); } /* 3 - CP memory Light sleep */ if (adev->cg_flags & AMD_CG_SUPPORT_GFX_CP_LS) { - def = data = RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_MEM_SLP_CNTL)); + def = data = RREG32_SOC15(GC, 0, mmCP_MEM_SLP_CNTL); data |= CP_MEM_SLP_CNTL__CP_MEM_LS_EN_MASK; if (def != data) - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_MEM_SLP_CNTL), data); + WREG32_SOC15(GC, 0, mmCP_MEM_SLP_CNTL, data); } } } else { /* 1 - MGCG_OVERRIDE */ - def = data = RREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_CGTT_MGCG_OVERRIDE)); + def = data = RREG32_SOC15(GC, 0, mmRLC_CGTT_MGCG_OVERRIDE); data |= (RLC_CGTT_MGCG_OVERRIDE__CPF_CGTT_SCLK_OVERRIDE_MASK | RLC_CGTT_MGCG_OVERRIDE__RLC_CGTT_SCLK_OVERRIDE_MASK | RLC_CGTT_MGCG_OVERRIDE__GRBM_CGTT_SCLK_OVERRIDE_MASK | RLC_CGTT_MGCG_OVERRIDE__GFXIP_MGCG_OVERRIDE_MASK | RLC_CGTT_MGCG_OVERRIDE__GFXIP_MGLS_OVERRIDE_MASK); if (def != data) - WREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_CGTT_MGCG_OVERRIDE), data); + WREG32_SOC15(GC, 0, mmRLC_CGTT_MGCG_OVERRIDE, data); /* 2 - disable MGLS in RLC */ - data = RREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_MEM_SLP_CNTL)); + data = RREG32_SOC15(GC, 0, mmRLC_MEM_SLP_CNTL); if (data & RLC_MEM_SLP_CNTL__RLC_MEM_LS_EN_MASK) { data &= ~RLC_MEM_SLP_CNTL__RLC_MEM_LS_EN_MASK; - WREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_MEM_SLP_CNTL), data); + WREG32_SOC15(GC, 0, mmRLC_MEM_SLP_CNTL, data); } /* 3 - disable MGLS in CP */ - data = RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_MEM_SLP_CNTL)); + data = RREG32_SOC15(GC, 0, mmCP_MEM_SLP_CNTL); if (data & CP_MEM_SLP_CNTL__CP_MEM_LS_EN_MASK) { data &= ~CP_MEM_SLP_CNTL__CP_MEM_LS_EN_MASK; - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_MEM_SLP_CNTL), data); + WREG32_SOC15(GC, 0, mmCP_MEM_SLP_CNTL, data); } } } @@ -2616,37 +2623,37 @@ static void gfx_v9_0_update_3d_clock_gating(struct amdgpu_device *adev, /* Enable 3D CGCG/CGLS */ if (enable && (adev->cg_flags & AMD_CG_SUPPORT_GFX_3D_CGCG)) { /* write cmd to clear cgcg/cgls ov */ - def = data = RREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_CGTT_MGCG_OVERRIDE)); + def = data = RREG32_SOC15(GC, 0, mmRLC_CGTT_MGCG_OVERRIDE); /* unset CGCG override */ data &= ~RLC_CGTT_MGCG_OVERRIDE__GFXIP_GFX3D_CG_OVERRIDE_MASK; /* update CGCG and CGLS override bits */ if (def != data) - WREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_CGTT_MGCG_OVERRIDE), data); + WREG32_SOC15(GC, 0, mmRLC_CGTT_MGCG_OVERRIDE, data); /* enable 3Dcgcg FSM(0x0020003f) */ - def = RREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_CGCG_CGLS_CTRL_3D)); + def = RREG32_SOC15(GC, 0, mmRLC_CGCG_CGLS_CTRL_3D); data = (0x2000 << RLC_CGCG_CGLS_CTRL_3D__CGCG_GFX_IDLE_THRESHOLD__SHIFT) | RLC_CGCG_CGLS_CTRL_3D__CGCG_EN_MASK; if (adev->cg_flags & AMD_CG_SUPPORT_GFX_3D_CGLS) data |= (0x000F << RLC_CGCG_CGLS_CTRL_3D__CGLS_REP_COMPANSAT_DELAY__SHIFT) | RLC_CGCG_CGLS_CTRL_3D__CGLS_EN_MASK; if (def != data) - WREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_CGCG_CGLS_CTRL_3D), data); + WREG32_SOC15(GC, 0, mmRLC_CGCG_CGLS_CTRL_3D, data); /* set IDLE_POLL_COUNT(0x00900100) */ - def = RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_RB_WPTR_POLL_CNTL)); + def = RREG32_SOC15(GC, 0, mmCP_RB_WPTR_POLL_CNTL); data = (0x0100 << CP_RB_WPTR_POLL_CNTL__POLL_FREQUENCY__SHIFT) | (0x0090 << CP_RB_WPTR_POLL_CNTL__IDLE_POLL_COUNT__SHIFT); if (def != data) - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_RB_WPTR_POLL_CNTL), data); + WREG32_SOC15(GC, 0, mmCP_RB_WPTR_POLL_CNTL, data); } else { /* Disable CGCG/CGLS */ - def = data = RREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_CGCG_CGLS_CTRL_3D)); + def = data = RREG32_SOC15(GC, 0, mmRLC_CGCG_CGLS_CTRL_3D); /* disable cgcg, cgls should be disabled */ data &= ~(RLC_CGCG_CGLS_CTRL_3D__CGCG_EN_MASK | RLC_CGCG_CGLS_CTRL_3D__CGLS_EN_MASK); /* disable cgcg and cgls in FSM */ if (def != data) - WREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_CGCG_CGLS_CTRL_3D), data); + WREG32_SOC15(GC, 0, mmRLC_CGCG_CGLS_CTRL_3D, data); } adev->gfx.rlc.funcs->exit_safe_mode(adev); @@ -2660,7 +2667,7 @@ static void gfx_v9_0_update_coarse_grain_clock_gating(struct amdgpu_device *adev adev->gfx.rlc.funcs->enter_safe_mode(adev); if (enable && (adev->cg_flags & AMD_CG_SUPPORT_GFX_CGCG)) { - def = data = RREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_CGTT_MGCG_OVERRIDE)); + def = data = RREG32_SOC15(GC, 0, mmRLC_CGTT_MGCG_OVERRIDE); /* unset CGCG override */ data &= ~RLC_CGTT_MGCG_OVERRIDE__GFXIP_CGCG_OVERRIDE_MASK; if (adev->cg_flags & AMD_CG_SUPPORT_GFX_CGLS) @@ -2669,31 +2676,31 @@ static void gfx_v9_0_update_coarse_grain_clock_gating(struct amdgpu_device *adev data |= RLC_CGTT_MGCG_OVERRIDE__GFXIP_CGLS_OVERRIDE_MASK; /* update CGCG and CGLS override bits */ if (def != data) - WREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_CGTT_MGCG_OVERRIDE), data); + WREG32_SOC15(GC, 0, mmRLC_CGTT_MGCG_OVERRIDE, data); /* enable cgcg FSM(0x0020003F) */ - def = RREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_CGCG_CGLS_CTRL)); + def = RREG32_SOC15(GC, 0, mmRLC_CGCG_CGLS_CTRL); data = (0x2000 << RLC_CGCG_CGLS_CTRL__CGCG_GFX_IDLE_THRESHOLD__SHIFT) | RLC_CGCG_CGLS_CTRL__CGCG_EN_MASK; if (adev->cg_flags & AMD_CG_SUPPORT_GFX_CGLS) data |= (0x000F << RLC_CGCG_CGLS_CTRL__CGLS_REP_COMPANSAT_DELAY__SHIFT) | RLC_CGCG_CGLS_CTRL__CGLS_EN_MASK; if (def != data) - WREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_CGCG_CGLS_CTRL), data); + WREG32_SOC15(GC, 0, mmRLC_CGCG_CGLS_CTRL, data); /* set IDLE_POLL_COUNT(0x00900100) */ - def = RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_RB_WPTR_POLL_CNTL)); + def = RREG32_SOC15(GC, 0, mmCP_RB_WPTR_POLL_CNTL); data = (0x0100 << CP_RB_WPTR_POLL_CNTL__POLL_FREQUENCY__SHIFT) | (0x0090 << CP_RB_WPTR_POLL_CNTL__IDLE_POLL_COUNT__SHIFT); if (def != data) - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_RB_WPTR_POLL_CNTL), data); + WREG32_SOC15(GC, 0, mmCP_RB_WPTR_POLL_CNTL, data); } else { - def = data = RREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_CGCG_CGLS_CTRL)); + def = data = RREG32_SOC15(GC, 0, mmRLC_CGCG_CGLS_CTRL); /* reset CGCG/CGLS bits */ data &= ~(RLC_CGCG_CGLS_CTRL__CGCG_EN_MASK | RLC_CGCG_CGLS_CTRL__CGLS_EN_MASK); /* disable cgcg and cgls in FSM */ if (def != data) - WREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_CGCG_CGLS_CTRL), data); + WREG32_SOC15(GC, 0, mmRLC_CGCG_CGLS_CTRL, data); } adev->gfx.rlc.funcs->exit_safe_mode(adev); @@ -2740,6 +2747,9 @@ static int gfx_v9_0_set_clockgating_state(void *handle, { struct amdgpu_device *adev = (struct amdgpu_device *)handle; + if (amdgpu_sriov_vf(adev)) + return 0; + switch (adev->asic_type) { case CHIP_VEGA10: gfx_v9_0_update_gfx_clock_gating(adev, @@ -2760,12 +2770,12 @@ static void gfx_v9_0_get_clockgating_state(void *handle, u32 *flags) *flags = 0; /* AMD_CG_SUPPORT_GFX_MGCG */ - data = RREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_CGTT_MGCG_OVERRIDE)); + data = RREG32_SOC15(GC, 0, mmRLC_CGTT_MGCG_OVERRIDE); if (!(data & RLC_CGTT_MGCG_OVERRIDE__GFXIP_MGCG_OVERRIDE_MASK)) *flags |= AMD_CG_SUPPORT_GFX_MGCG; /* AMD_CG_SUPPORT_GFX_CGCG */ - data = RREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_CGCG_CGLS_CTRL)); + data = RREG32_SOC15(GC, 0, mmRLC_CGCG_CGLS_CTRL); if (data & RLC_CGCG_CGLS_CTRL__CGCG_EN_MASK) *flags |= AMD_CG_SUPPORT_GFX_CGCG; @@ -2774,17 +2784,17 @@ static void gfx_v9_0_get_clockgating_state(void *handle, u32 *flags) *flags |= AMD_CG_SUPPORT_GFX_CGLS; /* AMD_CG_SUPPORT_GFX_RLC_LS */ - data = RREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_MEM_SLP_CNTL)); + data = RREG32_SOC15(GC, 0, mmRLC_MEM_SLP_CNTL); if (data & RLC_MEM_SLP_CNTL__RLC_MEM_LS_EN_MASK) *flags |= AMD_CG_SUPPORT_GFX_RLC_LS | AMD_CG_SUPPORT_GFX_MGLS; /* AMD_CG_SUPPORT_GFX_CP_LS */ - data = RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_MEM_SLP_CNTL)); + data = RREG32_SOC15(GC, 0, mmCP_MEM_SLP_CNTL); if (data & CP_MEM_SLP_CNTL__CP_MEM_LS_EN_MASK) *flags |= AMD_CG_SUPPORT_GFX_CP_LS | AMD_CG_SUPPORT_GFX_MGLS; /* AMD_CG_SUPPORT_GFX_3D_CGCG */ - data = RREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_CGCG_CGLS_CTRL_3D)); + data = RREG32_SOC15(GC, 0, mmRLC_CGCG_CGLS_CTRL_3D); if (data & RLC_CGCG_CGLS_CTRL_3D__CGCG_EN_MASK) *flags |= AMD_CG_SUPPORT_GFX_3D_CGCG; @@ -2807,8 +2817,8 @@ static u64 gfx_v9_0_ring_get_wptr_gfx(struct amdgpu_ring *ring) if (ring->use_doorbell) { wptr = atomic64_read((atomic64_t *)&adev->wb.wb[ring->wptr_offs]); } else { - wptr = RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_RB0_WPTR)); - wptr += (u64)RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_RB0_WPTR_HI)) << 32; + wptr = RREG32_SOC15(GC, 0, mmCP_RB0_WPTR); + wptr += (u64)RREG32_SOC15(GC, 0, mmCP_RB0_WPTR_HI) << 32; } return wptr; @@ -2823,8 +2833,8 @@ static void gfx_v9_0_ring_set_wptr_gfx(struct amdgpu_ring *ring) atomic64_set((atomic64_t*)&adev->wb.wb[ring->wptr_offs], ring->wptr); WDOORBELL64(ring->doorbell_index, ring->wptr); } else { - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_RB0_WPTR), lower_32_bits(ring->wptr)); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_RB0_WPTR_HI), upper_32_bits(ring->wptr)); + WREG32_SOC15(GC, 0, mmCP_RB0_WPTR, lower_32_bits(ring->wptr)); + WREG32_SOC15(GC, 0, mmCP_RB0_WPTR_HI, upper_32_bits(ring->wptr)); } } @@ -2956,35 +2966,29 @@ static void gfx_v9_0_ring_emit_pipeline_sync(struct amdgpu_ring *ring) static void gfx_v9_0_ring_emit_vm_flush(struct amdgpu_ring *ring, unsigned vm_id, uint64_t pd_addr) { + struct amdgpu_vmhub *hub = &ring->adev->vmhub[ring->funcs->vmhub]; int usepfp = (ring->funcs->type == AMDGPU_RING_TYPE_GFX); uint32_t req = ring->adev->gart.gart_funcs->get_invalidate_req(vm_id); - unsigned eng = ring->idx; - unsigned i; + unsigned eng = ring->vm_inv_eng; pd_addr = pd_addr | 0x1; /* valid bit */ /* now only use physical base address of PDE and valid */ BUG_ON(pd_addr & 0xFFFF00000000003EULL); - for (i = 0; i < AMDGPU_MAX_VMHUBS; ++i) { - struct amdgpu_vmhub *hub = &ring->adev->vmhub[i]; - - gfx_v9_0_write_data_to_reg(ring, usepfp, true, - hub->ctx0_ptb_addr_lo32 - + (2 * vm_id), - lower_32_bits(pd_addr)); + gfx_v9_0_write_data_to_reg(ring, usepfp, true, + hub->ctx0_ptb_addr_lo32 + (2 * vm_id), + lower_32_bits(pd_addr)); - gfx_v9_0_write_data_to_reg(ring, usepfp, true, - hub->ctx0_ptb_addr_hi32 - + (2 * vm_id), - upper_32_bits(pd_addr)); + gfx_v9_0_write_data_to_reg(ring, usepfp, true, + hub->ctx0_ptb_addr_hi32 + (2 * vm_id), + upper_32_bits(pd_addr)); - gfx_v9_0_write_data_to_reg(ring, usepfp, true, - hub->vm_inv_eng0_req + eng, req); + gfx_v9_0_write_data_to_reg(ring, usepfp, true, + hub->vm_inv_eng0_req + eng, req); - /* wait for the invalidate to complete */ - gfx_v9_0_wait_reg_mem(ring, 0, 0, 0, hub->vm_inv_eng0_ack + - eng, 0, 1 << vm_id, 1 << vm_id, 0x20); - } + /* wait for the invalidate to complete */ + gfx_v9_0_wait_reg_mem(ring, 0, 0, 0, hub->vm_inv_eng0_ack + + eng, 0, 1 << vm_id, 1 << vm_id, 0x20); /* compute doesn't have PFP */ if (usepfp) { @@ -3373,9 +3377,7 @@ static int gfx_v9_0_kiq_set_interrupt_state(struct amdgpu_device *adev, enum amdgpu_interrupt_state state) { uint32_t tmp, target; - struct amdgpu_ring *ring = (struct amdgpu_ring *)src->data; - - BUG_ON(!ring || (ring->funcs->type != AMDGPU_RING_TYPE_KIQ)); + struct amdgpu_ring *ring = &(adev->gfx.kiq.ring); if (ring->me == 1) target = SOC15_REG_OFFSET(GC, 0, mmCP_ME1_PIPE0_INT_CNTL); @@ -3386,20 +3388,20 @@ static int gfx_v9_0_kiq_set_interrupt_state(struct amdgpu_device *adev, switch (type) { case AMDGPU_CP_KIQ_IRQ_DRIVER0: if (state == AMDGPU_IRQ_STATE_DISABLE) { - tmp = RREG32(SOC15_REG_OFFSET(GC, 0, mmCPC_INT_CNTL)); + tmp = RREG32_SOC15(GC, 0, mmCPC_INT_CNTL); tmp = REG_SET_FIELD(tmp, CPC_INT_CNTL, GENERIC2_INT_ENABLE, 0); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCPC_INT_CNTL), tmp); + WREG32_SOC15(GC, 0, mmCPC_INT_CNTL, tmp); tmp = RREG32(target); tmp = REG_SET_FIELD(tmp, CP_ME2_PIPE0_INT_CNTL, GENERIC2_INT_ENABLE, 0); WREG32(target, tmp); } else { - tmp = RREG32(SOC15_REG_OFFSET(GC, 0, mmCPC_INT_CNTL)); + tmp = RREG32_SOC15(GC, 0, mmCPC_INT_CNTL); tmp = REG_SET_FIELD(tmp, CPC_INT_CNTL, GENERIC2_INT_ENABLE, 1); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCPC_INT_CNTL), tmp); + WREG32_SOC15(GC, 0, mmCPC_INT_CNTL, tmp); tmp = RREG32(target); tmp = REG_SET_FIELD(tmp, CP_ME2_PIPE0_INT_CNTL, @@ -3419,9 +3421,7 @@ static int gfx_v9_0_kiq_irq(struct amdgpu_device *adev, struct amdgpu_iv_entry *entry) { u8 me_id, pipe_id, queue_id; - struct amdgpu_ring *ring = (struct amdgpu_ring *)source->data; - - BUG_ON(!ring || (ring->funcs->type != AMDGPU_RING_TYPE_KIQ)); + struct amdgpu_ring *ring = &(adev->gfx.kiq.ring); me_id = (entry->ring_id & 0x0c) >> 2; pipe_id = (entry->ring_id & 0x03) >> 0; @@ -3456,13 +3456,14 @@ static const struct amdgpu_ring_funcs gfx_v9_0_ring_funcs_gfx = { .align_mask = 0xff, .nop = PACKET3(PACKET3_NOP, 0x3FFF), .support_64bit_ptrs = true, + .vmhub = AMDGPU_GFXHUB, .get_rptr = gfx_v9_0_ring_get_rptr_gfx, .get_wptr = gfx_v9_0_ring_get_wptr_gfx, .set_wptr = gfx_v9_0_ring_set_wptr_gfx, .emit_frame_size = /* totally 242 maximum if 16 IBs */ 5 + /* COND_EXEC */ 7 + /* PIPELINE_SYNC */ - 46 + /* VM_FLUSH */ + 24 + /* VM_FLUSH */ 8 + /* FENCE for VM_FLUSH */ 20 + /* GDS switch */ 4 + /* double SWITCH_BUFFER, @@ -3500,6 +3501,7 @@ static const struct amdgpu_ring_funcs gfx_v9_0_ring_funcs_compute = { .align_mask = 0xff, .nop = PACKET3(PACKET3_NOP, 0x3FFF), .support_64bit_ptrs = true, + .vmhub = AMDGPU_GFXHUB, .get_rptr = gfx_v9_0_ring_get_rptr_compute, .get_wptr = gfx_v9_0_ring_get_wptr_compute, .set_wptr = gfx_v9_0_ring_set_wptr_compute, @@ -3508,7 +3510,7 @@ static const struct amdgpu_ring_funcs gfx_v9_0_ring_funcs_compute = { 7 + /* gfx_v9_0_ring_emit_hdp_flush */ 5 + /* gfx_v9_0_ring_emit_hdp_invalidate */ 7 + /* gfx_v9_0_ring_emit_pipeline_sync */ - 64 + /* gfx_v9_0_ring_emit_vm_flush */ + 24 + /* gfx_v9_0_ring_emit_vm_flush */ 8 + 8 + 8, /* gfx_v9_0_ring_emit_fence x3 for user fence, vm fence */ .emit_ib_size = 4, /* gfx_v9_0_ring_emit_ib_compute */ .emit_ib = gfx_v9_0_ring_emit_ib_compute, @@ -3529,6 +3531,7 @@ static const struct amdgpu_ring_funcs gfx_v9_0_ring_funcs_kiq = { .align_mask = 0xff, .nop = PACKET3(PACKET3_NOP, 0x3FFF), .support_64bit_ptrs = true, + .vmhub = AMDGPU_GFXHUB, .get_rptr = gfx_v9_0_ring_get_rptr_compute, .get_wptr = gfx_v9_0_ring_get_wptr_compute, .set_wptr = gfx_v9_0_ring_set_wptr_compute, @@ -3537,7 +3540,7 @@ static const struct amdgpu_ring_funcs gfx_v9_0_ring_funcs_kiq = { 7 + /* gfx_v9_0_ring_emit_hdp_flush */ 5 + /* gfx_v9_0_ring_emit_hdp_invalidate */ 7 + /* gfx_v9_0_ring_emit_pipeline_sync */ - 64 + /* gfx_v9_0_ring_emit_vm_flush */ + 24 + /* gfx_v9_0_ring_emit_vm_flush */ 8 + 8 + 8, /* gfx_v9_0_ring_emit_fence_kiq x3 for user fence, vm fence */ .emit_ib_size = 4, /* gfx_v9_0_ring_emit_ib_compute */ .emit_ib = gfx_v9_0_ring_emit_ib_compute, @@ -3612,7 +3615,7 @@ static void gfx_v9_0_set_rlc_funcs(struct amdgpu_device *adev) static void gfx_v9_0_set_gds_init(struct amdgpu_device *adev) { /* init asci gds info */ - adev->gds.mem.total_size = RREG32(SOC15_REG_OFFSET(GC, 0, mmGDS_VMID0_SIZE)); + adev->gds.mem.total_size = RREG32_SOC15(GC, 0, mmGDS_VMID0_SIZE); adev->gds.gws.total_size = 64; adev->gds.oa.total_size = 16; @@ -3641,8 +3644,8 @@ static u32 gfx_v9_0_get_cu_active_bitmap(struct amdgpu_device *adev) { u32 data, mask; - data = RREG32(SOC15_REG_OFFSET(GC, 0, mmCC_GC_SHADER_ARRAY_CONFIG)); - data |= RREG32(SOC15_REG_OFFSET(GC, 0, mmGC_USER_SHADER_ARRAY_CONFIG)); + data = RREG32_SOC15(GC, 0, mmCC_GC_SHADER_ARRAY_CONFIG); + data |= RREG32_SOC15(GC, 0, mmGC_USER_SHADER_ARRAY_CONFIG); data &= CC_GC_SHADER_ARRAY_CONFIG__INACTIVE_CUS_MASK; data >>= CC_GC_SHADER_ARRAY_CONFIG__INACTIVE_CUS__SHIFT; @@ -3763,25 +3766,25 @@ static int gfx_v9_0_init_queue(struct amdgpu_ring *ring) eop_gpu_addr = adev->gfx.mec.hpd_eop_gpu_addr + (ring->queue * MEC_HPD_SIZE); eop_gpu_addr >>= 8; - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_EOP_BASE_ADDR), lower_32_bits(eop_gpu_addr)); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_EOP_BASE_ADDR_HI), upper_32_bits(eop_gpu_addr)); + WREG32_SOC15(GC, 0, mmCP_HQD_EOP_BASE_ADDR, lower_32_bits(eop_gpu_addr)); + WREG32_SOC15(GC, 0, mmCP_HQD_EOP_BASE_ADDR_HI, upper_32_bits(eop_gpu_addr)); mqd->cp_hqd_eop_base_addr_lo = lower_32_bits(eop_gpu_addr); mqd->cp_hqd_eop_base_addr_hi = upper_32_bits(eop_gpu_addr); /* set the EOP size, register value is 2^(EOP_SIZE+1) dwords */ - tmp = RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_EOP_CONTROL)); + tmp = RREG32_SOC15(GC, 0, mmCP_HQD_EOP_CONTROL); tmp = REG_SET_FIELD(tmp, CP_HQD_EOP_CONTROL, EOP_SIZE, (order_base_2(MEC_HPD_SIZE / 4) - 1)); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_EOP_CONTROL), tmp); + WREG32_SOC15(GC, 0, mmCP_HQD_EOP_CONTROL, tmp); /* enable doorbell? */ - tmp = RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_DOORBELL_CONTROL)); + tmp = RREG32_SOC15(GC, 0, mmCP_HQD_PQ_DOORBELL_CONTROL); if (use_doorbell) tmp = REG_SET_FIELD(tmp, CP_HQD_PQ_DOORBELL_CONTROL, DOORBELL_EN, 1); else tmp = REG_SET_FIELD(tmp, CP_HQD_PQ_DOORBELL_CONTROL, DOORBELL_EN, 0); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_DOORBELL_CONTROL), tmp); + WREG32_SOC15(GC, 0, mmCP_HQD_PQ_DOORBELL_CONTROL, tmp); mqd->cp_hqd_pq_doorbell_control = tmp; /* disable the queue if it's active */ @@ -3790,40 +3793,40 @@ static int gfx_v9_0_init_queue(struct amdgpu_ring *ring) mqd->cp_hqd_pq_rptr = 0; mqd->cp_hqd_pq_wptr_lo = 0; mqd->cp_hqd_pq_wptr_hi = 0; - if (RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_ACTIVE)) & 1) { - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_DEQUEUE_REQUEST), 1); + if (RREG32_SOC15(GC, 0, mmCP_HQD_ACTIVE) & 1) { + WREG32_SOC15(GC, 0, mmCP_HQD_DEQUEUE_REQUEST, 1); for (j = 0; j < adev->usec_timeout; j++) { - if (!(RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_ACTIVE)) & 1)) + if (!(RREG32_SOC15(GC, 0, mmCP_HQD_ACTIVE) & 1)) break; udelay(1); } - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_DEQUEUE_REQUEST), mqd->cp_hqd_dequeue_request); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_RPTR), mqd->cp_hqd_pq_rptr); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_WPTR_LO), mqd->cp_hqd_pq_wptr_lo); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_WPTR_HI), mqd->cp_hqd_pq_wptr_hi); + WREG32_SOC15(GC, 0, mmCP_HQD_DEQUEUE_REQUEST, mqd->cp_hqd_dequeue_request); + WREG32_SOC15(GC, 0, mmCP_HQD_PQ_RPTR, mqd->cp_hqd_pq_rptr); + WREG32_SOC15(GC, 0, mmCP_HQD_PQ_WPTR_LO, mqd->cp_hqd_pq_wptr_lo); + WREG32_SOC15(GC, 0, mmCP_HQD_PQ_WPTR_HI, mqd->cp_hqd_pq_wptr_hi); } /* set the pointer to the MQD */ mqd->cp_mqd_base_addr_lo = mqd_gpu_addr & 0xfffffffc; mqd->cp_mqd_base_addr_hi = upper_32_bits(mqd_gpu_addr); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_MQD_BASE_ADDR), mqd->cp_mqd_base_addr_lo); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_MQD_BASE_ADDR_HI), mqd->cp_mqd_base_addr_hi); + WREG32_SOC15(GC, 0, mmCP_MQD_BASE_ADDR, mqd->cp_mqd_base_addr_lo); + WREG32_SOC15(GC, 0, mmCP_MQD_BASE_ADDR_HI, mqd->cp_mqd_base_addr_hi); /* set MQD vmid to 0 */ - tmp = RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_MQD_CONTROL)); + tmp = RREG32_SOC15(GC, 0, mmCP_MQD_CONTROL); tmp = REG_SET_FIELD(tmp, CP_MQD_CONTROL, VMID, 0); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_MQD_CONTROL), tmp); + WREG32_SOC15(GC, 0, mmCP_MQD_CONTROL, tmp); mqd->cp_mqd_control = tmp; /* set the pointer to the HQD, this is similar CP_RB0_BASE/_HI */ hqd_gpu_addr = ring->gpu_addr >> 8; mqd->cp_hqd_pq_base_lo = hqd_gpu_addr; mqd->cp_hqd_pq_base_hi = upper_32_bits(hqd_gpu_addr); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_BASE), mqd->cp_hqd_pq_base_lo); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_BASE_HI), mqd->cp_hqd_pq_base_hi); + WREG32_SOC15(GC, 0, mmCP_HQD_PQ_BASE, mqd->cp_hqd_pq_base_lo); + WREG32_SOC15(GC, 0, mmCP_HQD_PQ_BASE_HI, mqd->cp_hqd_pq_base_hi); /* set up the HQD, this is similar to CP_RB0_CNTL */ - tmp = RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_CONTROL)); + tmp = RREG32_SOC15(GC, 0, mmCP_HQD_PQ_CONTROL); tmp = REG_SET_FIELD(tmp, CP_HQD_PQ_CONTROL, QUEUE_SIZE, (order_base_2(ring->ring_size / 4) - 1)); tmp = REG_SET_FIELD(tmp, CP_HQD_PQ_CONTROL, RPTR_BLOCK_SIZE, @@ -3835,7 +3838,7 @@ static int gfx_v9_0_init_queue(struct amdgpu_ring *ring) tmp = REG_SET_FIELD(tmp, CP_HQD_PQ_CONTROL, ROQ_PQ_IB_FLIP, 0); tmp = REG_SET_FIELD(tmp, CP_HQD_PQ_CONTROL, PRIV_STATE, 1); tmp = REG_SET_FIELD(tmp, CP_HQD_PQ_CONTROL, KMD_QUEUE, 1); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_CONTROL), tmp); + WREG32_SOC15(GC, 0, mmCP_HQD_PQ_CONTROL, tmp); mqd->cp_hqd_pq_control = tmp; /* set the wb address wether it's enabled or not */ @@ -3843,27 +3846,27 @@ static int gfx_v9_0_init_queue(struct amdgpu_ring *ring) mqd->cp_hqd_pq_rptr_report_addr_lo = wb_gpu_addr & 0xfffffffc; mqd->cp_hqd_pq_rptr_report_addr_hi = upper_32_bits(wb_gpu_addr) & 0xffff; - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_RPTR_REPORT_ADDR), + WREG32_SOC15(GC, 0, mmCP_HQD_PQ_RPTR_REPORT_ADDR, mqd->cp_hqd_pq_rptr_report_addr_lo); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_RPTR_REPORT_ADDR_HI), + WREG32_SOC15(GC, 0, mmCP_HQD_PQ_RPTR_REPORT_ADDR_HI, mqd->cp_hqd_pq_rptr_report_addr_hi); /* only used if CP_PQ_WPTR_POLL_CNTL.CP_PQ_WPTR_POLL_CNTL__EN_MASK=1 */ wb_gpu_addr = adev->wb.gpu_addr + (ring->wptr_offs * 4); mqd->cp_hqd_pq_wptr_poll_addr_lo = wb_gpu_addr & 0xfffffffc; mqd->cp_hqd_pq_wptr_poll_addr_hi = upper_32_bits(wb_gpu_addr) & 0xffff; - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_WPTR_POLL_ADDR), + WREG32_SOC15(GC, 0, mmCP_HQD_PQ_WPTR_POLL_ADDR, mqd->cp_hqd_pq_wptr_poll_addr_lo); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_WPTR_POLL_ADDR_HI), + WREG32_SOC15(GC, 0, mmCP_HQD_PQ_WPTR_POLL_ADDR_HI, mqd->cp_hqd_pq_wptr_poll_addr_hi); /* enable the doorbell if requested */ if (use_doorbell) { - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_MEC_DOORBELL_RANGE_LOWER), + WREG32_SOC15(GC, 0, mmCP_MEC_DOORBELL_RANGE_LOWER, (AMDGPU_DOORBELL64_KIQ * 2) << 2); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_MEC_DOORBELL_RANGE_UPPER), + WREG32_SOC15(GC, 0, mmCP_MEC_DOORBELL_RANGE_UPPER, (AMDGPU_DOORBELL64_MEC_RING7 * 2) << 2); - tmp = RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_DOORBELL_CONTROL)); + tmp = RREG32_SOC15(GC, 0, mmCP_HQD_PQ_DOORBELL_CONTROL); tmp = REG_SET_FIELD(tmp, CP_HQD_PQ_DOORBELL_CONTROL, DOORBELL_OFFSET, ring->doorbell_index); tmp = REG_SET_FIELD(tmp, CP_HQD_PQ_DOORBELL_CONTROL, DOORBELL_EN, 1); @@ -3874,25 +3877,25 @@ static int gfx_v9_0_init_queue(struct amdgpu_ring *ring) } else { mqd->cp_hqd_pq_doorbell_control = 0; } - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_DOORBELL_CONTROL), + WREG32_SOC15(GC, 0, mmCP_HQD_PQ_DOORBELL_CONTROL, mqd->cp_hqd_pq_doorbell_control); /* reset read and write pointers, similar to CP_RB0_WPTR/_RPTR */ - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_WPTR_LO), mqd->cp_hqd_pq_wptr_lo); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_WPTR_HI), mqd->cp_hqd_pq_wptr_hi); + WREG32_SOC15(GC, 0, mmCP_HQD_PQ_WPTR_LO, mqd->cp_hqd_pq_wptr_lo); + WREG32_SOC15(GC, 0, mmCP_HQD_PQ_WPTR_HI, mqd->cp_hqd_pq_wptr_hi); /* set the vmid for the queue */ mqd->cp_hqd_vmid = 0; - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_VMID), mqd->cp_hqd_vmid); + WREG32_SOC15(GC, 0, mmCP_HQD_VMID, mqd->cp_hqd_vmid); - tmp = RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PERSISTENT_STATE)); + tmp = RREG32_SOC15(GC, 0, mmCP_HQD_PERSISTENT_STATE); tmp = REG_SET_FIELD(tmp, CP_HQD_PERSISTENT_STATE, PRELOAD_SIZE, 0x53); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PERSISTENT_STATE), tmp); + WREG32_SOC15(GC, 0, mmCP_HQD_PERSISTENT_STATE, tmp); mqd->cp_hqd_persistent_state = tmp; /* activate the queue */ mqd->cp_hqd_active = 1; - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_ACTIVE), mqd->cp_hqd_active); + WREG32_SOC15(GC, 0, mmCP_HQD_ACTIVE, mqd->cp_hqd_active); soc15_grbm_select(adev, 0, 0, 0, 0); mutex_unlock(&adev->srbm_mutex); diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c index 631aef38126d..2a3983036a30 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c @@ -21,7 +21,7 @@ * */ #include <linux/firmware.h> -#include "drmP.h" +#include <drm/drmP.h> #include "amdgpu.h" #include "gmc_v6_0.h" #include "amdgpu_ucode.h" @@ -346,7 +346,8 @@ static int gmc_v6_0_mc_init(struct amdgpu_device *adev) * size equal to the 1024 or vram, whichever is larger. */ if (amdgpu_gart_size == -1) - adev->mc.gtt_size = max((1024ULL << 20), adev->mc.mc_vram_size); + adev->mc.gtt_size = max((AMDGPU_DEFAULT_GTT_SIZE_MB << 20), + adev->mc.mc_vram_size); else adev->mc.gtt_size = (uint64_t)amdgpu_gart_size << 20; @@ -621,7 +622,7 @@ static int gmc_v6_0_vm_init(struct amdgpu_device *adev) * amdgpu graphics/compute will use VMIDs 1-7 * amdkfd will use VMIDs 8-15 */ - adev->vm_manager.num_ids = AMDGPU_NUM_OF_VMIDS; + adev->vm_manager.id_mgr[0].num_ids = AMDGPU_NUM_OF_VMIDS; adev->vm_manager.num_level = 1; amdgpu_vm_manager_init(adev); diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c index 92abe12d92bb..6d347c1d2516 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c @@ -21,7 +21,7 @@ * */ #include <linux/firmware.h> -#include "drmP.h" +#include <drm/drmP.h> #include "amdgpu.h" #include "cikd.h" #include "cik.h" @@ -395,7 +395,8 @@ static int gmc_v7_0_mc_init(struct amdgpu_device *adev) * size equal to the 1024 or vram, whichever is larger. */ if (amdgpu_gart_size == -1) - adev->mc.gtt_size = max((1024ULL << 20), adev->mc.mc_vram_size); + adev->mc.gtt_size = max((AMDGPU_DEFAULT_GTT_SIZE_MB << 20), + adev->mc.mc_vram_size); else adev->mc.gtt_size = (uint64_t)amdgpu_gart_size << 20; @@ -746,7 +747,7 @@ static int gmc_v7_0_vm_init(struct amdgpu_device *adev) * amdgpu graphics/compute will use VMIDs 1-7 * amdkfd will use VMIDs 8-15 */ - adev->vm_manager.num_ids = AMDGPU_NUM_OF_VMIDS; + adev->vm_manager.id_mgr[0].num_ids = AMDGPU_NUM_OF_VMIDS; adev->vm_manager.num_level = 1; amdgpu_vm_manager_init(adev); diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c index f2ccefc66fd4..2784ff49cf56 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c @@ -21,7 +21,7 @@ * */ #include <linux/firmware.h> -#include "drmP.h" +#include <drm/drmP.h> #include "amdgpu.h" #include "gmc_v8_0.h" #include "amdgpu_ucode.h" @@ -557,7 +557,8 @@ static int gmc_v8_0_mc_init(struct amdgpu_device *adev) * size equal to the 1024 or vram, whichever is larger. */ if (amdgpu_gart_size == -1) - adev->mc.gtt_size = max((1024ULL << 20), adev->mc.mc_vram_size); + adev->mc.gtt_size = max((AMDGPU_DEFAULT_GTT_SIZE_MB << 20), + adev->mc.mc_vram_size); else adev->mc.gtt_size = (uint64_t)amdgpu_gart_size << 20; @@ -949,7 +950,7 @@ static int gmc_v8_0_vm_init(struct amdgpu_device *adev) * amdgpu graphics/compute will use VMIDs 1-7 * amdkfd will use VMIDs 8-15 */ - adev->vm_manager.num_ids = AMDGPU_NUM_OF_VMIDS; + adev->vm_manager.id_mgr[0].num_ids = AMDGPU_NUM_OF_VMIDS; adev->vm_manager.num_level = 1; amdgpu_vm_manager_init(adev); diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c index 3b045e0b114e..dc1e1c1d6b24 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c @@ -386,6 +386,23 @@ static int gmc_v9_0_early_init(void *handle) static int gmc_v9_0_late_init(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; + unsigned vm_inv_eng[AMDGPU_MAX_VMHUBS] = { 3, 3 }; + unsigned i; + + for(i = 0; i < adev->num_rings; ++i) { + struct amdgpu_ring *ring = adev->rings[i]; + unsigned vmhub = ring->funcs->vmhub; + + ring->vm_inv_eng = vm_inv_eng[vmhub]++; + dev_info(adev->dev, "ring %u(%s) uses VM inv eng %u on hub %u\n", + ring->idx, ring->name, ring->vm_inv_eng, + ring->funcs->vmhub); + } + + /* Engine 17 is used for GART flushes */ + for(i = 0; i < AMDGPU_MAX_VMHUBS; ++i) + BUG_ON(vm_inv_eng[i] > 17); + return amdgpu_irq_get(adev, &adev->mc.vm_fault, 0); } @@ -469,7 +486,8 @@ static int gmc_v9_0_mc_init(struct amdgpu_device *adev) * size equal to the 1024 or vram, whichever is larger. */ if (amdgpu_gart_size == -1) - adev->mc.gtt_size = max((1024ULL << 20), adev->mc.mc_vram_size); + adev->mc.gtt_size = max((AMDGPU_DEFAULT_GTT_SIZE_MB << 20), + adev->mc.mc_vram_size); else adev->mc.gtt_size = (uint64_t)amdgpu_gart_size << 20; @@ -519,7 +537,8 @@ static int gmc_v9_0_vm_init(struct amdgpu_device *adev) * amdgpu graphics/compute will use VMIDs 1-7 * amdkfd will use VMIDs 8-15 */ - adev->vm_manager.num_ids = AMDGPU_NUM_OF_VMIDS; + adev->vm_manager.id_mgr[AMDGPU_GFXHUB].num_ids = AMDGPU_NUM_OF_VMIDS; + adev->vm_manager.id_mgr[AMDGPU_MMHUB].num_ids = AMDGPU_NUM_OF_VMIDS; /* TODO: fix num_level for APU when updating vm size and block size */ if (adev->flags & AMD_IS_APU) diff --git a/drivers/gpu/drm/amd/amdgpu/iceland_ih.c b/drivers/gpu/drm/amd/amdgpu/iceland_ih.c index cb622add99a7..7a0ea27ac429 100644 --- a/drivers/gpu/drm/amd/amdgpu/iceland_ih.c +++ b/drivers/gpu/drm/amd/amdgpu/iceland_ih.c @@ -20,7 +20,7 @@ * OTHER DEALINGS IN THE SOFTWARE. * */ -#include "drmP.h" +#include <drm/drmP.h> #include "amdgpu.h" #include "amdgpu_ih.h" #include "vid.h" diff --git a/drivers/gpu/drm/amd/amdgpu/kv_dpm.c b/drivers/gpu/drm/amd/amdgpu/kv_dpm.c index 79a52ad2c80d..3bbf2ccfca89 100644 --- a/drivers/gpu/drm/amd/amdgpu/kv_dpm.c +++ b/drivers/gpu/drm/amd/amdgpu/kv_dpm.c @@ -21,7 +21,7 @@ * */ -#include "drmP.h" +#include <drm/drmP.h> #include "amdgpu.h" #include "amdgpu_pm.h" #include "cikd.h" diff --git a/drivers/gpu/drm/amd/amdgpu/kv_smc.c b/drivers/gpu/drm/amd/amdgpu/kv_smc.c index e6b7b42acfe1..b82e33c01571 100644 --- a/drivers/gpu/drm/amd/amdgpu/kv_smc.c +++ b/drivers/gpu/drm/amd/amdgpu/kv_smc.c @@ -22,7 +22,7 @@ * Authors: Alex Deucher */ -#include "drmP.h" +#include <drm/drmP.h> #include "amdgpu.h" #include "cikd.h" #include "kv_dpm.h" diff --git a/drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.c b/drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.c index 62684510ddcd..dbfe48d1207a 100644 --- a/drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.c +++ b/drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.c @@ -511,6 +511,9 @@ static int mmhub_v1_0_set_clockgating_state(void *handle, { struct amdgpu_device *adev = (struct amdgpu_device *)handle; + if (amdgpu_sriov_vf(adev)) + return 0; + switch (adev->asic_type) { case CHIP_VEGA10: mmhub_v1_0_update_medium_grain_clock_gating(adev, diff --git a/drivers/gpu/drm/amd/amdgpu/mmsch_v1_0.h b/drivers/gpu/drm/amd/amdgpu/mmsch_v1_0.h index 5f0fc8bf16a9..8af0bddf85e4 100644 --- a/drivers/gpu/drm/amd/amdgpu/mmsch_v1_0.h +++ b/drivers/gpu/drm/amd/amdgpu/mmsch_v1_0.h @@ -84,4 +84,61 @@ struct mmsch_v1_0_cmd_indirect_write { uint32_t reg_value; }; +static inline void mmsch_v1_0_insert_direct_wt(struct mmsch_v1_0_cmd_direct_write *direct_wt, + uint32_t *init_table, + uint32_t reg_offset, + uint32_t value) +{ + direct_wt->cmd_header.reg_offset = reg_offset; + direct_wt->reg_value = value; + memcpy((void *)init_table, direct_wt, sizeof(struct mmsch_v1_0_cmd_direct_write)); +} + +static inline void mmsch_v1_0_insert_direct_rd_mod_wt(struct mmsch_v1_0_cmd_direct_read_modify_write *direct_rd_mod_wt, + uint32_t *init_table, + uint32_t reg_offset, + uint32_t mask, uint32_t data) +{ + direct_rd_mod_wt->cmd_header.reg_offset = reg_offset; + direct_rd_mod_wt->mask_value = mask; + direct_rd_mod_wt->write_data = data; + memcpy((void *)init_table, direct_rd_mod_wt, + sizeof(struct mmsch_v1_0_cmd_direct_read_modify_write)); +} + +static inline void mmsch_v1_0_insert_direct_poll(struct mmsch_v1_0_cmd_direct_polling *direct_poll, + uint32_t *init_table, + uint32_t reg_offset, + uint32_t mask, uint32_t wait) +{ + direct_poll->cmd_header.reg_offset = reg_offset; + direct_poll->mask_value = mask; + direct_poll->wait_value = wait; + memcpy((void *)init_table, direct_poll, sizeof(struct mmsch_v1_0_cmd_direct_polling)); +} + +#define MMSCH_V1_0_INSERT_DIRECT_RD_MOD_WT(reg, mask, data) { \ + mmsch_v1_0_insert_direct_rd_mod_wt(&direct_rd_mod_wt, \ + init_table, (reg), \ + (mask), (data)); \ + init_table += sizeof(struct mmsch_v1_0_cmd_direct_read_modify_write)/4; \ + table_size += sizeof(struct mmsch_v1_0_cmd_direct_read_modify_write)/4; \ +} + +#define MMSCH_V1_0_INSERT_DIRECT_WT(reg, value) { \ + mmsch_v1_0_insert_direct_wt(&direct_wt, \ + init_table, (reg), \ + (value)); \ + init_table += sizeof(struct mmsch_v1_0_cmd_direct_write)/4; \ + table_size += sizeof(struct mmsch_v1_0_cmd_direct_write)/4; \ +} + +#define MMSCH_V1_0_INSERT_DIRECT_POLL(reg, mask, wait) { \ + mmsch_v1_0_insert_direct_poll(&direct_poll, \ + init_table, (reg), \ + (mask), (wait)); \ + init_table += sizeof(struct mmsch_v1_0_cmd_direct_polling)/4; \ + table_size += sizeof(struct mmsch_v1_0_cmd_direct_polling)/4; \ +} + #endif diff --git a/drivers/gpu/drm/amd/amdgpu/mxgpu_vi.c b/drivers/gpu/drm/amd/amdgpu/mxgpu_vi.c index 70a3dd13cb02..7bdc51b02326 100644 --- a/drivers/gpu/drm/amd/amdgpu/mxgpu_vi.c +++ b/drivers/gpu/drm/amd/amdgpu/mxgpu_vi.c @@ -368,9 +368,12 @@ static int xgpu_vi_mailbox_rcv_msg(struct amdgpu_device *adev, u32 reg; u32 mask = REG_FIELD_MASK(MAILBOX_CONTROL, RCV_MSG_VALID); - reg = RREG32_NO_KIQ(mmMAILBOX_CONTROL); - if (!(reg & mask)) - return -ENOENT; + /* workaround: host driver doesn't set VALID for CMPL now */ + if (event != IDH_FLR_NOTIFICATION_CMPL) { + reg = RREG32_NO_KIQ(mmMAILBOX_CONTROL); + if (!(reg & mask)) + return -ENOENT; + } reg = RREG32_NO_KIQ(mmMAILBOX_MSGBUF_RCV_DW0); if (reg != event) diff --git a/drivers/gpu/drm/amd/amdgpu/psp_v3_1.c b/drivers/gpu/drm/amd/amdgpu/psp_v3_1.c index c3588d1c7cb0..eef89abc0cee 100644 --- a/drivers/gpu/drm/amd/amdgpu/psp_v3_1.c +++ b/drivers/gpu/drm/amd/amdgpu/psp_v3_1.c @@ -24,7 +24,7 @@ */ #include <linux/firmware.h> -#include "drmP.h" +#include <drm/drmP.h> #include "amdgpu.h" #include "amdgpu_psp.h" #include "amdgpu_ucode.h" @@ -166,11 +166,8 @@ int psp_v3_1_bootloader_load_sysdrv(struct psp_context *psp) { int ret; uint32_t psp_gfxdrv_command_reg = 0; - struct amdgpu_bo *psp_sysdrv; - void *psp_sysdrv_virt = NULL; - uint64_t psp_sysdrv_mem; struct amdgpu_device *adev = psp->adev; - uint32_t size, sol_reg; + uint32_t sol_reg; /* Check sOS sign of life register to confirm sys driver and sOS * are already been loaded. @@ -185,27 +182,14 @@ int psp_v3_1_bootloader_load_sysdrv(struct psp_context *psp) if (ret) return ret; - /* - * Create a 1 meg GART memory to store the psp sys driver - * binary with a 1 meg aligned address - */ - size = (psp->sys_bin_size + (PSP_BOOTLOADER_1_MEG_ALIGNMENT - 1)) & - (~(PSP_BOOTLOADER_1_MEG_ALIGNMENT - 1)); - - ret = amdgpu_bo_create_kernel(adev, size, PSP_BOOTLOADER_1_MEG_ALIGNMENT, - AMDGPU_GEM_DOMAIN_GTT, - &psp_sysdrv, - &psp_sysdrv_mem, - &psp_sysdrv_virt); - if (ret) - return ret; + memset(psp->fw_pri_buf, 0, PSP_1_MEG); /* Copy PSP System Driver binary to memory */ - memcpy(psp_sysdrv_virt, psp->sys_start_addr, psp->sys_bin_size); + memcpy(psp->fw_pri_buf, psp->sys_start_addr, psp->sys_bin_size); /* Provide the sys driver to bootrom */ WREG32(SOC15_REG_OFFSET(MP0, 0, mmMP0_SMN_C2PMSG_36), - (uint32_t)(psp_sysdrv_mem >> 20)); + (uint32_t)(psp->fw_pri_mc_addr >> 20)); psp_gfxdrv_command_reg = 1 << 16; WREG32(SOC15_REG_OFFSET(MP0, 0, mmMP0_SMN_C2PMSG_35), psp_gfxdrv_command_reg); @@ -216,8 +200,6 @@ int psp_v3_1_bootloader_load_sysdrv(struct psp_context *psp) ret = psp_wait_for(psp, SOC15_REG_OFFSET(MP0, 0, mmMP0_SMN_C2PMSG_35), 0x80000000, 0x80000000, false); - amdgpu_bo_free_kernel(&psp_sysdrv, &psp_sysdrv_mem, &psp_sysdrv_virt); - return ret; } @@ -225,11 +207,8 @@ int psp_v3_1_bootloader_load_sos(struct psp_context *psp) { int ret; unsigned int psp_gfxdrv_command_reg = 0; - struct amdgpu_bo *psp_sos; - void *psp_sos_virt = NULL; - uint64_t psp_sos_mem; struct amdgpu_device *adev = psp->adev; - uint32_t size, sol_reg; + uint32_t sol_reg; /* Check sOS sign of life register to confirm sys driver and sOS * are already been loaded. @@ -244,23 +223,14 @@ int psp_v3_1_bootloader_load_sos(struct psp_context *psp) if (ret) return ret; - size = (psp->sos_bin_size + (PSP_BOOTLOADER_1_MEG_ALIGNMENT - 1)) & - (~((uint64_t)PSP_BOOTLOADER_1_MEG_ALIGNMENT - 1)); - - ret = amdgpu_bo_create_kernel(adev, size, PSP_BOOTLOADER_1_MEG_ALIGNMENT, - AMDGPU_GEM_DOMAIN_GTT, - &psp_sos, - &psp_sos_mem, - &psp_sos_virt); - if (ret) - return ret; + memset(psp->fw_pri_buf, 0, PSP_1_MEG); /* Copy Secure OS binary to PSP memory */ - memcpy(psp_sos_virt, psp->sos_start_addr, psp->sos_bin_size); + memcpy(psp->fw_pri_buf, psp->sos_start_addr, psp->sos_bin_size); /* Provide the PSP secure OS to bootrom */ WREG32(SOC15_REG_OFFSET(MP0, 0, mmMP0_SMN_C2PMSG_36), - (uint32_t)(psp_sos_mem >> 20)); + (uint32_t)(psp->fw_pri_mc_addr >> 20)); psp_gfxdrv_command_reg = 2 << 16; WREG32(SOC15_REG_OFFSET(MP0, 0, mmMP0_SMN_C2PMSG_35), psp_gfxdrv_command_reg); @@ -273,8 +243,6 @@ int psp_v3_1_bootloader_load_sos(struct psp_context *psp) 0, true); #endif - amdgpu_bo_free_kernel(&psp_sos, &psp_sos_mem, &psp_sos_virt); - return ret; } @@ -300,7 +268,6 @@ int psp_v3_1_prep_cmd_buf(struct amdgpu_firmware_info *ucode, struct psp_gfx_cmd int psp_v3_1_ring_init(struct psp_context *psp, enum psp_ring_type ring_type) { int ret = 0; - unsigned int psp_ring_reg = 0; struct psp_ring *ring; struct amdgpu_device *adev = psp->adev; @@ -320,6 +287,16 @@ int psp_v3_1_ring_init(struct psp_context *psp, enum psp_ring_type ring_type) return ret; } + return 0; +} + +int psp_v3_1_ring_create(struct psp_context *psp, enum psp_ring_type ring_type) +{ + int ret = 0; + unsigned int psp_ring_reg = 0; + struct psp_ring *ring = &psp->km_ring; + struct amdgpu_device *adev = psp->adev; + /* Write low address of the ring to C2PMSG_69 */ psp_ring_reg = lower_32_bits(ring->ring_mem_mc_addr); WREG32(SOC15_REG_OFFSET(MP0, 0, mmMP0_SMN_C2PMSG_69), psp_ring_reg); @@ -344,6 +321,33 @@ int psp_v3_1_ring_init(struct psp_context *psp, enum psp_ring_type ring_type) return ret; } +int psp_v3_1_ring_destroy(struct psp_context *psp, enum psp_ring_type ring_type) +{ + int ret = 0; + struct psp_ring *ring; + unsigned int psp_ring_reg = 0; + struct amdgpu_device *adev = psp->adev; + + ring = &psp->km_ring; + + /* Write the ring destroy command to C2PMSG_64 */ + psp_ring_reg = 3 << 16; + WREG32(SOC15_REG_OFFSET(MP0, 0, mmMP0_SMN_C2PMSG_64), psp_ring_reg); + + /* there might be handshake issue with hardware which needs delay */ + mdelay(20); + + /* Wait for response flag (bit 31) in C2PMSG_64 */ + ret = psp_wait_for(psp, SOC15_REG_OFFSET(MP0, 0, mmMP0_SMN_C2PMSG_64), + 0x80000000, 0x80000000, false); + + if (ring->ring_mem) + amdgpu_bo_free_kernel(&adev->firmware.rbuf, + &ring->ring_mem_mc_addr, + (void **)&ring->ring_mem); + return ret; +} + int psp_v3_1_cmd_submit(struct psp_context *psp, struct amdgpu_firmware_info *ucode, uint64_t cmd_buf_mc_addr, uint64_t fence_mc_addr, diff --git a/drivers/gpu/drm/amd/amdgpu/psp_v3_1.h b/drivers/gpu/drm/amd/amdgpu/psp_v3_1.h index e82eff741a08..9dcd0b25c4c6 100644 --- a/drivers/gpu/drm/amd/amdgpu/psp_v3_1.h +++ b/drivers/gpu/drm/amd/amdgpu/psp_v3_1.h @@ -39,6 +39,10 @@ extern int psp_v3_1_prep_cmd_buf(struct amdgpu_firmware_info *ucode, struct psp_gfx_cmd_resp *cmd); extern int psp_v3_1_ring_init(struct psp_context *psp, enum psp_ring_type ring_type); +extern int psp_v3_1_ring_create(struct psp_context *psp, + enum psp_ring_type ring_type); +extern int psp_v3_1_ring_destroy(struct psp_context *psp, + enum psp_ring_type ring_type); extern int psp_v3_1_cmd_submit(struct psp_context *psp, struct amdgpu_firmware_info *ucode, uint64_t cmd_buf_mc_addr, uint64_t fence_mc_addr, diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c b/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c index 21f38d882335..ecc70a730a54 100644 --- a/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c +++ b/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c @@ -48,8 +48,7 @@ static void sdma_v4_0_set_buffer_funcs(struct amdgpu_device *adev); static void sdma_v4_0_set_vm_pte_funcs(struct amdgpu_device *adev); static void sdma_v4_0_set_irq_funcs(struct amdgpu_device *adev); -static const u32 golden_settings_sdma_4[] = -{ +static const u32 golden_settings_sdma_4[] = { SOC15_REG_OFFSET(SDMA0, 0, mmSDMA0_CHICKEN_BITS), 0xfe931f07, 0x02831f07, SOC15_REG_OFFSET(SDMA0, 0, mmSDMA0_CLK_CTRL), 0xff000ff0, 0x3f000100, SOC15_REG_OFFSET(SDMA0, 0, mmSDMA0_GFX_IB_CNTL), 0x800f0100, 0x00000100, @@ -76,8 +75,7 @@ static const u32 golden_settings_sdma_4[] = SOC15_REG_OFFSET(SDMA1, 0, mmSDMA1_UTCL1_PAGE), 0x000003ff, 0x000003c0 }; -static const u32 golden_settings_sdma_vg10[] = -{ +static const u32 golden_settings_sdma_vg10[] = { SOC15_REG_OFFSET(SDMA0, 0, mmSDMA0_GB_ADDR_CONFIG), 0x0018773f, 0x00104002, SOC15_REG_OFFSET(SDMA0, 0, mmSDMA0_GB_ADDR_CONFIG_READ), 0x0018773f, 0x00104002, SOC15_REG_OFFSET(SDMA1, 0, mmSDMA1_GB_ADDR_CONFIG), 0x0018773f, 0x00104002, @@ -87,16 +85,17 @@ static const u32 golden_settings_sdma_vg10[] = static u32 sdma_v4_0_get_reg_offset(u32 instance, u32 internal_offset) { u32 base = 0; + switch (instance) { - case 0: - base = SDMA0_BASE.instance[0].segment[0]; - break; - case 1: - base = SDMA1_BASE.instance[0].segment[0]; - break; - default: - BUG(); - break; + case 0: + base = SDMA0_BASE.instance[0].segment[0]; + break; + case 1: + base = SDMA1_BASE.instance[0].segment[0]; + break; + default: + BUG(); + break; } return base + internal_offset; @@ -159,7 +158,8 @@ static int sdma_v4_0_init_microcode(struct amdgpu_device *adev) case CHIP_VEGA10: chip_name = "vega10"; break; - default: BUG(); + default: + BUG(); } for (i = 0; i < adev->sdma.num_instances; i++) { @@ -179,7 +179,7 @@ static int sdma_v4_0_init_microcode(struct amdgpu_device *adev) if (adev->sdma.instance[i].feature_version >= 20) adev->sdma.instance[i].burst_nop = true; DRM_DEBUG("psp_load == '%s'\n", - adev->firmware.load_type == AMDGPU_FW_LOAD_PSP? "true": "false"); + adev->firmware.load_type == AMDGPU_FW_LOAD_PSP ? "true" : "false"); if (adev->firmware.load_type == AMDGPU_FW_LOAD_PSP) { info = &adev->firmware.ucode[AMDGPU_UCODE_ID_SDMA0 + i]; @@ -192,9 +192,7 @@ static int sdma_v4_0_init_microcode(struct amdgpu_device *adev) } out: if (err) { - printk(KERN_ERR - "sdma_v4_0: Failed to load firmware \"%s\"\n", - fw_name); + DRM_ERROR("sdma_v4_0: Failed to load firmware \"%s\"\n", fw_name); for (i = 0; i < adev->sdma.num_instances; i++) { release_firmware(adev->sdma.instance[i].fw); adev->sdma.instance[i].fw = NULL; @@ -212,10 +210,10 @@ out: */ static uint64_t sdma_v4_0_ring_get_rptr(struct amdgpu_ring *ring) { - u64* rptr; + u64 *rptr; /* XXX check if swapping is necessary on BE */ - rptr =((u64*)&ring->adev->wb.wb[ring->rptr_offs]); + rptr = ((u64 *)&ring->adev->wb.wb[ring->rptr_offs]); DRM_DEBUG("rptr before shift == 0x%016llx\n", *rptr); return ((*rptr) >> 2); @@ -231,19 +229,20 @@ static uint64_t sdma_v4_0_ring_get_rptr(struct amdgpu_ring *ring) static uint64_t sdma_v4_0_ring_get_wptr(struct amdgpu_ring *ring) { struct amdgpu_device *adev = ring->adev; - u64* wptr = NULL; - uint64_t local_wptr=0; + u64 *wptr = NULL; + uint64_t local_wptr = 0; if (ring->use_doorbell) { /* XXX check if swapping is necessary on BE */ - wptr = ((u64*)&adev->wb.wb[ring->wptr_offs]); + wptr = ((u64 *)&adev->wb.wb[ring->wptr_offs]); DRM_DEBUG("wptr/doorbell before shift == 0x%016llx\n", *wptr); *wptr = (*wptr) >> 2; DRM_DEBUG("wptr/doorbell after shift == 0x%016llx\n", *wptr); } else { u32 lowbit, highbit; int me = (ring == &adev->sdma.instance[0].ring) ? 0 : 1; - wptr=&local_wptr; + + wptr = &local_wptr; lowbit = RREG32(sdma_v4_0_get_reg_offset(me, mmSDMA0_GFX_RB_WPTR)) >> 2; highbit = RREG32(sdma_v4_0_get_reg_offset(me, mmSDMA0_GFX_RB_WPTR_HI)) >> 2; @@ -285,12 +284,13 @@ static void sdma_v4_0_ring_set_wptr(struct amdgpu_ring *ring) WDOORBELL64(ring->doorbell_index, ring->wptr << 2); } else { int me = (ring == &ring->adev->sdma.instance[0].ring) ? 0 : 1; + DRM_DEBUG("Not using doorbell -- " "mmSDMA%i_GFX_RB_WPTR == 0x%08x " - "mmSDMA%i_GFX_RB_WPTR_HI == 0x%08x \n", - me, + "mmSDMA%i_GFX_RB_WPTR_HI == 0x%08x\n", me, lower_32_bits(ring->wptr << 2), + me, upper_32_bits(ring->wptr << 2)); WREG32(sdma_v4_0_get_reg_offset(me, mmSDMA0_GFX_RB_WPTR), lower_32_bits(ring->wptr << 2)); WREG32(sdma_v4_0_get_reg_offset(me, mmSDMA0_GFX_RB_WPTR_HI), upper_32_bits(ring->wptr << 2)); @@ -319,22 +319,22 @@ static void sdma_v4_0_ring_insert_nop(struct amdgpu_ring *ring, uint32_t count) * Schedule an IB in the DMA ring (VEGA10). */ static void sdma_v4_0_ring_emit_ib(struct amdgpu_ring *ring, - struct amdgpu_ib *ib, - unsigned vm_id, bool ctx_switch) + struct amdgpu_ib *ib, + unsigned vm_id, bool ctx_switch) { - u32 vmid = vm_id & 0xf; + u32 vmid = vm_id & 0xf; - /* IB packet must end on a 8 DW boundary */ - sdma_v4_0_ring_insert_nop(ring, (10 - (lower_32_bits(ring->wptr) & 7)) % 8); + /* IB packet must end on a 8 DW boundary */ + sdma_v4_0_ring_insert_nop(ring, (10 - (lower_32_bits(ring->wptr) & 7)) % 8); - amdgpu_ring_write(ring, SDMA_PKT_HEADER_OP(SDMA_OP_INDIRECT) | - SDMA_PKT_INDIRECT_HEADER_VMID(vmid)); - /* base must be 32 byte aligned */ - amdgpu_ring_write(ring, lower_32_bits(ib->gpu_addr) & 0xffffffe0); - amdgpu_ring_write(ring, upper_32_bits(ib->gpu_addr)); - amdgpu_ring_write(ring, ib->length_dw); - amdgpu_ring_write(ring, 0); - amdgpu_ring_write(ring, 0); + amdgpu_ring_write(ring, SDMA_PKT_HEADER_OP(SDMA_OP_INDIRECT) | + SDMA_PKT_INDIRECT_HEADER_VMID(vmid)); + /* base must be 32 byte aligned */ + amdgpu_ring_write(ring, lower_32_bits(ib->gpu_addr) & 0xffffffe0); + amdgpu_ring_write(ring, upper_32_bits(ib->gpu_addr)); + amdgpu_ring_write(ring, ib->length_dw); + amdgpu_ring_write(ring, 0); + amdgpu_ring_write(ring, 0); } @@ -523,7 +523,7 @@ static int sdma_v4_0_gfx_resume(struct amdgpu_device *adev) u32 doorbell; u32 doorbell_offset; u32 temp; - int i,r; + int i, r; for (i = 0; i < adev->sdma.num_instances; i++) { ring = &adev->sdma.instance[i].ring; @@ -572,7 +572,7 @@ static int sdma_v4_0_gfx_resume(struct amdgpu_device *adev) doorbell = RREG32(sdma_v4_0_get_reg_offset(i, mmSDMA0_GFX_DOORBELL)); doorbell_offset = RREG32(sdma_v4_0_get_reg_offset(i, mmSDMA0_GFX_DOORBELL_OFFSET)); - if (ring->use_doorbell){ + if (ring->use_doorbell) { doorbell = REG_SET_FIELD(doorbell, SDMA0_GFX_DOORBELL, ENABLE, 1); doorbell_offset = REG_SET_FIELD(doorbell_offset, SDMA0_GFX_DOORBELL_OFFSET, OFFSET, ring->doorbell_index); @@ -694,9 +694,7 @@ static int sdma_v4_0_load_microcode(struct amdgpu_device *adev) for (j = 0; j < fw_size; j++) - { WREG32(sdma_v4_0_get_reg_offset(i, mmSDMA0_UCODE_DATA), le32_to_cpup(fw_data++)); - } WREG32(sdma_v4_0_get_reg_offset(i, mmSDMA0_UCODE_ADDR), adev->sdma.instance[i].fw_version); } @@ -744,10 +742,8 @@ static int sdma_v4_0_start(struct amdgpu_device *adev) if (r) return r; r = sdma_v4_0_rlc_resume(adev); - if (r) - return r; - return 0; + return r; } /** @@ -797,9 +793,8 @@ static int sdma_v4_0_ring_test_ring(struct amdgpu_ring *ring) for (i = 0; i < adev->usec_timeout; i++) { tmp = le32_to_cpu(adev->wb.wb[index]); - if (tmp == 0xDEADBEEF) { + if (tmp == 0xDEADBEEF) break; - } DRM_UDELAY(1); } @@ -864,29 +859,29 @@ static int sdma_v4_0_ring_test_ib(struct amdgpu_ring *ring, long timeout) if (r) goto err1; - r = dma_fence_wait_timeout(f, false, timeout); - if (r == 0) { - DRM_ERROR("amdgpu: IB test timed out\n"); - r = -ETIMEDOUT; - goto err1; - } else if (r < 0) { - DRM_ERROR("amdgpu: fence wait failed (%ld).\n", r); - goto err1; - } - tmp = le32_to_cpu(adev->wb.wb[index]); - if (tmp == 0xDEADBEEF) { - DRM_INFO("ib test on ring %d succeeded\n", ring->idx); - r = 0; - } else { - DRM_ERROR("amdgpu: ib test failed (0x%08X)\n", tmp); - r = -EINVAL; - } + r = dma_fence_wait_timeout(f, false, timeout); + if (r == 0) { + DRM_ERROR("amdgpu: IB test timed out\n"); + r = -ETIMEDOUT; + goto err1; + } else if (r < 0) { + DRM_ERROR("amdgpu: fence wait failed (%ld).\n", r); + goto err1; + } + tmp = le32_to_cpu(adev->wb.wb[index]); + if (tmp == 0xDEADBEEF) { + DRM_INFO("ib test on ring %d succeeded\n", ring->idx); + r = 0; + } else { + DRM_ERROR("amdgpu: ib test failed (0x%08X)\n", tmp); + r = -EINVAL; + } err1: - amdgpu_ib_free(adev, &ib, NULL); - dma_fence_put(f); + amdgpu_ib_free(adev, &ib, NULL); + dma_fence_put(f); err0: - amdgpu_wb_free(adev, index); - return r; + amdgpu_wb_free(adev, index); + return r; } @@ -1039,44 +1034,40 @@ static void sdma_v4_0_ring_emit_pipeline_sync(struct amdgpu_ring *ring) static void sdma_v4_0_ring_emit_vm_flush(struct amdgpu_ring *ring, unsigned vm_id, uint64_t pd_addr) { + struct amdgpu_vmhub *hub = &ring->adev->vmhub[ring->funcs->vmhub]; uint32_t req = ring->adev->gart.gart_funcs->get_invalidate_req(vm_id); - unsigned eng = ring->idx; - unsigned i; + unsigned eng = ring->vm_inv_eng; pd_addr = pd_addr | 0x1; /* valid bit */ /* now only use physical base address of PDE and valid */ BUG_ON(pd_addr & 0xFFFF00000000003EULL); - for (i = 0; i < AMDGPU_MAX_VMHUBS; ++i) { - struct amdgpu_vmhub *hub = &ring->adev->vmhub[i]; - - amdgpu_ring_write(ring, SDMA_PKT_HEADER_OP(SDMA_OP_SRBM_WRITE) | - SDMA_PKT_SRBM_WRITE_HEADER_BYTE_EN(0xf)); - amdgpu_ring_write(ring, hub->ctx0_ptb_addr_lo32 + vm_id * 2); - amdgpu_ring_write(ring, lower_32_bits(pd_addr)); - - amdgpu_ring_write(ring, SDMA_PKT_HEADER_OP(SDMA_OP_SRBM_WRITE) | - SDMA_PKT_SRBM_WRITE_HEADER_BYTE_EN(0xf)); - amdgpu_ring_write(ring, hub->ctx0_ptb_addr_hi32 + vm_id * 2); - amdgpu_ring_write(ring, upper_32_bits(pd_addr)); - - /* flush TLB */ - amdgpu_ring_write(ring, SDMA_PKT_HEADER_OP(SDMA_OP_SRBM_WRITE) | - SDMA_PKT_SRBM_WRITE_HEADER_BYTE_EN(0xf)); - amdgpu_ring_write(ring, hub->vm_inv_eng0_req + eng); - amdgpu_ring_write(ring, req); - - /* wait for flush */ - amdgpu_ring_write(ring, SDMA_PKT_HEADER_OP(SDMA_OP_POLL_REGMEM) | - SDMA_PKT_POLL_REGMEM_HEADER_HDP_FLUSH(0) | - SDMA_PKT_POLL_REGMEM_HEADER_FUNC(3)); /* equal */ - amdgpu_ring_write(ring, (hub->vm_inv_eng0_ack + eng) << 2); - amdgpu_ring_write(ring, 0); - amdgpu_ring_write(ring, 1 << vm_id); /* reference */ - amdgpu_ring_write(ring, 1 << vm_id); /* mask */ - amdgpu_ring_write(ring, SDMA_PKT_POLL_REGMEM_DW5_RETRY_COUNT(0xfff) | - SDMA_PKT_POLL_REGMEM_DW5_INTERVAL(10)); - } + amdgpu_ring_write(ring, SDMA_PKT_HEADER_OP(SDMA_OP_SRBM_WRITE) | + SDMA_PKT_SRBM_WRITE_HEADER_BYTE_EN(0xf)); + amdgpu_ring_write(ring, hub->ctx0_ptb_addr_lo32 + vm_id * 2); + amdgpu_ring_write(ring, lower_32_bits(pd_addr)); + + amdgpu_ring_write(ring, SDMA_PKT_HEADER_OP(SDMA_OP_SRBM_WRITE) | + SDMA_PKT_SRBM_WRITE_HEADER_BYTE_EN(0xf)); + amdgpu_ring_write(ring, hub->ctx0_ptb_addr_hi32 + vm_id * 2); + amdgpu_ring_write(ring, upper_32_bits(pd_addr)); + + /* flush TLB */ + amdgpu_ring_write(ring, SDMA_PKT_HEADER_OP(SDMA_OP_SRBM_WRITE) | + SDMA_PKT_SRBM_WRITE_HEADER_BYTE_EN(0xf)); + amdgpu_ring_write(ring, hub->vm_inv_eng0_req + eng); + amdgpu_ring_write(ring, req); + + /* wait for flush */ + amdgpu_ring_write(ring, SDMA_PKT_HEADER_OP(SDMA_OP_POLL_REGMEM) | + SDMA_PKT_POLL_REGMEM_HEADER_HDP_FLUSH(0) | + SDMA_PKT_POLL_REGMEM_HEADER_FUNC(3)); /* equal */ + amdgpu_ring_write(ring, (hub->vm_inv_eng0_ack + eng) << 2); + amdgpu_ring_write(ring, 0); + amdgpu_ring_write(ring, 1 << vm_id); /* reference */ + amdgpu_ring_write(ring, 1 << vm_id); /* mask */ + amdgpu_ring_write(ring, SDMA_PKT_POLL_REGMEM_DW5_RETRY_COUNT(0xfff) | + SDMA_PKT_POLL_REGMEM_DW5_INTERVAL(10)); } static int sdma_v4_0_early_init(void *handle) @@ -1162,8 +1153,6 @@ static int sdma_v4_0_hw_init(void *handle) sdma_v4_0_init_golden_registers(adev); r = sdma_v4_0_start(adev); - if (r) - return r; return r; } @@ -1199,10 +1188,12 @@ static bool sdma_v4_0_is_idle(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; u32 i; + for (i = 0; i < adev->sdma.num_instances; i++) { u32 tmp = RREG32(sdma_v4_0_get_reg_offset(i, mmSDMA0_STATUS_REG)); + if (!(tmp & SDMA0_STATUS_REG__IDLE_MASK)) - return false; + return false; } return true; @@ -1211,8 +1202,9 @@ static bool sdma_v4_0_is_idle(void *handle) static int sdma_v4_0_wait_for_idle(void *handle) { unsigned i; - u32 sdma0,sdma1; + u32 sdma0, sdma1; struct amdgpu_device *adev = (struct amdgpu_device *)handle; + for (i = 0; i < adev->usec_timeout; i++) { sdma0 = RREG32(sdma_v4_0_get_reg_offset(0, mmSDMA0_STATUS_REG)); sdma1 = RREG32(sdma_v4_0_get_reg_offset(1, mmSDMA0_STATUS_REG)); @@ -1240,7 +1232,7 @@ static int sdma_v4_0_set_trap_irq_state(struct amdgpu_device *adev, u32 reg_offset = (type == AMDGPU_SDMA_IRQ_TRAP0) ? sdma_v4_0_get_reg_offset(0, mmSDMA0_CNTL) : - sdma_v4_0_get_reg_offset(1, mmSDMA0_CNTL); + sdma_v4_0_get_reg_offset(1, mmSDMA0_CNTL); sdma_cntl = RREG32(reg_offset); sdma_cntl = REG_SET_FIELD(sdma_cntl, SDMA0_CNTL, TRAP_ENABLE, @@ -1332,7 +1324,7 @@ static void sdma_v4_0_update_medium_grain_clock_gating( SDMA1_CLK_CTRL__SOFT_OVERRIDE2_MASK | SDMA1_CLK_CTRL__SOFT_OVERRIDE1_MASK | SDMA1_CLK_CTRL__SOFT_OVERRIDE0_MASK); - if(def != data) + if (def != data) WREG32(SOC15_REG_OFFSET(SDMA1, 0, mmSDMA1_CLK_CTRL), data); } } else { @@ -1382,17 +1374,17 @@ static void sdma_v4_0_update_medium_grain_light_sleep( /* 1-not override: enable sdma1 mem light sleep */ if (adev->asic_type == CHIP_VEGA10) { - def = data = RREG32(SOC15_REG_OFFSET(SDMA1, 0, mmSDMA1_POWER_CNTL)); - data |= SDMA1_POWER_CNTL__MEM_POWER_OVERRIDE_MASK; - if (def != data) - WREG32(SOC15_REG_OFFSET(SDMA1, 0, mmSDMA1_POWER_CNTL), data); + def = data = RREG32(SOC15_REG_OFFSET(SDMA1, 0, mmSDMA1_POWER_CNTL)); + data |= SDMA1_POWER_CNTL__MEM_POWER_OVERRIDE_MASK; + if (def != data) + WREG32(SOC15_REG_OFFSET(SDMA1, 0, mmSDMA1_POWER_CNTL), data); } } else { /* 0-override:disable sdma0 mem light sleep */ def = data = RREG32(SOC15_REG_OFFSET(SDMA0, 0, mmSDMA0_POWER_CNTL)); data &= ~SDMA0_POWER_CNTL__MEM_POWER_OVERRIDE_MASK; if (def != data) - WREG32(SOC15_REG_OFFSET(SDMA0, 0, mmSDMA0_POWER_CNTL), data); + WREG32(SOC15_REG_OFFSET(SDMA0, 0, mmSDMA0_POWER_CNTL), data); /* 0-override:disable sdma1 mem light sleep */ if (adev->asic_type == CHIP_VEGA10) { @@ -1473,6 +1465,7 @@ static const struct amdgpu_ring_funcs sdma_v4_0_ring_funcs = { .align_mask = 0xf, .nop = SDMA_PKT_NOP_HEADER_OP(SDMA_OP_NOP), .support_64bit_ptrs = true, + .vmhub = AMDGPU_MMHUB, .get_rptr = sdma_v4_0_ring_get_rptr, .get_wptr = sdma_v4_0_ring_get_wptr, .set_wptr = sdma_v4_0_ring_set_wptr, @@ -1480,7 +1473,7 @@ static const struct amdgpu_ring_funcs sdma_v4_0_ring_funcs = { 6 + /* sdma_v4_0_ring_emit_hdp_flush */ 3 + /* sdma_v4_0_ring_emit_hdp_invalidate */ 6 + /* sdma_v4_0_ring_emit_pipeline_sync */ - 36 + /* sdma_v4_0_ring_emit_vm_flush */ + 18 + /* sdma_v4_0_ring_emit_vm_flush */ 10 + 10 + 10, /* sdma_v4_0_ring_emit_fence x3 for user fence, vm fence */ .emit_ib_size = 7 + 6, /* sdma_v4_0_ring_emit_ib */ .emit_ib = sdma_v4_0_ring_emit_ib, @@ -1606,8 +1599,7 @@ static void sdma_v4_0_set_vm_pte_funcs(struct amdgpu_device *adev) } } -const struct amdgpu_ip_block_version sdma_v4_0_ip_block = -{ +const struct amdgpu_ip_block_version sdma_v4_0_ip_block = { .type = AMD_IP_BLOCK_TYPE_SDMA, .major = 4, .minor = 0, diff --git a/drivers/gpu/drm/amd/amdgpu/si.c b/drivers/gpu/drm/amd/amdgpu/si.c index c0b1aabf282f..2431639baf47 100644 --- a/drivers/gpu/drm/amd/amdgpu/si.c +++ b/drivers/gpu/drm/amd/amdgpu/si.c @@ -24,7 +24,7 @@ #include <linux/firmware.h> #include <linux/slab.h> #include <linux/module.h> -#include "drmP.h" +#include <drm/drmP.h> #include "amdgpu.h" #include "amdgpu_atombios.h" #include "amdgpu_ih.h" diff --git a/drivers/gpu/drm/amd/amdgpu/si_dpm.c b/drivers/gpu/drm/amd/amdgpu/si_dpm.c index 7c1c5d127281..a7ad8390981c 100644 --- a/drivers/gpu/drm/amd/amdgpu/si_dpm.c +++ b/drivers/gpu/drm/amd/amdgpu/si_dpm.c @@ -21,7 +21,7 @@ * */ -#include "drmP.h" +#include <drm/drmP.h> #include "amdgpu.h" #include "amdgpu_pm.h" #include "amdgpu_dpm.h" diff --git a/drivers/gpu/drm/amd/amdgpu/si_ih.c b/drivers/gpu/drm/amd/amdgpu/si_ih.c index e66084211c74..ce25e03a077d 100644 --- a/drivers/gpu/drm/amd/amdgpu/si_ih.c +++ b/drivers/gpu/drm/amd/amdgpu/si_ih.c @@ -20,7 +20,7 @@ * OTHER DEALINGS IN THE SOFTWARE. * */ -#include "drmP.h" +#include <drm/drmP.h> #include "amdgpu.h" #include "amdgpu_ih.h" #include "sid.h" diff --git a/drivers/gpu/drm/amd/amdgpu/si_smc.c b/drivers/gpu/drm/amd/amdgpu/si_smc.c index 0726bc3b6f90..4a2fd8b61940 100644 --- a/drivers/gpu/drm/amd/amdgpu/si_smc.c +++ b/drivers/gpu/drm/amd/amdgpu/si_smc.c @@ -23,7 +23,7 @@ */ #include <linux/firmware.h> -#include "drmP.h" +#include <drm/drmP.h> #include "amdgpu.h" #include "sid.h" #include "ppsmc.h" diff --git a/drivers/gpu/drm/amd/amdgpu/soc15.c b/drivers/gpu/drm/amd/amdgpu/soc15.c index 385de8617075..e945f8b07487 100644 --- a/drivers/gpu/drm/amd/amdgpu/soc15.c +++ b/drivers/gpu/drm/amd/amdgpu/soc15.c @@ -23,9 +23,9 @@ #include <linux/firmware.h> #include <linux/slab.h> #include <linux/module.h> -#include "drmP.h" +#include <drm/drmP.h> #include "amdgpu.h" -#include "amdgpu_atombios.h" +#include "amdgpu_atomfirmware.h" #include "amdgpu_ih.h" #include "amdgpu_uvd.h" #include "amdgpu_vce.h" @@ -405,11 +405,11 @@ static void soc15_gpu_pci_config_reset(struct amdgpu_device *adev) static int soc15_asic_reset(struct amdgpu_device *adev) { - amdgpu_atombios_scratch_regs_engine_hung(adev, true); + amdgpu_atomfirmware_scratch_regs_engine_hung(adev, true); soc15_gpu_pci_config_reset(adev); - amdgpu_atombios_scratch_regs_engine_hung(adev, false); + amdgpu_atomfirmware_scratch_regs_engine_hung(adev, false); return 0; } @@ -505,8 +505,7 @@ int soc15_set_ip_blocks(struct amdgpu_device *adev) amdgpu_ip_block_add(adev, &dce_virtual_ip_block); amdgpu_ip_block_add(adev, &gfx_v9_0_ip_block); amdgpu_ip_block_add(adev, &sdma_v4_0_ip_block); - if (!amdgpu_sriov_vf(adev)) - amdgpu_ip_block_add(adev, &uvd_v7_0_ip_block); + amdgpu_ip_block_add(adev, &uvd_v7_0_ip_block); amdgpu_ip_block_add(adev, &vce_v4_0_ip_block); break; default: diff --git a/drivers/gpu/drm/amd/amdgpu/soc15_common.h b/drivers/gpu/drm/amd/amdgpu/soc15_common.h index 2b96c806baa1..e8df6d820dbe 100644 --- a/drivers/gpu/drm/amd/amdgpu/soc15_common.h +++ b/drivers/gpu/drm/amd/amdgpu/soc15_common.h @@ -45,13 +45,31 @@ struct nbio_pcie_index_data { u32 index_offset; u32 data_offset; }; -// Register Access Macro + +/* Register Access Macros */ #define SOC15_REG_OFFSET(ip, inst, reg) (0 == reg##_BASE_IDX ? ip##_BASE__INST##inst##_SEG0 + reg : \ (1 == reg##_BASE_IDX ? ip##_BASE__INST##inst##_SEG1 + reg : \ (2 == reg##_BASE_IDX ? ip##_BASE__INST##inst##_SEG2 + reg : \ (3 == reg##_BASE_IDX ? ip##_BASE__INST##inst##_SEG3 + reg : \ (ip##_BASE__INST##inst##_SEG4 + reg))))) +#define WREG32_FIELD15(ip, idx, reg, field, val) \ + WREG32(SOC15_REG_OFFSET(ip, idx, mm##reg), (RREG32(SOC15_REG_OFFSET(ip, idx, mm##reg)) & ~REG_FIELD_MASK(reg, field)) | (val) << REG_FIELD_SHIFT(reg, field)) + +#define RREG32_SOC15(ip, inst, reg) \ + RREG32( (0 == reg##_BASE_IDX ? ip##_BASE__INST##inst##_SEG0 + reg : \ + (1 == reg##_BASE_IDX ? ip##_BASE__INST##inst##_SEG1 + reg : \ + (2 == reg##_BASE_IDX ? ip##_BASE__INST##inst##_SEG2 + reg : \ + (3 == reg##_BASE_IDX ? ip##_BASE__INST##inst##_SEG3 + reg : \ + (ip##_BASE__INST##inst##_SEG4 + reg)))))) + +#define WREG32_SOC15(ip, inst, reg, value) \ + WREG32( (0 == reg##_BASE_IDX ? ip##_BASE__INST##inst##_SEG0 + reg : \ + (1 == reg##_BASE_IDX ? ip##_BASE__INST##inst##_SEG1 + reg : \ + (2 == reg##_BASE_IDX ? ip##_BASE__INST##inst##_SEG2 + reg : \ + (3 == reg##_BASE_IDX ? ip##_BASE__INST##inst##_SEG3 + reg : \ + (ip##_BASE__INST##inst##_SEG4 + reg))))), value) + #endif diff --git a/drivers/gpu/drm/amd/amdgpu/tonga_ih.c b/drivers/gpu/drm/amd/amdgpu/tonga_ih.c index 3a5097ac2bb4..923df2c0e535 100644 --- a/drivers/gpu/drm/amd/amdgpu/tonga_ih.c +++ b/drivers/gpu/drm/amd/amdgpu/tonga_ih.c @@ -20,7 +20,7 @@ * OTHER DEALINGS IN THE SOFTWARE. * */ -#include "drmP.h" +#include <drm/drmP.h> #include "amdgpu.h" #include "amdgpu_ih.h" #include "vid.h" diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c index 9bcf01469282..eca8f6e01e97 100644 --- a/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c +++ b/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c @@ -27,10 +27,14 @@ #include "amdgpu_uvd.h" #include "soc15d.h" #include "soc15_common.h" +#include "mmsch_v1_0.h" #include "vega10/soc15ip.h" #include "vega10/UVD/uvd_7_0_offset.h" #include "vega10/UVD/uvd_7_0_sh_mask.h" +#include "vega10/VCE/vce_4_0_offset.h" +#include "vega10/VCE/vce_4_0_default.h" +#include "vega10/VCE/vce_4_0_sh_mask.h" #include "vega10/NBIF/nbif_6_1_offset.h" #include "vega10/HDP/hdp_4_0_offset.h" #include "vega10/MMHUB/mmhub_1_0_offset.h" @@ -41,6 +45,7 @@ static void uvd_v7_0_set_enc_ring_funcs(struct amdgpu_device *adev); static void uvd_v7_0_set_irq_funcs(struct amdgpu_device *adev); static int uvd_v7_0_start(struct amdgpu_device *adev); static void uvd_v7_0_stop(struct amdgpu_device *adev); +static int uvd_v7_0_sriov_start(struct amdgpu_device *adev); /** * uvd_v7_0_ring_get_rptr - get read pointer @@ -98,6 +103,9 @@ static uint64_t uvd_v7_0_enc_ring_get_wptr(struct amdgpu_ring *ring) { struct amdgpu_device *adev = ring->adev; + if (ring->use_doorbell) + return adev->wb.wb[ring->wptr_offs]; + if (ring == &adev->uvd.ring_enc[0]) return RREG32(SOC15_REG_OFFSET(UVD, 0, mmUVD_RB_WPTR)); else @@ -129,6 +137,13 @@ static void uvd_v7_0_enc_ring_set_wptr(struct amdgpu_ring *ring) { struct amdgpu_device *adev = ring->adev; + if (ring->use_doorbell) { + /* XXX check if swapping is necessary on BE */ + adev->wb.wb[ring->wptr_offs] = lower_32_bits(ring->wptr); + WDOORBELL32(ring->doorbell_index, lower_32_bits(ring->wptr)); + return; + } + if (ring == &adev->uvd.ring_enc[0]) WREG32(SOC15_REG_OFFSET(UVD, 0, mmUVD_RB_WPTR), lower_32_bits(ring->wptr)); @@ -353,7 +368,10 @@ static int uvd_v7_0_early_init(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; - adev->uvd.num_enc_rings = 2; + if (amdgpu_sriov_vf(adev)) + adev->uvd.num_enc_rings = 1; + else + adev->uvd.num_enc_rings = 2; uvd_v7_0_set_ring_funcs(adev); uvd_v7_0_set_enc_ring_funcs(adev); uvd_v7_0_set_irq_funcs(adev); @@ -406,21 +424,31 @@ static int uvd_v7_0_sw_init(void *handle) r = amdgpu_uvd_resume(adev); if (r) return r; + if (!amdgpu_sriov_vf(adev)) { + ring = &adev->uvd.ring; + sprintf(ring->name, "uvd"); + r = amdgpu_ring_init(adev, ring, 512, &adev->uvd.irq, 0); + if (r) + return r; + } - ring = &adev->uvd.ring; - sprintf(ring->name, "uvd"); - r = amdgpu_ring_init(adev, ring, 512, &adev->uvd.irq, 0); - if (r) - return r; for (i = 0; i < adev->uvd.num_enc_rings; ++i) { ring = &adev->uvd.ring_enc[i]; sprintf(ring->name, "uvd_enc%d", i); + if (amdgpu_sriov_vf(adev)) { + ring->use_doorbell = true; + ring->doorbell_index = AMDGPU_DOORBELL64_UVD_RING0_1 * 2; + } r = amdgpu_ring_init(adev, ring, 512, &adev->uvd.irq, 0); if (r) return r; } + r = amdgpu_virt_alloc_mm_table(adev); + if (r) + return r; + return r; } @@ -429,6 +457,8 @@ static int uvd_v7_0_sw_fini(void *handle) int i, r; struct amdgpu_device *adev = (struct amdgpu_device *)handle; + amdgpu_virt_free_mm_table(adev); + r = amdgpu_uvd_suspend(adev); if (r) return r; @@ -455,48 +485,53 @@ static int uvd_v7_0_hw_init(void *handle) uint32_t tmp; int i, r; - r = uvd_v7_0_start(adev); + if (amdgpu_sriov_vf(adev)) + r = uvd_v7_0_sriov_start(adev); + else + r = uvd_v7_0_start(adev); if (r) goto done; - ring->ready = true; - r = amdgpu_ring_test_ring(ring); - if (r) { - ring->ready = false; - goto done; - } + if (!amdgpu_sriov_vf(adev)) { + ring->ready = true; + r = amdgpu_ring_test_ring(ring); + if (r) { + ring->ready = false; + goto done; + } - r = amdgpu_ring_alloc(ring, 10); - if (r) { - DRM_ERROR("amdgpu: ring failed to lock UVD ring (%d).\n", r); - goto done; - } + r = amdgpu_ring_alloc(ring, 10); + if (r) { + DRM_ERROR("amdgpu: ring failed to lock UVD ring (%d).\n", r); + goto done; + } - tmp = PACKET0(SOC15_REG_OFFSET(UVD, 0, - mmUVD_SEMA_WAIT_FAULT_TIMEOUT_CNTL), 0); - amdgpu_ring_write(ring, tmp); - amdgpu_ring_write(ring, 0xFFFFF); + tmp = PACKET0(SOC15_REG_OFFSET(UVD, 0, + mmUVD_SEMA_WAIT_FAULT_TIMEOUT_CNTL), 0); + amdgpu_ring_write(ring, tmp); + amdgpu_ring_write(ring, 0xFFFFF); - tmp = PACKET0(SOC15_REG_OFFSET(UVD, 0, - mmUVD_SEMA_WAIT_INCOMPLETE_TIMEOUT_CNTL), 0); - amdgpu_ring_write(ring, tmp); - amdgpu_ring_write(ring, 0xFFFFF); + tmp = PACKET0(SOC15_REG_OFFSET(UVD, 0, + mmUVD_SEMA_WAIT_INCOMPLETE_TIMEOUT_CNTL), 0); + amdgpu_ring_write(ring, tmp); + amdgpu_ring_write(ring, 0xFFFFF); - tmp = PACKET0(SOC15_REG_OFFSET(UVD, 0, - mmUVD_SEMA_SIGNAL_INCOMPLETE_TIMEOUT_CNTL), 0); - amdgpu_ring_write(ring, tmp); - amdgpu_ring_write(ring, 0xFFFFF); + tmp = PACKET0(SOC15_REG_OFFSET(UVD, 0, + mmUVD_SEMA_SIGNAL_INCOMPLETE_TIMEOUT_CNTL), 0); + amdgpu_ring_write(ring, tmp); + amdgpu_ring_write(ring, 0xFFFFF); - /* Clear timeout status bits */ - amdgpu_ring_write(ring, PACKET0(SOC15_REG_OFFSET(UVD, 0, - mmUVD_SEMA_TIMEOUT_STATUS), 0)); - amdgpu_ring_write(ring, 0x8); + /* Clear timeout status bits */ + amdgpu_ring_write(ring, PACKET0(SOC15_REG_OFFSET(UVD, 0, + mmUVD_SEMA_TIMEOUT_STATUS), 0)); + amdgpu_ring_write(ring, 0x8); - amdgpu_ring_write(ring, PACKET0(SOC15_REG_OFFSET(UVD, 0, - mmUVD_SEMA_CNTL), 0)); - amdgpu_ring_write(ring, 3); + amdgpu_ring_write(ring, PACKET0(SOC15_REG_OFFSET(UVD, 0, + mmUVD_SEMA_CNTL), 0)); + amdgpu_ring_write(ring, 3); - amdgpu_ring_commit(ring); + amdgpu_ring_commit(ring); + } for (i = 0; i < adev->uvd.num_enc_rings; ++i) { ring = &adev->uvd.ring_enc[i]; @@ -618,6 +653,241 @@ static void uvd_v7_0_mc_resume(struct amdgpu_device *adev) WREG32(SOC15_REG_OFFSET(UVD, 0, mmUVD_GP_SCRATCH4), adev->uvd.max_handles); } +static int uvd_v7_0_mmsch_start(struct amdgpu_device *adev, + struct amdgpu_mm_table *table) +{ + uint32_t data = 0, loop; + uint64_t addr = table->gpu_addr; + struct mmsch_v1_0_init_header *header = (struct mmsch_v1_0_init_header *)table->cpu_addr; + uint32_t size; + + size = header->header_size + header->vce_table_size + header->uvd_table_size; + + /* 1, write to vce_mmsch_vf_ctx_addr_lo/hi register with GPU mc addr of memory descriptor location */ + WREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_MMSCH_VF_CTX_ADDR_LO), lower_32_bits(addr)); + WREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_MMSCH_VF_CTX_ADDR_HI), upper_32_bits(addr)); + + /* 2, update vmid of descriptor */ + data = RREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_MMSCH_VF_VMID)); + data &= ~VCE_MMSCH_VF_VMID__VF_CTX_VMID_MASK; + data |= (0 << VCE_MMSCH_VF_VMID__VF_CTX_VMID__SHIFT); /* use domain0 for MM scheduler */ + WREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_MMSCH_VF_VMID), data); + + /* 3, notify mmsch about the size of this descriptor */ + WREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_MMSCH_VF_CTX_SIZE), size); + + /* 4, set resp to zero */ + WREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_MMSCH_VF_MAILBOX_RESP), 0); + + /* 5, kick off the initialization and wait until VCE_MMSCH_VF_MAILBOX_RESP becomes non-zero */ + WREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_MMSCH_VF_MAILBOX_HOST), 0x10000001); + + data = RREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_MMSCH_VF_MAILBOX_RESP)); + loop = 1000; + while ((data & 0x10000002) != 0x10000002) { + udelay(10); + data = RREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_MMSCH_VF_MAILBOX_RESP)); + loop--; + if (!loop) + break; + } + + if (!loop) { + dev_err(adev->dev, "failed to init MMSCH, mmVCE_MMSCH_VF_MAILBOX_RESP = %x\n", data); + return -EBUSY; + } + + return 0; +} + +static int uvd_v7_0_sriov_start(struct amdgpu_device *adev) +{ + struct amdgpu_ring *ring; + uint32_t offset, size, tmp; + uint32_t table_size = 0; + struct mmsch_v1_0_cmd_direct_write direct_wt = { {0} }; + struct mmsch_v1_0_cmd_direct_read_modify_write direct_rd_mod_wt = { {0} }; + struct mmsch_v1_0_cmd_direct_polling direct_poll = { {0} }; + struct mmsch_v1_0_cmd_end end = { {0} }; + uint32_t *init_table = adev->virt.mm_table.cpu_addr; + struct mmsch_v1_0_init_header *header = (struct mmsch_v1_0_init_header *)init_table; + + direct_wt.cmd_header.command_type = MMSCH_COMMAND__DIRECT_REG_WRITE; + direct_rd_mod_wt.cmd_header.command_type = MMSCH_COMMAND__DIRECT_REG_READ_MODIFY_WRITE; + direct_poll.cmd_header.command_type = MMSCH_COMMAND__DIRECT_REG_POLLING; + end.cmd_header.command_type = MMSCH_COMMAND__END; + + if (header->uvd_table_offset == 0 && header->uvd_table_size == 0) { + header->version = MMSCH_VERSION; + header->header_size = sizeof(struct mmsch_v1_0_init_header) >> 2; + + if (header->vce_table_offset == 0 && header->vce_table_size == 0) + header->uvd_table_offset = header->header_size; + else + header->uvd_table_offset = header->vce_table_size + header->vce_table_offset; + + init_table += header->uvd_table_offset; + + ring = &adev->uvd.ring; + size = AMDGPU_GPU_PAGE_ALIGN(adev->uvd.fw->size + 4); + + /* disable clock gating */ + MMSCH_V1_0_INSERT_DIRECT_RD_MOD_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_POWER_STATUS), + ~UVD_POWER_STATUS__UVD_PG_MODE_MASK, 0); + MMSCH_V1_0_INSERT_DIRECT_RD_MOD_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_STATUS), + 0xFFFFFFFF, 0x00000004); + /* mc resume*/ + if (adev->firmware.load_type == AMDGPU_FW_LOAD_PSP) { + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_LOW), + lower_32_bits(adev->firmware.ucode[AMDGPU_UCODE_ID_UVD].mc_addr)); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_HIGH), + upper_32_bits(adev->firmware.ucode[AMDGPU_UCODE_ID_UVD].mc_addr)); + offset = 0; + } else { + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_LOW), + lower_32_bits(adev->uvd.gpu_addr)); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_HIGH), + upper_32_bits(adev->uvd.gpu_addr)); + offset = size; + } + + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_VCPU_CACHE_OFFSET0), + AMDGPU_UVD_FIRMWARE_OFFSET >> 3); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_VCPU_CACHE_SIZE0), size); + + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_VCPU_CACHE1_64BIT_BAR_LOW), + lower_32_bits(adev->uvd.gpu_addr + offset)); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_VCPU_CACHE1_64BIT_BAR_HIGH), + upper_32_bits(adev->uvd.gpu_addr + offset)); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_VCPU_CACHE_OFFSET1), (1 << 21)); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_VCPU_CACHE_SIZE1), AMDGPU_UVD_HEAP_SIZE); + + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_VCPU_CACHE2_64BIT_BAR_LOW), + lower_32_bits(adev->uvd.gpu_addr + offset + AMDGPU_UVD_HEAP_SIZE)); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_VCPU_CACHE2_64BIT_BAR_HIGH), + upper_32_bits(adev->uvd.gpu_addr + offset + AMDGPU_UVD_HEAP_SIZE)); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_VCPU_CACHE_OFFSET2), (2 << 21)); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_VCPU_CACHE_SIZE2), + AMDGPU_UVD_STACK_SIZE + (AMDGPU_UVD_SESSION_SIZE * 40)); + + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_UDEC_ADDR_CONFIG), + adev->gfx.config.gb_addr_config); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_UDEC_DB_ADDR_CONFIG), + adev->gfx.config.gb_addr_config); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_UDEC_DBW_ADDR_CONFIG), + adev->gfx.config.gb_addr_config); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_GP_SCRATCH4), adev->uvd.max_handles); + /* mc resume end*/ + + /* disable clock gating */ + MMSCH_V1_0_INSERT_DIRECT_RD_MOD_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_CGC_CTRL), + ~UVD_CGC_CTRL__DYN_CLOCK_MODE_MASK, 0); + + /* disable interupt */ + MMSCH_V1_0_INSERT_DIRECT_RD_MOD_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_MASTINT_EN), + ~UVD_MASTINT_EN__VCPU_EN_MASK, 0); + + /* stall UMC and register bus before resetting VCPU */ + MMSCH_V1_0_INSERT_DIRECT_RD_MOD_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_CTRL2), + ~UVD_LMI_CTRL2__STALL_ARB_UMC_MASK, + UVD_LMI_CTRL2__STALL_ARB_UMC_MASK); + + /* put LMI, VCPU, RBC etc... into reset */ + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_SOFT_RESET), + (uint32_t)(UVD_SOFT_RESET__LMI_SOFT_RESET_MASK | + UVD_SOFT_RESET__VCPU_SOFT_RESET_MASK | + UVD_SOFT_RESET__LBSI_SOFT_RESET_MASK | + UVD_SOFT_RESET__RBC_SOFT_RESET_MASK | + UVD_SOFT_RESET__CSM_SOFT_RESET_MASK | + UVD_SOFT_RESET__CXW_SOFT_RESET_MASK | + UVD_SOFT_RESET__TAP_SOFT_RESET_MASK | + UVD_SOFT_RESET__LMI_UMC_SOFT_RESET_MASK)); + + /* initialize UVD memory controller */ + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_CTRL), + (uint32_t)((0x40 << UVD_LMI_CTRL__WRITE_CLEAN_TIMER__SHIFT) | + UVD_LMI_CTRL__WRITE_CLEAN_TIMER_EN_MASK | + UVD_LMI_CTRL__DATA_COHERENCY_EN_MASK | + UVD_LMI_CTRL__VCPU_DATA_COHERENCY_EN_MASK | + UVD_LMI_CTRL__REQ_MODE_MASK | + 0x00100000L)); + + /* disable byte swapping */ + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_SWAP_CNTL), 0); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_MP_SWAP_CNTL), 0); + + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_MPC_SET_MUXA0), 0x40c2040); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_MPC_SET_MUXA1), 0x0); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_MPC_SET_MUXB0), 0x40c2040); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_MPC_SET_MUXB1), 0x0); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_MPC_SET_ALU), 0); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_MPC_SET_MUX), 0x88); + + /* take all subblocks out of reset, except VCPU */ + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_SOFT_RESET), + UVD_SOFT_RESET__VCPU_SOFT_RESET_MASK); + + /* enable VCPU clock */ + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_VCPU_CNTL), + UVD_VCPU_CNTL__CLK_EN_MASK); + + /* enable UMC */ + MMSCH_V1_0_INSERT_DIRECT_RD_MOD_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_CTRL2), + ~UVD_LMI_CTRL2__STALL_ARB_UMC_MASK, 0); + + /* boot up the VCPU */ + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_SOFT_RESET), 0); + + MMSCH_V1_0_INSERT_DIRECT_POLL(SOC15_REG_OFFSET(UVD, 0, mmUVD_STATUS), 0x02, 0x02); + + /* enable master interrupt */ + MMSCH_V1_0_INSERT_DIRECT_RD_MOD_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_MASTINT_EN), + ~(UVD_MASTINT_EN__VCPU_EN_MASK|UVD_MASTINT_EN__SYS_EN_MASK), + (UVD_MASTINT_EN__VCPU_EN_MASK|UVD_MASTINT_EN__SYS_EN_MASK)); + + /* clear the bit 4 of UVD_STATUS */ + MMSCH_V1_0_INSERT_DIRECT_RD_MOD_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_STATUS), + ~(2 << UVD_STATUS__VCPU_REPORT__SHIFT), 0); + + /* force RBC into idle state */ + size = order_base_2(ring->ring_size); + tmp = REG_SET_FIELD(0, UVD_RBC_RB_CNTL, RB_BUFSZ, size); + tmp = REG_SET_FIELD(tmp, UVD_RBC_RB_CNTL, RB_BLKSZ, 1); + tmp = REG_SET_FIELD(tmp, UVD_RBC_RB_CNTL, RB_NO_FETCH, 1); + tmp = REG_SET_FIELD(tmp, UVD_RBC_RB_CNTL, RB_WPTR_POLL_EN, 0); + tmp = REG_SET_FIELD(tmp, UVD_RBC_RB_CNTL, RB_NO_UPDATE, 1); + tmp = REG_SET_FIELD(tmp, UVD_RBC_RB_CNTL, RB_RPTR_WR_EN, 1); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_RBC_RB_CNTL), tmp); + + /* set the write pointer delay */ + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_RBC_RB_WPTR_CNTL), 0); + + /* set the wb address */ + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_RBC_RB_RPTR_ADDR), + (upper_32_bits(ring->gpu_addr) >> 2)); + + /* programm the RB_BASE for ring buffer */ + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_RBC_RB_64BIT_BAR_LOW), + lower_32_bits(ring->gpu_addr)); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_RBC_RB_64BIT_BAR_HIGH), + upper_32_bits(ring->gpu_addr)); + + ring->wptr = 0; + ring = &adev->uvd.ring_enc[0]; + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_RB_BASE_LO), ring->gpu_addr); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_RB_BASE_HI), upper_32_bits(ring->gpu_addr)); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_RB_SIZE), ring->ring_size / 4); + + /* add end packet */ + memcpy((void *)init_table, &end, sizeof(struct mmsch_v1_0_cmd_end)); + table_size += sizeof(struct mmsch_v1_0_cmd_end) / 4; + header->uvd_table_size = table_size; + + return uvd_v7_0_mmsch_start(adev, &adev->virt.mm_table); + } + return -EINVAL; /* already initializaed ? */ +} + /** * uvd_v7_0_start - start UVD block * @@ -1034,42 +1304,38 @@ static void uvd_v7_0_vm_reg_wait(struct amdgpu_ring *ring, static void uvd_v7_0_ring_emit_vm_flush(struct amdgpu_ring *ring, unsigned vm_id, uint64_t pd_addr) { + struct amdgpu_vmhub *hub = &ring->adev->vmhub[ring->funcs->vmhub]; uint32_t req = ring->adev->gart.gart_funcs->get_invalidate_req(vm_id); uint32_t data0, data1, mask; - unsigned eng = ring->idx; - unsigned i; + unsigned eng = ring->vm_inv_eng; pd_addr = pd_addr | 0x1; /* valid bit */ /* now only use physical base address of PDE and valid */ BUG_ON(pd_addr & 0xFFFF00000000003EULL); - for (i = 0; i < AMDGPU_MAX_VMHUBS; ++i) { - struct amdgpu_vmhub *hub = &ring->adev->vmhub[i]; - - data0 = (hub->ctx0_ptb_addr_hi32 + vm_id * 2) << 2; - data1 = upper_32_bits(pd_addr); - uvd_v7_0_vm_reg_write(ring, data0, data1); - - data0 = (hub->ctx0_ptb_addr_lo32 + vm_id * 2) << 2; - data1 = lower_32_bits(pd_addr); - uvd_v7_0_vm_reg_write(ring, data0, data1); - - data0 = (hub->ctx0_ptb_addr_lo32 + vm_id * 2) << 2; - data1 = lower_32_bits(pd_addr); - mask = 0xffffffff; - uvd_v7_0_vm_reg_wait(ring, data0, data1, mask); - - /* flush TLB */ - data0 = (hub->vm_inv_eng0_req + eng) << 2; - data1 = req; - uvd_v7_0_vm_reg_write(ring, data0, data1); - - /* wait for flush */ - data0 = (hub->vm_inv_eng0_ack + eng) << 2; - data1 = 1 << vm_id; - mask = 1 << vm_id; - uvd_v7_0_vm_reg_wait(ring, data0, data1, mask); - } + data0 = (hub->ctx0_ptb_addr_hi32 + vm_id * 2) << 2; + data1 = upper_32_bits(pd_addr); + uvd_v7_0_vm_reg_write(ring, data0, data1); + + data0 = (hub->ctx0_ptb_addr_lo32 + vm_id * 2) << 2; + data1 = lower_32_bits(pd_addr); + uvd_v7_0_vm_reg_write(ring, data0, data1); + + data0 = (hub->ctx0_ptb_addr_lo32 + vm_id * 2) << 2; + data1 = lower_32_bits(pd_addr); + mask = 0xffffffff; + uvd_v7_0_vm_reg_wait(ring, data0, data1, mask); + + /* flush TLB */ + data0 = (hub->vm_inv_eng0_req + eng) << 2; + data1 = req; + uvd_v7_0_vm_reg_write(ring, data0, data1); + + /* wait for flush */ + data0 = (hub->vm_inv_eng0_ack + eng) << 2; + data1 = 1 << vm_id; + mask = 1 << vm_id; + uvd_v7_0_vm_reg_wait(ring, data0, data1, mask); } static void uvd_v7_0_enc_ring_insert_end(struct amdgpu_ring *ring) @@ -1080,44 +1346,37 @@ static void uvd_v7_0_enc_ring_insert_end(struct amdgpu_ring *ring) static void uvd_v7_0_enc_ring_emit_vm_flush(struct amdgpu_ring *ring, unsigned int vm_id, uint64_t pd_addr) { + struct amdgpu_vmhub *hub = &ring->adev->vmhub[ring->funcs->vmhub]; uint32_t req = ring->adev->gart.gart_funcs->get_invalidate_req(vm_id); - unsigned eng = ring->idx; - unsigned i; + unsigned eng = ring->vm_inv_eng; pd_addr = pd_addr | 0x1; /* valid bit */ /* now only use physical base address of PDE and valid */ BUG_ON(pd_addr & 0xFFFF00000000003EULL); - for (i = 0; i < AMDGPU_MAX_VMHUBS; ++i) { - struct amdgpu_vmhub *hub = &ring->adev->vmhub[i]; - - amdgpu_ring_write(ring, HEVC_ENC_CMD_REG_WRITE); - amdgpu_ring_write(ring, - (hub->ctx0_ptb_addr_hi32 + vm_id * 2) << 2); - amdgpu_ring_write(ring, upper_32_bits(pd_addr)); - - amdgpu_ring_write(ring, HEVC_ENC_CMD_REG_WRITE); - amdgpu_ring_write(ring, - (hub->ctx0_ptb_addr_lo32 + vm_id * 2) << 2); - amdgpu_ring_write(ring, lower_32_bits(pd_addr)); - - amdgpu_ring_write(ring, HEVC_ENC_CMD_REG_WAIT); - amdgpu_ring_write(ring, - (hub->ctx0_ptb_addr_lo32 + vm_id * 2) << 2); - amdgpu_ring_write(ring, 0xffffffff); - amdgpu_ring_write(ring, lower_32_bits(pd_addr)); - - /* flush TLB */ - amdgpu_ring_write(ring, HEVC_ENC_CMD_REG_WRITE); - amdgpu_ring_write(ring, (hub->vm_inv_eng0_req + eng) << 2); - amdgpu_ring_write(ring, req); - - /* wait for flush */ - amdgpu_ring_write(ring, HEVC_ENC_CMD_REG_WAIT); - amdgpu_ring_write(ring, (hub->vm_inv_eng0_ack + eng) << 2); - amdgpu_ring_write(ring, 1 << vm_id); - amdgpu_ring_write(ring, 1 << vm_id); - } + amdgpu_ring_write(ring, HEVC_ENC_CMD_REG_WRITE); + amdgpu_ring_write(ring, (hub->ctx0_ptb_addr_hi32 + vm_id * 2) << 2); + amdgpu_ring_write(ring, upper_32_bits(pd_addr)); + + amdgpu_ring_write(ring, HEVC_ENC_CMD_REG_WRITE); + amdgpu_ring_write(ring, (hub->ctx0_ptb_addr_lo32 + vm_id * 2) << 2); + amdgpu_ring_write(ring, lower_32_bits(pd_addr)); + + amdgpu_ring_write(ring, HEVC_ENC_CMD_REG_WAIT); + amdgpu_ring_write(ring, (hub->ctx0_ptb_addr_lo32 + vm_id * 2) << 2); + amdgpu_ring_write(ring, 0xffffffff); + amdgpu_ring_write(ring, lower_32_bits(pd_addr)); + + /* flush TLB */ + amdgpu_ring_write(ring, HEVC_ENC_CMD_REG_WRITE); + amdgpu_ring_write(ring, (hub->vm_inv_eng0_req + eng) << 2); + amdgpu_ring_write(ring, req); + + /* wait for flush */ + amdgpu_ring_write(ring, HEVC_ENC_CMD_REG_WAIT); + amdgpu_ring_write(ring, (hub->vm_inv_eng0_ack + eng) << 2); + amdgpu_ring_write(ring, 1 << vm_id); + amdgpu_ring_write(ring, 1 << vm_id); } #if 0 @@ -1240,7 +1499,8 @@ static int uvd_v7_0_process_interrupt(struct amdgpu_device *adev, amdgpu_fence_process(&adev->uvd.ring_enc[0]); break; case 120: - amdgpu_fence_process(&adev->uvd.ring_enc[1]); + if (!amdgpu_sriov_vf(adev)) + amdgpu_fence_process(&adev->uvd.ring_enc[1]); break; default: DRM_ERROR("Unhandled interrupt: %d %d\n", @@ -1448,13 +1708,14 @@ static const struct amdgpu_ring_funcs uvd_v7_0_ring_vm_funcs = { .align_mask = 0xf, .nop = PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_NO_OP), 0), .support_64bit_ptrs = false, + .vmhub = AMDGPU_MMHUB, .get_rptr = uvd_v7_0_ring_get_rptr, .get_wptr = uvd_v7_0_ring_get_wptr, .set_wptr = uvd_v7_0_ring_set_wptr, .emit_frame_size = 2 + /* uvd_v7_0_ring_emit_hdp_flush */ 2 + /* uvd_v7_0_ring_emit_hdp_invalidate */ - 34 * AMDGPU_MAX_VMHUBS + /* uvd_v7_0_ring_emit_vm_flush */ + 34 + /* uvd_v7_0_ring_emit_vm_flush */ 14 + 14, /* uvd_v7_0_ring_emit_fence x2 vm fence */ .emit_ib_size = 8, /* uvd_v7_0_ring_emit_ib */ .emit_ib = uvd_v7_0_ring_emit_ib, @@ -1475,11 +1736,12 @@ static const struct amdgpu_ring_funcs uvd_v7_0_enc_ring_vm_funcs = { .align_mask = 0x3f, .nop = HEVC_ENC_CMD_NO_OP, .support_64bit_ptrs = false, + .vmhub = AMDGPU_MMHUB, .get_rptr = uvd_v7_0_enc_ring_get_rptr, .get_wptr = uvd_v7_0_enc_ring_get_wptr, .set_wptr = uvd_v7_0_enc_ring_set_wptr, .emit_frame_size = - 17 * AMDGPU_MAX_VMHUBS + /* uvd_v7_0_enc_ring_emit_vm_flush */ + 17 + /* uvd_v7_0_enc_ring_emit_vm_flush */ 5 + 5 + /* uvd_v7_0_enc_ring_emit_fence x2 vm fence */ 1, /* uvd_v7_0_enc_ring_insert_end */ .emit_ib_size = 5, /* uvd_v7_0_enc_ring_emit_ib */ diff --git a/drivers/gpu/drm/amd/amdgpu/vce_v4_0.c b/drivers/gpu/drm/amd/amdgpu/vce_v4_0.c index edde5fe938d6..139f964196b4 100644 --- a/drivers/gpu/drm/amd/amdgpu/vce_v4_0.c +++ b/drivers/gpu/drm/amd/amdgpu/vce_v4_0.c @@ -49,63 +49,6 @@ static void vce_v4_0_mc_resume(struct amdgpu_device *adev); static void vce_v4_0_set_ring_funcs(struct amdgpu_device *adev); static void vce_v4_0_set_irq_funcs(struct amdgpu_device *adev); -static inline void mmsch_insert_direct_wt(struct mmsch_v1_0_cmd_direct_write *direct_wt, - uint32_t *init_table, - uint32_t reg_offset, - uint32_t value) -{ - direct_wt->cmd_header.reg_offset = reg_offset; - direct_wt->reg_value = value; - memcpy((void *)init_table, direct_wt, sizeof(struct mmsch_v1_0_cmd_direct_write)); -} - -static inline void mmsch_insert_direct_rd_mod_wt(struct mmsch_v1_0_cmd_direct_read_modify_write *direct_rd_mod_wt, - uint32_t *init_table, - uint32_t reg_offset, - uint32_t mask, uint32_t data) -{ - direct_rd_mod_wt->cmd_header.reg_offset = reg_offset; - direct_rd_mod_wt->mask_value = mask; - direct_rd_mod_wt->write_data = data; - memcpy((void *)init_table, direct_rd_mod_wt, - sizeof(struct mmsch_v1_0_cmd_direct_read_modify_write)); -} - -static inline void mmsch_insert_direct_poll(struct mmsch_v1_0_cmd_direct_polling *direct_poll, - uint32_t *init_table, - uint32_t reg_offset, - uint32_t mask, uint32_t wait) -{ - direct_poll->cmd_header.reg_offset = reg_offset; - direct_poll->mask_value = mask; - direct_poll->wait_value = wait; - memcpy((void *)init_table, direct_poll, sizeof(struct mmsch_v1_0_cmd_direct_polling)); -} - -#define INSERT_DIRECT_RD_MOD_WT(reg, mask, data) { \ - mmsch_insert_direct_rd_mod_wt(&direct_rd_mod_wt, \ - init_table, (reg), \ - (mask), (data)); \ - init_table += sizeof(struct mmsch_v1_0_cmd_direct_read_modify_write)/4; \ - table_size += sizeof(struct mmsch_v1_0_cmd_direct_read_modify_write)/4; \ -} - -#define INSERT_DIRECT_WT(reg, value) { \ - mmsch_insert_direct_wt(&direct_wt, \ - init_table, (reg), \ - (value)); \ - init_table += sizeof(struct mmsch_v1_0_cmd_direct_write)/4; \ - table_size += sizeof(struct mmsch_v1_0_cmd_direct_write)/4; \ -} - -#define INSERT_DIRECT_POLL(reg, mask, wait) { \ - mmsch_insert_direct_poll(&direct_poll, \ - init_table, (reg), \ - (mask), (wait)); \ - init_table += sizeof(struct mmsch_v1_0_cmd_direct_polling)/4; \ - table_size += sizeof(struct mmsch_v1_0_cmd_direct_polling)/4; \ -} - /** * vce_v4_0_ring_get_rptr - get read pointer * @@ -280,60 +223,73 @@ static int vce_v4_0_sriov_start(struct amdgpu_device *adev) init_table += header->vce_table_offset; ring = &adev->vce.ring[0]; - INSERT_DIRECT_WT(SOC15_REG_OFFSET(VCE, 0, mmVCE_RB_RPTR), ring->wptr); - INSERT_DIRECT_WT(SOC15_REG_OFFSET(VCE, 0, mmVCE_RB_WPTR), ring->wptr); - INSERT_DIRECT_WT(SOC15_REG_OFFSET(VCE, 0, mmVCE_RB_BASE_LO), lower_32_bits(ring->gpu_addr)); - INSERT_DIRECT_WT(SOC15_REG_OFFSET(VCE, 0, mmVCE_RB_BASE_HI), upper_32_bits(ring->gpu_addr)); - INSERT_DIRECT_WT(SOC15_REG_OFFSET(VCE, 0, mmVCE_RB_SIZE), ring->ring_size / 4); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(VCE, 0, mmVCE_RB_BASE_LO), + lower_32_bits(ring->gpu_addr)); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(VCE, 0, mmVCE_RB_BASE_HI), + upper_32_bits(ring->gpu_addr)); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(VCE, 0, mmVCE_RB_SIZE), + ring->ring_size / 4); /* BEGING OF MC_RESUME */ - INSERT_DIRECT_RD_MOD_WT(SOC15_REG_OFFSET(VCE, 0, mmVCE_CLOCK_GATING_A), ~(1 << 16), 0); - INSERT_DIRECT_RD_MOD_WT(SOC15_REG_OFFSET(VCE, 0, mmVCE_UENC_CLOCK_GATING), ~0xFF9FF000, 0x1FF000); - INSERT_DIRECT_RD_MOD_WT(SOC15_REG_OFFSET(VCE, 0, mmVCE_UENC_REG_CLOCK_GATING), ~0x3F, 0x3F); - INSERT_DIRECT_WT(SOC15_REG_OFFSET(VCE, 0, mmVCE_CLOCK_GATING_B), 0x1FF); - - INSERT_DIRECT_WT(SOC15_REG_OFFSET(VCE, 0, mmVCE_LMI_CTRL), 0x398000); - INSERT_DIRECT_RD_MOD_WT(SOC15_REG_OFFSET(VCE, 0, mmVCE_LMI_CACHE_CTRL), ~0x1, 0); - INSERT_DIRECT_WT(SOC15_REG_OFFSET(VCE, 0, mmVCE_LMI_SWAP_CNTL), 0); - INSERT_DIRECT_WT(SOC15_REG_OFFSET(VCE, 0, mmVCE_LMI_SWAP_CNTL1), 0); - INSERT_DIRECT_WT(SOC15_REG_OFFSET(VCE, 0, mmVCE_LMI_VM_CTRL), 0); - - INSERT_DIRECT_WT(SOC15_REG_OFFSET(VCE, 0, mmVCE_LMI_VCPU_CACHE_40BIT_BAR0), adev->vce.gpu_addr >> 8); - INSERT_DIRECT_WT(SOC15_REG_OFFSET(VCE, 0, mmVCE_LMI_VCPU_CACHE_40BIT_BAR1), adev->vce.gpu_addr >> 8); - INSERT_DIRECT_WT(SOC15_REG_OFFSET(VCE, 0, mmVCE_LMI_VCPU_CACHE_40BIT_BAR2), adev->vce.gpu_addr >> 8); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(VCE, 0, mmVCE_LMI_CTRL), 0x398000); + MMSCH_V1_0_INSERT_DIRECT_RD_MOD_WT(SOC15_REG_OFFSET(VCE, 0, mmVCE_LMI_CACHE_CTRL), ~0x1, 0); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(VCE, 0, mmVCE_LMI_SWAP_CNTL), 0); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(VCE, 0, mmVCE_LMI_SWAP_CNTL1), 0); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(VCE, 0, mmVCE_LMI_VM_CTRL), 0); + + if (adev->firmware.load_type == AMDGPU_FW_LOAD_PSP) { + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(VCE, 0, mmVCE_LMI_VCPU_CACHE_40BIT_BAR0), + adev->firmware.ucode[AMDGPU_UCODE_ID_VCE].mc_addr >> 8); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(VCE, 0, mmVCE_LMI_VCPU_CACHE_40BIT_BAR1), + adev->firmware.ucode[AMDGPU_UCODE_ID_VCE].mc_addr >> 8); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(VCE, 0, mmVCE_LMI_VCPU_CACHE_40BIT_BAR2), + adev->firmware.ucode[AMDGPU_UCODE_ID_VCE].mc_addr >> 8); + } else { + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(VCE, 0, mmVCE_LMI_VCPU_CACHE_40BIT_BAR0), + adev->vce.gpu_addr >> 8); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(VCE, 0, mmVCE_LMI_VCPU_CACHE_40BIT_BAR1), + adev->vce.gpu_addr >> 8); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(VCE, 0, mmVCE_LMI_VCPU_CACHE_40BIT_BAR2), + adev->vce.gpu_addr >> 8); + } offset = AMDGPU_VCE_FIRMWARE_OFFSET; size = VCE_V4_0_FW_SIZE; - INSERT_DIRECT_WT(SOC15_REG_OFFSET(VCE, 0, mmVCE_VCPU_CACHE_OFFSET0), offset & 0x7FFFFFFF); - INSERT_DIRECT_WT(SOC15_REG_OFFSET(VCE, 0, mmVCE_VCPU_CACHE_SIZE0), size); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(VCE, 0, mmVCE_VCPU_CACHE_OFFSET0), + offset & 0x7FFFFFFF); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(VCE, 0, mmVCE_VCPU_CACHE_SIZE0), size); offset += size; size = VCE_V4_0_STACK_SIZE; - INSERT_DIRECT_WT(SOC15_REG_OFFSET(VCE, 0, mmVCE_VCPU_CACHE_OFFSET1), offset & 0x7FFFFFFF); - INSERT_DIRECT_WT(SOC15_REG_OFFSET(VCE, 0, mmVCE_VCPU_CACHE_SIZE1), size); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(VCE, 0, mmVCE_VCPU_CACHE_OFFSET1), + offset & 0x7FFFFFFF); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(VCE, 0, mmVCE_VCPU_CACHE_SIZE1), size); offset += size; size = VCE_V4_0_DATA_SIZE; - INSERT_DIRECT_WT(SOC15_REG_OFFSET(VCE, 0, mmVCE_VCPU_CACHE_OFFSET2), offset & 0x7FFFFFFF); - INSERT_DIRECT_WT(SOC15_REG_OFFSET(VCE, 0, mmVCE_VCPU_CACHE_SIZE2), size); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(VCE, 0, mmVCE_VCPU_CACHE_OFFSET2), + offset & 0x7FFFFFFF); + MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(VCE, 0, mmVCE_VCPU_CACHE_SIZE2), size); - INSERT_DIRECT_RD_MOD_WT(SOC15_REG_OFFSET(VCE, 0, mmVCE_LMI_CTRL2), ~0x100, 0); - INSERT_DIRECT_RD_MOD_WT(SOC15_REG_OFFSET(VCE, 0, mmVCE_SYS_INT_EN), - 0xffffffff, VCE_SYS_INT_EN__VCE_SYS_INT_TRAP_INTERRUPT_EN_MASK); + MMSCH_V1_0_INSERT_DIRECT_RD_MOD_WT(SOC15_REG_OFFSET(VCE, 0, mmVCE_LMI_CTRL2), ~0x100, 0); + MMSCH_V1_0_INSERT_DIRECT_RD_MOD_WT(SOC15_REG_OFFSET(VCE, 0, mmVCE_SYS_INT_EN), + 0xffffffff, VCE_SYS_INT_EN__VCE_SYS_INT_TRAP_INTERRUPT_EN_MASK); /* end of MC_RESUME */ - INSERT_DIRECT_RD_MOD_WT(SOC15_REG_OFFSET(VCE, 0, mmVCE_VCPU_CNTL), - ~0x200001, VCE_VCPU_CNTL__CLK_EN_MASK); - INSERT_DIRECT_RD_MOD_WT(SOC15_REG_OFFSET(VCE, 0, mmVCE_SOFT_RESET), - ~VCE_SOFT_RESET__ECPU_SOFT_RESET_MASK, 0); + MMSCH_V1_0_INSERT_DIRECT_RD_MOD_WT(SOC15_REG_OFFSET(VCE, 0, mmVCE_STATUS), + VCE_STATUS__JOB_BUSY_MASK, ~VCE_STATUS__JOB_BUSY_MASK); + MMSCH_V1_0_INSERT_DIRECT_RD_MOD_WT(SOC15_REG_OFFSET(VCE, 0, mmVCE_VCPU_CNTL), + ~0x200001, VCE_VCPU_CNTL__CLK_EN_MASK); + MMSCH_V1_0_INSERT_DIRECT_RD_MOD_WT(SOC15_REG_OFFSET(VCE, 0, mmVCE_SOFT_RESET), + ~VCE_SOFT_RESET__ECPU_SOFT_RESET_MASK, 0); - INSERT_DIRECT_POLL(SOC15_REG_OFFSET(VCE, 0, mmVCE_STATUS), - VCE_STATUS_VCPU_REPORT_FW_LOADED_MASK, - VCE_STATUS_VCPU_REPORT_FW_LOADED_MASK); + MMSCH_V1_0_INSERT_DIRECT_POLL(SOC15_REG_OFFSET(VCE, 0, mmVCE_STATUS), + VCE_STATUS_VCPU_REPORT_FW_LOADED_MASK, + VCE_STATUS_VCPU_REPORT_FW_LOADED_MASK); /* clear BUSY flag */ - INSERT_DIRECT_RD_MOD_WT(SOC15_REG_OFFSET(VCE, 0, mmVCE_STATUS), - ~VCE_STATUS__JOB_BUSY_MASK, 0); + MMSCH_V1_0_INSERT_DIRECT_RD_MOD_WT(SOC15_REG_OFFSET(VCE, 0, mmVCE_STATUS), + ~VCE_STATUS__JOB_BUSY_MASK, 0); /* add end packet */ memcpy((void *)init_table, &end, sizeof(struct mmsch_v1_0_cmd_end)); @@ -494,20 +450,9 @@ static int vce_v4_0_sw_init(void *handle) return r; } - if (amdgpu_sriov_vf(adev)) { - r = amdgpu_bo_create_kernel(adev, PAGE_SIZE, PAGE_SIZE, - AMDGPU_GEM_DOMAIN_VRAM, - &adev->virt.mm_table.bo, - &adev->virt.mm_table.gpu_addr, - (void *)&adev->virt.mm_table.cpu_addr); - if (!r) { - memset((void *)adev->virt.mm_table.cpu_addr, 0, PAGE_SIZE); - printk("mm table gpu addr = 0x%llx, cpu addr = %p. \n", - adev->virt.mm_table.gpu_addr, - adev->virt.mm_table.cpu_addr); - } + r = amdgpu_virt_alloc_mm_table(adev); + if (r) return r; - } return r; } @@ -518,10 +463,7 @@ static int vce_v4_0_sw_fini(void *handle) struct amdgpu_device *adev = (struct amdgpu_device *)handle; /* free MM table */ - if (amdgpu_sriov_vf(adev)) - amdgpu_bo_free_kernel(&adev->virt.mm_table.bo, - &adev->virt.mm_table.gpu_addr, - (void *)&adev->virt.mm_table.cpu_addr); + amdgpu_virt_free_mm_table(adev); r = amdgpu_vce_suspend(adev); if (r) @@ -973,44 +915,37 @@ static void vce_v4_0_ring_insert_end(struct amdgpu_ring *ring) static void vce_v4_0_emit_vm_flush(struct amdgpu_ring *ring, unsigned int vm_id, uint64_t pd_addr) { + struct amdgpu_vmhub *hub = &ring->adev->vmhub[ring->funcs->vmhub]; uint32_t req = ring->adev->gart.gart_funcs->get_invalidate_req(vm_id); - unsigned eng = ring->idx; - unsigned i; + unsigned eng = ring->vm_inv_eng; pd_addr = pd_addr | 0x1; /* valid bit */ /* now only use physical base address of PDE and valid */ BUG_ON(pd_addr & 0xFFFF00000000003EULL); - for (i = 0; i < AMDGPU_MAX_VMHUBS; ++i) { - struct amdgpu_vmhub *hub = &ring->adev->vmhub[i]; - - amdgpu_ring_write(ring, VCE_CMD_REG_WRITE); - amdgpu_ring_write(ring, - (hub->ctx0_ptb_addr_hi32 + vm_id * 2) << 2); - amdgpu_ring_write(ring, upper_32_bits(pd_addr)); - - amdgpu_ring_write(ring, VCE_CMD_REG_WRITE); - amdgpu_ring_write(ring, - (hub->ctx0_ptb_addr_lo32 + vm_id * 2) << 2); - amdgpu_ring_write(ring, lower_32_bits(pd_addr)); - - amdgpu_ring_write(ring, VCE_CMD_REG_WAIT); - amdgpu_ring_write(ring, - (hub->ctx0_ptb_addr_lo32 + vm_id * 2) << 2); - amdgpu_ring_write(ring, 0xffffffff); - amdgpu_ring_write(ring, lower_32_bits(pd_addr)); - - /* flush TLB */ - amdgpu_ring_write(ring, VCE_CMD_REG_WRITE); - amdgpu_ring_write(ring, (hub->vm_inv_eng0_req + eng) << 2); - amdgpu_ring_write(ring, req); - - /* wait for flush */ - amdgpu_ring_write(ring, VCE_CMD_REG_WAIT); - amdgpu_ring_write(ring, (hub->vm_inv_eng0_ack + eng) << 2); - amdgpu_ring_write(ring, 1 << vm_id); - amdgpu_ring_write(ring, 1 << vm_id); - } + amdgpu_ring_write(ring, VCE_CMD_REG_WRITE); + amdgpu_ring_write(ring, (hub->ctx0_ptb_addr_hi32 + vm_id * 2) << 2); + amdgpu_ring_write(ring, upper_32_bits(pd_addr)); + + amdgpu_ring_write(ring, VCE_CMD_REG_WRITE); + amdgpu_ring_write(ring, (hub->ctx0_ptb_addr_lo32 + vm_id * 2) << 2); + amdgpu_ring_write(ring, lower_32_bits(pd_addr)); + + amdgpu_ring_write(ring, VCE_CMD_REG_WAIT); + amdgpu_ring_write(ring, (hub->ctx0_ptb_addr_lo32 + vm_id * 2) << 2); + amdgpu_ring_write(ring, 0xffffffff); + amdgpu_ring_write(ring, lower_32_bits(pd_addr)); + + /* flush TLB */ + amdgpu_ring_write(ring, VCE_CMD_REG_WRITE); + amdgpu_ring_write(ring, (hub->vm_inv_eng0_req + eng) << 2); + amdgpu_ring_write(ring, req); + + /* wait for flush */ + amdgpu_ring_write(ring, VCE_CMD_REG_WAIT); + amdgpu_ring_write(ring, (hub->vm_inv_eng0_ack + eng) << 2); + amdgpu_ring_write(ring, 1 << vm_id); + amdgpu_ring_write(ring, 1 << vm_id); } static int vce_v4_0_set_interrupt_state(struct amdgpu_device *adev, @@ -1078,12 +1013,13 @@ static const struct amdgpu_ring_funcs vce_v4_0_ring_vm_funcs = { .align_mask = 0x3f, .nop = VCE_CMD_NO_OP, .support_64bit_ptrs = false, + .vmhub = AMDGPU_MMHUB, .get_rptr = vce_v4_0_ring_get_rptr, .get_wptr = vce_v4_0_ring_get_wptr, .set_wptr = vce_v4_0_ring_set_wptr, .parse_cs = amdgpu_vce_ring_parse_cs_vm, .emit_frame_size = - 17 * AMDGPU_MAX_VMHUBS + /* vce_v4_0_emit_vm_flush */ + 17 + /* vce_v4_0_emit_vm_flush */ 5 + 5 + /* amdgpu_vce_ring_emit_fence x2 vm fence */ 1, /* vce_v4_0_ring_insert_end */ .emit_ib_size = 5, /* vce_v4_0_ring_emit_ib */ diff --git a/drivers/gpu/drm/amd/amdgpu/vega10_ih.c b/drivers/gpu/drm/amd/amdgpu/vega10_ih.c index 071f56e439bb..3b9740fb2c41 100644 --- a/drivers/gpu/drm/amd/amdgpu/vega10_ih.c +++ b/drivers/gpu/drm/amd/amdgpu/vega10_ih.c @@ -20,7 +20,7 @@ * OTHER DEALINGS IN THE SOFTWARE. * */ -#include "drmP.h" +#include <drm/drmP.h> #include "amdgpu.h" #include "amdgpu_ih.h" #include "soc15.h" diff --git a/drivers/gpu/drm/amd/amdgpu/vi.c b/drivers/gpu/drm/amd/amdgpu/vi.c index b1132f5e84fc..3a187619286f 100644 --- a/drivers/gpu/drm/amd/amdgpu/vi.c +++ b/drivers/gpu/drm/amd/amdgpu/vi.c @@ -21,7 +21,7 @@ * */ #include <linux/slab.h> -#include "drmP.h" +#include <drm/drmP.h> #include "amdgpu.h" #include "amdgpu_atombios.h" #include "amdgpu_ih.h" diff --git a/drivers/gpu/drm/amd/amdkfd/Makefile b/drivers/gpu/drm/amd/amdkfd/Makefile index 7fc9b0f444cb..b400d5664252 100644 --- a/drivers/gpu/drm/amd/amdkfd/Makefile +++ b/drivers/gpu/drm/amd/amdkfd/Makefile @@ -2,7 +2,7 @@ # Makefile for Heterogenous System Architecture support for AMD GPU devices # -ccflags-y := -Iinclude/drm -Idrivers/gpu/drm/amd/include/ \ +ccflags-y := -Idrivers/gpu/drm/amd/include/ \ -Idrivers/gpu/drm/amd/include/asic_reg amdkfd-y := kfd_module.o kfd_device.o kfd_chardev.o kfd_topology.o \ diff --git a/drivers/gpu/drm/amd/include/amd_shared.h b/drivers/gpu/drm/amd/include/amd_shared.h index 2ccf44e580de..1d1ac1ef94f7 100644 --- a/drivers/gpu/drm/amd/include/amd_shared.h +++ b/drivers/gpu/drm/amd/include/amd_shared.h @@ -138,6 +138,12 @@ struct amd_pp_profile { uint8_t down_hyst; }; +enum amd_fan_ctrl_mode { + AMD_FAN_CTRL_NONE = 0, + AMD_FAN_CTRL_MANUAL = 1, + AMD_FAN_CTRL_AUTO = 2, +}; + /* CG flags */ #define AMD_CG_SUPPORT_GFX_MGCG (1 << 0) #define AMD_CG_SUPPORT_GFX_MGLS (1 << 1) diff --git a/drivers/gpu/drm/amd/include/cgs_common.h b/drivers/gpu/drm/amd/include/cgs_common.h index 17b9d41f3e87..0a94f749e3c0 100644 --- a/drivers/gpu/drm/amd/include/cgs_common.h +++ b/drivers/gpu/drm/amd/include/cgs_common.h @@ -54,20 +54,6 @@ enum cgs_ind_reg { }; /** - * enum cgs_clock - Clocks controlled by the SMU - */ -enum cgs_clock { - CGS_CLOCK__SCLK, - CGS_CLOCK__MCLK, - CGS_CLOCK__VCLK, - CGS_CLOCK__DCLK, - CGS_CLOCK__ECLK, - CGS_CLOCK__ACLK, - CGS_CLOCK__ICLK, - /* ... */ -}; - -/** * enum cgs_engine - Engines that can be statically power-gated */ enum cgs_engine { @@ -81,15 +67,6 @@ enum cgs_engine { /* ... */ }; -/** - * enum cgs_voltage_planes - Voltage planes for external camera HW - */ -enum cgs_voltage_planes { - CGS_VOLTAGE_PLANE__SENSOR0, - CGS_VOLTAGE_PLANE__SENSOR1, - /* ... */ -}; - /* * enum cgs_ucode_id - Firmware types for different IPs */ @@ -147,17 +124,6 @@ enum cgs_resource_type { }; /** - * struct cgs_clock_limits - Clock limits - * - * Clocks are specified in 10KHz units. - */ -struct cgs_clock_limits { - unsigned min; /**< Minimum supported frequency */ - unsigned max; /**< Maxumim supported frequency */ - unsigned sustainable; /**< Thermally sustainable frequency */ -}; - -/** * struct cgs_firmware_info - Firmware information */ struct cgs_firmware_info { @@ -221,54 +187,6 @@ struct cgs_acpi_method_info { }; /** - * cgs_gpu_mem_info() - Return information about memory heaps - * @cgs_device: opaque device handle - * @type: memory type - * @mc_start: Start MC address of the heap (output) - * @mc_size: MC address space size (output) - * @mem_size: maximum amount of memory available for allocation (output) - * - * This function returns information about memory heaps. The type - * parameter is used to select the memory heap. The mc_start and - * mc_size for GART heaps may be bigger than the memory available for - * allocation. - * - * mc_start and mc_size are undefined for non-contiguous FB memory - * types, since buffers allocated with these types may or may not be - * GART mapped. - * - * Return: 0 on success, -errno otherwise - */ -typedef int (*cgs_gpu_mem_info_t)(struct cgs_device *cgs_device, enum cgs_gpu_mem_type type, - uint64_t *mc_start, uint64_t *mc_size, - uint64_t *mem_size); - -/** - * cgs_gmap_kmem() - map kernel memory to GART aperture - * @cgs_device: opaque device handle - * @kmem: pointer to kernel memory - * @size: size to map - * @min_offset: minimum offset from start of GART aperture - * @max_offset: maximum offset from start of GART aperture - * @kmem_handle: kernel memory handle (output) - * @mcaddr: MC address (output) - * - * Return: 0 on success, -errno otherwise - */ -typedef int (*cgs_gmap_kmem_t)(struct cgs_device *cgs_device, void *kmem, uint64_t size, - uint64_t min_offset, uint64_t max_offset, - cgs_handle_t *kmem_handle, uint64_t *mcaddr); - -/** - * cgs_gunmap_kmem() - unmap kernel memory - * @cgs_device: opaque device handle - * @kmem_handle: kernel memory handle returned by gmap_kmem - * - * Return: 0 on success, -errno otherwise - */ -typedef int (*cgs_gunmap_kmem_t)(struct cgs_device *cgs_device, cgs_handle_t kmem_handle); - -/** * cgs_alloc_gpu_mem() - Allocate GPU memory * @cgs_device: opaque device handle * @type: memory type @@ -392,62 +310,6 @@ typedef void (*cgs_write_ind_register_t)(struct cgs_device *cgs_device, enum cgs unsigned index, uint32_t value); /** - * cgs_read_pci_config_byte() - Read byte from PCI configuration space - * @cgs_device: opaque device handle - * @addr: address - * - * Return: Value read - */ -typedef uint8_t (*cgs_read_pci_config_byte_t)(struct cgs_device *cgs_device, unsigned addr); - -/** - * cgs_read_pci_config_word() - Read word from PCI configuration space - * @cgs_device: opaque device handle - * @addr: address, must be word-aligned - * - * Return: Value read - */ -typedef uint16_t (*cgs_read_pci_config_word_t)(struct cgs_device *cgs_device, unsigned addr); - -/** - * cgs_read_pci_config_dword() - Read dword from PCI configuration space - * @cgs_device: opaque device handle - * @addr: address, must be dword-aligned - * - * Return: Value read - */ -typedef uint32_t (*cgs_read_pci_config_dword_t)(struct cgs_device *cgs_device, - unsigned addr); - -/** - * cgs_write_pci_config_byte() - Write byte to PCI configuration space - * @cgs_device: opaque device handle - * @addr: address - * @value: value to write - */ -typedef void (*cgs_write_pci_config_byte_t)(struct cgs_device *cgs_device, unsigned addr, - uint8_t value); - -/** - * cgs_write_pci_config_word() - Write byte to PCI configuration space - * @cgs_device: opaque device handle - * @addr: address, must be word-aligned - * @value: value to write - */ -typedef void (*cgs_write_pci_config_word_t)(struct cgs_device *cgs_device, unsigned addr, - uint16_t value); - -/** - * cgs_write_pci_config_dword() - Write byte to PCI configuration space - * @cgs_device: opaque device handle - * @addr: address, must be dword-aligned - * @value: value to write - */ -typedef void (*cgs_write_pci_config_dword_t)(struct cgs_device *cgs_device, unsigned addr, - uint32_t value); - - -/** * cgs_get_pci_resource() - provide access to a device resource (PCI BAR) * @cgs_device: opaque device handle * @resource_type: Type of Resource (MMIO, IO, ROM, FB, DOORBELL) @@ -501,87 +363,6 @@ typedef int (*cgs_atom_exec_cmd_table_t)(struct cgs_device *cgs_device, unsigned table, void *args); /** - * cgs_create_pm_request() - Create a power management request - * @cgs_device: opaque device handle - * @request: handle of created PM request (output) - * - * Return: 0 on success, -errno otherwise - */ -typedef int (*cgs_create_pm_request_t)(struct cgs_device *cgs_device, cgs_handle_t *request); - -/** - * cgs_destroy_pm_request() - Destroy a power management request - * @cgs_device: opaque device handle - * @request: handle of created PM request - * - * Return: 0 on success, -errno otherwise - */ -typedef int (*cgs_destroy_pm_request_t)(struct cgs_device *cgs_device, cgs_handle_t request); - -/** - * cgs_set_pm_request() - Activate or deactiveate a PM request - * @cgs_device: opaque device handle - * @request: PM request handle - * @active: 0 = deactivate, non-0 = activate - * - * While a PM request is active, its minimum clock requests are taken - * into account as the requested engines are powered up. When the - * request is inactive, the engines may be powered down and clocks may - * be lower, depending on other PM requests by other driver - * components. - * - * Return: 0 on success, -errno otherwise - */ -typedef int (*cgs_set_pm_request_t)(struct cgs_device *cgs_device, cgs_handle_t request, - int active); - -/** - * cgs_pm_request_clock() - Request a minimum frequency for a specific clock - * @cgs_device: opaque device handle - * @request: PM request handle - * @clock: which clock? - * @freq: requested min. frequency in 10KHz units (0 to clear request) - * - * Return: 0 on success, -errno otherwise - */ -typedef int (*cgs_pm_request_clock_t)(struct cgs_device *cgs_device, cgs_handle_t request, - enum cgs_clock clock, unsigned freq); - -/** - * cgs_pm_request_engine() - Request an engine to be powered up - * @cgs_device: opaque device handle - * @request: PM request handle - * @engine: which engine? - * @powered: 0 = powered down, non-0 = powered up - * - * Return: 0 on success, -errno otherwise - */ -typedef int (*cgs_pm_request_engine_t)(struct cgs_device *cgs_device, cgs_handle_t request, - enum cgs_engine engine, int powered); - -/** - * cgs_pm_query_clock_limits() - Query clock frequency limits - * @cgs_device: opaque device handle - * @clock: which clock? - * @limits: clock limits - * - * Return: 0 on success, -errno otherwise - */ -typedef int (*cgs_pm_query_clock_limits_t)(struct cgs_device *cgs_device, - enum cgs_clock clock, - struct cgs_clock_limits *limits); - -/** - * cgs_set_camera_voltages() - Apply specific voltages to PMIC voltage planes - * @cgs_device: opaque device handle - * @mask: bitmask of voltages to change (1<<CGS_VOLTAGE_PLANE__xyz|...) - * @voltages: pointer to array of voltage values in 1mV units - * - * Return: 0 on success, -errno otherwise - */ -typedef int (*cgs_set_camera_voltages_t)(struct cgs_device *cgs_device, uint32_t mask, - const uint32_t *voltages); -/** * cgs_get_firmware_info - Get the firmware information from core driver * @cgs_device: opaque device handle * @type: the firmware type @@ -627,9 +408,6 @@ typedef int (*cgs_enter_safe_mode)(struct cgs_device *cgs_device, bool en); struct cgs_ops { /* memory management calls (similar to KFD interface) */ - cgs_gpu_mem_info_t gpu_mem_info; - cgs_gmap_kmem_t gmap_kmem; - cgs_gunmap_kmem_t gunmap_kmem; cgs_alloc_gpu_mem_t alloc_gpu_mem; cgs_free_gpu_mem_t free_gpu_mem; cgs_gmap_gpu_mem_t gmap_gpu_mem; @@ -641,27 +419,12 @@ struct cgs_ops { cgs_write_register_t write_register; cgs_read_ind_register_t read_ind_register; cgs_write_ind_register_t write_ind_register; - /* PCI configuration space access */ - cgs_read_pci_config_byte_t read_pci_config_byte; - cgs_read_pci_config_word_t read_pci_config_word; - cgs_read_pci_config_dword_t read_pci_config_dword; - cgs_write_pci_config_byte_t write_pci_config_byte; - cgs_write_pci_config_word_t write_pci_config_word; - cgs_write_pci_config_dword_t write_pci_config_dword; /* PCI resources */ cgs_get_pci_resource_t get_pci_resource; /* ATOM BIOS */ cgs_atom_get_data_table_t atom_get_data_table; cgs_atom_get_cmd_table_revs_t atom_get_cmd_table_revs; cgs_atom_exec_cmd_table_t atom_exec_cmd_table; - /* Power management */ - cgs_create_pm_request_t create_pm_request; - cgs_destroy_pm_request_t destroy_pm_request; - cgs_set_pm_request_t set_pm_request; - cgs_pm_request_clock_t pm_request_clock; - cgs_pm_request_engine_t pm_request_engine; - cgs_pm_query_clock_limits_t pm_query_clock_limits; - cgs_set_camera_voltages_t set_camera_voltages; /* Firmware Info */ cgs_get_firmware_info get_firmware_info; cgs_rel_firmware rel_firmware; @@ -696,12 +459,6 @@ struct cgs_device #define CGS_OS_CALL(func,dev,...) \ (((struct cgs_device *)dev)->os_ops->func(dev, ##__VA_ARGS__)) -#define cgs_gpu_mem_info(dev,type,mc_start,mc_size,mem_size) \ - CGS_CALL(gpu_mem_info,dev,type,mc_start,mc_size,mem_size) -#define cgs_gmap_kmem(dev,kmem,size,min_off,max_off,kmem_handle,mcaddr) \ - CGS_CALL(gmap_kmem,dev,kmem,size,min_off,max_off,kmem_handle,mcaddr) -#define cgs_gunmap_kmem(dev,kmem_handle) \ - CGS_CALL(gunmap_kmem,dev,keme_handle) #define cgs_alloc_gpu_mem(dev,type,size,align,min_off,max_off,handle) \ CGS_CALL(alloc_gpu_mem,dev,type,size,align,min_off,max_off,handle) #define cgs_free_gpu_mem(dev,handle) \ @@ -724,19 +481,6 @@ struct cgs_device #define cgs_write_ind_register(dev,space,index,value) \ CGS_CALL(write_ind_register,dev,space,index,value) -#define cgs_read_pci_config_byte(dev,addr) \ - CGS_CALL(read_pci_config_byte,dev,addr) -#define cgs_read_pci_config_word(dev,addr) \ - CGS_CALL(read_pci_config_word,dev,addr) -#define cgs_read_pci_config_dword(dev,addr) \ - CGS_CALL(read_pci_config_dword,dev,addr) -#define cgs_write_pci_config_byte(dev,addr,value) \ - CGS_CALL(write_pci_config_byte,dev,addr,value) -#define cgs_write_pci_config_word(dev,addr,value) \ - CGS_CALL(write_pci_config_word,dev,addr,value) -#define cgs_write_pci_config_dword(dev,addr,value) \ - CGS_CALL(write_pci_config_dword,dev,addr,value) - #define cgs_atom_get_data_table(dev,table,size,frev,crev) \ CGS_CALL(atom_get_data_table,dev,table,size,frev,crev) #define cgs_atom_get_cmd_table_revs(dev,table,frev,crev) \ @@ -744,20 +488,6 @@ struct cgs_device #define cgs_atom_exec_cmd_table(dev,table,args) \ CGS_CALL(atom_exec_cmd_table,dev,table,args) -#define cgs_create_pm_request(dev,request) \ - CGS_CALL(create_pm_request,dev,request) -#define cgs_destroy_pm_request(dev,request) \ - CGS_CALL(destroy_pm_request,dev,request) -#define cgs_set_pm_request(dev,request,active) \ - CGS_CALL(set_pm_request,dev,request,active) -#define cgs_pm_request_clock(dev,request,clock,freq) \ - CGS_CALL(pm_request_clock,dev,request,clock,freq) -#define cgs_pm_request_engine(dev,request,engine,powered) \ - CGS_CALL(pm_request_engine,dev,request,engine,powered) -#define cgs_pm_query_clock_limits(dev,clock,limits) \ - CGS_CALL(pm_query_clock_limits,dev,clock,limits) -#define cgs_set_camera_voltages(dev,mask,voltages) \ - CGS_CALL(set_camera_voltages,dev,mask,voltages) #define cgs_get_firmware_info(dev, type, info) \ CGS_CALL(get_firmware_info, dev, type, info) #define cgs_rel_firmware(dev, type) \ diff --git a/drivers/gpu/drm/amd/powerplay/Makefile b/drivers/gpu/drm/amd/powerplay/Makefile index 043e6ebab575..4e132b936e3d 100644 --- a/drivers/gpu/drm/amd/powerplay/Makefile +++ b/drivers/gpu/drm/amd/powerplay/Makefile @@ -1,5 +1,5 @@ -subdir-ccflags-y += -Iinclude/drm \ +subdir-ccflags-y += \ -I$(FULL_AMD_PATH)/powerplay/inc/ \ -I$(FULL_AMD_PATH)/include/asic_reg \ -I$(FULL_AMD_PATH)/include \ diff --git a/drivers/gpu/drm/amd/powerplay/amd_powerplay.c b/drivers/gpu/drm/amd/powerplay/amd_powerplay.c index 9da5b0bb66d8..f73e80c4bf33 100644 --- a/drivers/gpu/drm/amd/powerplay/amd_powerplay.c +++ b/drivers/gpu/drm/amd/powerplay/amd_powerplay.c @@ -251,7 +251,9 @@ static int pp_suspend(void *handle) ret = pp_check(pp_handle); - if (ret != 0) + if (ret == PP_DPM_DISABLED) + return 0; + else if (ret != 0) return ret; eventmgr = pp_handle->eventmgr; diff --git a/drivers/gpu/drm/amd/powerplay/eventmgr/eventsubchains.c b/drivers/gpu/drm/amd/powerplay/eventmgr/eventsubchains.c index 9ef2d90e2886..b82c43af59ab 100644 --- a/drivers/gpu/drm/amd/powerplay/eventmgr/eventsubchains.c +++ b/drivers/gpu/drm/amd/powerplay/eventmgr/eventsubchains.c @@ -219,7 +219,7 @@ const pem_event_action notify_smu_suspend_tasks[] = { }; const pem_event_action disable_smc_firmware_ctf_tasks[] = { - /* PEM_Task_DisableSMCFirmwareCTF,*/ + pem_task_disable_smc_firmware_ctf, NULL }; diff --git a/drivers/gpu/drm/amd/powerplay/eventmgr/eventtasks.c b/drivers/gpu/drm/amd/powerplay/eventmgr/eventtasks.c index e04216ec7ee1..8c4ebaae1e0c 100644 --- a/drivers/gpu/drm/amd/powerplay/eventmgr/eventtasks.c +++ b/drivers/gpu/drm/amd/powerplay/eventmgr/eventtasks.c @@ -173,6 +173,11 @@ int pem_task_stop_asic_block_usage(struct pp_eventmgr *eventmgr, struct pem_even return 0; } +int pem_task_disable_smc_firmware_ctf(struct pp_eventmgr *eventmgr, struct pem_event_data *event_data) +{ + return phm_disable_smc_firmware_ctf(eventmgr->hwmgr); +} + int pem_task_setup_asic(struct pp_eventmgr *eventmgr, struct pem_event_data *event_data) { return phm_setup_asic(eventmgr->hwmgr); diff --git a/drivers/gpu/drm/amd/powerplay/eventmgr/eventtasks.h b/drivers/gpu/drm/amd/powerplay/eventmgr/eventtasks.h index 6c6297e3b598..37e7ca5a58e0 100644 --- a/drivers/gpu/drm/amd/powerplay/eventmgr/eventtasks.h +++ b/drivers/gpu/drm/amd/powerplay/eventmgr/eventtasks.h @@ -84,5 +84,6 @@ int pem_task_update_allowed_performance_levels(struct pp_eventmgr *eventmgr, str /*thermal */ int pem_task_initialize_thermal_controller(struct pp_eventmgr *eventmgr, struct pem_event_data *event_data); int pem_task_uninitialize_thermal_controller(struct pp_eventmgr *eventmgr, struct pem_event_data *event_data); +int pem_task_disable_smc_firmware_ctf(struct pp_eventmgr *eventmgr, struct pem_event_data *event_data); #endif /* _EVENT_TASKS_H_ */ diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/hardwaremanager.c b/drivers/gpu/drm/amd/powerplay/hwmgr/hardwaremanager.c index 23bba2c8b18e..fcc722ea7649 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/hardwaremanager.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/hardwaremanager.c @@ -501,3 +501,13 @@ int phm_get_max_high_clocks(struct pp_hwmgr *hwmgr, struct amd_pp_simple_clock_i return hwmgr->hwmgr_func->get_max_high_clocks(hwmgr, clocks); } + +int phm_disable_smc_firmware_ctf(struct pp_hwmgr *hwmgr) +{ + PHM_FUNC_CHECK(hwmgr); + + if (hwmgr->hwmgr_func->disable_smc_firmware_ctf == NULL) + return -EINVAL; + + return hwmgr->hwmgr_func->disable_smc_firmware_ctf(hwmgr); +} diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c index ff4ae3de6bb6..963a9e017a28 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c @@ -22,10 +22,10 @@ */ #include "pp_debug.h" -#include "linux/delay.h" -#include <linux/types.h> +#include <linux/delay.h> #include <linux/kernel.h> #include <linux/slab.h> +#include <linux/types.h> #include <drm/amdgpu_drm.h> #include "cgs_common.h" #include "power_state.h" diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/pp_acpi.c b/drivers/gpu/drm/amd/powerplay/hwmgr/pp_acpi.c index f5e8fda964f7..f6b4dd96c0ec 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/pp_acpi.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/pp_acpi.c @@ -21,8 +21,8 @@ * */ +#include <linux/delay.h> #include <linux/errno.h> -#include "linux/delay.h" #include "hwmgr.h" #include "amd_acpi.h" #include "pp_acpi.h" diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomfwctrl.c b/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomfwctrl.c index b71525f838e6..56023114ad6f 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomfwctrl.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomfwctrl.c @@ -314,52 +314,45 @@ int pp_atomfwctrl_get_avfs_information(struct pp_hwmgr *hwmgr, le32_to_cpu(profile->gb_vdroop_table_ckson_a2); param->ulGbFuseTableCksoffM1 = le32_to_cpu(profile->avfsgb_fuse_table_cksoff_m1); - param->usGbFuseTableCksoffM2 = + param->ulGbFuseTableCksoffM2 = le16_to_cpu(profile->avfsgb_fuse_table_cksoff_m2); param->ulGbFuseTableCksoffB = le32_to_cpu(profile->avfsgb_fuse_table_cksoff_b); param->ulGbFuseTableCksonM1 = le32_to_cpu(profile->avfsgb_fuse_table_ckson_m1); - param->usGbFuseTableCksonM2 = + param->ulGbFuseTableCksonM2 = le16_to_cpu(profile->avfsgb_fuse_table_ckson_m2); param->ulGbFuseTableCksonB = le32_to_cpu(profile->avfsgb_fuse_table_ckson_b); - param->usMaxVoltage025mv = - le16_to_cpu(profile->max_voltage_0_25mv); - param->ucEnableGbVdroopTableCksoff = - profile->enable_gb_vdroop_table_cksoff; + param->ucEnableGbVdroopTableCkson = profile->enable_gb_vdroop_table_ckson; - param->ucEnableGbFuseTableCksoff = - profile->enable_gb_fuse_table_cksoff; param->ucEnableGbFuseTableCkson = profile->enable_gb_fuse_table_ckson; param->usPsmAgeComfactor = le16_to_cpu(profile->psm_age_comfactor); - param->ucEnableApplyAvfsCksoffVoltage = - profile->enable_apply_avfs_cksoff_voltage; param->ulDispclk2GfxclkM1 = le32_to_cpu(profile->dispclk2gfxclk_a); - param->usDispclk2GfxclkM2 = + param->ulDispclk2GfxclkM2 = le16_to_cpu(profile->dispclk2gfxclk_b); param->ulDispclk2GfxclkB = le32_to_cpu(profile->dispclk2gfxclk_c); param->ulDcefclk2GfxclkM1 = le32_to_cpu(profile->dcefclk2gfxclk_a); - param->usDcefclk2GfxclkM2 = + param->ulDcefclk2GfxclkM2 = le16_to_cpu(profile->dcefclk2gfxclk_b); param->ulDcefclk2GfxclkB = le32_to_cpu(profile->dcefclk2gfxclk_c); param->ulPixelclk2GfxclkM1 = le32_to_cpu(profile->pixclk2gfxclk_a); - param->usPixelclk2GfxclkM2 = + param->ulPixelclk2GfxclkM2 = le16_to_cpu(profile->pixclk2gfxclk_b); param->ulPixelclk2GfxclkB = le32_to_cpu(profile->pixclk2gfxclk_c); param->ulPhyclk2GfxclkM1 = le32_to_cpu(profile->phyclk2gfxclk_a); - param->usPhyclk2GfxclkM2 = + param->ulPhyclk2GfxclkM2 = le16_to_cpu(profile->phyclk2gfxclk_b); param->ulPhyclk2GfxclkB = le32_to_cpu(profile->phyclk2gfxclk_c); @@ -394,3 +387,31 @@ int pp_atomfwctrl_get_gpio_information(struct pp_hwmgr *hwmgr, return 0; } + +int pp_atomfwctrl_get_vbios_bootup_values(struct pp_hwmgr *hwmgr, + struct pp_atomfwctrl_bios_boot_up_values *boot_values) +{ + struct atom_firmware_info_v3_1 *info = NULL; + uint16_t ix; + + ix = GetIndexIntoMasterDataTable(firmwareinfo); + info = (struct atom_firmware_info_v3_1 *) + cgs_atom_get_data_table(hwmgr->device, + ix, NULL, NULL, NULL); + + if (!info) { + pr_info("Error retrieving BIOS firmwareinfo!"); + return -EINVAL; + } + + boot_values->ulRevision = info->firmware_revision; + boot_values->ulGfxClk = info->bootup_sclk_in10khz; + boot_values->ulUClk = info->bootup_mclk_in10khz; + boot_values->ulSocClk = 0; + boot_values->usVddc = info->bootup_vddc_mv; + boot_values->usVddci = info->bootup_vddci_mv; + boot_values->usMvddc = info->bootup_mvddc_mv; + boot_values->usVddGfx = info->bootup_vddgfx_mv; + + return 0; +}
\ No newline at end of file diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomfwctrl.h b/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomfwctrl.h index 7efe9b96cb33..43a6711e3c06 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomfwctrl.h +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomfwctrl.h @@ -69,7 +69,7 @@ struct pp_atomfwctrl_clock_dividers_soc15 { struct pp_atomfwctrl_avfs_parameters { uint32_t ulMaxVddc; uint32_t ulMinVddc; - uint8_t ucMaxVidStep; + uint32_t ulMeanNsigmaAcontant0; uint32_t ulMeanNsigmaAcontant1; uint32_t ulMeanNsigmaAcontant2; @@ -82,30 +82,30 @@ struct pp_atomfwctrl_avfs_parameters { uint32_t ulGbVdroopTableCksonA0; uint32_t ulGbVdroopTableCksonA1; uint32_t ulGbVdroopTableCksonA2; + uint32_t ulGbFuseTableCksoffM1; - uint16_t usGbFuseTableCksoffM2; - uint32_t ulGbFuseTableCksoffB;\ + uint32_t ulGbFuseTableCksoffM2; + uint32_t ulGbFuseTableCksoffB; + uint32_t ulGbFuseTableCksonM1; - uint16_t usGbFuseTableCksonM2; + uint32_t ulGbFuseTableCksonM2; uint32_t ulGbFuseTableCksonB; - uint16_t usMaxVoltage025mv; - uint8_t ucEnableGbVdroopTableCksoff; + uint8_t ucEnableGbVdroopTableCkson; - uint8_t ucEnableGbFuseTableCksoff; uint8_t ucEnableGbFuseTableCkson; uint16_t usPsmAgeComfactor; - uint8_t ucEnableApplyAvfsCksoffVoltage; + uint32_t ulDispclk2GfxclkM1; - uint16_t usDispclk2GfxclkM2; + uint32_t ulDispclk2GfxclkM2; uint32_t ulDispclk2GfxclkB; uint32_t ulDcefclk2GfxclkM1; - uint16_t usDcefclk2GfxclkM2; + uint32_t ulDcefclk2GfxclkM2; uint32_t ulDcefclk2GfxclkB; uint32_t ulPixelclk2GfxclkM1; - uint16_t usPixelclk2GfxclkM2; + uint32_t ulPixelclk2GfxclkM2; uint32_t ulPixelclk2GfxclkB; uint32_t ulPhyclk2GfxclkM1; - uint16_t usPhyclk2GfxclkM2; + uint32_t ulPhyclk2GfxclkM2; uint32_t ulPhyclk2GfxclkB; }; @@ -119,6 +119,18 @@ struct pp_atomfwctrl_gpio_parameters { uint8_t ucFwCtfGpio; uint8_t ucFwCtfPolarity; }; + +struct pp_atomfwctrl_bios_boot_up_values { + uint32_t ulRevision; + uint32_t ulGfxClk; + uint32_t ulUClk; + uint32_t ulSocClk; + uint16_t usVddc; + uint16_t usVddci; + uint16_t usMvddc; + uint16_t usVddGfx; +}; + int pp_atomfwctrl_get_gpu_pll_dividers_vega10(struct pp_hwmgr *hwmgr, uint32_t clock_type, uint32_t clock_value, struct pp_atomfwctrl_clock_dividers_soc15 *dividers); @@ -136,5 +148,8 @@ int pp_atomfwctrl_get_avfs_information(struct pp_hwmgr *hwmgr, int pp_atomfwctrl_get_gpio_information(struct pp_hwmgr *hwmgr, struct pp_atomfwctrl_gpio_parameters *param); +int pp_atomfwctrl_get_vbios_bootup_values(struct pp_hwmgr *hwmgr, + struct pp_atomfwctrl_bios_boot_up_values *boot_values); + #endif diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c index 8f663ab56a80..975567f6813d 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c @@ -21,11 +21,11 @@ * */ #include "pp_debug.h" +#include <linux/delay.h> +#include <linux/fb.h> #include <linux/module.h> #include <linux/slab.h> -#include <linux/fb.h> #include <asm/div64.h> -#include "linux/delay.h" #include "pp_acpi.h" #include "ppatomctrl.h" #include "atombios.h" @@ -4334,26 +4334,31 @@ static int smu7_print_clock_levels(struct pp_hwmgr *hwmgr, static int smu7_set_fan_control_mode(struct pp_hwmgr *hwmgr, uint32_t mode) { - if (mode) { - /* stop auto-manage */ - if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, - PHM_PlatformCaps_MicrocodeFanControl)) - smu7_fan_ctrl_stop_smc_fan_control(hwmgr); - smu7_fan_ctrl_set_static_mode(hwmgr, mode); - } else - /* restart auto-manage */ - smu7_fan_ctrl_reset_fan_speed_to_default(hwmgr); + int result = 0; - return 0; + switch (mode) { + case AMD_FAN_CTRL_NONE: + result = smu7_fan_ctrl_set_fan_speed_percent(hwmgr, 100); + break; + case AMD_FAN_CTRL_MANUAL: + if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, + PHM_PlatformCaps_MicrocodeFanControl)) + result = smu7_fan_ctrl_stop_smc_fan_control(hwmgr); + break; + case AMD_FAN_CTRL_AUTO: + result = smu7_fan_ctrl_set_static_mode(hwmgr, mode); + if (!result) + result = smu7_fan_ctrl_start_smc_fan_control(hwmgr); + break; + default: + break; + } + return result; } static int smu7_get_fan_control_mode(struct pp_hwmgr *hwmgr) { - if (hwmgr->fan_ctrl_is_in_default_mode) - return hwmgr->fan_ctrl_default_mode; - else - return PHM_READ_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, - CG_FDO_CTRL2, FDO_PWM_MODE); + return hwmgr->fan_ctrl_enabled ? AMD_FAN_CTRL_AUTO : AMD_FAN_CTRL_MANUAL; } static int smu7_get_sclk_od(struct pp_hwmgr *hwmgr) @@ -4522,32 +4527,6 @@ static int smu7_get_clock_by_type(struct pp_hwmgr *hwmgr, enum amd_pp_clock_type return 0; } -static int smu7_request_firmware(struct pp_hwmgr *hwmgr) -{ - int ret; - struct cgs_firmware_info info = {0}; - - ret = cgs_get_firmware_info(hwmgr->device, - smu7_convert_fw_type_to_cgs(UCODE_ID_SMU), - &info); - if (ret || !info.kptr) - return -EINVAL; - - return 0; -} - -static int smu7_release_firmware(struct pp_hwmgr *hwmgr) -{ - int ret; - - ret = cgs_rel_firmware(hwmgr->device, - smu7_convert_fw_type_to_cgs(UCODE_ID_SMU)); - if (ret) - return -EINVAL; - - return 0; -} - static void smu7_find_min_clock_masks(struct pp_hwmgr *hwmgr, uint32_t *sclk_mask, uint32_t *mclk_mask, uint32_t min_sclk, uint32_t min_mclk) @@ -4691,10 +4670,9 @@ static const struct pp_hwmgr_func smu7_hwmgr_funcs = { .get_clock_by_type = smu7_get_clock_by_type, .read_sensor = smu7_read_sensor, .dynamic_state_management_disable = smu7_disable_dpm_tasks, - .request_firmware = smu7_request_firmware, - .release_firmware = smu7_release_firmware, .set_power_profile_state = smu7_set_power_profile_state, .avfs_control = smu7_avfs_control, + .disable_smc_firmware_ctf = smu7_thermal_disable_alert, }; uint8_t smu7_get_sleep_divider_id_from_clock(uint32_t clock, diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_thermal.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_thermal.c index 436ca5ce8248..baddb569a8b8 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_thermal.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_thermal.c @@ -112,10 +112,9 @@ int smu7_fan_ctrl_get_fan_speed_rpm(struct pp_hwmgr *hwmgr, uint32_t *speed) */ int smu7_fan_ctrl_set_static_mode(struct pp_hwmgr *hwmgr, uint32_t mode) { - if (hwmgr->fan_ctrl_is_in_default_mode) { hwmgr->fan_ctrl_default_mode = - PHM_READ_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, + PHM_READ_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, CG_FDO_CTRL2, FDO_PWM_MODE); hwmgr->tmin = PHM_READ_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, @@ -149,7 +148,7 @@ int smu7_fan_ctrl_set_default_mode(struct pp_hwmgr *hwmgr) return 0; } -static int smu7_fan_ctrl_start_smc_fan_control(struct pp_hwmgr *hwmgr) +int smu7_fan_ctrl_start_smc_fan_control(struct pp_hwmgr *hwmgr) { int result; @@ -179,6 +178,7 @@ static int smu7_fan_ctrl_start_smc_fan_control(struct pp_hwmgr *hwmgr) PPSMC_MSG_SetFanTemperatureTarget, hwmgr->thermal_controller. advanceFanControlParameters.ucTargetTemperature); + hwmgr->fan_ctrl_enabled = true; return result; } @@ -186,6 +186,7 @@ static int smu7_fan_ctrl_start_smc_fan_control(struct pp_hwmgr *hwmgr) int smu7_fan_ctrl_stop_smc_fan_control(struct pp_hwmgr *hwmgr) { + hwmgr->fan_ctrl_enabled = false; return smum_send_msg_to_smc(hwmgr->smumgr, PPSMC_StopFanControl); } @@ -280,7 +281,7 @@ int smu7_fan_ctrl_set_fan_speed_rpm(struct pp_hwmgr *hwmgr, uint32_t speed) PHM_WRITE_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, CG_TACH_STATUS, TACH_PERIOD, tach_period); - return smu7_fan_ctrl_set_static_mode(hwmgr, FDO_PWM_MODE_STATIC); + return smu7_fan_ctrl_set_static_mode(hwmgr, FDO_PWM_MODE_STATIC_RPM); } /** diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_thermal.h b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_thermal.h index 2ed774db42c7..ba71b608fa75 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_thermal.h +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_thermal.h @@ -54,6 +54,6 @@ extern int smu7_fan_ctrl_get_fan_speed_rpm(struct pp_hwmgr *hwmgr, uint32_t *spe extern int smu7_fan_ctrl_stop_smc_fan_control(struct pp_hwmgr *hwmgr); extern int smu7_thermal_enable_alert(struct pp_hwmgr *hwmgr); extern int smu7_thermal_disable_alert(struct pp_hwmgr *hwmgr); - +extern int smu7_fan_ctrl_start_smc_fan_control(struct pp_hwmgr *hwmgr); #endif diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c index 83949550edac..cd7bf6d3859a 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c @@ -20,10 +20,11 @@ * OTHER DEALINGS IN THE SOFTWARE. * */ + +#include <linux/delay.h> +#include <linux/fb.h> #include <linux/module.h> #include <linux/slab.h> -#include <linux/fb.h> -#include "linux/delay.h" #include "hwmgr.h" #include "amd_powerplay.h" @@ -111,6 +112,8 @@ static void vega10_set_default_registry_data(struct pp_hwmgr *hwmgr) hwmgr->feature_mask & PP_SOCCLK_DPM_MASK ? false : true; data->registry_data.mclk_dpm_key_disabled = hwmgr->feature_mask & PP_MCLK_DPM_MASK ? false : true; + data->registry_data.pcie_dpm_key_disabled = + hwmgr->feature_mask & PP_PCIE_DPM_MASK ? false : true; data->registry_data.dcefclk_dpm_key_disabled = hwmgr->feature_mask & PP_DCEFCLK_DPM_MASK ? false : true; @@ -121,7 +124,9 @@ static void vega10_set_default_registry_data(struct pp_hwmgr *hwmgr) data->registry_data.enable_tdc_limit_feature = 1; } - data->registry_data.pcie_dpm_key_disabled = 1; + data->registry_data.clock_stretcher_support = + hwmgr->feature_mask & PP_CLOCK_STRETCH_MASK ? false : true; + data->registry_data.disable_water_mark = 0; data->registry_data.fan_control_support = 1; @@ -1133,7 +1138,7 @@ static void vega10_setup_default_single_dpm_table(struct pp_hwmgr *hwmgr, int i; for (i = 0; i < dep_table->count; i++) { - if (i == 0 || dpm_table->dpm_levels[dpm_table->count - 1].value != + if (i == 0 || dpm_table->dpm_levels[dpm_table->count - 1].value <= dep_table->entries[i].clk) { dpm_table->dpm_levels[dpm_table->count].value = dep_table->entries[i].clk; @@ -1178,29 +1183,9 @@ static int vega10_setup_default_pcie_table(struct pp_hwmgr *hwmgr) else pcie_table->lclk[i] = bios_pcie_table->entries[i].pcie_sclk; - - pcie_table->count++; } - if (data->registry_data.pcieSpeedOverride) - pcie_table->pcie_gen[i] = data->registry_data.pcieSpeedOverride; - else - pcie_table->pcie_gen[i] = - bios_pcie_table->entries[bios_pcie_table->count - 1].gen_speed; - - if (data->registry_data.pcieLaneOverride) - pcie_table->pcie_lane[i] = data->registry_data.pcieLaneOverride; - else - pcie_table->pcie_lane[i] = - bios_pcie_table->entries[bios_pcie_table->count - 1].lane_width; - - if (data->registry_data.pcieClockOverride) - pcie_table->lclk[i] = data->registry_data.pcieClockOverride; - else - pcie_table->lclk[i] = - bios_pcie_table->entries[bios_pcie_table->count - 1].pcie_sclk; - - pcie_table->count++; + pcie_table->count = NUM_LINK_LEVELS; return 0; } @@ -1290,7 +1275,7 @@ static int vega10_setup_default_dpm_tables(struct pp_hwmgr *hwmgr) dpm_table = &(data->dpm_table.eclk_table); for (i = 0; i < dep_mm_table->count; i++) { if (i == 0 || dpm_table->dpm_levels - [dpm_table->count - 1].value != + [dpm_table->count - 1].value <= dep_mm_table->entries[i].eclk) { dpm_table->dpm_levels[dpm_table->count].value = dep_mm_table->entries[i].eclk; @@ -1306,7 +1291,7 @@ static int vega10_setup_default_dpm_tables(struct pp_hwmgr *hwmgr) dpm_table = &(data->dpm_table.vclk_table); for (i = 0; i < dep_mm_table->count; i++) { if (i == 0 || dpm_table->dpm_levels - [dpm_table->count - 1].value != + [dpm_table->count - 1].value <= dep_mm_table->entries[i].vclk) { dpm_table->dpm_levels[dpm_table->count].value = dep_mm_table->entries[i].vclk; @@ -1320,7 +1305,7 @@ static int vega10_setup_default_dpm_tables(struct pp_hwmgr *hwmgr) dpm_table = &(data->dpm_table.dclk_table); for (i = 0; i < dep_mm_table->count; i++) { if (i == 0 || dpm_table->dpm_levels - [dpm_table->count - 1].value != + [dpm_table->count - 1].value <= dep_mm_table->entries[i].dclk) { dpm_table->dpm_levels[dpm_table->count].value = dep_mm_table->entries[i].dclk; @@ -1432,9 +1417,7 @@ static int vega10_populate_ulv_state(struct pp_hwmgr *hwmgr) (struct phm_ppt_v2_information *)(hwmgr->pptable); data->smc_state_table.pp_table.UlvOffsetVid = - (uint8_t)(table_info->us_ulv_voltage_offset * - VOLTAGE_VID_OFFSET_SCALE2 / - VOLTAGE_VID_OFFSET_SCALE1); + (uint8_t)table_info->us_ulv_voltage_offset; data->smc_state_table.pp_table.UlvSmnclkDid = (uint8_t)(table_info->us_ulv_smnclk_did); @@ -1553,7 +1536,11 @@ static int vega10_populate_single_gfx_level(struct pp_hwmgr *hwmgr, current_gfxclk_level->FbMult = cpu_to_le32(dividers.ulPll_fb_mult); /* Spread FB Multiplier bit: bit 0:8 int, bit 31:16 frac */ - current_gfxclk_level->SsOn = dividers.ucPll_ss_enable; + if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, + PHM_PlatformCaps_EngineSpreadSpectrumSupport)) + current_gfxclk_level->SsOn = dividers.ucPll_ss_enable; + else + current_gfxclk_level->SsOn = 0; current_gfxclk_level->SsFbMult = cpu_to_le32(dividers.ulPll_ss_fbsmult); current_gfxclk_level->SsSlewFrac = @@ -2044,10 +2031,10 @@ static int vega10_populate_clock_stretcher_table(struct pp_hwmgr *hwmgr) table_info->vdd_dep_on_sclk; uint32_t i; - for (i = 0; dep_table->count; i++) { + for (i = 0; i < dep_table->count; i++) { pp_table->CksEnable[i] = dep_table->entries[i].cks_enable; - pp_table->CksVidOffset[i] = convert_to_vid( - dep_table->entries[i].cks_voffset); + pp_table->CksVidOffset[i] = (uint8_t)(dep_table->entries[i].cks_voffset + * VOLTAGE_VID_OFFSET_SCALE2 / VOLTAGE_VID_OFFSET_SCALE1); } return 0; @@ -2073,66 +2060,70 @@ static int vega10_populate_avfs_parameters(struct pp_hwmgr *hwmgr) result = pp_atomfwctrl_get_avfs_information(hwmgr, &avfs_params); if (!result) { pp_table->MinVoltageVid = (uint8_t) - convert_to_vid((uint16_t)(avfs_params.ulMaxVddc)); - pp_table->MaxVoltageVid = (uint8_t) convert_to_vid((uint16_t)(avfs_params.ulMinVddc)); - pp_table->BtcGbVdroopTableCksOn.a0 = - cpu_to_le32(avfs_params.ulGbVdroopTableCksonA0); - pp_table->BtcGbVdroopTableCksOn.a1 = - cpu_to_le32(avfs_params.ulGbVdroopTableCksonA1); - pp_table->BtcGbVdroopTableCksOn.a2 = - cpu_to_le32(avfs_params.ulGbVdroopTableCksonA2); + pp_table->MaxVoltageVid = (uint8_t) + convert_to_vid((uint16_t)(avfs_params.ulMaxVddc)); + + pp_table->AConstant[0] = cpu_to_le32(avfs_params.ulMeanNsigmaAcontant0); + pp_table->AConstant[1] = cpu_to_le32(avfs_params.ulMeanNsigmaAcontant1); + pp_table->AConstant[2] = cpu_to_le32(avfs_params.ulMeanNsigmaAcontant2); + pp_table->DC_tol_sigma = cpu_to_le16(avfs_params.usMeanNsigmaDcTolSigma); + pp_table->Platform_mean = cpu_to_le16(avfs_params.usMeanNsigmaPlatformMean); + pp_table->Platform_sigma = cpu_to_le16(avfs_params.usMeanNsigmaDcTolSigma); + pp_table->PSM_Age_CompFactor = cpu_to_le16(avfs_params.usPsmAgeComfactor); pp_table->BtcGbVdroopTableCksOff.a0 = cpu_to_le32(avfs_params.ulGbVdroopTableCksoffA0); + pp_table->BtcGbVdroopTableCksOff.a0_shift = 20; pp_table->BtcGbVdroopTableCksOff.a1 = cpu_to_le32(avfs_params.ulGbVdroopTableCksoffA1); + pp_table->BtcGbVdroopTableCksOff.a1_shift = 20; pp_table->BtcGbVdroopTableCksOff.a2 = cpu_to_le32(avfs_params.ulGbVdroopTableCksoffA2); + pp_table->BtcGbVdroopTableCksOff.a2_shift = 20; + + pp_table->OverrideBtcGbCksOn = avfs_params.ucEnableGbVdroopTableCkson; + pp_table->BtcGbVdroopTableCksOn.a0 = + cpu_to_le32(avfs_params.ulGbVdroopTableCksonA0); + pp_table->BtcGbVdroopTableCksOn.a0_shift = 20; + pp_table->BtcGbVdroopTableCksOn.a1 = + cpu_to_le32(avfs_params.ulGbVdroopTableCksonA1); + pp_table->BtcGbVdroopTableCksOn.a1_shift = 20; + pp_table->BtcGbVdroopTableCksOn.a2 = + cpu_to_le32(avfs_params.ulGbVdroopTableCksonA2); + pp_table->BtcGbVdroopTableCksOn.a2_shift = 20; pp_table->AvfsGbCksOn.m1 = cpu_to_le32(avfs_params.ulGbFuseTableCksonM1); pp_table->AvfsGbCksOn.m2 = - cpu_to_le16(avfs_params.usGbFuseTableCksonM2); + cpu_to_le16(avfs_params.ulGbFuseTableCksonM2); pp_table->AvfsGbCksOn.b = cpu_to_le32(avfs_params.ulGbFuseTableCksonB); pp_table->AvfsGbCksOn.m1_shift = 24; pp_table->AvfsGbCksOn.m2_shift = 12; + pp_table->AvfsGbCksOn.b_shift = 0; + pp_table->OverrideAvfsGbCksOn = + avfs_params.ucEnableGbFuseTableCkson; pp_table->AvfsGbCksOff.m1 = cpu_to_le32(avfs_params.ulGbFuseTableCksoffM1); pp_table->AvfsGbCksOff.m2 = - cpu_to_le16(avfs_params.usGbFuseTableCksoffM2); + cpu_to_le16(avfs_params.ulGbFuseTableCksoffM2); pp_table->AvfsGbCksOff.b = cpu_to_le32(avfs_params.ulGbFuseTableCksoffB); pp_table->AvfsGbCksOff.m1_shift = 24; pp_table->AvfsGbCksOff.m2_shift = 12; - - pp_table->AConstant[0] = - cpu_to_le32(avfs_params.ulMeanNsigmaAcontant0); - pp_table->AConstant[1] = - cpu_to_le32(avfs_params.ulMeanNsigmaAcontant1); - pp_table->AConstant[2] = - cpu_to_le32(avfs_params.ulMeanNsigmaAcontant2); - pp_table->DC_tol_sigma = - cpu_to_le16(avfs_params.usMeanNsigmaDcTolSigma); - pp_table->Platform_mean = - cpu_to_le16(avfs_params.usMeanNsigmaPlatformMean); - pp_table->PSM_Age_CompFactor = - cpu_to_le16(avfs_params.usPsmAgeComfactor); - pp_table->Platform_sigma = - cpu_to_le16(avfs_params.usMeanNsigmaDcTolSigma); - - for (i = 0; i < dep_table->count; i++) - pp_table->StaticVoltageOffsetVid[i] = (uint8_t) - (dep_table->entries[i].sclk_offset * + pp_table->AvfsGbCksOff.b_shift = 0; + + for (i = 0; i < dep_table->count; i++) { + if (dep_table->entries[i].sclk_offset == 0) + pp_table->StaticVoltageOffsetVid[i] = 248; + else + pp_table->StaticVoltageOffsetVid[i] = + (uint8_t)(dep_table->entries[i].sclk_offset * VOLTAGE_VID_OFFSET_SCALE2 / VOLTAGE_VID_OFFSET_SCALE1); - - pp_table->OverrideBtcGbCksOn = - avfs_params.ucEnableGbVdroopTableCkson; - pp_table->OverrideAvfsGbCksOn = - avfs_params.ucEnableGbFuseTableCkson; + } if ((PPREGKEY_VEGA10QUADRATICEQUATION_DFLT != data->disp_clk_quad_eqn_a) && @@ -2141,20 +2132,21 @@ static int vega10_populate_avfs_parameters(struct pp_hwmgr *hwmgr) pp_table->DisplayClock2Gfxclk[DSPCLK_DISPCLK].m1 = (int32_t)data->disp_clk_quad_eqn_a; pp_table->DisplayClock2Gfxclk[DSPCLK_DISPCLK].m2 = - (int16_t)data->disp_clk_quad_eqn_b; + (int32_t)data->disp_clk_quad_eqn_b; pp_table->DisplayClock2Gfxclk[DSPCLK_DISPCLK].b = (int32_t)data->disp_clk_quad_eqn_c; } else { pp_table->DisplayClock2Gfxclk[DSPCLK_DISPCLK].m1 = (int32_t)avfs_params.ulDispclk2GfxclkM1; pp_table->DisplayClock2Gfxclk[DSPCLK_DISPCLK].m2 = - (int16_t)avfs_params.usDispclk2GfxclkM2; + (int32_t)avfs_params.ulDispclk2GfxclkM2; pp_table->DisplayClock2Gfxclk[DSPCLK_DISPCLK].b = (int32_t)avfs_params.ulDispclk2GfxclkB; } pp_table->DisplayClock2Gfxclk[DSPCLK_DISPCLK].m1_shift = 24; pp_table->DisplayClock2Gfxclk[DSPCLK_DISPCLK].m2_shift = 12; + pp_table->DisplayClock2Gfxclk[DSPCLK_DISPCLK].b_shift = 12; if ((PPREGKEY_VEGA10QUADRATICEQUATION_DFLT != data->dcef_clk_quad_eqn_a) && @@ -2163,20 +2155,21 @@ static int vega10_populate_avfs_parameters(struct pp_hwmgr *hwmgr) pp_table->DisplayClock2Gfxclk[DSPCLK_DCEFCLK].m1 = (int32_t)data->dcef_clk_quad_eqn_a; pp_table->DisplayClock2Gfxclk[DSPCLK_DCEFCLK].m2 = - (int16_t)data->dcef_clk_quad_eqn_b; + (int32_t)data->dcef_clk_quad_eqn_b; pp_table->DisplayClock2Gfxclk[DSPCLK_DCEFCLK].b = (int32_t)data->dcef_clk_quad_eqn_c; } else { pp_table->DisplayClock2Gfxclk[DSPCLK_DCEFCLK].m1 = (int32_t)avfs_params.ulDcefclk2GfxclkM1; pp_table->DisplayClock2Gfxclk[DSPCLK_DCEFCLK].m2 = - (int16_t)avfs_params.usDcefclk2GfxclkM2; + (int32_t)avfs_params.ulDcefclk2GfxclkM2; pp_table->DisplayClock2Gfxclk[DSPCLK_DCEFCLK].b = (int32_t)avfs_params.ulDcefclk2GfxclkB; } pp_table->DisplayClock2Gfxclk[DSPCLK_DCEFCLK].m1_shift = 24; pp_table->DisplayClock2Gfxclk[DSPCLK_DCEFCLK].m2_shift = 12; + pp_table->DisplayClock2Gfxclk[DSPCLK_DCEFCLK].b_shift = 12; if ((PPREGKEY_VEGA10QUADRATICEQUATION_DFLT != data->pixel_clk_quad_eqn_a) && @@ -2185,21 +2178,21 @@ static int vega10_populate_avfs_parameters(struct pp_hwmgr *hwmgr) pp_table->DisplayClock2Gfxclk[DSPCLK_PIXCLK].m1 = (int32_t)data->pixel_clk_quad_eqn_a; pp_table->DisplayClock2Gfxclk[DSPCLK_PIXCLK].m2 = - (int16_t)data->pixel_clk_quad_eqn_b; + (int32_t)data->pixel_clk_quad_eqn_b; pp_table->DisplayClock2Gfxclk[DSPCLK_PIXCLK].b = (int32_t)data->pixel_clk_quad_eqn_c; } else { pp_table->DisplayClock2Gfxclk[DSPCLK_PIXCLK].m1 = (int32_t)avfs_params.ulPixelclk2GfxclkM1; pp_table->DisplayClock2Gfxclk[DSPCLK_PIXCLK].m2 = - (int16_t)avfs_params.usPixelclk2GfxclkM2; + (int32_t)avfs_params.ulPixelclk2GfxclkM2; pp_table->DisplayClock2Gfxclk[DSPCLK_PIXCLK].b = (int32_t)avfs_params.ulPixelclk2GfxclkB; } pp_table->DisplayClock2Gfxclk[DSPCLK_PIXCLK].m1_shift = 24; pp_table->DisplayClock2Gfxclk[DSPCLK_PIXCLK].m2_shift = 12; - + pp_table->DisplayClock2Gfxclk[DSPCLK_PIXCLK].b_shift = 12; if ((PPREGKEY_VEGA10QUADRATICEQUATION_DFLT != data->phy_clk_quad_eqn_a) && (PPREGKEY_VEGA10QUADRATICEQUATION_DFLT != @@ -2207,20 +2200,21 @@ static int vega10_populate_avfs_parameters(struct pp_hwmgr *hwmgr) pp_table->DisplayClock2Gfxclk[DSPCLK_PHYCLK].m1 = (int32_t)data->phy_clk_quad_eqn_a; pp_table->DisplayClock2Gfxclk[DSPCLK_PHYCLK].m2 = - (int16_t)data->phy_clk_quad_eqn_b; + (int32_t)data->phy_clk_quad_eqn_b; pp_table->DisplayClock2Gfxclk[DSPCLK_PHYCLK].b = (int32_t)data->phy_clk_quad_eqn_c; } else { pp_table->DisplayClock2Gfxclk[DSPCLK_PHYCLK].m1 = (int32_t)avfs_params.ulPhyclk2GfxclkM1; pp_table->DisplayClock2Gfxclk[DSPCLK_PHYCLK].m2 = - (int16_t)avfs_params.usPhyclk2GfxclkM2; + (int32_t)avfs_params.ulPhyclk2GfxclkM2; pp_table->DisplayClock2Gfxclk[DSPCLK_PHYCLK].b = (int32_t)avfs_params.ulPhyclk2GfxclkB; } pp_table->DisplayClock2Gfxclk[DSPCLK_PHYCLK].m1_shift = 24; pp_table->DisplayClock2Gfxclk[DSPCLK_PHYCLK].m2_shift = 12; + pp_table->DisplayClock2Gfxclk[DSPCLK_PHYCLK].b_shift = 12; } else { data->smu_features[GNLD_AVFS].supported = false; } @@ -2309,6 +2303,7 @@ static int vega10_init_smc_table(struct pp_hwmgr *hwmgr) (struct phm_ppt_v2_information *)(hwmgr->pptable); PPTable_t *pp_table = &(data->smc_state_table.pp_table); struct pp_atomfwctrl_voltage_table voltage_table; + struct pp_atomfwctrl_bios_boot_up_values boot_up_values; result = vega10_setup_default_dpm_tables(hwmgr); PP_ASSERT_WITH_CODE(!result, @@ -2331,6 +2326,7 @@ static int vega10_init_smc_table(struct pp_hwmgr *hwmgr) (uint8_t)(table_info->uc_vce_dpm_voltage_mode); pp_table->Mp0DpmVoltageMode = (uint8_t)(table_info->uc_mp0_dpm_voltage_mode); + pp_table->DisplayDpmVoltageMode = (uint8_t)(table_info->uc_dcef_dpm_voltage_mode); @@ -2372,14 +2368,31 @@ static int vega10_init_smc_table(struct pp_hwmgr *hwmgr) "Failed to initialize UVD Level!", return result); - if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, - PHM_PlatformCaps_ClockStretcher)) { + if (data->registry_data.clock_stretcher_support) { result = vega10_populate_clock_stretcher_table(hwmgr); PP_ASSERT_WITH_CODE(!result, "Failed to populate Clock Stretcher Table!", return result); } + result = pp_atomfwctrl_get_vbios_bootup_values(hwmgr, &boot_up_values); + if (!result) { + data->vbios_boot_state.vddc = boot_up_values.usVddc; + data->vbios_boot_state.vddci = boot_up_values.usVddci; + data->vbios_boot_state.mvddc = boot_up_values.usMvddc; + data->vbios_boot_state.gfx_clock = boot_up_values.ulGfxClk; + data->vbios_boot_state.mem_clock = boot_up_values.ulUClk; + data->vbios_boot_state.soc_clock = boot_up_values.ulSocClk; + if (0 != boot_up_values.usVddc) { + smum_send_msg_to_smc_with_parameter(hwmgr->smumgr, + PPSMC_MSG_SetFloorSocVoltage, + (boot_up_values.usVddc * 4)); + data->vbios_boot_state.bsoc_vddc_lock = true; + } else { + data->vbios_boot_state.bsoc_vddc_lock = false; + } + } + result = vega10_populate_avfs_parameters(hwmgr); PP_ASSERT_WITH_CODE(!result, "Failed to initialize AVFS Parameters!", @@ -2404,35 +2417,9 @@ static int vega10_init_smc_table(struct pp_hwmgr *hwmgr) PP_ASSERT_WITH_CODE(!result, "Failed to upload PPtable!", return result); - if (data->smu_features[GNLD_AVFS].supported) { - uint32_t features_enabled; - result = vega10_get_smc_features(hwmgr->smumgr, &features_enabled); - PP_ASSERT_WITH_CODE(!result, - "Failed to Retrieve Enabled Features!", - return result); - if (!(features_enabled & (1 << FEATURE_AVFS_BIT))) { - result = vega10_perform_btc(hwmgr->smumgr); - PP_ASSERT_WITH_CODE(!result, - "Failed to Perform BTC!", - return result); - result = vega10_avfs_enable(hwmgr, true); - PP_ASSERT_WITH_CODE(!result, - "Attempt to enable AVFS feature Failed!", - return result); - result = vega10_save_vft_table(hwmgr->smumgr, - (uint8_t *)&(data->smc_state_table.avfs_table)); - PP_ASSERT_WITH_CODE(!result, - "Attempt to save VFT table Failed!", + result = vega10_avfs_enable(hwmgr, true); + PP_ASSERT_WITH_CODE(!result, "Attempt to enable AVFS feature Failed!", return result); - } else { - data->smu_features[GNLD_AVFS].enabled = true; - result = vega10_restore_vft_table(hwmgr->smumgr, - (uint8_t *)&(data->smc_state_table.avfs_table)); - PP_ASSERT_WITH_CODE(!result, - "Attempt to restore VFT table Failed!", - return result;); - } - } return 0; } @@ -2457,6 +2444,26 @@ static int vega10_enable_thermal_protection(struct pp_hwmgr *hwmgr) return 0; } +static int vega10_disable_thermal_protection(struct pp_hwmgr *hwmgr) +{ + struct vega10_hwmgr *data = (struct vega10_hwmgr *)(hwmgr->backend); + + if (data->smu_features[GNLD_THERMAL].supported) { + if (!data->smu_features[GNLD_THERMAL].enabled) + pr_info("THERMAL Feature Already disabled!"); + + PP_ASSERT_WITH_CODE( + !vega10_enable_smc_features(hwmgr->smumgr, + false, + data->smu_features[GNLD_THERMAL].smu_feature_bitmap), + "disable THERMAL Feature Failed!", + return -1); + data->smu_features[GNLD_THERMAL].enabled = false; + } + + return 0; +} + static int vega10_enable_vrhot_feature(struct pp_hwmgr *hwmgr) { struct vega10_hwmgr *data = @@ -2535,6 +2542,37 @@ static int vega10_enable_deep_sleep_master_switch(struct pp_hwmgr *hwmgr) return 0; } +static int vega10_stop_dpm(struct pp_hwmgr *hwmgr, uint32_t bitmap) +{ + struct vega10_hwmgr *data = + (struct vega10_hwmgr *)(hwmgr->backend); + uint32_t i, feature_mask = 0; + + + if(data->smu_features[GNLD_LED_DISPLAY].supported == true){ + PP_ASSERT_WITH_CODE(!vega10_enable_smc_features(hwmgr->smumgr, + true, data->smu_features[GNLD_LED_DISPLAY].smu_feature_bitmap), + "Attempt to Enable LED DPM feature Failed!", return -EINVAL); + data->smu_features[GNLD_LED_DISPLAY].enabled = true; + } + + for (i = 0; i < GNLD_DPM_MAX; i++) { + if (data->smu_features[i].smu_feature_bitmap & bitmap) { + if (data->smu_features[i].supported) { + if (data->smu_features[i].enabled) { + feature_mask |= data->smu_features[i]. + smu_feature_bitmap; + data->smu_features[i].enabled = false; + } + } + } + } + + vega10_enable_smc_features(hwmgr->smumgr, false, feature_mask); + + return 0; +} + /** * @brief Tell SMC to enabled the supported DPMs. * @@ -2576,6 +2614,12 @@ static int vega10_start_dpm(struct pp_hwmgr *hwmgr, uint32_t bitmap) data->smu_features[GNLD_LED_DISPLAY].enabled = true; } + if (data->vbios_boot_state.bsoc_vddc_lock) { + smum_send_msg_to_smc_with_parameter(hwmgr->smumgr, + PPSMC_MSG_SetFloorSocVoltage, 0); + data->vbios_boot_state.bsoc_vddc_lock = false; + } + if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_Falcon_QuickTransition)) { if (data->smu_features[GNLD_ACDC].supported) { @@ -2602,8 +2646,6 @@ static int vega10_enable_dpm_tasks(struct pp_hwmgr *hwmgr) "Failed to configure telemetry!", return tmp_result); - vega10_set_tools_address(hwmgr->smumgr); - smum_send_msg_to_smc_with_parameter(hwmgr->smumgr, PPSMC_MSG_NumOfDisplays, 0); @@ -3880,32 +3922,36 @@ static int vega10_dpm_force_dpm_level(struct pp_hwmgr *hwmgr, static int vega10_set_fan_control_mode(struct pp_hwmgr *hwmgr, uint32_t mode) { - if (mode) { - /* stop auto-manage */ - if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, - PHM_PlatformCaps_MicrocodeFanControl)) - vega10_fan_ctrl_stop_smc_fan_control(hwmgr); - vega10_fan_ctrl_set_static_mode(hwmgr, mode); - } else - /* restart auto-manage */ - vega10_fan_ctrl_reset_fan_speed_to_default(hwmgr); + int result = 0; - return 0; + switch (mode) { + case AMD_FAN_CTRL_NONE: + result = vega10_fan_ctrl_set_fan_speed_percent(hwmgr, 100); + break; + case AMD_FAN_CTRL_MANUAL: + if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, + PHM_PlatformCaps_MicrocodeFanControl)) + result = vega10_fan_ctrl_stop_smc_fan_control(hwmgr); + break; + case AMD_FAN_CTRL_AUTO: + result = vega10_fan_ctrl_set_static_mode(hwmgr, mode); + if (!result) + result = vega10_fan_ctrl_start_smc_fan_control(hwmgr); + break; + default: + break; + } + return result; } static int vega10_get_fan_control_mode(struct pp_hwmgr *hwmgr) { - uint32_t reg; + struct vega10_hwmgr *data = (struct vega10_hwmgr *)(hwmgr->backend); - if (hwmgr->fan_ctrl_is_in_default_mode) { - return hwmgr->fan_ctrl_default_mode; - } else { - reg = soc15_get_register_offset(THM_HWID, 0, - mmCG_FDO_CTRL2_BASE_IDX, mmCG_FDO_CTRL2); - return (cgs_read_register(hwmgr->device, reg) & - CG_FDO_CTRL2__FDO_PWM_MODE_MASK) >> - CG_FDO_CTRL2__FDO_PWM_MODE__SHIFT; - } + if (data->smu_features[GNLD_FAN_CONTROL].enabled == false) + return AMD_FAN_CTRL_MANUAL; + else + return AMD_FAN_CTRL_AUTO; } static int vega10_get_dal_power_level(struct pp_hwmgr *hwmgr, @@ -4148,55 +4194,56 @@ static int vega10_force_clock_level(struct pp_hwmgr *hwmgr, switch (type) { case PP_SCLK: - if (data->registry_data.sclk_dpm_key_disabled) - break; - for (i = 0; i < 32; i++) { if (mask & (1 << i)) break; } + data->smc_state_table.gfx_boot_level = i; - PP_ASSERT_WITH_CODE(!smum_send_msg_to_smc_with_parameter( - hwmgr->smumgr, - PPSMC_MSG_SetSoftMinGfxclkByIndex, - i), - "Failed to set soft min sclk index!", - return -1); + for (i = 31; i >= 0; i--) { + if (mask & (1 << i)) + break; + } + data->smc_state_table.gfx_max_level = i; + + PP_ASSERT_WITH_CODE(!vega10_upload_dpm_bootup_level(hwmgr), + "Failed to upload boot level to lowest!", + return -EINVAL); + + PP_ASSERT_WITH_CODE(!vega10_upload_dpm_max_level(hwmgr), + "Failed to upload dpm max level to highest!", + return -EINVAL); break; case PP_MCLK: - if (data->registry_data.mclk_dpm_key_disabled) - break; - for (i = 0; i < 32; i++) { if (mask & (1 << i)) break; } - PP_ASSERT_WITH_CODE(!smum_send_msg_to_smc_with_parameter( - hwmgr->smumgr, - PPSMC_MSG_SetSoftMinUclkByIndex, - i), - "Failed to set soft min mclk index!", - return -1); - break; - - case PP_PCIE: - if (data->registry_data.pcie_dpm_key_disabled) - break; - for (i = 0; i < 32; i++) { if (mask & (1 << i)) break; } + data->smc_state_table.mem_boot_level = i; + + for (i = 31; i >= 0; i--) { + if (mask & (1 << i)) + break; + } + data->smc_state_table.mem_max_level = i; + + PP_ASSERT_WITH_CODE(!vega10_upload_dpm_bootup_level(hwmgr), + "Failed to upload boot level to lowest!", + return -EINVAL); + + PP_ASSERT_WITH_CODE(!vega10_upload_dpm_max_level(hwmgr), + "Failed to upload dpm max level to highest!", + return -EINVAL); - PP_ASSERT_WITH_CODE(!smum_send_msg_to_smc_with_parameter( - hwmgr->smumgr, - PPSMC_MSG_SetMinLinkDpmByIndex, - i), - "Failed to set min pcie index!", - return -1); break; + + case PP_PCIE: default: break; } @@ -4395,11 +4442,55 @@ vega10_check_smc_update_required_for_display_configuration(struct pp_hwmgr *hwmg return is_update_required; } +static int vega10_disable_dpm_tasks(struct pp_hwmgr *hwmgr) +{ + int tmp_result, result = 0; + + tmp_result = (vega10_is_dpm_running(hwmgr)) ? 0 : -1; + PP_ASSERT_WITH_CODE(tmp_result == 0, + "DPM is not running right now, no need to disable DPM!", + return 0); + + if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, + PHM_PlatformCaps_ThermalController)) + vega10_disable_thermal_protection(hwmgr); + + tmp_result = vega10_disable_power_containment(hwmgr); + PP_ASSERT_WITH_CODE((tmp_result == 0), + "Failed to disable power containment!", result = tmp_result); + + tmp_result = vega10_avfs_enable(hwmgr, false); + PP_ASSERT_WITH_CODE((tmp_result == 0), + "Failed to disable AVFS!", result = tmp_result); + + tmp_result = vega10_stop_dpm(hwmgr, SMC_DPM_FEATURES); + PP_ASSERT_WITH_CODE((tmp_result == 0), + "Failed to stop DPM!", result = tmp_result); + + return result; +} + +static int vega10_power_off_asic(struct pp_hwmgr *hwmgr) +{ + struct vega10_hwmgr *data = (struct vega10_hwmgr *)(hwmgr->backend); + int result; + + result = vega10_disable_dpm_tasks(hwmgr); + PP_ASSERT_WITH_CODE((0 == result), + "[disable_dpm_tasks] Failed to disable DPM!", + ); + data->water_marks_bitmap &= ~(WaterMarksLoaded); + + return result; +} + + static const struct pp_hwmgr_func vega10_hwmgr_funcs = { .backend_init = vega10_hwmgr_backend_init, .backend_fini = vega10_hwmgr_backend_fini, .asic_setup = vega10_setup_asic_task, .dynamic_state_management_enable = vega10_enable_dpm_tasks, + .dynamic_state_management_disable = vega10_disable_dpm_tasks, .get_num_of_pp_table_entries = vega10_get_number_of_powerplay_table_entries, .get_power_state_size = vega10_get_power_state_size, @@ -4439,6 +4530,8 @@ static const struct pp_hwmgr_func vega10_hwmgr_funcs = { .check_states_equal = vega10_check_states_equal, .check_smc_update_required_for_display_configuration = vega10_check_smc_update_required_for_display_configuration, + .power_off_asic = vega10_power_off_asic, + .disable_smc_firmware_ctf = vega10_thermal_disable_alert, }; int vega10_hwmgr_init(struct pp_hwmgr *hwmgr) diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.h b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.h index 83c67b9262ff..1912e086c0cf 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.h +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.h @@ -177,8 +177,11 @@ struct vega10_dpmlevel_enable_mask { }; struct vega10_vbios_boot_state { + bool bsoc_vddc_lock; uint16_t vddc; uint16_t vddci; + uint16_t mvddc; + uint16_t vdd_gfx; uint32_t gfx_clock; uint32_t mem_clock; uint32_t soc_clock; diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_powertune.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_powertune.c index f1e244cd2370..3f72268e99bb 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_powertune.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_powertune.c @@ -48,8 +48,8 @@ void vega10_initialize_power_tune_defaults(struct pp_hwmgr *hwmgr) table->Tliquid1Limit = cpu_to_le16(tdp_table->usTemperatureLimitLiquid1); table->Tliquid2Limit = cpu_to_le16(tdp_table->usTemperatureLimitLiquid2); table->TplxLimit = cpu_to_le16(tdp_table->usTemperatureLimitPlx); - table->LoadLineResistance = cpu_to_le16( - hwmgr->platform_descriptor.LoadLineSlope); + table->LoadLineResistance = + hwmgr->platform_descriptor.LoadLineSlope * 256; table->FitLimit = 0; /* Not used for Vega10 */ table->Liquid1_I2C_address = tdp_table->ucLiquid1_I2C_address; @@ -113,6 +113,29 @@ int vega10_enable_power_containment(struct pp_hwmgr *hwmgr) return result; } +int vega10_disable_power_containment(struct pp_hwmgr *hwmgr) +{ + struct vega10_hwmgr *data = + (struct vega10_hwmgr *)(hwmgr->backend); + + if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, + PHM_PlatformCaps_PowerContainment)) { + if (data->smu_features[GNLD_PPT].supported) + PP_ASSERT_WITH_CODE(!vega10_enable_smc_features(hwmgr->smumgr, + false, data->smu_features[GNLD_PPT].smu_feature_bitmap), + "Attempt to disable PPT feature Failed!", + data->smu_features[GNLD_PPT].supported = false); + + if (data->smu_features[GNLD_TDC].supported) + PP_ASSERT_WITH_CODE(!vega10_enable_smc_features(hwmgr->smumgr, + false, data->smu_features[GNLD_TDC].smu_feature_bitmap), + "Attempt to disable PPT feature Failed!", + data->smu_features[GNLD_TDC].supported = false); + } + + return 0; +} + static int vega10_set_overdrive_target_percentage(struct pp_hwmgr *hwmgr, uint32_t adjust_percent) { diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_powertune.h b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_powertune.h index d9662bf4a4b4..9ecaa27c0bb5 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_powertune.h +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_powertune.h @@ -60,6 +60,7 @@ int vega10_enable_smc_cac(struct pp_hwmgr *hwmgr); int vega10_enable_power_containment(struct pp_hwmgr *hwmgr); int vega10_set_power_limit(struct pp_hwmgr *hwmgr, uint32_t n); int vega10_power_control_set_level(struct pp_hwmgr *hwmgr); +int vega10_disable_power_containment(struct pp_hwmgr *hwmgr); #endif /* _VEGA10_POWERTUNE_H_ */ diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_processpptables.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_processpptables.c index 8b55ae01132d..00e95511e19a 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_processpptables.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_processpptables.c @@ -407,7 +407,7 @@ static int get_tdp_table( tdp_table->ucPlx_I2C_address = power_tune_table->ucPlx_I2C_address; tdp_table->ucPlx_I2C_Line = power_tune_table->ucPlx_I2C_LineSCL; tdp_table->ucPlx_I2C_LineSDA = power_tune_table->ucPlx_I2C_LineSDA; - hwmgr->platform_descriptor.LoadLineSlope = power_tune_table->usLoadLineResistance; + hwmgr->platform_descriptor.LoadLineSlope = le16_to_cpu(power_tune_table->usLoadLineResistance); } else { power_tune_table_v2 = (ATOM_Vega10_PowerTune_Table_V2 *)table; tdp_table->usMaximumPowerDeliveryLimit = le16_to_cpu(power_tune_table_v2->usSocketPowerLimit); @@ -453,7 +453,7 @@ static int get_tdp_table( tdp_table->ucPlx_I2C_LineSDA = sda; hwmgr->platform_descriptor.LoadLineSlope = - power_tune_table_v2->usLoadLineResistance; + le16_to_cpu(power_tune_table_v2->usLoadLineResistance); } *info_tdp_table = tdp_table; diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_thermal.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_thermal.c index f4d77b62e1ba..d5f53d04fa08 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_thermal.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_thermal.c @@ -381,14 +381,10 @@ int vega10_thermal_get_temperature(struct pp_hwmgr *hwmgr) temp = cgs_read_register(hwmgr->device, reg); - temp = (temp & CG_MULT_THERMAL_STATUS__CTF_TEMP_MASK) >> - CG_MULT_THERMAL_STATUS__CTF_TEMP__SHIFT; + temp = (temp & CG_MULT_THERMAL_STATUS__ASIC_MAX_TEMP_MASK) >> + CG_MULT_THERMAL_STATUS__ASIC_MAX_TEMP__SHIFT; - /* Bit 9 means the reading is lower than the lowest usable value. */ - if (temp & 0x200) - temp = VEGA10_THERMAL_MAXIMUM_TEMP_READING; - else - temp = temp & 0x1ff; + temp = temp & 0x1ff; temp *= PP_TEMPERATURE_UNITS_PER_CENTIGRADES; @@ -424,23 +420,28 @@ static int vega10_thermal_set_temperature_range(struct pp_hwmgr *hwmgr, mmTHM_THERMAL_INT_CTRL_BASE_IDX, mmTHM_THERMAL_INT_CTRL); val = cgs_read_register(hwmgr->device, reg); - val &= ~(THM_THERMAL_INT_CTRL__DIG_THERM_INTH_MASK); - val |= (high / PP_TEMPERATURE_UNITS_PER_CENTIGRADES) << - THM_THERMAL_INT_CTRL__DIG_THERM_INTH__SHIFT; - val &= ~(THM_THERMAL_INT_CTRL__DIG_THERM_INTL_MASK); - val |= (low / PP_TEMPERATURE_UNITS_PER_CENTIGRADES) << - THM_THERMAL_INT_CTRL__DIG_THERM_INTL__SHIFT; + + val &= (~THM_THERMAL_INT_CTRL__MAX_IH_CREDIT_MASK); + val |= (5 << THM_THERMAL_INT_CTRL__MAX_IH_CREDIT__SHIFT); + + val &= (~THM_THERMAL_INT_CTRL__THERM_IH_HW_ENA_MASK); + val |= (1 << THM_THERMAL_INT_CTRL__THERM_IH_HW_ENA__SHIFT); + + val &= (~THM_THERMAL_INT_CTRL__DIG_THERM_INTH_MASK); + val |= ((high / PP_TEMPERATURE_UNITS_PER_CENTIGRADES) + << THM_THERMAL_INT_CTRL__DIG_THERM_INTH__SHIFT); + + val &= (~THM_THERMAL_INT_CTRL__DIG_THERM_INTL_MASK); + val |= ((low / PP_TEMPERATURE_UNITS_PER_CENTIGRADES) + << THM_THERMAL_INT_CTRL__DIG_THERM_INTL__SHIFT); + + val = val & (~THM_THERMAL_INT_CTRL__THERM_TRIGGER_MASK_MASK); + cgs_write_register(hwmgr->device, reg, val); reg = soc15_get_register_offset(THM_HWID, 0, mmTHM_TCON_HTC_BASE_IDX, mmTHM_TCON_HTC); - val = cgs_read_register(hwmgr->device, reg); - val &= ~(THM_TCON_HTC__HTC_TMP_LMT_MASK); - val |= (high / PP_TEMPERATURE_UNITS_PER_CENTIGRADES) << - THM_TCON_HTC__HTC_TMP_LMT__SHIFT; - cgs_write_register(hwmgr->device, reg, val); - return 0; } @@ -482,18 +483,28 @@ static int vega10_thermal_initialize(struct pp_hwmgr *hwmgr) static int vega10_thermal_enable_alert(struct pp_hwmgr *hwmgr) { struct vega10_hwmgr *data = (struct vega10_hwmgr *)(hwmgr->backend); + uint32_t val = 0; + uint32_t reg; if (data->smu_features[GNLD_FW_CTF].supported) { if (data->smu_features[GNLD_FW_CTF].enabled) printk("[Thermal_EnableAlert] FW CTF Already Enabled!\n"); + + PP_ASSERT_WITH_CODE(!vega10_enable_smc_features(hwmgr->smumgr, + true, + data->smu_features[GNLD_FW_CTF].smu_feature_bitmap), + "Attempt to Enable FW CTF feature Failed!", + return -1); + data->smu_features[GNLD_FW_CTF].enabled = true; } - PP_ASSERT_WITH_CODE(!vega10_enable_smc_features(hwmgr->smumgr, - true, - data->smu_features[GNLD_FW_CTF].smu_feature_bitmap), - "Attempt to Enable FW CTF feature Failed!", - return -1); - data->smu_features[GNLD_FW_CTF].enabled = true; + val |= (1 << THM_THERMAL_INT_ENA__THERM_INTH_CLR__SHIFT); + val |= (1 << THM_THERMAL_INT_ENA__THERM_INTL_CLR__SHIFT); + val |= (1 << THM_THERMAL_INT_ENA__THERM_TRIGGER_CLR__SHIFT); + + reg = soc15_get_register_offset(THM_HWID, 0, mmTHM_THERMAL_INT_ENA_BASE_IDX, mmTHM_THERMAL_INT_ENA); + cgs_write_register(hwmgr->device, reg, val); + return 0; } @@ -501,21 +512,27 @@ static int vega10_thermal_enable_alert(struct pp_hwmgr *hwmgr) * Disable thermal alerts on the RV770 thermal controller. * @param hwmgr The address of the hardware manager. */ -static int vega10_thermal_disable_alert(struct pp_hwmgr *hwmgr) +int vega10_thermal_disable_alert(struct pp_hwmgr *hwmgr) { struct vega10_hwmgr *data = (struct vega10_hwmgr *)(hwmgr->backend); + uint32_t reg; if (data->smu_features[GNLD_FW_CTF].supported) { if (!data->smu_features[GNLD_FW_CTF].enabled) printk("[Thermal_EnableAlert] FW CTF Already disabled!\n"); - } - PP_ASSERT_WITH_CODE(!vega10_enable_smc_features(hwmgr->smumgr, + + PP_ASSERT_WITH_CODE(!vega10_enable_smc_features(hwmgr->smumgr, false, data->smu_features[GNLD_FW_CTF].smu_feature_bitmap), "Attempt to disable FW CTF feature Failed!", return -1); - data->smu_features[GNLD_FW_CTF].enabled = false; + data->smu_features[GNLD_FW_CTF].enabled = false; + } + + reg = soc15_get_register_offset(THM_HWID, 0, mmTHM_THERMAL_INT_ENA_BASE_IDX, mmTHM_THERMAL_INT_ENA); + cgs_write_register(hwmgr->device, reg, 0); + return 0; } @@ -561,6 +578,11 @@ int tf_vega10_thermal_setup_fan_table(struct pp_hwmgr *hwmgr, advanceFanControlParameters.ulMinFanSCLKAcousticLimit); table->FanTargetTemperature = hwmgr->thermal_controller. advanceFanControlParameters.usTMax; + + smum_send_msg_to_smc_with_parameter(hwmgr->smumgr, + PPSMC_MSG_SetFanTemperatureTarget, + (uint32_t)table->FanTargetTemperature); + table->FanPwmMin = hwmgr->thermal_controller. advanceFanControlParameters.usPWMMin * 255 / 100; table->FanTargetGfxclk = (uint16_t)(hwmgr->thermal_controller. diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_thermal.h b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_thermal.h index 8036808ec421..776f3a2effc0 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_thermal.h +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_thermal.h @@ -78,6 +78,8 @@ extern int vega10_fan_ctrl_get_fan_speed_rpm(struct pp_hwmgr *hwmgr, uint32_t *speed); extern int vega10_fan_ctrl_stop_smc_fan_control(struct pp_hwmgr *hwmgr); extern uint32_t smu7_get_xclk(struct pp_hwmgr *hwmgr); +extern int vega10_thermal_disable_alert(struct pp_hwmgr *hwmgr); +int vega10_fan_ctrl_start_smc_fan_control(struct pp_hwmgr *hwmgr); #endif diff --git a/drivers/gpu/drm/amd/powerplay/inc/hardwaremanager.h b/drivers/gpu/drm/amd/powerplay/inc/hardwaremanager.h index 5345b50761f4..a1ebe1014492 100644 --- a/drivers/gpu/drm/amd/powerplay/inc/hardwaremanager.h +++ b/drivers/gpu/drm/amd/powerplay/inc/hardwaremanager.h @@ -431,6 +431,6 @@ extern int phm_display_clock_voltage_request(struct pp_hwmgr *hwmgr, struct pp_display_clock_request *clock); extern int phm_get_max_high_clocks(struct pp_hwmgr *hwmgr, struct amd_pp_simple_clock_info *clocks); - +extern int phm_disable_smc_firmware_ctf(struct pp_hwmgr *hwmgr); #endif /* _HARDWARE_MANAGER_H_ */ diff --git a/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h b/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h index 320225dd3328..805b9df452a3 100644 --- a/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h +++ b/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h @@ -368,11 +368,10 @@ struct pp_hwmgr_func { int (*get_mclk_od)(struct pp_hwmgr *hwmgr); int (*set_mclk_od)(struct pp_hwmgr *hwmgr, uint32_t value); int (*read_sensor)(struct pp_hwmgr *hwmgr, int idx, void *value, int *size); - int (*request_firmware)(struct pp_hwmgr *hwmgr); - int (*release_firmware)(struct pp_hwmgr *hwmgr); int (*set_power_profile_state)(struct pp_hwmgr *hwmgr, struct amd_pp_profile *request); int (*avfs_control)(struct pp_hwmgr *hwmgr, bool enable); + int (*disable_smc_firmware_ctf)(struct pp_hwmgr *hwmgr); }; struct pp_table_func { @@ -765,6 +764,7 @@ struct pp_hwmgr { struct pp_thermal_controller_info thermal_controller; bool fan_ctrl_is_in_default_mode; uint32_t fan_ctrl_default_mode; + bool fan_ctrl_enabled; uint32_t tmin; struct phm_microcode_version_info microcode_version_info; uint32_t ps_size; diff --git a/drivers/gpu/drm/amd/powerplay/inc/smu9_driver_if.h b/drivers/gpu/drm/amd/powerplay/inc/smu9_driver_if.h index 2037910adcb1..d43f98a910b0 100644 --- a/drivers/gpu/drm/amd/powerplay/inc/smu9_driver_if.h +++ b/drivers/gpu/drm/amd/powerplay/inc/smu9_driver_if.h @@ -30,7 +30,7 @@ * SMU TEAM: Always increment the interface version if * any structure is changed in this file */ -#define SMU9_DRIVER_IF_VERSION 0xB +#define SMU9_DRIVER_IF_VERSION 0xD #define PPTABLE_V10_SMU_VERSION 1 @@ -302,7 +302,17 @@ typedef struct { uint32_t DpmLevelPowerDelta; - uint32_t Reserved[19]; + uint8_t EnableBoostState; + uint8_t AConstant_Shift; + uint8_t DC_tol_sigma_Shift; + uint8_t PSM_Age_CompFactor_Shift; + + uint16_t BoostStartTemperature; + uint16_t BoostStopTemperature; + + PllSetting_t GfxBoostState; + + uint32_t Reserved[14]; /* Padding - ignore */ uint32_t MmHubPadding[7]; /* SMU internal use */ @@ -464,4 +474,8 @@ typedef struct { #define DB_PCC_SHIFT 26 #define DB_EDC_SHIFT 27 +#define REMOVE_FMAX_MARGIN_BIT 0x0 +#define REMOVE_DCTOL_MARGIN_BIT 0x1 +#define REMOVE_PLATFORM_MARGIN_BIT 0x2 + #endif diff --git a/drivers/gpu/drm/amd/powerplay/inc/vega10_ppsmc.h b/drivers/gpu/drm/amd/powerplay/inc/vega10_ppsmc.h index 90beef35bba2..254974d3d371 100644 --- a/drivers/gpu/drm/amd/powerplay/inc/vega10_ppsmc.h +++ b/drivers/gpu/drm/amd/powerplay/inc/vega10_ppsmc.h @@ -122,7 +122,10 @@ typedef uint16_t PPSMC_Result; #define PPSMC_MSG_SetFanMinPwm 0x52 #define PPSMC_MSG_ConfigureGfxDidt 0x55 #define PPSMC_MSG_NumOfDisplays 0x56 -#define PPSMC_Message_Count 0x57 +#define PPSMC_MSG_ReadSerialNumTop32 0x58 +#define PPSMC_MSG_ReadSerialNumBottom32 0x59 +#define PPSMC_Message_Count 0x5A + typedef int PPSMC_Msg; diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/cz_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/cz_smumgr.c index 1f6744a443d4..39c7091866e8 100644 --- a/drivers/gpu/drm/amd/powerplay/smumgr/cz_smumgr.c +++ b/drivers/gpu/drm/amd/powerplay/smumgr/cz_smumgr.c @@ -20,11 +20,13 @@ * OTHER DEALINGS IN THE SOFTWARE. * */ -#include <linux/types.h> + +#include <linux/delay.h> +#include <linux/gfp.h> #include <linux/kernel.h> #include <linux/slab.h> -#include <linux/gfp.h> -#include "linux/delay.h" +#include <linux/types.h> + #include "cgs_common.h" #include "smu/smu_8_0_d.h" #include "smu/smu_8_0_sh_mask.h" diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/smumgr.c index c0d75766bbc8..2e954a44bac1 100644 --- a/drivers/gpu/drm/amd/powerplay/smumgr/smumgr.c +++ b/drivers/gpu/drm/amd/powerplay/smumgr/smumgr.c @@ -20,15 +20,16 @@ * OTHER DEALINGS IN THE SOFTWARE. * */ -#include <linux/types.h> + +#include <linux/delay.h> #include <linux/kernel.h> #include <linux/module.h> #include <linux/slab.h> +#include <linux/types.h> #include <drm/amdgpu_drm.h> #include "pp_instance.h" #include "smumgr.h" #include "cgs_common.h" -#include "linux/delay.h" MODULE_FIRMWARE("amdgpu/topaz_smc.bin"); MODULE_FIRMWARE("amdgpu/topaz_k_smc.bin"); diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/vega10_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/vega10_smumgr.c index 2685f02ab551..115f0e4b1603 100644 --- a/drivers/gpu/drm/amd/powerplay/smumgr/vega10_smumgr.c +++ b/drivers/gpu/drm/amd/powerplay/smumgr/vega10_smumgr.c @@ -74,18 +74,18 @@ static bool vega10_is_smc_ram_running(struct pp_smumgr *smumgr) return false; } -/** -* Check if SMC has responded to previous message. -* -* @param smumgr the address of the powerplay hardware manager. -* @return TRUE SMC has responded, FALSE otherwise. -*/ +/* + * Check if SMC has responded to previous message. + * + * @param smumgr the address of the powerplay hardware manager. + * @return TRUE SMC has responded, FALSE otherwise. + */ static uint32_t vega10_wait_for_response(struct pp_smumgr *smumgr) { uint32_t reg; if (!vega10_is_smc_ram_running(smumgr)) - return -1; + return -EINVAL; reg = soc15_get_register_offset(MP1_HWID, 0, mmMP1_SMN_C2PMSG_90_BASE_IDX, mmMP1_SMN_C2PMSG_90); @@ -96,20 +96,19 @@ static uint32_t vega10_wait_for_response(struct pp_smumgr *smumgr) return cgs_read_register(smumgr->device, reg); } -/** -* Send a message to the SMC, and do not wait for its response. -* -* @param smumgr the address of the powerplay hardware manager. -* @param msg the message to send. -* @return Always return 0. -*/ +/* + * Send a message to the SMC, and do not wait for its response. + * @param smumgr the address of the powerplay hardware manager. + * @param msg the message to send. + * @return Always return 0. + */ int vega10_send_msg_to_smc_without_waiting(struct pp_smumgr *smumgr, uint16_t msg) { uint32_t reg; if (!vega10_is_smc_ram_running(smumgr)) - return -1; + return -EINVAL; reg = soc15_get_register_offset(MP1_HWID, 0, mmMP1_SMN_C2PMSG_66_BASE_IDX, mmMP1_SMN_C2PMSG_66); @@ -118,19 +117,18 @@ int vega10_send_msg_to_smc_without_waiting(struct pp_smumgr *smumgr, return 0; } -/** -* Send a message to the SMC, and wait for its response. -* -* @param smumgr the address of the powerplay hardware manager. -* @param msg the message to send. -* @return The response that came from the SMC. -*/ +/* + * Send a message to the SMC, and wait for its response. + * @param smumgr the address of the powerplay hardware manager. + * @param msg the message to send. + * @return Always return 0. + */ int vega10_send_msg_to_smc(struct pp_smumgr *smumgr, uint16_t msg) { uint32_t reg; if (!vega10_is_smc_ram_running(smumgr)) - return -1; + return -EINVAL; vega10_wait_for_response(smumgr); @@ -140,19 +138,18 @@ int vega10_send_msg_to_smc(struct pp_smumgr *smumgr, uint16_t msg) vega10_send_msg_to_smc_without_waiting(smumgr, msg); - PP_ASSERT_WITH_CODE(vega10_wait_for_response(smumgr) == 1, - "Failed to send Message.", - return -1); + if (vega10_wait_for_response(smumgr) != 1) + pr_err("Failed to send message: 0x%x\n", msg); return 0; } -/** +/* * Send a message to the SMC with parameter * @param smumgr: the address of the powerplay hardware manager. * @param msg: the message to send. * @param parameter: the parameter to send - * @return The response that came from the SMC. + * @return Always return 0. */ int vega10_send_msg_to_smc_with_parameter(struct pp_smumgr *smumgr, uint16_t msg, uint32_t parameter) @@ -160,7 +157,7 @@ int vega10_send_msg_to_smc_with_parameter(struct pp_smumgr *smumgr, uint32_t reg; if (!vega10_is_smc_ram_running(smumgr)) - return -1; + return -EINVAL; vega10_wait_for_response(smumgr); @@ -174,22 +171,20 @@ int vega10_send_msg_to_smc_with_parameter(struct pp_smumgr *smumgr, vega10_send_msg_to_smc_without_waiting(smumgr, msg); - PP_ASSERT_WITH_CODE(vega10_wait_for_response(smumgr) == 1, - "Failed to send Message.", - return -1); + if (vega10_wait_for_response(smumgr) != 1) + pr_err("Failed to send message: 0x%x\n", msg); return 0; } -/** -* Send a message to the SMC with parameter, do not wait for response -* -* @param smumgr: the address of the powerplay hardware manager. -* @param msg: the message to send. -* @param parameter: the parameter to send -* @return The response that came from the SMC. -*/ +/* + * Send a message to the SMC with parameter, do not wait for response + * @param smumgr: the address of the powerplay hardware manager. + * @param msg: the message to send. + * @param parameter: the parameter to send + * @return The response that came from the SMC. + */ int vega10_send_msg_to_smc_with_parameter_without_waiting( struct pp_smumgr *smumgr, uint16_t msg, uint32_t parameter) { @@ -202,13 +197,12 @@ int vega10_send_msg_to_smc_with_parameter_without_waiting( return vega10_send_msg_to_smc_without_waiting(smumgr, msg); } -/** -* Retrieve an argument from SMC. -* -* @param smumgr the address of the powerplay hardware manager. -* @param arg pointer to store the argument from SMC. -* @return Always return 0. -*/ +/* + * Retrieve an argument from SMC. + * @param smumgr the address of the powerplay hardware manager. + * @param arg pointer to store the argument from SMC. + * @return Always return 0. + */ int vega10_read_arg_from_smc(struct pp_smumgr *smumgr, uint32_t *arg) { uint32_t reg; @@ -221,11 +215,11 @@ int vega10_read_arg_from_smc(struct pp_smumgr *smumgr, uint32_t *arg) return 0; } -/** -* Copy table from SMC into driver FB -* @param smumgr the address of the SMC manager -* @param table_id the driver's table ID to copy from -*/ +/* + * Copy table from SMC into driver FB + * @param smumgr the address of the SMC manager + * @param table_id the driver's table ID to copy from + */ int vega10_copy_table_from_smc(struct pp_smumgr *smumgr, uint8_t *table, int16_t table_id) { @@ -233,25 +227,25 @@ int vega10_copy_table_from_smc(struct pp_smumgr *smumgr, (struct vega10_smumgr *)(smumgr->backend); PP_ASSERT_WITH_CODE(table_id < MAX_SMU_TABLE, - "Invalid SMU Table ID!", return -1;); + "Invalid SMU Table ID!", return -EINVAL); PP_ASSERT_WITH_CODE(priv->smu_tables.entry[table_id].version != 0, - "Invalid SMU Table version!", return -1;); + "Invalid SMU Table version!", return -EINVAL); PP_ASSERT_WITH_CODE(priv->smu_tables.entry[table_id].size != 0, - "Invalid SMU Table Length!", return -1;); + "Invalid SMU Table Length!", return -EINVAL); PP_ASSERT_WITH_CODE(vega10_send_msg_to_smc_with_parameter(smumgr, PPSMC_MSG_SetDriverDramAddrHigh, priv->smu_tables.entry[table_id].table_addr_high) == 0, - "[CopyTableFromSMC] Attempt to Set Dram Addr High Failed!", return -1;); + "[CopyTableFromSMC] Attempt to Set Dram Addr High Failed!", return -EINVAL); PP_ASSERT_WITH_CODE(vega10_send_msg_to_smc_with_parameter(smumgr, PPSMC_MSG_SetDriverDramAddrLow, priv->smu_tables.entry[table_id].table_addr_low) == 0, "[CopyTableFromSMC] Attempt to Set Dram Addr Low Failed!", - return -1;); + return -EINVAL); PP_ASSERT_WITH_CODE(vega10_send_msg_to_smc_with_parameter(smumgr, PPSMC_MSG_TransferTableSmu2Dram, priv->smu_tables.entry[table_id].table_id) == 0, "[CopyTableFromSMC] Attempt to Transfer Table From SMU Failed!", - return -1;); + return -EINVAL); memcpy(table, priv->smu_tables.entry[table_id].table, priv->smu_tables.entry[table_id].size); @@ -259,11 +253,11 @@ int vega10_copy_table_from_smc(struct pp_smumgr *smumgr, return 0; } -/** -* Copy table from Driver FB into SMC -* @param smumgr the address of the SMC manager -* @param table_id the table to copy from -*/ +/* + * Copy table from Driver FB into SMC + * @param smumgr the address of the SMC manager + * @param table_id the table to copy from + */ int vega10_copy_table_to_smc(struct pp_smumgr *smumgr, uint8_t *table, int16_t table_id) { @@ -271,11 +265,11 @@ int vega10_copy_table_to_smc(struct pp_smumgr *smumgr, (struct vega10_smumgr *)(smumgr->backend); PP_ASSERT_WITH_CODE(table_id < MAX_SMU_TABLE, - "Invalid SMU Table ID!", return -1;); + "Invalid SMU Table ID!", return -EINVAL); PP_ASSERT_WITH_CODE(priv->smu_tables.entry[table_id].version != 0, - "Invalid SMU Table version!", return -1;); + "Invalid SMU Table version!", return -EINVAL); PP_ASSERT_WITH_CODE(priv->smu_tables.entry[table_id].size != 0, - "Invalid SMU Table Length!", return -1;); + "Invalid SMU Table Length!", return -EINVAL); memcpy(priv->smu_tables.entry[table_id].table, table, priv->smu_tables.entry[table_id].size); @@ -284,27 +278,18 @@ int vega10_copy_table_to_smc(struct pp_smumgr *smumgr, PPSMC_MSG_SetDriverDramAddrHigh, priv->smu_tables.entry[table_id].table_addr_high) == 0, "[CopyTableToSMC] Attempt to Set Dram Addr High Failed!", - return -1;); + return -EINVAL;); PP_ASSERT_WITH_CODE(vega10_send_msg_to_smc_with_parameter(smumgr, PPSMC_MSG_SetDriverDramAddrLow, priv->smu_tables.entry[table_id].table_addr_low) == 0, "[CopyTableToSMC] Attempt to Set Dram Addr Low Failed!", - return -1;); + return -EINVAL); PP_ASSERT_WITH_CODE(vega10_send_msg_to_smc_with_parameter(smumgr, PPSMC_MSG_TransferTableDram2Smu, priv->smu_tables.entry[table_id].table_id) == 0, "[CopyTableToSMC] Attempt to Transfer Table To SMU Failed!", - return -1;); - - return 0; -} + return -EINVAL); -int vega10_perform_btc(struct pp_smumgr *smumgr) -{ - PP_ASSERT_WITH_CODE(!vega10_send_msg_to_smc_with_parameter( - smumgr, PPSMC_MSG_RunBtc, 0), - "Attempt to run DC BTC Failed!", - return -1); return 0; } @@ -312,7 +297,7 @@ int vega10_save_vft_table(struct pp_smumgr *smumgr, uint8_t *avfs_table) { PP_ASSERT_WITH_CODE(avfs_table, "No access to SMC AVFS Table", - return -1); + return -EINVAL); return vega10_copy_table_from_smc(smumgr, avfs_table, AVFSTABLE); } @@ -321,7 +306,7 @@ int vega10_restore_vft_table(struct pp_smumgr *smumgr, uint8_t *avfs_table) { PP_ASSERT_WITH_CODE(avfs_table, "No access to SMC AVFS Table", - return -1); + return -EINVAL); return vega10_copy_table_to_smc(smumgr, avfs_table, AVFSTABLE); } @@ -339,13 +324,16 @@ int vega10_enable_smc_features(struct pp_smumgr *smumgr, int vega10_get_smc_features(struct pp_smumgr *smumgr, uint32_t *features_enabled) { + if (features_enabled == NULL) + return -EINVAL; + if (!vega10_send_msg_to_smc(smumgr, PPSMC_MSG_GetEnabledSmuFeatures)) { - if (!vega10_read_arg_from_smc(smumgr, features_enabled)) - return 0; + vega10_read_arg_from_smc(smumgr, features_enabled); + return 0; } - return -1; + return -EINVAL; } int vega10_set_tools_address(struct pp_smumgr *smumgr) @@ -372,25 +360,20 @@ static int vega10_verify_smc_interface(struct pp_smumgr *smumgr) PP_ASSERT_WITH_CODE(!vega10_send_msg_to_smc(smumgr, PPSMC_MSG_GetDriverIfVersion), "Attempt to get SMC IF Version Number Failed!", - return -1); - PP_ASSERT_WITH_CODE(!vega10_read_arg_from_smc(smumgr, - &smc_driver_if_version), - "Attempt to read SMC IF Version Number Failed!", - return -1); - - if (smc_driver_if_version != SMU9_DRIVER_IF_VERSION) - return -1; + return -EINVAL); + vega10_read_arg_from_smc(smumgr, &smc_driver_if_version); + + if (smc_driver_if_version != SMU9_DRIVER_IF_VERSION) { + pr_err("Your firmware(0x%x) doesn't match \ + SMU9_DRIVER_IF_VERSION(0x%x). \ + Please update your firmware!\n", + smc_driver_if_version, SMU9_DRIVER_IF_VERSION); + return -EINVAL; + } return 0; } -/** -* Write a 32bit value to the SMC SRAM space. -* ALL PARAMETERS ARE IN HOST BYTE ORDER. -* @param smumgr the address of the powerplay hardware manager. -* @param smc_addr the address in the SMC RAM to access. -* @param value to write to the SMC SRAM. -*/ static int vega10_smu_init(struct pp_smumgr *smumgr) { struct vega10_smumgr *priv; @@ -427,7 +410,7 @@ static int vega10_smu_init(struct pp_smumgr *smumgr) kfree(smumgr->backend); cgs_free_gpu_mem(smumgr->device, (cgs_handle_t)handle); - return -1); + return -EINVAL); priv->smu_tables.entry[PPTABLE].version = 0x01; priv->smu_tables.entry[PPTABLE].size = sizeof(PPTable_t); @@ -455,7 +438,7 @@ static int vega10_smu_init(struct pp_smumgr *smumgr) (cgs_handle_t)priv->smu_tables.entry[PPTABLE].handle); cgs_free_gpu_mem(smumgr->device, (cgs_handle_t)handle); - return -1); + return -EINVAL); priv->smu_tables.entry[WMTABLE].version = 0x01; priv->smu_tables.entry[WMTABLE].size = sizeof(Watermarks_t); @@ -485,7 +468,7 @@ static int vega10_smu_init(struct pp_smumgr *smumgr) (cgs_handle_t)priv->smu_tables.entry[WMTABLE].handle); cgs_free_gpu_mem(smumgr->device, (cgs_handle_t)handle); - return -1); + return -EINVAL); priv->smu_tables.entry[AVFSTABLE].version = 0x01; priv->smu_tables.entry[AVFSTABLE].size = sizeof(AvfsTable_t); @@ -497,7 +480,7 @@ static int vega10_smu_init(struct pp_smumgr *smumgr) priv->smu_tables.entry[AVFSTABLE].table = kaddr; priv->smu_tables.entry[AVFSTABLE].handle = handle; - tools_size = 0; + tools_size = 0x19000; if (tools_size) { smu_allocate_memory(smumgr->device, tools_size, @@ -517,9 +500,44 @@ static int vega10_smu_init(struct pp_smumgr *smumgr) smu_lower_32_bits(mc_addr); priv->smu_tables.entry[TOOLSTABLE].table = kaddr; priv->smu_tables.entry[TOOLSTABLE].handle = handle; + vega10_set_tools_address(smumgr); } } + /* allocate space for AVFS Fuse table */ + smu_allocate_memory(smumgr->device, + sizeof(AvfsFuseOverride_t), + CGS_GPU_MEM_TYPE__VISIBLE_CONTIG_FB, + PAGE_SIZE, + &mc_addr, + &kaddr, + &handle); + + PP_ASSERT_WITH_CODE(kaddr, + "[vega10_smu_init] Out of memory for avfs fuse table.", + kfree(smumgr->backend); + cgs_free_gpu_mem(smumgr->device, + (cgs_handle_t)priv->smu_tables.entry[PPTABLE].handle); + cgs_free_gpu_mem(smumgr->device, + (cgs_handle_t)priv->smu_tables.entry[WMTABLE].handle); + cgs_free_gpu_mem(smumgr->device, + (cgs_handle_t)priv->smu_tables.entry[AVFSTABLE].handle); + cgs_free_gpu_mem(smumgr->device, + (cgs_handle_t)priv->smu_tables.entry[TOOLSTABLE].handle); + cgs_free_gpu_mem(smumgr->device, + (cgs_handle_t)handle); + return -EINVAL); + + priv->smu_tables.entry[AVFSFUSETABLE].version = 0x01; + priv->smu_tables.entry[AVFSFUSETABLE].size = sizeof(AvfsFuseOverride_t); + priv->smu_tables.entry[AVFSFUSETABLE].table_id = TABLE_AVFS_FUSE_OVERRIDE; + priv->smu_tables.entry[AVFSFUSETABLE].table_addr_high = + smu_upper_32_bits(mc_addr); + priv->smu_tables.entry[AVFSFUSETABLE].table_addr_low = + smu_lower_32_bits(mc_addr); + priv->smu_tables.entry[AVFSFUSETABLE].table = kaddr; + priv->smu_tables.entry[AVFSFUSETABLE].handle = handle; + return 0; } @@ -538,6 +556,8 @@ static int vega10_smu_fini(struct pp_smumgr *smumgr) if (priv->smu_tables.entry[TOOLSTABLE].table) cgs_free_gpu_mem(smumgr->device, (cgs_handle_t)priv->smu_tables.entry[TOOLSTABLE].handle); + cgs_free_gpu_mem(smumgr->device, + (cgs_handle_t)priv->smu_tables.entry[AVFSFUSETABLE].handle); kfree(smumgr->backend); smumgr->backend = NULL; } @@ -548,7 +568,7 @@ static int vega10_start_smu(struct pp_smumgr *smumgr) { PP_ASSERT_WITH_CODE(!vega10_verify_smc_interface(smumgr), "Failed to verify SMC interface!", - return -1); + return -EINVAL); return 0; } diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/vega10_smumgr.h b/drivers/gpu/drm/amd/powerplay/smumgr/vega10_smumgr.h index ad050212426d..821425c1e4e0 100644 --- a/drivers/gpu/drm/amd/powerplay/smumgr/vega10_smumgr.h +++ b/drivers/gpu/drm/amd/powerplay/smumgr/vega10_smumgr.h @@ -30,6 +30,7 @@ enum smu_table_id { WMTABLE, AVFSTABLE, TOOLSTABLE, + AVFSFUSETABLE, MAX_SMU_TABLE, }; @@ -62,7 +63,6 @@ int vega10_get_smc_features(struct pp_smumgr *smumgr, uint32_t *features_enabled); int vega10_save_vft_table(struct pp_smumgr *smumgr, uint8_t *avfs_table); int vega10_restore_vft_table(struct pp_smumgr *smumgr, uint8_t *avfs_table); -int vega10_perform_btc(struct pp_smumgr *smumgr); int vega10_set_tools_address(struct pp_smumgr *smumgr); diff --git a/drivers/gpu/drm/amd/scheduler/gpu_scheduler.c b/drivers/gpu/drm/amd/scheduler/gpu_scheduler.c index acd882a188bc..fea96a765cf1 100644 --- a/drivers/gpu/drm/amd/scheduler/gpu_scheduler.c +++ b/drivers/gpu/drm/amd/scheduler/gpu_scheduler.c @@ -236,6 +236,23 @@ static void amd_sched_entity_clear_dep(struct dma_fence *f, struct dma_fence_cb dma_fence_put(f); } +bool amd_sched_dependency_optimized(struct dma_fence* fence, + struct amd_sched_entity *entity) +{ + struct amd_gpu_scheduler *sched = entity->sched; + struct amd_sched_fence *s_fence; + + if (!fence || dma_fence_is_signaled(fence)) + return false; + if (fence->context == entity->fence_context) + return true; + s_fence = to_amd_sched_fence(fence); + if (s_fence && s_fence->sched == sched) + return true; + + return false; +} + static bool amd_sched_entity_add_dependency_cb(struct amd_sched_entity *entity) { struct amd_gpu_scheduler *sched = entity->sched; @@ -387,7 +404,9 @@ void amd_sched_hw_job_reset(struct amd_gpu_scheduler *sched) spin_lock(&sched->job_list_lock); list_for_each_entry_reverse(s_job, &sched->ring_mirror_list, node) { - if (dma_fence_remove_callback(s_job->s_fence->parent, &s_job->s_fence->cb)) { + if (s_job->s_fence->parent && + dma_fence_remove_callback(s_job->s_fence->parent, + &s_job->s_fence->cb)) { dma_fence_put(s_job->s_fence->parent); s_job->s_fence->parent = NULL; } @@ -460,9 +479,9 @@ int amd_sched_job_init(struct amd_sched_job *job, job->sched = sched; job->s_entity = entity; job->s_fence = amd_sched_fence_create(entity, owner); - job->id = atomic64_inc_return(&sched->job_id_count); if (!job->s_fence) return -ENOMEM; + job->id = atomic64_inc_return(&sched->job_id_count); INIT_WORK(&job->finish_work, amd_sched_job_finish); INIT_LIST_HEAD(&job->node); diff --git a/drivers/gpu/drm/amd/scheduler/gpu_scheduler.h b/drivers/gpu/drm/amd/scheduler/gpu_scheduler.h index 0255c7f8a6d8..924d4a5899e1 100644 --- a/drivers/gpu/drm/amd/scheduler/gpu_scheduler.h +++ b/drivers/gpu/drm/amd/scheduler/gpu_scheduler.h @@ -158,4 +158,6 @@ int amd_sched_job_init(struct amd_sched_job *job, void *owner); void amd_sched_hw_job_reset(struct amd_gpu_scheduler *sched); void amd_sched_job_recovery(struct amd_gpu_scheduler *sched); +bool amd_sched_dependency_optimized(struct dma_fence* fence, + struct amd_sched_entity *entity); #endif diff --git a/drivers/gpu/drm/arm/malidp_drv.h b/drivers/gpu/drm/arm/malidp_drv.h index 040311ffcaec..2e2033140efc 100644 --- a/drivers/gpu/drm/arm/malidp_drv.h +++ b/drivers/gpu/drm/arm/malidp_drv.h @@ -65,6 +65,6 @@ void malidp_de_planes_destroy(struct drm_device *drm); int malidp_crtc_init(struct drm_device *drm); /* often used combination of rotational bits */ -#define MALIDP_ROTATED_MASK (DRM_ROTATE_90 | DRM_ROTATE_270) +#define MALIDP_ROTATED_MASK (DRM_MODE_ROTATE_90 | DRM_MODE_ROTATE_270) #endif /* __MALIDP_DRV_H__ */ diff --git a/drivers/gpu/drm/arm/malidp_planes.c b/drivers/gpu/drm/arm/malidp_planes.c index 814fda23cead..063a8d2b0be3 100644 --- a/drivers/gpu/drm/arm/malidp_planes.c +++ b/drivers/gpu/drm/arm/malidp_planes.c @@ -80,7 +80,7 @@ static void malidp_plane_reset(struct drm_plane *plane) state = kzalloc(sizeof(*state), GFP_KERNEL); if (state) { state->base.plane = plane; - state->base.rotation = DRM_ROTATE_0; + state->base.rotation = DRM_MODE_ROTATE_0; plane->state = &state->base; } } @@ -221,7 +221,7 @@ static int malidp_de_plane_check(struct drm_plane *plane, return ret; /* packed RGB888 / BGR888 can't be rotated or flipped */ - if (state->rotation != DRM_ROTATE_0 && + if (state->rotation != DRM_MODE_ROTATE_0 && (fb->format->format == DRM_FORMAT_RGB888 || fb->format->format == DRM_FORMAT_BGR888)) return -EINVAL; @@ -315,12 +315,12 @@ static void malidp_de_plane_update(struct drm_plane *plane, val &= ~LAYER_ROT_MASK; /* setup the rotation and axis flip bits */ - if (plane->state->rotation & DRM_ROTATE_MASK) - val |= ilog2(plane->state->rotation & DRM_ROTATE_MASK) << + if (plane->state->rotation & DRM_MODE_ROTATE_MASK) + val |= ilog2(plane->state->rotation & DRM_MODE_ROTATE_MASK) << LAYER_ROT_OFFSET; - if (plane->state->rotation & DRM_REFLECT_X) + if (plane->state->rotation & DRM_MODE_REFLECT_X) val |= LAYER_H_FLIP; - if (plane->state->rotation & DRM_REFLECT_Y) + if (plane->state->rotation & DRM_MODE_REFLECT_Y) val |= LAYER_V_FLIP; /* @@ -370,8 +370,8 @@ int malidp_de_planes_init(struct drm_device *drm) struct malidp_plane *plane = NULL; enum drm_plane_type plane_type; unsigned long crtcs = 1 << drm->mode_config.num_crtc; - unsigned long flags = DRM_ROTATE_0 | DRM_ROTATE_90 | DRM_ROTATE_180 | - DRM_ROTATE_270 | DRM_REFLECT_X | DRM_REFLECT_Y; + unsigned long flags = DRM_MODE_ROTATE_0 | DRM_MODE_ROTATE_90 | DRM_MODE_ROTATE_180 | + DRM_MODE_ROTATE_270 | DRM_MODE_REFLECT_X | DRM_MODE_REFLECT_Y; u32 *formats; int ret, i, j, n; @@ -420,7 +420,7 @@ int malidp_de_planes_init(struct drm_device *drm) continue; } - drm_plane_create_rotation_property(&plane->base, DRM_ROTATE_0, flags); + drm_plane_create_rotation_property(&plane->base, DRM_MODE_ROTATE_0, flags); malidp_hw_write(malidp->dev, MALIDP_ALPHA_LUT, plane->layer->base + MALIDP_LAYER_COMPOSE); } diff --git a/drivers/gpu/drm/armada/armada_overlay.c b/drivers/gpu/drm/armada/armada_overlay.c index 424e465ff407..e9a29df4b443 100644 --- a/drivers/gpu/drm/armada/armada_overlay.c +++ b/drivers/gpu/drm/armada/armada_overlay.c @@ -125,7 +125,7 @@ armada_ovl_plane_update(struct drm_plane *plane, struct drm_crtc *crtc, src_x, src_y, src_w, src_h); ret = drm_plane_helper_check_update(plane, crtc, fb, &src, &dest, &clip, - DRM_ROTATE_0, + DRM_MODE_ROTATE_0, 0, INT_MAX, true, false, &visible); if (ret) return ret; diff --git a/drivers/gpu/drm/ast/Makefile b/drivers/gpu/drm/ast/Makefile index 171aa0622b66..617fdd39519c 100644 --- a/drivers/gpu/drm/ast/Makefile +++ b/drivers/gpu/drm/ast/Makefile @@ -2,8 +2,6 @@ # Makefile for the drm device driver. This driver provides support for the # Direct Rendering Infrastructure (DRI) in XFree86 4.1.0 and higher. -ccflags-y := -Iinclude/drm - ast-y := ast_drv.o ast_main.o ast_mode.o ast_fb.o ast_ttm.o ast_post.o ast_dp501.o obj-$(CONFIG_DRM_AST) := ast.o diff --git a/drivers/gpu/drm/ast/ast_ttm.c b/drivers/gpu/drm/ast/ast_ttm.c index e879496b8a42..58084985e6cf 100644 --- a/drivers/gpu/drm/ast/ast_ttm.c +++ b/drivers/gpu/drm/ast/ast_ttm.c @@ -26,8 +26,9 @@ * Authors: Dave Airlie <airlied@redhat.com> */ #include <drm/drmP.h> +#include <drm/ttm/ttm_page_alloc.h> + #include "ast_drv.h" -#include <ttm/ttm_page_alloc.h> static inline struct ast_private * ast_bdev(struct ttm_bo_device *bd) diff --git a/drivers/gpu/drm/atmel-hlcdc/atmel_hlcdc_plane.c b/drivers/gpu/drm/atmel-hlcdc/atmel_hlcdc_plane.c index 29cc10d053eb..1124200bb280 100644 --- a/drivers/gpu/drm/atmel-hlcdc/atmel_hlcdc_plane.c +++ b/drivers/gpu/drm/atmel-hlcdc/atmel_hlcdc_plane.c @@ -678,8 +678,8 @@ static int atmel_hlcdc_plane_atomic_check(struct drm_plane *p, if (!state->bpp[i]) return -EINVAL; - switch (state->base.rotation & DRM_ROTATE_MASK) { - case DRM_ROTATE_90: + switch (state->base.rotation & DRM_MODE_ROTATE_MASK) { + case DRM_MODE_ROTATE_90: offset = ((y_offset + state->src_y + patched_src_w - 1) / ydiv) * fb->pitches[i]; offset += ((x_offset + state->src_x) / xdiv) * @@ -688,7 +688,7 @@ static int atmel_hlcdc_plane_atomic_check(struct drm_plane *p, fb->pitches[i]; state->pstride[i] = -fb->pitches[i] - state->bpp[i]; break; - case DRM_ROTATE_180: + case DRM_MODE_ROTATE_180: offset = ((y_offset + state->src_y + patched_src_h - 1) / ydiv) * fb->pitches[i]; offset += ((x_offset + state->src_x + patched_src_w - 1) / @@ -697,7 +697,7 @@ static int atmel_hlcdc_plane_atomic_check(struct drm_plane *p, state->bpp[i]) - fb->pitches[i]; state->pstride[i] = -2 * state->bpp[i]; break; - case DRM_ROTATE_270: + case DRM_MODE_ROTATE_270: offset = ((y_offset + state->src_y) / ydiv) * fb->pitches[i]; offset += ((x_offset + state->src_x + patched_src_h - 1) / @@ -707,7 +707,7 @@ static int atmel_hlcdc_plane_atomic_check(struct drm_plane *p, (2 * state->bpp[i]); state->pstride[i] = fb->pitches[i] - state->bpp[i]; break; - case DRM_ROTATE_0: + case DRM_MODE_ROTATE_0: default: offset = ((y_offset + state->src_y) / ydiv) * fb->pitches[i]; @@ -864,11 +864,11 @@ static int atmel_hlcdc_plane_init_properties(struct atmel_hlcdc_plane *plane, int ret; ret = drm_plane_create_rotation_property(&plane->base, - DRM_ROTATE_0, - DRM_ROTATE_0 | - DRM_ROTATE_90 | - DRM_ROTATE_180 | - DRM_ROTATE_270); + DRM_MODE_ROTATE_0, + DRM_MODE_ROTATE_0 | + DRM_MODE_ROTATE_90 | + DRM_MODE_ROTATE_180 | + DRM_MODE_ROTATE_270); if (ret) return ret; } diff --git a/drivers/gpu/drm/bochs/Makefile b/drivers/gpu/drm/bochs/Makefile index 844a55614920..98ef60a19e8f 100644 --- a/drivers/gpu/drm/bochs/Makefile +++ b/drivers/gpu/drm/bochs/Makefile @@ -1,4 +1,3 @@ -ccflags-y := -Iinclude/drm bochs-drm-y := bochs_drv.o bochs_mm.o bochs_kms.o bochs_fbdev.o bochs_hw.o obj-$(CONFIG_DRM_BOCHS) += bochs-drm.o diff --git a/drivers/gpu/drm/bochs/bochs.h b/drivers/gpu/drm/bochs/bochs.h index f626bab7f5e3..76c490c3cdbc 100644 --- a/drivers/gpu/drm/bochs/bochs.h +++ b/drivers/gpu/drm/bochs/bochs.h @@ -9,8 +9,8 @@ #include <drm/drm_gem.h> -#include <ttm/ttm_bo_driver.h> -#include <ttm/ttm_page_alloc.h> +#include <drm/ttm/ttm_bo_driver.h> +#include <drm/ttm/ttm_page_alloc.h> /* ---------------------------------------------------------------------- */ diff --git a/drivers/gpu/drm/bridge/Makefile b/drivers/gpu/drm/bridge/Makefile index 3fe2226ee2f2..defcf1e7ca1c 100644 --- a/drivers/gpu/drm/bridge/Makefile +++ b/drivers/gpu/drm/bridge/Makefile @@ -1,5 +1,3 @@ -ccflags-y := -Iinclude/drm - obj-$(CONFIG_DRM_ANALOGIX_ANX78XX) += analogix-anx78xx.o obj-$(CONFIG_DRM_DUMB_VGA_DAC) += dumb-vga-dac.o obj-$(CONFIG_DRM_LVDS_ENCODER) += lvds-encoder.o diff --git a/drivers/gpu/drm/bridge/nxp-ptn3460.c b/drivers/gpu/drm/bridge/nxp-ptn3460.c index 351704390d02..4f64e717e01b 100644 --- a/drivers/gpu/drm/bridge/nxp-ptn3460.c +++ b/drivers/gpu/drm/bridge/nxp-ptn3460.c @@ -20,15 +20,13 @@ #include <linux/module.h> #include <linux/of.h> #include <linux/of_gpio.h> - +#include <drm/drm_atomic_helper.h> +#include <drm/drm_crtc.h> +#include <drm/drm_crtc_helper.h> +#include <drm/drm_edid.h> #include <drm/drm_of.h> #include <drm/drm_panel.h> - -#include "drm_crtc.h" -#include "drm_crtc_helper.h" -#include "drm_atomic_helper.h" -#include "drm_edid.h" -#include "drmP.h" +#include <drm/drmP.h> #define PTN3460_EDID_ADDR 0x0 #define PTN3460_EDID_EMULATION_ADDR 0x84 diff --git a/drivers/gpu/drm/bridge/parade-ps8622.c b/drivers/gpu/drm/bridge/parade-ps8622.c index 1dcec3b97e67..6f22f9fec9bf 100644 --- a/drivers/gpu/drm/bridge/parade-ps8622.c +++ b/drivers/gpu/drm/bridge/parade-ps8622.c @@ -24,14 +24,12 @@ #include <linux/of_device.h> #include <linux/pm.h> #include <linux/regulator/consumer.h> - +#include <drm/drm_atomic_helper.h> +#include <drm/drm_crtc.h> +#include <drm/drm_crtc_helper.h> #include <drm/drm_of.h> #include <drm/drm_panel.h> - -#include "drmP.h" -#include "drm_crtc.h" -#include "drm_crtc_helper.h" -#include "drm_atomic_helper.h" +#include <drm/drmP.h> /* Brightness scale on the Parade chip */ #define PS8622_MAX_BRIGHTNESS 0xff diff --git a/drivers/gpu/drm/bridge/sii902x.c b/drivers/gpu/drm/bridge/sii902x.c index 9126d0306ab5..9b87067c022c 100644 --- a/drivers/gpu/drm/bridge/sii902x.c +++ b/drivers/gpu/drm/bridge/sii902x.c @@ -160,7 +160,7 @@ static int sii902x_get_modes(struct drm_connector *connector) time_before(jiffies, timeout)); if (!(status & SII902X_SYS_CTRL_DDC_BUS_GRTD)) { - dev_err(&sii902x->i2c->dev, "failed to acquire the i2c bus"); + dev_err(&sii902x->i2c->dev, "failed to acquire the i2c bus\n"); return -ETIMEDOUT; } @@ -202,7 +202,7 @@ static int sii902x_get_modes(struct drm_connector *connector) if (status & (SII902X_SYS_CTRL_DDC_BUS_REQ | SII902X_SYS_CTRL_DDC_BUS_GRTD)) { - dev_err(&sii902x->i2c->dev, "failed to release the i2c bus"); + dev_err(&sii902x->i2c->dev, "failed to release the i2c bus\n"); return -ETIMEDOUT; } @@ -298,7 +298,7 @@ static int sii902x_bridge_attach(struct drm_bridge *bridge) if (!drm_core_check_feature(drm, DRIVER_ATOMIC)) { dev_err(&sii902x->i2c->dev, - "sii902x driver is only compatible with DRM devices supporting atomic updates"); + "sii902x driver is only compatible with DRM devices supporting atomic updates\n"); return -ENOTSUPP; } diff --git a/drivers/gpu/drm/bridge/synopsys/dw-hdmi.c b/drivers/gpu/drm/bridge/synopsys/dw-hdmi.c index 4e1f54a675d8..8737de8c1c52 100644 --- a/drivers/gpu/drm/bridge/synopsys/dw-hdmi.c +++ b/drivers/gpu/drm/bridge/synopsys/dw-hdmi.c @@ -173,6 +173,8 @@ struct dw_hdmi { unsigned int reg_shift; struct regmap *regm; + void (*enable_audio)(struct dw_hdmi *hdmi); + void (*disable_audio)(struct dw_hdmi *hdmi); }; #define HDMI_IH_PHY_STAT0_RX_SENSE \ @@ -542,13 +544,41 @@ void dw_hdmi_set_sample_rate(struct dw_hdmi *hdmi, unsigned int rate) } EXPORT_SYMBOL_GPL(dw_hdmi_set_sample_rate); +static void hdmi_enable_audio_clk(struct dw_hdmi *hdmi, bool enable) +{ + hdmi_modb(hdmi, enable ? 0 : HDMI_MC_CLKDIS_AUDCLK_DISABLE, + HDMI_MC_CLKDIS_AUDCLK_DISABLE, HDMI_MC_CLKDIS); +} + +static void dw_hdmi_ahb_audio_enable(struct dw_hdmi *hdmi) +{ + hdmi_set_cts_n(hdmi, hdmi->audio_cts, hdmi->audio_n); +} + +static void dw_hdmi_ahb_audio_disable(struct dw_hdmi *hdmi) +{ + hdmi_set_cts_n(hdmi, hdmi->audio_cts, 0); +} + +static void dw_hdmi_i2s_audio_enable(struct dw_hdmi *hdmi) +{ + hdmi_set_cts_n(hdmi, hdmi->audio_cts, hdmi->audio_n); + hdmi_enable_audio_clk(hdmi, true); +} + +static void dw_hdmi_i2s_audio_disable(struct dw_hdmi *hdmi) +{ + hdmi_enable_audio_clk(hdmi, false); +} + void dw_hdmi_audio_enable(struct dw_hdmi *hdmi) { unsigned long flags; spin_lock_irqsave(&hdmi->audio_lock, flags); hdmi->audio_enable = true; - hdmi_set_cts_n(hdmi, hdmi->audio_cts, hdmi->audio_n); + if (hdmi->enable_audio) + hdmi->enable_audio(hdmi); spin_unlock_irqrestore(&hdmi->audio_lock, flags); } EXPORT_SYMBOL_GPL(dw_hdmi_audio_enable); @@ -559,7 +589,8 @@ void dw_hdmi_audio_disable(struct dw_hdmi *hdmi) spin_lock_irqsave(&hdmi->audio_lock, flags); hdmi->audio_enable = false; - hdmi_set_cts_n(hdmi, hdmi->audio_cts, 0); + if (hdmi->disable_audio) + hdmi->disable_audio(hdmi); spin_unlock_irqrestore(&hdmi->audio_lock, flags); } EXPORT_SYMBOL_GPL(dw_hdmi_audio_disable); @@ -1573,11 +1604,6 @@ static void dw_hdmi_enable_video_path(struct dw_hdmi *hdmi) HDMI_MC_FLOWCTRL); } -static void hdmi_enable_audio_clk(struct dw_hdmi *hdmi) -{ - hdmi_modb(hdmi, 0, HDMI_MC_CLKDIS_AUDCLK_DISABLE, HDMI_MC_CLKDIS); -} - /* Workaround to clear the overflow condition */ static void dw_hdmi_clear_overflow(struct dw_hdmi *hdmi) { @@ -1691,7 +1717,7 @@ static int dw_hdmi_setup(struct dw_hdmi *hdmi, struct drm_display_mode *mode) /* HDMI Initialization Step E - Configure audio */ hdmi_clk_regenerator_update_pixel_clock(hdmi); - hdmi_enable_audio_clk(hdmi); + hdmi_enable_audio_clk(hdmi, true); } /* not for DVI mode */ @@ -2403,6 +2429,8 @@ __dw_hdmi_probe(struct platform_device *pdev, audio.irq = irq; audio.hdmi = hdmi; audio.eld = hdmi->connector.eld; + hdmi->enable_audio = dw_hdmi_ahb_audio_enable; + hdmi->disable_audio = dw_hdmi_ahb_audio_disable; pdevinfo.name = "dw-hdmi-ahb-audio"; pdevinfo.data = &audio; @@ -2415,6 +2443,8 @@ __dw_hdmi_probe(struct platform_device *pdev, audio.hdmi = hdmi; audio.write = hdmi_writeb; audio.read = hdmi_readb; + hdmi->enable_audio = dw_hdmi_i2s_audio_enable; + hdmi->disable_audio = dw_hdmi_i2s_audio_disable; pdevinfo.name = "dw-hdmi-i2s-audio"; pdevinfo.data = &audio; diff --git a/drivers/gpu/drm/cirrus/Makefile b/drivers/gpu/drm/cirrus/Makefile index 69ffe7006d55..919c0a336c97 100644 --- a/drivers/gpu/drm/cirrus/Makefile +++ b/drivers/gpu/drm/cirrus/Makefile @@ -1,4 +1,3 @@ -ccflags-y := -Iinclude/drm cirrus-y := cirrus_main.o cirrus_mode.o \ cirrus_drv.o cirrus_fbdev.o cirrus_ttm.o diff --git a/drivers/gpu/drm/cirrus/cirrus_ttm.c b/drivers/gpu/drm/cirrus/cirrus_ttm.c index 93dbcd38355d..1ff1838c0d44 100644 --- a/drivers/gpu/drm/cirrus/cirrus_ttm.c +++ b/drivers/gpu/drm/cirrus/cirrus_ttm.c @@ -26,8 +26,9 @@ * Authors: Dave Airlie <airlied@redhat.com> */ #include <drm/drmP.h> +#include <drm/ttm/ttm_page_alloc.h> + #include "cirrus_drv.h" -#include <ttm/ttm_page_alloc.h> static inline struct cirrus_device * cirrus_bdev(struct ttm_bo_device *bd) diff --git a/drivers/gpu/drm/drm_atomic.c b/drivers/gpu/drm/drm_atomic.c index f32506a7c1d6..e1637011e18a 100644 --- a/drivers/gpu/drm/drm_atomic.c +++ b/drivers/gpu/drm/drm_atomic.c @@ -57,6 +57,7 @@ void drm_atomic_state_default_release(struct drm_atomic_state *state) kfree(state->connectors); kfree(state->crtcs); kfree(state->planes); + kfree(state->private_objs); } EXPORT_SYMBOL(drm_atomic_state_default_release); @@ -184,6 +185,17 @@ void drm_atomic_state_default_clear(struct drm_atomic_state *state) state->planes[i].ptr = NULL; state->planes[i].state = NULL; } + + for (i = 0; i < state->num_private_objs; i++) { + void *obj_state = state->private_objs[i].obj_state; + + state->private_objs[i].funcs->destroy_state(obj_state); + state->private_objs[i].obj = NULL; + state->private_objs[i].obj_state = NULL; + state->private_objs[i].funcs = NULL; + } + state->num_private_objs = 0; + } EXPORT_SYMBOL(drm_atomic_state_default_clear); @@ -425,7 +437,7 @@ drm_atomic_replace_property_blob(struct drm_property_blob **blob, } static int -drm_atomic_replace_property_blob_from_id(struct drm_crtc *crtc, +drm_atomic_replace_property_blob_from_id(struct drm_device *dev, struct drm_property_blob **blob, uint64_t blob_id, ssize_t expected_size, @@ -434,7 +446,7 @@ drm_atomic_replace_property_blob_from_id(struct drm_crtc *crtc, struct drm_property_blob *new_blob = NULL; if (blob_id != 0) { - new_blob = drm_property_lookup_blob(crtc->dev, blob_id); + new_blob = drm_property_lookup_blob(dev, blob_id); if (new_blob == NULL) return -EINVAL; @@ -483,7 +495,7 @@ int drm_atomic_crtc_set_property(struct drm_crtc *crtc, drm_property_blob_put(mode); return ret; } else if (property == config->degamma_lut_property) { - ret = drm_atomic_replace_property_blob_from_id(crtc, + ret = drm_atomic_replace_property_blob_from_id(dev, &state->degamma_lut, val, -1, @@ -491,7 +503,7 @@ int drm_atomic_crtc_set_property(struct drm_crtc *crtc, state->color_mgmt_changed |= replaced; return ret; } else if (property == config->ctm_property) { - ret = drm_atomic_replace_property_blob_from_id(crtc, + ret = drm_atomic_replace_property_blob_from_id(dev, &state->ctm, val, sizeof(struct drm_color_ctm), @@ -499,7 +511,7 @@ int drm_atomic_crtc_set_property(struct drm_crtc *crtc, state->color_mgmt_changed |= replaced; return ret; } else if (property == config->gamma_lut_property) { - ret = drm_atomic_replace_property_blob_from_id(crtc, + ret = drm_atomic_replace_property_blob_from_id(dev, &state->gamma_lut, val, -1, @@ -769,7 +781,7 @@ int drm_atomic_plane_set_property(struct drm_plane *plane, } else if (property == config->prop_src_h) { state->src_h = val; } else if (property == plane->rotation_property) { - if (!is_power_of_2(val & DRM_ROTATE_MASK)) + if (!is_power_of_2(val & DRM_MODE_ROTATE_MASK)) return -EINVAL; state->rotation = val; } else if (property == plane->zpos_property) { @@ -978,6 +990,59 @@ static void drm_atomic_plane_print_state(struct drm_printer *p, } /** + * drm_atomic_get_private_obj_state - get private object state + * @state: global atomic state + * @obj: private object to get the state for + * @funcs: pointer to the struct of function pointers that identify the object + * type + * + * This function returns the private object state for the given private object, + * allocating the state if needed. It does not grab any locks as the caller is + * expected to care of any required locking. + * + * RETURNS: + * + * Either the allocated state or the error code encoded into a pointer. + */ +void * +drm_atomic_get_private_obj_state(struct drm_atomic_state *state, void *obj, + const struct drm_private_state_funcs *funcs) +{ + int index, num_objs, i; + size_t size; + struct __drm_private_objs_state *arr; + + for (i = 0; i < state->num_private_objs; i++) + if (obj == state->private_objs[i].obj && + state->private_objs[i].obj_state) + return state->private_objs[i].obj_state; + + num_objs = state->num_private_objs + 1; + size = sizeof(*state->private_objs) * num_objs; + arr = krealloc(state->private_objs, size, GFP_KERNEL); + if (!arr) + return ERR_PTR(-ENOMEM); + + state->private_objs = arr; + index = state->num_private_objs; + memset(&state->private_objs[index], 0, sizeof(*state->private_objs)); + + state->private_objs[index].obj_state = funcs->duplicate_state(state, obj); + if (!state->private_objs[index].obj_state) + return ERR_PTR(-ENOMEM); + + state->private_objs[index].obj = obj; + state->private_objs[index].funcs = funcs; + state->num_private_objs = num_objs; + + DRM_DEBUG_ATOMIC("Added new private object state %p to %p\n", + state->private_objs[index].obj_state, state); + + return state->private_objs[index].obj_state; +} +EXPORT_SYMBOL(drm_atomic_get_private_obj_state); + +/** * drm_atomic_get_connector_state - get connector state * @state: global atomic state object * @connector: connector to get state object for @@ -1123,6 +1188,10 @@ int drm_atomic_connector_set_property(struct drm_connector *connector, */ if (state->link_status != DRM_LINK_STATUS_GOOD) state->link_status = val; + } else if (property == config->aspect_ratio_property) { + state->picture_aspect_ratio = val; + } else if (property == connector->scaling_mode_property) { + state->scaling_mode = val; } else if (connector->funcs->atomic_set_property) { return connector->funcs->atomic_set_property(connector, state, property, val); @@ -1199,6 +1268,10 @@ drm_atomic_connector_get_property(struct drm_connector *connector, *val = state->tv.hue; } else if (property == config->link_status_property) { *val = state->link_status; + } else if (property == config->aspect_ratio_property) { + *val = state->picture_aspect_ratio; + } else if (property == connector->scaling_mode_property) { + *val = state->scaling_mode; } else if (connector->funcs->atomic_get_property) { return connector->funcs->atomic_get_property(connector, state, property, val); @@ -1618,7 +1691,7 @@ int drm_atomic_commit(struct drm_atomic_state *state) if (ret) return ret; - DRM_DEBUG_ATOMIC("commiting %p\n", state); + DRM_DEBUG_ATOMIC("committing %p\n", state); return config->funcs->atomic_commit(state->dev, state, false); } @@ -1647,7 +1720,7 @@ int drm_atomic_nonblocking_commit(struct drm_atomic_state *state) if (ret) return ret; - DRM_DEBUG_ATOMIC("commiting %p nonblocking\n", state); + DRM_DEBUG_ATOMIC("committing %p nonblocking\n", state); return config->funcs->atomic_commit(state->dev, state, true); } diff --git a/drivers/gpu/drm/drm_atomic_helper.c b/drivers/gpu/drm/drm_atomic_helper.c index 8be9719284b0..636e561486a8 100644 --- a/drivers/gpu/drm/drm_atomic_helper.c +++ b/drivers/gpu/drm/drm_atomic_helper.c @@ -1070,8 +1070,8 @@ EXPORT_SYMBOL(drm_atomic_helper_commit_modeset_enables); * * Note that @pre_swap is needed since the point where we block for fences moves * around depending upon whether an atomic commit is blocking or - * non-blocking. For async commit all waiting needs to happen after - * drm_atomic_helper_swap_state() is called, but for synchronous commits we want + * non-blocking. For non-blocking commit all waiting needs to happen after + * drm_atomic_helper_swap_state() is called, but for blocking commits we want * to wait **before** we do anything that can't be easily rolled back. That is * before we call drm_atomic_helper_swap_state(). * @@ -2032,6 +2032,8 @@ void drm_atomic_helper_swap_state(struct drm_atomic_state *state, struct drm_plane *plane; struct drm_plane_state *old_plane_state, *new_plane_state; struct drm_crtc_commit *commit; + void *obj, *obj_state; + const struct drm_private_state_funcs *funcs; if (stall) { for_each_new_crtc_in_state(state, crtc, new_crtc_state, i) { @@ -2092,6 +2094,9 @@ void drm_atomic_helper_swap_state(struct drm_atomic_state *state, state->planes[i].state = old_plane_state; plane->state = new_plane_state; } + + __for_each_private_obj(state, obj, obj_state, i, funcs) + funcs->swap_state(obj, &state->private_objs[i].obj_state); } EXPORT_SYMBOL(drm_atomic_helper_swap_state); @@ -3220,7 +3225,7 @@ void drm_atomic_helper_plane_reset(struct drm_plane *plane) if (plane->state) { plane->state->plane = plane; - plane->state->rotation = DRM_ROTATE_0; + plane->state->rotation = DRM_MODE_ROTATE_0; } } EXPORT_SYMBOL(drm_atomic_helper_plane_reset); @@ -3517,7 +3522,8 @@ EXPORT_SYMBOL(drm_atomic_helper_connector_destroy_state); * * Implements support for legacy gamma correction table for drivers * that support color management through the DEGAMMA_LUT/GAMMA_LUT - * properties. + * properties. See drm_crtc_enable_color_mgmt() and the containing chapter for + * how the atomic color management and gamma tables work. */ int drm_atomic_helper_legacy_gamma_set(struct drm_crtc *crtc, u16 *red, u16 *green, u16 *blue, diff --git a/drivers/gpu/drm/drm_blend.c b/drivers/gpu/drm/drm_blend.c index a0d0d6843288..db6aeec50b82 100644 --- a/drivers/gpu/drm/drm_blend.c +++ b/drivers/gpu/drm/drm_blend.c @@ -119,17 +119,17 @@ * drm_property_create_bitmask()) called "rotation" and has the following * bitmask enumaration values: * - * DRM_ROTATE_0: + * DRM_MODE_ROTATE_0: * "rotate-0" - * DRM_ROTATE_90: + * DRM_MODE_ROTATE_90: * "rotate-90" - * DRM_ROTATE_180: + * DRM_MODE_ROTATE_180: * "rotate-180" - * DRM_ROTATE_270: + * DRM_MODE_ROTATE_270: * "rotate-270" - * DRM_REFLECT_X: + * DRM_MODE_REFLECT_X: * "reflect-x" - * DRM_REFELCT_Y: + * DRM_MODE_REFLECT_Y: * "reflect-y" * * Rotation is the specified amount in degrees in counter clockwise direction, @@ -142,17 +142,17 @@ int drm_plane_create_rotation_property(struct drm_plane *plane, unsigned int supported_rotations) { static const struct drm_prop_enum_list props[] = { - { __builtin_ffs(DRM_ROTATE_0) - 1, "rotate-0" }, - { __builtin_ffs(DRM_ROTATE_90) - 1, "rotate-90" }, - { __builtin_ffs(DRM_ROTATE_180) - 1, "rotate-180" }, - { __builtin_ffs(DRM_ROTATE_270) - 1, "rotate-270" }, - { __builtin_ffs(DRM_REFLECT_X) - 1, "reflect-x" }, - { __builtin_ffs(DRM_REFLECT_Y) - 1, "reflect-y" }, + { __builtin_ffs(DRM_MODE_ROTATE_0) - 1, "rotate-0" }, + { __builtin_ffs(DRM_MODE_ROTATE_90) - 1, "rotate-90" }, + { __builtin_ffs(DRM_MODE_ROTATE_180) - 1, "rotate-180" }, + { __builtin_ffs(DRM_MODE_ROTATE_270) - 1, "rotate-270" }, + { __builtin_ffs(DRM_MODE_REFLECT_X) - 1, "reflect-x" }, + { __builtin_ffs(DRM_MODE_REFLECT_Y) - 1, "reflect-y" }, }; struct drm_property *prop; - WARN_ON((supported_rotations & DRM_ROTATE_MASK) == 0); - WARN_ON(!is_power_of_2(rotation & DRM_ROTATE_MASK)); + WARN_ON((supported_rotations & DRM_MODE_ROTATE_MASK) == 0); + WARN_ON(!is_power_of_2(rotation & DRM_MODE_ROTATE_MASK)); WARN_ON(rotation & ~supported_rotations); prop = drm_property_create_bitmask(plane->dev, 0, "rotation", @@ -178,14 +178,14 @@ EXPORT_SYMBOL(drm_plane_create_rotation_property); * @supported_rotations: Supported rotations * * Attempt to simplify the rotation to a form that is supported. - * Eg. if the hardware supports everything except DRM_REFLECT_X + * Eg. if the hardware supports everything except DRM_MODE_REFLECT_X * one could call this function like this: * - * drm_rotation_simplify(rotation, DRM_ROTATE_0 | - * DRM_ROTATE_90 | DRM_ROTATE_180 | - * DRM_ROTATE_270 | DRM_REFLECT_Y); + * drm_rotation_simplify(rotation, DRM_MODE_ROTATE_0 | + * DRM_MODE_ROTATE_90 | DRM_MODE_ROTATE_180 | + * DRM_MODE_ROTATE_270 | DRM_MODE_REFLECT_Y); * - * to eliminate the DRM_ROTATE_X flag. Depending on what kind of + * to eliminate the DRM_MODE_ROTATE_X flag. Depending on what kind of * transforms the hardware supports, this function may not * be able to produce a supported transform, so the caller should * check the result afterwards. @@ -194,9 +194,10 @@ unsigned int drm_rotation_simplify(unsigned int rotation, unsigned int supported_rotations) { if (rotation & ~supported_rotations) { - rotation ^= DRM_REFLECT_X | DRM_REFLECT_Y; - rotation = (rotation & DRM_REFLECT_MASK) | - BIT((ffs(rotation & DRM_ROTATE_MASK) + 1) % 4); + rotation ^= DRM_MODE_REFLECT_X | DRM_MODE_REFLECT_Y; + rotation = (rotation & DRM_MODE_REFLECT_MASK) | + BIT((ffs(rotation & DRM_MODE_ROTATE_MASK) + 1) + % 4); } return rotation; diff --git a/drivers/gpu/drm/drm_color_mgmt.c b/drivers/gpu/drm/drm_color_mgmt.c index 533f3a3e6877..3eda500fc005 100644 --- a/drivers/gpu/drm/drm_color_mgmt.c +++ b/drivers/gpu/drm/drm_color_mgmt.c @@ -43,7 +43,8 @@ * * Setting this to NULL (blob property value set to 0) means a * linear/pass-thru gamma table should be used. This is generally the - * driver boot-up state too. + * driver boot-up state too. Drivers can access this blob through + * &drm_crtc_state.degamma_lut. * * “DEGAMMA_LUT_SIZE”: * Unsinged range property to give the size of the lookup table to be set @@ -60,7 +61,8 @@ * * Setting this to NULL (blob property value set to 0) means a * unit/pass-thru matrix should be used. This is generally the driver - * boot-up state too. + * boot-up state too. Drivers can access the blob for the color conversion + * matrix through &drm_crtc_state.ctm. * * “GAMMA_LUT”: * Blob property to set the gamma lookup table (LUT) mapping pixel data @@ -72,7 +74,8 @@ * * Setting this to NULL (blob property value set to 0) means a * linear/pass-thru gamma table should be used. This is generally the - * driver boot-up state too. + * driver boot-up state too. Drivers can access this blob through + * &drm_crtc_state.gamma_lut. * * “GAMMA_LUT_SIZE”: * Unsigned range property to give the size of the lookup table to be set diff --git a/drivers/gpu/drm/drm_connector.c b/drivers/gpu/drm/drm_connector.c index 9f847615ac74..5cd61aff7857 100644 --- a/drivers/gpu/drm/drm_connector.c +++ b/drivers/gpu/drm/drm_connector.c @@ -941,6 +941,10 @@ EXPORT_SYMBOL(drm_mode_create_tv_properties); * * Called by a driver the first time it's needed, must be attached to desired * connectors. + * + * Atomic drivers should use drm_connector_attach_scaling_mode_property() + * instead to correctly assign &drm_connector_state.picture_aspect_ratio + * in the atomic state. */ int drm_mode_create_scaling_mode_property(struct drm_device *dev) { @@ -961,6 +965,66 @@ int drm_mode_create_scaling_mode_property(struct drm_device *dev) EXPORT_SYMBOL(drm_mode_create_scaling_mode_property); /** + * drm_connector_attach_scaling_mode_property - attach atomic scaling mode property + * @connector: connector to attach scaling mode property on. + * @scaling_mode_mask: or'ed mask of BIT(%DRM_MODE_SCALE_\*). + * + * This is used to add support for scaling mode to atomic drivers. + * The scaling mode will be set to &drm_connector_state.picture_aspect_ratio + * and can be used from &drm_connector_helper_funcs->atomic_check for validation. + * + * This is the atomic version of drm_mode_create_scaling_mode_property(). + * + * Returns: + * Zero on success, negative errno on failure. + */ +int drm_connector_attach_scaling_mode_property(struct drm_connector *connector, + u32 scaling_mode_mask) +{ + struct drm_device *dev = connector->dev; + struct drm_property *scaling_mode_property; + int i, j = 0; + const unsigned valid_scaling_mode_mask = + (1U << ARRAY_SIZE(drm_scaling_mode_enum_list)) - 1; + + if (WARN_ON(hweight32(scaling_mode_mask) < 2 || + scaling_mode_mask & ~valid_scaling_mode_mask)) + return -EINVAL; + + scaling_mode_property = + drm_property_create(dev, DRM_MODE_PROP_ENUM, "scaling mode", + hweight32(scaling_mode_mask)); + + if (!scaling_mode_property) + return -ENOMEM; + + for (i = 0; i < ARRAY_SIZE(drm_scaling_mode_enum_list); i++) { + int ret; + + if (!(BIT(i) & scaling_mode_mask)) + continue; + + ret = drm_property_add_enum(scaling_mode_property, j++, + drm_scaling_mode_enum_list[i].type, + drm_scaling_mode_enum_list[i].name); + + if (ret) { + drm_property_destroy(dev, scaling_mode_property); + + return ret; + } + } + + drm_object_attach_property(&connector->base, + scaling_mode_property, 0); + + connector->scaling_mode_property = scaling_mode_property; + + return 0; +} +EXPORT_SYMBOL(drm_connector_attach_scaling_mode_property); + +/** * drm_mode_create_aspect_ratio_property - create aspect ratio property * @dev: DRM device * diff --git a/drivers/gpu/drm/drm_dp_mst_topology.c b/drivers/gpu/drm/drm_dp_mst_topology.c index d3fc7e4e85b7..222eb1a8549b 100644 --- a/drivers/gpu/drm/drm_dp_mst_topology.c +++ b/drivers/gpu/drm/drm_dp_mst_topology.c @@ -737,16 +737,16 @@ static void drm_dp_mst_put_payload_id(struct drm_dp_mst_topology_mgr *mgr, static bool check_txmsg_state(struct drm_dp_mst_topology_mgr *mgr, struct drm_dp_sideband_msg_tx *txmsg) { - bool ret; + unsigned int state; /* * All updates to txmsg->state are protected by mgr->qlock, and the two * cases we check here are terminal states. For those the barriers * provided by the wake_up/wait_event pair are enough. */ - ret = (txmsg->state == DRM_DP_SIDEBAND_TX_RX || - txmsg->state == DRM_DP_SIDEBAND_TX_TIMEOUT); - return ret; + state = READ_ONCE(txmsg->state); + return (state == DRM_DP_SIDEBAND_TX_RX || + state == DRM_DP_SIDEBAND_TX_TIMEOUT); } static int drm_dp_mst_wait_tx_reply(struct drm_dp_mst_branch *mstb, @@ -855,7 +855,7 @@ static void drm_dp_destroy_mst_branch_device(struct kref *kref) mutex_unlock(&mstb->mgr->qlock); if (wake_tx) - wake_up(&mstb->mgr->tx_waitq); + wake_up_all(&mstb->mgr->tx_waitq); kref_put(kref, drm_dp_free_mst_branch_device); } @@ -1510,7 +1510,7 @@ static void process_single_down_tx_qlock(struct drm_dp_mst_topology_mgr *mgr) if (txmsg->seqno != -1) txmsg->dst->tx_slots[txmsg->seqno] = NULL; txmsg->state = DRM_DP_SIDEBAND_TX_TIMEOUT; - wake_up(&mgr->tx_waitq); + wake_up_all(&mgr->tx_waitq); } } @@ -2258,7 +2258,7 @@ static int drm_dp_mst_handle_down_rep(struct drm_dp_mst_topology_mgr *mgr) mstb->tx_slots[slot] = NULL; mutex_unlock(&mgr->qlock); - wake_up(&mgr->tx_waitq); + wake_up_all(&mgr->tx_waitq); } return ret; } @@ -2498,6 +2498,81 @@ static int drm_dp_init_vcpi(struct drm_dp_mst_topology_mgr *mgr, } /** + * drm_dp_atomic_find_vcpi_slots() - Find and add vcpi slots to the state + * @state: global atomic state + * @mgr: MST topology manager for the port + * @port: port to find vcpi slots for + * @pbn: bandwidth required for the mode in PBN + * + * RETURNS: + * Total slots in the atomic state assigned for this port or error + */ +int drm_dp_atomic_find_vcpi_slots(struct drm_atomic_state *state, + struct drm_dp_mst_topology_mgr *mgr, + struct drm_dp_mst_port *port, int pbn) +{ + struct drm_dp_mst_topology_state *topology_state; + int req_slots; + + topology_state = drm_atomic_get_mst_topology_state(state, mgr); + if (topology_state == NULL) + return -ENOMEM; + + port = drm_dp_get_validated_port_ref(mgr, port); + if (port == NULL) + return -EINVAL; + req_slots = DIV_ROUND_UP(pbn, mgr->pbn_div); + DRM_DEBUG_KMS("vcpi slots req=%d, avail=%d\n", + req_slots, topology_state->avail_slots); + + if (req_slots > topology_state->avail_slots) { + drm_dp_put_port(port); + return -ENOSPC; + } + + topology_state->avail_slots -= req_slots; + DRM_DEBUG_KMS("vcpi slots avail=%d", topology_state->avail_slots); + + drm_dp_put_port(port); + return req_slots; +} +EXPORT_SYMBOL(drm_dp_atomic_find_vcpi_slots); + +/** + * drm_dp_atomic_release_vcpi_slots() - Release allocated vcpi slots + * @state: global atomic state + * @mgr: MST topology manager for the port + * @slots: number of vcpi slots to release + * + * RETURNS: + * 0 if @slots were added back to &drm_dp_mst_topology_state->avail_slots or + * negative error code + */ +int drm_dp_atomic_release_vcpi_slots(struct drm_atomic_state *state, + struct drm_dp_mst_topology_mgr *mgr, + int slots) +{ + struct drm_dp_mst_topology_state *topology_state; + + topology_state = drm_atomic_get_mst_topology_state(state, mgr); + if (topology_state == NULL) + return -ENOMEM; + + /* We cannot rely on port->vcpi.num_slots to update + * topology_state->avail_slots as the port may not exist if the parent + * branch device was unplugged. This should be fixed by tracking + * per-port slot allocation in drm_dp_mst_topology_state instead of + * depending on the caller to tell us how many slots to release. + */ + topology_state->avail_slots += slots; + DRM_DEBUG_KMS("vcpi slots released=%d, avail=%d\n", + slots, topology_state->avail_slots); + + return 0; +} +EXPORT_SYMBOL(drm_dp_atomic_release_vcpi_slots); + +/** * drm_dp_mst_allocate_vcpi() - Allocate a virtual channel * @mgr: manager for this port * @port: port to allocate a virtual channel for. @@ -2936,6 +3011,69 @@ static void drm_dp_destroy_connector_work(struct work_struct *work) (*mgr->cbs->hotplug)(mgr); } +void *drm_dp_mst_duplicate_state(struct drm_atomic_state *state, void *obj) +{ + struct drm_dp_mst_topology_mgr *mgr = obj; + struct drm_dp_mst_topology_state *new_mst_state; + + if (WARN_ON(!mgr->state)) + return NULL; + + new_mst_state = kmemdup(mgr->state, sizeof(*new_mst_state), GFP_KERNEL); + if (new_mst_state) + new_mst_state->state = state; + return new_mst_state; +} + +void drm_dp_mst_swap_state(void *obj, void **obj_state_ptr) +{ + struct drm_dp_mst_topology_mgr *mgr = obj; + struct drm_dp_mst_topology_state **topology_state_ptr; + + topology_state_ptr = (struct drm_dp_mst_topology_state **)obj_state_ptr; + + mgr->state->state = (*topology_state_ptr)->state; + swap(*topology_state_ptr, mgr->state); + mgr->state->state = NULL; +} + +void drm_dp_mst_destroy_state(void *obj_state) +{ + kfree(obj_state); +} + +static const struct drm_private_state_funcs mst_state_funcs = { + .duplicate_state = drm_dp_mst_duplicate_state, + .swap_state = drm_dp_mst_swap_state, + .destroy_state = drm_dp_mst_destroy_state, +}; + +/** + * drm_atomic_get_mst_topology_state: get MST topology state + * + * @state: global atomic state + * @mgr: MST topology manager, also the private object in this case + * + * This function wraps drm_atomic_get_priv_obj_state() passing in the MST atomic + * state vtable so that the private object state returned is that of a MST + * topology object. Also, drm_atomic_get_private_obj_state() expects the caller + * to care of the locking, so warn if don't hold the connection_mutex. + * + * RETURNS: + * + * The MST topology state or error pointer. + */ +struct drm_dp_mst_topology_state *drm_atomic_get_mst_topology_state(struct drm_atomic_state *state, + struct drm_dp_mst_topology_mgr *mgr) +{ + struct drm_device *dev = mgr->dev; + + WARN_ON(!drm_modeset_is_locked(&dev->mode_config.connection_mutex)); + return drm_atomic_get_private_obj_state(state, mgr, + &mst_state_funcs); +} +EXPORT_SYMBOL(drm_atomic_get_mst_topology_state); + /** * drm_dp_mst_topology_mgr_init - initialise a topology manager * @mgr: manager struct to initialise @@ -2980,6 +3118,15 @@ int drm_dp_mst_topology_mgr_init(struct drm_dp_mst_topology_mgr *mgr, if (test_calc_pbn_mode() < 0) DRM_ERROR("MST PBN self-test failed\n"); + mgr->state = kzalloc(sizeof(*mgr->state), GFP_KERNEL); + if (mgr->state == NULL) + return -ENOMEM; + mgr->state->mgr = mgr; + + /* max. time slots - one slot for MTP header */ + mgr->state->avail_slots = 63; + mgr->funcs = &mst_state_funcs; + return 0; } EXPORT_SYMBOL(drm_dp_mst_topology_mgr_init); @@ -3000,6 +3147,9 @@ void drm_dp_mst_topology_mgr_destroy(struct drm_dp_mst_topology_mgr *mgr) mutex_unlock(&mgr->payload_lock); mgr->dev = NULL; mgr->aux = NULL; + kfree(mgr->state); + mgr->state = NULL; + mgr->funcs = NULL; } EXPORT_SYMBOL(drm_dp_mst_topology_mgr_destroy); diff --git a/drivers/gpu/drm/drm_edid.c b/drivers/gpu/drm/drm_edid.c index fad3d44e4642..2e55599816aa 100644 --- a/drivers/gpu/drm/drm_edid.c +++ b/drivers/gpu/drm/drm_edid.c @@ -80,6 +80,8 @@ #define EDID_QUIRK_FORCE_12BPC (1 << 9) /* Force 6bpc */ #define EDID_QUIRK_FORCE_6BPC (1 << 10) +/* Force 10bpc */ +#define EDID_QUIRK_FORCE_10BPC (1 << 11) struct detailed_mode_closure { struct drm_connector *connector; @@ -122,6 +124,9 @@ static const struct edid_quirk { { "FCM", 13600, EDID_QUIRK_PREFER_LARGE_75 | EDID_QUIRK_DETAILED_IN_CM }, + /* LGD panel of HP zBook 17 G2, eDP 10 bpc, but reports unknown bpc */ + { "LGD", 764, EDID_QUIRK_FORCE_10BPC }, + /* LG Philips LCD LP154W01-A5 */ { "LPL", 0, EDID_QUIRK_DETAILED_USE_MAXIMUM_SIZE }, { "LPL", 0x2a00, EDID_QUIRK_DETAILED_USE_MAXIMUM_SIZE }, @@ -4244,6 +4249,9 @@ int drm_add_edid_modes(struct drm_connector *connector, struct edid *edid) if (quirks & EDID_QUIRK_FORCE_8BPC) connector->display_info.bpc = 8; + if (quirks & EDID_QUIRK_FORCE_10BPC) + connector->display_info.bpc = 10; + if (quirks & EDID_QUIRK_FORCE_12BPC) connector->display_info.bpc = 12; diff --git a/drivers/gpu/drm/drm_fb_cma_helper.c b/drivers/gpu/drm/drm_fb_cma_helper.c index 50abd1faf38f..53f9bdf470d7 100644 --- a/drivers/gpu/drm/drm_fb_cma_helper.c +++ b/drivers/gpu/drm/drm_fb_cma_helper.c @@ -189,7 +189,7 @@ struct drm_framebuffer *drm_fb_cma_create_with_funcs(struct drm_device *dev, obj = drm_gem_object_lookup(file_priv, mode_cmd->handles[i]); if (!obj) { dev_err(dev->dev, "Failed to lookup GEM object\n"); - ret = -ENXIO; + ret = -ENOENT; goto err_gem_object_put; } @@ -260,6 +260,33 @@ struct drm_gem_cma_object *drm_fb_cma_get_gem_obj(struct drm_framebuffer *fb, EXPORT_SYMBOL_GPL(drm_fb_cma_get_gem_obj); /** + * drm_fb_cma_get_gem_addr() - Get physical address for framebuffer + * @fb: The framebuffer + * @state: Which state of drm plane + * @plane: Which plane + * Return the CMA GEM address for given framebuffer. + * + * This function will usually be called from the PLANE callback functions. + */ +dma_addr_t drm_fb_cma_get_gem_addr(struct drm_framebuffer *fb, + struct drm_plane_state *state, + unsigned int plane) +{ + struct drm_fb_cma *fb_cma = to_fb_cma(fb); + dma_addr_t paddr; + + if (plane >= 4) + return 0; + + paddr = fb_cma->obj[plane]->paddr + fb->offsets[plane]; + paddr += fb->format->cpp[plane] * (state->src_x >> 16); + paddr += fb->pitches[plane] * (state->src_y >> 16); + + return paddr; +} +EXPORT_SYMBOL_GPL(drm_fb_cma_get_gem_addr); + +/** * drm_fb_cma_prepare_fb() - Prepare CMA framebuffer * @plane: Which plane * @state: Plane state attach fence to diff --git a/drivers/gpu/drm/drm_fb_helper.c b/drivers/gpu/drm/drm_fb_helper.c index a0ea3241c651..574af01d3ce9 100644 --- a/drivers/gpu/drm/drm_fb_helper.c +++ b/drivers/gpu/drm/drm_fb_helper.c @@ -378,7 +378,7 @@ retry: goto fail; } - plane_state->rotation = DRM_ROTATE_0; + plane_state->rotation = DRM_MODE_ROTATE_0; plane->old_fb = plane->fb; plane_mask |= 1 << drm_plane_index(plane); @@ -431,7 +431,7 @@ static int restore_fbdev_mode_legacy(struct drm_fb_helper *fb_helper) if (plane->rotation_property) drm_mode_plane_set_obj_prop(plane, plane->rotation_property, - DRM_ROTATE_0); + DRM_MODE_ROTATE_0); } for (i = 0; i < fb_helper->crtc_count; i++) { @@ -2446,7 +2446,7 @@ EXPORT_SYMBOL(drm_fb_helper_hotplug_event); int __init drm_fb_helper_modinit(void) { #if defined(CONFIG_FRAMEBUFFER_CONSOLE_MODULE) && !defined(CONFIG_EXPERT) - const char *name = "fbcon"; + const char name[] = "fbcon"; struct module *fbcon; mutex_lock(&module_mutex); diff --git a/drivers/gpu/drm/drm_file.c b/drivers/gpu/drm/drm_file.c index 3783b659cd38..caad93dab54b 100644 --- a/drivers/gpu/drm/drm_file.c +++ b/drivers/gpu/drm/drm_file.c @@ -351,9 +351,8 @@ void drm_lastclose(struct drm_device * dev) * * This function must be used by drivers as their &file_operations.release * method. It frees any resources associated with the open file, and calls the - * &drm_driver.preclose and &drm_driver.lastclose driver callbacks. If this is - * the last open file for the DRM device also proceeds to call the - * &drm_driver.lastclose driver callback. + * &drm_driver.postclose driver callback. If this is the last open file for the + * DRM device also proceeds to call the &drm_driver.lastclose driver callback. * * RETURNS: * @@ -373,7 +372,8 @@ int drm_release(struct inode *inode, struct file *filp) list_del(&file_priv->lhead); mutex_unlock(&dev->filelist_mutex); - if (dev->driver->preclose) + if (drm_core_check_feature(dev, DRIVER_LEGACY) && + dev->driver->preclose) dev->driver->preclose(dev, file_priv); /* ======================================================== diff --git a/drivers/gpu/drm/drm_gem.c b/drivers/gpu/drm/drm_gem.c index b1e28c944637..8dc11064253d 100644 --- a/drivers/gpu/drm/drm_gem.c +++ b/drivers/gpu/drm/drm_gem.c @@ -521,7 +521,7 @@ struct page **drm_gem_get_pages(struct drm_gem_object *obj) npages = obj->size >> PAGE_SHIFT; - pages = drm_malloc_ab(npages, sizeof(struct page *)); + pages = kvmalloc_array(npages, sizeof(struct page *), GFP_KERNEL); if (pages == NULL) return ERR_PTR(-ENOMEM); @@ -546,7 +546,7 @@ fail: while (i--) put_page(pages[i]); - drm_free_large(pages); + kvfree(pages); return ERR_CAST(p); } EXPORT_SYMBOL(drm_gem_get_pages); @@ -582,7 +582,7 @@ void drm_gem_put_pages(struct drm_gem_object *obj, struct page **pages, put_page(pages[i]); } - drm_free_large(pages); + kvfree(pages); } EXPORT_SYMBOL(drm_gem_put_pages); diff --git a/drivers/gpu/drm/drm_irq.c b/drivers/gpu/drm/drm_irq.c index 8c866cac62dd..c7debaad67f8 100644 --- a/drivers/gpu/drm/drm_irq.c +++ b/drivers/gpu/drm/drm_irq.c @@ -54,7 +54,7 @@ static bool drm_get_last_vbltimestamp(struct drm_device *dev, unsigned int pipe, - struct timeval *tvblank, unsigned flags); + struct timeval *tvblank, bool in_vblank_irq); static unsigned int drm_timestamp_precision = 20; /* Default to 20 usecs. */ @@ -138,7 +138,7 @@ static void drm_reset_vblank_timestamp(struct drm_device *dev, unsigned int pipe */ do { cur_vblank = __get_vblank_counter(dev, pipe); - rc = drm_get_last_vbltimestamp(dev, pipe, &t_vblank, 0); + rc = drm_get_last_vbltimestamp(dev, pipe, &t_vblank, false); } while (cur_vblank != __get_vblank_counter(dev, pipe) && --count > 0); /* @@ -171,7 +171,7 @@ static void drm_reset_vblank_timestamp(struct drm_device *dev, unsigned int pipe * device vblank fields. */ static void drm_update_vblank_count(struct drm_device *dev, unsigned int pipe, - unsigned long flags) + bool in_vblank_irq) { struct drm_vblank_crtc *vblank = &dev->vblank[pipe]; u32 cur_vblank, diff; @@ -194,7 +194,7 @@ static void drm_update_vblank_count(struct drm_device *dev, unsigned int pipe, */ do { cur_vblank = __get_vblank_counter(dev, pipe); - rc = drm_get_last_vbltimestamp(dev, pipe, &t_vblank, flags); + rc = drm_get_last_vbltimestamp(dev, pipe, &t_vblank, in_vblank_irq); } while (cur_vblank != __get_vblank_counter(dev, pipe) && --count > 0); if (dev->max_vblank_count != 0) { @@ -214,13 +214,13 @@ static void drm_update_vblank_count(struct drm_device *dev, unsigned int pipe, */ diff = DIV_ROUND_CLOSEST_ULL(diff_ns, framedur_ns); - if (diff == 0 && flags & DRM_CALLED_FROM_VBLIRQ) + if (diff == 0 && in_vblank_irq) DRM_DEBUG_VBL("crtc %u: Redundant vblirq ignored." " diff_ns = %lld, framedur_ns = %d)\n", pipe, (long long) diff_ns, framedur_ns); } else { /* some kind of default for drivers w/o accurate vbl timestamping */ - diff = (flags & DRM_CALLED_FROM_VBLIRQ) != 0; + diff = in_vblank_irq ? 1 : 0; } /* @@ -253,7 +253,7 @@ static void drm_update_vblank_count(struct drm_device *dev, unsigned int pipe, * Otherwise reinitialize delayed at next vblank interrupt and assign 0 * for now, to mark the vblanktimestamp as invalid. */ - if (!rc && (flags & DRM_CALLED_FROM_VBLIRQ) == 0) + if (!rc && in_vblank_irq) t_vblank = (struct timeval) {0, 0}; store_vblank(dev, pipe, diff, &t_vblank, cur_vblank); @@ -291,7 +291,7 @@ u32 drm_accurate_vblank_count(struct drm_crtc *crtc) spin_lock_irqsave(&dev->vblank_time_lock, flags); - drm_update_vblank_count(dev, pipe, 0); + drm_update_vblank_count(dev, pipe, false); vblank = drm_vblank_count(dev, pipe); spin_unlock_irqrestore(&dev->vblank_time_lock, flags); @@ -349,7 +349,7 @@ static void vblank_disable_and_save(struct drm_device *dev, unsigned int pipe) * this time. This makes the count account for the entire time * between drm_crtc_vblank_on() and drm_crtc_vblank_off(). */ - drm_update_vblank_count(dev, pipe, 0); + drm_update_vblank_count(dev, pipe, false); spin_unlock_irqrestore(&dev->vblank_time_lock, irqflags); } @@ -684,6 +684,7 @@ void drm_calc_timestamping_constants(struct drm_crtc *crtc, vblank->linedur_ns = linedur_ns; vblank->framedur_ns = framedur_ns; + vblank->hwmode = *mode; DRM_DEBUG("crtc %u: hwmode: htotal %d, vtotal %d, vdisplay %d\n", crtc->base.id, mode->crtc_htotal, @@ -700,10 +701,10 @@ EXPORT_SYMBOL(drm_calc_timestamping_constants); * @max_error: Desired maximum allowable error in timestamps (nanosecs) * On return contains true maximum error of timestamp * @vblank_time: Pointer to struct timeval which should receive the timestamp - * @flags: Flags to pass to driver: - * 0 = Default, - * DRM_CALLED_FROM_VBLIRQ = If function is called from vbl IRQ handler - * @mode: mode which defines the scanout timings + * @in_vblank_irq: + * True when called from drm_crtc_handle_vblank(). Some drivers + * need to apply some workarounds for gpu-specific vblank irq quirks + * if flag is set. * * Implements calculation of exact vblank timestamps from given drm_display_mode * timings and current video scanout position of a CRTC. This can be called from @@ -723,52 +724,62 @@ EXPORT_SYMBOL(drm_calc_timestamping_constants); * returns as no operation if a doublescan or interlaced video mode is * active. Higher level code is expected to handle this. * - * Returns: - * Negative value on error, failure or if not supported in current - * video mode: - * - * -EINVAL Invalid CRTC. - * -EAGAIN Temporary unavailable, e.g., called before initial modeset. - * -ENOTSUPP Function not supported in current display mode. - * -EIO Failed, e.g., due to failed scanout position query. + * This function can be used to implement the &drm_driver.get_vblank_timestamp + * directly, if the driver implements the &drm_driver.get_scanout_position hook. * - * Returns or'ed positive status flags on success: + * Note that atomic drivers must call drm_calc_timestamping_constants() before + * enabling a CRTC. The atomic helpers already take care of that in + * drm_atomic_helper_update_legacy_modeset_state(). * - * DRM_VBLANKTIME_SCANOUTPOS_METHOD - Signal this method used for timestamping. - * DRM_VBLANKTIME_INVBL - Timestamp taken while scanout was in vblank interval. + * Returns: * + * Returns true on success, and false on failure, i.e. when no accurate + * timestamp could be acquired. */ -int drm_calc_vbltimestamp_from_scanoutpos(struct drm_device *dev, - unsigned int pipe, - int *max_error, - struct timeval *vblank_time, - unsigned flags, - const struct drm_display_mode *mode) +bool drm_calc_vbltimestamp_from_scanoutpos(struct drm_device *dev, + unsigned int pipe, + int *max_error, + struct timeval *vblank_time, + bool in_vblank_irq) { struct timeval tv_etime; ktime_t stime, etime; - unsigned int vbl_status; - int ret = DRM_VBLANKTIME_SCANOUTPOS_METHOD; + bool vbl_status; + struct drm_crtc *crtc; + const struct drm_display_mode *mode; + struct drm_vblank_crtc *vblank = &dev->vblank[pipe]; int vpos, hpos, i; int delta_ns, duration_ns; - if (pipe >= dev->num_crtcs) { + if (!drm_core_check_feature(dev, DRIVER_MODESET)) + return false; + + crtc = drm_crtc_from_index(dev, pipe); + + if (pipe >= dev->num_crtcs || !crtc) { DRM_ERROR("Invalid crtc %u\n", pipe); - return -EINVAL; + return false; } /* Scanout position query not supported? Should not happen. */ if (!dev->driver->get_scanout_position) { DRM_ERROR("Called from driver w/o get_scanout_position()!?\n"); - return -EIO; + return false; } + if (drm_drv_uses_atomic_modeset(dev)) + mode = &vblank->hwmode; + else + mode = &crtc->hwmode; + /* If mode timing undefined, just return as no-op: * Happens during initial modesetting of a crtc. */ if (mode->crtc_clock == 0) { DRM_DEBUG("crtc %u: Noop due to uninitialized mode.\n", pipe); - return -EAGAIN; + WARN_ON_ONCE(drm_drv_uses_atomic_modeset(dev)); + + return false; } /* Get current scanout position with system timestamp. @@ -783,16 +794,17 @@ int drm_calc_vbltimestamp_from_scanoutpos(struct drm_device *dev, * Get vertical and horizontal scanout position vpos, hpos, * and bounding timestamps stime, etime, pre/post query. */ - vbl_status = dev->driver->get_scanout_position(dev, pipe, flags, + vbl_status = dev->driver->get_scanout_position(dev, pipe, + in_vblank_irq, &vpos, &hpos, &stime, &etime, mode); /* Return as no-op if scanout query unsupported or failed. */ - if (!(vbl_status & DRM_SCANOUTPOS_VALID)) { - DRM_DEBUG("crtc %u : scanoutpos query failed [0x%x].\n", - pipe, vbl_status); - return -EIO; + if (!vbl_status) { + DRM_DEBUG("crtc %u : scanoutpos query failed.\n", + pipe); + return false; } /* Compute uncertainty in timestamp of scanout position query. */ @@ -830,13 +842,13 @@ int drm_calc_vbltimestamp_from_scanoutpos(struct drm_device *dev, etime = ktime_sub_ns(etime, delta_ns); *vblank_time = ktime_to_timeval(etime); - DRM_DEBUG_VBL("crtc %u : v 0x%x p(%d,%d)@ %ld.%ld -> %ld.%ld [e %d us, %d rep]\n", - pipe, vbl_status, hpos, vpos, + DRM_DEBUG_VBL("crtc %u : v p(%d,%d)@ %ld.%ld -> %ld.%ld [e %d us, %d rep]\n", + pipe, hpos, vpos, (long)tv_etime.tv_sec, (long)tv_etime.tv_usec, (long)vblank_time->tv_sec, (long)vblank_time->tv_usec, duration_ns/1000, i); - return ret; + return true; } EXPORT_SYMBOL(drm_calc_vbltimestamp_from_scanoutpos); @@ -854,9 +866,10 @@ static struct timeval get_drm_timestamp(void) * @dev: DRM device * @pipe: index of CRTC whose vblank timestamp to retrieve * @tvblank: Pointer to target struct timeval which should receive the timestamp - * @flags: Flags to pass to driver: - * 0 = Default, - * DRM_CALLED_FROM_VBLIRQ = If function is called from vbl IRQ handler + * @in_vblank_irq: + * True when called from drm_crtc_handle_vblank(). Some drivers + * need to apply some workarounds for gpu-specific vblank irq quirks + * if flag is set. * * Fetches the system timestamp corresponding to the time of the most recent * vblank interval on specified CRTC. May call into kms-driver to @@ -870,27 +883,25 @@ static struct timeval get_drm_timestamp(void) */ static bool drm_get_last_vbltimestamp(struct drm_device *dev, unsigned int pipe, - struct timeval *tvblank, unsigned flags) + struct timeval *tvblank, bool in_vblank_irq) { - int ret; + bool ret = false; /* Define requested maximum error on timestamps (nanoseconds). */ int max_error = (int) drm_timestamp_precision * 1000; /* Query driver if possible and precision timestamping enabled. */ - if (dev->driver->get_vblank_timestamp && (max_error > 0)) { + if (dev->driver->get_vblank_timestamp && (max_error > 0)) ret = dev->driver->get_vblank_timestamp(dev, pipe, &max_error, - tvblank, flags); - if (ret > 0) - return true; - } + tvblank, in_vblank_irq); /* GPU high precision timestamp query unsupported or failed. * Return current monotonic/gettimeofday timestamp as best estimate. */ - *tvblank = get_drm_timestamp(); + if (!ret) + *tvblank = get_drm_timestamp(); - return false; + return ret; } /** @@ -1329,6 +1340,10 @@ void drm_crtc_vblank_off(struct drm_crtc *crtc) send_vblank_event(dev, e, seq, &now); } spin_unlock_irqrestore(&dev->event_lock, irqflags); + + /* Will be reset by the modeset helpers when re-enabling the crtc by + * calling drm_calc_timestamping_constants(). */ + vblank->hwmode.crtc_clock = 0; } EXPORT_SYMBOL(drm_crtc_vblank_off); @@ -1760,7 +1775,7 @@ bool drm_handle_vblank(struct drm_device *dev, unsigned int pipe) return false; } - drm_update_vblank_count(dev, pipe, DRM_CALLED_FROM_VBLIRQ); + drm_update_vblank_count(dev, pipe, true); spin_unlock(&dev->vblank_time_lock); diff --git a/drivers/gpu/drm/drm_plane_helper.c b/drivers/gpu/drm/drm_plane_helper.c index b84a295230fc..06aee1741e96 100644 --- a/drivers/gpu/drm/drm_plane_helper.c +++ b/drivers/gpu/drm/drm_plane_helper.c @@ -336,7 +336,7 @@ int drm_primary_helper_update(struct drm_plane *plane, struct drm_crtc *crtc, ret = drm_plane_helper_check_update(plane, crtc, fb, &src, &dest, &clip, - DRM_ROTATE_0, + DRM_MODE_ROTATE_0, DRM_PLANE_HELPER_NO_SCALING, DRM_PLANE_HELPER_NO_SCALING, false, false, &visible); @@ -381,6 +381,7 @@ EXPORT_SYMBOL(drm_primary_helper_update); /** * drm_primary_helper_disable() - Helper for primary plane disable * @plane: plane to disable + * @ctx: lock acquire context, not used here * * Provides a default plane disable handler for primary planes. This is handler * is called in response to a userspace SetPlane operation on the plane with a @@ -510,12 +511,10 @@ int drm_plane_helper_commit(struct drm_plane *plane, if (plane_funcs->cleanup_fb) plane_funcs->cleanup_fb(plane, plane_state); out: - if (plane_state) { - if (plane->funcs->atomic_destroy_state) - plane->funcs->atomic_destroy_state(plane, plane_state); - else - drm_atomic_helper_plane_destroy_state(plane, plane_state); - } + if (plane->funcs->atomic_destroy_state) + plane->funcs->atomic_destroy_state(plane, plane_state); + else + drm_atomic_helper_plane_destroy_state(plane, plane_state); return ret; } diff --git a/drivers/gpu/drm/drm_prime.c b/drivers/gpu/drm/drm_prime.c index 954eb848b5e2..22408badc617 100644 --- a/drivers/gpu/drm/drm_prime.c +++ b/drivers/gpu/drm/drm_prime.c @@ -595,15 +595,18 @@ out_unlock: EXPORT_SYMBOL(drm_gem_prime_handle_to_fd); /** - * drm_gem_prime_import - helper library implementation of the import callback + * drm_gem_prime_import_dev - core implementation of the import callback * @dev: drm_device to import into * @dma_buf: dma-buf object to import + * @attach_dev: struct device to dma_buf attach * - * This is the implementation of the gem_prime_import functions for GEM drivers - * using the PRIME helpers. + * This is the core of drm_gem_prime_import. It's designed to be called by + * drivers who want to use a different device structure than dev->dev for + * attaching via dma_buf. */ -struct drm_gem_object *drm_gem_prime_import(struct drm_device *dev, - struct dma_buf *dma_buf) +struct drm_gem_object *drm_gem_prime_import_dev(struct drm_device *dev, + struct dma_buf *dma_buf, + struct device *attach_dev) { struct dma_buf_attachment *attach; struct sg_table *sgt; @@ -625,7 +628,7 @@ struct drm_gem_object *drm_gem_prime_import(struct drm_device *dev, if (!dev->driver->gem_prime_import_sg_table) return ERR_PTR(-EINVAL); - attach = dma_buf_attach(dma_buf, dev->dev); + attach = dma_buf_attach(dma_buf, attach_dev); if (IS_ERR(attach)) return ERR_CAST(attach); @@ -655,6 +658,21 @@ fail_detach: return ERR_PTR(ret); } +EXPORT_SYMBOL(drm_gem_prime_import_dev); + +/** + * drm_gem_prime_import - helper library implementation of the import callback + * @dev: drm_device to import into + * @dma_buf: dma-buf object to import + * + * This is the implementation of the gem_prime_import functions for GEM drivers + * using the PRIME helpers. + */ +struct drm_gem_object *drm_gem_prime_import(struct drm_device *dev, + struct dma_buf *dma_buf) +{ + return drm_gem_prime_import_dev(dev, dma_buf, dev->dev); +} EXPORT_SYMBOL(drm_gem_prime_import); /** diff --git a/drivers/gpu/drm/drm_rect.c b/drivers/gpu/drm/drm_rect.c index bc5575960ebc..9817c1445ba9 100644 --- a/drivers/gpu/drm/drm_rect.c +++ b/drivers/gpu/drm/drm_rect.c @@ -310,38 +310,38 @@ void drm_rect_rotate(struct drm_rect *r, { struct drm_rect tmp; - if (rotation & (DRM_REFLECT_X | DRM_REFLECT_Y)) { + if (rotation & (DRM_MODE_REFLECT_X | DRM_MODE_REFLECT_Y)) { tmp = *r; - if (rotation & DRM_REFLECT_X) { + if (rotation & DRM_MODE_REFLECT_X) { r->x1 = width - tmp.x2; r->x2 = width - tmp.x1; } - if (rotation & DRM_REFLECT_Y) { + if (rotation & DRM_MODE_REFLECT_Y) { r->y1 = height - tmp.y2; r->y2 = height - tmp.y1; } } - switch (rotation & DRM_ROTATE_MASK) { - case DRM_ROTATE_0: + switch (rotation & DRM_MODE_ROTATE_MASK) { + case DRM_MODE_ROTATE_0: break; - case DRM_ROTATE_90: + case DRM_MODE_ROTATE_90: tmp = *r; r->x1 = tmp.y1; r->x2 = tmp.y2; r->y1 = width - tmp.x2; r->y2 = width - tmp.x1; break; - case DRM_ROTATE_180: + case DRM_MODE_ROTATE_180: tmp = *r; r->x1 = width - tmp.x2; r->x2 = width - tmp.x1; r->y1 = height - tmp.y2; r->y2 = height - tmp.y1; break; - case DRM_ROTATE_270: + case DRM_MODE_ROTATE_270: tmp = *r; r->x1 = height - tmp.y2; r->x2 = height - tmp.y1; @@ -373,8 +373,8 @@ EXPORT_SYMBOL(drm_rect_rotate); * them when doing a rotatation and its inverse. * That is, if you do :: * - * drm_rotate(&r, width, height, rotation); - * drm_rotate_inv(&r, width, height, rotation); + * DRM_MODE_PROP_ROTATE(&r, width, height, rotation); + * DRM_MODE_ROTATE_inv(&r, width, height, rotation); * * you will always get back the original rectangle. */ @@ -384,24 +384,24 @@ void drm_rect_rotate_inv(struct drm_rect *r, { struct drm_rect tmp; - switch (rotation & DRM_ROTATE_MASK) { - case DRM_ROTATE_0: + switch (rotation & DRM_MODE_ROTATE_MASK) { + case DRM_MODE_ROTATE_0: break; - case DRM_ROTATE_90: + case DRM_MODE_ROTATE_90: tmp = *r; r->x1 = width - tmp.y2; r->x2 = width - tmp.y1; r->y1 = tmp.x1; r->y2 = tmp.x2; break; - case DRM_ROTATE_180: + case DRM_MODE_ROTATE_180: tmp = *r; r->x1 = width - tmp.x2; r->x2 = width - tmp.x1; r->y1 = height - tmp.y2; r->y2 = height - tmp.y1; break; - case DRM_ROTATE_270: + case DRM_MODE_ROTATE_270: tmp = *r; r->x1 = tmp.y1; r->x2 = tmp.y2; @@ -412,15 +412,15 @@ void drm_rect_rotate_inv(struct drm_rect *r, break; } - if (rotation & (DRM_REFLECT_X | DRM_REFLECT_Y)) { + if (rotation & (DRM_MODE_REFLECT_X | DRM_MODE_REFLECT_Y)) { tmp = *r; - if (rotation & DRM_REFLECT_X) { + if (rotation & DRM_MODE_REFLECT_X) { r->x1 = width - tmp.x2; r->x2 = width - tmp.x1; } - if (rotation & DRM_REFLECT_Y) { + if (rotation & DRM_MODE_REFLECT_Y) { r->y1 = height - tmp.y2; r->y2 = height - tmp.y1; } diff --git a/drivers/gpu/drm/etnaviv/etnaviv_dump.c b/drivers/gpu/drm/etnaviv/etnaviv_dump.c index d019b5e311cc..2d955d7d7b6d 100644 --- a/drivers/gpu/drm/etnaviv/etnaviv_dump.c +++ b/drivers/gpu/drm/etnaviv/etnaviv_dump.c @@ -161,8 +161,8 @@ void etnaviv_core_dump(struct etnaviv_gpu *gpu) file_size += sizeof(*iter.hdr) * n_obj; /* Allocate the file in vmalloc memory, it's likely to be big */ - iter.start = __vmalloc(file_size, GFP_KERNEL | __GFP_HIGHMEM | - __GFP_NOWARN | __GFP_NORETRY, PAGE_KERNEL); + iter.start = __vmalloc(file_size, GFP_KERNEL | __GFP_NOWARN | __GFP_NORETRY, + PAGE_KERNEL); if (!iter.start) { dev_warn(gpu->dev, "failed to allocate devcoredump file\n"); return; diff --git a/drivers/gpu/drm/etnaviv/etnaviv_gem.c b/drivers/gpu/drm/etnaviv/etnaviv_gem.c index fd56f92f3469..d6fb724fc3cc 100644 --- a/drivers/gpu/drm/etnaviv/etnaviv_gem.c +++ b/drivers/gpu/drm/etnaviv/etnaviv_gem.c @@ -748,7 +748,7 @@ static struct page **etnaviv_gem_userptr_do_get_pages( uintptr_t ptr; unsigned int flags = 0; - pvec = drm_malloc_ab(npages, sizeof(struct page *)); + pvec = kvmalloc_array(npages, sizeof(struct page *), GFP_KERNEL); if (!pvec) return ERR_PTR(-ENOMEM); @@ -772,7 +772,7 @@ static struct page **etnaviv_gem_userptr_do_get_pages( if (ret < 0) { release_pages(pvec, pinned, 0); - drm_free_large(pvec); + kvfree(pvec); return ERR_PTR(ret); } @@ -823,7 +823,7 @@ static int etnaviv_gem_userptr_get_pages(struct etnaviv_gem_object *etnaviv_obj) mm = get_task_mm(etnaviv_obj->userptr.task); pinned = 0; if (mm == current->mm) { - pvec = drm_malloc_ab(npages, sizeof(struct page *)); + pvec = kvmalloc_array(npages, sizeof(struct page *), GFP_KERNEL); if (!pvec) { mmput(mm); return -ENOMEM; @@ -832,7 +832,7 @@ static int etnaviv_gem_userptr_get_pages(struct etnaviv_gem_object *etnaviv_obj) pinned = __get_user_pages_fast(etnaviv_obj->userptr.ptr, npages, !etnaviv_obj->userptr.ro, pvec); if (pinned < 0) { - drm_free_large(pvec); + kvfree(pvec); mmput(mm); return pinned; } @@ -845,7 +845,7 @@ static int etnaviv_gem_userptr_get_pages(struct etnaviv_gem_object *etnaviv_obj) } release_pages(pvec, pinned, 0); - drm_free_large(pvec); + kvfree(pvec); work = kmalloc(sizeof(*work), GFP_KERNEL); if (!work) { @@ -879,7 +879,7 @@ static void etnaviv_gem_userptr_release(struct etnaviv_gem_object *etnaviv_obj) int npages = etnaviv_obj->base.size >> PAGE_SHIFT; release_pages(etnaviv_obj->pages, npages, 0); - drm_free_large(etnaviv_obj->pages); + kvfree(etnaviv_obj->pages); } put_task_struct(etnaviv_obj->userptr.task); } diff --git a/drivers/gpu/drm/etnaviv/etnaviv_gem_prime.c b/drivers/gpu/drm/etnaviv/etnaviv_gem_prime.c index 62b47972a52e..367bf952f61a 100644 --- a/drivers/gpu/drm/etnaviv/etnaviv_gem_prime.c +++ b/drivers/gpu/drm/etnaviv/etnaviv_gem_prime.c @@ -87,7 +87,7 @@ static void etnaviv_gem_prime_release(struct etnaviv_gem_object *etnaviv_obj) * ours, just free the array we allocated: */ if (etnaviv_obj->pages) - drm_free_large(etnaviv_obj->pages); + kvfree(etnaviv_obj->pages); drm_prime_gem_destroy(&etnaviv_obj->base, etnaviv_obj->sgt); } @@ -128,7 +128,7 @@ struct drm_gem_object *etnaviv_gem_prime_import_sg_table(struct drm_device *dev, npages = size / PAGE_SIZE; etnaviv_obj->sgt = sgt; - etnaviv_obj->pages = drm_malloc_ab(npages, sizeof(struct page *)); + etnaviv_obj->pages = kvmalloc_array(npages, sizeof(struct page *), GFP_KERNEL); if (!etnaviv_obj->pages) { ret = -ENOMEM; goto fail; diff --git a/drivers/gpu/drm/etnaviv/etnaviv_gem_submit.c b/drivers/gpu/drm/etnaviv/etnaviv_gem_submit.c index e1909429837e..a13930e1d8c9 100644 --- a/drivers/gpu/drm/etnaviv/etnaviv_gem_submit.c +++ b/drivers/gpu/drm/etnaviv/etnaviv_gem_submit.c @@ -343,9 +343,9 @@ int etnaviv_ioctl_gem_submit(struct drm_device *dev, void *data, * Copy the command submission and bo array to kernel space in * one go, and do this outside of any locks. */ - bos = drm_malloc_ab(args->nr_bos, sizeof(*bos)); - relocs = drm_malloc_ab(args->nr_relocs, sizeof(*relocs)); - stream = drm_malloc_ab(1, args->stream_size); + bos = kvmalloc_array(args->nr_bos, sizeof(*bos), GFP_KERNEL); + relocs = kvmalloc_array(args->nr_relocs, sizeof(*relocs), GFP_KERNEL); + stream = kvmalloc_array(1, args->stream_size, GFP_KERNEL); cmdbuf = etnaviv_cmdbuf_new(gpu->cmdbuf_suballoc, ALIGN(args->stream_size, 8) + 8, args->nr_bos); @@ -487,11 +487,11 @@ err_submit_cmds: if (cmdbuf) etnaviv_cmdbuf_free(cmdbuf); if (stream) - drm_free_large(stream); + kvfree(stream); if (bos) - drm_free_large(bos); + kvfree(bos); if (relocs) - drm_free_large(relocs); + kvfree(relocs); return ret; } diff --git a/drivers/gpu/drm/exynos/exynos_drm_drv.c b/drivers/gpu/drm/exynos/exynos_drm_drv.c index 09d3c4c3c858..50294a7bd29d 100644 --- a/drivers/gpu/drm/exynos/exynos_drm_drv.c +++ b/drivers/gpu/drm/exynos/exynos_drm_drv.c @@ -82,14 +82,9 @@ err_file_priv_free: return ret; } -static void exynos_drm_preclose(struct drm_device *dev, - struct drm_file *file) -{ - exynos_drm_subdrv_close(dev, file); -} - static void exynos_drm_postclose(struct drm_device *dev, struct drm_file *file) { + exynos_drm_subdrv_close(dev, file); kfree(file->driver_priv); file->driver_priv = NULL; } @@ -145,7 +140,6 @@ static struct drm_driver exynos_drm_driver = { .driver_features = DRIVER_MODESET | DRIVER_GEM | DRIVER_PRIME | DRIVER_ATOMIC | DRIVER_RENDER, .open = exynos_drm_open, - .preclose = exynos_drm_preclose, .lastclose = exynos_drm_lastclose, .postclose = exynos_drm_postclose, .gem_free_object_unlocked = exynos_drm_gem_free_object, diff --git a/drivers/gpu/drm/exynos/exynos_drm_gem.c b/drivers/gpu/drm/exynos/exynos_drm_gem.c index 55a1579d11b3..c23479be4850 100644 --- a/drivers/gpu/drm/exynos/exynos_drm_gem.c +++ b/drivers/gpu/drm/exynos/exynos_drm_gem.c @@ -59,7 +59,8 @@ static int exynos_drm_alloc_buf(struct exynos_drm_gem *exynos_gem) nr_pages = exynos_gem->size >> PAGE_SHIFT; - exynos_gem->pages = drm_calloc_large(nr_pages, sizeof(struct page *)); + exynos_gem->pages = kvmalloc_array(nr_pages, sizeof(struct page *), + GFP_KERNEL | __GFP_ZERO); if (!exynos_gem->pages) { DRM_ERROR("failed to allocate pages.\n"); return -ENOMEM; @@ -101,7 +102,7 @@ err_dma_free: dma_free_attrs(to_dma_dev(dev), exynos_gem->size, exynos_gem->cookie, exynos_gem->dma_addr, exynos_gem->dma_attrs); err_free: - drm_free_large(exynos_gem->pages); + kvfree(exynos_gem->pages); return ret; } @@ -122,7 +123,7 @@ static void exynos_drm_free_buf(struct exynos_drm_gem *exynos_gem) (dma_addr_t)exynos_gem->dma_addr, exynos_gem->dma_attrs); - drm_free_large(exynos_gem->pages); + kvfree(exynos_gem->pages); } static int exynos_drm_gem_handle_create(struct drm_gem_object *obj, @@ -559,7 +560,7 @@ exynos_drm_gem_prime_import_sg_table(struct drm_device *dev, exynos_gem->dma_addr = sg_dma_address(sgt->sgl); npages = exynos_gem->size >> PAGE_SHIFT; - exynos_gem->pages = drm_malloc_ab(npages, sizeof(struct page *)); + exynos_gem->pages = kvmalloc_array(npages, sizeof(struct page *), GFP_KERNEL); if (!exynos_gem->pages) { ret = -ENOMEM; goto err; @@ -588,7 +589,7 @@ exynos_drm_gem_prime_import_sg_table(struct drm_device *dev, return &exynos_gem->base; err_free_large: - drm_free_large(exynos_gem->pages); + kvfree(exynos_gem->pages); err: drm_gem_object_release(&exynos_gem->base); kfree(exynos_gem); diff --git a/drivers/gpu/drm/exynos/exynos_hdmi.c b/drivers/gpu/drm/exynos/exynos_hdmi.c index 52438404c8c9..1ff6ab6371e8 100644 --- a/drivers/gpu/drm/exynos/exynos_hdmi.c +++ b/drivers/gpu/drm/exynos/exynos_hdmi.c @@ -43,6 +43,8 @@ #include <drm/exynos_drm.h> +#include <media/cec-notifier.h> + #include "exynos_drm_crtc.h" #define HOTPLUG_DEBOUNCE_MS 1100 @@ -118,6 +120,7 @@ struct hdmi_context { bool dvi_mode; struct delayed_work hotplug_work; struct drm_display_mode current_mode; + struct cec_notifier *notifier; const struct hdmi_driver_data *drv_data; void __iomem *regs; @@ -821,6 +824,7 @@ static enum drm_connector_status hdmi_detect(struct drm_connector *connector, if (gpiod_get_value(hdata->hpd_gpio)) return connector_status_connected; + cec_notifier_set_phys_addr(hdata->notifier, CEC_PHYS_ADDR_INVALID); return connector_status_disconnected; } @@ -859,6 +863,7 @@ static int hdmi_get_modes(struct drm_connector *connector) edid->width_cm, edid->height_cm); drm_mode_connector_update_edid_property(connector, edid); + cec_notifier_set_phys_addr_from_edid(hdata->notifier, edid); ret = drm_add_edid_modes(connector, edid); @@ -1501,6 +1506,7 @@ static void hdmi_disable(struct drm_encoder *encoder) if (funcs && funcs->disable) (*funcs->disable)(crtc); + cec_notifier_set_phys_addr(hdata->notifier, CEC_PHYS_ADDR_INVALID); cancel_delayed_work(&hdata->hotplug_work); hdmiphy_disable(hdata); @@ -1880,15 +1886,22 @@ static int hdmi_probe(struct platform_device *pdev) } } + hdata->notifier = cec_notifier_get(&pdev->dev); + if (hdata->notifier == NULL) { + ret = -ENOMEM; + goto err_hdmiphy; + } + pm_runtime_enable(dev); ret = component_add(&pdev->dev, &hdmi_component_ops); if (ret) - goto err_disable_pm_runtime; + goto err_notifier_put; return ret; -err_disable_pm_runtime: +err_notifier_put: + cec_notifier_put(hdata->notifier); pm_runtime_disable(dev); err_hdmiphy: @@ -1907,9 +1920,11 @@ static int hdmi_remove(struct platform_device *pdev) struct hdmi_context *hdata = platform_get_drvdata(pdev); cancel_delayed_work_sync(&hdata->hotplug_work); + cec_notifier_set_phys_addr(hdata->notifier, CEC_PHYS_ADDR_INVALID); component_del(&pdev->dev, &hdmi_component_ops); + cec_notifier_put(hdata->notifier); pm_runtime_disable(&pdev->dev); if (!IS_ERR(hdata->reg_hdmi_en)) diff --git a/drivers/gpu/drm/gma500/Makefile b/drivers/gpu/drm/gma500/Makefile index 190e55f2f891..c1c8dc18aa53 100644 --- a/drivers/gpu/drm/gma500/Makefile +++ b/drivers/gpu/drm/gma500/Makefile @@ -1,7 +1,6 @@ # # KMS driver for the GMA500 # -ccflags-y += -I$(srctree)/include/drm gma500_gfx-y += \ accel_2d.o \ diff --git a/drivers/gpu/drm/gma500/gtt.c b/drivers/gpu/drm/gma500/gtt.c index 3f4f424196b2..3949b0990916 100644 --- a/drivers/gpu/drm/gma500/gtt.c +++ b/drivers/gpu/drm/gma500/gtt.c @@ -21,6 +21,7 @@ #include <drm/drmP.h> #include <linux/shmem_fs.h> +#include <asm/set_memory.h> #include "psb_drv.h" #include "blitter.h" diff --git a/drivers/gpu/drm/gma500/mdfld_tpo_vid.c b/drivers/gpu/drm/gma500/mdfld_tpo_vid.c index d8d4170725b2..a9420bf9a419 100644 --- a/drivers/gpu/drm/gma500/mdfld_tpo_vid.c +++ b/drivers/gpu/drm/gma500/mdfld_tpo_vid.c @@ -30,55 +30,20 @@ static struct drm_display_mode *tpo_vid_get_config_mode(struct drm_device *dev) { struct drm_display_mode *mode; - struct drm_psb_private *dev_priv = dev->dev_private; - struct oaktrail_timing_info *ti = &dev_priv->gct_data.DTD; - bool use_gct = false; mode = kzalloc(sizeof(*mode), GFP_KERNEL); if (!mode) return NULL; - if (use_gct) { - mode->hdisplay = (ti->hactive_hi << 8) | ti->hactive_lo; - mode->vdisplay = (ti->vactive_hi << 8) | ti->vactive_lo; - mode->hsync_start = mode->hdisplay + - ((ti->hsync_offset_hi << 8) | - ti->hsync_offset_lo); - mode->hsync_end = mode->hsync_start + - ((ti->hsync_pulse_width_hi << 8) | - ti->hsync_pulse_width_lo); - mode->htotal = mode->hdisplay + ((ti->hblank_hi << 8) | - ti->hblank_lo); - mode->vsync_start = - mode->vdisplay + ((ti->vsync_offset_hi << 8) | - ti->vsync_offset_lo); - mode->vsync_end = - mode->vsync_start + ((ti->vsync_pulse_width_hi << 8) | - ti->vsync_pulse_width_lo); - mode->vtotal = mode->vdisplay + - ((ti->vblank_hi << 8) | ti->vblank_lo); - mode->clock = ti->pixel_clock * 10; - - dev_dbg(dev->dev, "hdisplay is %d\n", mode->hdisplay); - dev_dbg(dev->dev, "vdisplay is %d\n", mode->vdisplay); - dev_dbg(dev->dev, "HSS is %d\n", mode->hsync_start); - dev_dbg(dev->dev, "HSE is %d\n", mode->hsync_end); - dev_dbg(dev->dev, "htotal is %d\n", mode->htotal); - dev_dbg(dev->dev, "VSS is %d\n", mode->vsync_start); - dev_dbg(dev->dev, "VSE is %d\n", mode->vsync_end); - dev_dbg(dev->dev, "vtotal is %d\n", mode->vtotal); - dev_dbg(dev->dev, "clock is %d\n", mode->clock); - } else { - mode->hdisplay = 864; - mode->vdisplay = 480; - mode->hsync_start = 873; - mode->hsync_end = 876; - mode->htotal = 887; - mode->vsync_start = 487; - mode->vsync_end = 490; - mode->vtotal = 499; - mode->clock = 33264; - } + mode->hdisplay = 864; + mode->vdisplay = 480; + mode->hsync_start = 873; + mode->hsync_end = 876; + mode->htotal = 887; + mode->vsync_start = 487; + mode->vsync_end = 490; + mode->vtotal = 499; + mode->clock = 33264; drm_mode_set_name(mode); drm_mode_set_crtcinfo(mode, 0); diff --git a/drivers/gpu/drm/gma500/psb_drv.c b/drivers/gpu/drm/gma500/psb_drv.c index 5ee93ff55608..1f9b35afefee 100644 --- a/drivers/gpu/drm/gma500/psb_drv.c +++ b/drivers/gpu/drm/gma500/psb_drv.c @@ -35,6 +35,7 @@ #include <linux/pm_runtime.h> #include <acpi/video.h> #include <linux/module.h> +#include <asm/set_memory.h> static struct drm_driver driver; static int psb_pci_probe(struct pci_dev *pdev, const struct pci_device_id *ent); diff --git a/drivers/gpu/drm/hisilicon/hibmc/Makefile b/drivers/gpu/drm/hisilicon/hibmc/Makefile index f2e04c035673..3df726696372 100644 --- a/drivers/gpu/drm/hisilicon/hibmc/Makefile +++ b/drivers/gpu/drm/hisilicon/hibmc/Makefile @@ -1,4 +1,3 @@ -ccflags-y := -Iinclude/drm hibmc-drm-y := hibmc_drm_drv.o hibmc_drm_de.o hibmc_drm_vdac.o hibmc_drm_fbdev.o hibmc_ttm.o obj-$(CONFIG_DRM_HISI_HIBMC) += hibmc-drm.o diff --git a/drivers/gpu/drm/hisilicon/hibmc/hibmc_ttm.c b/drivers/gpu/drm/hisilicon/hibmc/hibmc_ttm.c index 20732b62d4c9..ac457c779caa 100644 --- a/drivers/gpu/drm/hisilicon/hibmc/hibmc_ttm.c +++ b/drivers/gpu/drm/hisilicon/hibmc/hibmc_ttm.c @@ -17,7 +17,7 @@ */ #include <drm/drm_atomic_helper.h> -#include <ttm/ttm_page_alloc.h> +#include <drm/ttm/ttm_page_alloc.h> #include "hibmc_drm_drv.h" diff --git a/drivers/gpu/drm/i2c/Makefile b/drivers/gpu/drm/i2c/Makefile index 43aa33baebed..a77acfc1852e 100644 --- a/drivers/gpu/drm/i2c/Makefile +++ b/drivers/gpu/drm/i2c/Makefile @@ -1,5 +1,3 @@ -ccflags-y := -Iinclude/drm - ch7006-y := ch7006_drv.o ch7006_mode.o obj-$(CONFIG_DRM_I2C_CH7006) += ch7006.o diff --git a/drivers/gpu/drm/i810/Makefile b/drivers/gpu/drm/i810/Makefile index 43844ecafcc5..639f8596c978 100644 --- a/drivers/gpu/drm/i810/Makefile +++ b/drivers/gpu/drm/i810/Makefile @@ -2,7 +2,6 @@ # Makefile for the drm device driver. This driver provides support for the # Direct Rendering Infrastructure (DRI) in XFree86 4.1.0 and higher. -ccflags-y := -Iinclude/drm i810-y := i810_drv.o i810_dma.o obj-$(CONFIG_DRM_I810) += i810.o diff --git a/drivers/gpu/drm/i915/Kconfig.debug b/drivers/gpu/drm/i915/Kconfig.debug index d4860c3a3bf9..78c5c049a347 100644 --- a/drivers/gpu/drm/i915/Kconfig.debug +++ b/drivers/gpu/drm/i915/Kconfig.debug @@ -99,3 +99,16 @@ config DRM_I915_LOW_LEVEL_TRACEPOINTS and also analyze the request dependency resolving timeline. If in doubt, say "N". + +config DRM_I915_DEBUG_VBLANK_EVADE + bool "Enable extra debug warnings for vblank evasion" + depends on DRM_I915 + default n + help + Choose this option to turn on extra debug warnings for the + vblank evade mechanism. This gives a warning every time the + the deadline allotted for the vblank evade critical section + is exceeded, even if there isn't an actual risk of missing + the vblank. + + If in doubt, say "N". diff --git a/drivers/gpu/drm/i915/i915_debugfs.c b/drivers/gpu/drm/i915/i915_debugfs.c index 0615237d1cea..7e0816ccdc21 100644 --- a/drivers/gpu/drm/i915/i915_debugfs.c +++ b/drivers/gpu/drm/i915/i915_debugfs.c @@ -229,7 +229,7 @@ static int i915_gem_stolen_list_info(struct seq_file *m, void *data) int ret; total = READ_ONCE(dev_priv->mm.object_count); - objects = drm_malloc_ab(total, sizeof(*objects)); + objects = kvmalloc_array(total, sizeof(*objects), GFP_KERNEL); if (!objects) return -ENOMEM; @@ -274,7 +274,7 @@ static int i915_gem_stolen_list_info(struct seq_file *m, void *data) mutex_unlock(&dev->struct_mutex); out: - drm_free_large(objects); + kvfree(objects); return ret; } @@ -3118,17 +3118,17 @@ static const char *plane_rotation(unsigned int rotation) { static char buf[48]; /* - * According to doc only one DRM_ROTATE_ is allowed but this + * According to doc only one DRM_MODE_ROTATE_ is allowed but this * will print them all to visualize if the values are misused */ snprintf(buf, sizeof(buf), "%s%s%s%s%s%s(0x%08x)", - (rotation & DRM_ROTATE_0) ? "0 " : "", - (rotation & DRM_ROTATE_90) ? "90 " : "", - (rotation & DRM_ROTATE_180) ? "180 " : "", - (rotation & DRM_ROTATE_270) ? "270 " : "", - (rotation & DRM_REFLECT_X) ? "FLIPX " : "", - (rotation & DRM_REFLECT_Y) ? "FLIPY " : "", + (rotation & DRM_MODE_ROTATE_0) ? "0 " : "", + (rotation & DRM_MODE_ROTATE_90) ? "90 " : "", + (rotation & DRM_MODE_ROTATE_180) ? "180 " : "", + (rotation & DRM_MODE_ROTATE_270) ? "270 " : "", + (rotation & DRM_MODE_REFLECT_X) ? "FLIPX " : "", + (rotation & DRM_MODE_REFLECT_Y) ? "FLIPY " : "", rotation); return buf; diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c index 7572dcdf4745..7ab47a84671f 100644 --- a/drivers/gpu/drm/i915/i915_gem.c +++ b/drivers/gpu/drm/i915/i915_gem.c @@ -2556,7 +2556,7 @@ static void *i915_gem_object_map(const struct drm_i915_gem_object *obj, if (n_pages > ARRAY_SIZE(stack_pages)) { /* Too big for stack -- allocate temporary array instead */ - pages = drm_malloc_gfp(n_pages, sizeof(*pages), GFP_TEMPORARY); + pages = kvmalloc_array(n_pages, sizeof(*pages), GFP_TEMPORARY); if (!pages) return NULL; } @@ -2578,7 +2578,7 @@ static void *i915_gem_object_map(const struct drm_i915_gem_object *obj, addr = vmap(pages, n_pages, 0, pgprot); if (pages != stack_pages) - drm_free_large(pages); + kvfree(pages); return addr; } @@ -4854,7 +4854,7 @@ i915_gem_load_init(struct drm_i915_private *dev_priv) dev_priv->requests = KMEM_CACHE(drm_i915_gem_request, SLAB_HWCACHE_ALIGN | SLAB_RECLAIM_ACCOUNT | - SLAB_DESTROY_BY_RCU); + SLAB_TYPESAFE_BY_RCU); if (!dev_priv->requests) goto err_vmas; diff --git a/drivers/gpu/drm/i915/i915_gem_execbuffer.c b/drivers/gpu/drm/i915/i915_gem_execbuffer.c index af1965774e7b..04211c970b9f 100644 --- a/drivers/gpu/drm/i915/i915_gem_execbuffer.c +++ b/drivers/gpu/drm/i915/i915_gem_execbuffer.c @@ -1019,11 +1019,11 @@ i915_gem_execbuffer_relocate_slow(struct drm_device *dev, for (i = 0; i < count; i++) total += exec[i].relocation_count; - reloc_offset = drm_malloc_ab(count, sizeof(*reloc_offset)); - reloc = drm_malloc_ab(total, sizeof(*reloc)); + reloc_offset = kvmalloc_array(count, sizeof(*reloc_offset), GFP_KERNEL); + reloc = kvmalloc_array(total, sizeof(*reloc), GFP_KERNEL); if (reloc == NULL || reloc_offset == NULL) { - drm_free_large(reloc); - drm_free_large(reloc_offset); + kvfree(reloc); + kvfree(reloc_offset); mutex_lock(&dev->struct_mutex); return -ENOMEM; } @@ -1099,8 +1099,8 @@ i915_gem_execbuffer_relocate_slow(struct drm_device *dev, */ err: - drm_free_large(reloc); - drm_free_large(reloc_offset); + kvfree(reloc); + kvfree(reloc_offset); return ret; } @@ -1871,13 +1871,13 @@ i915_gem_execbuffer(struct drm_device *dev, void *data, } /* Copy in the exec list from userland */ - exec_list = drm_malloc_ab(sizeof(*exec_list), args->buffer_count); - exec2_list = drm_malloc_ab(sizeof(*exec2_list), args->buffer_count); + exec_list = kvmalloc_array(sizeof(*exec_list), args->buffer_count, GFP_KERNEL); + exec2_list = kvmalloc_array(sizeof(*exec2_list), args->buffer_count, GFP_KERNEL); if (exec_list == NULL || exec2_list == NULL) { DRM_DEBUG("Failed to allocate exec list for %d buffers\n", args->buffer_count); - drm_free_large(exec_list); - drm_free_large(exec2_list); + kvfree(exec_list); + kvfree(exec2_list); return -ENOMEM; } ret = copy_from_user(exec_list, @@ -1886,8 +1886,8 @@ i915_gem_execbuffer(struct drm_device *dev, void *data, if (ret != 0) { DRM_DEBUG("copy %d exec entries failed %d\n", args->buffer_count, ret); - drm_free_large(exec_list); - drm_free_large(exec2_list); + kvfree(exec_list); + kvfree(exec2_list); return -EFAULT; } @@ -1936,8 +1936,8 @@ i915_gem_execbuffer(struct drm_device *dev, void *data, } } - drm_free_large(exec_list); - drm_free_large(exec2_list); + kvfree(exec_list); + kvfree(exec2_list); return ret; } @@ -1955,7 +1955,7 @@ i915_gem_execbuffer2(struct drm_device *dev, void *data, return -EINVAL; } - exec2_list = drm_malloc_gfp(args->buffer_count, + exec2_list = kvmalloc_array(args->buffer_count, sizeof(*exec2_list), GFP_TEMPORARY); if (exec2_list == NULL) { @@ -1969,7 +1969,7 @@ i915_gem_execbuffer2(struct drm_device *dev, void *data, if (ret != 0) { DRM_DEBUG("copy %d exec entries failed %d\n", args->buffer_count, ret); - drm_free_large(exec2_list); + kvfree(exec2_list); return -EFAULT; } @@ -1996,6 +1996,6 @@ i915_gem_execbuffer2(struct drm_device *dev, void *data, } } - drm_free_large(exec2_list); + kvfree(exec2_list); return ret; } diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.c b/drivers/gpu/drm/i915/i915_gem_gtt.c index f52068aee247..0c1008a2bbda 100644 --- a/drivers/gpu/drm/i915/i915_gem_gtt.c +++ b/drivers/gpu/drm/i915/i915_gem_gtt.c @@ -31,6 +31,8 @@ #include <linux/seq_file.h> #include <linux/stop_machine.h> +#include <asm/set_memory.h> + #include <drm/drmP.h> #include <drm/i915_drm.h> @@ -3207,7 +3209,7 @@ intel_rotate_pages(struct intel_rotation_info *rot_info, int ret = -ENOMEM; /* Allocate a temporary list of source pages for random access. */ - page_addr_list = drm_malloc_gfp(n_pages, + page_addr_list = kvmalloc_array(n_pages, sizeof(dma_addr_t), GFP_TEMPORARY); if (!page_addr_list) @@ -3240,14 +3242,14 @@ intel_rotate_pages(struct intel_rotation_info *rot_info, DRM_DEBUG_KMS("Created rotated page mapping for object size %zu (%ux%u tiles, %u pages)\n", obj->base.size, rot_info->plane[0].width, rot_info->plane[0].height, size); - drm_free_large(page_addr_list); + kvfree(page_addr_list); return st; err_sg_alloc: kfree(st); err_st_alloc: - drm_free_large(page_addr_list); + kvfree(page_addr_list); DRM_DEBUG_KMS("Failed to create rotated mapping for object size %zu! (%ux%u tiles, %u pages)\n", obj->base.size, rot_info->plane[0].width, rot_info->plane[0].height, size); diff --git a/drivers/gpu/drm/i915/i915_gem_request.h b/drivers/gpu/drm/i915/i915_gem_request.h index 8c508bd9088e..7b7c84369d78 100644 --- a/drivers/gpu/drm/i915/i915_gem_request.h +++ b/drivers/gpu/drm/i915/i915_gem_request.h @@ -533,7 +533,7 @@ static inline struct drm_i915_gem_request * __i915_gem_active_get_rcu(const struct i915_gem_active *active) { /* Performing a lockless retrieval of the active request is super - * tricky. SLAB_DESTROY_BY_RCU merely guarantees that the backing + * tricky. SLAB_TYPESAFE_BY_RCU merely guarantees that the backing * slab of request objects will not be freed whilst we hold the * RCU read lock. It does not guarantee that the request itself * will not be freed and then *reused*. Viz, diff --git a/drivers/gpu/drm/i915/i915_gem_userptr.c b/drivers/gpu/drm/i915/i915_gem_userptr.c index 58ccf8b8ca1c..1a0ce1dc68f5 100644 --- a/drivers/gpu/drm/i915/i915_gem_userptr.c +++ b/drivers/gpu/drm/i915/i915_gem_userptr.c @@ -507,7 +507,7 @@ __i915_gem_userptr_get_pages_worker(struct work_struct *_work) ret = -ENOMEM; pinned = 0; - pvec = drm_malloc_gfp(npages, sizeof(struct page *), GFP_TEMPORARY); + pvec = kvmalloc_array(npages, sizeof(struct page *), GFP_TEMPORARY); if (pvec != NULL) { struct mm_struct *mm = obj->userptr.mm->mm; unsigned int flags = 0; @@ -555,7 +555,7 @@ __i915_gem_userptr_get_pages_worker(struct work_struct *_work) mutex_unlock(&obj->mm.lock); release_pages(pvec, pinned, 0); - drm_free_large(pvec); + kvfree(pvec); i915_gem_object_put(obj); put_task_struct(work->task); @@ -642,7 +642,7 @@ i915_gem_userptr_get_pages(struct drm_i915_gem_object *obj) pinned = 0; if (mm == current->mm) { - pvec = drm_malloc_gfp(num_pages, sizeof(struct page *), + pvec = kvmalloc_array(num_pages, sizeof(struct page *), GFP_TEMPORARY | __GFP_NORETRY | __GFP_NOWARN); @@ -669,7 +669,7 @@ i915_gem_userptr_get_pages(struct drm_i915_gem_object *obj) if (IS_ERR(pages)) release_pages(pvec, pinned, 0); - drm_free_large(pvec); + kvfree(pvec); return pages; } diff --git a/drivers/gpu/drm/i915/i915_irq.c b/drivers/gpu/drm/i915/i915_irq.c index d63a2ba3bc4f..7b7f55a28eec 100644 --- a/drivers/gpu/drm/i915/i915_irq.c +++ b/drivers/gpu/drm/i915/i915_irq.c @@ -720,9 +720,7 @@ static u32 i915_get_vblank_counter(struct drm_device *dev, unsigned int pipe) struct drm_i915_private *dev_priv = to_i915(dev); i915_reg_t high_frame, low_frame; u32 high1, high2, low, pixel, vbl_start, hsync_start, htotal; - struct intel_crtc *intel_crtc = intel_get_crtc_for_pipe(dev_priv, - pipe); - const struct drm_display_mode *mode = &intel_crtc->base.hwmode; + const struct drm_display_mode *mode = &dev->vblank[pipe].hwmode; unsigned long irqflags; htotal = mode->crtc_htotal; @@ -779,13 +777,17 @@ static int __intel_get_crtc_scanline(struct intel_crtc *crtc) { struct drm_device *dev = crtc->base.dev; struct drm_i915_private *dev_priv = to_i915(dev); - const struct drm_display_mode *mode = &crtc->base.hwmode; + const struct drm_display_mode *mode; + struct drm_vblank_crtc *vblank; enum pipe pipe = crtc->pipe; int position, vtotal; if (!crtc->active) return -1; + vblank = &crtc->base.dev->vblank[drm_crtc_index(&crtc->base)]; + mode = &vblank->hwmode; + vtotal = mode->crtc_vtotal; if (mode->flags & DRM_MODE_FLAG_INTERLACE) vtotal /= 2; @@ -827,10 +829,10 @@ static int __intel_get_crtc_scanline(struct intel_crtc *crtc) return (position + crtc->scanline_offset) % vtotal; } -static int i915_get_crtc_scanoutpos(struct drm_device *dev, unsigned int pipe, - unsigned int flags, int *vpos, int *hpos, - ktime_t *stime, ktime_t *etime, - const struct drm_display_mode *mode) +static bool i915_get_crtc_scanoutpos(struct drm_device *dev, unsigned int pipe, + bool in_vblank_irq, int *vpos, int *hpos, + ktime_t *stime, ktime_t *etime, + const struct drm_display_mode *mode) { struct drm_i915_private *dev_priv = to_i915(dev); struct intel_crtc *intel_crtc = intel_get_crtc_for_pipe(dev_priv, @@ -838,13 +840,12 @@ static int i915_get_crtc_scanoutpos(struct drm_device *dev, unsigned int pipe, int position; int vbl_start, vbl_end, hsync_start, htotal, vtotal; bool in_vbl = true; - int ret = 0; unsigned long irqflags; if (WARN_ON(!mode->crtc_clock)) { DRM_DEBUG_DRIVER("trying to get scanoutpos for disabled " "pipe %c\n", pipe_name(pipe)); - return 0; + return false; } htotal = mode->crtc_htotal; @@ -859,8 +860,6 @@ static int i915_get_crtc_scanoutpos(struct drm_device *dev, unsigned int pipe, vtotal /= 2; } - ret |= DRM_SCANOUTPOS_VALID | DRM_SCANOUTPOS_ACCURATE; - /* * Lock uncore.lock, as we will do multiple timing critical raw * register reads, potentially with preemption disabled, so the @@ -944,11 +943,7 @@ static int i915_get_crtc_scanoutpos(struct drm_device *dev, unsigned int pipe, *hpos = position - (*vpos * htotal); } - /* In vblank? */ - if (in_vbl) - ret |= DRM_SCANOUTPOS_IN_VBLANK; - - return ret; + return true; } int intel_get_crtc_scanline(struct intel_crtc *crtc) @@ -964,37 +959,6 @@ int intel_get_crtc_scanline(struct intel_crtc *crtc) return position; } -static int i915_get_vblank_timestamp(struct drm_device *dev, unsigned int pipe, - int *max_error, - struct timeval *vblank_time, - unsigned flags) -{ - struct drm_i915_private *dev_priv = to_i915(dev); - struct intel_crtc *crtc; - - if (pipe >= INTEL_INFO(dev_priv)->num_pipes) { - DRM_ERROR("Invalid crtc %u\n", pipe); - return -EINVAL; - } - - /* Get drm_crtc to timestamp: */ - crtc = intel_get_crtc_for_pipe(dev_priv, pipe); - if (crtc == NULL) { - DRM_ERROR("Invalid crtc %u\n", pipe); - return -EINVAL; - } - - if (!crtc->base.hwmode.crtc_clock) { - DRM_DEBUG_KMS("crtc %u is disabled\n", pipe); - return -EBUSY; - } - - /* Helper routine in DRM core does all the work: */ - return drm_calc_vbltimestamp_from_scanoutpos(dev, pipe, max_error, - vblank_time, flags, - &crtc->base.hwmode); -} - static void ironlake_rps_change_irq_handler(struct drm_i915_private *dev_priv) { u32 busy_up, busy_down, max_avg, min_avg; @@ -4297,7 +4261,7 @@ void intel_irq_init(struct drm_i915_private *dev_priv) dev_priv->hotplug.hpd_storm_threshold = HPD_STORM_DEFAULT_THRESHOLD; - dev->driver->get_vblank_timestamp = i915_get_vblank_timestamp; + dev->driver->get_vblank_timestamp = drm_calc_vbltimestamp_from_scanoutpos; dev->driver->get_scanout_position = i915_get_crtc_scanoutpos; if (IS_CHERRYVIEW(dev_priv)) { diff --git a/drivers/gpu/drm/i915/intel_atomic_plane.c b/drivers/gpu/drm/i915/intel_atomic_plane.c index 4a0ed0278ae9..4325cb0a04f5 100644 --- a/drivers/gpu/drm/i915/intel_atomic_plane.c +++ b/drivers/gpu/drm/i915/intel_atomic_plane.c @@ -55,7 +55,7 @@ intel_create_plane_state(struct drm_plane *plane) return NULL; state->base.plane = plane; - state->base.rotation = DRM_ROTATE_0; + state->base.rotation = DRM_MODE_ROTATE_0; state->ckey.flags = I915_SET_COLORKEY_NONE; return state; @@ -162,8 +162,8 @@ int intel_plane_atomic_check_with_state(struct intel_crtc_state *crtc_state, /* CHV ignores the mirror bit when the rotate bit is set :( */ if (IS_CHERRYVIEW(dev_priv) && - state->rotation & DRM_ROTATE_180 && - state->rotation & DRM_REFLECT_X) { + state->rotation & DRM_MODE_ROTATE_180 && + state->rotation & DRM_MODE_REFLECT_X) { DRM_DEBUG_KMS("Cannot rotate and reflect at the same time\n"); return -EINVAL; } diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index 55c2c149ad0b..7fa21df5bcd7 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -2486,7 +2486,7 @@ intel_fill_fb_info(struct drm_i915_private *dev_priv, offset = _intel_compute_tile_offset(dev_priv, &x, &y, fb, i, fb->pitches[i], - DRM_ROTATE_0, tile_size); + DRM_MODE_ROTATE_0, tile_size); offset /= tile_size; if (fb->modifier != DRM_FORMAT_MOD_LINEAR) { @@ -2521,7 +2521,7 @@ intel_fill_fb_info(struct drm_i915_private *dev_priv, drm_rect_rotate(&r, rot_info->plane[i].width * tile_width, rot_info->plane[i].height * tile_height, - DRM_ROTATE_270); + DRM_MODE_ROTATE_270); x = r.x1; y = r.y1; @@ -2957,7 +2957,7 @@ int skl_check_plane_surface(struct intel_plane_state *plane_state) if (drm_rotation_90_or_270(rotation)) drm_rect_rotate(&plane_state->base.src, fb->width << 16, fb->height << 16, - DRM_ROTATE_270); + DRM_MODE_ROTATE_270); /* * Handle the AUX surface first since @@ -3033,10 +3033,10 @@ static u32 i9xx_plane_ctl(const struct intel_crtc_state *crtc_state, fb->modifier == I915_FORMAT_MOD_X_TILED) dspcntr |= DISPPLANE_TILED; - if (rotation & DRM_ROTATE_180) + if (rotation & DRM_MODE_ROTATE_180) dspcntr |= DISPPLANE_ROTATE_180; - if (rotation & DRM_REFLECT_X) + if (rotation & DRM_MODE_REFLECT_X) dspcntr |= DISPPLANE_MIRROR; return dspcntr; @@ -3064,10 +3064,10 @@ int i9xx_check_plane_surface(struct intel_plane_state *plane_state) int src_w = drm_rect_width(&plane_state->base.src) >> 16; int src_h = drm_rect_height(&plane_state->base.src) >> 16; - if (rotation & DRM_ROTATE_180) { + if (rotation & DRM_MODE_ROTATE_180) { src_x += src_w - 1; src_y += src_h - 1; - } else if (rotation & DRM_REFLECT_X) { + } else if (rotation & DRM_MODE_REFLECT_X) { src_x += src_w - 1; } } @@ -3285,17 +3285,17 @@ static u32 skl_plane_ctl_tiling(uint64_t fb_modifier) static u32 skl_plane_ctl_rotation(unsigned int rotation) { switch (rotation) { - case DRM_ROTATE_0: + case DRM_MODE_ROTATE_0: break; /* - * DRM_ROTATE_ is counter clockwise to stay compatible with Xrandr + * DRM_MODE_ROTATE_ is counter clockwise to stay compatible with Xrandr * while i915 HW rotation is clockwise, thats why this swapping. */ - case DRM_ROTATE_90: + case DRM_MODE_ROTATE_90: return PLANE_CTL_ROTATE_270; - case DRM_ROTATE_180: + case DRM_MODE_ROTATE_180: return PLANE_CTL_ROTATE_180; - case DRM_ROTATE_270: + case DRM_MODE_ROTATE_270: return PLANE_CTL_ROTATE_90; default: MISSING_CASE(rotation); @@ -4683,7 +4683,7 @@ int skl_update_scaler_crtc(struct intel_crtc_state *state) const struct drm_display_mode *adjusted_mode = &state->base.adjusted_mode; return skl_update_scaler(state, !state->base.active, SKL_CRTC_INDEX, - &state->scaler_state.scaler_id, DRM_ROTATE_0, + &state->scaler_state.scaler_id, DRM_MODE_ROTATE_0, state->pipe_src_w, state->pipe_src_h, adjusted_mode->crtc_hdisplay, adjusted_mode->crtc_vdisplay); } @@ -9161,7 +9161,7 @@ static u32 intel_cursor_base(const struct intel_plane_state *plane_state) /* ILK+ do this automagically */ if (HAS_GMCH_DISPLAY(dev_priv) && - plane_state->base.rotation & DRM_ROTATE_180) + plane_state->base.rotation & DRM_MODE_ROTATE_180) base += (plane_state->base.crtc_h * plane_state->base.crtc_w - 1) * fb->format->cpp[0]; @@ -9383,7 +9383,7 @@ static u32 i9xx_cursor_ctl(const struct intel_crtc_state *crtc_state, return 0; } - if (plane_state->base.rotation & DRM_ROTATE_180) + if (plane_state->base.rotation & DRM_MODE_ROTATE_180) cntl |= CURSOR_ROTATE_180; return cntl; @@ -9416,7 +9416,7 @@ static bool i9xx_cursor_size_ok(const struct intel_plane_state *plane_state) * cursors. */ if (HAS_CUR_FBC(dev_priv) && - plane_state->base.rotation & DRM_ROTATE_0) { + plane_state->base.rotation & DRM_MODE_ROTATE_0) { if (height < 8 || height > width) return false; } else { @@ -11630,12 +11630,6 @@ intel_modeset_update_crtc_state(struct drm_atomic_state *state) for_each_new_crtc_in_state(state, crtc, new_crtc_state, i) { to_intel_crtc(crtc)->config = to_intel_crtc_state(new_crtc_state); - /* Update hwmode for vblank functions */ - if (new_crtc_state->active) - crtc->hwmode = new_crtc_state->adjusted_mode; - else - crtc->hwmode.crtc_clock = 0; - /* * Update legacy state to satisfy fbc code. This can * be removed when fbc uses the atomic state. @@ -13803,22 +13797,22 @@ intel_primary_plane_create(struct drm_i915_private *dev_priv, enum pipe pipe) if (INTEL_GEN(dev_priv) >= 9) { supported_rotations = - DRM_ROTATE_0 | DRM_ROTATE_90 | - DRM_ROTATE_180 | DRM_ROTATE_270; + DRM_MODE_ROTATE_0 | DRM_MODE_ROTATE_90 | + DRM_MODE_ROTATE_180 | DRM_MODE_ROTATE_270; } else if (IS_CHERRYVIEW(dev_priv) && pipe == PIPE_B) { supported_rotations = - DRM_ROTATE_0 | DRM_ROTATE_180 | - DRM_REFLECT_X; + DRM_MODE_ROTATE_0 | DRM_MODE_ROTATE_180 | + DRM_MODE_REFLECT_X; } else if (INTEL_GEN(dev_priv) >= 4) { supported_rotations = - DRM_ROTATE_0 | DRM_ROTATE_180; + DRM_MODE_ROTATE_0 | DRM_MODE_ROTATE_180; } else { - supported_rotations = DRM_ROTATE_0; + supported_rotations = DRM_MODE_ROTATE_0; } if (INTEL_GEN(dev_priv) >= 4) drm_plane_create_rotation_property(&primary->base, - DRM_ROTATE_0, + DRM_MODE_ROTATE_0, supported_rotations); drm_plane_helper_add(&primary->base, &intel_plane_helper_funcs); @@ -13888,9 +13882,9 @@ intel_cursor_plane_create(struct drm_i915_private *dev_priv, if (INTEL_GEN(dev_priv) >= 4) drm_plane_create_rotation_property(&cursor->base, - DRM_ROTATE_0, - DRM_ROTATE_0 | - DRM_ROTATE_180); + DRM_MODE_ROTATE_0, + DRM_MODE_ROTATE_0 | + DRM_MODE_ROTATE_180); if (INTEL_GEN(dev_priv) >= 9) state->scaler_id = -1; @@ -15527,8 +15521,6 @@ static void intel_modeset_readout_hw_state(struct drm_device *dev) to_intel_crtc_state(crtc->base.state); int pixclk = 0; - crtc->base.hwmode = crtc_state->base.adjusted_mode; - memset(&crtc->base.mode, 0, sizeof(crtc->base.mode)); if (crtc_state->base.active) { intel_mode_from_pipe_config(&crtc->base.mode, crtc_state); @@ -15558,7 +15550,8 @@ static void intel_modeset_readout_hw_state(struct drm_device *dev) if (IS_BROADWELL(dev_priv) && crtc_state->ips_enabled) pixclk = DIV_ROUND_UP(pixclk * 100, 95); - drm_calc_timestamping_constants(&crtc->base, &crtc->base.hwmode); + drm_calc_timestamping_constants(&crtc->base, + &crtc_state->base.adjusted_mode); update_scanline_offset(crtc); } diff --git a/drivers/gpu/drm/i915/intel_dp_mst.c b/drivers/gpu/drm/i915/intel_dp_mst.c index 5af22a7c11bf..3715386e4272 100644 --- a/drivers/gpu/drm/i915/intel_dp_mst.c +++ b/drivers/gpu/drm/i915/intel_dp_mst.c @@ -39,7 +39,7 @@ static bool intel_dp_mst_compute_config(struct intel_encoder *encoder, struct intel_dp *intel_dp = &intel_dig_port->dp; struct intel_connector *connector = to_intel_connector(conn_state->connector); - struct drm_atomic_state *state; + struct drm_atomic_state *state = pipe_config->base.state; int bpp; int lane_count, slots; const struct drm_display_mode *adjusted_mode = &pipe_config->base.adjusted_mode; @@ -61,16 +61,21 @@ static bool intel_dp_mst_compute_config(struct intel_encoder *encoder, pipe_config->lane_count = lane_count; pipe_config->pipe_bpp = bpp; - pipe_config->port_clock = intel_dp_max_link_rate(intel_dp); - state = pipe_config->base.state; + pipe_config->port_clock = intel_dp_max_link_rate(intel_dp); if (drm_dp_mst_port_has_audio(&intel_dp->mst_mgr, connector->port)) pipe_config->has_audio = true; - mst_pbn = drm_dp_calc_pbn_mode(adjusted_mode->crtc_clock, bpp); + mst_pbn = drm_dp_calc_pbn_mode(adjusted_mode->crtc_clock, bpp); pipe_config->pbn = mst_pbn; - slots = drm_dp_find_vcpi_slots(&intel_dp->mst_mgr, mst_pbn); + + slots = drm_dp_atomic_find_vcpi_slots(state, &intel_dp->mst_mgr, + connector->port, mst_pbn); + if (slots < 0) { + DRM_DEBUG_KMS("failed finding vcpi slots:%d\n", slots); + return false; + } intel_link_compute_m_n(bpp, lane_count, adjusted_mode->crtc_clock, @@ -80,7 +85,38 @@ static bool intel_dp_mst_compute_config(struct intel_encoder *encoder, pipe_config->dp_m_n.tu = slots; return true; +} +static int intel_dp_mst_atomic_check(struct drm_connector *connector, + struct drm_connector_state *new_conn_state) +{ + struct drm_atomic_state *state = new_conn_state->state; + struct drm_connector_state *old_conn_state; + struct drm_crtc *old_crtc; + struct drm_crtc_state *crtc_state; + int slots, ret = 0; + + old_conn_state = drm_atomic_get_old_connector_state(state, connector); + old_crtc = old_conn_state->crtc; + if (!old_crtc) + return ret; + + crtc_state = drm_atomic_get_new_crtc_state(state, old_crtc); + slots = to_intel_crtc_state(crtc_state)->dp_m_n.tu; + if (drm_atomic_crtc_needs_modeset(crtc_state) && slots > 0) { + struct drm_dp_mst_topology_mgr *mgr; + struct drm_encoder *old_encoder; + + old_encoder = old_conn_state->best_encoder; + mgr = &enc_to_mst(old_encoder)->primary->dp.mst_mgr; + + ret = drm_dp_atomic_release_vcpi_slots(state, mgr, slots); + if (ret) + DRM_DEBUG_KMS("failed releasing %d vcpi slots:%d\n", slots, ret); + else + to_intel_crtc_state(crtc_state)->dp_m_n.tu = 0; + } + return ret; } static void intel_mst_disable_dp(struct intel_encoder *encoder, @@ -378,6 +414,7 @@ static const struct drm_connector_helper_funcs intel_dp_mst_connector_helper_fun .mode_valid = intel_dp_mst_mode_valid, .atomic_best_encoder = intel_mst_atomic_best_encoder, .best_encoder = intel_mst_best_encoder, + .atomic_check = intel_dp_mst_atomic_check, }; static void intel_dp_mst_encoder_destroy(struct drm_encoder *encoder) diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h index cc1370686193..bd500977b3fc 100644 --- a/drivers/gpu/drm/i915/intel_drv.h +++ b/drivers/gpu/drm/i915/intel_drv.h @@ -902,7 +902,6 @@ struct intel_hdmi { bool has_audio; enum hdmi_force_audio force_audio; bool rgb_quant_range_selectable; - enum hdmi_picture_aspect aspect_ratio; struct intel_connector *attached_connector; void (*write_infoframe)(struct drm_encoder *encoder, const struct intel_crtc_state *crtc_state, diff --git a/drivers/gpu/drm/i915/intel_fbc.c b/drivers/gpu/drm/i915/intel_fbc.c index 64e44d9fa6a5..ff2fc5bc4af4 100644 --- a/drivers/gpu/drm/i915/intel_fbc.c +++ b/drivers/gpu/drm/i915/intel_fbc.c @@ -801,7 +801,7 @@ static bool intel_fbc_can_activate(struct intel_crtc *crtc) return false; } if (INTEL_GEN(dev_priv) <= 4 && !IS_G4X(dev_priv) && - cache->plane.rotation != DRM_ROTATE_0) { + cache->plane.rotation != DRM_MODE_ROTATE_0) { fbc->no_fbc_reason = "rotation unsupported"; return false; } diff --git a/drivers/gpu/drm/i915/intel_fbdev.c b/drivers/gpu/drm/i915/intel_fbdev.c index 332254a8eebe..03347c6ae599 100644 --- a/drivers/gpu/drm/i915/intel_fbdev.c +++ b/drivers/gpu/drm/i915/intel_fbdev.c @@ -211,7 +211,7 @@ static int intelfb_create(struct drm_fb_helper *helper, * This also validates that any existing fb inherited from the * BIOS is suitable for own access. */ - vma = intel_pin_and_fence_fb_obj(&ifbdev->fb->base, DRM_ROTATE_0); + vma = intel_pin_and_fence_fb_obj(&ifbdev->fb->base, DRM_MODE_ROTATE_0); if (IS_ERR(vma)) { ret = PTR_ERR(vma); goto out_unlock; diff --git a/drivers/gpu/drm/i915/intel_hdmi.c b/drivers/gpu/drm/i915/intel_hdmi.c index 52f0b2d5fad2..58d690393b29 100644 --- a/drivers/gpu/drm/i915/intel_hdmi.c +++ b/drivers/gpu/drm/i915/intel_hdmi.c @@ -1408,7 +1408,7 @@ bool intel_hdmi_compute_config(struct intel_encoder *encoder, } /* Set user selected PAR to incoming mode's member */ - adjusted_mode->picture_aspect_ratio = intel_hdmi->aspect_ratio; + adjusted_mode->picture_aspect_ratio = conn_state->picture_aspect_ratio; pipe_config->lane_count = 4; @@ -1654,19 +1654,7 @@ intel_hdmi_set_property(struct drm_connector *connector, } if (property == connector->dev->mode_config.aspect_ratio_property) { - switch (val) { - case DRM_MODE_PICTURE_ASPECT_NONE: - intel_hdmi->aspect_ratio = HDMI_PICTURE_ASPECT_NONE; - break; - case DRM_MODE_PICTURE_ASPECT_4_3: - intel_hdmi->aspect_ratio = HDMI_PICTURE_ASPECT_4_3; - break; - case DRM_MODE_PICTURE_ASPECT_16_9: - intel_hdmi->aspect_ratio = HDMI_PICTURE_ASPECT_16_9; - break; - default: - return -EINVAL; - } + connector->state->picture_aspect_ratio = val; goto done; } @@ -1828,7 +1816,7 @@ intel_hdmi_add_properties(struct intel_hdmi *intel_hdmi, struct drm_connector *c intel_attach_broadcast_rgb_property(connector); intel_hdmi->color_range_auto = true; intel_attach_aspect_ratio_property(connector); - intel_hdmi->aspect_ratio = HDMI_PICTURE_ASPECT_NONE; + connector->state->picture_aspect_ratio = HDMI_PICTURE_ASPECT_NONE; } /* diff --git a/drivers/gpu/drm/i915/intel_sdvo.c b/drivers/gpu/drm/i915/intel_sdvo.c index 496b24c03222..6cc181203135 100644 --- a/drivers/gpu/drm/i915/intel_sdvo.c +++ b/drivers/gpu/drm/i915/intel_sdvo.c @@ -107,11 +107,6 @@ struct intel_sdvo { bool color_range_auto; /** - * HDMI user specified aspect ratio - */ - enum hdmi_picture_aspect aspect_ratio; - - /** * This is set if we're going to treat the device as TV-out. * * While we have these nice friendly flags for output types that ought @@ -1186,7 +1181,7 @@ static bool intel_sdvo_compute_config(struct intel_encoder *encoder, /* Set user selected PAR to incoming mode's member */ if (intel_sdvo->is_hdmi) - adjusted_mode->picture_aspect_ratio = intel_sdvo->aspect_ratio; + adjusted_mode->picture_aspect_ratio = conn_state->picture_aspect_ratio; return true; } @@ -2067,19 +2062,7 @@ intel_sdvo_set_property(struct drm_connector *connector, } if (property == connector->dev->mode_config.aspect_ratio_property) { - switch (val) { - case DRM_MODE_PICTURE_ASPECT_NONE: - intel_sdvo->aspect_ratio = HDMI_PICTURE_ASPECT_NONE; - break; - case DRM_MODE_PICTURE_ASPECT_4_3: - intel_sdvo->aspect_ratio = HDMI_PICTURE_ASPECT_4_3; - break; - case DRM_MODE_PICTURE_ASPECT_16_9: - intel_sdvo->aspect_ratio = HDMI_PICTURE_ASPECT_16_9; - break; - default: - return -EINVAL; - } + connector->state->picture_aspect_ratio = val; goto done; } @@ -2418,7 +2401,7 @@ intel_sdvo_add_hdmi_properties(struct intel_sdvo *intel_sdvo, intel_sdvo->color_range_auto = true; } intel_attach_aspect_ratio_property(&connector->base.base); - intel_sdvo->aspect_ratio = HDMI_PICTURE_ASPECT_NONE; + connector->base.base.state->picture_aspect_ratio = HDMI_PICTURE_ASPECT_NONE; } static struct intel_sdvo_connector *intel_sdvo_connector_alloc(void) diff --git a/drivers/gpu/drm/i915/intel_sprite.c b/drivers/gpu/drm/i915/intel_sprite.c index 9dfd5b343497..c4bf19364e49 100644 --- a/drivers/gpu/drm/i915/intel_sprite.c +++ b/drivers/gpu/drm/i915/intel_sprite.c @@ -198,12 +198,15 @@ void intel_pipe_update_end(struct intel_crtc *crtc, struct intel_flip_work *work ktime_us_delta(end_vbl_time, crtc->debug.start_vbl_time), crtc->debug.min_vbl, crtc->debug.max_vbl, crtc->debug.scanline_start, scanline_end); - } else if (ktime_us_delta(end_vbl_time, crtc->debug.start_vbl_time) > - VBLANK_EVASION_TIME_US) + } +#ifdef CONFIG_DRM_I915_DEBUG_VBLANK_EVADE + else if (ktime_us_delta(end_vbl_time, crtc->debug.start_vbl_time) > + VBLANK_EVASION_TIME_US) DRM_WARN("Atomic update on pipe (%c) took %lld us, max time under evasion is %u us\n", pipe_name(pipe), ktime_us_delta(end_vbl_time, crtc->debug.start_vbl_time), VBLANK_EVASION_TIME_US); +#endif } static void @@ -391,10 +394,10 @@ static u32 vlv_sprite_ctl(const struct intel_crtc_state *crtc_state, if (fb->modifier == I915_FORMAT_MOD_X_TILED) sprctl |= SP_TILED; - if (rotation & DRM_ROTATE_180) + if (rotation & DRM_MODE_ROTATE_180) sprctl |= SP_ROTATE_180; - if (rotation & DRM_REFLECT_X) + if (rotation & DRM_MODE_REFLECT_X) sprctl |= SP_MIRROR; if (key->flags & I915_SET_COLORKEY_SOURCE) @@ -522,7 +525,7 @@ static u32 ivb_sprite_ctl(const struct intel_crtc_state *crtc_state, if (fb->modifier == I915_FORMAT_MOD_X_TILED) sprctl |= SPRITE_TILED; - if (rotation & DRM_ROTATE_180) + if (rotation & DRM_MODE_ROTATE_180) sprctl |= SPRITE_ROTATE_180; if (key->flags & I915_SET_COLORKEY_DESTINATION) @@ -659,7 +662,7 @@ static u32 g4x_sprite_ctl(const struct intel_crtc_state *crtc_state, if (fb->modifier == I915_FORMAT_MOD_X_TILED) dvscntr |= DVS_TILED; - if (rotation & DRM_ROTATE_180) + if (rotation & DRM_MODE_ROTATE_180) dvscntr |= DVS_ROTATE_180; if (key->flags & I915_SET_COLORKEY_DESTINATION) @@ -1124,15 +1127,15 @@ intel_sprite_plane_create(struct drm_i915_private *dev_priv, if (INTEL_GEN(dev_priv) >= 9) { supported_rotations = - DRM_ROTATE_0 | DRM_ROTATE_90 | - DRM_ROTATE_180 | DRM_ROTATE_270; + DRM_MODE_ROTATE_0 | DRM_MODE_ROTATE_90 | + DRM_MODE_ROTATE_180 | DRM_MODE_ROTATE_270; } else if (IS_CHERRYVIEW(dev_priv) && pipe == PIPE_B) { supported_rotations = - DRM_ROTATE_0 | DRM_ROTATE_180 | - DRM_REFLECT_X; + DRM_MODE_ROTATE_0 | DRM_MODE_ROTATE_180 | + DRM_MODE_REFLECT_X; } else { supported_rotations = - DRM_ROTATE_0 | DRM_ROTATE_180; + DRM_MODE_ROTATE_0 | DRM_MODE_ROTATE_180; } intel_plane->pipe = pipe; @@ -1159,7 +1162,7 @@ intel_sprite_plane_create(struct drm_i915_private *dev_priv, goto fail; drm_plane_create_rotation_property(&intel_plane->base, - DRM_ROTATE_0, + DRM_MODE_ROTATE_0, supported_rotations); drm_plane_helper_add(&intel_plane->base, &intel_plane_helper_funcs); diff --git a/drivers/gpu/drm/i915/selftests/intel_breadcrumbs.c b/drivers/gpu/drm/i915/selftests/intel_breadcrumbs.c index 19860a372d90..7276194c04f7 100644 --- a/drivers/gpu/drm/i915/selftests/intel_breadcrumbs.c +++ b/drivers/gpu/drm/i915/selftests/intel_breadcrumbs.c @@ -117,7 +117,7 @@ static int igt_random_insert_remove(void *arg) mock_engine_reset(engine); - waiters = drm_malloc_gfp(count, sizeof(*waiters), GFP_TEMPORARY); + waiters = kvmalloc_array(count, sizeof(*waiters), GFP_TEMPORARY); if (!waiters) goto out_engines; @@ -169,7 +169,7 @@ out_order: out_bitmap: kfree(bitmap); out_waiters: - drm_free_large(waiters); + kvfree(waiters); out_engines: mock_engine_flush(engine); return err; @@ -187,7 +187,7 @@ static int igt_insert_complete(void *arg) mock_engine_reset(engine); - waiters = drm_malloc_gfp(count, sizeof(*waiters), GFP_TEMPORARY); + waiters = kvmalloc_array(count, sizeof(*waiters), GFP_TEMPORARY); if (!waiters) goto out_engines; @@ -254,7 +254,7 @@ static int igt_insert_complete(void *arg) out_bitmap: kfree(bitmap); out_waiters: - drm_free_large(waiters); + kvfree(waiters); out_engines: mock_engine_flush(engine); return err; @@ -368,7 +368,7 @@ static int igt_wakeup(void *arg) mock_engine_reset(engine); - waiters = drm_malloc_gfp(count, sizeof(*waiters), GFP_TEMPORARY); + waiters = kvmalloc_array(count, sizeof(*waiters), GFP_TEMPORARY); if (!waiters) goto out_engines; @@ -454,7 +454,7 @@ out_waiters: put_task_struct(waiters[n].tsk); } - drm_free_large(waiters); + kvfree(waiters); out_engines: mock_engine_flush(engine); return err; diff --git a/drivers/gpu/drm/i915/selftests/mock_gem_device.c b/drivers/gpu/drm/i915/selftests/mock_gem_device.c index 4af00000ce91..627e2aa09766 100644 --- a/drivers/gpu/drm/i915/selftests/mock_gem_device.c +++ b/drivers/gpu/drm/i915/selftests/mock_gem_device.c @@ -178,7 +178,7 @@ struct drm_i915_private *mock_gem_device(void) i915->requests = KMEM_CACHE(mock_request, SLAB_HWCACHE_ALIGN | SLAB_RECLAIM_ACCOUNT | - SLAB_DESTROY_BY_RCU); + SLAB_TYPESAFE_BY_RCU); if (!i915->requests) goto err_vmas; diff --git a/drivers/gpu/drm/imx/ipuv3-plane.c b/drivers/gpu/drm/imx/ipuv3-plane.c index d63e853a0300..49546222c6d3 100644 --- a/drivers/gpu/drm/imx/ipuv3-plane.c +++ b/drivers/gpu/drm/imx/ipuv3-plane.c @@ -273,7 +273,7 @@ void ipu_plane_state_reset(struct drm_plane *plane) if (ipu_state) { ipu_state->base.plane = plane; - ipu_state->base.rotation = DRM_ROTATE_0; + ipu_state->base.rotation = DRM_MODE_ROTATE_0; } plane->state = &ipu_state->base; diff --git a/drivers/gpu/drm/mga/Makefile b/drivers/gpu/drm/mga/Makefile index 60684785c203..49e972c2f787 100644 --- a/drivers/gpu/drm/mga/Makefile +++ b/drivers/gpu/drm/mga/Makefile @@ -2,7 +2,6 @@ # Makefile for the drm device driver. This driver provides support for the # Direct Rendering Infrastructure (DRI) in XFree86 4.1.0 and higher. -ccflags-y := -Iinclude/drm mga-y := mga_drv.o mga_dma.o mga_state.o mga_warp.o mga_irq.o mga-$(CONFIG_COMPAT) += mga_ioc32.o diff --git a/drivers/gpu/drm/mgag200/Makefile b/drivers/gpu/drm/mgag200/Makefile index a9a0300f09fc..3d91d1d6c45d 100644 --- a/drivers/gpu/drm/mgag200/Makefile +++ b/drivers/gpu/drm/mgag200/Makefile @@ -1,4 +1,3 @@ -ccflags-y := -Iinclude/drm mgag200-y := mgag200_main.o mgag200_mode.o mgag200_cursor.o \ mgag200_drv.o mgag200_fb.o mgag200_i2c.o mgag200_ttm.o diff --git a/drivers/gpu/drm/mgag200/mgag200_ttm.c b/drivers/gpu/drm/mgag200/mgag200_ttm.c index 565a217b46f2..3e7e1cd31395 100644 --- a/drivers/gpu/drm/mgag200/mgag200_ttm.c +++ b/drivers/gpu/drm/mgag200/mgag200_ttm.c @@ -26,8 +26,9 @@ * Authors: Dave Airlie <airlied@redhat.com> */ #include <drm/drmP.h> +#include <drm/ttm/ttm_page_alloc.h> + #include "mgag200_drv.h" -#include <ttm/ttm_page_alloc.h> static inline struct mga_device * mgag200_bdev(struct ttm_bo_device *bd) diff --git a/drivers/gpu/drm/msm/Makefile b/drivers/gpu/drm/msm/Makefile index 5241ac8803ba..33008fa1be9b 100644 --- a/drivers/gpu/drm/msm/Makefile +++ b/drivers/gpu/drm/msm/Makefile @@ -1,4 +1,4 @@ -ccflags-y := -Iinclude/drm -Idrivers/gpu/drm/msm +ccflags-y := -Idrivers/gpu/drm/msm ccflags-$(CONFIG_DRM_MSM_DSI) += -Idrivers/gpu/drm/msm/dsi msm-y := \ diff --git a/drivers/gpu/drm/msm/dsi/dsi.h b/drivers/gpu/drm/msm/dsi/dsi.h index 32369975d155..9e6017387efb 100644 --- a/drivers/gpu/drm/msm/dsi/dsi.h +++ b/drivers/gpu/drm/msm/dsi/dsi.h @@ -17,9 +17,9 @@ #include <linux/of_platform.h> #include <linux/platform_device.h> -#include "drm_crtc.h" -#include "drm_mipi_dsi.h" -#include "drm_panel.h" +#include <drm/drm_crtc.h> +#include <drm/drm_mipi_dsi.h> +#include <drm/drm_panel.h> #include "msm_drv.h" diff --git a/drivers/gpu/drm/msm/edp/edp.h b/drivers/gpu/drm/msm/edp/edp.h index ba5bedde5241..e0f5818ec9ca 100644 --- a/drivers/gpu/drm/msm/edp/edp.h +++ b/drivers/gpu/drm/msm/edp/edp.h @@ -18,9 +18,9 @@ #include <linux/interrupt.h> #include <linux/kernel.h> #include <linux/platform_device.h> +#include <drm/drm_crtc.h> +#include <drm/drm_dp_helper.h> -#include "drm_crtc.h" -#include "drm_dp_helper.h" #include "msm_drv.h" #define edp_read(offset) msm_readl((offset)) diff --git a/drivers/gpu/drm/msm/edp/edp_ctrl.c b/drivers/gpu/drm/msm/edp/edp_ctrl.c index 149bfe7ddd82..e32a4a4f3797 100644 --- a/drivers/gpu/drm/msm/edp/edp_ctrl.c +++ b/drivers/gpu/drm/msm/edp/edp_ctrl.c @@ -14,10 +14,10 @@ #include <linux/clk.h> #include <linux/gpio/consumer.h> #include <linux/regulator/consumer.h> +#include <drm/drm_crtc.h> +#include <drm/drm_dp_helper.h> +#include <drm/drm_edid.h> -#include "drm_crtc.h" -#include "drm_dp_helper.h" -#include "drm_edid.h" #include "edp.h" #include "edp.xml.h" diff --git a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_crtc.c b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_crtc.c index f29194a74a19..698e514203c6 100644 --- a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_crtc.c +++ b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_crtc.c @@ -15,12 +15,12 @@ * this program. If not, see <http://www.gnu.org/licenses/>. */ -#include "mdp4_kms.h" - +#include <drm/drm_crtc.h> +#include <drm/drm_crtc_helper.h> +#include <drm/drm_flip_work.h> #include <drm/drm_mode.h> -#include "drm_crtc.h" -#include "drm_crtc_helper.h" -#include "drm_flip_work.h" + +#include "mdp4_kms.h" struct mdp4_crtc { struct drm_crtc base; diff --git a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_dsi_encoder.c b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_dsi_encoder.c index 106f0e772595..6a1ebdace391 100644 --- a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_dsi_encoder.c +++ b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_dsi_encoder.c @@ -17,10 +17,10 @@ * this program. If not, see <http://www.gnu.org/licenses/>. */ -#include "mdp4_kms.h" +#include <drm/drm_crtc.h> +#include <drm/drm_crtc_helper.h> -#include "drm_crtc.h" -#include "drm_crtc_helper.h" +#include "mdp4_kms.h" struct mdp4_dsi_encoder { struct drm_encoder base; diff --git a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_dtv_encoder.c b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_dtv_encoder.c index 24258e3025e3..ba8e587f734b 100644 --- a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_dtv_encoder.c +++ b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_dtv_encoder.c @@ -15,11 +15,10 @@ * this program. If not, see <http://www.gnu.org/licenses/>. */ -#include "mdp4_kms.h" - -#include "drm_crtc.h" -#include "drm_crtc_helper.h" +#include <drm/drm_crtc.h> +#include <drm/drm_crtc_helper.h> +#include "mdp4_kms.h" struct mdp4_dtv_encoder { struct drm_encoder base; diff --git a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_kms.h b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_kms.h index 62712ca164ee..c413779d488a 100644 --- a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_kms.h +++ b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_kms.h @@ -18,12 +18,14 @@ #ifndef __MDP4_KMS_H__ #define __MDP4_KMS_H__ +#include <drm/drm_panel.h> + #include "msm_drv.h" #include "msm_kms.h" #include "mdp/mdp_kms.h" #include "mdp4.xml.h" -#include "drm_panel.h" +struct device_node; struct mdp4_kms { struct mdp_kms base; diff --git a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_lcdc_encoder.c b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_lcdc_encoder.c index a06b064f86c1..4a645926edb7 100644 --- a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_lcdc_encoder.c +++ b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_lcdc_encoder.c @@ -16,10 +16,10 @@ * this program. If not, see <http://www.gnu.org/licenses/>. */ -#include "mdp4_kms.h" +#include <drm/drm_crtc.h> +#include <drm/drm_crtc_helper.h> -#include "drm_crtc.h" -#include "drm_crtc_helper.h" +#include "mdp4_kms.h" struct mdp4_lcdc_encoder { struct drm_encoder base; diff --git a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_cmd_encoder.c b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_cmd_encoder.c index 8dafc7bdba48..aa7402e03f67 100644 --- a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_cmd_encoder.c +++ b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_cmd_encoder.c @@ -11,10 +11,10 @@ * GNU General Public License for more details. */ -#include "mdp5_kms.h" +#include <drm/drm_crtc.h> +#include <drm/drm_crtc_helper.h> -#include "drm_crtc.h" -#include "drm_crtc_helper.h" +#include "mdp5_kms.h" static struct mdp5_kms *get_kms(struct drm_encoder *encoder) { diff --git a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_crtc.c b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_crtc.c index 9217e0d6e93e..0764a6498110 100644 --- a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_crtc.c +++ b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_crtc.c @@ -16,13 +16,13 @@ * this program. If not, see <http://www.gnu.org/licenses/>. */ -#include "mdp5_kms.h" - #include <linux/sort.h> #include <drm/drm_mode.h> -#include "drm_crtc.h" -#include "drm_crtc_helper.h" -#include "drm_flip_work.h" +#include <drm/drm_crtc.h> +#include <drm/drm_crtc_helper.h> +#include <drm/drm_flip_work.h> + +#include "mdp5_kms.h" #define CURSOR_WIDTH 64 #define CURSOR_HEIGHT 64 diff --git a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_encoder.c b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_encoder.c index c2ab0f033031..97f3294fbfc6 100644 --- a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_encoder.c +++ b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_encoder.c @@ -16,10 +16,10 @@ * this program. If not, see <http://www.gnu.org/licenses/>. */ -#include "mdp5_kms.h" +#include <drm/drm_crtc.h> +#include <drm/drm_crtc_helper.h> -#include "drm_crtc.h" -#include "drm_crtc_helper.h" +#include "mdp5_kms.h" static struct mdp5_kms *get_kms(struct drm_encoder *encoder) { diff --git a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_kms.c b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_kms.c index d3d6b4cae1e6..e2b3346ead48 100644 --- a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_kms.c +++ b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_kms.c @@ -527,31 +527,28 @@ static struct drm_encoder *get_encoder_from_crtc(struct drm_crtc *crtc) return NULL; } -static int mdp5_get_scanoutpos(struct drm_device *dev, unsigned int pipe, - unsigned int flags, int *vpos, int *hpos, - ktime_t *stime, ktime_t *etime, - const struct drm_display_mode *mode) +static bool mdp5_get_scanoutpos(struct drm_device *dev, unsigned int pipe, + bool in_vblank_irq, int *vpos, int *hpos, + ktime_t *stime, ktime_t *etime, + const struct drm_display_mode *mode) { struct msm_drm_private *priv = dev->dev_private; struct drm_crtc *crtc; struct drm_encoder *encoder; int line, vsw, vbp, vactive_start, vactive_end, vfp_end; - int ret = 0; crtc = priv->crtcs[pipe]; if (!crtc) { DRM_ERROR("Invalid crtc %d\n", pipe); - return 0; + return false; } encoder = get_encoder_from_crtc(crtc); if (!encoder) { DRM_ERROR("no encoder found for crtc %d\n", pipe); - return 0; + return false; } - ret |= DRM_SCANOUTPOS_VALID | DRM_SCANOUTPOS_ACCURATE; - vsw = mode->crtc_vsync_end - mode->crtc_vsync_start; vbp = mode->crtc_vtotal - mode->crtc_vsync_end; @@ -575,10 +572,8 @@ static int mdp5_get_scanoutpos(struct drm_device *dev, unsigned int pipe, if (line < vactive_start) { line -= vactive_start; - ret |= DRM_SCANOUTPOS_IN_VBLANK; } else if (line > vactive_end) { line = line - vfp_end - vactive_start; - ret |= DRM_SCANOUTPOS_IN_VBLANK; } else { line -= vactive_start; } @@ -589,31 +584,7 @@ static int mdp5_get_scanoutpos(struct drm_device *dev, unsigned int pipe, if (etime) *etime = ktime_get(); - return ret; -} - -static int mdp5_get_vblank_timestamp(struct drm_device *dev, unsigned int pipe, - int *max_error, - struct timeval *vblank_time, - unsigned flags) -{ - struct msm_drm_private *priv = dev->dev_private; - struct drm_crtc *crtc; - - if (pipe < 0 || pipe >= priv->num_crtcs) { - DRM_ERROR("Invalid crtc %d\n", pipe); - return -EINVAL; - } - - crtc = priv->crtcs[pipe]; - if (!crtc) { - DRM_ERROR("Invalid crtc %d\n", pipe); - return -EINVAL; - } - - return drm_calc_vbltimestamp_from_scanoutpos(dev, pipe, max_error, - vblank_time, flags, - &crtc->mode); + return true; } static u32 mdp5_get_vblank_counter(struct drm_device *dev, unsigned int pipe) @@ -725,7 +696,7 @@ struct msm_kms *mdp5_kms_init(struct drm_device *dev) dev->mode_config.max_width = 0xffff; dev->mode_config.max_height = 0xffff; - dev->driver->get_vblank_timestamp = mdp5_get_vblank_timestamp; + dev->driver->get_vblank_timestamp = drm_calc_vbltimestamp_from_scanoutpos; dev->driver->get_scanout_position = mdp5_get_scanoutpos; dev->driver->get_vblank_counter = mdp5_get_vblank_counter; dev->max_vblank_count = 0xffffffff; diff --git a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_plane.c b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_plane.c index a38c5fe6cc19..5e7d9af4cba8 100644 --- a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_plane.c +++ b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_plane.c @@ -67,11 +67,11 @@ static void mdp5_plane_install_rotation_property(struct drm_device *dev, struct drm_plane *plane) { drm_plane_create_rotation_property(plane, - DRM_ROTATE_0, - DRM_ROTATE_0 | - DRM_ROTATE_180 | - DRM_REFLECT_X | - DRM_REFLECT_Y); + DRM_MODE_ROTATE_0, + DRM_MODE_ROTATE_0 | + DRM_MODE_ROTATE_180 | + DRM_MODE_REFLECT_X | + DRM_MODE_REFLECT_Y); } /* helper to install properties which are common to planes and crtcs */ @@ -369,14 +369,14 @@ static int mdp5_plane_atomic_check_with_state(struct drm_crtc_state *crtc_state, caps |= MDP_PIPE_CAP_SCALE; rotation = drm_rotation_simplify(state->rotation, - DRM_ROTATE_0 | - DRM_REFLECT_X | - DRM_REFLECT_Y); + DRM_MODE_ROTATE_0 | + DRM_MODE_REFLECT_X | + DRM_MODE_REFLECT_Y); - if (rotation & DRM_REFLECT_X) + if (rotation & DRM_MODE_REFLECT_X) caps |= MDP_PIPE_CAP_HFLIP; - if (rotation & DRM_REFLECT_Y) + if (rotation & DRM_MODE_REFLECT_Y) caps |= MDP_PIPE_CAP_VFLIP; if (plane->type == DRM_PLANE_TYPE_CURSOR) @@ -970,11 +970,11 @@ static int mdp5_plane_mode_set(struct drm_plane *plane, DBG("scale config = %x", config); rotation = drm_rotation_simplify(pstate->rotation, - DRM_ROTATE_0 | - DRM_REFLECT_X | - DRM_REFLECT_Y); - hflip = !!(rotation & DRM_REFLECT_X); - vflip = !!(rotation & DRM_REFLECT_Y); + DRM_MODE_ROTATE_0 | + DRM_MODE_REFLECT_X | + DRM_MODE_REFLECT_Y); + hflip = !!(rotation & DRM_MODE_REFLECT_X); + vflip = !!(rotation & DRM_MODE_REFLECT_Y); spin_lock_irqsave(&mdp5_plane->pipe_lock, flags); diff --git a/drivers/gpu/drm/msm/msm_fb.c b/drivers/gpu/drm/msm/msm_fb.c index 5cf165c9c3a9..ba2733a95a4f 100644 --- a/drivers/gpu/drm/msm/msm_fb.c +++ b/drivers/gpu/drm/msm/msm_fb.c @@ -15,12 +15,12 @@ * this program. If not, see <http://www.gnu.org/licenses/>. */ +#include <drm/drm_crtc.h> +#include <drm/drm_crtc_helper.h> + #include "msm_drv.h" #include "msm_kms.h" -#include "drm_crtc.h" -#include "drm_crtc_helper.h" - struct msm_framebuffer { struct drm_framebuffer base; const struct msm_format *format; diff --git a/drivers/gpu/drm/msm/msm_fbdev.c b/drivers/gpu/drm/msm/msm_fbdev.c index 951e40faf6e8..feea8ba4e05b 100644 --- a/drivers/gpu/drm/msm/msm_fbdev.c +++ b/drivers/gpu/drm/msm/msm_fbdev.c @@ -15,10 +15,10 @@ * this program. If not, see <http://www.gnu.org/licenses/>. */ -#include "msm_drv.h" +#include <drm/drm_crtc.h> +#include <drm/drm_fb_helper.h> -#include "drm_crtc.h" -#include "drm_fb_helper.h" +#include "msm_drv.h" #include "msm_gem.h" extern int msm_gem_mmap_obj(struct drm_gem_object *obj, diff --git a/drivers/gpu/drm/msm/msm_gem.c b/drivers/gpu/drm/msm/msm_gem.c index 68e509b3b9e4..465dab942afa 100644 --- a/drivers/gpu/drm/msm/msm_gem.c +++ b/drivers/gpu/drm/msm/msm_gem.c @@ -50,13 +50,13 @@ static struct page **get_pages_vram(struct drm_gem_object *obj, struct page **p; int ret, i; - p = drm_malloc_ab(npages, sizeof(struct page *)); + p = kvmalloc_array(npages, sizeof(struct page *), GFP_KERNEL); if (!p) return ERR_PTR(-ENOMEM); ret = drm_mm_insert_node(&priv->vram.mm, msm_obj->vram_node, npages); if (ret) { - drm_free_large(p); + kvfree(p); return ERR_PTR(ret); } @@ -127,7 +127,7 @@ static void put_pages(struct drm_gem_object *obj) drm_gem_put_pages(obj, msm_obj->pages, true, false); else { drm_mm_remove_node(msm_obj->vram_node); - drm_free_large(msm_obj->pages); + kvfree(msm_obj->pages); } msm_obj->pages = NULL; @@ -707,7 +707,7 @@ void msm_gem_free_object(struct drm_gem_object *obj) * ours, just free the array we allocated: */ if (msm_obj->pages) - drm_free_large(msm_obj->pages); + kvfree(msm_obj->pages); drm_prime_gem_destroy(obj, msm_obj->sgt); } else { @@ -863,7 +863,7 @@ struct drm_gem_object *msm_gem_import(struct drm_device *dev, msm_obj = to_msm_bo(obj); msm_obj->sgt = sgt; - msm_obj->pages = drm_malloc_ab(npages, sizeof(struct page *)); + msm_obj->pages = kvmalloc_array(npages, sizeof(struct page *), GFP_KERNEL); if (!msm_obj->pages) { ret = -ENOMEM; goto fail; diff --git a/drivers/gpu/drm/nouveau/Kbuild b/drivers/gpu/drm/nouveau/Kbuild index fde6e3656636..2e9ce53ae3a8 100644 --- a/drivers/gpu/drm/nouveau/Kbuild +++ b/drivers/gpu/drm/nouveau/Kbuild @@ -1,4 +1,3 @@ -ccflags-y := -Iinclude/drm ccflags-y += -I$(src)/include ccflags-y += -I$(src)/include/nvkm ccflags-y += -I$(src)/nvkm diff --git a/drivers/gpu/drm/nouveau/nouveau_display.c b/drivers/gpu/drm/nouveau/nouveau_display.c index 21b10f9840c9..6718c84fb862 100644 --- a/drivers/gpu/drm/nouveau/nouveau_display.c +++ b/drivers/gpu/drm/nouveau/nouveau_display.c @@ -98,7 +98,7 @@ calc(int blanks, int blanke, int total, int line) return line; } -static int +static bool nouveau_display_scanoutpos_head(struct drm_crtc *crtc, int *vpos, int *hpos, ktime_t *stime, ktime_t *etime) { @@ -111,16 +111,16 @@ nouveau_display_scanoutpos_head(struct drm_crtc *crtc, int *vpos, int *hpos, }; struct nouveau_display *disp = nouveau_display(crtc->dev); struct drm_vblank_crtc *vblank = &crtc->dev->vblank[drm_crtc_index(crtc)]; - int ret, retry = 20; + int retry = 20; + bool ret = false; do { ret = nvif_mthd(&disp->disp, 0, &args, sizeof(args)); if (ret != 0) - return 0; + return false; if (args.scan.vline) { - ret |= DRM_SCANOUTPOS_ACCURATE; - ret |= DRM_SCANOUTPOS_VALID; + ret = true; break; } @@ -133,14 +133,12 @@ nouveau_display_scanoutpos_head(struct drm_crtc *crtc, int *vpos, int *hpos, if (stime) *stime = ns_to_ktime(args.scan.time[0]); if (etime) *etime = ns_to_ktime(args.scan.time[1]); - if (*vpos < 0) - ret |= DRM_SCANOUTPOS_IN_VBLANK; return ret; } -int +bool nouveau_display_scanoutpos(struct drm_device *dev, unsigned int pipe, - unsigned int flags, int *vpos, int *hpos, + bool in_vblank_irq, int *vpos, int *hpos, ktime_t *stime, ktime_t *etime, const struct drm_display_mode *mode) { @@ -153,28 +151,7 @@ nouveau_display_scanoutpos(struct drm_device *dev, unsigned int pipe, } } - return 0; -} - -int -nouveau_display_vblstamp(struct drm_device *dev, unsigned int pipe, - int *max_error, struct timeval *time, unsigned flags) -{ - struct drm_crtc *crtc; - - list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) { - if (nouveau_crtc(crtc)->index == pipe) { - struct drm_display_mode *mode; - if (drm_drv_uses_atomic_modeset(dev)) - mode = &crtc->state->adjusted_mode; - else - mode = &crtc->hwmode; - return drm_calc_vbltimestamp_from_scanoutpos(dev, - pipe, max_error, time, flags, mode); - } - } - - return -EINVAL; + return false; } static void diff --git a/drivers/gpu/drm/nouveau/nouveau_display.h b/drivers/gpu/drm/nouveau/nouveau_display.h index e1d772d39488..201aec2ea5b8 100644 --- a/drivers/gpu/drm/nouveau/nouveau_display.h +++ b/drivers/gpu/drm/nouveau/nouveau_display.h @@ -68,11 +68,9 @@ int nouveau_display_suspend(struct drm_device *dev, bool runtime); void nouveau_display_resume(struct drm_device *dev, bool runtime); int nouveau_display_vblank_enable(struct drm_device *, unsigned int); void nouveau_display_vblank_disable(struct drm_device *, unsigned int); -int nouveau_display_scanoutpos(struct drm_device *, unsigned int, - unsigned int, int *, int *, ktime_t *, - ktime_t *, const struct drm_display_mode *); -int nouveau_display_vblstamp(struct drm_device *, unsigned int, int *, - struct timeval *, unsigned); +bool nouveau_display_scanoutpos(struct drm_device *, unsigned int, + bool, int *, int *, ktime_t *, + ktime_t *, const struct drm_display_mode *); int nouveau_crtc_page_flip(struct drm_crtc *crtc, struct drm_framebuffer *fb, struct drm_pending_vblank_event *event, diff --git a/drivers/gpu/drm/nouveau/nouveau_drm.c b/drivers/gpu/drm/nouveau/nouveau_drm.c index 2b6ac24ce690..c3dc75fee700 100644 --- a/drivers/gpu/drm/nouveau/nouveau_drm.c +++ b/drivers/gpu/drm/nouveau/nouveau_drm.c @@ -29,8 +29,8 @@ #include <linux/pm_runtime.h> #include <linux/vga_switcheroo.h> -#include "drmP.h" -#include "drm_crtc_helper.h" +#include <drm/drmP.h> +#include <drm/drm_crtc_helper.h> #include <core/gpuobj.h> #include <core/option.h> @@ -881,7 +881,7 @@ done: } static void -nouveau_drm_preclose(struct drm_device *dev, struct drm_file *fpriv) +nouveau_drm_postclose(struct drm_device *dev, struct drm_file *fpriv) { struct nouveau_cli *cli = nouveau_cli(fpriv); struct nouveau_drm *drm = nouveau_drm(dev); @@ -897,12 +897,6 @@ nouveau_drm_preclose(struct drm_device *dev, struct drm_file *fpriv) list_del(&cli->head); mutex_unlock(&drm->client.mutex); -} - -static void -nouveau_drm_postclose(struct drm_device *dev, struct drm_file *fpriv) -{ - struct nouveau_cli *cli = nouveau_cli(fpriv); nouveau_cli_fini(cli); kfree(cli); pm_runtime_mark_last_busy(dev->dev); @@ -974,7 +968,6 @@ driver_stub = { .load = nouveau_drm_load, .unload = nouveau_drm_unload, .open = nouveau_drm_open, - .preclose = nouveau_drm_preclose, .postclose = nouveau_drm_postclose, .lastclose = nouveau_vga_lastclose, @@ -985,7 +978,7 @@ driver_stub = { .enable_vblank = nouveau_display_vblank_enable, .disable_vblank = nouveau_display_vblank_disable, .get_scanout_position = nouveau_display_scanoutpos, - .get_vblank_timestamp = nouveau_display_vblstamp, + .get_vblank_timestamp = drm_calc_vbltimestamp_from_scanoutpos, .ioctls = nouveau_ioctls, .num_ioctls = ARRAY_SIZE(nouveau_ioctls), diff --git a/drivers/gpu/drm/nouveau/nouveau_drv.h b/drivers/gpu/drm/nouveau/nouveau_drv.h index eadec2f49ad3..aaa25641fed6 100644 --- a/drivers/gpu/drm/nouveau/nouveau_drv.h +++ b/drivers/gpu/drm/nouveau/nouveau_drv.h @@ -43,7 +43,7 @@ #include <nvif/device.h> #include <nvif/ioctl.h> -#include <drmP.h> +#include <drm/drmP.h> #include <drm/ttm/ttm_bo_api.h> #include <drm/ttm/ttm_bo_driver.h> diff --git a/drivers/gpu/drm/nouveau/nouveau_gem.c b/drivers/gpu/drm/nouveau/nouveau_gem.c index ca5397beb357..2170534101ca 100644 --- a/drivers/gpu/drm/nouveau/nouveau_gem.c +++ b/drivers/gpu/drm/nouveau/nouveau_gem.c @@ -568,9 +568,7 @@ u_memcpya(uint64_t user, unsigned nmemb, unsigned size) size *= nmemb; - mem = kmalloc(size, GFP_KERNEL | __GFP_NOWARN); - if (!mem) - mem = vmalloc(size); + mem = kvmalloc(size, GFP_KERNEL); if (!mem) return ERR_PTR(-ENOMEM); diff --git a/drivers/gpu/drm/nouveau/nouveau_ttm.c b/drivers/gpu/drm/nouveau/nouveau_ttm.c index 13e5cc5f07fe..999c35a25498 100644 --- a/drivers/gpu/drm/nouveau/nouveau_ttm.c +++ b/drivers/gpu/drm/nouveau/nouveau_ttm.c @@ -28,7 +28,7 @@ #include "nouveau_ttm.h" #include "nouveau_gem.h" -#include "drm_legacy.h" +#include <drm/drm_legacy.h> #include <core/tegra.h> diff --git a/drivers/gpu/drm/nouveau/nv50_display.c b/drivers/gpu/drm/nouveau/nv50_display.c index 0e58537352fe..9303daa79aba 100644 --- a/drivers/gpu/drm/nouveau/nv50_display.c +++ b/drivers/gpu/drm/nouveau/nv50_display.c @@ -831,8 +831,7 @@ nv50_wndw_atomic_check_release(struct nv50_wndw *wndw, static int nv50_wndw_atomic_check_acquire(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw, - struct nv50_head_atom *asyh, - u32 pflip_flags) + struct nv50_head_atom *asyh) { struct nouveau_framebuffer *fb = nouveau_framebuffer(asyw->state.fb); struct nouveau_drm *drm = nouveau_drm(wndw->plane.dev); @@ -848,7 +847,10 @@ nv50_wndw_atomic_check_acquire(struct nv50_wndw *wndw, asyw->image.h = fb->base.height; asyw->image.kind = (fb->nvbo->tile_flags & 0x0000ff00) >> 8; - asyw->interval = pflip_flags & DRM_MODE_PAGE_FLIP_ASYNC ? 0 : 1; + if (asyh->state.pageflip_flags & DRM_MODE_PAGE_FLIP_ASYNC) + asyw->interval = 0; + else + asyw->interval = 1; if (asyw->image.kind) { asyw->image.layout = 0; @@ -887,7 +889,6 @@ nv50_wndw_atomic_check(struct drm_plane *plane, struct drm_plane_state *state) struct nv50_head_atom *harm = NULL, *asyh = NULL; bool varm = false, asyv = false, asym = false; int ret; - u32 pflip_flags = 0; NV_ATOMIC(drm, "%s atomic_check\n", plane->name); if (asyw->state.crtc) { @@ -896,7 +897,6 @@ nv50_wndw_atomic_check(struct drm_plane *plane, struct drm_plane_state *state) return PTR_ERR(asyh); asym = drm_atomic_crtc_needs_modeset(&asyh->state); asyv = asyh->state.active; - pflip_flags = asyh->state.pageflip_flags; } if (armw->state.crtc) { @@ -912,12 +912,9 @@ nv50_wndw_atomic_check(struct drm_plane *plane, struct drm_plane_state *state) if (memcmp(&armw->point, &asyw->point, sizeof(asyw->point))) asyw->set.point = true; - if (!varm || asym || armw->state.fb != asyw->state.fb) { - ret = nv50_wndw_atomic_check_acquire( - wndw, asyw, asyh, pflip_flags); - if (ret) - return ret; - } + ret = nv50_wndw_atomic_check_acquire(wndw, asyw, asyh); + if (ret) + return ret; } else if (varm) { nv50_wndw_atomic_check_release(wndw, asyw, harm); @@ -1036,7 +1033,7 @@ nv50_wndw_reset(struct drm_plane *plane) plane->funcs->atomic_destroy_state(plane, plane->state); plane->state = &asyw->state; plane->state->plane = plane; - plane->state->rotation = DRM_ROTATE_0; + plane->state->rotation = DRM_MODE_ROTATE_0; } static void @@ -1122,9 +1119,13 @@ static void nv50_curs_prepare(struct nv50_wndw *wndw, struct nv50_head_atom *asyh, struct nv50_wndw_atom *asyw) { - asyh->curs.handle = nv50_disp(wndw->plane.dev)->mast.base.vram.handle; - asyh->curs.offset = asyw->image.offset; - asyh->set.curs = asyh->curs.visible; + u32 handle = nv50_disp(wndw->plane.dev)->mast.base.vram.handle; + u32 offset = asyw->image.offset; + if (asyh->curs.handle != handle || asyh->curs.offset != offset) { + asyh->curs.handle = handle; + asyh->curs.offset = offset; + asyh->set.curs = asyh->curs.visible; + } } static void diff --git a/drivers/gpu/drm/nouveau/nvkm/core/object.c b/drivers/gpu/drm/nouveau/nvkm/core/object.c index 89d2e9da11c7..acd76fd4f6d8 100644 --- a/drivers/gpu/drm/nouveau/nvkm/core/object.c +++ b/drivers/gpu/drm/nouveau/nvkm/core/object.c @@ -295,7 +295,7 @@ nvkm_object_ctor(const struct nvkm_object_func *func, INIT_LIST_HEAD(&object->head); INIT_LIST_HEAD(&object->tree); RB_CLEAR_NODE(&object->node); - WARN_ON(oclass->engine && !object->engine); + WARN_ON(IS_ERR(object->engine)); } int diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramgf100.c b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramgf100.c index c63975907c90..4a9bd4f1cb93 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramgf100.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramgf100.c @@ -638,7 +638,6 @@ gf100_ram_ctor(const struct nvkm_ram_func *func, struct nvkm_fb *fb, return ret; } - ram->ranks = (nvkm_rd32(device, 0x10f200) & 0x00000004) ? 2 : 1; return 0; } diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/therm/base.c b/drivers/gpu/drm/nouveau/nvkm/subdev/therm/base.c index df949fa7d05d..be691a7b972f 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/therm/base.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/therm/base.c @@ -146,7 +146,7 @@ nvkm_therm_update(struct nvkm_therm *therm, int mode) poll = false; } - if (list_empty(&therm->alarm.head) && poll) + if (poll) nvkm_timer_alarm(tmr, 1000000000ULL, &therm->alarm); spin_unlock_irqrestore(&therm->lock, flags); diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/therm/fan.c b/drivers/gpu/drm/nouveau/nvkm/subdev/therm/fan.c index 91198d79393a..e2feccec25f5 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/therm/fan.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/therm/fan.c @@ -83,7 +83,7 @@ nvkm_fan_update(struct nvkm_fan *fan, bool immediate, int target) spin_unlock_irqrestore(&fan->lock, flags); /* schedule next fan update, if not at target speed already */ - if (list_empty(&fan->alarm.head) && target != duty) { + if (target != duty) { u16 bump_period = fan->bios.bump_period; u16 slow_down_period = fan->bios.slow_down_period; u64 delay; diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/therm/fantog.c b/drivers/gpu/drm/nouveau/nvkm/subdev/therm/fantog.c index 59701b7a6597..ff9fbe7950e5 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/therm/fantog.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/therm/fantog.c @@ -53,7 +53,7 @@ nvkm_fantog_update(struct nvkm_fantog *fan, int percent) duty = !nvkm_gpio_get(gpio, 0, DCB_GPIO_FAN, 0xff); nvkm_gpio_set(gpio, 0, DCB_GPIO_FAN, 0xff, duty); - if (list_empty(&fan->alarm.head) && percent != (duty * 100)) { + if (percent != (duty * 100)) { u64 next_change = (percent * fan->period_us) / 100; if (!duty) next_change = fan->period_us - next_change; diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/therm/temp.c b/drivers/gpu/drm/nouveau/nvkm/subdev/therm/temp.c index b9703c02d8ca..9a79e91fdfdc 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/therm/temp.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/therm/temp.c @@ -185,7 +185,7 @@ alarm_timer_callback(struct nvkm_alarm *alarm) spin_unlock_irqrestore(&therm->sensor.alarm_program_lock, flags); /* schedule the next poll in one second */ - if (therm->func->temp_get(therm) >= 0 && list_empty(&alarm->head)) + if (therm->func->temp_get(therm) >= 0) nvkm_timer_alarm(tmr, 1000000000ULL, alarm); } diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/timer/base.c b/drivers/gpu/drm/nouveau/nvkm/subdev/timer/base.c index 07dc82bfe346..f2a86eae0a0d 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/timer/base.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/timer/base.c @@ -36,23 +36,29 @@ nvkm_timer_alarm_trigger(struct nvkm_timer *tmr) unsigned long flags; LIST_HEAD(exec); - /* move any due alarms off the pending list */ + /* Process pending alarms. */ spin_lock_irqsave(&tmr->lock, flags); list_for_each_entry_safe(alarm, atemp, &tmr->alarms, head) { - if (alarm->timestamp <= nvkm_timer_read(tmr)) - list_move_tail(&alarm->head, &exec); + /* Have we hit the earliest alarm that hasn't gone off? */ + if (alarm->timestamp > nvkm_timer_read(tmr)) { + /* Schedule it. If we didn't race, we're done. */ + tmr->func->alarm_init(tmr, alarm->timestamp); + if (alarm->timestamp > nvkm_timer_read(tmr)) + break; + } + + /* Move to completed list. We'll drop the lock before + * executing the callback so it can reschedule itself. + */ + list_move_tail(&alarm->head, &exec); } - /* reschedule interrupt for next alarm time */ - if (!list_empty(&tmr->alarms)) { - alarm = list_first_entry(&tmr->alarms, typeof(*alarm), head); - tmr->func->alarm_init(tmr, alarm->timestamp); - } else { + /* Shut down interrupt if no more pending alarms. */ + if (list_empty(&tmr->alarms)) tmr->func->alarm_fini(tmr); - } spin_unlock_irqrestore(&tmr->lock, flags); - /* execute any pending alarm handlers */ + /* Execute completed callbacks. */ list_for_each_entry_safe(alarm, atemp, &exec, head) { list_del_init(&alarm->head); alarm->func(alarm); @@ -65,24 +71,37 @@ nvkm_timer_alarm(struct nvkm_timer *tmr, u32 nsec, struct nvkm_alarm *alarm) struct nvkm_alarm *list; unsigned long flags; - alarm->timestamp = nvkm_timer_read(tmr) + nsec; - - /* append new alarm to list, in soonest-alarm-first order */ + /* Remove alarm from pending list. + * + * This both protects against the corruption of the list, + * and implements alarm rescheduling/cancellation. + */ spin_lock_irqsave(&tmr->lock, flags); - if (!nsec) { - if (!list_empty(&alarm->head)) - list_del(&alarm->head); - } else { + list_del_init(&alarm->head); + + if (nsec) { + /* Insert into pending list, ordered earliest to latest. */ + alarm->timestamp = nvkm_timer_read(tmr) + nsec; list_for_each_entry(list, &tmr->alarms, head) { if (list->timestamp > alarm->timestamp) break; } + list_add_tail(&alarm->head, &list->head); + + /* Update HW if this is now the earliest alarm. */ + list = list_first_entry(&tmr->alarms, typeof(*list), head); + if (list == alarm) { + tmr->func->alarm_init(tmr, alarm->timestamp); + /* This shouldn't happen if callers aren't stupid. + * + * Worst case scenario is that it'll take roughly + * 4 seconds for the next alarm to trigger. + */ + WARN_ON(alarm->timestamp <= nvkm_timer_read(tmr)); + } } spin_unlock_irqrestore(&tmr->lock, flags); - - /* process pending alarms */ - nvkm_timer_alarm_trigger(tmr); } void diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/timer/nv04.c b/drivers/gpu/drm/nouveau/nvkm/subdev/timer/nv04.c index 7b9ce87f0617..7f48249f41de 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/timer/nv04.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/timer/nv04.c @@ -76,8 +76,8 @@ nv04_timer_intr(struct nvkm_timer *tmr) u32 stat = nvkm_rd32(device, NV04_PTIMER_INTR_0); if (stat & 0x00000001) { - nvkm_timer_alarm_trigger(tmr); nvkm_wr32(device, NV04_PTIMER_INTR_0, 0x00000001); + nvkm_timer_alarm_trigger(tmr); stat &= ~0x00000001; } diff --git a/drivers/gpu/drm/omapdrm/Makefile b/drivers/gpu/drm/omapdrm/Makefile index 48b7b750c05c..b391be7ecb6c 100644 --- a/drivers/gpu/drm/omapdrm/Makefile +++ b/drivers/gpu/drm/omapdrm/Makefile @@ -6,7 +6,6 @@ obj-y += dss/ obj-y += displays/ -ccflags-y := -Iinclude/drm omapdrm-y := omap_drv.o \ omap_irq.o \ omap_debugfs.o \ diff --git a/drivers/gpu/drm/omapdrm/omap_drv.c b/drivers/gpu/drm/omapdrm/omap_drv.c index e1f47f0b3ccf..663e930a7b0f 100644 --- a/drivers/gpu/drm/omapdrm/omap_drv.c +++ b/drivers/gpu/drm/omapdrm/omap_drv.c @@ -577,7 +577,7 @@ static void dev_lastclose(struct drm_device *dev) drm_object_property_set_value(&crtc->base, crtc->primary->rotation_property, - DRM_ROTATE_0); + DRM_MODE_ROTATE_0); } for (i = 0; i < priv->num_planes; i++) { @@ -588,7 +588,7 @@ static void dev_lastclose(struct drm_device *dev) drm_object_property_set_value(&plane->base, plane->rotation_property, - DRM_ROTATE_0); + DRM_MODE_ROTATE_0); } if (priv->fbdev) { diff --git a/drivers/gpu/drm/omapdrm/omap_fb.c b/drivers/gpu/drm/omapdrm/omap_fb.c index 29dc677dd4d3..5ca0537bb427 100644 --- a/drivers/gpu/drm/omapdrm/omap_fb.c +++ b/drivers/gpu/drm/omapdrm/omap_fb.c @@ -167,30 +167,30 @@ void omap_framebuffer_update_scanout(struct drm_framebuffer *fb, uint32_t w = win->src_w; uint32_t h = win->src_h; - switch (win->rotation & DRM_ROTATE_MASK) { + switch (win->rotation & DRM_MODE_ROTATE_MASK) { default: dev_err(fb->dev->dev, "invalid rotation: %02x", (uint32_t)win->rotation); /* fallthru to default to no rotation */ case 0: - case DRM_ROTATE_0: + case DRM_MODE_ROTATE_0: orient = 0; break; - case DRM_ROTATE_90: + case DRM_MODE_ROTATE_90: orient = MASK_XY_FLIP | MASK_X_INVERT; break; - case DRM_ROTATE_180: + case DRM_MODE_ROTATE_180: orient = MASK_X_INVERT | MASK_Y_INVERT; break; - case DRM_ROTATE_270: + case DRM_MODE_ROTATE_270: orient = MASK_XY_FLIP | MASK_Y_INVERT; break; } - if (win->rotation & DRM_REFLECT_X) + if (win->rotation & DRM_MODE_REFLECT_X) orient ^= MASK_X_INVERT; - if (win->rotation & DRM_REFLECT_Y) + if (win->rotation & DRM_MODE_REFLECT_Y) orient ^= MASK_Y_INVERT; /* adjust x,y offset for flip/invert: */ @@ -205,9 +205,9 @@ void omap_framebuffer_update_scanout(struct drm_framebuffer *fb, info->rotation_type = OMAP_DSS_ROT_TILER; info->screen_width = omap_gem_tiled_stride(plane->bo, orient); } else { - switch (win->rotation & DRM_ROTATE_MASK) { + switch (win->rotation & DRM_MODE_ROTATE_MASK) { case 0: - case DRM_ROTATE_0: + case DRM_MODE_ROTATE_0: /* OK */ break; diff --git a/drivers/gpu/drm/omapdrm/omap_plane.c b/drivers/gpu/drm/omapdrm/omap_plane.c index 9168154d749e..d3d6818c68f8 100644 --- a/drivers/gpu/drm/omapdrm/omap_plane.c +++ b/drivers/gpu/drm/omapdrm/omap_plane.c @@ -141,7 +141,7 @@ static void omap_plane_atomic_disable(struct drm_plane *plane, struct omap_plane_state *omap_state = to_omap_plane_state(plane->state); struct omap_plane *omap_plane = to_omap_plane(plane); - plane->state->rotation = DRM_ROTATE_0; + plane->state->rotation = DRM_MODE_ROTATE_0; omap_state->zorder = plane->type == DRM_PLANE_TYPE_PRIMARY ? 0 : omap_plane->id; @@ -177,7 +177,7 @@ static int omap_plane_atomic_check(struct drm_plane *plane, if (state->crtc_y + state->crtc_h > crtc_state->adjusted_mode.vdisplay) return -EINVAL; - if (state->rotation != DRM_ROTATE_0 && + if (state->rotation != DRM_MODE_ROTATE_0 && !omap_framebuffer_supports_rotation(state->fb)) return -EINVAL; @@ -213,15 +213,15 @@ void omap_plane_install_properties(struct drm_plane *plane, if (priv->has_dmm) { if (!plane->rotation_property) drm_plane_create_rotation_property(plane, - DRM_ROTATE_0, - DRM_ROTATE_0 | DRM_ROTATE_90 | - DRM_ROTATE_180 | DRM_ROTATE_270 | - DRM_REFLECT_X | DRM_REFLECT_Y); + DRM_MODE_ROTATE_0, + DRM_MODE_ROTATE_0 | DRM_MODE_ROTATE_90 | + DRM_MODE_ROTATE_180 | DRM_MODE_ROTATE_270 | + DRM_MODE_REFLECT_X | DRM_MODE_REFLECT_Y); /* Attach the rotation property also to the crtc object */ if (plane->rotation_property && obj != &plane->base) drm_object_attach_property(obj, plane->rotation_property, - DRM_ROTATE_0); + DRM_MODE_ROTATE_0); } drm_object_attach_property(obj, priv->zorder_prop, 0); @@ -273,7 +273,7 @@ static void omap_plane_reset(struct drm_plane *plane) */ omap_state->zorder = plane->type == DRM_PLANE_TYPE_PRIMARY ? 0 : omap_plane->id; - omap_state->base.rotation = DRM_ROTATE_0; + omap_state->base.rotation = DRM_MODE_ROTATE_0; plane->state = &omap_state->base; plane->state->plane = plane; diff --git a/drivers/gpu/drm/pl111/Kconfig b/drivers/gpu/drm/pl111/Kconfig new file mode 100644 index 000000000000..309f4fd52de7 --- /dev/null +++ b/drivers/gpu/drm/pl111/Kconfig @@ -0,0 +1,13 @@ +config DRM_PL111 + tristate "DRM Support for PL111 CLCD Controller" + depends on DRM + depends on ARM || ARM64 || COMPILE_TEST + depends on COMMON_CLK + select DRM_KMS_HELPER + select DRM_KMS_CMA_HELPER + select DRM_GEM_CMA_HELPER + select VT_HW_CONSOLE_BINDING if FRAMEBUFFER_CONSOLE + help + Choose this option for DRM support for the PL111 CLCD controller. + If M is selected the module will be called pl111_drm. + diff --git a/drivers/gpu/drm/pl111/Makefile b/drivers/gpu/drm/pl111/Makefile new file mode 100644 index 000000000000..59483d610ef5 --- /dev/null +++ b/drivers/gpu/drm/pl111/Makefile @@ -0,0 +1,7 @@ +pl111_drm-y += pl111_connector.o \ + pl111_display.o \ + pl111_drv.o + +pl111_drm-$(CONFIG_DEBUG_FS) += pl111_debugfs.o + +obj-$(CONFIG_DRM_PL111) += pl111_drm.o diff --git a/drivers/gpu/drm/pl111/pl111_connector.c b/drivers/gpu/drm/pl111/pl111_connector.c new file mode 100644 index 000000000000..3f213d7e7692 --- /dev/null +++ b/drivers/gpu/drm/pl111/pl111_connector.c @@ -0,0 +1,127 @@ +/* + * (C) COPYRIGHT 2012-2013 ARM Limited. All rights reserved. + * + * Parts of this file were based on sources as follows: + * + * Copyright (c) 2006-2008 Intel Corporation + * Copyright (c) 2007 Dave Airlie <airlied@linux.ie> + * Copyright (C) 2011 Texas Instruments + * + * This program is free software and is provided to you under the terms of the + * GNU General Public License version 2 as published by the Free Software + * Foundation, and any use by you of this program is subject to the terms of + * such GNU licence. + * + */ + +/** + * pl111_drm_connector.c + * Implementation of the connector functions for PL111 DRM + */ +#include <linux/amba/clcd-regs.h> +#include <linux/version.h> +#include <linux/shmem_fs.h> +#include <linux/dma-buf.h> + +#include <drm/drmP.h> +#include <drm/drm_atomic_helper.h> +#include <drm/drm_crtc_helper.h> +#include <drm/drm_of.h> +#include <drm/drm_panel.h> + +#include "pl111_drm.h" + +static void pl111_connector_destroy(struct drm_connector *connector) +{ + struct pl111_drm_connector *pl111_connector = + to_pl111_connector(connector); + + if (pl111_connector->panel) + drm_panel_detach(pl111_connector->panel); + + drm_connector_unregister(connector); + drm_connector_cleanup(connector); +} + +static enum drm_connector_status pl111_connector_detect(struct drm_connector + *connector, bool force) +{ + struct pl111_drm_connector *pl111_connector = + to_pl111_connector(connector); + + return (pl111_connector->panel ? + connector_status_connected : + connector_status_disconnected); +} + +static int pl111_connector_helper_get_modes(struct drm_connector *connector) +{ + struct pl111_drm_connector *pl111_connector = + to_pl111_connector(connector); + + if (!pl111_connector->panel) + return 0; + + return drm_panel_get_modes(pl111_connector->panel); +} + +const struct drm_connector_funcs connector_funcs = { + .fill_modes = drm_helper_probe_single_connector_modes, + .destroy = pl111_connector_destroy, + .detect = pl111_connector_detect, + .dpms = drm_atomic_helper_connector_dpms, + .reset = drm_atomic_helper_connector_reset, + .atomic_duplicate_state = drm_atomic_helper_connector_duplicate_state, + .atomic_destroy_state = drm_atomic_helper_connector_destroy_state, +}; + +const struct drm_connector_helper_funcs connector_helper_funcs = { + .get_modes = pl111_connector_helper_get_modes, +}; + +/* Walks the OF graph to find the panel node and then asks DRM to look + * up the panel. + */ +static struct drm_panel *pl111_get_panel(struct device *dev) +{ + struct device_node *endpoint, *panel_node; + struct device_node *np = dev->of_node; + struct drm_panel *panel; + + endpoint = of_graph_get_next_endpoint(np, NULL); + if (!endpoint) { + dev_err(dev, "no endpoint to fetch panel\n"); + return NULL; + } + + /* don't proceed if we have an endpoint but no panel_node tied to it */ + panel_node = of_graph_get_remote_port_parent(endpoint); + of_node_put(endpoint); + if (!panel_node) { + dev_err(dev, "no valid panel node\n"); + return NULL; + } + + panel = of_drm_find_panel(panel_node); + of_node_put(panel_node); + + return panel; +} + +int pl111_connector_init(struct drm_device *dev) +{ + struct pl111_drm_dev_private *priv = dev->dev_private; + struct pl111_drm_connector *pl111_connector = &priv->connector; + struct drm_connector *connector = &pl111_connector->connector; + + drm_connector_init(dev, connector, &connector_funcs, + DRM_MODE_CONNECTOR_DPI); + drm_connector_helper_add(connector, &connector_helper_funcs); + + pl111_connector->panel = pl111_get_panel(dev->dev); + if (pl111_connector->panel) + drm_panel_attach(pl111_connector->panel, connector); + + return 0; +} + diff --git a/drivers/gpu/drm/pl111/pl111_debugfs.c b/drivers/gpu/drm/pl111/pl111_debugfs.c new file mode 100644 index 000000000000..0d9dee199b2c --- /dev/null +++ b/drivers/gpu/drm/pl111/pl111_debugfs.c @@ -0,0 +1,55 @@ +/* + * Copyright © 2017 Broadcom + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ + +#include <linux/amba/clcd-regs.h> +#include <linux/seq_file.h> +#include <drm/drm_debugfs.h> +#include <drm/drmP.h> +#include "pl111_drm.h" + +#define REGDEF(reg) { reg, #reg } +static const struct { + u32 reg; + const char *name; +} pl111_reg_defs[] = { + REGDEF(CLCD_TIM0), + REGDEF(CLCD_TIM1), + REGDEF(CLCD_TIM2), + REGDEF(CLCD_TIM3), + REGDEF(CLCD_UBAS), + REGDEF(CLCD_PL111_CNTL), + REGDEF(CLCD_PL111_IENB), +}; + +int pl111_debugfs_regs(struct seq_file *m, void *unused) +{ + struct drm_info_node *node = (struct drm_info_node *)m->private; + struct drm_device *dev = node->minor->dev; + struct pl111_drm_dev_private *priv = dev->dev_private; + int i; + + for (i = 0; i < ARRAY_SIZE(pl111_reg_defs); i++) { + seq_printf(m, "%s (0x%04x): 0x%08x\n", + pl111_reg_defs[i].name, pl111_reg_defs[i].reg, + readl(priv->regs + pl111_reg_defs[i].reg)); + } + + return 0; +} + +static const struct drm_info_list pl111_debugfs_list[] = { + {"regs", pl111_debugfs_regs, 0}, +}; + +int +pl111_debugfs_init(struct drm_minor *minor) +{ + return drm_debugfs_create_files(pl111_debugfs_list, + ARRAY_SIZE(pl111_debugfs_list), + minor->debugfs_root, minor); +} diff --git a/drivers/gpu/drm/pl111/pl111_display.c b/drivers/gpu/drm/pl111/pl111_display.c new file mode 100644 index 000000000000..3e0a4fa73ddb --- /dev/null +++ b/drivers/gpu/drm/pl111/pl111_display.c @@ -0,0 +1,476 @@ +/* + * (C) COPYRIGHT 2012-2013 ARM Limited. All rights reserved. + * + * Parts of this file were based on sources as follows: + * + * Copyright (c) 2006-2008 Intel Corporation + * Copyright (c) 2007 Dave Airlie <airlied@linux.ie> + * Copyright (C) 2011 Texas Instruments + * + * This program is free software and is provided to you under the terms of the + * GNU General Public License version 2 as published by the Free Software + * Foundation, and any use by you of this program is subject to the terms of + * such GNU licence. + * + */ + +#include <linux/amba/clcd-regs.h> +#include <linux/clk.h> +#include <linux/version.h> +#include <linux/dma-buf.h> +#include <linux/of_graph.h> + +#include <drm/drmP.h> +#include <drm/drm_panel.h> +#include <drm/drm_gem_cma_helper.h> +#include <drm/drm_fb_cma_helper.h> + +#include "pl111_drm.h" + +irqreturn_t pl111_irq(int irq, void *data) +{ + struct pl111_drm_dev_private *priv = data; + u32 irq_stat; + irqreturn_t status = IRQ_NONE; + + irq_stat = readl(priv->regs + CLCD_PL111_MIS); + + if (!irq_stat) + return IRQ_NONE; + + if (irq_stat & CLCD_IRQ_NEXTBASE_UPDATE) { + drm_crtc_handle_vblank(&priv->pipe.crtc); + + status = IRQ_HANDLED; + } + + /* Clear the interrupt once done */ + writel(irq_stat, priv->regs + CLCD_PL111_ICR); + + return status; +} + +static u32 pl111_get_fb_offset(struct drm_plane_state *pstate) +{ + struct drm_framebuffer *fb = pstate->fb; + struct drm_gem_cma_object *obj = drm_fb_cma_get_gem_obj(fb, 0); + + return (obj->paddr + + fb->offsets[0] + + fb->format->cpp[0] * pstate->src_x + + fb->pitches[0] * pstate->src_y); +} + +static int pl111_display_check(struct drm_simple_display_pipe *pipe, + struct drm_plane_state *pstate, + struct drm_crtc_state *cstate) +{ + const struct drm_display_mode *mode = &cstate->mode; + struct drm_framebuffer *old_fb = pipe->plane.state->fb; + struct drm_framebuffer *fb = pstate->fb; + + if (mode->hdisplay % 16) + return -EINVAL; + + if (fb) { + u32 offset = pl111_get_fb_offset(pstate); + + /* FB base address must be dword aligned. */ + if (offset & 3) + return -EINVAL; + + /* There's no pitch register -- the mode's hdisplay + * controls it. + */ + if (fb->pitches[0] != mode->hdisplay * fb->format->cpp[0]) + return -EINVAL; + + /* We can't change the FB format in a flicker-free + * manner (and only update it during CRTC enable). + */ + if (old_fb && old_fb->format != fb->format) + cstate->mode_changed = true; + } + + return 0; +} + +static void pl111_display_enable(struct drm_simple_display_pipe *pipe, + struct drm_crtc_state *cstate) +{ + struct drm_crtc *crtc = &pipe->crtc; + struct drm_plane *plane = &pipe->plane; + struct drm_device *drm = crtc->dev; + struct pl111_drm_dev_private *priv = drm->dev_private; + const struct drm_display_mode *mode = &cstate->mode; + struct drm_framebuffer *fb = plane->state->fb; + struct drm_connector *connector = &priv->connector.connector; + u32 cntl; + u32 ppl, hsw, hfp, hbp; + u32 lpp, vsw, vfp, vbp; + u32 cpl, tim2; + int ret; + + ret = clk_set_rate(priv->clk, mode->clock * 1000); + if (ret) { + dev_err(drm->dev, + "Failed to set pixel clock rate to %d: %d\n", + mode->clock * 1000, ret); + } + + clk_prepare_enable(priv->clk); + + ppl = (mode->hdisplay / 16) - 1; + hsw = mode->hsync_end - mode->hsync_start - 1; + hfp = mode->hsync_start - mode->hdisplay - 1; + hbp = mode->htotal - mode->hsync_end - 1; + + lpp = mode->vdisplay - 1; + vsw = mode->vsync_end - mode->vsync_start - 1; + vfp = mode->vsync_start - mode->vdisplay; + vbp = mode->vtotal - mode->vsync_end; + + cpl = mode->hdisplay - 1; + + writel((ppl << 2) | + (hsw << 8) | + (hfp << 16) | + (hbp << 24), + priv->regs + CLCD_TIM0); + writel(lpp | + (vsw << 10) | + (vfp << 16) | + (vbp << 24), + priv->regs + CLCD_TIM1); + + spin_lock(&priv->tim2_lock); + + tim2 = readl(priv->regs + CLCD_TIM2); + tim2 &= (TIM2_BCD | TIM2_PCD_LO_MASK | TIM2_PCD_HI_MASK); + + if (mode->flags & DRM_MODE_FLAG_NHSYNC) + tim2 |= TIM2_IHS; + + if (mode->flags & DRM_MODE_FLAG_NVSYNC) + tim2 |= TIM2_IVS; + + if (connector->display_info.bus_flags & DRM_BUS_FLAG_DE_LOW) + tim2 |= TIM2_IOE; + + if (connector->display_info.bus_flags & DRM_BUS_FLAG_PIXDATA_NEGEDGE) + tim2 |= TIM2_IPC; + + tim2 |= cpl << 16; + writel(tim2, priv->regs + CLCD_TIM2); + spin_unlock(&priv->tim2_lock); + + writel(0, priv->regs + CLCD_TIM3); + + drm_panel_prepare(priv->connector.panel); + + /* Enable and Power Up */ + cntl = CNTL_LCDEN | CNTL_LCDTFT | CNTL_LCDPWR | CNTL_LCDVCOMP(1); + + /* Note that the the hardware's format reader takes 'r' from + * the low bit, while DRM formats list channels from high bit + * to low bit as you read left to right. + */ + switch (fb->format->format) { + case DRM_FORMAT_ABGR8888: + case DRM_FORMAT_XBGR8888: + cntl |= CNTL_LCDBPP24; + break; + case DRM_FORMAT_ARGB8888: + case DRM_FORMAT_XRGB8888: + cntl |= CNTL_LCDBPP24 | CNTL_BGR; + break; + case DRM_FORMAT_BGR565: + cntl |= CNTL_LCDBPP16_565; + break; + case DRM_FORMAT_RGB565: + cntl |= CNTL_LCDBPP16_565 | CNTL_BGR; + break; + case DRM_FORMAT_ABGR1555: + case DRM_FORMAT_XBGR1555: + cntl |= CNTL_LCDBPP16; + break; + case DRM_FORMAT_ARGB1555: + case DRM_FORMAT_XRGB1555: + cntl |= CNTL_LCDBPP16 | CNTL_BGR; + break; + case DRM_FORMAT_ABGR4444: + case DRM_FORMAT_XBGR4444: + cntl |= CNTL_LCDBPP16_444; + break; + case DRM_FORMAT_ARGB4444: + case DRM_FORMAT_XRGB4444: + cntl |= CNTL_LCDBPP16_444 | CNTL_BGR; + break; + default: + WARN_ONCE(true, "Unknown FB format 0x%08x\n", + fb->format->format); + break; + } + + writel(cntl, priv->regs + CLCD_PL111_CNTL); + + drm_panel_enable(priv->connector.panel); + + drm_crtc_vblank_on(crtc); +} + +void pl111_display_disable(struct drm_simple_display_pipe *pipe) +{ + struct drm_crtc *crtc = &pipe->crtc; + struct drm_device *drm = crtc->dev; + struct pl111_drm_dev_private *priv = drm->dev_private; + + drm_crtc_vblank_off(crtc); + + drm_panel_disable(priv->connector.panel); + + /* Disable and Power Down */ + writel(0, priv->regs + CLCD_PL111_CNTL); + + drm_panel_unprepare(priv->connector.panel); + + clk_disable_unprepare(priv->clk); +} + +static void pl111_display_update(struct drm_simple_display_pipe *pipe, + struct drm_plane_state *old_pstate) +{ + struct drm_crtc *crtc = &pipe->crtc; + struct drm_device *drm = crtc->dev; + struct pl111_drm_dev_private *priv = drm->dev_private; + struct drm_pending_vblank_event *event = crtc->state->event; + struct drm_plane *plane = &pipe->plane; + struct drm_plane_state *pstate = plane->state; + struct drm_framebuffer *fb = pstate->fb; + + if (fb) { + u32 addr = pl111_get_fb_offset(pstate); + + writel(addr, priv->regs + CLCD_UBAS); + } + + if (event) { + crtc->state->event = NULL; + + spin_lock_irq(&crtc->dev->event_lock); + if (crtc->state->active && drm_crtc_vblank_get(crtc) == 0) + drm_crtc_arm_vblank_event(crtc, event); + else + drm_crtc_send_vblank_event(crtc, event); + spin_unlock_irq(&crtc->dev->event_lock); + } +} + +int pl111_enable_vblank(struct drm_device *drm, unsigned int crtc) +{ + struct pl111_drm_dev_private *priv = drm->dev_private; + + writel(CLCD_IRQ_NEXTBASE_UPDATE, priv->regs + CLCD_PL111_IENB); + + return 0; +} + +void pl111_disable_vblank(struct drm_device *drm, unsigned int crtc) +{ + struct pl111_drm_dev_private *priv = drm->dev_private; + + writel(0, priv->regs + CLCD_PL111_IENB); +} + +static int pl111_display_prepare_fb(struct drm_simple_display_pipe *pipe, + struct drm_plane_state *plane_state) +{ + return drm_fb_cma_prepare_fb(&pipe->plane, plane_state); +} + +static const struct drm_simple_display_pipe_funcs pl111_display_funcs = { + .check = pl111_display_check, + .enable = pl111_display_enable, + .disable = pl111_display_disable, + .update = pl111_display_update, + .prepare_fb = pl111_display_prepare_fb, +}; + +static int pl111_clk_div_choose_div(struct clk_hw *hw, unsigned long rate, + unsigned long *prate, bool set_parent) +{ + int best_div = 1, div; + struct clk_hw *parent = clk_hw_get_parent(hw); + unsigned long best_prate = 0; + unsigned long best_diff = ~0ul; + int max_div = (1 << (TIM2_PCD_LO_BITS + TIM2_PCD_HI_BITS)) - 1; + + for (div = 1; div < max_div; div++) { + unsigned long this_prate, div_rate, diff; + + if (set_parent) + this_prate = clk_hw_round_rate(parent, rate * div); + else + this_prate = *prate; + div_rate = DIV_ROUND_UP_ULL(this_prate, div); + diff = abs(rate - div_rate); + + if (diff < best_diff) { + best_div = div; + best_diff = diff; + best_prate = this_prate; + } + } + + *prate = best_prate; + return best_div; +} + +static long pl111_clk_div_round_rate(struct clk_hw *hw, unsigned long rate, + unsigned long *prate) +{ + int div = pl111_clk_div_choose_div(hw, rate, prate, true); + + return DIV_ROUND_UP_ULL(*prate, div); +} + +static unsigned long pl111_clk_div_recalc_rate(struct clk_hw *hw, + unsigned long prate) +{ + struct pl111_drm_dev_private *priv = + container_of(hw, struct pl111_drm_dev_private, clk_div); + u32 tim2 = readl(priv->regs + CLCD_TIM2); + int div; + + if (tim2 & TIM2_BCD) + return prate; + + div = tim2 & TIM2_PCD_LO_MASK; + div |= (tim2 & TIM2_PCD_HI_MASK) >> + (TIM2_PCD_HI_SHIFT - TIM2_PCD_LO_BITS); + div += 2; + + return DIV_ROUND_UP_ULL(prate, div); +} + +static int pl111_clk_div_set_rate(struct clk_hw *hw, unsigned long rate, + unsigned long prate) +{ + struct pl111_drm_dev_private *priv = + container_of(hw, struct pl111_drm_dev_private, clk_div); + int div = pl111_clk_div_choose_div(hw, rate, &prate, false); + u32 tim2; + + spin_lock(&priv->tim2_lock); + tim2 = readl(priv->regs + CLCD_TIM2); + tim2 &= ~(TIM2_BCD | TIM2_PCD_LO_MASK | TIM2_PCD_HI_MASK); + + if (div == 1) { + tim2 |= TIM2_BCD; + } else { + div -= 2; + tim2 |= div & TIM2_PCD_LO_MASK; + tim2 |= (div >> TIM2_PCD_LO_BITS) << TIM2_PCD_HI_SHIFT; + } + + writel(tim2, priv->regs + CLCD_TIM2); + spin_unlock(&priv->tim2_lock); + + return 0; +} + +static const struct clk_ops pl111_clk_div_ops = { + .recalc_rate = pl111_clk_div_recalc_rate, + .round_rate = pl111_clk_div_round_rate, + .set_rate = pl111_clk_div_set_rate, +}; + +static int +pl111_init_clock_divider(struct drm_device *drm) +{ + struct pl111_drm_dev_private *priv = drm->dev_private; + struct clk *parent = devm_clk_get(drm->dev, "clcdclk"); + struct clk_hw *div = &priv->clk_div; + const char *parent_name; + struct clk_init_data init = { + .name = "pl111_div", + .ops = &pl111_clk_div_ops, + .parent_names = &parent_name, + .num_parents = 1, + .flags = CLK_SET_RATE_PARENT, + }; + int ret; + + if (IS_ERR(parent)) { + dev_err(drm->dev, "CLCD: unable to get clcdclk.\n"); + return PTR_ERR(parent); + } + parent_name = __clk_get_name(parent); + + spin_lock_init(&priv->tim2_lock); + div->init = &init; + + ret = devm_clk_hw_register(drm->dev, div); + + priv->clk = div->clk; + return ret; +} + +int pl111_display_init(struct drm_device *drm) +{ + struct pl111_drm_dev_private *priv = drm->dev_private; + struct device *dev = drm->dev; + struct device_node *endpoint; + u32 tft_r0b0g0[3]; + int ret; + static const u32 formats[] = { + DRM_FORMAT_ABGR8888, + DRM_FORMAT_XBGR8888, + DRM_FORMAT_ARGB8888, + DRM_FORMAT_XRGB8888, + DRM_FORMAT_BGR565, + DRM_FORMAT_RGB565, + DRM_FORMAT_ABGR1555, + DRM_FORMAT_XBGR1555, + DRM_FORMAT_ARGB1555, + DRM_FORMAT_XRGB1555, + DRM_FORMAT_ABGR4444, + DRM_FORMAT_XBGR4444, + DRM_FORMAT_ARGB4444, + DRM_FORMAT_XRGB4444, + }; + + endpoint = of_graph_get_next_endpoint(dev->of_node, NULL); + if (!endpoint) + return -ENODEV; + + if (of_property_read_u32_array(endpoint, + "arm,pl11x,tft-r0g0b0-pads", + tft_r0b0g0, + ARRAY_SIZE(tft_r0b0g0)) != 0) { + dev_err(dev, "arm,pl11x,tft-r0g0b0-pads should be 3 ints\n"); + of_node_put(endpoint); + return -ENOENT; + } + of_node_put(endpoint); + + if (tft_r0b0g0[0] != 0 || + tft_r0b0g0[1] != 8 || + tft_r0b0g0[2] != 16) { + dev_err(dev, "arm,pl11x,tft-r0g0b0-pads != [0,8,16] not yet supported\n"); + return -EINVAL; + } + + ret = pl111_init_clock_divider(drm); + if (ret) + return ret; + + ret = drm_simple_display_pipe_init(drm, &priv->pipe, + &pl111_display_funcs, + formats, ARRAY_SIZE(formats), + &priv->connector.connector); + if (ret) + return ret; + + return 0; +} diff --git a/drivers/gpu/drm/pl111/pl111_drm.h b/drivers/gpu/drm/pl111/pl111_drm.h new file mode 100644 index 000000000000..5c685bfc8fdc --- /dev/null +++ b/drivers/gpu/drm/pl111/pl111_drm.h @@ -0,0 +1,67 @@ +/* + * + * (C) COPYRIGHT 2012-2013 ARM Limited. All rights reserved. + * + * + * Parts of this file were based on sources as follows: + * + * Copyright (c) 2006-2008 Intel Corporation + * Copyright (c) 2007 Dave Airlie <airlied@linux.ie> + * Copyright (C) 2011 Texas Instruments + * + * This program is free software and is provided to you under the terms of the + * GNU General Public License version 2 as published by the Free Software + * Foundation, and any use by you of this program is subject to the terms of + * such GNU licence. + * + */ + +#ifndef _PL111_DRM_H_ +#define _PL111_DRM_H_ + +#include <drm/drm_gem.h> +#include <drm/drm_simple_kms_helper.h> +#include <linux/clk-provider.h> + +#define CLCD_IRQ_NEXTBASE_UPDATE BIT(2) + +struct drm_minor; + +struct pl111_drm_connector { + struct drm_connector connector; + struct drm_panel *panel; +}; + +struct pl111_drm_dev_private { + struct drm_device *drm; + + struct pl111_drm_connector connector; + struct drm_simple_display_pipe pipe; + struct drm_fbdev_cma *fbdev; + + void *regs; + /* The pixel clock (a reference to our clock divider off of CLCDCLK). */ + struct clk *clk; + /* pl111's internal clock divider. */ + struct clk_hw clk_div; + /* Lock to sync access to CLCD_TIM2 between the common clock + * subsystem and pl111_display_enable(). + */ + spinlock_t tim2_lock; +}; + +#define to_pl111_connector(x) \ + container_of(x, struct pl111_drm_connector, connector) + +int pl111_display_init(struct drm_device *dev); +int pl111_enable_vblank(struct drm_device *drm, unsigned int crtc); +void pl111_disable_vblank(struct drm_device *drm, unsigned int crtc); +irqreturn_t pl111_irq(int irq, void *data); +int pl111_connector_init(struct drm_device *dev); +int pl111_encoder_init(struct drm_device *dev); +int pl111_dumb_create(struct drm_file *file_priv, + struct drm_device *dev, + struct drm_mode_create_dumb *args); +int pl111_debugfs_init(struct drm_minor *minor); + +#endif /* _PL111_DRM_H_ */ diff --git a/drivers/gpu/drm/pl111/pl111_drv.c b/drivers/gpu/drm/pl111/pl111_drv.c new file mode 100644 index 000000000000..e96efad37d27 --- /dev/null +++ b/drivers/gpu/drm/pl111/pl111_drv.c @@ -0,0 +1,269 @@ +/* + * (C) COPYRIGHT 2012-2013 ARM Limited. All rights reserved. + * + * Parts of this file were based on sources as follows: + * + * Copyright (c) 2006-2008 Intel Corporation + * Copyright (c) 2007 Dave Airlie <airlied@linux.ie> + * Copyright (C) 2011 Texas Instruments + * + * This program is free software and is provided to you under the terms of the + * GNU General Public License version 2 as published by the Free Software + * Foundation, and any use by you of this program is subject to the terms of + * such GNU licence. + * + */ + +/** + * DOC: ARM PrimeCell PL111 CLCD Driver + * + * The PL111 is a simple LCD controller that can support TFT and STN + * displays. This driver exposes a standard KMS interface for them. + * + * This driver uses the same Device Tree binding as the fbdev CLCD + * driver. While the fbdev driver supports panels that may be + * connected to the CLCD internally to the CLCD driver, in DRM the + * panels get split out to drivers/gpu/drm/panels/. This means that, + * in converting from using fbdev to using DRM, you also need to write + * a panel driver (which may be as simple as an entry in + * panel-simple.c). + * + * The driver currently doesn't expose the cursor. The DRM API for + * cursors requires support for 64x64 ARGB8888 cursor images, while + * the hardware can only support 64x64 monochrome with masking + * cursors. While one could imagine trying to hack something together + * to look at the ARGB8888 and program reasonable in monochrome, we + * just don't expose the cursor at all instead, and leave cursor + * support to the X11 software cursor layer. + * + * TODO: + * + * - Fix race between setting plane base address and getting IRQ for + * vsync firing the pageflip completion. + * + * - Expose the correct set of formats we can support based on the + * "arm,pl11x,tft-r0g0b0-pads" DT property. + * + * - Use the "max-memory-bandwidth" DT property to filter the + * supported formats. + * + * - Read back hardware state at boot to skip reprogramming the + * hardware when doing a no-op modeset. + * + * - Use the CLKSEL bit to support switching between the two external + * clock parents. + */ + +#include <linux/amba/bus.h> +#include <linux/amba/clcd-regs.h> +#include <linux/version.h> +#include <linux/shmem_fs.h> +#include <linux/dma-buf.h> +#include <linux/module.h> +#include <linux/slab.h> + +#include <drm/drmP.h> +#include <drm/drm_atomic_helper.h> +#include <drm/drm_crtc_helper.h> +#include <drm/drm_gem_cma_helper.h> +#include <drm/drm_fb_cma_helper.h> + +#include "pl111_drm.h" + +#define DRIVER_DESC "DRM module for PL111" + +static struct drm_mode_config_funcs mode_config_funcs = { + .fb_create = drm_fb_cma_create, + .atomic_check = drm_atomic_helper_check, + .atomic_commit = drm_atomic_helper_commit, +}; + +static int pl111_modeset_init(struct drm_device *dev) +{ + struct drm_mode_config *mode_config; + struct pl111_drm_dev_private *priv = dev->dev_private; + int ret = 0; + + drm_mode_config_init(dev); + mode_config = &dev->mode_config; + mode_config->funcs = &mode_config_funcs; + mode_config->min_width = 1; + mode_config->max_width = 1024; + mode_config->min_height = 1; + mode_config->max_height = 768; + + ret = pl111_connector_init(dev); + if (ret) { + dev_err(dev->dev, "Failed to create pl111_drm_connector\n"); + goto out_config; + } + + /* Don't actually attach if we didn't find a drm_panel + * attached to us. This will allow a kernel to include both + * the fbdev pl111 driver and this one, and choose between + * them based on which subsystem has support for the panel. + */ + if (!priv->connector.panel) { + dev_info(dev->dev, + "Disabling due to lack of DRM panel device.\n"); + ret = -ENODEV; + goto out_config; + } + + ret = pl111_display_init(dev); + if (ret != 0) { + dev_err(dev->dev, "Failed to init display\n"); + goto out_config; + } + + ret = drm_vblank_init(dev, 1); + if (ret != 0) { + dev_err(dev->dev, "Failed to init vblank\n"); + goto out_config; + } + + drm_mode_config_reset(dev); + + priv->fbdev = drm_fbdev_cma_init(dev, 32, + dev->mode_config.num_connector); + + drm_kms_helper_poll_init(dev); + + goto finish; + +out_config: + drm_mode_config_cleanup(dev); +finish: + return ret; +} + +DEFINE_DRM_GEM_CMA_FOPS(drm_fops); + +static void pl111_lastclose(struct drm_device *dev) +{ + struct pl111_drm_dev_private *priv = dev->dev_private; + + drm_fbdev_cma_restore_mode(priv->fbdev); +} + +static struct drm_driver pl111_drm_driver = { + .driver_features = + DRIVER_MODESET | DRIVER_GEM | DRIVER_PRIME | DRIVER_ATOMIC, + .lastclose = pl111_lastclose, + .ioctls = NULL, + .fops = &drm_fops, + .name = "pl111", + .desc = DRIVER_DESC, + .date = "20170317", + .major = 1, + .minor = 0, + .patchlevel = 0, + .dumb_create = drm_gem_cma_dumb_create, + .dumb_destroy = drm_gem_dumb_destroy, + .dumb_map_offset = drm_gem_cma_dumb_map_offset, + .gem_free_object = drm_gem_cma_free_object, + .gem_vm_ops = &drm_gem_cma_vm_ops, + + .enable_vblank = pl111_enable_vblank, + .disable_vblank = pl111_disable_vblank, + + .prime_handle_to_fd = drm_gem_prime_handle_to_fd, + .prime_fd_to_handle = drm_gem_prime_fd_to_handle, + .gem_prime_import = drm_gem_prime_import, + .gem_prime_import_sg_table = drm_gem_cma_prime_import_sg_table, + .gem_prime_export = drm_gem_prime_export, + .gem_prime_get_sg_table = drm_gem_cma_prime_get_sg_table, + +#if defined(CONFIG_DEBUG_FS) + .debugfs_init = pl111_debugfs_init, +#endif +}; + +#ifdef CONFIG_ARM_AMBA +static int pl111_amba_probe(struct amba_device *amba_dev, + const struct amba_id *id) +{ + struct device *dev = &amba_dev->dev; + struct pl111_drm_dev_private *priv; + struct drm_device *drm; + int ret; + + priv = devm_kzalloc(dev, sizeof(*priv), GFP_KERNEL); + if (!priv) + return -ENOMEM; + + drm = drm_dev_alloc(&pl111_drm_driver, dev); + if (IS_ERR(drm)) + return PTR_ERR(drm); + amba_set_drvdata(amba_dev, drm); + priv->drm = drm; + drm->dev_private = priv; + + priv->regs = devm_ioremap_resource(dev, &amba_dev->res); + if (IS_ERR(priv->regs)) { + dev_err(dev, "%s failed mmio\n", __func__); + return PTR_ERR(priv->regs); + } + + /* turn off interrupts before requesting the irq */ + writel(0, priv->regs + CLCD_PL111_IENB); + + ret = devm_request_irq(dev, amba_dev->irq[0], pl111_irq, 0, + "pl111", priv); + if (ret != 0) { + dev_err(dev, "%s failed irq %d\n", __func__, ret); + return ret; + } + + ret = pl111_modeset_init(drm); + if (ret != 0) + goto dev_unref; + + ret = drm_dev_register(drm, 0); + if (ret < 0) + goto dev_unref; + + return 0; + +dev_unref: + drm_dev_unref(drm); + return ret; +} + +static int pl111_amba_remove(struct amba_device *amba_dev) +{ + struct drm_device *drm = amba_get_drvdata(amba_dev); + struct pl111_drm_dev_private *priv = drm->dev_private; + + drm_dev_unregister(drm); + if (priv->fbdev) + drm_fbdev_cma_fini(priv->fbdev); + drm_mode_config_cleanup(drm); + drm_dev_unref(drm); + + return 0; +} + +static struct amba_id pl111_id_table[] = { + { + .id = 0x00041111, + .mask = 0x000fffff, + }, + {0, 0}, +}; + +static struct amba_driver pl111_amba_driver = { + .drv = { + .name = "drm-clcd-pl111", + }, + .probe = pl111_amba_probe, + .remove = pl111_amba_remove, + .id_table = pl111_id_table, +}; + +module_amba_driver(pl111_amba_driver); +#endif /* CONFIG_ARM_AMBA */ + +MODULE_DESCRIPTION(DRIVER_DESC); +MODULE_AUTHOR("ARM Ltd."); +MODULE_LICENSE("GPL"); diff --git a/drivers/gpu/drm/qxl/Makefile b/drivers/gpu/drm/qxl/Makefile index bacc4aff1201..33a7d0c434b7 100644 --- a/drivers/gpu/drm/qxl/Makefile +++ b/drivers/gpu/drm/qxl/Makefile @@ -2,8 +2,6 @@ # Makefile for the drm device driver. This driver provides support for the # Direct Rendering Infrastructure (DRI) in XFree86 4.1.0 and higher. -ccflags-y := -Iinclude/drm - qxl-y := qxl_drv.o qxl_kms.o qxl_display.o qxl_ttm.o qxl_fb.o qxl_object.o qxl_gem.o qxl_cmd.o qxl_image.o qxl_draw.o qxl_debugfs.o qxl_irq.o qxl_dumb.o qxl_ioctl.o qxl_release.o qxl_prime.o obj-$(CONFIG_DRM_QXL)+= qxl.o diff --git a/drivers/gpu/drm/qxl/qxl_debugfs.c b/drivers/gpu/drm/qxl/qxl_debugfs.c index ffe821b61f7d..15c84068d3fb 100644 --- a/drivers/gpu/drm/qxl/qxl_debugfs.c +++ b/drivers/gpu/drm/qxl/qxl_debugfs.c @@ -30,7 +30,7 @@ #include <linux/debugfs.h> -#include "drmP.h" +#include <drm/drmP.h> #include "qxl_drv.h" #include "qxl_object.h" diff --git a/drivers/gpu/drm/qxl/qxl_display.c b/drivers/gpu/drm/qxl/qxl_display.c index 058340a002c2..f5ef81595f5a 100644 --- a/drivers/gpu/drm/qxl/qxl_display.c +++ b/drivers/gpu/drm/qxl/qxl_display.c @@ -23,16 +23,15 @@ * Alon Levy */ - #include <linux/crc32.h> - -#include "qxl_drv.h" -#include "qxl_object.h" -#include "drm_crtc_helper.h" +#include <drm/drm_crtc_helper.h> #include <drm/drm_plane_helper.h> #include <drm/drm_atomic_helper.h> #include <drm/drm_atomic.h> +#include "qxl_drv.h" +#include "qxl_object.h" + static bool qxl_head_enabled(struct qxl_head *head) { return head->width && head->height; diff --git a/drivers/gpu/drm/qxl/qxl_drv.c b/drivers/gpu/drm/qxl/qxl_drv.c index abf7b8360361..c2fc201d9e1b 100644 --- a/drivers/gpu/drm/qxl/qxl_drv.c +++ b/drivers/gpu/drm/qxl/qxl_drv.c @@ -31,9 +31,9 @@ #include <linux/module.h> #include <linux/console.h> -#include "drmP.h" -#include "drm/drm.h" -#include "drm_crtc_helper.h" +#include <drm/drmP.h> +#include <drm/drm.h> +#include <drm/drm_crtc_helper.h> #include "qxl_drv.h" #include "qxl_object.h" diff --git a/drivers/gpu/drm/qxl/qxl_drv.h b/drivers/gpu/drm/qxl/qxl_drv.h index 5ea290a33a68..3591d2330a09 100644 --- a/drivers/gpu/drm/qxl/qxl_drv.h +++ b/drivers/gpu/drm/qxl/qxl_drv.h @@ -36,20 +36,18 @@ #include <linux/firmware.h> #include <linux/platform_device.h> -#include "drmP.h" -#include "drm_crtc.h" -#include <ttm/ttm_bo_api.h> -#include <ttm/ttm_bo_driver.h> -#include <ttm/ttm_placement.h> -#include <ttm/ttm_module.h> - +#include <drm/drm_crtc.h> #include <drm/drm_encoder.h> #include <drm/drm_gem.h> - +#include <drm/drmP.h> +#include <drm/ttm/ttm_bo_api.h> +#include <drm/ttm/ttm_bo_driver.h> /* just for ttm_validate_buffer */ -#include <ttm/ttm_execbuf_util.h> - +#include <drm/ttm/ttm_execbuf_util.h> +#include <drm/ttm/ttm_module.h> +#include <drm/ttm/ttm_placement.h> #include <drm/qxl_drm.h> + #include "qxl_dev.h" #define DRIVER_AUTHOR "Dave Airlie" diff --git a/drivers/gpu/drm/qxl/qxl_fb.c b/drivers/gpu/drm/qxl/qxl_fb.c index 14e2a49a4dcf..573e7e9a5f98 100644 --- a/drivers/gpu/drm/qxl/qxl_fb.c +++ b/drivers/gpu/drm/qxl/qxl_fb.c @@ -25,14 +25,15 @@ */ #include <linux/module.h> -#include "drmP.h" -#include "drm/drm.h" -#include "drm/drm_crtc.h" -#include "drm/drm_crtc_helper.h" +#include <drm/drmP.h> +#include <drm/drm.h> +#include <drm/drm_crtc.h> +#include <drm/drm_crtc_helper.h> +#include <drm/drm_fb_helper.h> + #include "qxl_drv.h" #include "qxl_object.h" -#include "drm_fb_helper.h" #define QXL_DIRTY_DELAY (HZ / 30) diff --git a/drivers/gpu/drm/qxl/qxl_gem.c b/drivers/gpu/drm/qxl/qxl_gem.c index 3f185c4da5b7..85f546719adb 100644 --- a/drivers/gpu/drm/qxl/qxl_gem.c +++ b/drivers/gpu/drm/qxl/qxl_gem.c @@ -23,8 +23,9 @@ * Alon Levy */ -#include "drmP.h" -#include "drm/drm.h" +#include <drm/drmP.h> +#include <drm/drm.h> + #include "qxl_drv.h" #include "qxl_object.h" diff --git a/drivers/gpu/drm/qxl/qxl_ttm.c b/drivers/gpu/drm/qxl/qxl_ttm.c index 0fdedee4509d..87fc1dbd0a2f 100644 --- a/drivers/gpu/drm/qxl/qxl_ttm.c +++ b/drivers/gpu/drm/qxl/qxl_ttm.c @@ -23,11 +23,11 @@ * Alon Levy */ -#include <ttm/ttm_bo_api.h> -#include <ttm/ttm_bo_driver.h> -#include <ttm/ttm_placement.h> -#include <ttm/ttm_page_alloc.h> -#include <ttm/ttm_module.h> +#include <drm/ttm/ttm_bo_api.h> +#include <drm/ttm/ttm_bo_driver.h> +#include <drm/ttm/ttm_placement.h> +#include <drm/ttm/ttm_page_alloc.h> +#include <drm/ttm/ttm_module.h> #include <drm/drmP.h> #include <drm/drm.h> #include <drm/qxl_drm.h> diff --git a/drivers/gpu/drm/r128/Makefile b/drivers/gpu/drm/r128/Makefile index 1cc72ae3a880..1a6700ebaf09 100644 --- a/drivers/gpu/drm/r128/Makefile +++ b/drivers/gpu/drm/r128/Makefile @@ -2,7 +2,6 @@ # Makefile for the drm device driver. This driver provides support for the # Direct Rendering Infrastructure (DRI) in XFree86 4.1.0 and higher. -ccflags-y := -Iinclude/drm r128-y := r128_drv.o r128_cce.o r128_state.o r128_irq.o r128-$(CONFIG_COMPAT) += r128_ioc32.o diff --git a/drivers/gpu/drm/radeon/Makefile b/drivers/gpu/drm/radeon/Makefile index 08bd17d3925c..a5d3cd3ecb5f 100644 --- a/drivers/gpu/drm/radeon/Makefile +++ b/drivers/gpu/drm/radeon/Makefile @@ -2,7 +2,7 @@ # Makefile for the drm device driver. This driver provides support for the # Direct Rendering Infrastructure (DRI) in XFree86 4.1.0 and higher. -ccflags-y := -Iinclude/drm -Idrivers/gpu/drm/amd/include +ccflags-y := -Idrivers/gpu/drm/amd/include hostprogs-y := mkregtable clean-files := rn50_reg_safe.h r100_reg_safe.h r200_reg_safe.h rv515_reg_safe.h r300_reg_safe.h r420_reg_safe.h rs600_reg_safe.h r600_reg_safe.h evergreen_reg_safe.h cayman_reg_safe.h diff --git a/drivers/gpu/drm/radeon/btc_dpm.c b/drivers/gpu/drm/radeon/btc_dpm.c index 38e5123708e7..95652e643da1 100644 --- a/drivers/gpu/drm/radeon/btc_dpm.c +++ b/drivers/gpu/drm/radeon/btc_dpm.c @@ -22,7 +22,7 @@ * Authors: Alex Deucher */ -#include "drmP.h" +#include <drm/drmP.h> #include "radeon.h" #include "radeon_asic.h" #include "btcd.h" diff --git a/drivers/gpu/drm/radeon/ci_dpm.c b/drivers/gpu/drm/radeon/ci_dpm.c index 7ba450832e6b..f0cf99783c62 100644 --- a/drivers/gpu/drm/radeon/ci_dpm.c +++ b/drivers/gpu/drm/radeon/ci_dpm.c @@ -22,7 +22,7 @@ */ #include <linux/firmware.h> -#include "drmP.h" +#include <drm/drmP.h> #include "radeon.h" #include "radeon_asic.h" #include "radeon_ucode.h" diff --git a/drivers/gpu/drm/radeon/ci_smc.c b/drivers/gpu/drm/radeon/ci_smc.c index 24760ee3063e..3356a21d97ec 100644 --- a/drivers/gpu/drm/radeon/ci_smc.c +++ b/drivers/gpu/drm/radeon/ci_smc.c @@ -23,7 +23,7 @@ */ #include <linux/firmware.h> -#include "drmP.h" +#include <drm/drmP.h> #include "radeon.h" #include "cikd.h" #include "ppsmc.h" diff --git a/drivers/gpu/drm/radeon/cik.c b/drivers/gpu/drm/radeon/cik.c index 53710dd7d5dd..e368ce22bcc4 100644 --- a/drivers/gpu/drm/radeon/cik.c +++ b/drivers/gpu/drm/radeon/cik.c @@ -24,7 +24,7 @@ #include <linux/firmware.h> #include <linux/slab.h> #include <linux/module.h> -#include "drmP.h" +#include <drm/drmP.h> #include "radeon.h" #include "radeon_asic.h" #include "radeon_audio.h" @@ -9150,23 +9150,10 @@ static u32 dce8_latency_watermark(struct dce8_wm_params *wm) a.full = dfixed_const(available_bandwidth); b.full = dfixed_const(wm->num_heads); a.full = dfixed_div(a, b); + tmp = div_u64((u64) dmif_size * (u64) wm->disp_clk, mc_latency + 512); + tmp = min(dfixed_trunc(a), tmp); - b.full = dfixed_const(mc_latency + 512); - c.full = dfixed_const(wm->disp_clk); - b.full = dfixed_div(b, c); - - c.full = dfixed_const(dmif_size); - b.full = dfixed_div(c, b); - - tmp = min(dfixed_trunc(a), dfixed_trunc(b)); - - b.full = dfixed_const(1000); - c.full = dfixed_const(wm->disp_clk); - b.full = dfixed_div(c, b); - c.full = dfixed_const(wm->bytes_per_pixel); - b.full = dfixed_mul(b, c); - - lb_fill_bw = min(tmp, dfixed_trunc(b)); + lb_fill_bw = min(tmp, wm->disp_clk * wm->bytes_per_pixel / 1000); a.full = dfixed_const(max_src_lines_per_dst_line * wm->src_width * wm->bytes_per_pixel); b.full = dfixed_const(1000); @@ -9274,14 +9261,14 @@ static void dce8_program_watermarks(struct radeon_device *rdev, { struct drm_display_mode *mode = &radeon_crtc->base.mode; struct dce8_wm_params wm_low, wm_high; - u32 pixel_period; + u32 active_time; u32 line_time = 0; u32 latency_watermark_a = 0, latency_watermark_b = 0; u32 tmp, wm_mask; if (radeon_crtc->base.enabled && num_heads && mode) { - pixel_period = 1000000 / (u32)mode->clock; - line_time = min((u32)mode->crtc_htotal * pixel_period, (u32)65535); + active_time = 1000000UL * (u32)mode->crtc_hdisplay / (u32)mode->clock; + line_time = min((u32) (1000000UL * (u32)mode->crtc_htotal / (u32)mode->clock), (u32)65535); /* watermark for high clocks */ if ((rdev->pm.pm_method == PM_METHOD_DPM) && @@ -9297,7 +9284,7 @@ static void dce8_program_watermarks(struct radeon_device *rdev, wm_high.disp_clk = mode->clock; wm_high.src_width = mode->crtc_hdisplay; - wm_high.active_time = mode->crtc_hdisplay * pixel_period; + wm_high.active_time = active_time; wm_high.blank_time = line_time - wm_high.active_time; wm_high.interlaced = false; if (mode->flags & DRM_MODE_FLAG_INTERLACE) @@ -9337,7 +9324,7 @@ static void dce8_program_watermarks(struct radeon_device *rdev, wm_low.disp_clk = mode->clock; wm_low.src_width = mode->crtc_hdisplay; - wm_low.active_time = mode->crtc_hdisplay * pixel_period; + wm_low.active_time = active_time; wm_low.blank_time = line_time - wm_low.active_time; wm_low.interlaced = false; if (mode->flags & DRM_MODE_FLAG_INTERLACE) diff --git a/drivers/gpu/drm/radeon/cypress_dpm.c b/drivers/gpu/drm/radeon/cypress_dpm.c index a4edd0702718..3eb7899a4035 100644 --- a/drivers/gpu/drm/radeon/cypress_dpm.c +++ b/drivers/gpu/drm/radeon/cypress_dpm.c @@ -22,7 +22,7 @@ * Authors: Alex Deucher */ -#include "drmP.h" +#include <drm/drmP.h> #include "radeon.h" #include "radeon_asic.h" #include "evergreend.h" diff --git a/drivers/gpu/drm/radeon/evergreen.c b/drivers/gpu/drm/radeon/evergreen.c index d1b1e0cc3c25..f130ec41ee4b 100644 --- a/drivers/gpu/drm/radeon/evergreen.c +++ b/drivers/gpu/drm/radeon/evergreen.c @@ -2188,13 +2188,7 @@ static u32 evergreen_latency_watermark(struct evergreen_wm_params *wm) b.full = dfixed_const(wm->num_heads); a.full = dfixed_div(a, b); - b.full = dfixed_const(1000); - c.full = dfixed_const(wm->disp_clk); - b.full = dfixed_div(c, b); - c.full = dfixed_const(wm->bytes_per_pixel); - b.full = dfixed_mul(b, c); - - lb_fill_bw = min(dfixed_trunc(a), dfixed_trunc(b)); + lb_fill_bw = min(dfixed_trunc(a), wm->disp_clk * wm->bytes_per_pixel / 1000); a.full = dfixed_const(max_src_lines_per_dst_line * wm->src_width * wm->bytes_per_pixel); b.full = dfixed_const(1000); @@ -2261,7 +2255,7 @@ static void evergreen_program_watermarks(struct radeon_device *rdev, struct drm_display_mode *mode = &radeon_crtc->base.mode; struct evergreen_wm_params wm_low, wm_high; u32 dram_channels; - u32 pixel_period; + u32 active_time; u32 line_time = 0; u32 latency_watermark_a = 0, latency_watermark_b = 0; u32 priority_a_mark = 0, priority_b_mark = 0; @@ -2272,8 +2266,8 @@ static void evergreen_program_watermarks(struct radeon_device *rdev, fixed20_12 a, b, c; if (radeon_crtc->base.enabled && num_heads && mode) { - pixel_period = 1000000 / (u32)mode->clock; - line_time = min((u32)mode->crtc_htotal * pixel_period, (u32)65535); + active_time = 1000000UL * (u32)mode->crtc_hdisplay / (u32)mode->clock; + line_time = min((u32) (1000000UL * (u32)mode->crtc_htotal / (u32)mode->clock), (u32)65535); priority_a_cnt = 0; priority_b_cnt = 0; dram_channels = evergreen_get_number_of_dram_channels(rdev); @@ -2291,7 +2285,7 @@ static void evergreen_program_watermarks(struct radeon_device *rdev, wm_high.disp_clk = mode->clock; wm_high.src_width = mode->crtc_hdisplay; - wm_high.active_time = mode->crtc_hdisplay * pixel_period; + wm_high.active_time = active_time; wm_high.blank_time = line_time - wm_high.active_time; wm_high.interlaced = false; if (mode->flags & DRM_MODE_FLAG_INTERLACE) @@ -2318,7 +2312,7 @@ static void evergreen_program_watermarks(struct radeon_device *rdev, wm_low.disp_clk = mode->clock; wm_low.src_width = mode->crtc_hdisplay; - wm_low.active_time = mode->crtc_hdisplay * pixel_period; + wm_low.active_time = active_time; wm_low.blank_time = line_time - wm_low.active_time; wm_low.interlaced = false; if (mode->flags & DRM_MODE_FLAG_INTERLACE) diff --git a/drivers/gpu/drm/radeon/kv_dpm.c b/drivers/gpu/drm/radeon/kv_dpm.c index a7e978677937..ae1529b0ef6f 100644 --- a/drivers/gpu/drm/radeon/kv_dpm.c +++ b/drivers/gpu/drm/radeon/kv_dpm.c @@ -21,7 +21,7 @@ * */ -#include "drmP.h" +#include <drm/drmP.h> #include "radeon.h" #include "cikd.h" #include "r600_dpm.h" diff --git a/drivers/gpu/drm/radeon/kv_smc.c b/drivers/gpu/drm/radeon/kv_smc.c index 0000b59a6d05..af60bd32a287 100644 --- a/drivers/gpu/drm/radeon/kv_smc.c +++ b/drivers/gpu/drm/radeon/kv_smc.c @@ -22,7 +22,7 @@ * Authors: Alex Deucher */ -#include "drmP.h" +#include <drm/drmP.h> #include "radeon.h" #include "cikd.h" #include "kv_dpm.h" diff --git a/drivers/gpu/drm/radeon/ni_dpm.c b/drivers/gpu/drm/radeon/ni_dpm.c index 4a601f990562..9416e72f86aa 100644 --- a/drivers/gpu/drm/radeon/ni_dpm.c +++ b/drivers/gpu/drm/radeon/ni_dpm.c @@ -21,7 +21,7 @@ * */ -#include "drmP.h" +#include <drm/drmP.h> #include "radeon.h" #include "radeon_asic.h" #include "nid.h" diff --git a/drivers/gpu/drm/radeon/r420.c b/drivers/gpu/drm/radeon/r420.c index 3eb0c4f9f796..45e1d4e60759 100644 --- a/drivers/gpu/drm/radeon/r420.c +++ b/drivers/gpu/drm/radeon/r420.c @@ -203,6 +203,7 @@ static void r420_clock_resume(struct radeon_device *rdev) static void r420_cp_errata_init(struct radeon_device *rdev) { + int r; struct radeon_ring *ring = &rdev->ring[RADEON_RING_TYPE_GFX_INDEX]; /* RV410 and R420 can lock up if CP DMA to host memory happens @@ -212,7 +213,8 @@ static void r420_cp_errata_init(struct radeon_device *rdev) * of the CP init, apparently. */ radeon_scratch_get(rdev, &rdev->config.r300.resync_scratch); - radeon_ring_lock(rdev, ring, 8); + r = radeon_ring_lock(rdev, ring, 8); + WARN_ON(r); radeon_ring_write(ring, PACKET0(R300_CP_RESYNC_ADDR, 1)); radeon_ring_write(ring, rdev->config.r300.resync_scratch); radeon_ring_write(ring, 0xDEADBEEF); @@ -221,12 +223,14 @@ static void r420_cp_errata_init(struct radeon_device *rdev) static void r420_cp_errata_fini(struct radeon_device *rdev) { + int r; struct radeon_ring *ring = &rdev->ring[RADEON_RING_TYPE_GFX_INDEX]; /* Catch the RESYNC we dispatched all the way back, * at the very beginning of the CP init. */ - radeon_ring_lock(rdev, ring, 8); + r = radeon_ring_lock(rdev, ring, 8); + WARN_ON(r); radeon_ring_write(ring, PACKET0(R300_RB3D_DSTCACHE_CTLSTAT, 0)); radeon_ring_write(ring, R300_RB3D_DC_FINISH); radeon_ring_unlock_commit(rdev, ring, false); diff --git a/drivers/gpu/drm/radeon/r600_dpm.c b/drivers/gpu/drm/radeon/r600_dpm.c index c7fc1dbfd192..31d1b4710844 100644 --- a/drivers/gpu/drm/radeon/r600_dpm.c +++ b/drivers/gpu/drm/radeon/r600_dpm.c @@ -22,7 +22,7 @@ * Authors: Alex Deucher */ -#include "drmP.h" +#include <drm/drmP.h> #include "radeon.h" #include "radeon_asic.h" #include "r600d.h" diff --git a/drivers/gpu/drm/radeon/radeon.h b/drivers/gpu/drm/radeon/radeon.h index c1c8e2208a21..342e3b1fb9c7 100644 --- a/drivers/gpu/drm/radeon/radeon.h +++ b/drivers/gpu/drm/radeon/radeon.h @@ -68,11 +68,11 @@ #include <linux/hashtable.h> #include <linux/dma-fence.h> -#include <ttm/ttm_bo_api.h> -#include <ttm/ttm_bo_driver.h> -#include <ttm/ttm_placement.h> -#include <ttm/ttm_module.h> -#include <ttm/ttm_execbuf_util.h> +#include <drm/ttm/ttm_bo_api.h> +#include <drm/ttm/ttm_bo_driver.h> +#include <drm/ttm/ttm_placement.h> +#include <drm/ttm/ttm_module.h> +#include <drm/ttm/ttm_execbuf_util.h> #include <drm/drm_gem.h> diff --git a/drivers/gpu/drm/radeon/radeon_cs.c b/drivers/gpu/drm/radeon/radeon_cs.c index df6b58c08544..00b22af70f5c 100644 --- a/drivers/gpu/drm/radeon/radeon_cs.c +++ b/drivers/gpu/drm/radeon/radeon_cs.c @@ -87,7 +87,8 @@ static int radeon_cs_parser_relocs(struct radeon_cs_parser *p) p->dma_reloc_idx = 0; /* FIXME: we assume that each relocs use 4 dwords */ p->nrelocs = chunk->length_dw / 4; - p->relocs = drm_calloc_large(p->nrelocs, sizeof(struct radeon_bo_list)); + p->relocs = kvmalloc_array(p->nrelocs, sizeof(struct radeon_bo_list), + GFP_KERNEL | __GFP_ZERO); if (p->relocs == NULL) { return -ENOMEM; } @@ -117,11 +118,13 @@ static int radeon_cs_parser_relocs(struct radeon_cs_parser *p) priority = (r->flags & RADEON_RELOC_PRIO_MASK) * 2 + !!r->write_domain; - /* the first reloc of an UVD job is the msg and that must be in - VRAM, also but everything into VRAM on AGP cards and older - IGP chips to avoid image corruptions */ + /* The first reloc of an UVD job is the msg and that must be in + * VRAM, the second reloc is the DPB and for WMV that must be in + * VRAM as well. Also put everything into VRAM on AGP cards and older + * IGP chips to avoid image corruptions + */ if (p->ring == R600_RING_TYPE_UVD_INDEX && - (i == 0 || pci_find_capability(p->rdev->ddev->pdev, + (i <= 0 || pci_find_capability(p->rdev->ddev->pdev, PCI_CAP_ID_AGP) || p->rdev->family == CHIP_RS780 || p->rdev->family == CHIP_RS880)) { @@ -339,7 +342,7 @@ int radeon_cs_parser_init(struct radeon_cs_parser *p, void *data) continue; } - p->chunks[i].kdata = drm_malloc_ab(size, sizeof(uint32_t)); + p->chunks[i].kdata = kvmalloc_array(size, sizeof(uint32_t), GFP_KERNEL); size *= sizeof(uint32_t); if (p->chunks[i].kdata == NULL) { return -ENOMEM; @@ -438,10 +441,10 @@ static void radeon_cs_parser_fini(struct radeon_cs_parser *parser, int error, bo } } kfree(parser->track); - drm_free_large(parser->relocs); - drm_free_large(parser->vm_bos); + kvfree(parser->relocs); + kvfree(parser->vm_bos); for (i = 0; i < parser->nchunks; i++) - drm_free_large(parser->chunks[i].kdata); + kvfree(parser->chunks[i].kdata); kfree(parser->chunks); kfree(parser->chunks_array); radeon_ib_free(parser->rdev, &parser->ib); diff --git a/drivers/gpu/drm/radeon/radeon_drv.c b/drivers/gpu/drm/radeon/radeon_drv.c index 93d45aa5c3d4..6f906abd612b 100644 --- a/drivers/gpu/drm/radeon/radeon_drv.c +++ b/drivers/gpu/drm/radeon/radeon_drv.c @@ -41,7 +41,7 @@ #include <drm/drm_gem.h> #include <drm/drm_fb_helper.h> -#include "drm_crtc_helper.h" +#include <drm/drm_crtc_helper.h> #include "radeon_kfd.h" /* @@ -115,10 +115,6 @@ int radeon_resume_kms(struct drm_device *dev, bool resume, bool fbcon); u32 radeon_get_vblank_counter_kms(struct drm_device *dev, unsigned int pipe); int radeon_enable_vblank_kms(struct drm_device *dev, unsigned int pipe); void radeon_disable_vblank_kms(struct drm_device *dev, unsigned int pipe); -int radeon_get_vblank_timestamp_kms(struct drm_device *dev, unsigned int pipe, - int *max_error, - struct timeval *vblank_time, - unsigned flags); void radeon_driver_irq_preinstall_kms(struct drm_device *dev); int radeon_driver_irq_postinstall_kms(struct drm_device *dev); void radeon_driver_irq_uninstall_kms(struct drm_device *dev); @@ -530,6 +526,16 @@ static const struct file_operations radeon_driver_kms_fops = { #endif }; +static bool +radeon_get_crtc_scanout_position(struct drm_device *dev, unsigned int pipe, + bool in_vblank_irq, int *vpos, int *hpos, + ktime_t *stime, ktime_t *etime, + const struct drm_display_mode *mode) +{ + return radeon_get_crtc_scanoutpos(dev, pipe, 0, vpos, hpos, + stime, etime, mode); +} + static struct drm_driver kms_driver = { .driver_features = DRIVER_USE_AGP | @@ -544,8 +550,8 @@ static struct drm_driver kms_driver = { .get_vblank_counter = radeon_get_vblank_counter_kms, .enable_vblank = radeon_enable_vblank_kms, .disable_vblank = radeon_disable_vblank_kms, - .get_vblank_timestamp = radeon_get_vblank_timestamp_kms, - .get_scanout_position = radeon_get_crtc_scanoutpos, + .get_vblank_timestamp = drm_calc_vbltimestamp_from_scanoutpos, + .get_scanout_position = radeon_get_crtc_scanout_position, .irq_preinstall = radeon_driver_irq_preinstall_kms, .irq_postinstall = radeon_driver_irq_postinstall_kms, .irq_uninstall = radeon_driver_irq_uninstall_kms, diff --git a/drivers/gpu/drm/radeon/radeon_gart.c b/drivers/gpu/drm/radeon/radeon_gart.c index c4777c8d0312..0b3ec35515f3 100644 --- a/drivers/gpu/drm/radeon/radeon_gart.c +++ b/drivers/gpu/drm/radeon/radeon_gart.c @@ -27,6 +27,9 @@ */ #include <drm/drmP.h> #include <drm/radeon_drm.h> +#ifdef CONFIG_X86 +#include <asm/set_memory.h> +#endif #include "radeon.h" /* diff --git a/drivers/gpu/drm/radeon/radeon_gem.c b/drivers/gpu/drm/radeon/radeon_gem.c index dddb372de2b9..574bf7e6b118 100644 --- a/drivers/gpu/drm/radeon/radeon_gem.c +++ b/drivers/gpu/drm/radeon/radeon_gem.c @@ -587,7 +587,7 @@ error_unreserve: ttm_eu_backoff_reservation(&ticket, &list); error_free: - drm_free_large(vm_bos); + kvfree(vm_bos); if (r && r != -ERESTARTSYS) DRM_ERROR("Couldn't update BO_VA (%d)\n", r); diff --git a/drivers/gpu/drm/radeon/radeon_kms.c b/drivers/gpu/drm/radeon/radeon_kms.c index e3e7cb1d10a2..6a68d440bc44 100644 --- a/drivers/gpu/drm/radeon/radeon_kms.c +++ b/drivers/gpu/drm/radeon/radeon_kms.c @@ -858,43 +858,6 @@ void radeon_disable_vblank_kms(struct drm_device *dev, int crtc) spin_unlock_irqrestore(&rdev->irq.lock, irqflags); } -/** - * radeon_get_vblank_timestamp_kms - get vblank timestamp - * - * @dev: drm dev pointer - * @crtc: crtc to get the timestamp for - * @max_error: max error - * @vblank_time: time value - * @flags: flags passed to the driver - * - * Gets the timestamp on the requested crtc based on the - * scanout position. (all asics). - * Returns postive status flags on success, negative error on failure. - */ -int radeon_get_vblank_timestamp_kms(struct drm_device *dev, int crtc, - int *max_error, - struct timeval *vblank_time, - unsigned flags) -{ - struct drm_crtc *drmcrtc; - struct radeon_device *rdev = dev->dev_private; - - if (crtc < 0 || crtc >= dev->num_crtcs) { - DRM_ERROR("Invalid crtc %d\n", crtc); - return -EINVAL; - } - - /* Get associated drm_crtc: */ - drmcrtc = &rdev->mode_info.crtcs[crtc]->base; - if (!drmcrtc) - return -EINVAL; - - /* Helper routine in DRM core does all the work: */ - return drm_calc_vbltimestamp_from_scanoutpos(dev, crtc, max_error, - vblank_time, flags, - &drmcrtc->hwmode); -} - const struct drm_ioctl_desc radeon_ioctls_kms[] = { DRM_IOCTL_DEF_DRV(RADEON_CP_INIT, drm_invalid_op, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY), DRM_IOCTL_DEF_DRV(RADEON_CP_START, drm_invalid_op, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY), diff --git a/drivers/gpu/drm/radeon/radeon_mode.h b/drivers/gpu/drm/radeon/radeon_mode.h index ad282648fc8b..00f5ec5c12c7 100644 --- a/drivers/gpu/drm/radeon/radeon_mode.h +++ b/drivers/gpu/drm/radeon/radeon_mode.h @@ -691,6 +691,9 @@ struct atom_voltage_table }; /* Driver internal use only flags of radeon_get_crtc_scanoutpos() */ +#define DRM_SCANOUTPOS_VALID (1 << 0) +#define DRM_SCANOUTPOS_IN_VBLANK (1 << 1) +#define DRM_SCANOUTPOS_ACCURATE (1 << 2) #define USE_REAL_VBLANKSTART (1 << 30) #define GET_DISTANCE_TO_VBLANKSTART (1 << 31) diff --git a/drivers/gpu/drm/radeon/radeon_object.c b/drivers/gpu/drm/radeon/radeon_object.c index bec2ec056de4..8b722297a05c 100644 --- a/drivers/gpu/drm/radeon/radeon_object.c +++ b/drivers/gpu/drm/radeon/radeon_object.c @@ -81,7 +81,7 @@ static void radeon_ttm_bo_destroy(struct ttm_buffer_object *tbo) list_del_init(&bo->list); mutex_unlock(&bo->rdev->gem.mutex); radeon_bo_clear_surface_reg(bo); - WARN_ON(!list_empty(&bo->va)); + WARN_ON_ONCE(!list_empty(&bo->va)); drm_gem_object_release(&bo->gem_base); kfree(bo); } diff --git a/drivers/gpu/drm/radeon/radeon_ring.c b/drivers/gpu/drm/radeon/radeon_ring.c index 8c7872339c2a..84802b201bef 100644 --- a/drivers/gpu/drm/radeon/radeon_ring.c +++ b/drivers/gpu/drm/radeon/radeon_ring.c @@ -314,7 +314,7 @@ unsigned radeon_ring_backup(struct radeon_device *rdev, struct radeon_ring *ring } /* and then save the content of the ring */ - *data = drm_malloc_ab(size, sizeof(uint32_t)); + *data = kvmalloc_array(size, sizeof(uint32_t), GFP_KERNEL); if (!*data) { mutex_unlock(&rdev->ring_lock); return 0; @@ -356,7 +356,7 @@ int radeon_ring_restore(struct radeon_device *rdev, struct radeon_ring *ring, } radeon_ring_unlock_commit(rdev, ring, false); - drm_free_large(data); + kvfree(data); return 0; } diff --git a/drivers/gpu/drm/radeon/radeon_test.c b/drivers/gpu/drm/radeon/radeon_test.c index 4fdc7bda7a7d..f5e9abfadb56 100644 --- a/drivers/gpu/drm/radeon/radeon_test.c +++ b/drivers/gpu/drm/radeon/radeon_test.c @@ -298,7 +298,12 @@ static int radeon_test_create_and_emit_fence(struct radeon_device *rdev, DRM_ERROR("Failed to lock ring A %d\n", ring->idx); return r; } - radeon_fence_emit(rdev, fence, ring->idx); + r = radeon_fence_emit(rdev, fence, ring->idx); + if (r) { + DRM_ERROR("Failed to emit fence\n"); + radeon_ring_unlock_undo(rdev, ring); + return r; + } radeon_ring_unlock_commit(rdev, ring, false); } return 0; diff --git a/drivers/gpu/drm/radeon/radeon_ttm.c b/drivers/gpu/drm/radeon/radeon_ttm.c index 8b7623b5a624..faa021396da3 100644 --- a/drivers/gpu/drm/radeon/radeon_ttm.c +++ b/drivers/gpu/drm/radeon/radeon_ttm.c @@ -29,11 +29,11 @@ * Thomas Hellstrom <thomas-at-tungstengraphics-dot-com> * Dave Airlie */ -#include <ttm/ttm_bo_api.h> -#include <ttm/ttm_bo_driver.h> -#include <ttm/ttm_placement.h> -#include <ttm/ttm_module.h> -#include <ttm/ttm_page_alloc.h> +#include <drm/ttm/ttm_bo_api.h> +#include <drm/ttm/ttm_bo_driver.h> +#include <drm/ttm/ttm_placement.h> +#include <drm/ttm/ttm_module.h> +#include <drm/ttm/ttm_page_alloc.h> #include <drm/drmP.h> #include <drm/radeon_drm.h> #include <linux/seq_file.h> diff --git a/drivers/gpu/drm/radeon/radeon_uvd.c b/drivers/gpu/drm/radeon/radeon_uvd.c index d34d1cf33895..7431eb4a11b7 100644 --- a/drivers/gpu/drm/radeon/radeon_uvd.c +++ b/drivers/gpu/drm/radeon/radeon_uvd.c @@ -621,7 +621,7 @@ static int radeon_uvd_cs_reloc(struct radeon_cs_parser *p, } /* TODO: is this still necessary on NI+ ? */ - if ((cmd == 0 || cmd == 0x3) && + if ((cmd == 0 || cmd == 1 || cmd == 0x3) && (start >> 28) != (p->rdev->uvd.gpu_addr >> 28)) { DRM_ERROR("msg/fb buffer %LX-%LX out of 256MB segment!\n", start, end); diff --git a/drivers/gpu/drm/radeon/radeon_vm.c b/drivers/gpu/drm/radeon/radeon_vm.c index a1358748cea5..5f68245579a3 100644 --- a/drivers/gpu/drm/radeon/radeon_vm.c +++ b/drivers/gpu/drm/radeon/radeon_vm.c @@ -132,8 +132,8 @@ struct radeon_bo_list *radeon_vm_get_bos(struct radeon_device *rdev, struct radeon_bo_list *list; unsigned i, idx; - list = drm_malloc_ab(vm->max_pde_used + 2, - sizeof(struct radeon_bo_list)); + list = kvmalloc_array(vm->max_pde_used + 2, + sizeof(struct radeon_bo_list), GFP_KERNEL); if (!list) return NULL; diff --git a/drivers/gpu/drm/radeon/rs780_dpm.c b/drivers/gpu/drm/radeon/rs780_dpm.c index 94b48fc1e266..b5e4e09a8996 100644 --- a/drivers/gpu/drm/radeon/rs780_dpm.c +++ b/drivers/gpu/drm/radeon/rs780_dpm.c @@ -22,7 +22,7 @@ * Authors: Alex Deucher */ -#include "drmP.h" +#include <drm/drmP.h> #include "radeon.h" #include "radeon_asic.h" #include "rs780d.h" diff --git a/drivers/gpu/drm/radeon/rv6xx_dpm.c b/drivers/gpu/drm/radeon/rv6xx_dpm.c index 25e29303b119..d91aa3944593 100644 --- a/drivers/gpu/drm/radeon/rv6xx_dpm.c +++ b/drivers/gpu/drm/radeon/rv6xx_dpm.c @@ -22,7 +22,7 @@ * Authors: Alex Deucher */ -#include "drmP.h" +#include <drm/drmP.h> #include "radeon.h" #include "radeon_asic.h" #include "rv6xxd.h" diff --git a/drivers/gpu/drm/radeon/rv730_dpm.c b/drivers/gpu/drm/radeon/rv730_dpm.c index d37ba2cb886e..38fdb4152e2a 100644 --- a/drivers/gpu/drm/radeon/rv730_dpm.c +++ b/drivers/gpu/drm/radeon/rv730_dpm.c @@ -22,7 +22,7 @@ * Authors: Alex Deucher */ -#include "drmP.h" +#include <drm/drmP.h> #include "radeon.h" #include "rv730d.h" #include "r600_dpm.h" diff --git a/drivers/gpu/drm/radeon/rv740_dpm.c b/drivers/gpu/drm/radeon/rv740_dpm.c index 4b850824fe06..afd597ec5085 100644 --- a/drivers/gpu/drm/radeon/rv740_dpm.c +++ b/drivers/gpu/drm/radeon/rv740_dpm.c @@ -22,7 +22,7 @@ * Authors: Alex Deucher */ -#include "drmP.h" +#include <drm/drmP.h> #include "radeon.h" #include "rv740d.h" #include "r600_dpm.h" diff --git a/drivers/gpu/drm/radeon/rv770_dpm.c b/drivers/gpu/drm/radeon/rv770_dpm.c index a010decf59af..cb2a7ec4e217 100644 --- a/drivers/gpu/drm/radeon/rv770_dpm.c +++ b/drivers/gpu/drm/radeon/rv770_dpm.c @@ -22,7 +22,7 @@ * Authors: Alex Deucher */ -#include "drmP.h" +#include <drm/drmP.h> #include "radeon.h" #include "radeon_asic.h" #include "rv770d.h" diff --git a/drivers/gpu/drm/radeon/rv770_smc.c b/drivers/gpu/drm/radeon/rv770_smc.c index b2a224407365..2b7ddee3984c 100644 --- a/drivers/gpu/drm/radeon/rv770_smc.c +++ b/drivers/gpu/drm/radeon/rv770_smc.c @@ -23,7 +23,7 @@ */ #include <linux/firmware.h> -#include "drmP.h" +#include <drm/drmP.h> #include "radeon.h" #include "rv770d.h" #include "rv770_dpm.h" diff --git a/drivers/gpu/drm/radeon/si.c b/drivers/gpu/drm/radeon/si.c index 528e5a49a214..ceee87f029d9 100644 --- a/drivers/gpu/drm/radeon/si.c +++ b/drivers/gpu/drm/radeon/si.c @@ -2204,23 +2204,10 @@ static u32 dce6_latency_watermark(struct dce6_wm_params *wm) a.full = dfixed_const(available_bandwidth); b.full = dfixed_const(wm->num_heads); a.full = dfixed_div(a, b); + tmp = div_u64((u64) dmif_size * (u64) wm->disp_clk, mc_latency + 512); + tmp = min(dfixed_trunc(a), tmp); - b.full = dfixed_const(mc_latency + 512); - c.full = dfixed_const(wm->disp_clk); - b.full = dfixed_div(b, c); - - c.full = dfixed_const(dmif_size); - b.full = dfixed_div(c, b); - - tmp = min(dfixed_trunc(a), dfixed_trunc(b)); - - b.full = dfixed_const(1000); - c.full = dfixed_const(wm->disp_clk); - b.full = dfixed_div(c, b); - c.full = dfixed_const(wm->bytes_per_pixel); - b.full = dfixed_mul(b, c); - - lb_fill_bw = min(tmp, dfixed_trunc(b)); + lb_fill_bw = min(tmp, wm->disp_clk * wm->bytes_per_pixel / 1000); a.full = dfixed_const(max_src_lines_per_dst_line * wm->src_width * wm->bytes_per_pixel); b.full = dfixed_const(1000); @@ -2287,7 +2274,7 @@ static void dce6_program_watermarks(struct radeon_device *rdev, struct drm_display_mode *mode = &radeon_crtc->base.mode; struct dce6_wm_params wm_low, wm_high; u32 dram_channels; - u32 pixel_period; + u32 active_time; u32 line_time = 0; u32 latency_watermark_a = 0, latency_watermark_b = 0; u32 priority_a_mark = 0, priority_b_mark = 0; @@ -2297,8 +2284,8 @@ static void dce6_program_watermarks(struct radeon_device *rdev, fixed20_12 a, b, c; if (radeon_crtc->base.enabled && num_heads && mode) { - pixel_period = 1000000 / (u32)mode->clock; - line_time = min((u32)mode->crtc_htotal * pixel_period, (u32)65535); + active_time = 1000000UL * (u32)mode->crtc_hdisplay / (u32)mode->clock; + line_time = min((u32) (1000000UL * (u32)mode->crtc_htotal / (u32)mode->clock), (u32)65535); priority_a_cnt = 0; priority_b_cnt = 0; @@ -2320,7 +2307,7 @@ static void dce6_program_watermarks(struct radeon_device *rdev, wm_high.disp_clk = mode->clock; wm_high.src_width = mode->crtc_hdisplay; - wm_high.active_time = mode->crtc_hdisplay * pixel_period; + wm_high.active_time = active_time; wm_high.blank_time = line_time - wm_high.active_time; wm_high.interlaced = false; if (mode->flags & DRM_MODE_FLAG_INTERLACE) @@ -2347,7 +2334,7 @@ static void dce6_program_watermarks(struct radeon_device *rdev, wm_low.disp_clk = mode->clock; wm_low.src_width = mode->crtc_hdisplay; - wm_low.active_time = mode->crtc_hdisplay * pixel_period; + wm_low.active_time = active_time; wm_low.blank_time = line_time - wm_low.active_time; wm_low.interlaced = false; if (mode->flags & DRM_MODE_FLAG_INTERLACE) diff --git a/drivers/gpu/drm/radeon/si_dpm.c b/drivers/gpu/drm/radeon/si_dpm.c index c7af9fdd20c7..ee3e74266a13 100644 --- a/drivers/gpu/drm/radeon/si_dpm.c +++ b/drivers/gpu/drm/radeon/si_dpm.c @@ -21,7 +21,7 @@ * */ -#include "drmP.h" +#include <drm/drmP.h> #include "radeon.h" #include "radeon_asic.h" #include "sid.h" diff --git a/drivers/gpu/drm/radeon/si_smc.c b/drivers/gpu/drm/radeon/si_smc.c index e5bb92f16775..51155abda8d8 100644 --- a/drivers/gpu/drm/radeon/si_smc.c +++ b/drivers/gpu/drm/radeon/si_smc.c @@ -23,7 +23,7 @@ */ #include <linux/firmware.h> -#include "drmP.h" +#include <drm/drmP.h> #include "radeon.h" #include "sid.h" #include "ppsmc.h" diff --git a/drivers/gpu/drm/radeon/sumo_dpm.c b/drivers/gpu/drm/radeon/sumo_dpm.c index f0d5c1724f55..fd4804829e46 100644 --- a/drivers/gpu/drm/radeon/sumo_dpm.c +++ b/drivers/gpu/drm/radeon/sumo_dpm.c @@ -21,7 +21,7 @@ * */ -#include "drmP.h" +#include <drm/drmP.h> #include "radeon.h" #include "radeon_asic.h" #include "sumod.h" diff --git a/drivers/gpu/drm/radeon/sumo_smc.c b/drivers/gpu/drm/radeon/sumo_smc.c index fb081d2ae374..cc051be42362 100644 --- a/drivers/gpu/drm/radeon/sumo_smc.c +++ b/drivers/gpu/drm/radeon/sumo_smc.c @@ -21,7 +21,7 @@ * */ -#include "drmP.h" +#include <drm/drmP.h> #include "radeon.h" #include "sumod.h" #include "sumo_dpm.h" diff --git a/drivers/gpu/drm/radeon/trinity_dpm.c b/drivers/gpu/drm/radeon/trinity_dpm.c index 6730367ac228..2ef7c4e5e495 100644 --- a/drivers/gpu/drm/radeon/trinity_dpm.c +++ b/drivers/gpu/drm/radeon/trinity_dpm.c @@ -21,7 +21,7 @@ * */ -#include "drmP.h" +#include <drm/drmP.h> #include "radeon.h" #include "radeon_asic.h" #include "trinityd.h" diff --git a/drivers/gpu/drm/radeon/trinity_smc.c b/drivers/gpu/drm/radeon/trinity_smc.c index 99dd0455334d..0310e36e3159 100644 --- a/drivers/gpu/drm/radeon/trinity_smc.c +++ b/drivers/gpu/drm/radeon/trinity_smc.c @@ -21,7 +21,7 @@ * */ -#include "drmP.h" +#include <drm/drmP.h> #include "radeon.h" #include "trinityd.h" #include "trinity_dpm.h" diff --git a/drivers/gpu/drm/rockchip/analogix_dp-rockchip.c b/drivers/gpu/drm/rockchip/analogix_dp-rockchip.c index d8fa7a9c9240..1bccd827d2e4 100644 --- a/drivers/gpu/drm/rockchip/analogix_dp-rockchip.c +++ b/drivers/gpu/drm/rockchip/analogix_dp-rockchip.c @@ -104,26 +104,18 @@ static void analogix_dp_psr_work(struct work_struct *work) { struct rockchip_dp_device *dp = container_of(work, typeof(*dp), psr_work); - struct drm_crtc *crtc = dp->encoder.crtc; - int psr_state = dp->psr_state; - int vact_end; int ret; unsigned long flags; - if (!crtc) - return; - - vact_end = crtc->mode.vtotal - crtc->mode.vsync_start + crtc->mode.vdisplay; - - ret = rockchip_drm_wait_line_flag(dp->encoder.crtc, vact_end, - PSR_WAIT_LINE_FLAG_TIMEOUT_MS); + ret = rockchip_drm_wait_vact_end(dp->encoder.crtc, + PSR_WAIT_LINE_FLAG_TIMEOUT_MS); if (ret) { dev_err(dp->dev, "line flag interrupt did not arrive\n"); return; } spin_lock_irqsave(&dp->psr_lock, flags); - if (psr_state == EDP_VSC_PSR_STATE_ACTIVE) + if (dp->psr_state == EDP_VSC_PSR_STATE_ACTIVE) analogix_dp_enable_psr(dp->dev); else analogix_dp_disable_psr(dp->dev); diff --git a/drivers/gpu/drm/rockchip/rockchip_drm_drv.h b/drivers/gpu/drm/rockchip/rockchip_drm_drv.h index a48fcce3f5f6..47905faf5586 100644 --- a/drivers/gpu/drm/rockchip/rockchip_drm_drv.h +++ b/drivers/gpu/drm/rockchip/rockchip_drm_drv.h @@ -62,8 +62,7 @@ int rockchip_drm_dma_attach_device(struct drm_device *drm_dev, struct device *dev); void rockchip_drm_dma_detach_device(struct drm_device *drm_dev, struct device *dev); -int rockchip_drm_wait_line_flag(struct drm_crtc *crtc, unsigned int line_num, - unsigned int mstimeout); +int rockchip_drm_wait_vact_end(struct drm_crtc *crtc, unsigned int mstimeout); extern struct platform_driver cdn_dp_driver; extern struct platform_driver dw_hdmi_rockchip_pltfm_driver; diff --git a/drivers/gpu/drm/rockchip/rockchip_drm_vop.c b/drivers/gpu/drm/rockchip/rockchip_drm_vop.c index 3f7a82d1e095..40a5e6ef6f2c 100644 --- a/drivers/gpu/drm/rockchip/rockchip_drm_vop.c +++ b/drivers/gpu/drm/rockchip/rockchip_drm_vop.c @@ -468,7 +468,7 @@ static bool vop_line_flag_irq_is_enabled(struct vop *vop) return !!line_flag_irq; } -static void vop_line_flag_irq_enable(struct vop *vop, int line_num) +static void vop_line_flag_irq_enable(struct vop *vop) { unsigned long flags; @@ -477,7 +477,6 @@ static void vop_line_flag_irq_enable(struct vop *vop, int line_num) spin_lock_irqsave(&vop->irq_lock, flags); - VOP_CTRL_SET(vop, line_flag_num[0], line_num); VOP_INTR_SET_TYPE(vop, clear, LINE_FLAG_INTR, 1); VOP_INTR_SET_TYPE(vop, enable, LINE_FLAG_INTR, 1); @@ -981,6 +980,8 @@ static void vop_crtc_enable(struct drm_crtc *crtc) VOP_CTRL_SET(vop, vact_st_end, val); VOP_CTRL_SET(vop, vpost_st_end, val); + VOP_CTRL_SET(vop, line_flag_num[0], vact_end); + clk_set_rate(vop->dclk, adjusted_mode->clock * 1000); VOP_CTRL_SET(vop, standby, 0); @@ -1507,19 +1508,16 @@ static void vop_win_init(struct vop *vop) } /** - * rockchip_drm_wait_line_flag - acqiure the give line flag event + * rockchip_drm_wait_vact_end * @crtc: CRTC to enable line flag - * @line_num: interested line number * @mstimeout: millisecond for timeout * - * Driver would hold here until the interested line flag interrupt have - * happened or timeout to wait. + * Wait for vact_end line flag irq or timeout. * * Returns: * Zero on success, negative errno on failure. */ -int rockchip_drm_wait_line_flag(struct drm_crtc *crtc, unsigned int line_num, - unsigned int mstimeout) +int rockchip_drm_wait_vact_end(struct drm_crtc *crtc, unsigned int mstimeout) { struct vop *vop = to_vop(crtc); unsigned long jiffies_left; @@ -1527,14 +1525,14 @@ int rockchip_drm_wait_line_flag(struct drm_crtc *crtc, unsigned int line_num, if (!crtc || !vop->is_enabled) return -ENODEV; - if (line_num > crtc->mode.vtotal || mstimeout <= 0) + if (mstimeout <= 0) return -EINVAL; if (vop_line_flag_irq_is_enabled(vop)) return -EBUSY; reinit_completion(&vop->line_flag_completion); - vop_line_flag_irq_enable(vop, line_num); + vop_line_flag_irq_enable(vop); jiffies_left = wait_for_completion_timeout(&vop->line_flag_completion, msecs_to_jiffies(mstimeout)); @@ -1547,7 +1545,7 @@ int rockchip_drm_wait_line_flag(struct drm_crtc *crtc, unsigned int line_num, return 0; } -EXPORT_SYMBOL(rockchip_drm_wait_line_flag); +EXPORT_SYMBOL(rockchip_drm_wait_vact_end); static int vop_bind(struct device *dev, struct device *master, void *data) { diff --git a/drivers/gpu/drm/savage/Makefile b/drivers/gpu/drm/savage/Makefile index d8f84ac7bb26..cfd436bb28e4 100644 --- a/drivers/gpu/drm/savage/Makefile +++ b/drivers/gpu/drm/savage/Makefile @@ -2,7 +2,6 @@ # Makefile for the drm device driver. This driver provides support for the # Direct Rendering Infrastructure (DRI) in XFree86 4.1.0 and higher. -ccflags-y = -Iinclude/drm savage-y := savage_drv.o savage_bci.o savage_state.o obj-$(CONFIG_DRM_SAVAGE)+= savage.o diff --git a/drivers/gpu/drm/selftests/test-drm_mm.c b/drivers/gpu/drm/selftests/test-drm_mm.c index fa356f5dae27..dfdd858eda0a 100644 --- a/drivers/gpu/drm/selftests/test-drm_mm.c +++ b/drivers/gpu/drm/selftests/test-drm_mm.c @@ -514,6 +514,8 @@ static int igt_reserve(void *ignored) ret = __igt_reserve(count, size + 1); if (ret) return ret; + + cond_resched(); } return 0; @@ -712,6 +714,10 @@ static int igt_insert(void *ignored) return ret; ret = __igt_insert(count, size + 1, false); + if (ret) + return ret; + + cond_resched(); } return 0; @@ -741,6 +747,10 @@ static int igt_replace(void *ignored) return ret; ret = __igt_insert(count, size + 1, true); + if (ret) + return ret; + + cond_resched(); } return 0; @@ -1011,6 +1021,8 @@ static int igt_insert_range(void *ignored) ret = __igt_insert_range(count, size, max/4+1, 3*max/4-1); if (ret) return ret; + + cond_resched(); } return 0; @@ -1056,6 +1068,7 @@ static int igt_align(void *ignored) drm_mm_for_each_node_safe(node, next, &mm) drm_mm_remove_node(node); DRM_MM_BUG_ON(!drm_mm_clean(&mm)); + cond_resched(); } ret = 0; @@ -1097,6 +1110,8 @@ static int igt_align_pot(int max) align, bit); goto out; } + + cond_resched(); } ret = 0; @@ -1471,6 +1486,8 @@ static int igt_evict(void *ignored) goto out; } } + + cond_resched(); } ret = 0; @@ -1566,6 +1583,8 @@ static int igt_evict_range(void *ignored) goto out; } } + + cond_resched(); } ret = 0; @@ -1683,6 +1702,7 @@ static int igt_topdown(void *ignored) drm_mm_for_each_node_safe(node, next, &mm) drm_mm_remove_node(node); DRM_MM_BUG_ON(!drm_mm_clean(&mm)); + cond_resched(); } ret = 0; @@ -1783,6 +1803,7 @@ static int igt_bottomup(void *ignored) drm_mm_for_each_node_safe(node, next, &mm) drm_mm_remove_node(node); DRM_MM_BUG_ON(!drm_mm_clean(&mm)); + cond_resched(); } ret = 0; @@ -1970,6 +1991,8 @@ static int igt_color(void *ignored) drm_mm_remove_node(node); kfree(node); } + + cond_resched(); } ret = 0; @@ -2047,6 +2070,7 @@ static int evict_color(struct drm_mm *mm, } } + cond_resched(); return 0; } @@ -2132,6 +2156,8 @@ static int igt_color_evict(void *ignored) goto out; } } + + cond_resched(); } ret = 0; @@ -2231,6 +2257,8 @@ static int igt_color_evict_range(void *ignored) goto out; } } + + cond_resched(); } ret = 0; diff --git a/drivers/gpu/drm/sis/Makefile b/drivers/gpu/drm/sis/Makefile index 441c061c3ad0..7bf4c130c8fd 100644 --- a/drivers/gpu/drm/sis/Makefile +++ b/drivers/gpu/drm/sis/Makefile @@ -2,7 +2,6 @@ # Makefile for the drm device driver. This driver provides support for the # Direct Rendering Infrastructure (DRI) in XFree86 4.1.0 and higher. -ccflags-y = -Iinclude/drm sis-y := sis_drv.o sis_mm.o obj-$(CONFIG_DRM_SIS) += sis.o diff --git a/drivers/gpu/drm/sti/sti_compositor.c b/drivers/gpu/drm/sti/sti_compositor.c index f62041fe8412..6e4bf68262db 100644 --- a/drivers/gpu/drm/sti/sti_compositor.c +++ b/drivers/gpu/drm/sti/sti_compositor.c @@ -89,7 +89,7 @@ static int sti_compositor_bind(struct device *dev, /* Nothing to do, wait for the second round */ break; default: - DRM_ERROR("Unknow subdev compoment type\n"); + DRM_ERROR("Unknown subdev component type\n"); return 1; } } @@ -129,7 +129,7 @@ static int sti_compositor_bind(struct device *dev, } break; default: - DRM_ERROR("Unknown subdev compoment type\n"); + DRM_ERROR("Unknown subdev component type\n"); return 1; } diff --git a/drivers/gpu/drm/sti/sti_cursor.c b/drivers/gpu/drm/sti/sti_cursor.c index cca75bddb9ad..5b3a41f74f21 100644 --- a/drivers/gpu/drm/sti/sti_cursor.c +++ b/drivers/gpu/drm/sti/sti_cursor.c @@ -33,7 +33,7 @@ #define STI_CURS_MAX_SIZE 128 /* - * pixmap dma buffer stucture + * pixmap dma buffer structure * * @paddr: physical address * @size: buffer size @@ -121,8 +121,7 @@ static int cursor_dbg_show(struct seq_file *s, void *data) cursor_dbg_cml(s, cursor, readl(cursor->regs + CUR_CML)); DBGFS_DUMP(CUR_AWS); DBGFS_DUMP(CUR_AWE); - seq_puts(s, "\n"); - + seq_putc(s, '\n'); return 0; } diff --git a/drivers/gpu/drm/sti/sti_dvo.c b/drivers/gpu/drm/sti/sti_dvo.c index bb23318a44b7..24ebc6b2f34d 100644 --- a/drivers/gpu/drm/sti/sti_dvo.c +++ b/drivers/gpu/drm/sti/sti_dvo.c @@ -186,8 +186,7 @@ static int dvo_dbg_show(struct seq_file *s, void *data) DBGFS_DUMP(DVO_LUT_PROG_MID); DBGFS_DUMP(DVO_LUT_PROG_HIGH); dvo_dbg_awg_microcode(s, dvo->regs + DVO_DIGSYNC_INSTR_I); - seq_puts(s, "\n"); - + seq_putc(s, '\n'); return 0; } diff --git a/drivers/gpu/drm/sti/sti_gdp.c b/drivers/gpu/drm/sti/sti_gdp.c index 88f16cdf6a4b..5ee0503945c8 100644 --- a/drivers/gpu/drm/sti/sti_gdp.c +++ b/drivers/gpu/drm/sti/sti_gdp.c @@ -149,7 +149,7 @@ static void gdp_dbg_ctl(struct seq_file *s, int val) seq_puts(s, "\tColor:"); for (i = 0; i < ARRAY_SIZE(gdp_format_to_str); i++) { if (gdp_format_to_str[i].format == (val & 0x1F)) { - seq_printf(s, gdp_format_to_str[i].name); + seq_puts(s, gdp_format_to_str[i].name); break; } } @@ -266,8 +266,7 @@ static void gdp_node_dump_node(struct seq_file *s, struct sti_gdp_node *node) seq_printf(s, "\n\tKEY2 0x%08X", node->gam_gdp_key2); seq_printf(s, "\n\tPPT 0x%08X", node->gam_gdp_ppt); gdp_dbg_ppt(s, node->gam_gdp_ppt); - seq_printf(s, "\n\tCML 0x%08X", node->gam_gdp_cml); - seq_puts(s, "\n"); + seq_printf(s, "\n\tCML 0x%08X\n", node->gam_gdp_cml); } static int gdp_node_dbg_show(struct seq_file *s, void *arg) diff --git a/drivers/gpu/drm/sti/sti_hda.c b/drivers/gpu/drm/sti/sti_hda.c index 0c0a75bc8bc3..d6ed909d9d75 100644 --- a/drivers/gpu/drm/sti/sti_hda.c +++ b/drivers/gpu/drm/sti/sti_hda.c @@ -320,8 +320,7 @@ static void hda_dbg_awg_microcode(struct seq_file *s, void __iomem *reg) { unsigned int i; - seq_puts(s, "\n\n"); - seq_puts(s, " HDA AWG microcode:"); + seq_puts(s, "\n\n HDA AWG microcode:"); for (i = 0; i < AWG_MAX_INST; i++) { if (i % 8 == 0) seq_printf(s, "\n %04X:", i); @@ -333,8 +332,7 @@ static void hda_dbg_video_dacs_ctrl(struct seq_file *s, void __iomem *reg) { u32 val = readl(reg); - seq_puts(s, "\n"); - seq_printf(s, "\n %-25s 0x%08X", "VIDEO_DACS_CONTROL", val); + seq_printf(s, "\n\n %-25s 0x%08X", "VIDEO_DACS_CONTROL", val); seq_puts(s, "\tHD DACs "); seq_puts(s, val & DAC_CFG_HD_HZUVW_OFF_MASK ? "disabled" : "enabled"); } @@ -356,8 +354,7 @@ static int hda_dbg_show(struct seq_file *s, void *data) hda_dbg_awg_microcode(s, hda->regs + HDA_SYNC_AWGI); if (hda->video_dacs_ctrl) hda_dbg_video_dacs_ctrl(s, hda->video_dacs_ctrl); - seq_puts(s, "\n"); - + seq_putc(s, '\n'); return 0; } diff --git a/drivers/gpu/drm/sti/sti_hdmi.c b/drivers/gpu/drm/sti/sti_hdmi.c index ce2dcba679d5..a59c95a8081b 100644 --- a/drivers/gpu/drm/sti/sti_hdmi.c +++ b/drivers/gpu/drm/sti/sti_hdmi.c @@ -592,7 +592,7 @@ static void hdmi_dbg_cfg(struct seq_file *s, int val) { int tmp; - seq_puts(s, "\t"); + seq_putc(s, '\t'); tmp = val & HDMI_CFG_HDMI_NOT_DVI; DBGFS_PRINT_STR("mode:", tmp ? "HDMI" : "DVI"); seq_puts(s, "\t\t\t\t\t"); @@ -616,7 +616,7 @@ static void hdmi_dbg_sta(struct seq_file *s, int val) { int tmp; - seq_puts(s, "\t"); + seq_putc(s, '\t'); tmp = (val & HDMI_STA_DLL_LCK); DBGFS_PRINT_STR("pll:", tmp ? "locked" : "not locked"); seq_puts(s, "\t\t\t\t\t"); @@ -632,7 +632,7 @@ static void hdmi_dbg_sw_di_cfg(struct seq_file *s, int val) "once every field", "once every frame"}; - seq_puts(s, "\t"); + seq_putc(s, '\t'); tmp = (val & HDMI_IFRAME_CFG_DI_N(HDMI_IFRAME_MASK, 1)); DBGFS_PRINT_STR("Data island 1:", en_di[tmp]); seq_puts(s, "\t\t\t\t\t"); @@ -664,16 +664,16 @@ static int hdmi_dbg_show(struct seq_file *s, void *data) DBGFS_DUMP("\n", HDMI_STA); hdmi_dbg_sta(s, hdmi_read(hdmi, HDMI_STA)); DBGFS_DUMP("", HDMI_ACTIVE_VID_XMIN); - seq_puts(s, "\t"); + seq_putc(s, '\t'); DBGFS_PRINT_INT("Xmin:", hdmi_read(hdmi, HDMI_ACTIVE_VID_XMIN)); DBGFS_DUMP("", HDMI_ACTIVE_VID_XMAX); - seq_puts(s, "\t"); + seq_putc(s, '\t'); DBGFS_PRINT_INT("Xmax:", hdmi_read(hdmi, HDMI_ACTIVE_VID_XMAX)); DBGFS_DUMP("", HDMI_ACTIVE_VID_YMIN); - seq_puts(s, "\t"); + seq_putc(s, '\t'); DBGFS_PRINT_INT("Ymin:", hdmi_read(hdmi, HDMI_ACTIVE_VID_YMIN)); DBGFS_DUMP("", HDMI_ACTIVE_VID_YMAX); - seq_puts(s, "\t"); + seq_putc(s, '\t'); DBGFS_PRINT_INT("Ymax:", hdmi_read(hdmi, HDMI_ACTIVE_VID_YMAX)); DBGFS_DUMP("", HDMI_SW_DI_CFG); hdmi_dbg_sw_di_cfg(s, hdmi_read(hdmi, HDMI_SW_DI_CFG)); @@ -692,8 +692,7 @@ static int hdmi_dbg_show(struct seq_file *s, void *data) DBGFS_DUMP_DI(HDMI_SW_DI_N_PKT_WORD4, HDMI_IFRAME_SLOT_AVI); DBGFS_DUMP_DI(HDMI_SW_DI_N_PKT_WORD5, HDMI_IFRAME_SLOT_AVI); DBGFS_DUMP_DI(HDMI_SW_DI_N_PKT_WORD6, HDMI_IFRAME_SLOT_AVI); - seq_puts(s, "\n"); - seq_printf(s, "\n AUDIO Infoframe (Data Island slot N=%d):", + seq_printf(s, "\n\n AUDIO Infoframe (Data Island slot N=%d):", HDMI_IFRAME_SLOT_AUDIO); DBGFS_DUMP_DI(HDMI_SW_DI_N_HEAD_WORD, HDMI_IFRAME_SLOT_AUDIO); DBGFS_DUMP_DI(HDMI_SW_DI_N_PKT_WORD0, HDMI_IFRAME_SLOT_AUDIO); @@ -703,8 +702,7 @@ static int hdmi_dbg_show(struct seq_file *s, void *data) DBGFS_DUMP_DI(HDMI_SW_DI_N_PKT_WORD4, HDMI_IFRAME_SLOT_AUDIO); DBGFS_DUMP_DI(HDMI_SW_DI_N_PKT_WORD5, HDMI_IFRAME_SLOT_AUDIO); DBGFS_DUMP_DI(HDMI_SW_DI_N_PKT_WORD6, HDMI_IFRAME_SLOT_AUDIO); - seq_puts(s, "\n"); - seq_printf(s, "\n VENDOR SPECIFIC Infoframe (Data Island slot N=%d):", + seq_printf(s, "\n\n VENDOR SPECIFIC Infoframe (Data Island slot N=%d):", HDMI_IFRAME_SLOT_VENDOR); DBGFS_DUMP_DI(HDMI_SW_DI_N_HEAD_WORD, HDMI_IFRAME_SLOT_VENDOR); DBGFS_DUMP_DI(HDMI_SW_DI_N_PKT_WORD0, HDMI_IFRAME_SLOT_VENDOR); @@ -714,8 +712,7 @@ static int hdmi_dbg_show(struct seq_file *s, void *data) DBGFS_DUMP_DI(HDMI_SW_DI_N_PKT_WORD4, HDMI_IFRAME_SLOT_VENDOR); DBGFS_DUMP_DI(HDMI_SW_DI_N_PKT_WORD5, HDMI_IFRAME_SLOT_VENDOR); DBGFS_DUMP_DI(HDMI_SW_DI_N_PKT_WORD6, HDMI_IFRAME_SLOT_VENDOR); - seq_puts(s, "\n"); - + seq_putc(s, '\n'); return 0; } @@ -771,6 +768,8 @@ static void sti_hdmi_disable(struct drm_bridge *bridge) clk_disable_unprepare(hdmi->clk_pix); hdmi->enabled = false; + + cec_notifier_set_phys_addr(hdmi->notifier, CEC_PHYS_ADDR_INVALID); } /** @@ -973,6 +972,7 @@ static int sti_hdmi_connector_get_modes(struct drm_connector *connector) DRM_DEBUG_KMS("%s : %dx%d cm\n", (hdmi->hdmi_monitor ? "hdmi monitor" : "dvi monitor"), edid->width_cm, edid->height_cm); + cec_notifier_set_phys_addr_from_edid(hdmi->notifier, edid); count = drm_add_edid_modes(connector, edid); drm_mode_connector_update_edid_property(connector, edid); @@ -1035,6 +1035,7 @@ sti_hdmi_connector_detect(struct drm_connector *connector, bool force) } DRM_DEBUG_DRIVER("hdmi cable disconnected\n"); + cec_notifier_set_phys_addr(hdmi->notifier, CEC_PHYS_ADDR_INVALID); return connector_status_disconnected; } @@ -1423,6 +1424,10 @@ static int sti_hdmi_probe(struct platform_device *pdev) goto release_adapter; } + hdmi->notifier = cec_notifier_get(&pdev->dev); + if (!hdmi->notifier) + goto release_adapter; + hdmi->reset = devm_reset_control_get(dev, "hdmi"); /* Take hdmi out of reset */ if (!IS_ERR(hdmi->reset)) @@ -1442,11 +1447,14 @@ static int sti_hdmi_remove(struct platform_device *pdev) { struct sti_hdmi *hdmi = dev_get_drvdata(&pdev->dev); + cec_notifier_set_phys_addr(hdmi->notifier, CEC_PHYS_ADDR_INVALID); + i2c_put_adapter(hdmi->ddc_adapt); if (hdmi->audio_pdev) platform_device_unregister(hdmi->audio_pdev); component_del(&pdev->dev, &sti_hdmi_ops); + cec_notifier_put(hdmi->notifier); return 0; } diff --git a/drivers/gpu/drm/sti/sti_hdmi.h b/drivers/gpu/drm/sti/sti_hdmi.h index 407012350f1a..c6469b56ce7e 100644 --- a/drivers/gpu/drm/sti/sti_hdmi.h +++ b/drivers/gpu/drm/sti/sti_hdmi.h @@ -11,6 +11,7 @@ #include <linux/platform_device.h> #include <drm/drmP.h> +#include <media/cec-notifier.h> #define HDMI_STA 0x0010 #define HDMI_STA_DLL_LCK BIT(5) @@ -64,6 +65,7 @@ static const struct drm_prop_enum_list colorspace_mode_names[] = { * @audio_pdev: ASoC hdmi-codec platform device * @audio: hdmi audio parameters. * @drm_connector: hdmi connector + * @notifier: hotplug detect notifier */ struct sti_hdmi { struct device dev; @@ -89,6 +91,7 @@ struct sti_hdmi { struct platform_device *audio_pdev; struct hdmi_audio_params audio; struct drm_connector *drm_connector; + struct cec_notifier *notifier; }; u32 hdmi_read(struct sti_hdmi *hdmi, int offset); diff --git a/drivers/gpu/drm/sti/sti_hqvdp.c b/drivers/gpu/drm/sti/sti_hqvdp.c index 66f843148ef7..a1c161f77804 100644 --- a/drivers/gpu/drm/sti/sti_hqvdp.c +++ b/drivers/gpu/drm/sti/sti_hqvdp.c @@ -625,8 +625,7 @@ static int hqvdp_dbg_show(struct seq_file *s, void *data) hqvdp_dbg_dump_cmd(s, (struct sti_hqvdp_cmd *)virt); } - seq_puts(s, "\n"); - + seq_putc(s, '\n'); return 0; } @@ -1357,12 +1356,12 @@ static int sti_hqvdp_probe(struct platform_device *pdev) /* Get Memory resources */ res = platform_get_resource(pdev, IORESOURCE_MEM, 0); - if (res == NULL) { + if (!res) { DRM_ERROR("Get memory resource failed\n"); return -ENXIO; } hqvdp->regs = devm_ioremap(dev, res->start, resource_size(res)); - if (hqvdp->regs == NULL) { + if (!hqvdp->regs) { DRM_ERROR("Register mapping failed\n"); return -ENXIO; } diff --git a/drivers/gpu/drm/sti/sti_mixer.c b/drivers/gpu/drm/sti/sti_mixer.c index 4ddc58f7fe2e..2bd1d46fe1cd 100644 --- a/drivers/gpu/drm/sti/sti_mixer.c +++ b/drivers/gpu/drm/sti/sti_mixer.c @@ -162,8 +162,7 @@ static int mixer_dbg_show(struct seq_file *s, void *arg) DBGFS_DUMP(GAM_MIXER_MBP); DBGFS_DUMP(GAM_MIXER_MX0); mixer_dbg_mxn(s, mixer->regs + GAM_MIXER_MX0); - seq_puts(s, "\n"); - + seq_putc(s, '\n'); return 0; } diff --git a/drivers/gpu/drm/sti/sti_tvout.c b/drivers/gpu/drm/sti/sti_tvout.c index 8b8ea717c121..8959fcc743a8 100644 --- a/drivers/gpu/drm/sti/sti_tvout.c +++ b/drivers/gpu/drm/sti/sti_tvout.c @@ -459,7 +459,7 @@ static void tvout_dbg_vip(struct seq_file *s, int val) "Aux (color matrix by-passed)", "", "", "", "", "", "Force value"}; - seq_puts(s, "\t"); + seq_putc(s, '\t'); mask = TVO_VIP_REORDER_MASK << TVO_VIP_REORDER_R_SHIFT; r = (val & mask) >> TVO_VIP_REORDER_R_SHIFT; mask = TVO_VIP_REORDER_MASK << TVO_VIP_REORDER_G_SHIFT; @@ -558,8 +558,7 @@ static int tvout_dbg_show(struct seq_file *s, void *data) DBGFS_DUMP(TVO_CSC_AUX_M6); DBGFS_DUMP(TVO_CSC_AUX_M7); DBGFS_DUMP(TVO_AUX_IN_VID_FORMAT); - seq_puts(s, "\n"); - + seq_putc(s, '\n'); return 0; } @@ -847,7 +846,7 @@ static int sti_tvout_probe(struct platform_device *pdev) tvout->dev = dev; - /* get Memory ressources */ + /* get memory resources */ res = platform_get_resource_byname(pdev, IORESOURCE_MEM, "tvout-reg"); if (!res) { DRM_ERROR("Invalid glue resource\n"); diff --git a/drivers/gpu/drm/sti/sti_vid.c b/drivers/gpu/drm/sti/sti_vid.c index 2ad59892b57e..577a3341d3c1 100644 --- a/drivers/gpu/drm/sti/sti_vid.c +++ b/drivers/gpu/drm/sti/sti_vid.c @@ -61,7 +61,7 @@ static void vid_dbg_ctl(struct seq_file *s, int val) { val = val >> 30; - seq_puts(s, "\t"); + seq_putc(s, '\t'); if (!(val & 1)) seq_puts(s, "NOT "); @@ -114,8 +114,7 @@ static int vid_dbg_show(struct seq_file *s, void *arg) DBGFS_DUMP(VID_BC); DBGFS_DUMP(VID_TINT); DBGFS_DUMP(VID_CSAT); - seq_puts(s, "\n"); - + seq_putc(s, '\n'); return 0; } diff --git a/drivers/gpu/drm/stm/Kconfig b/drivers/gpu/drm/stm/Kconfig new file mode 100644 index 000000000000..2c4817fb0890 --- /dev/null +++ b/drivers/gpu/drm/stm/Kconfig @@ -0,0 +1,16 @@ +config DRM_STM + tristate "DRM Support for STMicroelectronics SoC Series" + depends on DRM && (ARCH_STM32 || ARCH_MULTIPLATFORM) + select DRM_KMS_HELPER + select DRM_GEM_CMA_HELPER + select DRM_KMS_CMA_HELPER + select DRM_PANEL + select VIDEOMODE_HELPERS + select FB_PROVIDE_GET_FB_UNMAPPED_AREA + default y + + help + Enable support for the on-chip display controller on + STMicroelectronics STM32 MCUs. + To compile this driver as a module, choose M here: the module + will be called stm-drm. diff --git a/drivers/gpu/drm/stm/Makefile b/drivers/gpu/drm/stm/Makefile new file mode 100644 index 000000000000..a09ecf450218 --- /dev/null +++ b/drivers/gpu/drm/stm/Makefile @@ -0,0 +1,5 @@ +stm-drm-y := \ + drv.o \ + ltdc.o + +obj-$(CONFIG_DRM_STM) += stm-drm.o diff --git a/drivers/gpu/drm/stm/drv.c b/drivers/gpu/drm/stm/drv.c new file mode 100644 index 000000000000..83ab48f1fd00 --- /dev/null +++ b/drivers/gpu/drm/stm/drv.c @@ -0,0 +1,221 @@ +/* + * Copyright (C) STMicroelectronics SA 2017 + * + * Authors: Philippe Cornu <philippe.cornu@st.com> + * Yannick Fertre <yannick.fertre@st.com> + * Fabien Dessenne <fabien.dessenne@st.com> + * Mickael Reulier <mickael.reulier@st.com> + * + * License terms: GNU General Public License (GPL), version 2 + */ + +#include <linux/component.h> +#include <linux/of_platform.h> + +#include <drm/drm_atomic.h> +#include <drm/drm_atomic_helper.h> +#include <drm/drm_crtc_helper.h> +#include <drm/drm_fb_cma_helper.h> +#include <drm/drm_gem_cma_helper.h> + +#include "ltdc.h" + +#define DRIVER_NAME "stm" +#define DRIVER_DESC "STMicroelectronics SoC DRM" +#define DRIVER_DATE "20170330" +#define DRIVER_MAJOR 1 +#define DRIVER_MINOR 0 +#define DRIVER_PATCH_LEVEL 0 + +#define STM_MAX_FB_WIDTH 2048 +#define STM_MAX_FB_HEIGHT 2048 /* same as width to handle orientation */ + +static void drv_output_poll_changed(struct drm_device *ddev) +{ + struct ltdc_device *ldev = ddev->dev_private; + + drm_fbdev_cma_hotplug_event(ldev->fbdev); +} + +static const struct drm_mode_config_funcs drv_mode_config_funcs = { + .fb_create = drm_fb_cma_create, + .output_poll_changed = drv_output_poll_changed, + .atomic_check = drm_atomic_helper_check, + .atomic_commit = drm_atomic_helper_commit, +}; + +static void drv_lastclose(struct drm_device *ddev) +{ + struct ltdc_device *ldev = ddev->dev_private; + + DRM_DEBUG("%s\n", __func__); + + drm_fbdev_cma_restore_mode(ldev->fbdev); +} + +DEFINE_DRM_GEM_CMA_FOPS(drv_driver_fops); + +static struct drm_driver drv_driver = { + .driver_features = DRIVER_MODESET | DRIVER_GEM | DRIVER_PRIME | + DRIVER_ATOMIC, + .lastclose = drv_lastclose, + .name = DRIVER_NAME, + .desc = DRIVER_DESC, + .date = DRIVER_DATE, + .major = DRIVER_MAJOR, + .minor = DRIVER_MINOR, + .patchlevel = DRIVER_PATCH_LEVEL, + .fops = &drv_driver_fops, + .dumb_create = drm_gem_cma_dumb_create, + .dumb_map_offset = drm_gem_cma_dumb_map_offset, + .dumb_destroy = drm_gem_dumb_destroy, + .prime_handle_to_fd = drm_gem_prime_handle_to_fd, + .prime_fd_to_handle = drm_gem_prime_fd_to_handle, + .gem_free_object_unlocked = drm_gem_cma_free_object, + .gem_vm_ops = &drm_gem_cma_vm_ops, + .gem_prime_export = drm_gem_prime_export, + .gem_prime_import = drm_gem_prime_import, + .gem_prime_get_sg_table = drm_gem_cma_prime_get_sg_table, + .gem_prime_import_sg_table = drm_gem_cma_prime_import_sg_table, + .gem_prime_vmap = drm_gem_cma_prime_vmap, + .gem_prime_vunmap = drm_gem_cma_prime_vunmap, + .gem_prime_mmap = drm_gem_cma_prime_mmap, + .enable_vblank = ltdc_crtc_enable_vblank, + .disable_vblank = ltdc_crtc_disable_vblank, +}; + +static int drv_load(struct drm_device *ddev) +{ + struct platform_device *pdev = to_platform_device(ddev->dev); + struct drm_fbdev_cma *fbdev; + struct ltdc_device *ldev; + int ret; + + DRM_DEBUG("%s\n", __func__); + + ldev = devm_kzalloc(ddev->dev, sizeof(*ldev), GFP_KERNEL); + if (!ldev) + return -ENOMEM; + + ddev->dev_private = (void *)ldev; + + drm_mode_config_init(ddev); + + /* + * set max width and height as default value. + * this value would be used to check framebuffer size limitation + * at drm_mode_addfb(). + */ + ddev->mode_config.min_width = 0; + ddev->mode_config.min_height = 0; + ddev->mode_config.max_width = STM_MAX_FB_WIDTH; + ddev->mode_config.max_height = STM_MAX_FB_HEIGHT; + ddev->mode_config.funcs = &drv_mode_config_funcs; + + ret = ltdc_load(ddev); + if (ret) + goto err; + + drm_mode_config_reset(ddev); + drm_kms_helper_poll_init(ddev); + + if (ddev->mode_config.num_connector) { + ldev = ddev->dev_private; + fbdev = drm_fbdev_cma_init(ddev, 16, + ddev->mode_config.num_connector); + if (IS_ERR(fbdev)) { + DRM_DEBUG("Warning: fails to create fbdev\n"); + fbdev = NULL; + } + ldev->fbdev = fbdev; + } + + platform_set_drvdata(pdev, ddev); + + return 0; +err: + drm_mode_config_cleanup(ddev); + return ret; +} + +static void drv_unload(struct drm_device *ddev) +{ + struct ltdc_device *ldev = ddev->dev_private; + + DRM_DEBUG("%s\n", __func__); + + if (ldev->fbdev) { + drm_fbdev_cma_fini(ldev->fbdev); + ldev->fbdev = NULL; + } + drm_kms_helper_poll_fini(ddev); + ltdc_unload(ddev); + drm_mode_config_cleanup(ddev); +} + +static int stm_drm_platform_probe(struct platform_device *pdev) +{ + struct device *dev = &pdev->dev; + struct drm_device *ddev; + int ret; + + DRM_DEBUG("%s\n", __func__); + + dma_set_coherent_mask(dev, DMA_BIT_MASK(32)); + + ddev = drm_dev_alloc(&drv_driver, dev); + if (IS_ERR(ddev)) + return PTR_ERR(ddev); + + ret = drv_load(ddev); + if (ret) + goto err_unref; + + ret = drm_dev_register(ddev, 0); + if (ret) + goto err_unref; + + return 0; + +err_unref: + drm_dev_unref(ddev); + + return ret; +} + +static int stm_drm_platform_remove(struct platform_device *pdev) +{ + struct drm_device *ddev = platform_get_drvdata(pdev); + + DRM_DEBUG("%s\n", __func__); + + drm_dev_unregister(ddev); + drv_unload(ddev); + drm_dev_unref(ddev); + + return 0; +} + +static const struct of_device_id drv_dt_ids[] = { + { .compatible = "st,stm32-ltdc"}, + { /* end node */ }, +}; +MODULE_DEVICE_TABLE(of, drv_dt_ids); + +static struct platform_driver stm_drm_platform_driver = { + .probe = stm_drm_platform_probe, + .remove = stm_drm_platform_remove, + .driver = { + .name = DRIVER_NAME, + .of_match_table = drv_dt_ids, + }, +}; + +module_platform_driver(stm_drm_platform_driver); + +MODULE_AUTHOR("Philippe Cornu <philippe.cornu@st.com>"); +MODULE_AUTHOR("Yannick Fertre <yannick.fertre@st.com>"); +MODULE_AUTHOR("Fabien Dessenne <fabien.dessenne@st.com>"); +MODULE_AUTHOR("Mickael Reulier <mickael.reulier@st.com>"); +MODULE_DESCRIPTION("STMicroelectronics ST DRM LTDC driver"); +MODULE_LICENSE("GPL v2"); diff --git a/drivers/gpu/drm/stm/ltdc.c b/drivers/gpu/drm/stm/ltdc.c new file mode 100644 index 000000000000..700cc0800e51 --- /dev/null +++ b/drivers/gpu/drm/stm/ltdc.c @@ -0,0 +1,1160 @@ +/* + * Copyright (C) STMicroelectronics SA 2017 + * + * Authors: Philippe Cornu <philippe.cornu@st.com> + * Yannick Fertre <yannick.fertre@st.com> + * Fabien Dessenne <fabien.dessenne@st.com> + * Mickael Reulier <mickael.reulier@st.com> + * + * License terms: GNU General Public License (GPL), version 2 + */ + +#include <linux/clk.h> +#include <linux/component.h> +#include <linux/of_address.h> +#include <linux/of_graph.h> +#include <linux/reset.h> + +#include <drm/drm_atomic.h> +#include <drm/drm_atomic_helper.h> +#include <drm/drm_crtc_helper.h> +#include <drm/drm_fb_cma_helper.h> +#include <drm/drm_gem_cma_helper.h> +#include <drm/drm_of.h> +#include <drm/drm_panel.h> +#include <drm/drm_plane_helper.h> + +#include <video/videomode.h> + +#include "ltdc.h" + +#define NB_CRTC 1 +#define CRTC_MASK GENMASK(NB_CRTC - 1, 0) + +#define MAX_IRQ 4 + +#define HWVER_10200 0x010200 +#define HWVER_10300 0x010300 +#define HWVER_20101 0x020101 + +/* + * The address of some registers depends on the HW version: such registers have + * an extra offset specified with reg_ofs. + */ +#define REG_OFS_NONE 0 +#define REG_OFS_4 4 /* Insertion of "Layer Configuration 2" reg */ +#define REG_OFS (ldev->caps.reg_ofs) +#define LAY_OFS 0x80 /* Register Offset between 2 layers */ + +/* Global register offsets */ +#define LTDC_IDR 0x0000 /* IDentification */ +#define LTDC_LCR 0x0004 /* Layer Count */ +#define LTDC_SSCR 0x0008 /* Synchronization Size Configuration */ +#define LTDC_BPCR 0x000C /* Back Porch Configuration */ +#define LTDC_AWCR 0x0010 /* Active Width Configuration */ +#define LTDC_TWCR 0x0014 /* Total Width Configuration */ +#define LTDC_GCR 0x0018 /* Global Control */ +#define LTDC_GC1R 0x001C /* Global Configuration 1 */ +#define LTDC_GC2R 0x0020 /* Global Configuration 2 */ +#define LTDC_SRCR 0x0024 /* Shadow Reload Configuration */ +#define LTDC_GACR 0x0028 /* GAmma Correction */ +#define LTDC_BCCR 0x002C /* Background Color Configuration */ +#define LTDC_IER 0x0034 /* Interrupt Enable */ +#define LTDC_ISR 0x0038 /* Interrupt Status */ +#define LTDC_ICR 0x003C /* Interrupt Clear */ +#define LTDC_LIPCR 0x0040 /* Line Interrupt Position Configuration */ +#define LTDC_CPSR 0x0044 /* Current Position Status */ +#define LTDC_CDSR 0x0048 /* Current Display Status */ + +/* Layer register offsets */ +#define LTDC_L1LC1R (0x0080) /* L1 Layer Configuration 1 */ +#define LTDC_L1LC2R (0x0084) /* L1 Layer Configuration 2 */ +#define LTDC_L1CR (0x0084 + REG_OFS) /* L1 Control */ +#define LTDC_L1WHPCR (0x0088 + REG_OFS) /* L1 Window Hor Position Config */ +#define LTDC_L1WVPCR (0x008C + REG_OFS) /* L1 Window Vert Position Config */ +#define LTDC_L1CKCR (0x0090 + REG_OFS) /* L1 Color Keying Configuration */ +#define LTDC_L1PFCR (0x0094 + REG_OFS) /* L1 Pixel Format Configuration */ +#define LTDC_L1CACR (0x0098 + REG_OFS) /* L1 Constant Alpha Config */ +#define LTDC_L1DCCR (0x009C + REG_OFS) /* L1 Default Color Configuration */ +#define LTDC_L1BFCR (0x00A0 + REG_OFS) /* L1 Blend Factors Configuration */ +#define LTDC_L1FBBCR (0x00A4 + REG_OFS) /* L1 FrameBuffer Bus Control */ +#define LTDC_L1AFBCR (0x00A8 + REG_OFS) /* L1 AuxFB Control */ +#define LTDC_L1CFBAR (0x00AC + REG_OFS) /* L1 Color FrameBuffer Address */ +#define LTDC_L1CFBLR (0x00B0 + REG_OFS) /* L1 Color FrameBuffer Length */ +#define LTDC_L1CFBLNR (0x00B4 + REG_OFS) /* L1 Color FrameBuffer Line Nb */ +#define LTDC_L1AFBAR (0x00B8 + REG_OFS) /* L1 AuxFB Address */ +#define LTDC_L1AFBLR (0x00BC + REG_OFS) /* L1 AuxFB Length */ +#define LTDC_L1AFBLNR (0x00C0 + REG_OFS) /* L1 AuxFB Line Number */ +#define LTDC_L1CLUTWR (0x00C4 + REG_OFS) /* L1 CLUT Write */ +#define LTDC_L1YS1R (0x00E0 + REG_OFS) /* L1 YCbCr Scale 1 */ +#define LTDC_L1YS2R (0x00E4 + REG_OFS) /* L1 YCbCr Scale 2 */ + +/* Bit definitions */ +#define SSCR_VSH GENMASK(10, 0) /* Vertical Synchronization Height */ +#define SSCR_HSW GENMASK(27, 16) /* Horizontal Synchronization Width */ + +#define BPCR_AVBP GENMASK(10, 0) /* Accumulated Vertical Back Porch */ +#define BPCR_AHBP GENMASK(27, 16) /* Accumulated Horizontal Back Porch */ + +#define AWCR_AAH GENMASK(10, 0) /* Accumulated Active Height */ +#define AWCR_AAW GENMASK(27, 16) /* Accumulated Active Width */ + +#define TWCR_TOTALH GENMASK(10, 0) /* TOTAL Height */ +#define TWCR_TOTALW GENMASK(27, 16) /* TOTAL Width */ + +#define GCR_LTDCEN BIT(0) /* LTDC ENable */ +#define GCR_DEN BIT(16) /* Dither ENable */ +#define GCR_PCPOL BIT(28) /* Pixel Clock POLarity */ +#define GCR_DEPOL BIT(29) /* Data Enable POLarity */ +#define GCR_VSPOL BIT(30) /* Vertical Synchro POLarity */ +#define GCR_HSPOL BIT(31) /* Horizontal Synchro POLarity */ + +#define GC1R_WBCH GENMASK(3, 0) /* Width of Blue CHannel output */ +#define GC1R_WGCH GENMASK(7, 4) /* Width of Green Channel output */ +#define GC1R_WRCH GENMASK(11, 8) /* Width of Red Channel output */ +#define GC1R_PBEN BIT(12) /* Precise Blending ENable */ +#define GC1R_DT GENMASK(15, 14) /* Dithering Technique */ +#define GC1R_GCT GENMASK(19, 17) /* Gamma Correction Technique */ +#define GC1R_SHREN BIT(21) /* SHadow Registers ENabled */ +#define GC1R_BCP BIT(22) /* Background Colour Programmable */ +#define GC1R_BBEN BIT(23) /* Background Blending ENabled */ +#define GC1R_LNIP BIT(24) /* Line Number IRQ Position */ +#define GC1R_TP BIT(25) /* Timing Programmable */ +#define GC1R_IPP BIT(26) /* IRQ Polarity Programmable */ +#define GC1R_SPP BIT(27) /* Sync Polarity Programmable */ +#define GC1R_DWP BIT(28) /* Dither Width Programmable */ +#define GC1R_STREN BIT(29) /* STatus Registers ENabled */ +#define GC1R_BMEN BIT(31) /* Blind Mode ENabled */ + +#define GC2R_EDCA BIT(0) /* External Display Control Ability */ +#define GC2R_STSAEN BIT(1) /* Slave Timing Sync Ability ENabled */ +#define GC2R_DVAEN BIT(2) /* Dual-View Ability ENabled */ +#define GC2R_DPAEN BIT(3) /* Dual-Port Ability ENabled */ +#define GC2R_BW GENMASK(6, 4) /* Bus Width (log2 of nb of bytes) */ +#define GC2R_EDCEN BIT(7) /* External Display Control ENabled */ + +#define SRCR_IMR BIT(0) /* IMmediate Reload */ +#define SRCR_VBR BIT(1) /* Vertical Blanking Reload */ + +#define BCCR_BCBLACK 0x00 /* Background Color BLACK */ +#define BCCR_BCBLUE GENMASK(7, 0) /* Background Color BLUE */ +#define BCCR_BCGREEN GENMASK(15, 8) /* Background Color GREEN */ +#define BCCR_BCRED GENMASK(23, 16) /* Background Color RED */ +#define BCCR_BCWHITE GENMASK(23, 0) /* Background Color WHITE */ + +#define IER_LIE BIT(0) /* Line Interrupt Enable */ +#define IER_FUIE BIT(1) /* Fifo Underrun Interrupt Enable */ +#define IER_TERRIE BIT(2) /* Transfer ERRor Interrupt Enable */ +#define IER_RRIE BIT(3) /* Register Reload Interrupt enable */ + +#define ISR_LIF BIT(0) /* Line Interrupt Flag */ +#define ISR_FUIF BIT(1) /* Fifo Underrun Interrupt Flag */ +#define ISR_TERRIF BIT(2) /* Transfer ERRor Interrupt Flag */ +#define ISR_RRIF BIT(3) /* Register Reload Interrupt Flag */ + +#define LXCR_LEN BIT(0) /* Layer ENable */ +#define LXCR_COLKEN BIT(1) /* Color Keying Enable */ +#define LXCR_CLUTEN BIT(4) /* Color Look-Up Table ENable */ + +#define LXWHPCR_WHSTPOS GENMASK(11, 0) /* Window Horizontal StarT POSition */ +#define LXWHPCR_WHSPPOS GENMASK(27, 16) /* Window Horizontal StoP POSition */ + +#define LXWVPCR_WVSTPOS GENMASK(10, 0) /* Window Vertical StarT POSition */ +#define LXWVPCR_WVSPPOS GENMASK(26, 16) /* Window Vertical StoP POSition */ + +#define LXPFCR_PF GENMASK(2, 0) /* Pixel Format */ + +#define LXCACR_CONSTA GENMASK(7, 0) /* CONSTant Alpha */ + +#define LXBFCR_BF2 GENMASK(2, 0) /* Blending Factor 2 */ +#define LXBFCR_BF1 GENMASK(10, 8) /* Blending Factor 1 */ + +#define LXCFBLR_CFBLL GENMASK(12, 0) /* Color Frame Buffer Line Length */ +#define LXCFBLR_CFBP GENMASK(28, 16) /* Color Frame Buffer Pitch in bytes */ + +#define LXCFBLNR_CFBLN GENMASK(10, 0) /* Color Frame Buffer Line Number */ + +#define HSPOL_AL 0 /* Horizontal Sync POLarity Active Low */ +#define VSPOL_AL 0 /* Vertical Sync POLarity Active Low */ +#define DEPOL_AL 0 /* Data Enable POLarity Active Low */ +#define PCPOL_IPC 0 /* Input Pixel Clock */ +#define HSPOL_AH GCR_HSPOL /* Horizontal Sync POLarity Active High */ +#define VSPOL_AH GCR_VSPOL /* Vertical Sync POLarity Active High */ +#define DEPOL_AH GCR_DEPOL /* Data Enable POLarity Active High */ +#define PCPOL_IIPC GCR_PCPOL /* Inverted Input Pixel Clock */ +#define CONSTA_MAX 0xFF /* CONSTant Alpha MAX= 1.0 */ +#define BF1_PAXCA 0x600 /* Pixel Alpha x Constant Alpha */ +#define BF1_CA 0x400 /* Constant Alpha */ +#define BF2_1PAXCA 0x007 /* 1 - (Pixel Alpha x Constant Alpha) */ +#define BF2_1CA 0x005 /* 1 - Constant Alpha */ + +#define NB_PF 8 /* Max nb of HW pixel format */ + +enum ltdc_pix_fmt { + PF_NONE, + /* RGB formats */ + PF_ARGB8888, /* ARGB [32 bits] */ + PF_RGBA8888, /* RGBA [32 bits] */ + PF_RGB888, /* RGB [24 bits] */ + PF_RGB565, /* RGB [16 bits] */ + PF_ARGB1555, /* ARGB A:1 bit RGB:15 bits [16 bits] */ + PF_ARGB4444, /* ARGB A:4 bits R/G/B: 4 bits each [16 bits] */ + /* Indexed formats */ + PF_L8, /* Indexed 8 bits [8 bits] */ + PF_AL44, /* Alpha:4 bits + indexed 4 bits [8 bits] */ + PF_AL88 /* Alpha:8 bits + indexed 8 bits [16 bits] */ +}; + +/* The index gives the encoding of the pixel format for an HW version */ +static const enum ltdc_pix_fmt ltdc_pix_fmt_a0[NB_PF] = { + PF_ARGB8888, /* 0x00 */ + PF_RGB888, /* 0x01 */ + PF_RGB565, /* 0x02 */ + PF_ARGB1555, /* 0x03 */ + PF_ARGB4444, /* 0x04 */ + PF_L8, /* 0x05 */ + PF_AL44, /* 0x06 */ + PF_AL88 /* 0x07 */ +}; + +static const enum ltdc_pix_fmt ltdc_pix_fmt_a1[NB_PF] = { + PF_ARGB8888, /* 0x00 */ + PF_RGB888, /* 0x01 */ + PF_RGB565, /* 0x02 */ + PF_RGBA8888, /* 0x03 */ + PF_AL44, /* 0x04 */ + PF_L8, /* 0x05 */ + PF_ARGB1555, /* 0x06 */ + PF_ARGB4444 /* 0x07 */ +}; + +static inline u32 reg_read(void __iomem *base, u32 reg) +{ + return readl_relaxed(base + reg); +} + +static inline void reg_write(void __iomem *base, u32 reg, u32 val) +{ + writel_relaxed(val, base + reg); +} + +static inline void reg_set(void __iomem *base, u32 reg, u32 mask) +{ + reg_write(base, reg, reg_read(base, reg) | mask); +} + +static inline void reg_clear(void __iomem *base, u32 reg, u32 mask) +{ + reg_write(base, reg, reg_read(base, reg) & ~mask); +} + +static inline void reg_update_bits(void __iomem *base, u32 reg, u32 mask, + u32 val) +{ + reg_write(base, reg, (reg_read(base, reg) & ~mask) | val); +} + +static inline struct ltdc_device *crtc_to_ltdc(struct drm_crtc *crtc) +{ + return (struct ltdc_device *)crtc->dev->dev_private; +} + +static inline struct ltdc_device *plane_to_ltdc(struct drm_plane *plane) +{ + return (struct ltdc_device *)plane->dev->dev_private; +} + +static inline struct ltdc_device *encoder_to_ltdc(struct drm_encoder *enc) +{ + return (struct ltdc_device *)enc->dev->dev_private; +} + +static inline struct ltdc_device *connector_to_ltdc(struct drm_connector *con) +{ + return (struct ltdc_device *)con->dev->dev_private; +} + +static inline enum ltdc_pix_fmt to_ltdc_pixelformat(u32 drm_fmt) +{ + enum ltdc_pix_fmt pf; + + switch (drm_fmt) { + case DRM_FORMAT_ARGB8888: + case DRM_FORMAT_XRGB8888: + pf = PF_ARGB8888; + break; + case DRM_FORMAT_RGBA8888: + case DRM_FORMAT_RGBX8888: + pf = PF_RGBA8888; + break; + case DRM_FORMAT_RGB888: + pf = PF_RGB888; + break; + case DRM_FORMAT_RGB565: + pf = PF_RGB565; + break; + case DRM_FORMAT_ARGB1555: + case DRM_FORMAT_XRGB1555: + pf = PF_ARGB1555; + break; + case DRM_FORMAT_ARGB4444: + case DRM_FORMAT_XRGB4444: + pf = PF_ARGB4444; + break; + case DRM_FORMAT_C8: + pf = PF_L8; + break; + default: + pf = PF_NONE; + break; + /* Note: There are no DRM_FORMAT for AL44 and AL88 */ + } + + return pf; +} + +static inline u32 to_drm_pixelformat(enum ltdc_pix_fmt pf) +{ + switch (pf) { + case PF_ARGB8888: + return DRM_FORMAT_ARGB8888; + case PF_RGBA8888: + return DRM_FORMAT_RGBA8888; + case PF_RGB888: + return DRM_FORMAT_RGB888; + case PF_RGB565: + return DRM_FORMAT_RGB565; + case PF_ARGB1555: + return DRM_FORMAT_ARGB1555; + case PF_ARGB4444: + return DRM_FORMAT_ARGB4444; + case PF_L8: + return DRM_FORMAT_C8; + case PF_AL44: /* No DRM support */ + case PF_AL88: /* No DRM support */ + case PF_NONE: + default: + return 0; + } +} + +static irqreturn_t ltdc_irq_thread(int irq, void *arg) +{ + struct drm_device *ddev = arg; + struct ltdc_device *ldev = ddev->dev_private; + struct drm_crtc *crtc = drm_crtc_from_index(ddev, 0); + + /* Line IRQ : trigger the vblank event */ + if (ldev->irq_status & ISR_LIF) + drm_crtc_handle_vblank(crtc); + + /* Save FIFO Underrun & Transfer Error status */ + mutex_lock(&ldev->err_lock); + if (ldev->irq_status & ISR_FUIF) + ldev->error_status |= ISR_FUIF; + if (ldev->irq_status & ISR_TERRIF) + ldev->error_status |= ISR_TERRIF; + mutex_unlock(&ldev->err_lock); + + return IRQ_HANDLED; +} + +static irqreturn_t ltdc_irq(int irq, void *arg) +{ + struct drm_device *ddev = arg; + struct ltdc_device *ldev = ddev->dev_private; + + /* Read & Clear the interrupt status */ + ldev->irq_status = reg_read(ldev->regs, LTDC_ISR); + reg_write(ldev->regs, LTDC_ICR, ldev->irq_status); + + return IRQ_WAKE_THREAD; +} + +/* + * DRM_CRTC + */ + +static void ltdc_crtc_load_lut(struct drm_crtc *crtc) +{ + struct ltdc_device *ldev = crtc_to_ltdc(crtc); + unsigned int i, lay; + + for (lay = 0; lay < ldev->caps.nb_layers; lay++) + for (i = 0; i < 256; i++) + reg_write(ldev->regs, LTDC_L1CLUTWR + lay * LAY_OFS, + ldev->clut[i]); +} + +static void ltdc_crtc_enable(struct drm_crtc *crtc) +{ + struct ltdc_device *ldev = crtc_to_ltdc(crtc); + + DRM_DEBUG_DRIVER("\n"); + + /* Sets the background color value */ + reg_write(ldev->regs, LTDC_BCCR, BCCR_BCBLACK); + + /* Enable IRQ */ + reg_set(ldev->regs, LTDC_IER, IER_RRIE | IER_FUIE | IER_TERRIE); + + /* Immediately commit the planes */ + reg_set(ldev->regs, LTDC_SRCR, SRCR_IMR); + + /* Enable LTDC */ + reg_set(ldev->regs, LTDC_GCR, GCR_LTDCEN); + + drm_crtc_vblank_on(crtc); +} + +static void ltdc_crtc_disable(struct drm_crtc *crtc) +{ + struct ltdc_device *ldev = crtc_to_ltdc(crtc); + + DRM_DEBUG_DRIVER("\n"); + + drm_crtc_vblank_off(crtc); + + /* disable LTDC */ + reg_clear(ldev->regs, LTDC_GCR, GCR_LTDCEN); + + /* disable IRQ */ + reg_clear(ldev->regs, LTDC_IER, IER_RRIE | IER_FUIE | IER_TERRIE); + + /* immediately commit disable of layers before switching off LTDC */ + reg_set(ldev->regs, LTDC_SRCR, SRCR_IMR); +} + +static void ltdc_crtc_mode_set_nofb(struct drm_crtc *crtc) +{ + struct ltdc_device *ldev = crtc_to_ltdc(crtc); + struct drm_display_mode *mode = &crtc->state->adjusted_mode; + struct videomode vm; + int rate = mode->clock * 1000; + u32 hsync, vsync, accum_hbp, accum_vbp, accum_act_w, accum_act_h; + u32 total_width, total_height; + u32 val; + + drm_display_mode_to_videomode(mode, &vm); + + DRM_DEBUG_DRIVER("CRTC:%d mode:%s\n", crtc->base.id, mode->name); + DRM_DEBUG_DRIVER("Video mode: %dx%d", vm.hactive, vm.vactive); + DRM_DEBUG_DRIVER(" hfp %d hbp %d hsl %d vfp %d vbp %d vsl %d\n", + vm.hfront_porch, vm.hback_porch, vm.hsync_len, + vm.vfront_porch, vm.vback_porch, vm.vsync_len); + + /* Convert video timings to ltdc timings */ + hsync = vm.hsync_len - 1; + vsync = vm.vsync_len - 1; + accum_hbp = hsync + vm.hback_porch; + accum_vbp = vsync + vm.vback_porch; + accum_act_w = accum_hbp + vm.hactive; + accum_act_h = accum_vbp + vm.vactive; + total_width = accum_act_w + vm.hfront_porch; + total_height = accum_act_h + vm.vfront_porch; + + clk_disable(ldev->pixel_clk); + + if (clk_set_rate(ldev->pixel_clk, rate) < 0) { + DRM_ERROR("Cannot set rate (%dHz) for pixel clk\n", rate); + return; + } + + clk_enable(ldev->pixel_clk); + + /* Configures the HS, VS, DE and PC polarities. */ + val = HSPOL_AL | VSPOL_AL | DEPOL_AL | PCPOL_IPC; + + if (vm.flags & DISPLAY_FLAGS_HSYNC_HIGH) + val |= HSPOL_AH; + + if (vm.flags & DISPLAY_FLAGS_VSYNC_HIGH) + val |= VSPOL_AH; + + if (vm.flags & DISPLAY_FLAGS_DE_HIGH) + val |= DEPOL_AH; + + if (vm.flags & DISPLAY_FLAGS_PIXDATA_NEGEDGE) + val |= PCPOL_IIPC; + + reg_update_bits(ldev->regs, LTDC_GCR, + GCR_HSPOL | GCR_VSPOL | GCR_DEPOL | GCR_PCPOL, val); + + /* Set Synchronization size */ + val = (hsync << 16) | vsync; + reg_update_bits(ldev->regs, LTDC_SSCR, SSCR_VSH | SSCR_HSW, val); + + /* Set Accumulated Back porch */ + val = (accum_hbp << 16) | accum_vbp; + reg_update_bits(ldev->regs, LTDC_BPCR, BPCR_AVBP | BPCR_AHBP, val); + + /* Set Accumulated Active Width */ + val = (accum_act_w << 16) | accum_act_h; + reg_update_bits(ldev->regs, LTDC_AWCR, AWCR_AAW | AWCR_AAH, val); + + /* Set total width & height */ + val = (total_width << 16) | total_height; + reg_update_bits(ldev->regs, LTDC_TWCR, TWCR_TOTALH | TWCR_TOTALW, val); + + reg_write(ldev->regs, LTDC_LIPCR, (accum_act_h + 1)); +} + +static void ltdc_crtc_atomic_flush(struct drm_crtc *crtc, + struct drm_crtc_state *old_crtc_state) +{ + struct ltdc_device *ldev = crtc_to_ltdc(crtc); + struct drm_pending_vblank_event *event = crtc->state->event; + + DRM_DEBUG_ATOMIC("\n"); + + /* Commit shadow registers = update planes at next vblank */ + reg_set(ldev->regs, LTDC_SRCR, SRCR_VBR); + + if (event) { + crtc->state->event = NULL; + + spin_lock_irq(&crtc->dev->event_lock); + if (drm_crtc_vblank_get(crtc) == 0) + drm_crtc_arm_vblank_event(crtc, event); + else + drm_crtc_send_vblank_event(crtc, event); + spin_unlock_irq(&crtc->dev->event_lock); + } +} + +static struct drm_crtc_helper_funcs ltdc_crtc_helper_funcs = { + .load_lut = ltdc_crtc_load_lut, + .enable = ltdc_crtc_enable, + .disable = ltdc_crtc_disable, + .mode_set_nofb = ltdc_crtc_mode_set_nofb, + .atomic_flush = ltdc_crtc_atomic_flush, +}; + +int ltdc_crtc_enable_vblank(struct drm_device *ddev, unsigned int pipe) +{ + struct ltdc_device *ldev = ddev->dev_private; + + DRM_DEBUG_DRIVER("\n"); + reg_set(ldev->regs, LTDC_IER, IER_LIE); + + return 0; +} + +void ltdc_crtc_disable_vblank(struct drm_device *ddev, unsigned int pipe) +{ + struct ltdc_device *ldev = ddev->dev_private; + + DRM_DEBUG_DRIVER("\n"); + reg_clear(ldev->regs, LTDC_IER, IER_LIE); +} + +static struct drm_crtc_funcs ltdc_crtc_funcs = { + .destroy = drm_crtc_cleanup, + .set_config = drm_atomic_helper_set_config, + .page_flip = drm_atomic_helper_page_flip, + .reset = drm_atomic_helper_crtc_reset, + .atomic_duplicate_state = drm_atomic_helper_crtc_duplicate_state, + .atomic_destroy_state = drm_atomic_helper_crtc_destroy_state, +}; + +/* + * DRM_PLANE + */ + +static int ltdc_plane_atomic_check(struct drm_plane *plane, + struct drm_plane_state *state) +{ + struct drm_framebuffer *fb = state->fb; + u32 src_x, src_y, src_w, src_h; + + DRM_DEBUG_DRIVER("\n"); + + if (!fb) + return 0; + + /* convert src_ from 16:16 format */ + src_x = state->src_x >> 16; + src_y = state->src_y >> 16; + src_w = state->src_w >> 16; + src_h = state->src_h >> 16; + + /* Reject scaling */ + if ((src_w != state->crtc_w) || (src_h != state->crtc_h)) { + DRM_ERROR("Scaling is not supported"); + return -EINVAL; + } + + return 0; +} + +static void ltdc_plane_atomic_update(struct drm_plane *plane, + struct drm_plane_state *oldstate) +{ + struct ltdc_device *ldev = plane_to_ltdc(plane); + struct drm_plane_state *state = plane->state; + struct drm_framebuffer *fb = state->fb; + u32 lofs = plane->index * LAY_OFS; + u32 x0 = state->crtc_x; + u32 x1 = state->crtc_x + state->crtc_w - 1; + u32 y0 = state->crtc_y; + u32 y1 = state->crtc_y + state->crtc_h - 1; + u32 src_x, src_y, src_w, src_h; + u32 val, pitch_in_bytes, line_length, paddr, ahbp, avbp, bpcr; + enum ltdc_pix_fmt pf; + + if (!state->crtc || !fb) { + DRM_DEBUG_DRIVER("fb or crtc NULL"); + return; + } + + /* convert src_ from 16:16 format */ + src_x = state->src_x >> 16; + src_y = state->src_y >> 16; + src_w = state->src_w >> 16; + src_h = state->src_h >> 16; + + DRM_DEBUG_DRIVER( + "plane:%d fb:%d (%dx%d)@(%d,%d) -> (%dx%d)@(%d,%d)\n", + plane->base.id, fb->base.id, + src_w, src_h, src_x, src_y, + state->crtc_w, state->crtc_h, state->crtc_x, state->crtc_y); + + bpcr = reg_read(ldev->regs, LTDC_BPCR); + ahbp = (bpcr & BPCR_AHBP) >> 16; + avbp = bpcr & BPCR_AVBP; + + /* Configures the horizontal start and stop position */ + val = ((x1 + 1 + ahbp) << 16) + (x0 + 1 + ahbp); + reg_update_bits(ldev->regs, LTDC_L1WHPCR + lofs, + LXWHPCR_WHSTPOS | LXWHPCR_WHSPPOS, val); + + /* Configures the vertical start and stop position */ + val = ((y1 + 1 + avbp) << 16) + (y0 + 1 + avbp); + reg_update_bits(ldev->regs, LTDC_L1WVPCR + lofs, + LXWVPCR_WVSTPOS | LXWVPCR_WVSPPOS, val); + + /* Specifies the pixel format */ + pf = to_ltdc_pixelformat(fb->format->format); + for (val = 0; val < NB_PF; val++) + if (ldev->caps.pix_fmt_hw[val] == pf) + break; + + if (val == NB_PF) { + DRM_ERROR("Pixel format %.4s not supported\n", + (char *)&fb->format->format); + val = 0; /* set by default ARGB 32 bits */ + } + reg_update_bits(ldev->regs, LTDC_L1PFCR + lofs, LXPFCR_PF, val); + + /* Configures the color frame buffer pitch in bytes & line length */ + pitch_in_bytes = fb->pitches[0]; + line_length = drm_format_plane_cpp(fb->format->format, 0) * + (x1 - x0 + 1) + (ldev->caps.bus_width >> 3) - 1; + val = ((pitch_in_bytes << 16) | line_length); + reg_update_bits(ldev->regs, LTDC_L1CFBLR + lofs, + LXCFBLR_CFBLL | LXCFBLR_CFBP, val); + + /* Specifies the constant alpha value */ + val = CONSTA_MAX; + reg_update_bits(ldev->regs, LTDC_L1CACR + lofs, + LXCACR_CONSTA, val); + + /* Specifies the blending factors */ + val = BF1_PAXCA | BF2_1PAXCA; + reg_update_bits(ldev->regs, LTDC_L1BFCR + lofs, + LXBFCR_BF2 | LXBFCR_BF1, val); + + /* Configures the frame buffer line number */ + val = y1 - y0 + 1; + reg_update_bits(ldev->regs, LTDC_L1CFBLNR + lofs, + LXCFBLNR_CFBLN, val); + + /* Sets the FB address */ + paddr = (u32)drm_fb_cma_get_gem_addr(fb, state, 0); + + DRM_DEBUG_DRIVER("fb: phys 0x%08x", paddr); + reg_write(ldev->regs, LTDC_L1CFBAR + lofs, paddr); + + /* Enable layer and CLUT if needed */ + val = fb->format->format == DRM_FORMAT_C8 ? LXCR_CLUTEN : 0; + val |= LXCR_LEN; + reg_update_bits(ldev->regs, LTDC_L1CR + lofs, + LXCR_LEN | LXCR_CLUTEN, val); + + mutex_lock(&ldev->err_lock); + if (ldev->error_status & ISR_FUIF) { + DRM_DEBUG_DRIVER("Fifo underrun\n"); + ldev->error_status &= ~ISR_FUIF; + } + if (ldev->error_status & ISR_TERRIF) { + DRM_DEBUG_DRIVER("Transfer error\n"); + ldev->error_status &= ~ISR_TERRIF; + } + mutex_unlock(&ldev->err_lock); +} + +static void ltdc_plane_atomic_disable(struct drm_plane *plane, + struct drm_plane_state *oldstate) +{ + struct ltdc_device *ldev = plane_to_ltdc(plane); + u32 lofs = plane->index * LAY_OFS; + + /* disable layer */ + reg_clear(ldev->regs, LTDC_L1CR + lofs, LXCR_LEN); + + DRM_DEBUG_DRIVER("CRTC:%d plane:%d\n", + oldstate->crtc->base.id, plane->base.id); +} + +static struct drm_plane_funcs ltdc_plane_funcs = { + .update_plane = drm_atomic_helper_update_plane, + .disable_plane = drm_atomic_helper_disable_plane, + .destroy = drm_plane_cleanup, + .set_property = drm_atomic_helper_plane_set_property, + .reset = drm_atomic_helper_plane_reset, + .atomic_duplicate_state = drm_atomic_helper_plane_duplicate_state, + .atomic_destroy_state = drm_atomic_helper_plane_destroy_state, +}; + +static const struct drm_plane_helper_funcs ltdc_plane_helper_funcs = { + .atomic_check = ltdc_plane_atomic_check, + .atomic_update = ltdc_plane_atomic_update, + .atomic_disable = ltdc_plane_atomic_disable, +}; + +static struct drm_plane *ltdc_plane_create(struct drm_device *ddev, + enum drm_plane_type type) +{ + unsigned long possible_crtcs = CRTC_MASK; + struct ltdc_device *ldev = ddev->dev_private; + struct device *dev = ddev->dev; + struct drm_plane *plane; + unsigned int i, nb_fmt = 0; + u32 formats[NB_PF]; + u32 drm_fmt; + int ret; + + /* Get supported pixel formats */ + for (i = 0; i < NB_PF; i++) { + drm_fmt = to_drm_pixelformat(ldev->caps.pix_fmt_hw[i]); + if (!drm_fmt) + continue; + formats[nb_fmt++] = drm_fmt; + } + + plane = devm_kzalloc(dev, sizeof(*plane), GFP_KERNEL); + if (!plane) + return 0; + + ret = drm_universal_plane_init(ddev, plane, possible_crtcs, + <dc_plane_funcs, formats, nb_fmt, + type, NULL); + if (ret < 0) + return 0; + + drm_plane_helper_add(plane, <dc_plane_helper_funcs); + + DRM_DEBUG_DRIVER("plane:%d created\n", plane->base.id); + + return plane; +} + +static void ltdc_plane_destroy_all(struct drm_device *ddev) +{ + struct drm_plane *plane, *plane_temp; + + list_for_each_entry_safe(plane, plane_temp, + &ddev->mode_config.plane_list, head) + drm_plane_cleanup(plane); +} + +static int ltdc_crtc_init(struct drm_device *ddev, struct drm_crtc *crtc) +{ + struct ltdc_device *ldev = ddev->dev_private; + struct drm_plane *primary, *overlay; + unsigned int i; + int res; + + primary = ltdc_plane_create(ddev, DRM_PLANE_TYPE_PRIMARY); + if (!primary) { + DRM_ERROR("Can not create primary plane\n"); + return -EINVAL; + } + + res = drm_crtc_init_with_planes(ddev, crtc, primary, NULL, + <dc_crtc_funcs, NULL); + if (res) { + DRM_ERROR("Can not initialize CRTC\n"); + goto cleanup; + } + + drm_crtc_helper_add(crtc, <dc_crtc_helper_funcs); + + DRM_DEBUG_DRIVER("CRTC:%d created\n", crtc->base.id); + + /* Add planes. Note : the first layer is used by primary plane */ + for (i = 1; i < ldev->caps.nb_layers; i++) { + overlay = ltdc_plane_create(ddev, DRM_PLANE_TYPE_OVERLAY); + if (!overlay) { + res = -ENOMEM; + DRM_ERROR("Can not create overlay plane %d\n", i); + goto cleanup; + } + } + + return 0; + +cleanup: + ltdc_plane_destroy_all(ddev); + return res; +} + +/* + * DRM_ENCODER + */ + +static void ltdc_rgb_encoder_enable(struct drm_encoder *encoder) +{ + struct ltdc_device *ldev = encoder_to_ltdc(encoder); + + DRM_DEBUG_DRIVER("\n"); + + drm_panel_prepare(ldev->panel); + drm_panel_enable(ldev->panel); +} + +static void ltdc_rgb_encoder_disable(struct drm_encoder *encoder) +{ + struct ltdc_device *ldev = encoder_to_ltdc(encoder); + + DRM_DEBUG_DRIVER("\n"); + + drm_panel_disable(ldev->panel); + drm_panel_unprepare(ldev->panel); +} + +static const struct drm_encoder_helper_funcs ltdc_rgb_encoder_helper_funcs = { + .enable = ltdc_rgb_encoder_enable, + .disable = ltdc_rgb_encoder_disable, +}; + +static const struct drm_encoder_funcs ltdc_rgb_encoder_funcs = { + .destroy = drm_encoder_cleanup, +}; + +static struct drm_encoder *ltdc_rgb_encoder_create(struct drm_device *ddev) +{ + struct drm_encoder *encoder; + + encoder = devm_kzalloc(ddev->dev, sizeof(*encoder), GFP_KERNEL); + if (!encoder) + return NULL; + + encoder->possible_crtcs = CRTC_MASK; + encoder->possible_clones = 0; /* No cloning support */ + + drm_encoder_init(ddev, encoder, <dc_rgb_encoder_funcs, + DRM_MODE_ENCODER_DPI, NULL); + + drm_encoder_helper_add(encoder, <dc_rgb_encoder_helper_funcs); + + DRM_DEBUG_DRIVER("RGB encoder:%d created\n", encoder->base.id); + + return encoder; +} + +/* + * DRM_CONNECTOR + */ + +static int ltdc_rgb_connector_get_modes(struct drm_connector *connector) +{ + struct drm_device *ddev = connector->dev; + struct ltdc_device *ldev = ddev->dev_private; + int ret = 0; + + DRM_DEBUG_DRIVER("\n"); + + if (ldev->panel) + ret = drm_panel_get_modes(ldev->panel); + + return ret < 0 ? 0 : ret; +} + +static struct drm_connector_helper_funcs ltdc_rgb_connector_helper_funcs = { + .get_modes = ltdc_rgb_connector_get_modes, +}; + +static enum drm_connector_status +ltdc_rgb_connector_detect(struct drm_connector *connector, bool force) +{ + struct ltdc_device *ldev = connector_to_ltdc(connector); + + return ldev->panel ? connector_status_connected : + connector_status_disconnected; +} + +static void ltdc_rgb_connector_destroy(struct drm_connector *connector) +{ + DRM_DEBUG_DRIVER("\n"); + + drm_connector_unregister(connector); + drm_connector_cleanup(connector); +} + +static const struct drm_connector_funcs ltdc_rgb_connector_funcs = { + .dpms = drm_atomic_helper_connector_dpms, + .fill_modes = drm_helper_probe_single_connector_modes, + .detect = ltdc_rgb_connector_detect, + .destroy = ltdc_rgb_connector_destroy, + .reset = drm_atomic_helper_connector_reset, + .atomic_duplicate_state = drm_atomic_helper_connector_duplicate_state, + .atomic_destroy_state = drm_atomic_helper_connector_destroy_state, +}; + +struct drm_connector *ltdc_rgb_connector_create(struct drm_device *ddev) +{ + struct drm_connector *connector; + int err; + + connector = devm_kzalloc(ddev->dev, sizeof(*connector), GFP_KERNEL); + if (!connector) { + DRM_ERROR("Failed to allocate connector\n"); + return NULL; + } + + connector->polled = DRM_CONNECTOR_POLL_HPD; + + err = drm_connector_init(ddev, connector, <dc_rgb_connector_funcs, + DRM_MODE_CONNECTOR_DPI); + if (err) { + DRM_ERROR("Failed to initialize connector\n"); + return NULL; + } + + drm_connector_helper_add(connector, <dc_rgb_connector_helper_funcs); + + DRM_DEBUG_DRIVER("RGB connector:%d created\n", connector->base.id); + + return connector; +} + +static int ltdc_get_caps(struct drm_device *ddev) +{ + struct ltdc_device *ldev = ddev->dev_private; + u32 bus_width_log2, lcr, gc2r; + + /* at least 1 layer must be managed */ + lcr = reg_read(ldev->regs, LTDC_LCR); + + ldev->caps.nb_layers = max_t(int, lcr, 1); + + /* set data bus width */ + gc2r = reg_read(ldev->regs, LTDC_GC2R); + bus_width_log2 = (gc2r & GC2R_BW) >> 4; + ldev->caps.bus_width = 8 << bus_width_log2; + ldev->caps.hw_version = reg_read(ldev->regs, LTDC_IDR); + + switch (ldev->caps.hw_version) { + case HWVER_10200: + case HWVER_10300: + ldev->caps.reg_ofs = REG_OFS_NONE; + ldev->caps.pix_fmt_hw = ltdc_pix_fmt_a0; + break; + case HWVER_20101: + ldev->caps.reg_ofs = REG_OFS_4; + ldev->caps.pix_fmt_hw = ltdc_pix_fmt_a1; + break; + default: + return -ENODEV; + } + + return 0; +} + +static struct drm_panel *ltdc_get_panel(struct drm_device *ddev) +{ + struct device *dev = ddev->dev; + struct device_node *np = dev->of_node; + struct device_node *entity, *port = NULL; + struct drm_panel *panel = NULL; + + DRM_DEBUG_DRIVER("\n"); + + /* + * Parse ltdc node to get remote port and find RGB panel / HDMI slave + * If a dsi or a bridge (hdmi, lvds...) is connected to ltdc, + * a remote port & RGB panel will not be found. + */ + for_each_endpoint_of_node(np, entity) { + if (!of_device_is_available(entity)) + continue; + + port = of_graph_get_remote_port_parent(entity); + if (port) { + panel = of_drm_find_panel(port); + of_node_put(port); + if (panel) { + DRM_DEBUG_DRIVER("remote panel %s\n", + port->full_name); + } else { + DRM_DEBUG_DRIVER("panel missing\n"); + of_node_put(entity); + } + } + } + + return panel; +} + +int ltdc_load(struct drm_device *ddev) +{ + struct platform_device *pdev = to_platform_device(ddev->dev); + struct ltdc_device *ldev = ddev->dev_private; + struct device *dev = ddev->dev; + struct device_node *np = dev->of_node; + struct drm_encoder *encoder; + struct drm_connector *connector = NULL; + struct drm_crtc *crtc; + struct reset_control *rstc; + struct resource res; + int irq, ret, i; + + DRM_DEBUG_DRIVER("\n"); + + ldev->panel = ltdc_get_panel(ddev); + if (!ldev->panel) + return -EPROBE_DEFER; + + rstc = of_reset_control_get(np, NULL); + + mutex_init(&ldev->err_lock); + + ldev->pixel_clk = devm_clk_get(dev, "lcd"); + if (IS_ERR(ldev->pixel_clk)) { + DRM_ERROR("Unable to get lcd clock\n"); + return -ENODEV; + } + + if (clk_prepare_enable(ldev->pixel_clk)) { + DRM_ERROR("Unable to prepare pixel clock\n"); + return -ENODEV; + } + + if (of_address_to_resource(np, 0, &res)) { + DRM_ERROR("Unable to get resource\n"); + return -ENODEV; + } + + ldev->regs = devm_ioremap_resource(dev, &res); + if (IS_ERR(ldev->regs)) { + DRM_ERROR("Unable to get ltdc registers\n"); + return PTR_ERR(ldev->regs); + } + + for (i = 0; i < MAX_IRQ; i++) { + irq = platform_get_irq(pdev, i); + if (irq < 0) + continue; + + ret = devm_request_threaded_irq(dev, irq, ltdc_irq, + ltdc_irq_thread, IRQF_ONESHOT, + dev_name(dev), ddev); + if (ret) { + DRM_ERROR("Failed to register LTDC interrupt\n"); + return ret; + } + } + + if (!IS_ERR(rstc)) + reset_control_deassert(rstc); + + /* Disable interrupts */ + reg_clear(ldev->regs, LTDC_IER, + IER_LIE | IER_RRIE | IER_FUIE | IER_TERRIE); + + ret = ltdc_get_caps(ddev); + if (ret) { + DRM_ERROR("hardware identifier (0x%08x) not supported!\n", + ldev->caps.hw_version); + return ret; + } + + DRM_INFO("ltdc hw version 0x%08x - ready\n", ldev->caps.hw_version); + + if (ldev->panel) { + encoder = ltdc_rgb_encoder_create(ddev); + if (!encoder) { + DRM_ERROR("Failed to create RGB encoder\n"); + ret = -EINVAL; + goto err; + } + + connector = ltdc_rgb_connector_create(ddev); + if (!connector) { + DRM_ERROR("Failed to create RGB connector\n"); + ret = -EINVAL; + goto err; + } + + ret = drm_mode_connector_attach_encoder(connector, encoder); + if (ret) { + DRM_ERROR("Failed to attach connector to encoder\n"); + goto err; + } + + drm_panel_attach(ldev->panel, connector); + } + + crtc = devm_kzalloc(dev, sizeof(*crtc), GFP_KERNEL); + if (!crtc) { + DRM_ERROR("Failed to allocate crtc\n"); + ret = -ENOMEM; + goto err; + } + + ret = ltdc_crtc_init(ddev, crtc); + if (ret) { + DRM_ERROR("Failed to init crtc\n"); + goto err; + } + + ret = drm_vblank_init(ddev, NB_CRTC); + if (ret) { + DRM_ERROR("Failed calling drm_vblank_init()\n"); + goto err; + } + + /* Allow usage of vblank without having to call drm_irq_install */ + ddev->irq_enabled = 1; + + return 0; +err: + if (ldev->panel) + drm_panel_detach(ldev->panel); + + clk_disable_unprepare(ldev->pixel_clk); + + return ret; +} + +void ltdc_unload(struct drm_device *ddev) +{ + struct ltdc_device *ldev = ddev->dev_private; + + DRM_DEBUG_DRIVER("\n"); + + drm_vblank_cleanup(ddev); + + if (ldev->panel) + drm_panel_detach(ldev->panel); + + clk_disable_unprepare(ldev->pixel_clk); +} + +MODULE_AUTHOR("Philippe Cornu <philippe.cornu@st.com>"); +MODULE_AUTHOR("Yannick Fertre <yannick.fertre@st.com>"); +MODULE_AUTHOR("Fabien Dessenne <fabien.dessenne@st.com>"); +MODULE_AUTHOR("Mickael Reulier <mickael.reulier@st.com>"); +MODULE_DESCRIPTION("STMicroelectronics ST DRM LTDC driver"); +MODULE_LICENSE("GPL v2"); diff --git a/drivers/gpu/drm/stm/ltdc.h b/drivers/gpu/drm/stm/ltdc.h new file mode 100644 index 000000000000..d7a9c736ac1e --- /dev/null +++ b/drivers/gpu/drm/stm/ltdc.h @@ -0,0 +1,40 @@ +/* + * Copyright (C) STMicroelectronics SA 2017 + * + * Authors: Philippe Cornu <philippe.cornu@st.com> + * Yannick Fertre <yannick.fertre@st.com> + * Fabien Dessenne <fabien.dessenne@st.com> + * Mickael Reulier <mickael.reulier@st.com> + * + * License terms: GNU General Public License (GPL), version 2 + */ + +#ifndef _LTDC_H_ +#define _LTDC_H_ + +struct ltdc_caps { + u32 hw_version; /* hardware version */ + u32 nb_layers; /* number of supported layers */ + u32 reg_ofs; /* register offset for applicable regs */ + u32 bus_width; /* bus width (32 or 64 bits) */ + const u32 *pix_fmt_hw; /* supported pixel formats */ +}; + +struct ltdc_device { + struct drm_fbdev_cma *fbdev; + void __iomem *regs; + struct clk *pixel_clk; /* lcd pixel clock */ + struct drm_panel *panel; + struct mutex err_lock; /* protecting error_status */ + struct ltdc_caps caps; + u32 clut[256]; /* color look up table */ + u32 error_status; + u32 irq_status; +}; + +int ltdc_crtc_enable_vblank(struct drm_device *dev, unsigned int pipe); +void ltdc_crtc_disable_vblank(struct drm_device *dev, unsigned int pipe); +int ltdc_load(struct drm_device *ddev); +void ltdc_unload(struct drm_device *ddev); + +#endif diff --git a/drivers/gpu/drm/tdfx/Makefile b/drivers/gpu/drm/tdfx/Makefile index 0379f294b32a..74bd4ae32348 100644 --- a/drivers/gpu/drm/tdfx/Makefile +++ b/drivers/gpu/drm/tdfx/Makefile @@ -2,7 +2,6 @@ # Makefile for the drm device driver. This driver provides support for the # Direct Rendering Infrastructure (DRI) in XFree86 4.1.0 and higher. -ccflags-y := -Iinclude/drm tdfx-y := tdfx_drv.o obj-$(CONFIG_DRM_TDFX) += tdfx.o diff --git a/drivers/gpu/drm/tegra/Kconfig b/drivers/gpu/drm/tegra/Kconfig index bbf5a4b7e0b6..2db29d67193d 100644 --- a/drivers/gpu/drm/tegra/Kconfig +++ b/drivers/gpu/drm/tegra/Kconfig @@ -7,6 +7,7 @@ config DRM_TEGRA select DRM_MIPI_DSI select DRM_PANEL select TEGRA_HOST1X + select IOMMU_IOVA if IOMMU_SUPPORT help Choose this option if you have an NVIDIA Tegra SoC. diff --git a/drivers/gpu/drm/tegra/Makefile b/drivers/gpu/drm/tegra/Makefile index 2c66a8db9da4..6af3a9ad6565 100644 --- a/drivers/gpu/drm/tegra/Makefile +++ b/drivers/gpu/drm/tegra/Makefile @@ -13,6 +13,8 @@ tegra-drm-y := \ sor.o \ dpaux.o \ gr2d.o \ - gr3d.o + gr3d.o \ + falcon.o \ + vic.o obj-$(CONFIG_DRM_TEGRA) += tegra-drm.o diff --git a/drivers/gpu/drm/tegra/drm.c b/drivers/gpu/drm/tegra/drm.c index dba4e090d3df..51c48a8e00ec 100644 --- a/drivers/gpu/drm/tegra/drm.c +++ b/drivers/gpu/drm/tegra/drm.c @@ -1,13 +1,15 @@ /* * Copyright (C) 2012 Avionic Design GmbH - * Copyright (C) 2012-2013 NVIDIA CORPORATION. All rights reserved. + * Copyright (C) 2012-2016 NVIDIA CORPORATION. All rights reserved. * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License version 2 as * published by the Free Software Foundation. */ +#include <linux/bitops.h> #include <linux/host1x.h> +#include <linux/idr.h> #include <linux/iommu.h> #include <drm/drm_atomic.h> @@ -23,8 +25,11 @@ #define DRIVER_MINOR 0 #define DRIVER_PATCHLEVEL 0 +#define CARVEOUT_SZ SZ_64M + struct tegra_drm_file { - struct list_head contexts; + struct idr contexts; + struct mutex lock; }; static void tegra_atomic_schedule(struct tegra_drm *tegra, @@ -126,8 +131,9 @@ static int tegra_drm_load(struct drm_device *drm, unsigned long flags) return -ENOMEM; if (iommu_present(&platform_bus_type)) { + u64 carveout_start, carveout_end, gem_start, gem_end; struct iommu_domain_geometry *geometry; - u64 start, end; + unsigned long order; tegra->domain = iommu_domain_alloc(&platform_bus_type); if (!tegra->domain) { @@ -136,12 +142,26 @@ static int tegra_drm_load(struct drm_device *drm, unsigned long flags) } geometry = &tegra->domain->geometry; - start = geometry->aperture_start; - end = geometry->aperture_end; - - DRM_DEBUG_DRIVER("IOMMU aperture initialized (%#llx-%#llx)\n", - start, end); - drm_mm_init(&tegra->mm, start, end - start + 1); + gem_start = geometry->aperture_start; + gem_end = geometry->aperture_end - CARVEOUT_SZ; + carveout_start = gem_end + 1; + carveout_end = geometry->aperture_end; + + order = __ffs(tegra->domain->pgsize_bitmap); + init_iova_domain(&tegra->carveout.domain, 1UL << order, + carveout_start >> order, + carveout_end >> order); + + tegra->carveout.shift = iova_shift(&tegra->carveout.domain); + tegra->carveout.limit = carveout_end >> tegra->carveout.shift; + + drm_mm_init(&tegra->mm, gem_start, gem_end - gem_start + 1); + mutex_init(&tegra->mm_lock); + + DRM_DEBUG("IOMMU apertures:\n"); + DRM_DEBUG(" GEM: %#llx-%#llx\n", gem_start, gem_end); + DRM_DEBUG(" Carveout: %#llx-%#llx\n", carveout_start, + carveout_end); } mutex_init(&tegra->clients_lock); @@ -161,6 +181,8 @@ static int tegra_drm_load(struct drm_device *drm, unsigned long flags) drm->mode_config.max_width = 4096; drm->mode_config.max_height = 4096; + drm->mode_config.allow_fb_modifiers = true; + drm->mode_config.funcs = &tegra_drm_mode_funcs; err = tegra_drm_fb_prepare(drm); @@ -208,6 +230,8 @@ config: if (tegra->domain) { iommu_domain_free(tegra->domain); drm_mm_takedown(&tegra->mm); + mutex_destroy(&tegra->mm_lock); + put_iova_domain(&tegra->carveout.domain); } free: kfree(tegra); @@ -232,6 +256,8 @@ static void tegra_drm_unload(struct drm_device *drm) if (tegra->domain) { iommu_domain_free(tegra->domain); drm_mm_takedown(&tegra->mm); + mutex_destroy(&tegra->mm_lock); + put_iova_domain(&tegra->carveout.domain); } kfree(tegra); @@ -245,7 +271,8 @@ static int tegra_drm_open(struct drm_device *drm, struct drm_file *filp) if (!fpriv) return -ENOMEM; - INIT_LIST_HEAD(&fpriv->contexts); + idr_init(&fpriv->contexts); + mutex_init(&fpriv->lock); filp->driver_priv = fpriv; return 0; @@ -424,21 +451,16 @@ fail: #ifdef CONFIG_DRM_TEGRA_STAGING -static struct tegra_drm_context *tegra_drm_get_context(__u64 context) -{ - return (struct tegra_drm_context *)(uintptr_t)context; -} - -static bool tegra_drm_file_owns_context(struct tegra_drm_file *file, - struct tegra_drm_context *context) +static struct tegra_drm_context * +tegra_drm_file_get_context(struct tegra_drm_file *file, u32 id) { - struct tegra_drm_context *ctx; + struct tegra_drm_context *context; - list_for_each_entry(ctx, &file->contexts, list) - if (ctx == context) - return true; + mutex_lock(&file->lock); + context = idr_find(&file->contexts, id); + mutex_unlock(&file->lock); - return false; + return context; } static int tegra_gem_create(struct drm_device *drm, void *data, @@ -519,6 +541,28 @@ static int tegra_syncpt_wait(struct drm_device *drm, void *data, &args->value); } +static int tegra_client_open(struct tegra_drm_file *fpriv, + struct tegra_drm_client *client, + struct tegra_drm_context *context) +{ + int err; + + err = client->ops->open_channel(client, context); + if (err < 0) + return err; + + err = idr_alloc(&fpriv->contexts, context, 0, 0, GFP_KERNEL); + if (err < 0) { + client->ops->close_channel(context); + return err; + } + + context->client = client; + context->id = err; + + return 0; +} + static int tegra_open_channel(struct drm_device *drm, void *data, struct drm_file *file) { @@ -533,19 +577,22 @@ static int tegra_open_channel(struct drm_device *drm, void *data, if (!context) return -ENOMEM; + mutex_lock(&fpriv->lock); + list_for_each_entry(client, &tegra->clients, list) if (client->base.class == args->client) { - err = client->ops->open_channel(client, context); - if (err) + err = tegra_client_open(fpriv, client, context); + if (err < 0) break; - list_add(&context->list, &fpriv->contexts); - args->context = (uintptr_t)context; - context->client = client; - return 0; + args->context = context->id; + break; } - kfree(context); + if (err < 0) + kfree(context); + + mutex_unlock(&fpriv->lock); return err; } @@ -555,16 +602,22 @@ static int tegra_close_channel(struct drm_device *drm, void *data, struct tegra_drm_file *fpriv = file->driver_priv; struct drm_tegra_close_channel *args = data; struct tegra_drm_context *context; + int err = 0; - context = tegra_drm_get_context(args->context); + mutex_lock(&fpriv->lock); - if (!tegra_drm_file_owns_context(fpriv, context)) - return -EINVAL; + context = tegra_drm_file_get_context(fpriv, args->context); + if (!context) { + err = -EINVAL; + goto unlock; + } - list_del(&context->list); + idr_remove(&fpriv->contexts, context->id); tegra_drm_context_free(context); - return 0; +unlock: + mutex_unlock(&fpriv->lock); + return err; } static int tegra_get_syncpt(struct drm_device *drm, void *data, @@ -574,19 +627,27 @@ static int tegra_get_syncpt(struct drm_device *drm, void *data, struct drm_tegra_get_syncpt *args = data; struct tegra_drm_context *context; struct host1x_syncpt *syncpt; + int err = 0; - context = tegra_drm_get_context(args->context); + mutex_lock(&fpriv->lock); - if (!tegra_drm_file_owns_context(fpriv, context)) - return -ENODEV; + context = tegra_drm_file_get_context(fpriv, args->context); + if (!context) { + err = -ENODEV; + goto unlock; + } - if (args->index >= context->client->base.num_syncpts) - return -EINVAL; + if (args->index >= context->client->base.num_syncpts) { + err = -EINVAL; + goto unlock; + } syncpt = context->client->base.syncpts[args->index]; args->id = host1x_syncpt_id(syncpt); - return 0; +unlock: + mutex_unlock(&fpriv->lock); + return err; } static int tegra_submit(struct drm_device *drm, void *data, @@ -595,13 +656,21 @@ static int tegra_submit(struct drm_device *drm, void *data, struct tegra_drm_file *fpriv = file->driver_priv; struct drm_tegra_submit *args = data; struct tegra_drm_context *context; + int err; - context = tegra_drm_get_context(args->context); + mutex_lock(&fpriv->lock); + + context = tegra_drm_file_get_context(fpriv, args->context); + if (!context) { + err = -ENODEV; + goto unlock; + } - if (!tegra_drm_file_owns_context(fpriv, context)) - return -ENODEV; + err = context->client->ops->submit(context, args, drm, file); - return context->client->ops->submit(context, args, drm, file); +unlock: + mutex_unlock(&fpriv->lock); + return err; } static int tegra_get_syncpt_base(struct drm_device *drm, void *data, @@ -612,24 +681,34 @@ static int tegra_get_syncpt_base(struct drm_device *drm, void *data, struct tegra_drm_context *context; struct host1x_syncpt_base *base; struct host1x_syncpt *syncpt; + int err = 0; - context = tegra_drm_get_context(args->context); + mutex_lock(&fpriv->lock); - if (!tegra_drm_file_owns_context(fpriv, context)) - return -ENODEV; + context = tegra_drm_file_get_context(fpriv, args->context); + if (!context) { + err = -ENODEV; + goto unlock; + } - if (args->syncpt >= context->client->base.num_syncpts) - return -EINVAL; + if (args->syncpt >= context->client->base.num_syncpts) { + err = -EINVAL; + goto unlock; + } syncpt = context->client->base.syncpts[args->syncpt]; base = host1x_syncpt_get_base(syncpt); - if (!base) - return -ENXIO; + if (!base) { + err = -ENXIO; + goto unlock; + } args->id = host1x_syncpt_base_id(base); - return 0; +unlock: + mutex_unlock(&fpriv->lock); + return err; } static int tegra_gem_set_tiling(struct drm_device *drm, void *data, @@ -804,14 +883,25 @@ static const struct file_operations tegra_drm_fops = { .llseek = noop_llseek, }; -static void tegra_drm_preclose(struct drm_device *drm, struct drm_file *file) +static int tegra_drm_context_cleanup(int id, void *p, void *data) +{ + struct tegra_drm_context *context = p; + + tegra_drm_context_free(context); + + return 0; +} + +static void tegra_drm_postclose(struct drm_device *drm, struct drm_file *file) { struct tegra_drm_file *fpriv = file->driver_priv; - struct tegra_drm_context *context, *tmp; - list_for_each_entry_safe(context, tmp, &fpriv->contexts, list) - tegra_drm_context_free(context); + mutex_lock(&fpriv->lock); + idr_for_each(&fpriv->contexts, tegra_drm_context_cleanup, NULL); + mutex_unlock(&fpriv->lock); + idr_destroy(&fpriv->contexts); + mutex_destroy(&fpriv->lock); kfree(fpriv); } @@ -844,7 +934,9 @@ static int tegra_debugfs_iova(struct seq_file *s, void *data) struct tegra_drm *tegra = drm->dev_private; struct drm_printer p = drm_seq_file_printer(s); + mutex_lock(&tegra->mm_lock); drm_mm_print(&tegra->mm, &p); + mutex_unlock(&tegra->mm_lock); return 0; } @@ -868,7 +960,7 @@ static struct drm_driver tegra_drm_driver = { .load = tegra_drm_load, .unload = tegra_drm_unload, .open = tegra_drm_open, - .preclose = tegra_drm_preclose, + .postclose = tegra_drm_postclose, .lastclose = tegra_drm_lastclose, #if defined(CONFIG_DEBUG_FS) @@ -919,6 +1011,84 @@ int tegra_drm_unregister_client(struct tegra_drm *tegra, return 0; } +void *tegra_drm_alloc(struct tegra_drm *tegra, size_t size, + dma_addr_t *dma) +{ + struct iova *alloc; + void *virt; + gfp_t gfp; + int err; + + if (tegra->domain) + size = iova_align(&tegra->carveout.domain, size); + else + size = PAGE_ALIGN(size); + + gfp = GFP_KERNEL | __GFP_ZERO; + if (!tegra->domain) { + /* + * Many units only support 32-bit addresses, even on 64-bit + * SoCs. If there is no IOMMU to translate into a 32-bit IO + * virtual address space, force allocations to be in the + * lower 32-bit range. + */ + gfp |= GFP_DMA; + } + + virt = (void *)__get_free_pages(gfp, get_order(size)); + if (!virt) + return ERR_PTR(-ENOMEM); + + if (!tegra->domain) { + /* + * If IOMMU is disabled, devices address physical memory + * directly. + */ + *dma = virt_to_phys(virt); + return virt; + } + + alloc = alloc_iova(&tegra->carveout.domain, + size >> tegra->carveout.shift, + tegra->carveout.limit, true); + if (!alloc) { + err = -EBUSY; + goto free_pages; + } + + *dma = iova_dma_addr(&tegra->carveout.domain, alloc); + err = iommu_map(tegra->domain, *dma, virt_to_phys(virt), + size, IOMMU_READ | IOMMU_WRITE); + if (err < 0) + goto free_iova; + + return virt; + +free_iova: + __free_iova(&tegra->carveout.domain, alloc); +free_pages: + free_pages((unsigned long)virt, get_order(size)); + + return ERR_PTR(err); +} + +void tegra_drm_free(struct tegra_drm *tegra, size_t size, void *virt, + dma_addr_t dma) +{ + if (tegra->domain) + size = iova_align(&tegra->carveout.domain, size); + else + size = PAGE_ALIGN(size); + + if (tegra->domain) { + iommu_unmap(tegra->domain, dma, size); + free_iova(&tegra->carveout.domain, + iova_pfn(&tegra->carveout.domain, dma)); + } + + free_pages((unsigned long)virt, get_order(size)); +} + static int host1x_drm_probe(struct host1x_device *dev) { struct drm_driver *driver = &tegra_drm_driver; @@ -1003,11 +1173,13 @@ static const struct of_device_id host1x_drm_subdevs[] = { { .compatible = "nvidia,tegra124-sor", }, { .compatible = "nvidia,tegra124-hdmi", }, { .compatible = "nvidia,tegra124-dsi", }, + { .compatible = "nvidia,tegra124-vic", }, { .compatible = "nvidia,tegra132-dsi", }, { .compatible = "nvidia,tegra210-dc", }, { .compatible = "nvidia,tegra210-dsi", }, { .compatible = "nvidia,tegra210-sor", }, { .compatible = "nvidia,tegra210-sor1", }, + { .compatible = "nvidia,tegra210-vic", }, { /* sentinel */ } }; @@ -1029,6 +1201,7 @@ static struct platform_driver * const drivers[] = { &tegra_sor_driver, &tegra_gr2d_driver, &tegra_gr3d_driver, + &tegra_vic_driver, }; static int __init host1x_drm_init(void) diff --git a/drivers/gpu/drm/tegra/drm.h b/drivers/gpu/drm/tegra/drm.h index 5747accb2271..85aa2e3d9d4e 100644 --- a/drivers/gpu/drm/tegra/drm.h +++ b/drivers/gpu/drm/tegra/drm.h @@ -12,6 +12,7 @@ #include <uapi/drm/tegra_drm.h> #include <linux/host1x.h> +#include <linux/iova.h> #include <linux/of_gpio.h> #include <drm/drmP.h> @@ -42,8 +43,15 @@ struct tegra_drm { struct drm_device *drm; struct iommu_domain *domain; + struct mutex mm_lock; struct drm_mm mm; + struct { + struct iova_domain domain; + unsigned long shift; + unsigned long limit; + } carveout; + struct mutex clients_lock; struct list_head clients; @@ -67,7 +75,7 @@ struct tegra_drm_client; struct tegra_drm_context { struct tegra_drm_client *client; struct host1x_channel *channel; - struct list_head list; + unsigned int id; }; struct tegra_drm_client_ops { @@ -105,6 +113,10 @@ int tegra_drm_unregister_client(struct tegra_drm *tegra, int tegra_drm_init(struct tegra_drm *tegra, struct drm_device *drm); int tegra_drm_exit(struct tegra_drm *tegra); +void *tegra_drm_alloc(struct tegra_drm *tegra, size_t size, dma_addr_t *iova); +void tegra_drm_free(struct tegra_drm *tegra, size_t size, void *virt, + dma_addr_t iova); + struct tegra_dc_soc_info; struct tegra_output; @@ -283,5 +295,6 @@ extern struct platform_driver tegra_dpaux_driver; extern struct platform_driver tegra_sor_driver; extern struct platform_driver tegra_gr2d_driver; extern struct platform_driver tegra_gr3d_driver; +extern struct platform_driver tegra_vic_driver; #endif /* HOST1X_DRM_H */ diff --git a/drivers/gpu/drm/tegra/falcon.c b/drivers/gpu/drm/tegra/falcon.c new file mode 100644 index 000000000000..f685e72949d1 --- /dev/null +++ b/drivers/gpu/drm/tegra/falcon.c @@ -0,0 +1,259 @@ +/* + * Copyright (c) 2015, NVIDIA Corporation. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ + +#include <linux/platform_device.h> +#include <linux/dma-mapping.h> +#include <linux/firmware.h> +#include <linux/pci_ids.h> +#include <linux/iopoll.h> + +#include "falcon.h" +#include "drm.h" + +enum falcon_memory { + FALCON_MEMORY_IMEM, + FALCON_MEMORY_DATA, +}; + +static void falcon_writel(struct falcon *falcon, u32 value, u32 offset) +{ + writel(value, falcon->regs + offset); +} + +int falcon_wait_idle(struct falcon *falcon) +{ + u32 value; + + return readl_poll_timeout(falcon->regs + FALCON_IDLESTATE, value, + (value == 0), 10, 100000); +} + +static int falcon_dma_wait_idle(struct falcon *falcon) +{ + u32 value; + + return readl_poll_timeout(falcon->regs + FALCON_DMATRFCMD, value, + (value & FALCON_DMATRFCMD_IDLE), 10, 100000); +} + +static int falcon_copy_chunk(struct falcon *falcon, + phys_addr_t base, + unsigned long offset, + enum falcon_memory target) +{ + u32 cmd = FALCON_DMATRFCMD_SIZE_256B; + + if (target == FALCON_MEMORY_IMEM) + cmd |= FALCON_DMATRFCMD_IMEM; + + falcon_writel(falcon, offset, FALCON_DMATRFMOFFS); + falcon_writel(falcon, base, FALCON_DMATRFFBOFFS); + falcon_writel(falcon, cmd, FALCON_DMATRFCMD); + + return falcon_dma_wait_idle(falcon); +} + +static void falcon_copy_firmware_image(struct falcon *falcon, + const struct firmware *firmware) +{ + u32 *firmware_vaddr = falcon->firmware.vaddr; + dma_addr_t daddr; + size_t i; + int err; + + /* copy the whole thing taking into account endianness */ + for (i = 0; i < firmware->size / sizeof(u32); i++) + firmware_vaddr[i] = le32_to_cpu(((u32 *)firmware->data)[i]); + + /* ensure that caches are flushed and falcon can see the firmware */ + daddr = dma_map_single(falcon->dev, firmware_vaddr, + falcon->firmware.size, DMA_TO_DEVICE); + err = dma_mapping_error(falcon->dev, daddr); + if (err) { + dev_err(falcon->dev, "failed to map firmware: %d\n", err); + return; + } + dma_sync_single_for_device(falcon->dev, daddr, + falcon->firmware.size, DMA_TO_DEVICE); + dma_unmap_single(falcon->dev, daddr, falcon->firmware.size, + DMA_TO_DEVICE); +} + +static int falcon_parse_firmware_image(struct falcon *falcon) +{ + struct falcon_fw_bin_header_v1 *bin = (void *)falcon->firmware.vaddr; + struct falcon_fw_os_header_v1 *os; + + /* endian problems would show up right here */ + if (bin->magic != PCI_VENDOR_ID_NVIDIA) { + dev_err(falcon->dev, "incorrect firmware magic\n"); + return -EINVAL; + } + + /* currently only version 1 is supported */ + if (bin->version != 1) { + dev_err(falcon->dev, "unsupported firmware version\n"); + return -EINVAL; + } + + /* check that the firmware size is consistent */ + if (bin->size > falcon->firmware.size) { + dev_err(falcon->dev, "firmware image size inconsistency\n"); + return -EINVAL; + } + + os = falcon->firmware.vaddr + bin->os_header_offset; + + falcon->firmware.bin_data.size = bin->os_size; + falcon->firmware.bin_data.offset = bin->os_data_offset; + falcon->firmware.code.offset = os->code_offset; + falcon->firmware.code.size = os->code_size; + falcon->firmware.data.offset = os->data_offset; + falcon->firmware.data.size = os->data_size; + + return 0; +} + +int falcon_read_firmware(struct falcon *falcon, const char *name) +{ + int err; + + /* request_firmware prints error if it fails */ + err = request_firmware(&falcon->firmware.firmware, name, falcon->dev); + if (err < 0) + return err; + + return 0; +} + +int falcon_load_firmware(struct falcon *falcon) +{ + const struct firmware *firmware = falcon->firmware.firmware; + int err; + + falcon->firmware.size = firmware->size; + + /* allocate iova space for the firmware */ + falcon->firmware.vaddr = falcon->ops->alloc(falcon, firmware->size, + &falcon->firmware.paddr); + if (!falcon->firmware.vaddr) { + dev_err(falcon->dev, "dma memory mapping failed\n"); + return -ENOMEM; + } + + /* copy firmware image into local area. this also ensures endianness */ + falcon_copy_firmware_image(falcon, firmware); + + /* parse the image data */ + err = falcon_parse_firmware_image(falcon); + if (err < 0) { + dev_err(falcon->dev, "failed to parse firmware image\n"); + goto err_setup_firmware_image; + } + + release_firmware(firmware); + falcon->firmware.firmware = NULL; + + return 0; + +err_setup_firmware_image: + falcon->ops->free(falcon, falcon->firmware.size, + falcon->firmware.paddr, falcon->firmware.vaddr); + + return err; +} + +int falcon_init(struct falcon *falcon) +{ + /* check mandatory ops */ + if (!falcon->ops || !falcon->ops->alloc || !falcon->ops->free) + return -EINVAL; + + falcon->firmware.vaddr = NULL; + + return 0; +} + +void falcon_exit(struct falcon *falcon) +{ + if (falcon->firmware.firmware) { + release_firmware(falcon->firmware.firmware); + falcon->firmware.firmware = NULL; + } + + if (falcon->firmware.vaddr) { + falcon->ops->free(falcon, falcon->firmware.size, + falcon->firmware.paddr, + falcon->firmware.vaddr); + falcon->firmware.vaddr = NULL; + } +} + +int falcon_boot(struct falcon *falcon) +{ + unsigned long offset; + int err; + + if (!falcon->firmware.vaddr) + return -EINVAL; + + falcon_writel(falcon, 0, FALCON_DMACTL); + + /* setup the address of the binary data so Falcon can access it later */ + falcon_writel(falcon, (falcon->firmware.paddr + + falcon->firmware.bin_data.offset) >> 8, + FALCON_DMATRFBASE); + + /* copy the data segment into Falcon internal memory */ + for (offset = 0; offset < falcon->firmware.data.size; offset += 256) + falcon_copy_chunk(falcon, + falcon->firmware.data.offset + offset, + offset, FALCON_MEMORY_DATA); + + /* copy the first code segment into Falcon internal memory */ + falcon_copy_chunk(falcon, falcon->firmware.code.offset, + 0, FALCON_MEMORY_IMEM); + + /* setup falcon interrupts */ + falcon_writel(falcon, FALCON_IRQMSET_EXT(0xff) | + FALCON_IRQMSET_SWGEN1 | + FALCON_IRQMSET_SWGEN0 | + FALCON_IRQMSET_EXTERR | + FALCON_IRQMSET_HALT | + FALCON_IRQMSET_WDTMR, + FALCON_IRQMSET); + falcon_writel(falcon, FALCON_IRQDEST_EXT(0xff) | + FALCON_IRQDEST_SWGEN1 | + FALCON_IRQDEST_SWGEN0 | + FALCON_IRQDEST_EXTERR | + FALCON_IRQDEST_HALT, + FALCON_IRQDEST); + + /* enable interface */ + falcon_writel(falcon, FALCON_ITFEN_MTHDEN | + FALCON_ITFEN_CTXEN, + FALCON_ITFEN); + + /* boot falcon */ + falcon_writel(falcon, 0x00000000, FALCON_BOOTVEC); + falcon_writel(falcon, FALCON_CPUCTL_STARTCPU, FALCON_CPUCTL); + + err = falcon_wait_idle(falcon); + if (err < 0) { + dev_err(falcon->dev, "Falcon boot failed due to timeout\n"); + return err; + } + + return 0; +} + +void falcon_execute_method(struct falcon *falcon, u32 method, u32 data) +{ + falcon_writel(falcon, method >> 2, FALCON_UCLASS_METHOD_OFFSET); + falcon_writel(falcon, data, FALCON_UCLASS_METHOD_DATA); +} diff --git a/drivers/gpu/drm/tegra/falcon.h b/drivers/gpu/drm/tegra/falcon.h new file mode 100644 index 000000000000..4504ed5a199e --- /dev/null +++ b/drivers/gpu/drm/tegra/falcon.h @@ -0,0 +1,127 @@ +/* + * Copyright (c) 2015, NVIDIA Corporation. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ + +#ifndef _FALCON_H_ +#define _FALCON_H_ + +#include <linux/types.h> + +#define FALCON_UCLASS_METHOD_OFFSET 0x00000040 + +#define FALCON_UCLASS_METHOD_DATA 0x00000044 + +#define FALCON_IRQMSET 0x00001010 +#define FALCON_IRQMSET_WDTMR (1 << 1) +#define FALCON_IRQMSET_HALT (1 << 4) +#define FALCON_IRQMSET_EXTERR (1 << 5) +#define FALCON_IRQMSET_SWGEN0 (1 << 6) +#define FALCON_IRQMSET_SWGEN1 (1 << 7) +#define FALCON_IRQMSET_EXT(v) (((v) & 0xff) << 8) + +#define FALCON_IRQDEST 0x0000101c +#define FALCON_IRQDEST_HALT (1 << 4) +#define FALCON_IRQDEST_EXTERR (1 << 5) +#define FALCON_IRQDEST_SWGEN0 (1 << 6) +#define FALCON_IRQDEST_SWGEN1 (1 << 7) +#define FALCON_IRQDEST_EXT(v) (((v) & 0xff) << 8) + +#define FALCON_ITFEN 0x00001048 +#define FALCON_ITFEN_CTXEN (1 << 0) +#define FALCON_ITFEN_MTHDEN (1 << 1) + +#define FALCON_IDLESTATE 0x0000104c + +#define FALCON_CPUCTL 0x00001100 +#define FALCON_CPUCTL_STARTCPU (1 << 1) + +#define FALCON_BOOTVEC 0x00001104 + +#define FALCON_DMACTL 0x0000110c +#define FALCON_DMACTL_DMEM_SCRUBBING (1 << 1) +#define FALCON_DMACTL_IMEM_SCRUBBING (1 << 2) + +#define FALCON_DMATRFBASE 0x00001110 + +#define FALCON_DMATRFMOFFS 0x00001114 + +#define FALCON_DMATRFCMD 0x00001118 +#define FALCON_DMATRFCMD_IDLE (1 << 1) +#define FALCON_DMATRFCMD_IMEM (1 << 4) +#define FALCON_DMATRFCMD_SIZE_256B (6 << 8) + +#define FALCON_DMATRFFBOFFS 0x0000111c + +struct falcon_fw_bin_header_v1 { + u32 magic; /* 0x10de */ + u32 version; /* version of bin format (1) */ + u32 size; /* entire image size including this header */ + u32 os_header_offset; + u32 os_data_offset; + u32 os_size; +}; + +struct falcon_fw_os_app_v1 { + u32 offset; + u32 size; +}; + +struct falcon_fw_os_header_v1 { + u32 code_offset; + u32 code_size; + u32 data_offset; + u32 data_size; +}; + +struct falcon; + +struct falcon_ops { + void *(*alloc)(struct falcon *falcon, size_t size, + dma_addr_t *paddr); + void (*free)(struct falcon *falcon, size_t size, + dma_addr_t paddr, void *vaddr); +}; + +struct falcon_firmware_section { + unsigned long offset; + size_t size; +}; + +struct falcon_firmware { + /* Firmware after it is read but not loaded */ + const struct firmware *firmware; + + /* Raw firmware data */ + dma_addr_t paddr; + void *vaddr; + size_t size; + + /* Parsed firmware information */ + struct falcon_firmware_section bin_data; + struct falcon_firmware_section data; + struct falcon_firmware_section code; +}; + +struct falcon { + /* Set by falcon client */ + struct device *dev; + void __iomem *regs; + const struct falcon_ops *ops; + void *data; + + struct falcon_firmware firmware; +}; + +int falcon_init(struct falcon *falcon); +void falcon_exit(struct falcon *falcon); +int falcon_read_firmware(struct falcon *falcon, const char *firmware_name); +int falcon_load_firmware(struct falcon *falcon); +int falcon_boot(struct falcon *falcon); +void falcon_execute_method(struct falcon *falcon, u32 method, u32 data); +int falcon_wait_idle(struct falcon *falcon); + +#endif /* _FALCON_H_ */ diff --git a/drivers/gpu/drm/tegra/fb.c b/drivers/gpu/drm/tegra/fb.c index c61d67d16ce3..25acb73ee728 100644 --- a/drivers/gpu/drm/tegra/fb.c +++ b/drivers/gpu/drm/tegra/fb.c @@ -52,9 +52,26 @@ int tegra_fb_get_tiling(struct drm_framebuffer *framebuffer, struct tegra_bo_tiling *tiling) { struct tegra_fb *fb = to_tegra_fb(framebuffer); - - /* TODO: handle YUV formats? */ - *tiling = fb->planes[0]->tiling; + uint64_t modifier = fb->base.modifier; + + switch (fourcc_mod_tegra_mod(modifier)) { + case NV_FORMAT_MOD_TEGRA_TILED: + tiling->mode = TEGRA_BO_TILING_MODE_TILED; + tiling->value = 0; + break; + + case NV_FORMAT_MOD_TEGRA_16BX2_BLOCK(0): + tiling->mode = TEGRA_BO_TILING_MODE_BLOCK; + tiling->value = fourcc_mod_tegra_param(modifier); + if (tiling->value > 5) + return -EINVAL; + break; + + default: + /* TODO: handle YUV formats? */ + *tiling = fb->planes[0]->tiling; + break; + } return 0; } diff --git a/drivers/gpu/drm/tegra/gem.c b/drivers/gpu/drm/tegra/gem.c index 8672f5d2f237..424569b53e57 100644 --- a/drivers/gpu/drm/tegra/gem.c +++ b/drivers/gpu/drm/tegra/gem.c @@ -128,12 +128,14 @@ static int tegra_bo_iommu_map(struct tegra_drm *tegra, struct tegra_bo *bo) if (!bo->mm) return -ENOMEM; + mutex_lock(&tegra->mm_lock); + err = drm_mm_insert_node_generic(&tegra->mm, bo->mm, bo->gem.size, PAGE_SIZE, 0, 0); if (err < 0) { dev_err(tegra->drm->dev, "out of I/O virtual memory: %zd\n", err); - goto free; + goto unlock; } bo->paddr = bo->mm->start; @@ -147,11 +149,14 @@ static int tegra_bo_iommu_map(struct tegra_drm *tegra, struct tegra_bo *bo) bo->size = err; + mutex_unlock(&tegra->mm_lock); + return 0; remove: drm_mm_remove_node(bo->mm); -free: +unlock: + mutex_unlock(&tegra->mm_lock); kfree(bo->mm); return err; } @@ -161,8 +166,11 @@ static int tegra_bo_iommu_unmap(struct tegra_drm *tegra, struct tegra_bo *bo) if (!bo->mm) return 0; + mutex_lock(&tegra->mm_lock); iommu_unmap(tegra->domain, bo->paddr, bo->size); drm_mm_remove_node(bo->mm); + mutex_unlock(&tegra->mm_lock); + kfree(bo->mm); return 0; diff --git a/drivers/gpu/drm/tegra/vic.c b/drivers/gpu/drm/tegra/vic.c new file mode 100644 index 000000000000..cd804e404a11 --- /dev/null +++ b/drivers/gpu/drm/tegra/vic.c @@ -0,0 +1,396 @@ +/* + * Copyright (c) 2015, NVIDIA Corporation. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ + +#include <linux/clk.h> +#include <linux/host1x.h> +#include <linux/iommu.h> +#include <linux/module.h> +#include <linux/of.h> +#include <linux/of_device.h> +#include <linux/of_platform.h> +#include <linux/platform_device.h> +#include <linux/pm_runtime.h> +#include <linux/reset.h> + +#include <soc/tegra/pmc.h> + +#include "drm.h" +#include "falcon.h" +#include "vic.h" + +struct vic_config { + const char *firmware; +}; + +struct vic { + struct falcon falcon; + bool booted; + + void __iomem *regs; + struct tegra_drm_client client; + struct host1x_channel *channel; + struct iommu_domain *domain; + struct device *dev; + struct clk *clk; + + /* Platform configuration */ + const struct vic_config *config; +}; + +static inline struct vic *to_vic(struct tegra_drm_client *client) +{ + return container_of(client, struct vic, client); +} + +static void vic_writel(struct vic *vic, u32 value, unsigned int offset) +{ + writel(value, vic->regs + offset); +} + +static int vic_runtime_resume(struct device *dev) +{ + struct vic *vic = dev_get_drvdata(dev); + + return clk_prepare_enable(vic->clk); +} + +static int vic_runtime_suspend(struct device *dev) +{ + struct vic *vic = dev_get_drvdata(dev); + + clk_disable_unprepare(vic->clk); + + vic->booted = false; + + return 0; +} + +static int vic_boot(struct vic *vic) +{ + u32 fce_ucode_size, fce_bin_data_offset; + void *hdr; + int err = 0; + + if (vic->booted) + return 0; + + /* setup clockgating registers */ + vic_writel(vic, CG_IDLE_CG_DLY_CNT(4) | + CG_IDLE_CG_EN | + CG_WAKEUP_DLY_CNT(4), + NV_PVIC_MISC_PRI_VIC_CG); + + err = falcon_boot(&vic->falcon); + if (err < 0) + return err; + + hdr = vic->falcon.firmware.vaddr; + fce_bin_data_offset = *(u32 *)(hdr + VIC_UCODE_FCE_DATA_OFFSET); + hdr = vic->falcon.firmware.vaddr + + *(u32 *)(hdr + VIC_UCODE_FCE_HEADER_OFFSET); + fce_ucode_size = *(u32 *)(hdr + FCE_UCODE_SIZE_OFFSET); + + falcon_execute_method(&vic->falcon, VIC_SET_APPLICATION_ID, 1); + falcon_execute_method(&vic->falcon, VIC_SET_FCE_UCODE_SIZE, + fce_ucode_size); + falcon_execute_method(&vic->falcon, VIC_SET_FCE_UCODE_OFFSET, + (vic->falcon.firmware.paddr + fce_bin_data_offset) + >> 8); + + err = falcon_wait_idle(&vic->falcon); + if (err < 0) { + dev_err(vic->dev, + "failed to set application ID and FCE base\n"); + return err; + } + + vic->booted = true; + + return 0; +} + +static void *vic_falcon_alloc(struct falcon *falcon, size_t size, + dma_addr_t *iova) +{ + struct tegra_drm *tegra = falcon->data; + + return tegra_drm_alloc(tegra, size, iova); +} + +static void vic_falcon_free(struct falcon *falcon, size_t size, + dma_addr_t iova, void *va) +{ + struct tegra_drm *tegra = falcon->data; + + return tegra_drm_free(tegra, size, va, iova); +} + +static const struct falcon_ops vic_falcon_ops = { + .alloc = vic_falcon_alloc, + .free = vic_falcon_free +}; + +static int vic_init(struct host1x_client *client) +{ + struct tegra_drm_client *drm = host1x_to_drm_client(client); + struct drm_device *dev = dev_get_drvdata(client->parent); + struct tegra_drm *tegra = dev->dev_private; + struct vic *vic = to_vic(drm); + int err; + + if (tegra->domain) { + err = iommu_attach_device(tegra->domain, vic->dev); + if (err < 0) { + dev_err(vic->dev, "failed to attach to domain: %d\n", + err); + return err; + } + + vic->domain = tegra->domain; + } + + if (!vic->falcon.data) { + vic->falcon.data = tegra; + err = falcon_load_firmware(&vic->falcon); + if (err < 0) + goto detach_device; + } + + vic->channel = host1x_channel_request(client->dev); + if (!vic->channel) { + err = -ENOMEM; + goto detach_device; + } + + client->syncpts[0] = host1x_syncpt_request(client->dev, 0); + if (!client->syncpts[0]) { + err = -ENOMEM; + goto free_channel; + } + + err = tegra_drm_register_client(tegra, drm); + if (err < 0) + goto free_syncpt; + + return 0; + +free_syncpt: + host1x_syncpt_free(client->syncpts[0]); +free_channel: + host1x_channel_free(vic->channel); +detach_device: + if (tegra->domain) + iommu_detach_device(tegra->domain, vic->dev); + + return err; +} + +static int vic_exit(struct host1x_client *client) +{ + struct tegra_drm_client *drm = host1x_to_drm_client(client); + struct drm_device *dev = dev_get_drvdata(client->parent); + struct tegra_drm *tegra = dev->dev_private; + struct vic *vic = to_vic(drm); + int err; + + err = tegra_drm_unregister_client(tegra, drm); + if (err < 0) + return err; + + host1x_syncpt_free(client->syncpts[0]); + host1x_channel_free(vic->channel); + + if (vic->domain) { + iommu_detach_device(vic->domain, vic->dev); + vic->domain = NULL; + } + + return 0; +} + +static const struct host1x_client_ops vic_client_ops = { + .init = vic_init, + .exit = vic_exit, +}; + +static int vic_open_channel(struct tegra_drm_client *client, + struct tegra_drm_context *context) +{ + struct vic *vic = to_vic(client); + int err; + + err = pm_runtime_get_sync(vic->dev); + if (err < 0) + return err; + + err = vic_boot(vic); + if (err < 0) { + pm_runtime_put(vic->dev); + return err; + } + + context->channel = host1x_channel_get(vic->channel); + if (!context->channel) { + pm_runtime_put(vic->dev); + return -ENOMEM; + } + + return 0; +} + +static void vic_close_channel(struct tegra_drm_context *context) +{ + struct vic *vic = to_vic(context->client); + + host1x_channel_put(context->channel); + + pm_runtime_put(vic->dev); +} + +static const struct tegra_drm_client_ops vic_ops = { + .open_channel = vic_open_channel, + .close_channel = vic_close_channel, + .submit = tegra_drm_submit, +}; + +static const struct vic_config vic_t124_config = { + .firmware = "nvidia/tegra124/vic03_ucode.bin", +}; + +static const struct vic_config vic_t210_config = { + .firmware = "nvidia/tegra210/vic04_ucode.bin", +}; + +static const struct of_device_id vic_match[] = { + { .compatible = "nvidia,tegra124-vic", .data = &vic_t124_config }, + { .compatible = "nvidia,tegra210-vic", .data = &vic_t210_config }, + { }, +}; + +static int vic_probe(struct platform_device *pdev) +{ + struct vic_config *vic_config = NULL; + struct device *dev = &pdev->dev; + struct host1x_syncpt **syncpts; + struct resource *regs; + const struct of_device_id *match; + struct vic *vic; + int err; + + match = of_match_device(vic_match, dev); + vic_config = (struct vic_config *)match->data; + + vic = devm_kzalloc(dev, sizeof(*vic), GFP_KERNEL); + if (!vic) + return -ENOMEM; + + syncpts = devm_kzalloc(dev, sizeof(*syncpts), GFP_KERNEL); + if (!syncpts) + return -ENOMEM; + + regs = platform_get_resource(pdev, IORESOURCE_MEM, 0); + if (!regs) { + dev_err(&pdev->dev, "failed to get registers\n"); + return -ENXIO; + } + + vic->regs = devm_ioremap_resource(dev, regs); + if (IS_ERR(vic->regs)) + return PTR_ERR(vic->regs); + + vic->clk = devm_clk_get(dev, NULL); + if (IS_ERR(vic->clk)) { + dev_err(&pdev->dev, "failed to get clock\n"); + return PTR_ERR(vic->clk); + } + + vic->falcon.dev = dev; + vic->falcon.regs = vic->regs; + vic->falcon.ops = &vic_falcon_ops; + + err = falcon_init(&vic->falcon); + if (err < 0) + return err; + + err = falcon_read_firmware(&vic->falcon, vic_config->firmware); + if (err < 0) + goto exit_falcon; + + platform_set_drvdata(pdev, vic); + + INIT_LIST_HEAD(&vic->client.base.list); + vic->client.base.ops = &vic_client_ops; + vic->client.base.dev = dev; + vic->client.base.class = HOST1X_CLASS_VIC; + vic->client.base.syncpts = syncpts; + vic->client.base.num_syncpts = 1; + vic->dev = dev; + vic->config = vic_config; + + INIT_LIST_HEAD(&vic->client.list); + vic->client.ops = &vic_ops; + + err = host1x_client_register(&vic->client.base); + if (err < 0) { + dev_err(dev, "failed to register host1x client: %d\n", err); + platform_set_drvdata(pdev, NULL); + goto exit_falcon; + } + + pm_runtime_enable(&pdev->dev); + if (!pm_runtime_enabled(&pdev->dev)) { + err = vic_runtime_resume(&pdev->dev); + if (err < 0) + goto unregister_client; + } + + return 0; + +unregister_client: + host1x_client_unregister(&vic->client.base); +exit_falcon: + falcon_exit(&vic->falcon); + + return err; +} + +static int vic_remove(struct platform_device *pdev) +{ + struct vic *vic = platform_get_drvdata(pdev); + int err; + + err = host1x_client_unregister(&vic->client.base); + if (err < 0) { + dev_err(&pdev->dev, "failed to unregister host1x client: %d\n", + err); + return err; + } + + if (pm_runtime_enabled(&pdev->dev)) + pm_runtime_disable(&pdev->dev); + else + vic_runtime_suspend(&pdev->dev); + + falcon_exit(&vic->falcon); + + return 0; +} + +static const struct dev_pm_ops vic_pm_ops = { + SET_RUNTIME_PM_OPS(vic_runtime_suspend, vic_runtime_resume, NULL) +}; + +struct platform_driver tegra_vic_driver = { + .driver = { + .name = "tegra-vic", + .of_match_table = vic_match, + .pm = &vic_pm_ops + }, + .probe = vic_probe, + .remove = vic_remove, +}; diff --git a/drivers/gpu/drm/tegra/vic.h b/drivers/gpu/drm/tegra/vic.h new file mode 100644 index 000000000000..21844817a7e1 --- /dev/null +++ b/drivers/gpu/drm/tegra/vic.h @@ -0,0 +1,31 @@ +/* + * Copyright (c) 2015, NVIDIA Corporation. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ + +#ifndef TEGRA_VIC_H +#define TEGRA_VIC_H + +/* VIC methods */ + +#define VIC_SET_APPLICATION_ID 0x00000200 +#define VIC_SET_FCE_UCODE_SIZE 0x0000071C +#define VIC_SET_FCE_UCODE_OFFSET 0x0000072C + +/* VIC registers */ + +#define NV_PVIC_MISC_PRI_VIC_CG 0x000016d0 +#define CG_IDLE_CG_DLY_CNT(val) ((val & 0x3f) << 0) +#define CG_IDLE_CG_EN (1 << 6) +#define CG_WAKEUP_DLY_CNT(val) ((val & 0xf) << 16) + +/* Firmware offsets */ + +#define VIC_UCODE_FCE_HEADER_OFFSET (6*4) +#define VIC_UCODE_FCE_DATA_OFFSET (7*4) +#define FCE_UCODE_SIZE_OFFSET (2*4) + +#endif /* TEGRA_VIC_H */ diff --git a/drivers/gpu/drm/tilcdc/Makefile b/drivers/gpu/drm/tilcdc/Makefile index 6f675175a9e5..55ebd516728f 100644 --- a/drivers/gpu/drm/tilcdc/Makefile +++ b/drivers/gpu/drm/tilcdc/Makefile @@ -1,4 +1,3 @@ -ccflags-y := -Iinclude/drm ifeq (, $(findstring -W,$(EXTRA_CFLAGS))) ccflags-y += -Werror endif diff --git a/drivers/gpu/drm/tilcdc/tilcdc_drv.c b/drivers/gpu/drm/tilcdc/tilcdc_drv.c index d7ae5be56d12..d67e18983a7d 100644 --- a/drivers/gpu/drm/tilcdc/tilcdc_drv.c +++ b/drivers/gpu/drm/tilcdc/tilcdc_drv.c @@ -22,6 +22,7 @@ #include <linux/suspend.h> #include <drm/drm_atomic.h> #include <drm/drm_atomic_helper.h> +#include <drm/drm_fb_helper.h> #include "tilcdc_drv.h" #include "tilcdc_regs.h" @@ -29,8 +30,6 @@ #include "tilcdc_panel.h" #include "tilcdc_external.h" -#include "drm_fb_helper.h" - static LIST_HEAD(module_list); static const u32 tilcdc_rev1_formats[] = { DRM_FORMAT_RGB565 }; diff --git a/drivers/gpu/drm/ttm/Makefile b/drivers/gpu/drm/ttm/Makefile index f92325800f8a..4d0c938ff4b2 100644 --- a/drivers/gpu/drm/ttm/Makefile +++ b/drivers/gpu/drm/ttm/Makefile @@ -1,7 +1,6 @@ # # Makefile for the drm device driver. This driver provides support for the -ccflags-y := -Iinclude/drm ttm-y := ttm_memory.o ttm_tt.o ttm_bo.o \ ttm_bo_util.o ttm_bo_vm.o ttm_module.o \ ttm_object.o ttm_lock.o ttm_execbuf_util.o ttm_page_alloc.o \ diff --git a/drivers/gpu/drm/ttm/ttm_bo.c b/drivers/gpu/drm/ttm/ttm_bo.c index e44626a2e698..a6d7fcb99c0b 100644 --- a/drivers/gpu/drm/ttm/ttm_bo.c +++ b/drivers/gpu/drm/ttm/ttm_bo.c @@ -1394,7 +1394,7 @@ EXPORT_SYMBOL(ttm_bo_evict_mm); int ttm_bo_init_mm(struct ttm_bo_device *bdev, unsigned type, unsigned long p_size) { - int ret = -EINVAL; + int ret; struct ttm_mem_type_manager *man; unsigned i; @@ -1412,7 +1412,6 @@ int ttm_bo_init_mm(struct ttm_bo_device *bdev, unsigned type, return ret; man->bdev = bdev; - ret = 0; if (type != TTM_PL_SYSTEM) { ret = (*man->func->init)(man, p_size); if (ret) diff --git a/drivers/gpu/drm/ttm/ttm_bo_vm.c b/drivers/gpu/drm/ttm/ttm_bo_vm.c index 9f53df95f35c..b442d12f2f7d 100644 --- a/drivers/gpu/drm/ttm/ttm_bo_vm.c +++ b/drivers/gpu/drm/ttm/ttm_bo_vm.c @@ -30,9 +30,9 @@ #define pr_fmt(fmt) "[TTM] " fmt -#include <ttm/ttm_module.h> -#include <ttm/ttm_bo_driver.h> -#include <ttm/ttm_placement.h> +#include <drm/ttm/ttm_module.h> +#include <drm/ttm/ttm_bo_driver.h> +#include <drm/ttm/ttm_placement.h> #include <drm/drm_vma_manager.h> #include <linux/mm.h> #include <linux/pfn_t.h> diff --git a/drivers/gpu/drm/ttm/ttm_page_alloc.c b/drivers/gpu/drm/ttm/ttm_page_alloc.c index a37de5db5731..eeddc1e48409 100644 --- a/drivers/gpu/drm/ttm/ttm_page_alloc.c +++ b/drivers/gpu/drm/ttm/ttm_page_alloc.c @@ -51,6 +51,9 @@ #if IS_ENABLED(CONFIG_AGP) #include <asm/agp.h> #endif +#ifdef CONFIG_X86 +#include <asm/set_memory.h> +#endif #define NUM_PAGES_TO_ALLOC (PAGE_SIZE/sizeof(struct page *)) #define SMALL_ALLOCATION 16 diff --git a/drivers/gpu/drm/ttm/ttm_page_alloc_dma.c b/drivers/gpu/drm/ttm/ttm_page_alloc_dma.c index cec4b4baa179..90ddbdca93bd 100644 --- a/drivers/gpu/drm/ttm/ttm_page_alloc_dma.c +++ b/drivers/gpu/drm/ttm/ttm_page_alloc_dma.c @@ -53,6 +53,9 @@ #if IS_ENABLED(CONFIG_AGP) #include <asm/agp.h> #endif +#ifdef CONFIG_X86 +#include <asm/set_memory.h> +#endif #define NUM_PAGES_TO_ALLOC (PAGE_SIZE/sizeof(struct page *)) #define SMALL_ALLOCATION 4 diff --git a/drivers/gpu/drm/ttm/ttm_tt.c b/drivers/gpu/drm/ttm/ttm_tt.c index aee3c00f836e..8ebc8d3560c3 100644 --- a/drivers/gpu/drm/ttm/ttm_tt.c +++ b/drivers/gpu/drm/ttm/ttm_tt.c @@ -39,25 +39,29 @@ #include <linux/slab.h> #include <linux/export.h> #include <drm/drm_cache.h> -#include <drm/drm_mem_util.h> #include <drm/ttm/ttm_module.h> #include <drm/ttm/ttm_bo_driver.h> #include <drm/ttm/ttm_placement.h> #include <drm/ttm/ttm_page_alloc.h> +#ifdef CONFIG_X86 +#include <asm/set_memory.h> +#endif /** * Allocates storage for pointers to the pages that back the ttm. */ static void ttm_tt_alloc_page_directory(struct ttm_tt *ttm) { - ttm->pages = drm_calloc_large(ttm->num_pages, sizeof(void*)); + ttm->pages = kvmalloc_array(ttm->num_pages, sizeof(void*), + GFP_KERNEL | __GFP_ZERO); } static void ttm_dma_tt_alloc_page_directory(struct ttm_dma_tt *ttm) { - ttm->ttm.pages = drm_calloc_large(ttm->ttm.num_pages, + ttm->ttm.pages = kvmalloc_array(ttm->ttm.num_pages, sizeof(*ttm->ttm.pages) + - sizeof(*ttm->dma_address)); + sizeof(*ttm->dma_address), + GFP_KERNEL | __GFP_ZERO); ttm->dma_address = (void *) (ttm->ttm.pages + ttm->ttm.num_pages); } @@ -205,7 +209,7 @@ EXPORT_SYMBOL(ttm_tt_init); void ttm_tt_fini(struct ttm_tt *ttm) { - drm_free_large(ttm->pages); + kvfree(ttm->pages); ttm->pages = NULL; } EXPORT_SYMBOL(ttm_tt_fini); @@ -240,7 +244,7 @@ void ttm_dma_tt_fini(struct ttm_dma_tt *ttm_dma) { struct ttm_tt *ttm = &ttm_dma->ttm; - drm_free_large(ttm->pages); + kvfree(ttm->pages); ttm->pages = NULL; ttm_dma->dma_address = NULL; } diff --git a/drivers/gpu/drm/udl/Makefile b/drivers/gpu/drm/udl/Makefile index 195bcac0b6c8..36f2e825102b 100644 --- a/drivers/gpu/drm/udl/Makefile +++ b/drivers/gpu/drm/udl/Makefile @@ -1,6 +1,3 @@ - -ccflags-y := -Iinclude/drm - udl-y := udl_drv.o udl_modeset.o udl_connector.o udl_encoder.o udl_main.o udl_fb.o udl_transfer.o udl_gem.o udl_dmabuf.o obj-$(CONFIG_DRM_UDL) := udl.o diff --git a/drivers/gpu/drm/udl/udl_dmabuf.c b/drivers/gpu/drm/udl/udl_dmabuf.c index ed0e636243b2..2e031a894813 100644 --- a/drivers/gpu/drm/udl/udl_dmabuf.c +++ b/drivers/gpu/drm/udl/udl_dmabuf.c @@ -228,7 +228,7 @@ static int udl_prime_create(struct drm_device *dev, return -ENOMEM; obj->sg = sg; - obj->pages = drm_malloc_ab(npages, sizeof(struct page *)); + obj->pages = kvmalloc_array(npages, sizeof(struct page *), GFP_KERNEL); if (obj->pages == NULL) { DRM_ERROR("obj pages is NULL %d\n", npages); return -ENOMEM; diff --git a/drivers/gpu/drm/udl/udl_fb.c b/drivers/gpu/drm/udl/udl_fb.c index d05abc69e305..4a6500362564 100644 --- a/drivers/gpu/drm/udl/udl_fb.c +++ b/drivers/gpu/drm/udl/udl_fb.c @@ -37,7 +37,7 @@ struct udl_fbdev { }; #define DL_ALIGN_UP(x, a) ALIGN(x, a) -#define DL_ALIGN_DOWN(x, a) ALIGN(x-(a-1), a) +#define DL_ALIGN_DOWN(x, a) ALIGN_DOWN(x, a) /** Read the red component (0..255) of a 32 bpp colour. */ #define DLO_RGB_GETRED(col) (uint8_t)((col) & 0xFF) diff --git a/drivers/gpu/drm/udl/udl_gem.c b/drivers/gpu/drm/udl/udl_gem.c index 775c50e4f02c..db9ceceba30e 100644 --- a/drivers/gpu/drm/udl/udl_gem.c +++ b/drivers/gpu/drm/udl/udl_gem.c @@ -146,7 +146,7 @@ int udl_gem_get_pages(struct udl_gem_object *obj) void udl_gem_put_pages(struct udl_gem_object *obj) { if (obj->base.import_attach) { - drm_free_large(obj->pages); + kvfree(obj->pages); obj->pages = NULL; return; } diff --git a/drivers/gpu/drm/vc4/Kconfig b/drivers/gpu/drm/vc4/Kconfig index 973b4203c0b2..b16aefe4a8d3 100644 --- a/drivers/gpu/drm/vc4/Kconfig +++ b/drivers/gpu/drm/vc4/Kconfig @@ -1,6 +1,6 @@ config DRM_VC4 tristate "Broadcom VC4 Graphics" - depends on ARCH_BCM2835 || COMPILE_TEST + depends on ARCH_BCM || ARCH_BCM2835 || COMPILE_TEST depends on DRM depends on SND && SND_SOC depends on COMMON_CLK diff --git a/drivers/gpu/drm/vc4/Makefile b/drivers/gpu/drm/vc4/Makefile index 61f45d122bd0..25bd5d30415d 100644 --- a/drivers/gpu/drm/vc4/Makefile +++ b/drivers/gpu/drm/vc4/Makefile @@ -1,5 +1,3 @@ -ccflags-y := -Iinclude/drm - # Please keep these build lists sorted! # core driver code @@ -9,6 +7,7 @@ vc4-y := \ vc4_drv.o \ vc4_dpi.o \ vc4_dsi.o \ + vc4_fence.o \ vc4_kms.o \ vc4_gem.o \ vc4_hdmi.o \ diff --git a/drivers/gpu/drm/vc4/vc4_bo.c b/drivers/gpu/drm/vc4/vc4_bo.c index af29432a6471..80b2f9e55c5c 100644 --- a/drivers/gpu/drm/vc4/vc4_bo.c +++ b/drivers/gpu/drm/vc4/vc4_bo.c @@ -19,6 +19,8 @@ * rendering can return quickly. */ +#include <linux/dma-buf.h> + #include "vc4_drv.h" #include "uapi/drm/vc4_drm.h" @@ -88,6 +90,10 @@ static void vc4_bo_destroy(struct vc4_bo *bo) vc4->bo_stats.num_allocated--; vc4->bo_stats.size_allocated -= obj->size; + + if (bo->resv == &bo->_resv) + reservation_object_fini(bo->resv); + drm_gem_cma_free_object(obj); } @@ -244,8 +250,12 @@ struct vc4_bo *vc4_bo_create(struct drm_device *dev, size_t unaligned_size, return ERR_PTR(-ENOMEM); } } + bo = to_vc4_bo(&cma_obj->base); - return to_vc4_bo(&cma_obj->base); + bo->resv = &bo->_resv; + reservation_object_init(bo->resv); + + return bo; } int vc4_dumb_create(struct drm_file *file_priv, @@ -369,6 +379,13 @@ static void vc4_bo_cache_time_timer(unsigned long data) schedule_work(&vc4->bo_cache.time_work); } +struct reservation_object *vc4_prime_res_obj(struct drm_gem_object *obj) +{ + struct vc4_bo *bo = to_vc4_bo(obj); + + return bo->resv; +} + struct dma_buf * vc4_prime_export(struct drm_device *dev, struct drm_gem_object *obj, int flags) { @@ -440,6 +457,24 @@ void *vc4_prime_vmap(struct drm_gem_object *obj) return drm_gem_cma_prime_vmap(obj); } +struct drm_gem_object * +vc4_prime_import_sg_table(struct drm_device *dev, + struct dma_buf_attachment *attach, + struct sg_table *sgt) +{ + struct drm_gem_object *obj; + struct vc4_bo *bo; + + obj = drm_gem_cma_prime_import_sg_table(dev, attach, sgt); + if (IS_ERR(obj)) + return obj; + + bo = to_vc4_bo(obj); + bo->resv = attach->dmabuf->resv; + + return obj; +} + int vc4_create_bo_ioctl(struct drm_device *dev, void *data, struct drm_file *file_priv) { diff --git a/drivers/gpu/drm/vc4/vc4_crtc.c b/drivers/gpu/drm/vc4/vc4_crtc.c index d86c8cce3182..0bfc4d88e4c2 100644 --- a/drivers/gpu/drm/vc4/vc4_crtc.c +++ b/drivers/gpu/drm/vc4/vc4_crtc.c @@ -32,13 +32,13 @@ * ones that set the clock. */ -#include "drm_atomic.h" -#include "drm_atomic_helper.h" -#include "drm_crtc_helper.h" -#include "linux/clk.h" -#include "drm_fb_cma_helper.h" -#include "linux/component.h" -#include "linux/of_device.h" +#include <drm/drm_atomic.h> +#include <drm/drm_atomic_helper.h> +#include <drm/drm_crtc_helper.h> +#include <linux/clk.h> +#include <drm/drm_fb_cma_helper.h> +#include <linux/component.h> +#include <linux/of_device.h> #include "vc4_drv.h" #include "vc4_regs.h" @@ -151,10 +151,10 @@ int vc4_crtc_debugfs_regs(struct seq_file *m, void *unused) } #endif -int vc4_crtc_get_scanoutpos(struct drm_device *dev, unsigned int crtc_id, - unsigned int flags, int *vpos, int *hpos, - ktime_t *stime, ktime_t *etime, - const struct drm_display_mode *mode) +bool vc4_crtc_get_scanoutpos(struct drm_device *dev, unsigned int crtc_id, + bool in_vblank_irq, int *vpos, int *hpos, + ktime_t *stime, ktime_t *etime, + const struct drm_display_mode *mode) { struct vc4_dev *vc4 = to_vc4_dev(dev); struct drm_crtc *crtc = drm_crtc_from_index(dev, crtc_id); @@ -162,7 +162,7 @@ int vc4_crtc_get_scanoutpos(struct drm_device *dev, unsigned int crtc_id, u32 val; int fifo_lines; int vblank_lines; - int ret = 0; + bool ret = false; /* preempt_disable_rt() should go right here in PREEMPT_RT patchset. */ @@ -198,7 +198,7 @@ int vc4_crtc_get_scanoutpos(struct drm_device *dev, unsigned int crtc_id, fifo_lines = vc4_crtc->cob_size / mode->crtc_hdisplay; if (fifo_lines > 0) - ret |= DRM_SCANOUTPOS_VALID; + ret = true; /* HVS more than fifo_lines into frame for compositing? */ if (*vpos > fifo_lines) { @@ -216,7 +216,6 @@ int vc4_crtc_get_scanoutpos(struct drm_device *dev, unsigned int crtc_id, */ *vpos -= fifo_lines + 1; - ret |= DRM_SCANOUTPOS_ACCURATE; return ret; } @@ -229,10 +228,9 @@ int vc4_crtc_get_scanoutpos(struct drm_device *dev, unsigned int crtc_id, * We can't get meaningful readings wrt. scanline position of the PV * and need to make things up in a approximative but consistent way. */ - ret |= DRM_SCANOUTPOS_IN_VBLANK; vblank_lines = mode->vtotal - mode->vdisplay; - if (flags & DRM_CALLED_FROM_VBLIRQ) { + if (in_vblank_irq) { /* * Assume the irq handler got called close to first * line of vblank, so PV has about a full vblank @@ -254,9 +252,10 @@ int vc4_crtc_get_scanoutpos(struct drm_device *dev, unsigned int crtc_id, * we are at the very beginning of vblank, as the hvs just * started refilling, and the stime and etime timestamps * truly correspond to start of vblank. + * + * Unfortunately there's no way to report this to upper levels + * and make it more useful. */ - if ((val & SCALER_DISPSTATX_FULL) != SCALER_DISPSTATX_FULL) - ret |= DRM_SCANOUTPOS_ACCURATE; } else { /* * No clue where we are inside vblank. Return a vpos of zero, @@ -270,19 +269,6 @@ int vc4_crtc_get_scanoutpos(struct drm_device *dev, unsigned int crtc_id, return ret; } -int vc4_crtc_get_vblank_timestamp(struct drm_device *dev, unsigned int crtc_id, - int *max_error, struct timeval *vblank_time, - unsigned flags) -{ - struct drm_crtc *crtc = drm_crtc_from_index(dev, crtc_id); - struct drm_crtc_state *state = crtc->state; - - /* Helper routine in DRM core does all the work: */ - return drm_calc_vbltimestamp_from_scanoutpos(dev, crtc_id, max_error, - vblank_time, flags, - &state->adjusted_mode); -} - static void vc4_crtc_destroy(struct drm_crtc *crtc) { drm_crtc_cleanup(crtc); diff --git a/drivers/gpu/drm/vc4/vc4_dpi.c b/drivers/gpu/drm/vc4/vc4_dpi.c index c6d703903fd9..39d68080873c 100644 --- a/drivers/gpu/drm/vc4/vc4_dpi.c +++ b/drivers/gpu/drm/vc4/vc4_dpi.c @@ -22,14 +22,14 @@ * ALT2 function. */ -#include "drm_atomic_helper.h" -#include "drm_crtc_helper.h" -#include "drm_edid.h" -#include "drm_panel.h" -#include "linux/clk.h" -#include "linux/component.h" -#include "linux/of_graph.h" -#include "linux/of_platform.h" +#include <drm/drm_atomic_helper.h> +#include <drm/drm_crtc_helper.h> +#include <drm/drm_edid.h> +#include <drm/drm_panel.h> +#include <linux/clk.h> +#include <linux/component.h> +#include <linux/of_graph.h> +#include <linux/of_platform.h> #include "vc4_drv.h" #include "vc4_regs.h" diff --git a/drivers/gpu/drm/vc4/vc4_drv.c b/drivers/gpu/drm/vc4/vc4_drv.c index 61e674baf3a6..136bb4213dc0 100644 --- a/drivers/gpu/drm/vc4/vc4_drv.c +++ b/drivers/gpu/drm/vc4/vc4_drv.c @@ -31,7 +31,7 @@ #include <linux/of_platform.h> #include <linux/platform_device.h> #include <linux/pm_runtime.h> -#include "drm_fb_cma_helper.h" +#include <drm/drm_fb_cma_helper.h> #include <drm/drm_fb_helper.h> #include "uapi/drm/vc4_drm.h" @@ -154,7 +154,7 @@ static struct drm_driver vc4_drm_driver = { .irq_uninstall = vc4_irq_uninstall, .get_scanout_position = vc4_crtc_get_scanoutpos, - .get_vblank_timestamp = vc4_crtc_get_vblank_timestamp, + .get_vblank_timestamp = drm_calc_vbltimestamp_from_scanoutpos, #if defined(CONFIG_DEBUG_FS) .debugfs_init = vc4_debugfs_init, @@ -168,8 +168,9 @@ static struct drm_driver vc4_drm_driver = { .prime_fd_to_handle = drm_gem_prime_fd_to_handle, .gem_prime_import = drm_gem_prime_import, .gem_prime_export = vc4_prime_export, + .gem_prime_res_obj = vc4_prime_res_obj, .gem_prime_get_sg_table = drm_gem_cma_prime_get_sg_table, - .gem_prime_import_sg_table = drm_gem_cma_prime_import_sg_table, + .gem_prime_import_sg_table = vc4_prime_import_sg_table, .gem_prime_vmap = vc4_prime_vmap, .gem_prime_vunmap = drm_gem_cma_prime_vunmap, .gem_prime_mmap = vc4_prime_mmap, @@ -334,6 +335,7 @@ static int vc4_platform_drm_remove(struct platform_device *pdev) static const struct of_device_id vc4_of_match[] = { { .compatible = "brcm,bcm2835-vc4", }, + { .compatible = "brcm,cygnus-vc4", }, {}, }; MODULE_DEVICE_TABLE(of, vc4_of_match); diff --git a/drivers/gpu/drm/vc4/vc4_drv.h b/drivers/gpu/drm/vc4/vc4_drv.h index dffce6293d87..a97556f7ccba 100644 --- a/drivers/gpu/drm/vc4/vc4_drv.h +++ b/drivers/gpu/drm/vc4/vc4_drv.h @@ -6,10 +6,10 @@ * published by the Free Software Foundation. */ -#include "drmP.h" -#include "drm_gem_cma_helper.h" - +#include <linux/reservation.h> +#include <drm/drmP.h> #include <drm/drm_encoder.h> +#include <drm/drm_gem_cma_helper.h> struct vc4_dev { struct drm_device *dev; @@ -56,6 +56,8 @@ struct vc4_dev { /* Protects bo_cache and the BO stats. */ struct mutex bo_lock; + uint64_t dma_fence_context; + /* Sequence number for the last job queued in bin_job_list. * Starts at 0 (no jobs emitted). */ @@ -95,12 +97,23 @@ struct vc4_dev { */ struct list_head seqno_cb_list; - /* The binner overflow memory that's currently set up in - * BPOA/BPOS registers. When overflow occurs and a new one is - * allocated, the previous one will be moved to - * vc4->current_exec's free list. + /* The memory used for storing binner tile alloc, tile state, + * and overflow memory allocations. This is freed when V3D + * powers down. + */ + struct vc4_bo *bin_bo; + + /* Size of blocks allocated within bin_bo. */ + uint32_t bin_alloc_size; + + /* Bitmask of the bin_alloc_size chunks in bin_bo that are + * used. */ - struct vc4_bo *overflow_mem; + uint32_t bin_alloc_used; + + /* Bitmask of the current bin_alloc used for overflow memory. */ + uint32_t bin_alloc_overflow; + struct work_struct overflow_mem_work; int power_refcount; @@ -150,6 +163,10 @@ struct vc4_bo { * DRM_IOCTL_VC4_CREATE_SHADER_BO. */ struct vc4_validated_shader_info *validated_shader; + + /* normally (resv == &_resv) except for imported bo's */ + struct reservation_object *resv; + struct reservation_object _resv; }; static inline struct vc4_bo * @@ -158,6 +175,19 @@ to_vc4_bo(struct drm_gem_object *bo) return (struct vc4_bo *)bo; } +struct vc4_fence { + struct dma_fence base; + struct drm_device *dev; + /* vc4 seqno for signaled() test */ + uint64_t seqno; +}; + +static inline struct vc4_fence * +to_vc4_fence(struct dma_fence *fence) +{ + return (struct vc4_fence *)fence; +} + struct vc4_seqno_cb { struct work_struct work; uint64_t seqno; @@ -168,6 +198,7 @@ struct vc4_v3d { struct vc4_dev *vc4; struct platform_device *pdev; void __iomem *regs; + struct clk *clk; }; struct vc4_hvs { @@ -230,6 +261,8 @@ struct vc4_exec_info { /* Latest write_seqno of any BO that binning depends on. */ uint64_t bin_dep_seqno; + struct dma_fence *fence; + /* Last current addresses the hardware was processing when the * hangcheck timer checked on us. */ @@ -293,8 +326,12 @@ struct vc4_exec_info { bool found_increment_semaphore_packet; bool found_flush; uint8_t bin_tiles_x, bin_tiles_y; - struct drm_gem_cma_object *tile_bo; + /* Physical address of the start of the tile alloc array + * (where each tile's binned CL will start) + */ uint32_t tile_alloc_offset; + /* Bitmask of which binner slots are freed when this job completes. */ + uint32_t bin_slots; /** * Computed addresses pointing into exec_bo where we start the @@ -436,7 +473,11 @@ int vc4_mmap_bo_ioctl(struct drm_device *dev, void *data, int vc4_get_hang_state_ioctl(struct drm_device *dev, void *data, struct drm_file *file_priv); int vc4_mmap(struct file *filp, struct vm_area_struct *vma); +struct reservation_object *vc4_prime_res_obj(struct drm_gem_object *obj); int vc4_prime_mmap(struct drm_gem_object *obj, struct vm_area_struct *vma); +struct drm_gem_object *vc4_prime_import_sg_table(struct drm_device *dev, + struct dma_buf_attachment *attach, + struct sg_table *sgt); void *vc4_prime_vmap(struct drm_gem_object *obj); void vc4_bo_cache_init(struct drm_device *dev); void vc4_bo_cache_destroy(struct drm_device *dev); @@ -446,13 +487,10 @@ int vc4_bo_stats_debugfs(struct seq_file *m, void *arg); extern struct platform_driver vc4_crtc_driver; bool vc4_event_pending(struct drm_crtc *crtc); int vc4_crtc_debugfs_regs(struct seq_file *m, void *arg); -int vc4_crtc_get_scanoutpos(struct drm_device *dev, unsigned int crtc_id, - unsigned int flags, int *vpos, int *hpos, - ktime_t *stime, ktime_t *etime, - const struct drm_display_mode *mode); -int vc4_crtc_get_vblank_timestamp(struct drm_device *dev, unsigned int crtc_id, - int *max_error, struct timeval *vblank_time, - unsigned flags); +bool vc4_crtc_get_scanoutpos(struct drm_device *dev, unsigned int crtc_id, + bool in_vblank_irq, int *vpos, int *hpos, + ktime_t *stime, ktime_t *etime, + const struct drm_display_mode *mode); /* vc4_debugfs.c */ int vc4_debugfs_init(struct drm_minor *minor); @@ -468,6 +506,9 @@ int vc4_dpi_debugfs_regs(struct seq_file *m, void *unused); extern struct platform_driver vc4_dsi_driver; int vc4_dsi_debugfs_regs(struct seq_file *m, void *unused); +/* vc4_fence.c */ +extern const struct dma_fence_ops vc4_fence_ops; + /* vc4_gem.c */ void vc4_gem_init(struct drm_device *dev); void vc4_gem_destroy(struct drm_device *dev); @@ -522,6 +563,7 @@ void vc4_plane_async_set_fb(struct drm_plane *plane, extern struct platform_driver vc4_v3d_driver; int vc4_v3d_debugfs_ident(struct seq_file *m, void *unused); int vc4_v3d_debugfs_regs(struct seq_file *m, void *unused); +int vc4_v3d_get_bin_slot(struct vc4_dev *vc4); /* vc4_validate.c */ int diff --git a/drivers/gpu/drm/vc4/vc4_dsi.c b/drivers/gpu/drm/vc4/vc4_dsi.c index 160f981d1cf4..0ef41df3915f 100644 --- a/drivers/gpu/drm/vc4/vc4_dsi.c +++ b/drivers/gpu/drm/vc4/vc4_dsi.c @@ -29,20 +29,20 @@ * hopefully present. */ -#include "drm_atomic_helper.h" -#include "drm_crtc_helper.h" -#include "drm_edid.h" -#include "drm_mipi_dsi.h" -#include "drm_panel.h" -#include "linux/clk.h" -#include "linux/clk-provider.h" -#include "linux/completion.h" -#include "linux/component.h" -#include "linux/dmaengine.h" -#include "linux/i2c.h" -#include "linux/of_address.h" -#include "linux/of_platform.h" -#include "linux/pm_runtime.h" +#include <drm/drm_atomic_helper.h> +#include <drm/drm_crtc_helper.h> +#include <drm/drm_edid.h> +#include <drm/drm_mipi_dsi.h> +#include <drm/drm_panel.h> +#include <linux/clk.h> +#include <linux/clk-provider.h> +#include <linux/completion.h> +#include <linux/component.h> +#include <linux/dmaengine.h> +#include <linux/i2c.h> +#include <linux/of_address.h> +#include <linux/of_platform.h> +#include <linux/pm_runtime.h> #include "vc4_drv.h" #include "vc4_regs.h" @@ -519,7 +519,8 @@ struct vc4_dsi { /* DSI channel for the panel we're connected to. */ u32 channel; u32 lanes; - enum mipi_dsi_pixel_format format; + u32 format; + u32 divider; u32 mode_flags; /* Input clock from CPRMAN to the digital PHY, for the DSI @@ -906,13 +907,67 @@ static void vc4_dsi_encoder_disable(struct drm_encoder *encoder) pm_runtime_put(dev); } +/* Extends the mode's blank intervals to handle BCM2835's integer-only + * DSI PLL divider. + * + * On 2835, PLLD is set to 2Ghz, and may not be changed by the display + * driver since most peripherals are hanging off of the PLLD_PER + * divider. PLLD_DSI1, which drives our DSI bit clock (and therefore + * the pixel clock), only has an integer divider off of DSI. + * + * To get our panel mode to refresh at the expected 60Hz, we need to + * extend the horizontal blank time. This means we drive a + * higher-than-expected clock rate to the panel, but that's what the + * firmware does too. + */ +static bool vc4_dsi_encoder_mode_fixup(struct drm_encoder *encoder, + const struct drm_display_mode *mode, + struct drm_display_mode *adjusted_mode) +{ + struct vc4_dsi_encoder *vc4_encoder = to_vc4_dsi_encoder(encoder); + struct vc4_dsi *dsi = vc4_encoder->dsi; + struct clk *phy_parent = clk_get_parent(dsi->pll_phy_clock); + unsigned long parent_rate = clk_get_rate(phy_parent); + unsigned long pixel_clock_hz = mode->clock * 1000; + unsigned long pll_clock = pixel_clock_hz * dsi->divider; + int divider; + + /* Find what divider gets us a faster clock than the requested + * pixel clock. + */ + for (divider = 1; divider < 8; divider++) { + if (parent_rate / divider < pll_clock) { + divider--; + break; + } + } + + /* Now that we've picked a PLL divider, calculate back to its + * pixel clock. + */ + pll_clock = parent_rate / divider; + pixel_clock_hz = pll_clock / dsi->divider; + + /* Round up the clk_set_rate() request slightly, since + * PLLD_DSI1 is an integer divider and its rate selection will + * never round up. + */ + adjusted_mode->clock = pixel_clock_hz / 1000 + 1; + + /* Given the new pixel clock, adjust HFP to keep vrefresh the same. */ + adjusted_mode->htotal = pixel_clock_hz / (mode->vrefresh * mode->vtotal); + adjusted_mode->hsync_end += adjusted_mode->htotal - mode->htotal; + adjusted_mode->hsync_start += adjusted_mode->htotal - mode->htotal; + + return true; +} + static void vc4_dsi_encoder_enable(struct drm_encoder *encoder) { - struct drm_display_mode *mode = &encoder->crtc->mode; + struct drm_display_mode *mode = &encoder->crtc->state->adjusted_mode; struct vc4_dsi_encoder *vc4_encoder = to_vc4_dsi_encoder(encoder); struct vc4_dsi *dsi = vc4_encoder->dsi; struct device *dev = &dsi->pdev->dev; - u32 format = 0, divider = 0; bool debug_dump_regs = false; unsigned long hs_clock; u32 ui_ns; @@ -940,26 +995,7 @@ static void vc4_dsi_encoder_enable(struct drm_encoder *encoder) vc4_dsi_dump_regs(dsi); } - switch (dsi->format) { - case MIPI_DSI_FMT_RGB888: - format = DSI_PFORMAT_RGB888; - divider = 24 / dsi->lanes; - break; - case MIPI_DSI_FMT_RGB666: - format = DSI_PFORMAT_RGB666; - divider = 24 / dsi->lanes; - break; - case MIPI_DSI_FMT_RGB666_PACKED: - format = DSI_PFORMAT_RGB666_PACKED; - divider = 18 / dsi->lanes; - break; - case MIPI_DSI_FMT_RGB565: - format = DSI_PFORMAT_RGB565; - divider = 16 / dsi->lanes; - break; - } - - phy_clock = pixel_clock_hz * divider; + phy_clock = pixel_clock_hz * dsi->divider; ret = clk_set_rate(dsi->pll_phy_clock, phy_clock); if (ret) { dev_err(&dsi->pdev->dev, @@ -1134,8 +1170,9 @@ static void vc4_dsi_encoder_enable(struct drm_encoder *encoder) if (dsi->mode_flags & MIPI_DSI_MODE_VIDEO) { DSI_PORT_WRITE(DISP0_CTRL, - VC4_SET_FIELD(divider, DSI_DISP0_PIX_CLK_DIV) | - VC4_SET_FIELD(format, DSI_DISP0_PFORMAT) | + VC4_SET_FIELD(dsi->divider, + DSI_DISP0_PIX_CLK_DIV) | + VC4_SET_FIELD(dsi->format, DSI_DISP0_PFORMAT) | VC4_SET_FIELD(DSI_DISP0_LP_STOP_PERFRAME, DSI_DISP0_LP_STOP_CTRL) | DSI_DISP0_ST_END | @@ -1347,9 +1384,31 @@ static int vc4_dsi_host_attach(struct mipi_dsi_host *host, dsi->lanes = device->lanes; dsi->channel = device->channel; - dsi->format = device->format; dsi->mode_flags = device->mode_flags; + switch (device->format) { + case MIPI_DSI_FMT_RGB888: + dsi->format = DSI_PFORMAT_RGB888; + dsi->divider = 24 / dsi->lanes; + break; + case MIPI_DSI_FMT_RGB666: + dsi->format = DSI_PFORMAT_RGB666; + dsi->divider = 24 / dsi->lanes; + break; + case MIPI_DSI_FMT_RGB666_PACKED: + dsi->format = DSI_PFORMAT_RGB666_PACKED; + dsi->divider = 18 / dsi->lanes; + break; + case MIPI_DSI_FMT_RGB565: + dsi->format = DSI_PFORMAT_RGB565; + dsi->divider = 16 / dsi->lanes; + break; + default: + dev_err(&dsi->pdev->dev, "Unknown DSI format: %d.\n", + dsi->format); + return 0; + } + if (!(dsi->mode_flags & MIPI_DSI_MODE_VIDEO)) { dev_err(&dsi->pdev->dev, "Only VIDEO mode panels supported currently.\n"); @@ -1397,6 +1456,7 @@ static const struct mipi_dsi_host_ops vc4_dsi_host_ops = { static const struct drm_encoder_helper_funcs vc4_dsi_encoder_helper_funcs = { .disable = vc4_dsi_encoder_disable, .enable = vc4_dsi_encoder_enable, + .mode_fixup = vc4_dsi_encoder_mode_fixup, }; static const struct of_device_id vc4_dsi_dt_match[] = { diff --git a/drivers/gpu/drm/vc4/vc4_fence.c b/drivers/gpu/drm/vc4/vc4_fence.c new file mode 100644 index 000000000000..dbf5a5a5d5f5 --- /dev/null +++ b/drivers/gpu/drm/vc4/vc4_fence.c @@ -0,0 +1,56 @@ +/* + * Copyright © 2017 Broadcom + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice (including the next + * paragraph) shall be included in all copies or substantial portions of the + * Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS + * IN THE SOFTWARE. + */ + +#include "vc4_drv.h" + +static const char *vc4_fence_get_driver_name(struct dma_fence *fence) +{ + return "vc4"; +} + +static const char *vc4_fence_get_timeline_name(struct dma_fence *fence) +{ + return "vc4-v3d"; +} + +static bool vc4_fence_enable_signaling(struct dma_fence *fence) +{ + return true; +} + +static bool vc4_fence_signaled(struct dma_fence *fence) +{ + struct vc4_fence *f = to_vc4_fence(fence); + struct vc4_dev *vc4 = to_vc4_dev(f->dev); + + return vc4->finished_seqno >= f->seqno; +} + +const struct dma_fence_ops vc4_fence_ops = { + .get_driver_name = vc4_fence_get_driver_name, + .get_timeline_name = vc4_fence_get_timeline_name, + .enable_signaling = vc4_fence_enable_signaling, + .signaled = vc4_fence_signaled, + .wait = dma_fence_default_wait, + .release = dma_fence_free, +}; diff --git a/drivers/gpu/drm/vc4/vc4_gem.c b/drivers/gpu/drm/vc4/vc4_gem.c index e9c381c42139..9dc7646d49ed 100644 --- a/drivers/gpu/drm/vc4/vc4_gem.c +++ b/drivers/gpu/drm/vc4/vc4_gem.c @@ -463,6 +463,8 @@ vc4_update_bo_seqnos(struct vc4_exec_info *exec, uint64_t seqno) for (i = 0; i < exec->bo_count; i++) { bo = to_vc4_bo(&exec->bo[i]->base); bo->seqno = seqno; + + reservation_object_add_shared_fence(bo->resv, exec->fence); } list_for_each_entry(bo, &exec->unref_list, unref_head) { @@ -472,7 +474,103 @@ vc4_update_bo_seqnos(struct vc4_exec_info *exec, uint64_t seqno) for (i = 0; i < exec->rcl_write_bo_count; i++) { bo = to_vc4_bo(&exec->rcl_write_bo[i]->base); bo->write_seqno = seqno; + + reservation_object_add_excl_fence(bo->resv, exec->fence); + } +} + +static void +vc4_unlock_bo_reservations(struct drm_device *dev, + struct vc4_exec_info *exec, + struct ww_acquire_ctx *acquire_ctx) +{ + int i; + + for (i = 0; i < exec->bo_count; i++) { + struct vc4_bo *bo = to_vc4_bo(&exec->bo[i]->base); + + ww_mutex_unlock(&bo->resv->lock); + } + + ww_acquire_fini(acquire_ctx); +} + +/* Takes the reservation lock on all the BOs being referenced, so that + * at queue submit time we can update the reservations. + * + * We don't lock the RCL the tile alloc/state BOs, or overflow memory + * (all of which are on exec->unref_list). They're entirely private + * to vc4, so we don't attach dma-buf fences to them. + */ +static int +vc4_lock_bo_reservations(struct drm_device *dev, + struct vc4_exec_info *exec, + struct ww_acquire_ctx *acquire_ctx) +{ + int contended_lock = -1; + int i, ret; + struct vc4_bo *bo; + + ww_acquire_init(acquire_ctx, &reservation_ww_class); + +retry: + if (contended_lock != -1) { + bo = to_vc4_bo(&exec->bo[contended_lock]->base); + ret = ww_mutex_lock_slow_interruptible(&bo->resv->lock, + acquire_ctx); + if (ret) { + ww_acquire_done(acquire_ctx); + return ret; + } + } + + for (i = 0; i < exec->bo_count; i++) { + if (i == contended_lock) + continue; + + bo = to_vc4_bo(&exec->bo[i]->base); + + ret = ww_mutex_lock_interruptible(&bo->resv->lock, acquire_ctx); + if (ret) { + int j; + + for (j = 0; j < i; j++) { + bo = to_vc4_bo(&exec->bo[j]->base); + ww_mutex_unlock(&bo->resv->lock); + } + + if (contended_lock != -1 && contended_lock >= i) { + bo = to_vc4_bo(&exec->bo[contended_lock]->base); + + ww_mutex_unlock(&bo->resv->lock); + } + + if (ret == -EDEADLK) { + contended_lock = i; + goto retry; + } + + ww_acquire_done(acquire_ctx); + return ret; + } } + + ww_acquire_done(acquire_ctx); + + /* Reserve space for our shared (read-only) fence references, + * before we commit the CL to the hardware. + */ + for (i = 0; i < exec->bo_count; i++) { + bo = to_vc4_bo(&exec->bo[i]->base); + + ret = reservation_object_reserve_shared(bo->resv); + if (ret) { + vc4_unlock_bo_reservations(dev, exec, acquire_ctx); + return ret; + } + } + + return 0; } /* Queues a struct vc4_exec_info for execution. If no job is @@ -484,19 +582,34 @@ vc4_update_bo_seqnos(struct vc4_exec_info *exec, uint64_t seqno) * then bump the end address. That's a change for a later date, * though. */ -static void -vc4_queue_submit(struct drm_device *dev, struct vc4_exec_info *exec) +static int +vc4_queue_submit(struct drm_device *dev, struct vc4_exec_info *exec, + struct ww_acquire_ctx *acquire_ctx) { struct vc4_dev *vc4 = to_vc4_dev(dev); uint64_t seqno; unsigned long irqflags; + struct vc4_fence *fence; + + fence = kzalloc(sizeof(*fence), GFP_KERNEL); + if (!fence) + return -ENOMEM; + fence->dev = dev; spin_lock_irqsave(&vc4->job_lock, irqflags); seqno = ++vc4->emit_seqno; exec->seqno = seqno; + + dma_fence_init(&fence->base, &vc4_fence_ops, &vc4->job_lock, + vc4->dma_fence_context, exec->seqno); + fence->seqno = exec->seqno; + exec->fence = &fence->base; + vc4_update_bo_seqnos(exec, seqno); + vc4_unlock_bo_reservations(dev, exec, acquire_ctx); + list_add_tail(&exec->head, &vc4->bin_job_list); /* If no job was executing, kick ours off. Otherwise, it'll @@ -509,6 +622,8 @@ vc4_queue_submit(struct drm_device *dev, struct vc4_exec_info *exec) } spin_unlock_irqrestore(&vc4->job_lock, irqflags); + + return 0; } /** @@ -545,14 +660,15 @@ vc4_cl_lookup_bos(struct drm_device *dev, return -EINVAL; } - exec->bo = drm_calloc_large(exec->bo_count, - sizeof(struct drm_gem_cma_object *)); + exec->bo = kvmalloc_array(exec->bo_count, + sizeof(struct drm_gem_cma_object *), + GFP_KERNEL | __GFP_ZERO); if (!exec->bo) { DRM_ERROR("Failed to allocate validated BO pointers\n"); return -ENOMEM; } - handles = drm_malloc_ab(exec->bo_count, sizeof(uint32_t)); + handles = kvmalloc_array(exec->bo_count, sizeof(uint32_t), GFP_KERNEL); if (!handles) { ret = -ENOMEM; DRM_ERROR("Failed to allocate incoming GEM handles\n"); @@ -584,7 +700,7 @@ vc4_cl_lookup_bos(struct drm_device *dev, spin_unlock(&file_priv->table_lock); fail: - drm_free_large(handles); + kvfree(handles); return ret; } @@ -622,7 +738,7 @@ vc4_get_bcl(struct drm_device *dev, struct vc4_exec_info *exec) * read the contents back for validation, and I think the * bo->vaddr is uncached access. */ - temp = drm_malloc_ab(temp_size, 1); + temp = kvmalloc_array(temp_size, 1, GFP_KERNEL); if (!temp) { DRM_ERROR("Failed to allocate storage for copying " "in bin/render CLs.\n"); @@ -697,7 +813,7 @@ vc4_get_bcl(struct drm_device *dev, struct vc4_exec_info *exec) ret = vc4_wait_for_seqno(dev, exec->bin_dep_seqno, ~0ull, true); fail: - drm_free_large(temp); + kvfree(temp); return ret; } @@ -705,12 +821,19 @@ static void vc4_complete_exec(struct drm_device *dev, struct vc4_exec_info *exec) { struct vc4_dev *vc4 = to_vc4_dev(dev); + unsigned long irqflags; unsigned i; + /* If we got force-completed because of GPU reset rather than + * through our IRQ handler, signal the fence now. + */ + if (exec->fence) + dma_fence_signal(exec->fence); + if (exec->bo) { for (i = 0; i < exec->bo_count; i++) drm_gem_object_unreference_unlocked(&exec->bo[i]->base); - drm_free_large(exec->bo); + kvfree(exec->bo); } while (!list_empty(&exec->unref_list)) { @@ -720,6 +843,11 @@ vc4_complete_exec(struct drm_device *dev, struct vc4_exec_info *exec) drm_gem_object_unreference_unlocked(&bo->base.base); } + /* Free up the allocation of any bin slots we used. */ + spin_lock_irqsave(&vc4->job_lock, irqflags); + vc4->bin_alloc_used &= ~exec->bin_slots; + spin_unlock_irqrestore(&vc4->job_lock, irqflags); + mutex_lock(&vc4->power_lock); if (--vc4->power_refcount == 0) { pm_runtime_mark_last_busy(&vc4->v3d->pdev->dev); @@ -874,6 +1002,7 @@ vc4_submit_cl_ioctl(struct drm_device *dev, void *data, struct vc4_dev *vc4 = to_vc4_dev(dev); struct drm_vc4_submit_cl *args = data; struct vc4_exec_info *exec; + struct ww_acquire_ctx acquire_ctx; int ret = 0; if ((args->flags & ~VC4_SUBMIT_CL_USE_CLEAR_COLOR) != 0) { @@ -888,13 +1017,16 @@ vc4_submit_cl_ioctl(struct drm_device *dev, void *data, } mutex_lock(&vc4->power_lock); - if (vc4->power_refcount++ == 0) + if (vc4->power_refcount++ == 0) { ret = pm_runtime_get_sync(&vc4->v3d->pdev->dev); - mutex_unlock(&vc4->power_lock); - if (ret < 0) { - kfree(exec); - return ret; + if (ret < 0) { + mutex_unlock(&vc4->power_lock); + vc4->power_refcount--; + kfree(exec); + return ret; + } } + mutex_unlock(&vc4->power_lock); exec->args = args; INIT_LIST_HEAD(&exec->unref_list); @@ -916,12 +1048,18 @@ vc4_submit_cl_ioctl(struct drm_device *dev, void *data, if (ret) goto fail; + ret = vc4_lock_bo_reservations(dev, exec, &acquire_ctx); + if (ret) + goto fail; + /* Clear this out of the struct we'll be putting in the queue, * since it's part of our stack. */ exec->args = NULL; - vc4_queue_submit(dev, exec); + ret = vc4_queue_submit(dev, exec, &acquire_ctx); + if (ret) + goto fail; /* Return the seqno for our job. */ args->seqno = vc4->emit_seqno; @@ -939,6 +1077,8 @@ vc4_gem_init(struct drm_device *dev) { struct vc4_dev *vc4 = to_vc4_dev(dev); + vc4->dma_fence_context = dma_fence_context_alloc(1); + INIT_LIST_HEAD(&vc4->bin_job_list); INIT_LIST_HEAD(&vc4->render_job_list); INIT_LIST_HEAD(&vc4->job_done_list); @@ -968,9 +1108,9 @@ vc4_gem_destroy(struct drm_device *dev) /* V3D should already have disabled its interrupt and cleared * the overflow allocation registers. Now free the object. */ - if (vc4->overflow_mem) { - drm_gem_object_unreference_unlocked(&vc4->overflow_mem->base.base); - vc4->overflow_mem = NULL; + if (vc4->bin_bo) { + drm_gem_object_put_unlocked(&vc4->bin_bo->base.base); + vc4->bin_bo = NULL; } if (vc4->hang_state) diff --git a/drivers/gpu/drm/vc4/vc4_hdmi.c b/drivers/gpu/drm/vc4/vc4_hdmi.c index e9cbe269710b..ed63d4e85762 100644 --- a/drivers/gpu/drm/vc4/vc4_hdmi.c +++ b/drivers/gpu/drm/vc4/vc4_hdmi.c @@ -42,20 +42,21 @@ * encoder block has CEC support. */ -#include "drm_atomic_helper.h" -#include "drm_crtc_helper.h" -#include "drm_edid.h" -#include "linux/clk.h" -#include "linux/component.h" -#include "linux/i2c.h" -#include "linux/of_address.h" -#include "linux/of_gpio.h" -#include "linux/of_platform.h" -#include "linux/rational.h" -#include "sound/dmaengine_pcm.h" -#include "sound/pcm_drm_eld.h" -#include "sound/pcm_params.h" -#include "sound/soc.h" +#include <drm/drm_atomic_helper.h> +#include <drm/drm_crtc_helper.h> +#include <drm/drm_edid.h> +#include <linux/clk.h> +#include <linux/component.h> +#include <linux/i2c.h> +#include <linux/of_address.h> +#include <linux/of_gpio.h> +#include <linux/of_platform.h> +#include <linux/pm_runtime.h> +#include <linux/rational.h> +#include <sound/dmaengine_pcm.h> +#include <sound/pcm_drm_eld.h> +#include <sound/pcm_params.h> +#include <sound/soc.h> #include "vc4_drv.h" #include "vc4_regs.h" @@ -449,13 +450,38 @@ static void vc4_hdmi_set_infoframes(struct drm_encoder *encoder) vc4_hdmi_set_spd_infoframe(encoder); } -static void vc4_hdmi_encoder_mode_set(struct drm_encoder *encoder, - struct drm_display_mode *unadjusted_mode, - struct drm_display_mode *mode) +static void vc4_hdmi_encoder_disable(struct drm_encoder *encoder) +{ + struct drm_device *dev = encoder->dev; + struct vc4_dev *vc4 = to_vc4_dev(dev); + struct vc4_hdmi *hdmi = vc4->hdmi; + int ret; + + HDMI_WRITE(VC4_HDMI_RAM_PACKET_CONFIG, 0); + + HDMI_WRITE(VC4_HDMI_TX_PHY_RESET_CTL, 0xf << 16); + HD_WRITE(VC4_HD_VID_CTL, + HD_READ(VC4_HD_VID_CTL) & ~VC4_HD_VID_CTL_ENABLE); + + HD_WRITE(VC4_HD_M_CTL, VC4_HD_M_SW_RST); + udelay(1); + HD_WRITE(VC4_HD_M_CTL, 0); + + clk_disable_unprepare(hdmi->hsm_clock); + clk_disable_unprepare(hdmi->pixel_clock); + + ret = pm_runtime_put(&hdmi->pdev->dev); + if (ret < 0) + DRM_ERROR("Failed to release power domain: %d\n", ret); +} + +static void vc4_hdmi_encoder_enable(struct drm_encoder *encoder) { + struct drm_display_mode *mode = &encoder->crtc->state->adjusted_mode; struct vc4_hdmi_encoder *vc4_encoder = to_vc4_hdmi_encoder(encoder); struct drm_device *dev = encoder->dev; struct vc4_dev *vc4 = to_vc4_dev(dev); + struct vc4_hdmi *hdmi = vc4->hdmi; bool debug_dump_regs = false; bool hsync_pos = mode->flags & DRM_MODE_FLAG_PHSYNC; bool vsync_pos = mode->flags & DRM_MODE_FLAG_PVSYNC; @@ -475,6 +501,64 @@ static void vc4_hdmi_encoder_mode_set(struct drm_encoder *encoder, interlaced, VC4_HDMI_VERTB_VBP)); u32 csc_ctl; + int ret; + + ret = pm_runtime_get_sync(&hdmi->pdev->dev); + if (ret < 0) { + DRM_ERROR("Failed to retain power domain: %d\n", ret); + return; + } + + /* This is the rate that is set by the firmware. The number + * needs to be a bit higher than the pixel clock rate + * (generally 148.5Mhz). + */ + ret = clk_set_rate(hdmi->hsm_clock, 163682864); + if (ret) { + DRM_ERROR("Failed to set HSM clock rate: %d\n", ret); + return; + } + + ret = clk_set_rate(hdmi->pixel_clock, + mode->clock * 1000 * + ((mode->flags & DRM_MODE_FLAG_DBLCLK) ? 2 : 1)); + if (ret) { + DRM_ERROR("Failed to set pixel clock rate: %d\n", ret); + return; + } + + ret = clk_prepare_enable(hdmi->pixel_clock); + if (ret) { + DRM_ERROR("Failed to turn on pixel clock: %d\n", ret); + return; + } + + ret = clk_prepare_enable(hdmi->hsm_clock); + if (ret) { + DRM_ERROR("Failed to turn on HDMI state machine clock: %d\n", + ret); + clk_disable_unprepare(hdmi->pixel_clock); + return; + } + + HD_WRITE(VC4_HD_M_CTL, VC4_HD_M_SW_RST); + udelay(1); + HD_WRITE(VC4_HD_M_CTL, 0); + + HD_WRITE(VC4_HD_M_CTL, VC4_HD_M_ENABLE); + + HDMI_WRITE(VC4_HDMI_SW_RESET_CONTROL, + VC4_HDMI_SW_RESET_HDMI | + VC4_HDMI_SW_RESET_FORMAT_DETECT); + + HDMI_WRITE(VC4_HDMI_SW_RESET_CONTROL, 0); + + /* PHY should be in reset, like + * vc4_hdmi_encoder_disable() does. + */ + HDMI_WRITE(VC4_HDMI_TX_PHY_RESET_CTL, 0xf << 16); + + HDMI_WRITE(VC4_HDMI_TX_PHY_RESET_CTL, 0); if (debug_dump_regs) { DRM_INFO("HDMI regs before:\n"); @@ -483,9 +567,6 @@ static void vc4_hdmi_encoder_mode_set(struct drm_encoder *encoder, HD_WRITE(VC4_HD_VID_CTL, 0); - clk_set_rate(vc4->hdmi->pixel_clock, mode->clock * 1000 * - ((mode->flags & DRM_MODE_FLAG_DBLCLK) ? 2 : 1)); - HDMI_WRITE(VC4_HDMI_SCHEDULER_CONTROL, HDMI_READ(VC4_HDMI_SCHEDULER_CONTROL) | VC4_HDMI_SCHEDULER_CONTROL_MANUAL_FORMAT | @@ -559,28 +640,6 @@ static void vc4_hdmi_encoder_mode_set(struct drm_encoder *encoder, DRM_INFO("HDMI regs after:\n"); vc4_hdmi_dump_regs(dev); } -} - -static void vc4_hdmi_encoder_disable(struct drm_encoder *encoder) -{ - struct drm_device *dev = encoder->dev; - struct vc4_dev *vc4 = to_vc4_dev(dev); - - HDMI_WRITE(VC4_HDMI_RAM_PACKET_CONFIG, 0); - - HDMI_WRITE(VC4_HDMI_TX_PHY_RESET_CTL, 0xf << 16); - HD_WRITE(VC4_HD_VID_CTL, - HD_READ(VC4_HD_VID_CTL) & ~VC4_HD_VID_CTL_ENABLE); -} - -static void vc4_hdmi_encoder_enable(struct drm_encoder *encoder) -{ - struct vc4_hdmi_encoder *vc4_encoder = to_vc4_hdmi_encoder(encoder); - struct drm_device *dev = encoder->dev; - struct vc4_dev *vc4 = to_vc4_dev(dev); - int ret; - - HDMI_WRITE(VC4_HDMI_TX_PHY_RESET_CTL, 0); HD_WRITE(VC4_HD_VID_CTL, HD_READ(VC4_HD_VID_CTL) | @@ -646,7 +705,6 @@ static void vc4_hdmi_encoder_enable(struct drm_encoder *encoder) } static const struct drm_encoder_helper_funcs vc4_hdmi_encoder_helper_funcs = { - .mode_set = vc4_hdmi_encoder_mode_set, .disable = vc4_hdmi_encoder_disable, .enable = vc4_hdmi_encoder_enable, }; @@ -1147,33 +1205,6 @@ static int vc4_hdmi_bind(struct device *dev, struct device *master, void *data) return -EPROBE_DEFER; } - /* Enable the clocks at startup. We can't quite recover from - * turning off the pixel clock during disable/enables yet, so - * it's always running. - */ - ret = clk_prepare_enable(hdmi->pixel_clock); - if (ret) { - DRM_ERROR("Failed to turn on pixel clock: %d\n", ret); - goto err_put_i2c; - } - - /* This is the rate that is set by the firmware. The number - * needs to be a bit higher than the pixel clock rate - * (generally 148.5Mhz). - */ - ret = clk_set_rate(hdmi->hsm_clock, 163682864); - if (ret) { - DRM_ERROR("Failed to set HSM clock rate: %d\n", ret); - goto err_unprepare_pix; - } - - ret = clk_prepare_enable(hdmi->hsm_clock); - if (ret) { - DRM_ERROR("Failed to turn on HDMI state machine clock: %d\n", - ret); - goto err_unprepare_pix; - } - /* Only use the GPIO HPD pin if present in the DT, otherwise * we'll use the HDMI core's register. */ @@ -1185,7 +1216,7 @@ static int vc4_hdmi_bind(struct device *dev, struct device *master, void *data) &hpd_gpio_flags); if (hdmi->hpd_gpio < 0) { ret = hdmi->hpd_gpio; - goto err_unprepare_hsm; + goto err_put_i2c; } hdmi->hpd_active_low = hpd_gpio_flags & OF_GPIO_ACTIVE_LOW; @@ -1193,25 +1224,7 @@ static int vc4_hdmi_bind(struct device *dev, struct device *master, void *data) vc4->hdmi = hdmi; - /* HDMI core must be enabled. */ - if (!(HD_READ(VC4_HD_M_CTL) & VC4_HD_M_ENABLE)) { - HD_WRITE(VC4_HD_M_CTL, VC4_HD_M_SW_RST); - udelay(1); - HD_WRITE(VC4_HD_M_CTL, 0); - - HD_WRITE(VC4_HD_M_CTL, VC4_HD_M_ENABLE); - - HDMI_WRITE(VC4_HDMI_SW_RESET_CONTROL, - VC4_HDMI_SW_RESET_HDMI | - VC4_HDMI_SW_RESET_FORMAT_DETECT); - - HDMI_WRITE(VC4_HDMI_SW_RESET_CONTROL, 0); - - /* PHY should be in reset, like - * vc4_hdmi_encoder_disable() does. - */ - HDMI_WRITE(VC4_HDMI_TX_PHY_RESET_CTL, 0xf << 16); - } + pm_runtime_enable(dev); drm_encoder_init(drm, hdmi->encoder, &vc4_hdmi_encoder_funcs, DRM_MODE_ENCODER_TMDS, NULL); @@ -1231,10 +1244,7 @@ static int vc4_hdmi_bind(struct device *dev, struct device *master, void *data) err_destroy_encoder: vc4_hdmi_encoder_destroy(hdmi->encoder); -err_unprepare_hsm: - clk_disable_unprepare(hdmi->hsm_clock); -err_unprepare_pix: - clk_disable_unprepare(hdmi->pixel_clock); + pm_runtime_disable(dev); err_put_i2c: put_device(&hdmi->ddc->dev); @@ -1253,8 +1263,8 @@ static void vc4_hdmi_unbind(struct device *dev, struct device *master, vc4_hdmi_connector_destroy(hdmi->connector); vc4_hdmi_encoder_destroy(hdmi->encoder); - clk_disable_unprepare(hdmi->pixel_clock); - clk_disable_unprepare(hdmi->hsm_clock); + pm_runtime_disable(dev); + put_device(&hdmi->ddc->dev); vc4->hdmi = NULL; diff --git a/drivers/gpu/drm/vc4/vc4_hvs.c b/drivers/gpu/drm/vc4/vc4_hvs.c index fd421ba3c5d7..2b62fc5b8d85 100644 --- a/drivers/gpu/drm/vc4/vc4_hvs.c +++ b/drivers/gpu/drm/vc4/vc4_hvs.c @@ -22,7 +22,7 @@ * each CRTC. */ -#include "linux/component.h" +#include <linux/component.h> #include "vc4_drv.h" #include "vc4_regs.h" diff --git a/drivers/gpu/drm/vc4/vc4_irq.c b/drivers/gpu/drm/vc4/vc4_irq.c index cdc6e6760705..7d7af3a93d94 100644 --- a/drivers/gpu/drm/vc4/vc4_irq.c +++ b/drivers/gpu/drm/vc4/vc4_irq.c @@ -59,50 +59,45 @@ vc4_overflow_mem_work(struct work_struct *work) { struct vc4_dev *vc4 = container_of(work, struct vc4_dev, overflow_mem_work); - struct drm_device *dev = vc4->dev; - struct vc4_bo *bo; + struct vc4_bo *bo = vc4->bin_bo; + int bin_bo_slot; + struct vc4_exec_info *exec; + unsigned long irqflags; - bo = vc4_bo_create(dev, 256 * 1024, true); - if (IS_ERR(bo)) { + bin_bo_slot = vc4_v3d_get_bin_slot(vc4); + if (bin_bo_slot < 0) { DRM_ERROR("Couldn't allocate binner overflow mem\n"); return; } - /* If there's a job executing currently, then our previous - * overflow allocation is getting used in that job and we need - * to queue it to be released when the job is done. But if no - * job is executing at all, then we can free the old overflow - * object direcctly. - * - * No lock necessary for this pointer since we're the only - * ones that update the pointer, and our workqueue won't - * reenter. - */ - if (vc4->overflow_mem) { - struct vc4_exec_info *current_exec; - unsigned long irqflags; - - spin_lock_irqsave(&vc4->job_lock, irqflags); - current_exec = vc4_first_bin_job(vc4); - if (!current_exec) - current_exec = vc4_last_render_job(vc4); - if (current_exec) { - vc4->overflow_mem->seqno = current_exec->seqno; - list_add_tail(&vc4->overflow_mem->unref_head, - ¤t_exec->unref_list); - vc4->overflow_mem = NULL; + spin_lock_irqsave(&vc4->job_lock, irqflags); + + if (vc4->bin_alloc_overflow) { + /* If we had overflow memory allocated previously, + * then that chunk will free when the current bin job + * is done. If we don't have a bin job running, then + * the chunk will be done whenever the list of render + * jobs has drained. + */ + exec = vc4_first_bin_job(vc4); + if (!exec) + exec = vc4_last_render_job(vc4); + if (exec) { + exec->bin_slots |= vc4->bin_alloc_overflow; + } else { + /* There's nothing queued in the hardware, so + * the old slot is free immediately. + */ + vc4->bin_alloc_used &= ~vc4->bin_alloc_overflow; } - spin_unlock_irqrestore(&vc4->job_lock, irqflags); } + vc4->bin_alloc_overflow = BIT(bin_bo_slot); - if (vc4->overflow_mem) - drm_gem_object_unreference_unlocked(&vc4->overflow_mem->base.base); - vc4->overflow_mem = bo; - - V3D_WRITE(V3D_BPOA, bo->base.paddr); + V3D_WRITE(V3D_BPOA, bo->base.paddr + bin_bo_slot * vc4->bin_alloc_size); V3D_WRITE(V3D_BPOS, bo->base.base.size); V3D_WRITE(V3D_INTCTL, V3D_INT_OUTOMEM); V3D_WRITE(V3D_INTENA, V3D_INT_OUTOMEM); + spin_unlock_irqrestore(&vc4->job_lock, irqflags); } static void @@ -142,6 +137,10 @@ vc4_irq_finish_render_job(struct drm_device *dev) vc4->finished_seqno++; list_move_tail(&exec->head, &vc4->job_done_list); + if (exec->fence) { + dma_fence_signal_locked(exec->fence); + exec->fence = NULL; + } vc4_submit_next_render_job(dev); wake_up_all(&vc4->job_wait_queue); diff --git a/drivers/gpu/drm/vc4/vc4_kms.c b/drivers/gpu/drm/vc4/vc4_kms.c index ad7925a9e0ea..928d191ef90f 100644 --- a/drivers/gpu/drm/vc4/vc4_kms.c +++ b/drivers/gpu/drm/vc4/vc4_kms.c @@ -14,12 +14,12 @@ * crtc, HDMI encoder). */ -#include "drm_crtc.h" -#include "drm_atomic.h" -#include "drm_atomic_helper.h" -#include "drm_crtc_helper.h" -#include "drm_plane_helper.h" -#include "drm_fb_cma_helper.h" +#include <drm/drm_crtc.h> +#include <drm/drm_atomic.h> +#include <drm/drm_atomic_helper.h> +#include <drm/drm_crtc_helper.h> +#include <drm/drm_plane_helper.h> +#include <drm/drm_fb_cma_helper.h> #include "vc4_drv.h" static void vc4_output_poll_changed(struct drm_device *dev) @@ -230,10 +230,12 @@ int vc4_kms_load(struct drm_device *dev) drm_mode_config_reset(dev); - vc4->fbdev = drm_fbdev_cma_init(dev, 32, - dev->mode_config.num_connector); - if (IS_ERR(vc4->fbdev)) - vc4->fbdev = NULL; + if (dev->mode_config.num_connector) { + vc4->fbdev = drm_fbdev_cma_init(dev, 32, + dev->mode_config.num_connector); + if (IS_ERR(vc4->fbdev)) + vc4->fbdev = NULL; + } drm_kms_helper_poll_init(dev); diff --git a/drivers/gpu/drm/vc4/vc4_plane.c b/drivers/gpu/drm/vc4/vc4_plane.c index d34cd5393a9b..da18dec21696 100644 --- a/drivers/gpu/drm/vc4/vc4_plane.c +++ b/drivers/gpu/drm/vc4/vc4_plane.c @@ -18,12 +18,13 @@ * into the region of the HVS that it has allocated for us. */ +#include <drm/drm_atomic.h> +#include <drm/drm_atomic_helper.h> +#include <drm/drm_fb_cma_helper.h> +#include <drm/drm_plane_helper.h> + #include "vc4_drv.h" #include "vc4_regs.h" -#include "drm_atomic.h" -#include "drm_atomic_helper.h" -#include "drm_fb_cma_helper.h" -#include "drm_plane_helper.h" enum vc4_scaling_mode { VC4_SCALING_NONE, diff --git a/drivers/gpu/drm/vc4/vc4_render_cl.c b/drivers/gpu/drm/vc4/vc4_render_cl.c index 4339471f517f..5dc19429d4ae 100644 --- a/drivers/gpu/drm/vc4/vc4_render_cl.c +++ b/drivers/gpu/drm/vc4/vc4_render_cl.c @@ -182,8 +182,7 @@ static void emit_tile(struct vc4_exec_info *exec, if (has_bin) { rcl_u8(setup, VC4_PACKET_BRANCH_TO_SUB_LIST); - rcl_u32(setup, (exec->tile_bo->paddr + - exec->tile_alloc_offset + + rcl_u32(setup, (exec->tile_alloc_offset + (y * exec->bin_tiles_x + x) * 32)); } diff --git a/drivers/gpu/drm/vc4/vc4_v3d.c b/drivers/gpu/drm/vc4/vc4_v3d.c index 7cc346ad9b0b..5ae5518e605b 100644 --- a/drivers/gpu/drm/vc4/vc4_v3d.c +++ b/drivers/gpu/drm/vc4/vc4_v3d.c @@ -16,8 +16,9 @@ * this program. If not, see <http://www.gnu.org/licenses/>. */ -#include "linux/component.h" -#include "linux/pm_runtime.h" +#include <linux/clk.h> +#include <linux/component.h> +#include <linux/pm_runtime.h> #include "vc4_drv.h" #include "vc4_regs.h" @@ -156,6 +157,144 @@ static void vc4_v3d_init_hw(struct drm_device *dev) V3D_WRITE(V3D_VPMBASE, 0); } +int vc4_v3d_get_bin_slot(struct vc4_dev *vc4) +{ + struct drm_device *dev = vc4->dev; + unsigned long irqflags; + int slot; + uint64_t seqno = 0; + struct vc4_exec_info *exec; + +try_again: + spin_lock_irqsave(&vc4->job_lock, irqflags); + slot = ffs(~vc4->bin_alloc_used); + if (slot != 0) { + /* Switch from ffs() bit index to a 0-based index. */ + slot--; + vc4->bin_alloc_used |= BIT(slot); + spin_unlock_irqrestore(&vc4->job_lock, irqflags); + return slot; + } + + /* Couldn't find an open slot. Wait for render to complete + * and try again. + */ + exec = vc4_last_render_job(vc4); + if (exec) + seqno = exec->seqno; + spin_unlock_irqrestore(&vc4->job_lock, irqflags); + + if (seqno) { + int ret = vc4_wait_for_seqno(dev, seqno, ~0ull, true); + + if (ret == 0) + goto try_again; + + return ret; + } + + return -ENOMEM; +} + +/** + * vc4_allocate_bin_bo() - allocates the memory that will be used for + * tile binning. + * + * The binner has a limitation that the addresses in the tile state + * buffer that point into the tile alloc buffer or binner overflow + * memory only have 28 bits (256MB), and the top 4 on the bus for + * tile alloc references end up coming from the tile state buffer's + * address. + * + * To work around this, we allocate a single large buffer while V3D is + * in use, make sure that it has the top 4 bits constant across its + * entire extent, and then put the tile state, tile alloc, and binner + * overflow memory inside that buffer. + * + * This creates a limitation where we may not be able to execute a job + * if it doesn't fit within the buffer that we allocated up front. + * However, it turns out that 16MB is "enough for anybody", and + * real-world applications run into allocation failures from the + * overall CMA pool before they make scenes complicated enough to run + * out of bin space. + */ +int +vc4_allocate_bin_bo(struct drm_device *drm) +{ + struct vc4_dev *vc4 = to_vc4_dev(drm); + struct vc4_v3d *v3d = vc4->v3d; + uint32_t size = 16 * 1024 * 1024; + int ret = 0; + struct list_head list; + + /* We may need to try allocating more than once to get a BO + * that doesn't cross 256MB. Track the ones we've allocated + * that failed so far, so that we can free them when we've got + * one that succeeded (if we freed them right away, our next + * allocation would probably be the same chunk of memory). + */ + INIT_LIST_HEAD(&list); + + while (true) { + struct vc4_bo *bo = vc4_bo_create(drm, size, true); + + if (IS_ERR(bo)) { + ret = PTR_ERR(bo); + + dev_err(&v3d->pdev->dev, + "Failed to allocate memory for tile binning: " + "%d. You may need to enable CMA or give it " + "more memory.", + ret); + break; + } + + /* Check if this BO won't trigger the addressing bug. */ + if ((bo->base.paddr & 0xf0000000) == + ((bo->base.paddr + bo->base.base.size - 1) & 0xf0000000)) { + vc4->bin_bo = bo; + + /* Set up for allocating 512KB chunks of + * binner memory. The biggest allocation we + * need to do is for the initial tile alloc + + * tile state buffer. We can render to a + * maximum of ((2048*2048) / (32*32) = 4096 + * tiles in a frame (until we do floating + * point rendering, at which point it would be + * 8192). Tile state is 48b/tile (rounded to + * a page), and tile alloc is 32b/tile + * (rounded to a page), plus a page of extra, + * for a total of 320kb for our worst-case. + * We choose 512kb so that it divides evenly + * into our 16MB, and the rest of the 512kb + * will be used as storage for the overflow + * from the initial 32b CL per bin. + */ + vc4->bin_alloc_size = 512 * 1024; + vc4->bin_alloc_used = 0; + vc4->bin_alloc_overflow = 0; + WARN_ON_ONCE(sizeof(vc4->bin_alloc_used) * 8 != + bo->base.base.size / vc4->bin_alloc_size); + + break; + } + + /* Put it on the list to free later, and try again. */ + list_add(&bo->unref_head, &list); + } + + /* Free all the BOs we allocated but didn't choose. */ + while (!list_empty(&list)) { + struct vc4_bo *bo = list_last_entry(&list, + struct vc4_bo, unref_head); + + list_del(&bo->unref_head); + drm_gem_object_put_unlocked(&bo->base.base); + } + + return ret; +} + #ifdef CONFIG_PM static int vc4_v3d_runtime_suspend(struct device *dev) { @@ -164,6 +303,11 @@ static int vc4_v3d_runtime_suspend(struct device *dev) vc4_irq_uninstall(vc4->dev); + drm_gem_object_put_unlocked(&vc4->bin_bo->base.base); + vc4->bin_bo = NULL; + + clk_disable_unprepare(v3d->clk); + return 0; } @@ -171,6 +315,15 @@ static int vc4_v3d_runtime_resume(struct device *dev) { struct vc4_v3d *v3d = dev_get_drvdata(dev); struct vc4_dev *vc4 = v3d->vc4; + int ret; + + ret = vc4_allocate_bin_bo(vc4->dev); + if (ret) + return ret; + + ret = clk_prepare_enable(v3d->clk); + if (ret != 0) + return ret; vc4_v3d_init_hw(vc4->dev); vc4_irq_postinstall(vc4->dev); @@ -202,12 +355,38 @@ static int vc4_v3d_bind(struct device *dev, struct device *master, void *data) vc4->v3d = v3d; v3d->vc4 = vc4; + v3d->clk = devm_clk_get(dev, NULL); + if (IS_ERR(v3d->clk)) { + int ret = PTR_ERR(v3d->clk); + + if (ret == -ENOENT) { + /* bcm2835 didn't have a clock reference in the DT. */ + ret = 0; + v3d->clk = NULL; + } else { + if (ret != -EPROBE_DEFER) + dev_err(dev, "Failed to get V3D clock: %d\n", + ret); + return ret; + } + } + if (V3D_READ(V3D_IDENT0) != V3D_EXPECTED_IDENT0) { DRM_ERROR("V3D_IDENT0 read 0x%08x instead of 0x%08x\n", V3D_READ(V3D_IDENT0), V3D_EXPECTED_IDENT0); return -EINVAL; } + ret = clk_prepare_enable(v3d->clk); + if (ret != 0) + return ret; + + ret = vc4_allocate_bin_bo(drm); + if (ret) { + clk_disable_unprepare(v3d->clk); + return ret; + } + /* Reset the binner overflow address/size at setup, to be sure * we don't reuse an old one. */ @@ -271,6 +450,7 @@ static int vc4_v3d_dev_remove(struct platform_device *pdev) static const struct of_device_id vc4_v3d_dt_match[] = { { .compatible = "brcm,bcm2835-v3d" }, + { .compatible = "brcm,cygnus-v3d" }, { .compatible = "brcm,vc4-v3d" }, {} }; diff --git a/drivers/gpu/drm/vc4/vc4_validate.c b/drivers/gpu/drm/vc4/vc4_validate.c index da6f1e138e8d..814b512c6b9a 100644 --- a/drivers/gpu/drm/vc4/vc4_validate.c +++ b/drivers/gpu/drm/vc4/vc4_validate.c @@ -172,7 +172,8 @@ vc4_check_tex_size(struct vc4_exec_info *exec, struct drm_gem_cma_object *fbo, * our math. */ if (width > 4096 || height > 4096) { - DRM_ERROR("Surface dimesions (%d,%d) too large", width, height); + DRM_ERROR("Surface dimensions (%d,%d) too large", + width, height); return false; } @@ -348,10 +349,11 @@ static int validate_tile_binning_config(VALIDATE_ARGS) { struct drm_device *dev = exec->exec_bo->base.dev; - struct vc4_bo *tile_bo; + struct vc4_dev *vc4 = to_vc4_dev(dev); uint8_t flags; - uint32_t tile_state_size, tile_alloc_size; - uint32_t tile_count; + uint32_t tile_state_size; + uint32_t tile_count, bin_addr; + int bin_slot; if (exec->found_tile_binning_mode_config_packet) { DRM_ERROR("Duplicate VC4_PACKET_TILE_BINNING_MODE_CONFIG\n"); @@ -377,13 +379,28 @@ validate_tile_binning_config(VALIDATE_ARGS) return -EINVAL; } + bin_slot = vc4_v3d_get_bin_slot(vc4); + if (bin_slot < 0) { + if (bin_slot != -EINTR && bin_slot != -ERESTARTSYS) { + DRM_ERROR("Failed to allocate binner memory: %d\n", + bin_slot); + } + return bin_slot; + } + + /* The slot we allocated will only be used by this job, and is + * free when the job completes rendering. + */ + exec->bin_slots |= BIT(bin_slot); + bin_addr = vc4->bin_bo->base.paddr + bin_slot * vc4->bin_alloc_size; + /* The tile state data array is 48 bytes per tile, and we put it at * the start of a BO containing both it and the tile alloc. */ tile_state_size = 48 * tile_count; /* Since the tile alloc array will follow us, align. */ - exec->tile_alloc_offset = roundup(tile_state_size, 4096); + exec->tile_alloc_offset = bin_addr + roundup(tile_state_size, 4096); *(uint8_t *)(validated + 14) = ((flags & ~(VC4_BIN_CONFIG_ALLOC_INIT_BLOCK_SIZE_MASK | @@ -394,35 +411,13 @@ validate_tile_binning_config(VALIDATE_ARGS) VC4_SET_FIELD(VC4_BIN_CONFIG_ALLOC_BLOCK_SIZE_128, VC4_BIN_CONFIG_ALLOC_BLOCK_SIZE)); - /* Initial block size. */ - tile_alloc_size = 32 * tile_count; - - /* - * The initial allocation gets rounded to the next 256 bytes before - * the hardware starts fulfilling further allocations. - */ - tile_alloc_size = roundup(tile_alloc_size, 256); - - /* Add space for the extra allocations. This is what gets used first, - * before overflow memory. It must have at least 4096 bytes, but we - * want to avoid overflow memory usage if possible. - */ - tile_alloc_size += 1024 * 1024; - - tile_bo = vc4_bo_create(dev, exec->tile_alloc_offset + tile_alloc_size, - true); - exec->tile_bo = &tile_bo->base; - if (IS_ERR(exec->tile_bo)) - return PTR_ERR(exec->tile_bo); - list_add_tail(&tile_bo->unref_head, &exec->unref_list); - /* tile alloc address. */ - *(uint32_t *)(validated + 0) = (exec->tile_bo->paddr + - exec->tile_alloc_offset); + *(uint32_t *)(validated + 0) = exec->tile_alloc_offset; /* tile alloc size. */ - *(uint32_t *)(validated + 4) = tile_alloc_size; + *(uint32_t *)(validated + 4) = (bin_addr + vc4->bin_alloc_size - + exec->tile_alloc_offset); /* tile state address. */ - *(uint32_t *)(validated + 8) = exec->tile_bo->paddr; + *(uint32_t *)(validated + 8) = bin_addr; return 0; } diff --git a/drivers/gpu/drm/vgem/Makefile b/drivers/gpu/drm/vgem/Makefile index bfcdea1330e6..cb5d413b9c93 100644 --- a/drivers/gpu/drm/vgem/Makefile +++ b/drivers/gpu/drm/vgem/Makefile @@ -1,4 +1,3 @@ -ccflags-y := -Iinclude/drm vgem-y := vgem_drv.o vgem_fence.o obj-$(CONFIG_DRM_VGEM) += vgem.o diff --git a/drivers/gpu/drm/vgem/vgem_drv.c b/drivers/gpu/drm/vgem/vgem_drv.c index 9fee38a942c4..18f401b442c2 100644 --- a/drivers/gpu/drm/vgem/vgem_drv.c +++ b/drivers/gpu/drm/vgem/vgem_drv.c @@ -42,10 +42,20 @@ #define DRIVER_MAJOR 1 #define DRIVER_MINOR 0 +static struct vgem_device { + struct drm_device drm; + struct platform_device *platform; +} *vgem_device; + static void vgem_gem_free_object(struct drm_gem_object *obj) { struct drm_vgem_gem_object *vgem_obj = to_vgem_bo(obj); + kvfree(vgem_obj->pages); + + if (obj->import_attach) + drm_prime_gem_destroy(obj, vgem_obj->table); + drm_gem_object_release(obj); kfree(vgem_obj); } @@ -56,26 +66,49 @@ static int vgem_gem_fault(struct vm_fault *vmf) struct drm_vgem_gem_object *obj = vma->vm_private_data; /* We don't use vmf->pgoff since that has the fake offset */ unsigned long vaddr = vmf->address; - struct page *page; - - page = shmem_read_mapping_page(file_inode(obj->base.filp)->i_mapping, - (vaddr - vma->vm_start) >> PAGE_SHIFT); - if (!IS_ERR(page)) { - vmf->page = page; - return 0; - } else switch (PTR_ERR(page)) { - case -ENOSPC: - case -ENOMEM: - return VM_FAULT_OOM; - case -EBUSY: - return VM_FAULT_RETRY; - case -EFAULT: - case -EINVAL: - return VM_FAULT_SIGBUS; - default: - WARN_ON_ONCE(PTR_ERR(page)); - return VM_FAULT_SIGBUS; + int ret; + loff_t num_pages; + pgoff_t page_offset; + page_offset = (vaddr - vma->vm_start) >> PAGE_SHIFT; + + num_pages = DIV_ROUND_UP(obj->base.size, PAGE_SIZE); + + if (page_offset > num_pages) + return VM_FAULT_SIGBUS; + + if (obj->pages) { + get_page(obj->pages[page_offset]); + vmf->page = obj->pages[page_offset]; + ret = 0; + } else { + struct page *page; + + page = shmem_read_mapping_page( + file_inode(obj->base.filp)->i_mapping, + page_offset); + if (!IS_ERR(page)) { + vmf->page = page; + ret = 0; + } else switch (PTR_ERR(page)) { + case -ENOSPC: + case -ENOMEM: + ret = VM_FAULT_OOM; + break; + case -EBUSY: + ret = VM_FAULT_RETRY; + break; + case -EFAULT: + case -EINVAL: + ret = VM_FAULT_SIGBUS; + break; + default: + WARN_ON(PTR_ERR(page)); + ret = VM_FAULT_SIGBUS; + break; + } + } + return ret; } static const struct vm_operations_struct vgem_gem_vm_ops = { @@ -112,12 +145,8 @@ static void vgem_postclose(struct drm_device *dev, struct drm_file *file) kfree(vfile); } -/* ioctls */ - -static struct drm_gem_object *vgem_gem_create(struct drm_device *dev, - struct drm_file *file, - unsigned int *handle, - unsigned long size) +static struct drm_vgem_gem_object *__vgem_gem_create(struct drm_device *dev, + unsigned long size) { struct drm_vgem_gem_object *obj; int ret; @@ -127,8 +156,31 @@ static struct drm_gem_object *vgem_gem_create(struct drm_device *dev, return ERR_PTR(-ENOMEM); ret = drm_gem_object_init(dev, &obj->base, roundup(size, PAGE_SIZE)); - if (ret) - goto err_free; + if (ret) { + kfree(obj); + return ERR_PTR(ret); + } + + return obj; +} + +static void __vgem_gem_destroy(struct drm_vgem_gem_object *obj) +{ + drm_gem_object_release(&obj->base); + kfree(obj); +} + +static struct drm_gem_object *vgem_gem_create(struct drm_device *dev, + struct drm_file *file, + unsigned int *handle, + unsigned long size) +{ + struct drm_vgem_gem_object *obj; + int ret; + + obj = __vgem_gem_create(dev, size); + if (IS_ERR(obj)) + return ERR_CAST(obj); ret = drm_gem_handle_create(file, &obj->base, handle); drm_gem_object_unreference_unlocked(&obj->base); @@ -137,9 +189,8 @@ static struct drm_gem_object *vgem_gem_create(struct drm_device *dev, return &obj->base; -err_free: - kfree(obj); err: + __vgem_gem_destroy(obj); return ERR_PTR(ret); } @@ -256,6 +307,37 @@ static struct sg_table *vgem_prime_get_sg_table(struct drm_gem_object *obj) return st; } +static struct drm_gem_object* vgem_prime_import(struct drm_device *dev, + struct dma_buf *dma_buf) +{ + struct vgem_device *vgem = container_of(dev, typeof(*vgem), drm); + + return drm_gem_prime_import_dev(dev, dma_buf, &vgem->platform->dev); +} + +static struct drm_gem_object *vgem_prime_import_sg_table(struct drm_device *dev, + struct dma_buf_attachment *attach, struct sg_table *sg) +{ + struct drm_vgem_gem_object *obj; + int npages; + + obj = __vgem_gem_create(dev, attach->dmabuf->size); + if (IS_ERR(obj)) + return ERR_CAST(obj); + + npages = PAGE_ALIGN(attach->dmabuf->size) / PAGE_SIZE; + + obj->table = sg; + obj->pages = kvmalloc_array(npages, sizeof(struct page *), GFP_KERNEL); + if (!obj->pages) { + __vgem_gem_destroy(obj); + return ERR_PTR(-ENOMEM); + } + drm_prime_sg_to_page_addr_arrays(obj->table, obj->pages, NULL, + npages); + return &obj->base; +} + static void *vgem_prime_vmap(struct drm_gem_object *obj) { long n_pages = obj->size >> PAGE_SHIFT; @@ -300,8 +382,19 @@ static int vgem_prime_mmap(struct drm_gem_object *obj, return 0; } +static void vgem_release(struct drm_device *dev) +{ + struct vgem_device *vgem = container_of(dev, typeof(*vgem), drm); + + platform_device_unregister(vgem->platform); + drm_dev_fini(&vgem->drm); + + kfree(vgem); +} + static struct drm_driver vgem_driver = { .driver_features = DRIVER_GEM | DRIVER_PRIME, + .release = vgem_release, .open = vgem_open, .postclose = vgem_postclose, .gem_free_object_unlocked = vgem_gem_free_object, @@ -314,8 +407,11 @@ static struct drm_driver vgem_driver = { .dumb_map_offset = vgem_gem_dumb_map, .prime_handle_to_fd = drm_gem_prime_handle_to_fd, + .prime_fd_to_handle = drm_gem_prime_fd_to_handle, .gem_prime_pin = vgem_prime_pin, + .gem_prime_import = vgem_prime_import, .gem_prime_export = drm_gem_prime_export, + .gem_prime_import_sg_table = vgem_prime_import_sg_table, .gem_prime_get_sg_table = vgem_prime_get_sg_table, .gem_prime_vmap = vgem_prime_vmap, .gem_prime_vunmap = vgem_prime_vunmap, @@ -328,34 +424,48 @@ static struct drm_driver vgem_driver = { .minor = DRIVER_MINOR, }; -static struct drm_device *vgem_device; - static int __init vgem_init(void) { int ret; - vgem_device = drm_dev_alloc(&vgem_driver, NULL); - if (IS_ERR(vgem_device)) { - ret = PTR_ERR(vgem_device); - goto out; + vgem_device = kzalloc(sizeof(*vgem_device), GFP_KERNEL); + if (!vgem_device) + return -ENOMEM; + + ret = drm_dev_init(&vgem_device->drm, &vgem_driver, NULL); + if (ret) + goto out_free; + + vgem_device->platform = + platform_device_register_simple("vgem", -1, NULL, 0); + if (IS_ERR(vgem_device->platform)) { + ret = PTR_ERR(vgem_device->platform); + goto out_fini; } - ret = drm_dev_register(vgem_device, 0); + dma_coerce_mask_and_coherent(&vgem_device->platform->dev, + DMA_BIT_MASK(64)); + + /* Final step: expose the device/driver to userspace */ + ret = drm_dev_register(&vgem_device->drm, 0); if (ret) - goto out_unref; + goto out_unregister; return 0; -out_unref: - drm_dev_unref(vgem_device); -out: +out_unregister: + platform_device_unregister(vgem_device->platform); +out_fini: + drm_dev_fini(&vgem_device->drm); +out_free: + kfree(vgem_device); return ret; } static void __exit vgem_exit(void) { - drm_dev_unregister(vgem_device); - drm_dev_unref(vgem_device); + drm_dev_unregister(&vgem_device->drm); + drm_dev_unref(&vgem_device->drm); } module_init(vgem_init); diff --git a/drivers/gpu/drm/vgem/vgem_drv.h b/drivers/gpu/drm/vgem/vgem_drv.h index cb59c7ab98b9..1aae01419112 100644 --- a/drivers/gpu/drm/vgem/vgem_drv.h +++ b/drivers/gpu/drm/vgem/vgem_drv.h @@ -43,6 +43,8 @@ struct vgem_file { #define to_vgem_bo(x) container_of(x, struct drm_vgem_gem_object, base) struct drm_vgem_gem_object { struct drm_gem_object base; + struct page **pages; + struct sg_table *table; }; int vgem_fence_open(struct vgem_file *file); diff --git a/drivers/gpu/drm/via/Makefile b/drivers/gpu/drm/via/Makefile index d59e258e2c13..751fa8b8a014 100644 --- a/drivers/gpu/drm/via/Makefile +++ b/drivers/gpu/drm/via/Makefile @@ -2,7 +2,6 @@ # Makefile for the drm device driver. This driver provides support for the # Direct Rendering Infrastructure (DRI) in XFree86 4.1.0 and higher. -ccflags-y := -Iinclude/drm via-y := via_irq.o via_drv.o via_map.o via_mm.o via_dma.o via_verifier.o via_video.o via_dmablit.o obj-$(CONFIG_DRM_VIA) +=via.o diff --git a/drivers/gpu/drm/virtio/Makefile b/drivers/gpu/drm/virtio/Makefile index 3fb8eac1084f..7684f613bdc3 100644 --- a/drivers/gpu/drm/virtio/Makefile +++ b/drivers/gpu/drm/virtio/Makefile @@ -2,8 +2,6 @@ # Makefile for the drm device driver. This driver provides support for the # Direct Rendering Infrastructure (DRI) in XFree86 4.1.0 and higher. -ccflags-y := -Iinclude/drm - virtio-gpu-y := virtgpu_drv.o virtgpu_kms.o virtgpu_drm_bus.o virtgpu_gem.o \ virtgpu_fb.o virtgpu_display.o virtgpu_vq.o virtgpu_ttm.o \ virtgpu_fence.o virtgpu_object.o virtgpu_debugfs.o virtgpu_plane.o \ diff --git a/drivers/gpu/drm/virtio/virtgpu_debugfs.c b/drivers/gpu/drm/virtio/virtgpu_debugfs.c index f51240aa720d..73dc99046c43 100644 --- a/drivers/gpu/drm/virtio/virtgpu_debugfs.c +++ b/drivers/gpu/drm/virtio/virtgpu_debugfs.c @@ -24,8 +24,8 @@ */ #include <linux/debugfs.h> +#include <drm/drmP.h> -#include "drmP.h" #include "virtgpu_drv.h" static int diff --git a/drivers/gpu/drm/virtio/virtgpu_drv.c b/drivers/gpu/drm/virtio/virtgpu_drv.c index 2d29b0141545..63d35c7e416c 100644 --- a/drivers/gpu/drm/virtio/virtgpu_drv.c +++ b/drivers/gpu/drm/virtio/virtgpu_drv.c @@ -29,8 +29,8 @@ #include <linux/module.h> #include <linux/console.h> #include <linux/pci.h> -#include "drmP.h" -#include "drm/drm.h" +#include <drm/drmP.h> +#include <drm/drm.h> #include "virtgpu_drv.h" static struct drm_driver driver; diff --git a/drivers/gpu/drm/virtio/virtgpu_drv.h b/drivers/gpu/drm/virtio/virtgpu_drv.h index 1328185bfd59..3a66abb8fd50 100644 --- a/drivers/gpu/drm/virtio/virtgpu_drv.h +++ b/drivers/gpu/drm/virtio/virtgpu_drv.h @@ -36,10 +36,10 @@ #include <drm/drm_atomic.h> #include <drm/drm_crtc_helper.h> #include <drm/drm_encoder.h> -#include <ttm/ttm_bo_api.h> -#include <ttm/ttm_bo_driver.h> -#include <ttm/ttm_placement.h> -#include <ttm/ttm_module.h> +#include <drm/ttm/ttm_bo_api.h> +#include <drm/ttm/ttm_bo_driver.h> +#include <drm/ttm/ttm_placement.h> +#include <drm/ttm/ttm_module.h> #define DRIVER_NAME "virtio_gpu" #define DRIVER_DESC "virtio GPU" diff --git a/drivers/gpu/drm/virtio/virtgpu_ioctl.c b/drivers/gpu/drm/virtio/virtgpu_ioctl.c index 61f3a963af95..b94bd5440e57 100644 --- a/drivers/gpu/drm/virtio/virtgpu_ioctl.c +++ b/drivers/gpu/drm/virtio/virtgpu_ioctl.c @@ -26,9 +26,10 @@ */ #include <drm/drmP.h> -#include "virtgpu_drv.h" #include <drm/virtgpu_drm.h> -#include "ttm/ttm_execbuf_util.h" +#include <drm/ttm/ttm_execbuf_util.h> + +#include "virtgpu_drv.h" static void convert_to_hw_box(struct virtio_gpu_box *dst, const struct drm_virtgpu_3d_box *src) @@ -119,13 +120,14 @@ static int virtio_gpu_execbuffer_ioctl(struct drm_device *dev, void *data, INIT_LIST_HEAD(&validate_list); if (exbuf->num_bo_handles) { - bo_handles = drm_malloc_ab(exbuf->num_bo_handles, - sizeof(uint32_t)); - buflist = drm_calloc_large(exbuf->num_bo_handles, - sizeof(struct ttm_validate_buffer)); + bo_handles = kvmalloc_array(exbuf->num_bo_handles, + sizeof(uint32_t), GFP_KERNEL); + buflist = kvmalloc_array(exbuf->num_bo_handles, + sizeof(struct ttm_validate_buffer), + GFP_KERNEL | __GFP_ZERO); if (!bo_handles || !buflist) { - drm_free_large(bo_handles); - drm_free_large(buflist); + kvfree(bo_handles); + kvfree(buflist); return -ENOMEM; } @@ -133,16 +135,16 @@ static int virtio_gpu_execbuffer_ioctl(struct drm_device *dev, void *data, if (copy_from_user(bo_handles, user_bo_handles, exbuf->num_bo_handles * sizeof(uint32_t))) { ret = -EFAULT; - drm_free_large(bo_handles); - drm_free_large(buflist); + kvfree(bo_handles); + kvfree(buflist); return ret; } for (i = 0; i < exbuf->num_bo_handles; i++) { gobj = drm_gem_object_lookup(drm_file, bo_handles[i]); if (!gobj) { - drm_free_large(bo_handles); - drm_free_large(buflist); + kvfree(bo_handles); + kvfree(buflist); return -ENOENT; } @@ -151,7 +153,7 @@ static int virtio_gpu_execbuffer_ioctl(struct drm_device *dev, void *data, list_add(&buflist[i].head, &validate_list); } - drm_free_large(bo_handles); + kvfree(bo_handles); } ret = virtio_gpu_object_list_validate(&ticket, &validate_list); @@ -171,7 +173,7 @@ static int virtio_gpu_execbuffer_ioctl(struct drm_device *dev, void *data, /* fence the command bo */ virtio_gpu_unref_list(&validate_list); - drm_free_large(buflist); + kvfree(buflist); dma_fence_put(&fence->f); return 0; @@ -179,7 +181,7 @@ out_unresv: ttm_eu_backoff_reservation(&ticket, &validate_list); out_free: virtio_gpu_unref_list(&validate_list); - drm_free_large(buflist); + kvfree(buflist); return ret; } diff --git a/drivers/gpu/drm/virtio/virtgpu_kms.c b/drivers/gpu/drm/virtio/virtgpu_kms.c index 491866865c33..1e1c90b30d4a 100644 --- a/drivers/gpu/drm/virtio/virtgpu_kms.c +++ b/drivers/gpu/drm/virtio/virtgpu_kms.c @@ -175,8 +175,7 @@ int virtio_gpu_driver_load(struct drm_device *dev, unsigned long flags) DRM_INFO("virgl 3d acceleration not supported by guest\n"); #endif - ret = vgdev->vdev->config->find_vqs(vgdev->vdev, 2, vqs, - callbacks, names, NULL); + ret = virtio_find_vqs(vgdev->vdev, 2, vqs, callbacks, names, NULL); if (ret) { DRM_ERROR("failed to find virt queues\n"); goto err_vqs; diff --git a/drivers/gpu/drm/virtio/virtgpu_ttm.c b/drivers/gpu/drm/virtio/virtgpu_ttm.c index 4e8e27d50922..c1f2af4ca4ca 100644 --- a/drivers/gpu/drm/virtio/virtgpu_ttm.c +++ b/drivers/gpu/drm/virtio/virtgpu_ttm.c @@ -25,11 +25,11 @@ * OTHER DEALINGS IN THE SOFTWARE. */ -#include <ttm/ttm_bo_api.h> -#include <ttm/ttm_bo_driver.h> -#include <ttm/ttm_placement.h> -#include <ttm/ttm_page_alloc.h> -#include <ttm/ttm_module.h> +#include <drm/ttm/ttm_bo_api.h> +#include <drm/ttm/ttm_bo_driver.h> +#include <drm/ttm/ttm_placement.h> +#include <drm/ttm/ttm_page_alloc.h> +#include <drm/ttm/ttm_module.h> #include <drm/drmP.h> #include <drm/drm.h> #include <drm/virtgpu_drm.h> diff --git a/drivers/gpu/drm/vmwgfx/Makefile b/drivers/gpu/drm/vmwgfx/Makefile index aac17a640cce..a365330bbb82 100644 --- a/drivers/gpu/drm/vmwgfx/Makefile +++ b/drivers/gpu/drm/vmwgfx/Makefile @@ -1,6 +1,3 @@ - -ccflags-y := -Iinclude/drm - vmwgfx-y := vmwgfx_execbuf.o vmwgfx_gmr.o vmwgfx_kms.o vmwgfx_drv.o \ vmwgfx_fb.o vmwgfx_ioctl.o vmwgfx_resource.o vmwgfx_buffer.o \ vmwgfx_fifo.o vmwgfx_irq.o vmwgfx_ldu.o vmwgfx_ttm_glue.o \ diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_cmdbuf.c b/drivers/gpu/drm/vmwgfx/vmwgfx_cmdbuf.c index 77cb7c627e09..99a7f4ab7d97 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_cmdbuf.c +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_cmdbuf.c @@ -25,8 +25,9 @@ * **************************************************************************/ +#include <drm/ttm/ttm_bo_api.h> + #include "vmwgfx_drv.h" -#include "ttm/ttm_bo_api.h" /* * Size of inline command buffers. Try to make sure that a page size is a diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_context.c b/drivers/gpu/drm/vmwgfx/vmwgfx_context.c index 443d1ed00de7..bcc6d4136c87 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_context.c +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_context.c @@ -25,10 +25,11 @@ * **************************************************************************/ +#include <drm/ttm/ttm_placement.h> + #include "vmwgfx_drv.h" #include "vmwgfx_resource_priv.h" #include "vmwgfx_binding.h" -#include "ttm/ttm_placement.h" struct vmw_user_context { struct ttm_base_object base; diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_cotable.c b/drivers/gpu/drm/vmwgfx/vmwgfx_cotable.c index 265c81e6cf39..6c026d75c180 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_cotable.c +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_cotable.c @@ -30,9 +30,10 @@ * whenever the backing MOB is evicted. */ +#include <drm/ttm/ttm_placement.h> + #include "vmwgfx_drv.h" #include "vmwgfx_resource_priv.h" -#include <ttm/ttm_placement.h> #include "vmwgfx_so.h" /** diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_kms.c b/drivers/gpu/drm/vmwgfx/vmwgfx_kms.c index ef9f3a2a4030..a8876b070168 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_kms.c +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_kms.c @@ -566,7 +566,7 @@ int vmw_du_primary_plane_atomic_check(struct drm_plane *plane, ret = drm_plane_helper_check_update(plane, state->crtc, new_fb, &src, &dest, &clip, - DRM_ROTATE_0, + DRM_MODE_ROTATE_0, DRM_PLANE_HELPER_NO_SCALING, DRM_PLANE_HELPER_NO_SCALING, false, true, &visible); @@ -845,7 +845,7 @@ void vmw_du_plane_reset(struct drm_plane *plane) plane->state = &vps->base; plane->state->plane = plane; - plane->state->rotation = DRM_ROTATE_0; + plane->state->rotation = DRM_MODE_ROTATE_0; } diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_msg.c b/drivers/gpu/drm/vmwgfx/vmwgfx_msg.c index e57a0bad7a62..6063c9636d4a 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_msg.c +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_msg.c @@ -30,7 +30,7 @@ #include <linux/kernel.h> #include <linux/frame.h> #include <asm/hypervisor.h> -#include "drmP.h" +#include <drm/drmP.h> #include "vmwgfx_msg.h" diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_shader.c b/drivers/gpu/drm/vmwgfx/vmwgfx_shader.c index 92f8b1d04f0f..68f135c5b0d8 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_shader.c +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_shader.c @@ -25,10 +25,11 @@ * **************************************************************************/ +#include <drm/ttm/ttm_placement.h> + #include "vmwgfx_drv.h" #include "vmwgfx_resource_priv.h" #include "vmwgfx_binding.h" -#include "ttm/ttm_placement.h" struct vmw_shader { struct vmw_resource res; diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_surface.c b/drivers/gpu/drm/vmwgfx/vmwgfx_surface.c index 7681341fe32b..5900cff5bbc3 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_surface.c +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_surface.c @@ -25,11 +25,12 @@ * **************************************************************************/ +#include <drm/ttm/ttm_placement.h> + #include "vmwgfx_drv.h" #include "vmwgfx_resource_priv.h" #include "vmwgfx_so.h" #include "vmwgfx_binding.h" -#include <ttm/ttm_placement.h> #include "device_include/svga3d_surfacedefs.h" diff --git a/drivers/gpu/drm/zte/Makefile b/drivers/gpu/drm/zte/Makefile index 01352b56c418..9df7766a7f9d 100644 --- a/drivers/gpu/drm/zte/Makefile +++ b/drivers/gpu/drm/zte/Makefile @@ -3,6 +3,7 @@ zxdrm-y := \ zx_hdmi.o \ zx_plane.o \ zx_tvenc.o \ + zx_vga.o \ zx_vou.o obj-$(CONFIG_DRM_ZTE) += zxdrm.o diff --git a/drivers/gpu/drm/zte/zx_common_regs.h b/drivers/gpu/drm/zte/zx_common_regs.h new file mode 100644 index 000000000000..2afd80664c51 --- /dev/null +++ b/drivers/gpu/drm/zte/zx_common_regs.h @@ -0,0 +1,31 @@ +/* + * Copyright (C) 2017 Sanechips Technology Co., Ltd. + * Copyright 2017 Linaro Ltd. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ + +#ifndef __ZX_COMMON_REGS_H__ +#define __ZX_COMMON_REGS_H__ + +/* CSC registers */ +#define CSC_CTRL0 0x30 +#define CSC_COV_MODE_SHIFT 16 +#define CSC_COV_MODE_MASK (0xffff << CSC_COV_MODE_SHIFT) +#define CSC_BT601_IMAGE_RGB2YCBCR 0 +#define CSC_BT601_IMAGE_YCBCR2RGB 1 +#define CSC_BT601_VIDEO_RGB2YCBCR 2 +#define CSC_BT601_VIDEO_YCBCR2RGB 3 +#define CSC_BT709_IMAGE_RGB2YCBCR 4 +#define CSC_BT709_IMAGE_YCBCR2RGB 5 +#define CSC_BT709_VIDEO_RGB2YCBCR 6 +#define CSC_BT709_VIDEO_YCBCR2RGB 7 +#define CSC_BT2020_IMAGE_RGB2YCBCR 8 +#define CSC_BT2020_IMAGE_YCBCR2RGB 9 +#define CSC_BT2020_VIDEO_RGB2YCBCR 10 +#define CSC_BT2020_VIDEO_YCBCR2RGB 11 +#define CSC_WORK_ENABLE BIT(0) + +#endif /* __ZX_COMMON_REGS_H__ */ diff --git a/drivers/gpu/drm/zte/zx_drm_drv.c b/drivers/gpu/drm/zte/zx_drm_drv.c index 614e854f6be5..490aafc99610 100644 --- a/drivers/gpu/drm/zte/zx_drm_drv.c +++ b/drivers/gpu/drm/zte/zx_drm_drv.c @@ -233,6 +233,7 @@ static struct platform_driver *drivers[] = { &zx_crtc_driver, &zx_hdmi_driver, &zx_tvenc_driver, + &zx_vga_driver, &zx_drm_platform_driver, }; diff --git a/drivers/gpu/drm/zte/zx_drm_drv.h b/drivers/gpu/drm/zte/zx_drm_drv.h index 5ca035b079c7..2a8cdc5f8be4 100644 --- a/drivers/gpu/drm/zte/zx_drm_drv.h +++ b/drivers/gpu/drm/zte/zx_drm_drv.h @@ -14,6 +14,7 @@ extern struct platform_driver zx_crtc_driver; extern struct platform_driver zx_hdmi_driver; extern struct platform_driver zx_tvenc_driver; +extern struct platform_driver zx_vga_driver; static inline u32 zx_readl(void __iomem *reg) { diff --git a/drivers/gpu/drm/zte/zx_plane.c b/drivers/gpu/drm/zte/zx_plane.c index d646ac931663..4a6252720c10 100644 --- a/drivers/gpu/drm/zte/zx_plane.c +++ b/drivers/gpu/drm/zte/zx_plane.c @@ -16,6 +16,7 @@ #include <drm/drm_plane_helper.h> #include <drm/drmP.h> +#include "zx_common_regs.h" #include "zx_drm_drv.h" #include "zx_plane.h" #include "zx_plane_regs.h" diff --git a/drivers/gpu/drm/zte/zx_plane_regs.h b/drivers/gpu/drm/zte/zx_plane_regs.h index 65f271aeabed..9c655f59f9f7 100644 --- a/drivers/gpu/drm/zte/zx_plane_regs.h +++ b/drivers/gpu/drm/zte/zx_plane_regs.h @@ -77,24 +77,6 @@ #define LUMA_STRIDE(x) (((x) << LUMA_STRIDE_SHIFT) & LUMA_STRIDE_MASK) #define CHROMA_STRIDE(x) (((x) << CHROMA_STRIDE_SHIFT) & CHROMA_STRIDE_MASK) -/* CSC registers */ -#define CSC_CTRL0 0x30 -#define CSC_COV_MODE_SHIFT 16 -#define CSC_COV_MODE_MASK (0xffff << CSC_COV_MODE_SHIFT) -#define CSC_BT601_IMAGE_RGB2YCBCR 0 -#define CSC_BT601_IMAGE_YCBCR2RGB 1 -#define CSC_BT601_VIDEO_RGB2YCBCR 2 -#define CSC_BT601_VIDEO_YCBCR2RGB 3 -#define CSC_BT709_IMAGE_RGB2YCBCR 4 -#define CSC_BT709_IMAGE_YCBCR2RGB 5 -#define CSC_BT709_VIDEO_RGB2YCBCR 6 -#define CSC_BT709_VIDEO_YCBCR2RGB 7 -#define CSC_BT2020_IMAGE_RGB2YCBCR 8 -#define CSC_BT2020_IMAGE_YCBCR2RGB 9 -#define CSC_BT2020_VIDEO_RGB2YCBCR 10 -#define CSC_BT2020_VIDEO_YCBCR2RGB 11 -#define CSC_WORK_ENABLE BIT(0) - /* RSZ registers */ #define RSZ_SRC_CFG 0x00 #define RSZ_DEST_CFG 0x04 diff --git a/drivers/gpu/drm/zte/zx_vga.c b/drivers/gpu/drm/zte/zx_vga.c new file mode 100644 index 000000000000..1e0811f775cb --- /dev/null +++ b/drivers/gpu/drm/zte/zx_vga.c @@ -0,0 +1,531 @@ +/* + * Copyright (C) 2017 Sanechips Technology Co., Ltd. + * Copyright 2017 Linaro Ltd. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ + +#include <linux/clk.h> +#include <linux/component.h> +#include <linux/mfd/syscon.h> +#include <linux/regmap.h> + +#include <drm/drm_atomic_helper.h> +#include <drm/drm_crtc_helper.h> +#include <drm/drmP.h> + +#include "zx_drm_drv.h" +#include "zx_vga_regs.h" +#include "zx_vou.h" + +struct zx_vga_pwrctrl { + struct regmap *regmap; + u32 reg; + u32 mask; +}; + +struct zx_vga_i2c { + struct i2c_adapter adap; + struct mutex lock; +}; + +struct zx_vga { + struct drm_connector connector; + struct drm_encoder encoder; + struct zx_vga_i2c *ddc; + struct device *dev; + void __iomem *mmio; + struct clk *i2c_wclk; + struct zx_vga_pwrctrl pwrctrl; + struct completion complete; + bool connected; +}; + +#define to_zx_vga(x) container_of(x, struct zx_vga, x) + +static void zx_vga_encoder_enable(struct drm_encoder *encoder) +{ + struct zx_vga *vga = to_zx_vga(encoder); + struct zx_vga_pwrctrl *pwrctrl = &vga->pwrctrl; + + /* Set bit to power up VGA DACs */ + regmap_update_bits(pwrctrl->regmap, pwrctrl->reg, pwrctrl->mask, + pwrctrl->mask); + + vou_inf_enable(VOU_VGA, encoder->crtc); +} + +static void zx_vga_encoder_disable(struct drm_encoder *encoder) +{ + struct zx_vga *vga = to_zx_vga(encoder); + struct zx_vga_pwrctrl *pwrctrl = &vga->pwrctrl; + + vou_inf_disable(VOU_VGA, encoder->crtc); + + /* Clear bit to power down VGA DACs */ + regmap_update_bits(pwrctrl->regmap, pwrctrl->reg, pwrctrl->mask, 0); +} + +static const struct drm_encoder_helper_funcs zx_vga_encoder_helper_funcs = { + .enable = zx_vga_encoder_enable, + .disable = zx_vga_encoder_disable, +}; + +static const struct drm_encoder_funcs zx_vga_encoder_funcs = { + .destroy = drm_encoder_cleanup, +}; + +static int zx_vga_connector_get_modes(struct drm_connector *connector) +{ + struct zx_vga *vga = to_zx_vga(connector); + struct edid *edid; + int ret; + + /* + * Clear both detection bits to switch I2C bus from device + * detecting to EDID reading. + */ + zx_writel(vga->mmio + VGA_AUTO_DETECT_SEL, 0); + + edid = drm_get_edid(connector, &vga->ddc->adap); + if (!edid) { + /* + * If EDID reading fails, we set the device state into + * disconnected. Locking is not required here, since the + * VGA_AUTO_DETECT_SEL register write in irq handler cannot + * be triggered when both detection bits are cleared as above. + */ + zx_writel(vga->mmio + VGA_AUTO_DETECT_SEL, + VGA_DETECT_SEL_NO_DEVICE); + vga->connected = false; + return 0; + } + + /* + * As edid reading succeeds, device must be connected, so we set + * up detection bit for unplug interrupt here. + */ + zx_writel(vga->mmio + VGA_AUTO_DETECT_SEL, VGA_DETECT_SEL_HAS_DEVICE); + + drm_mode_connector_update_edid_property(connector, edid); + ret = drm_add_edid_modes(connector, edid); + kfree(edid); + + return ret; +} + +static enum drm_mode_status +zx_vga_connector_mode_valid(struct drm_connector *connector, + struct drm_display_mode *mode) +{ + return MODE_OK; +} + +static struct drm_connector_helper_funcs zx_vga_connector_helper_funcs = { + .get_modes = zx_vga_connector_get_modes, + .mode_valid = zx_vga_connector_mode_valid, +}; + +static enum drm_connector_status +zx_vga_connector_detect(struct drm_connector *connector, bool force) +{ + struct zx_vga *vga = to_zx_vga(connector); + + return vga->connected ? connector_status_connected : + connector_status_disconnected; +} + +static const struct drm_connector_funcs zx_vga_connector_funcs = { + .dpms = drm_atomic_helper_connector_dpms, + .fill_modes = drm_helper_probe_single_connector_modes, + .detect = zx_vga_connector_detect, + .destroy = drm_connector_cleanup, + .reset = drm_atomic_helper_connector_reset, + .atomic_duplicate_state = drm_atomic_helper_connector_duplicate_state, + .atomic_destroy_state = drm_atomic_helper_connector_destroy_state, +}; + +static int zx_vga_register(struct drm_device *drm, struct zx_vga *vga) +{ + struct drm_encoder *encoder = &vga->encoder; + struct drm_connector *connector = &vga->connector; + struct device *dev = vga->dev; + int ret; + + encoder->possible_crtcs = VOU_CRTC_MASK; + + ret = drm_encoder_init(drm, encoder, &zx_vga_encoder_funcs, + DRM_MODE_ENCODER_DAC, NULL); + if (ret) { + DRM_DEV_ERROR(dev, "failed to init encoder: %d\n", ret); + return ret; + }; + + drm_encoder_helper_add(encoder, &zx_vga_encoder_helper_funcs); + + vga->connector.polled = DRM_CONNECTOR_POLL_HPD; + + ret = drm_connector_init(drm, connector, &zx_vga_connector_funcs, + DRM_MODE_CONNECTOR_VGA); + if (ret) { + DRM_DEV_ERROR(dev, "failed to init connector: %d\n", ret); + goto clean_encoder; + }; + + drm_connector_helper_add(connector, &zx_vga_connector_helper_funcs); + + ret = drm_mode_connector_attach_encoder(connector, encoder); + if (ret) { + DRM_DEV_ERROR(dev, "failed to attach encoder: %d\n", ret); + goto clean_connector; + }; + + return 0; + +clean_connector: + drm_connector_cleanup(connector); +clean_encoder: + drm_encoder_cleanup(encoder); + return ret; +} + +static int zx_vga_pwrctrl_init(struct zx_vga *vga) +{ + struct zx_vga_pwrctrl *pwrctrl = &vga->pwrctrl; + struct device *dev = vga->dev; + struct of_phandle_args out_args; + struct regmap *regmap; + int ret; + + ret = of_parse_phandle_with_fixed_args(dev->of_node, + "zte,vga-power-control", 2, 0, &out_args); + if (ret) + return ret; + + regmap = syscon_node_to_regmap(out_args.np); + if (IS_ERR(regmap)) { + ret = PTR_ERR(regmap); + goto out; + } + + pwrctrl->regmap = regmap; + pwrctrl->reg = out_args.args[0]; + pwrctrl->mask = out_args.args[1]; + +out: + of_node_put(out_args.np); + return ret; +} + +static int zx_vga_i2c_read(struct zx_vga *vga, struct i2c_msg *msg) +{ + int len = msg->len; + u8 *buf = msg->buf; + u32 offset = 0; + int i; + + reinit_completion(&vga->complete); + + /* Select combo write */ + zx_writel_mask(vga->mmio + VGA_CMD_CFG, VGA_CMD_COMBO, VGA_CMD_COMBO); + zx_writel_mask(vga->mmio + VGA_CMD_CFG, VGA_CMD_RW, 0); + + while (len > 0) { + u32 cnt; + + /* Clear RX FIFO */ + zx_writel_mask(vga->mmio + VGA_RXF_CTRL, VGA_RX_FIFO_CLEAR, + VGA_RX_FIFO_CLEAR); + + /* Data offset to read from */ + zx_writel(vga->mmio + VGA_SUB_ADDR, offset); + + /* Kick off the transfer */ + zx_writel_mask(vga->mmio + VGA_CMD_CFG, VGA_CMD_TRANS, + VGA_CMD_TRANS); + + if (!wait_for_completion_timeout(&vga->complete, + msecs_to_jiffies(1000))) { + DRM_DEV_ERROR(vga->dev, "transfer timeout\n"); + return -ETIMEDOUT; + } + + cnt = zx_readl(vga->mmio + VGA_RXF_STATUS); + cnt = (cnt & VGA_RXF_COUNT_MASK) >> VGA_RXF_COUNT_SHIFT; + /* FIFO status may report more data than we need to read */ + cnt = min_t(u32, len, cnt); + + for (i = 0; i < cnt; i++) + *buf++ = zx_readl(vga->mmio + VGA_DATA); + + len -= cnt; + offset += cnt; + } + + return 0; +} + +static int zx_vga_i2c_write(struct zx_vga *vga, struct i2c_msg *msg) +{ + /* + * The DDC I2C adapter is only for reading EDID data, so we assume + * that the write to this adapter must be the EDID data offset. + */ + if ((msg->len != 1) || ((msg->addr != DDC_ADDR))) + return -EINVAL; + + /* Hardware will take care of the slave address shifting */ + zx_writel(vga->mmio + VGA_DEVICE_ADDR, msg->addr); + + return 0; +} + +static int zx_vga_i2c_xfer(struct i2c_adapter *adap, struct i2c_msg *msgs, + int num) +{ + struct zx_vga *vga = i2c_get_adapdata(adap); + struct zx_vga_i2c *ddc = vga->ddc; + int ret = 0; + int i; + + mutex_lock(&ddc->lock); + + for (i = 0; i < num; i++) { + if (msgs[i].flags & I2C_M_RD) + ret = zx_vga_i2c_read(vga, &msgs[i]); + else + ret = zx_vga_i2c_write(vga, &msgs[i]); + + if (ret < 0) + break; + } + + if (!ret) + ret = num; + + mutex_unlock(&ddc->lock); + + return ret; +} + +static u32 zx_vga_i2c_func(struct i2c_adapter *adapter) +{ + return I2C_FUNC_I2C | I2C_FUNC_SMBUS_EMUL; +} + +static const struct i2c_algorithm zx_vga_algorithm = { + .master_xfer = zx_vga_i2c_xfer, + .functionality = zx_vga_i2c_func, +}; + +static int zx_vga_ddc_register(struct zx_vga *vga) +{ + struct device *dev = vga->dev; + struct i2c_adapter *adap; + struct zx_vga_i2c *ddc; + int ret; + + ddc = devm_kzalloc(dev, sizeof(*ddc), GFP_KERNEL); + if (!ddc) + return -ENOMEM; + + vga->ddc = ddc; + mutex_init(&ddc->lock); + + adap = &ddc->adap; + adap->owner = THIS_MODULE; + adap->class = I2C_CLASS_DDC; + adap->dev.parent = dev; + adap->algo = &zx_vga_algorithm; + snprintf(adap->name, sizeof(adap->name), "zx vga i2c"); + + ret = i2c_add_adapter(adap); + if (ret) { + DRM_DEV_ERROR(dev, "failed to add I2C adapter: %d\n", ret); + return ret; + } + + i2c_set_adapdata(adap, vga); + + return 0; +} + +static irqreturn_t zx_vga_irq_thread(int irq, void *dev_id) +{ + struct zx_vga *vga = dev_id; + + drm_helper_hpd_irq_event(vga->connector.dev); + + return IRQ_HANDLED; +} + +static irqreturn_t zx_vga_irq_handler(int irq, void *dev_id) +{ + struct zx_vga *vga = dev_id; + u32 status; + + status = zx_readl(vga->mmio + VGA_I2C_STATUS); + + /* Clear interrupt status */ + zx_writel_mask(vga->mmio + VGA_I2C_STATUS, VGA_CLEAR_IRQ, + VGA_CLEAR_IRQ); + + if (status & VGA_DEVICE_CONNECTED) { + /* + * Since VGA_DETECT_SEL bits need to be reset for switching DDC + * bus from device detection to EDID read, rather than setting + * up HAS_DEVICE bit here, we need to do that in .get_modes + * hook for unplug detecting after EDID read succeeds. + */ + vga->connected = true; + return IRQ_WAKE_THREAD; + } + + if (status & VGA_DEVICE_DISCONNECTED) { + zx_writel(vga->mmio + VGA_AUTO_DETECT_SEL, + VGA_DETECT_SEL_NO_DEVICE); + vga->connected = false; + return IRQ_WAKE_THREAD; + } + + if (status & VGA_TRANS_DONE) { + complete(&vga->complete); + return IRQ_HANDLED; + } + + return IRQ_NONE; +} + +static void zx_vga_hw_init(struct zx_vga *vga) +{ + unsigned long ref = clk_get_rate(vga->i2c_wclk); + int div; + + /* + * Set up I2C fast speed divider per formula below to get 400kHz. + * scl = ref / ((div + 1) * 4) + */ + div = DIV_ROUND_UP(ref / 1000, 400 * 4) - 1; + zx_writel(vga->mmio + VGA_CLK_DIV_FS, div); + + /* Set up device detection */ + zx_writel(vga->mmio + VGA_AUTO_DETECT_PARA, 0x80); + zx_writel(vga->mmio + VGA_AUTO_DETECT_SEL, VGA_DETECT_SEL_NO_DEVICE); + + /* + * We need to poke monitor via DDC bus to get connection irq + * start working. + */ + zx_writel(vga->mmio + VGA_DEVICE_ADDR, DDC_ADDR); + zx_writel_mask(vga->mmio + VGA_CMD_CFG, VGA_CMD_TRANS, VGA_CMD_TRANS); +} + +static int zx_vga_bind(struct device *dev, struct device *master, void *data) +{ + struct platform_device *pdev = to_platform_device(dev); + struct drm_device *drm = data; + struct resource *res; + struct zx_vga *vga; + int irq; + int ret; + + vga = devm_kzalloc(dev, sizeof(*vga), GFP_KERNEL); + if (!vga) + return -ENOMEM; + + vga->dev = dev; + dev_set_drvdata(dev, vga); + + res = platform_get_resource(pdev, IORESOURCE_MEM, 0); + vga->mmio = devm_ioremap_resource(dev, res); + if (IS_ERR(vga->mmio)) + return PTR_ERR(vga->mmio); + + irq = platform_get_irq(pdev, 0); + if (irq < 0) + return irq; + + vga->i2c_wclk = devm_clk_get(dev, "i2c_wclk"); + if (IS_ERR(vga->i2c_wclk)) { + ret = PTR_ERR(vga->i2c_wclk); + DRM_DEV_ERROR(dev, "failed to get i2c_wclk: %d\n", ret); + return ret; + } + + ret = zx_vga_pwrctrl_init(vga); + if (ret) { + DRM_DEV_ERROR(dev, "failed to init power control: %d\n", ret); + return ret; + } + + ret = zx_vga_ddc_register(vga); + if (ret) { + DRM_DEV_ERROR(dev, "failed to register ddc: %d\n", ret); + return ret; + } + + ret = zx_vga_register(drm, vga); + if (ret) { + DRM_DEV_ERROR(dev, "failed to register vga: %d\n", ret); + return ret; + } + + init_completion(&vga->complete); + + ret = devm_request_threaded_irq(dev, irq, zx_vga_irq_handler, + zx_vga_irq_thread, IRQF_SHARED, + dev_name(dev), vga); + if (ret) { + DRM_DEV_ERROR(dev, "failed to request threaded irq: %d\n", ret); + return ret; + } + + ret = clk_prepare_enable(vga->i2c_wclk); + if (ret) + return ret; + + zx_vga_hw_init(vga); + + return 0; +} + +static void zx_vga_unbind(struct device *dev, struct device *master, + void *data) +{ + struct zx_vga *vga = dev_get_drvdata(dev); + + clk_disable_unprepare(vga->i2c_wclk); +} + +static const struct component_ops zx_vga_component_ops = { + .bind = zx_vga_bind, + .unbind = zx_vga_unbind, +}; + +static int zx_vga_probe(struct platform_device *pdev) +{ + return component_add(&pdev->dev, &zx_vga_component_ops); +} + +static int zx_vga_remove(struct platform_device *pdev) +{ + component_del(&pdev->dev, &zx_vga_component_ops); + return 0; +} + +static const struct of_device_id zx_vga_of_match[] = { + { .compatible = "zte,zx296718-vga", }, + { /* end */ }, +}; +MODULE_DEVICE_TABLE(of, zx_vga_of_match); + +struct platform_driver zx_vga_driver = { + .probe = zx_vga_probe, + .remove = zx_vga_remove, + .driver = { + .name = "zx-vga", + .of_match_table = zx_vga_of_match, + }, +}; diff --git a/drivers/gpu/drm/zte/zx_vga_regs.h b/drivers/gpu/drm/zte/zx_vga_regs.h new file mode 100644 index 000000000000..feaa345fe6a6 --- /dev/null +++ b/drivers/gpu/drm/zte/zx_vga_regs.h @@ -0,0 +1,36 @@ +/* + * Copyright (C) 2017 Sanechips Technology Co., Ltd. + * Copyright 2017 Linaro Ltd. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ + +#ifndef __ZX_VGA_REGS_H__ +#define __ZX_VGA_REGS_H__ + +#define VGA_CMD_CFG 0x04 +#define VGA_CMD_TRANS BIT(6) +#define VGA_CMD_COMBO BIT(5) +#define VGA_CMD_RW BIT(4) +#define VGA_SUB_ADDR 0x0c +#define VGA_DEVICE_ADDR 0x10 +#define VGA_CLK_DIV_FS 0x14 +#define VGA_RXF_CTRL 0x20 +#define VGA_RX_FIFO_CLEAR BIT(7) +#define VGA_DATA 0x24 +#define VGA_I2C_STATUS 0x28 +#define VGA_DEVICE_DISCONNECTED BIT(7) +#define VGA_DEVICE_CONNECTED BIT(6) +#define VGA_CLEAR_IRQ BIT(4) +#define VGA_TRANS_DONE BIT(0) +#define VGA_RXF_STATUS 0x30 +#define VGA_RXF_COUNT_SHIFT 2 +#define VGA_RXF_COUNT_MASK GENMASK(7, 2) +#define VGA_AUTO_DETECT_PARA 0x34 +#define VGA_AUTO_DETECT_SEL 0x38 +#define VGA_DETECT_SEL_HAS_DEVICE BIT(1) +#define VGA_DETECT_SEL_NO_DEVICE BIT(0) + +#endif /* __ZX_VGA_REGS_H__ */ diff --git a/drivers/gpu/drm/zte/zx_vou.c b/drivers/gpu/drm/zte/zx_vou.c index b500c8dd0d9d..5fbd10b60ee5 100644 --- a/drivers/gpu/drm/zte/zx_vou.c +++ b/drivers/gpu/drm/zte/zx_vou.c @@ -23,6 +23,7 @@ #include <drm/drm_plane_helper.h> #include <drm/drmP.h> +#include "zx_common_regs.h" #include "zx_drm_drv.h" #include "zx_plane.h" #include "zx_vou.h" @@ -122,6 +123,8 @@ struct zx_crtc { struct drm_plane *primary; struct zx_vou_hw *vou; void __iomem *chnreg; + void __iomem *chncsc; + void __iomem *dither; const struct zx_crtc_regs *regs; const struct zx_crtc_bits *bits; enum vou_chn_type chn_type; @@ -204,6 +207,11 @@ static struct vou_inf vou_infs[] = { .clocks_en_bits = BIT(15), .clocks_sel_bits = BIT(11) | BIT(0), }, + [VOU_VGA] = { + .data_sel = VOU_RGB_888, + .clocks_en_bits = BIT(1), + .clocks_sel_bits = BIT(10), + }, }; static inline struct zx_vou_hw *crtc_to_vou(struct drm_crtc *crtc) @@ -227,9 +235,26 @@ void vou_inf_enable(enum vou_inf_id id, struct drm_crtc *crtc) struct zx_crtc *zcrtc = to_zx_crtc(crtc); struct zx_vou_hw *vou = zcrtc->vou; struct vou_inf *inf = &vou_infs[id]; + void __iomem *dither = zcrtc->dither; + void __iomem *csc = zcrtc->chncsc; bool is_main = zcrtc->chn_type == VOU_CHN_MAIN; u32 data_sel_shift = id << 1; + if (inf->data_sel != VOU_YUV444) { + /* Enable channel CSC for RGB output */ + zx_writel_mask(csc + CSC_CTRL0, CSC_COV_MODE_MASK, + CSC_BT709_IMAGE_YCBCR2RGB << CSC_COV_MODE_SHIFT); + zx_writel_mask(csc + CSC_CTRL0, CSC_WORK_ENABLE, + CSC_WORK_ENABLE); + + /* Bypass Dither block for RGB output */ + zx_writel_mask(dither + OSD_DITHER_CTRL0, DITHER_BYSPASS, + DITHER_BYSPASS); + } else { + zx_writel_mask(csc + CSC_CTRL0, CSC_WORK_ENABLE, 0); + zx_writel_mask(dither + OSD_DITHER_CTRL0, DITHER_BYSPASS, 0); + } + /* Select data format */ zx_writel_mask(vou->vouctl + VOU_INF_DATA_SEL, 0x3 << data_sel_shift, inf->data_sel << data_sel_shift); @@ -525,20 +550,24 @@ static int zx_crtc_init(struct drm_device *drm, struct zx_vou_hw *vou, if (chn_type == VOU_CHN_MAIN) { zplane->layer = vou->osd + MAIN_GL_OFFSET; - zplane->csc = vou->osd + MAIN_CSC_OFFSET; + zplane->csc = vou->osd + MAIN_GL_CSC_OFFSET; zplane->hbsc = vou->osd + MAIN_HBSC_OFFSET; zplane->rsz = vou->otfppu + MAIN_RSZ_OFFSET; zplane->bits = &zx_gl_bits[0]; zcrtc->chnreg = vou->osd + OSD_MAIN_CHN; + zcrtc->chncsc = vou->osd + MAIN_CHN_CSC_OFFSET; + zcrtc->dither = vou->osd + MAIN_DITHER_OFFSET; zcrtc->regs = &main_crtc_regs; zcrtc->bits = &main_crtc_bits; } else { zplane->layer = vou->osd + AUX_GL_OFFSET; - zplane->csc = vou->osd + AUX_CSC_OFFSET; + zplane->csc = vou->osd + AUX_GL_CSC_OFFSET; zplane->hbsc = vou->osd + AUX_HBSC_OFFSET; zplane->rsz = vou->otfppu + AUX_RSZ_OFFSET; zplane->bits = &zx_gl_bits[1]; zcrtc->chnreg = vou->osd + OSD_AUX_CHN; + zcrtc->chncsc = vou->osd + AUX_CHN_CSC_OFFSET; + zcrtc->dither = vou->osd + AUX_DITHER_OFFSET; zcrtc->regs = &aux_crtc_regs; zcrtc->bits = &aux_crtc_bits; } @@ -705,9 +734,6 @@ static void vou_hw_init(struct zx_vou_hw *vou) /* Release reset for all VOU modules */ zx_writel(vou->vouctl + VOU_SOFT_RST, ~0); - /* Enable clock auto-gating for all VOU modules */ - zx_writel(vou->vouctl + VOU_CLK_REQEN, ~0); - /* Enable all VOU module clocks */ zx_writel(vou->vouctl + VOU_CLK_EN, ~0); diff --git a/drivers/gpu/drm/zte/zx_vou_regs.h b/drivers/gpu/drm/zte/zx_vou_regs.h index c066ef123434..5a218351b497 100644 --- a/drivers/gpu/drm/zte/zx_vou_regs.h +++ b/drivers/gpu/drm/zte/zx_vou_regs.h @@ -13,13 +13,17 @@ /* Sub-module offset */ #define MAIN_GL_OFFSET 0x130 -#define MAIN_CSC_OFFSET 0x580 +#define MAIN_GL_CSC_OFFSET 0x580 +#define MAIN_CHN_CSC_OFFSET 0x6c0 #define MAIN_HBSC_OFFSET 0x820 +#define MAIN_DITHER_OFFSET 0x960 #define MAIN_RSZ_OFFSET 0x600 /* OTFPPU sub-module */ #define AUX_GL_OFFSET 0x200 -#define AUX_CSC_OFFSET 0x5d0 +#define AUX_GL_CSC_OFFSET 0x5d0 +#define AUX_CHN_CSC_OFFSET 0x710 #define AUX_HBSC_OFFSET 0x860 +#define AUX_DITHER_OFFSET 0x970 #define AUX_RSZ_OFFSET 0x800 #define OSD_VL0_OFFSET 0x040 @@ -78,6 +82,10 @@ #define CHN_INTERLACE_BUF_CTRL 0x24 #define CHN_INTERLACE_EN BIT(2) +/* Dither registers */ +#define OSD_DITHER_CTRL0 0x00 +#define DITHER_BYSPASS BIT(31) + /* TIMING_CTRL registers */ #define TIMING_TC_ENABLE 0x04 #define AUX_TC_EN BIT(1) diff --git a/drivers/gpu/host1x/Kconfig b/drivers/gpu/host1x/Kconfig index b2fd029d67b3..91916326957f 100644 --- a/drivers/gpu/host1x/Kconfig +++ b/drivers/gpu/host1x/Kconfig @@ -1,6 +1,7 @@ config TEGRA_HOST1X tristate "NVIDIA Tegra host1x driver" depends on ARCH_TEGRA || (ARM && COMPILE_TEST) + select IOMMU_IOVA if IOMMU_SUPPORT help Driver for the NVIDIA Tegra host1x hardware. diff --git a/drivers/gpu/host1x/bus.c b/drivers/gpu/host1x/bus.c index eeb021fe6410..561831e1ae2c 100644 --- a/drivers/gpu/host1x/bus.c +++ b/drivers/gpu/host1x/bus.c @@ -267,37 +267,6 @@ static int host1x_device_match(struct device *dev, struct device_driver *drv) return strcmp(dev_name(dev), drv->name) == 0; } -static int host1x_device_probe(struct device *dev) -{ - struct host1x_driver *driver = to_host1x_driver(dev->driver); - struct host1x_device *device = to_host1x_device(dev); - - if (driver->probe) - return driver->probe(device); - - return 0; -} - -static int host1x_device_remove(struct device *dev) -{ - struct host1x_driver *driver = to_host1x_driver(dev->driver); - struct host1x_device *device = to_host1x_device(dev); - - if (driver->remove) - return driver->remove(device); - - return 0; -} - -static void host1x_device_shutdown(struct device *dev) -{ - struct host1x_driver *driver = to_host1x_driver(dev->driver); - struct host1x_device *device = to_host1x_device(dev); - - if (driver->shutdown) - driver->shutdown(device); -} - static const struct dev_pm_ops host1x_device_pm_ops = { .suspend = pm_generic_suspend, .resume = pm_generic_resume, @@ -310,9 +279,6 @@ static const struct dev_pm_ops host1x_device_pm_ops = { struct bus_type host1x_bus_type = { .name = "host1x", .match = host1x_device_match, - .probe = host1x_device_probe, - .remove = host1x_device_remove, - .shutdown = host1x_device_shutdown, .pm = &host1x_device_pm_ops, }; @@ -516,6 +482,37 @@ int host1x_unregister(struct host1x *host1x) return 0; } +static int host1x_device_probe(struct device *dev) +{ + struct host1x_driver *driver = to_host1x_driver(dev->driver); + struct host1x_device *device = to_host1x_device(dev); + + if (driver->probe) + return driver->probe(device); + + return 0; +} + +static int host1x_device_remove(struct device *dev) +{ + struct host1x_driver *driver = to_host1x_driver(dev->driver); + struct host1x_device *device = to_host1x_device(dev); + + if (driver->remove) + return driver->remove(device); + + return 0; +} + +static void host1x_device_shutdown(struct device *dev) +{ + struct host1x_driver *driver = to_host1x_driver(dev->driver); + struct host1x_device *device = to_host1x_device(dev); + + if (driver->shutdown) + driver->shutdown(device); +} + int host1x_driver_register_full(struct host1x_driver *driver, struct module *owner) { @@ -536,6 +533,9 @@ int host1x_driver_register_full(struct host1x_driver *driver, driver->driver.bus = &host1x_bus_type; driver->driver.owner = owner; + driver->driver.probe = host1x_device_probe; + driver->driver.remove = host1x_device_remove; + driver->driver.shutdown = host1x_device_shutdown; return driver_register(&driver->driver); } diff --git a/drivers/gpu/host1x/cdma.c b/drivers/gpu/host1x/cdma.c index c5d82a8a2ec9..28541b280739 100644 --- a/drivers/gpu/host1x/cdma.c +++ b/drivers/gpu/host1x/cdma.c @@ -51,9 +51,15 @@ static void host1x_pushbuffer_destroy(struct push_buffer *pb) struct host1x_cdma *cdma = pb_to_cdma(pb); struct host1x *host1x = cdma_to_host1x(cdma); - if (pb->phys != 0) - dma_free_wc(host1x->dev, pb->size_bytes + 4, pb->mapped, - pb->phys); + if (!pb->phys) + return; + + if (host1x->domain) { + iommu_unmap(host1x->domain, pb->dma, pb->alloc_size); + free_iova(&host1x->iova, iova_pfn(&host1x->iova, pb->dma)); + } + + dma_free_wc(host1x->dev, pb->alloc_size, pb->mapped, pb->phys); pb->mapped = NULL; pb->phys = 0; @@ -66,28 +72,64 @@ static int host1x_pushbuffer_init(struct push_buffer *pb) { struct host1x_cdma *cdma = pb_to_cdma(pb); struct host1x *host1x = cdma_to_host1x(cdma); + struct iova *alloc; + u32 size; + int err; pb->mapped = NULL; pb->phys = 0; - pb->size_bytes = HOST1X_PUSHBUFFER_SLOTS * 8; + pb->size = HOST1X_PUSHBUFFER_SLOTS * 8; + + size = pb->size + 4; /* initialize buffer pointers */ - pb->fence = pb->size_bytes - 8; + pb->fence = pb->size - 8; pb->pos = 0; - /* allocate and map pushbuffer memory */ - pb->mapped = dma_alloc_wc(host1x->dev, pb->size_bytes + 4, &pb->phys, - GFP_KERNEL); - if (!pb->mapped) - goto fail; + if (host1x->domain) { + unsigned long shift; + + size = iova_align(&host1x->iova, size); + + pb->mapped = dma_alloc_wc(host1x->dev, size, &pb->phys, + GFP_KERNEL); + if (!pb->mapped) + return -ENOMEM; + + shift = iova_shift(&host1x->iova); + alloc = alloc_iova(&host1x->iova, size >> shift, + host1x->iova_end >> shift, true); + if (!alloc) { + err = -ENOMEM; + goto iommu_free_mem; + } + + pb->dma = iova_dma_addr(&host1x->iova, alloc); + err = iommu_map(host1x->domain, pb->dma, pb->phys, size, + IOMMU_READ); + if (err) + goto iommu_free_iova; + } else { + pb->mapped = dma_alloc_wc(host1x->dev, size, &pb->phys, + GFP_KERNEL); + if (!pb->mapped) + return -ENOMEM; + + pb->dma = pb->phys; + } + + pb->alloc_size = size; host1x_hw_pushbuffer_init(host1x, pb); return 0; -fail: - host1x_pushbuffer_destroy(pb); - return -ENOMEM; +iommu_free_iova: + __free_iova(&host1x->iova, alloc); +iommu_free_mem: + dma_free_wc(host1x->dev, pb->alloc_size, pb->mapped, pb->phys); + + return err; } /* @@ -101,7 +143,7 @@ static void host1x_pushbuffer_push(struct push_buffer *pb, u32 op1, u32 op2) WARN_ON(pb->pos == pb->fence); *(p++) = op1; *(p++) = op2; - pb->pos = (pb->pos + 8) & (pb->size_bytes - 1); + pb->pos = (pb->pos + 8) & (pb->size - 1); } /* @@ -111,7 +153,7 @@ static void host1x_pushbuffer_push(struct push_buffer *pb, u32 op1, u32 op2) static void host1x_pushbuffer_pop(struct push_buffer *pb, unsigned int slots) { /* Advance the next write position */ - pb->fence = (pb->fence + slots * 8) & (pb->size_bytes - 1); + pb->fence = (pb->fence + slots * 8) & (pb->size - 1); } /* @@ -119,7 +161,7 @@ static void host1x_pushbuffer_pop(struct push_buffer *pb, unsigned int slots) */ static u32 host1x_pushbuffer_space(struct push_buffer *pb) { - return ((pb->fence - pb->pos) & (pb->size_bytes - 1)) / 8; + return ((pb->fence - pb->pos) & (pb->size - 1)) / 8; } /* diff --git a/drivers/gpu/host1x/cdma.h b/drivers/gpu/host1x/cdma.h index 470087af8fe5..ec170a78f4e1 100644 --- a/drivers/gpu/host1x/cdma.h +++ b/drivers/gpu/host1x/cdma.h @@ -43,10 +43,12 @@ struct host1x_job; struct push_buffer { void *mapped; /* mapped pushbuffer memory */ - dma_addr_t phys; /* physical address of pushbuffer */ + dma_addr_t dma; /* device address of pushbuffer */ + phys_addr_t phys; /* physical address of pushbuffer */ u32 fence; /* index we've written */ u32 pos; /* index to write to */ - u32 size_bytes; + u32 size; + u32 alloc_size; }; struct buffer_timeout { diff --git a/drivers/gpu/host1x/dev.c b/drivers/gpu/host1x/dev.c index a62317af76ad..f05ebb14fa63 100644 --- a/drivers/gpu/host1x/dev.c +++ b/drivers/gpu/host1x/dev.c @@ -16,23 +16,25 @@ * along with this program. If not, see <http://www.gnu.org/licenses/>. */ -#include <linux/module.h> -#include <linux/list.h> -#include <linux/slab.h> -#include <linux/of.h> -#include <linux/of_device.h> #include <linux/clk.h> -#include <linux/io.h> #include <linux/dma-mapping.h> +#include <linux/io.h> +#include <linux/list.h> +#include <linux/module.h> +#include <linux/of_device.h> +#include <linux/of.h> +#include <linux/slab.h> #define CREATE_TRACE_POINTS #include <trace/events/host1x.h> +#undef CREATE_TRACE_POINTS #include "bus.h" -#include "dev.h" -#include "intr.h" #include "channel.h" #include "debug.h" +#include "dev.h" +#include "intr.h" + #include "hw/host1x01.h" #include "hw/host1x02.h" #include "hw/host1x04.h" @@ -168,22 +170,56 @@ static int host1x_probe(struct platform_device *pdev) return err; } + host->rst = devm_reset_control_get(&pdev->dev, "host1x"); + if (IS_ERR(host->rst)) { + err = PTR_ERR(host->clk); + dev_err(&pdev->dev, "failed to get reset: %d\n", err); + return err; + } + + if (iommu_present(&platform_bus_type)) { + struct iommu_domain_geometry *geometry; + unsigned long order; + + host->domain = iommu_domain_alloc(&platform_bus_type); + if (!host->domain) + return -ENOMEM; + + err = iommu_attach_device(host->domain, &pdev->dev); + if (err) + goto fail_free_domain; + + geometry = &host->domain->geometry; + + order = __ffs(host->domain->pgsize_bitmap); + init_iova_domain(&host->iova, 1UL << order, + geometry->aperture_start >> order, + geometry->aperture_end >> order); + host->iova_end = geometry->aperture_end; + } + err = host1x_channel_list_init(host); if (err) { dev_err(&pdev->dev, "failed to initialize channel list\n"); - return err; + goto fail_detach_device; } err = clk_prepare_enable(host->clk); if (err < 0) { dev_err(&pdev->dev, "failed to enable clock\n"); - return err; + goto fail_detach_device; + } + + err = reset_control_deassert(host->rst); + if (err < 0) { + dev_err(&pdev->dev, "failed to deassert reset: %d\n", err); + goto fail_unprepare_disable; } err = host1x_syncpt_init(host); if (err) { dev_err(&pdev->dev, "failed to initialize syncpts\n"); - goto fail_unprepare_disable; + goto fail_reset_assert; } err = host1x_intr_init(host, syncpt_irq); @@ -204,8 +240,19 @@ fail_deinit_intr: host1x_intr_deinit(host); fail_deinit_syncpt: host1x_syncpt_deinit(host); +fail_reset_assert: + reset_control_assert(host->rst); fail_unprepare_disable: clk_disable_unprepare(host->clk); +fail_detach_device: + if (host->domain) { + put_iova_domain(&host->iova); + iommu_detach_device(host->domain, &pdev->dev); + } +fail_free_domain: + if (host->domain) + iommu_domain_free(host->domain); + return err; } @@ -216,8 +263,15 @@ static int host1x_remove(struct platform_device *pdev) host1x_unregister(host); host1x_intr_deinit(host); host1x_syncpt_deinit(host); + reset_control_assert(host->rst); clk_disable_unprepare(host->clk); + if (host->domain) { + put_iova_domain(&host->iova); + iommu_detach_device(host->domain, &pdev->dev); + iommu_domain_free(host->domain); + } + return 0; } diff --git a/drivers/gpu/host1x/dev.h b/drivers/gpu/host1x/dev.h index 06dd4f85125f..229d08b6a45e 100644 --- a/drivers/gpu/host1x/dev.h +++ b/drivers/gpu/host1x/dev.h @@ -17,14 +17,17 @@ #ifndef HOST1X_DEV_H #define HOST1X_DEV_H -#include <linux/platform_device.h> #include <linux/device.h> +#include <linux/iommu.h> +#include <linux/iova.h> +#include <linux/platform_device.h> +#include <linux/reset.h> +#include "cdma.h" #include "channel.h" -#include "syncpt.h" #include "intr.h" -#include "cdma.h" #include "job.h" +#include "syncpt.h" struct host1x_syncpt; struct host1x_syncpt_base; @@ -107,6 +110,11 @@ struct host1x { struct host1x_syncpt_base *bases; struct device *dev; struct clk *clk; + struct reset_control *rst; + + struct iommu_domain *domain; + struct iova_domain iova; + dma_addr_t iova_end; struct mutex intr_mutex; int intr_syncpt_irq; diff --git a/drivers/gpu/host1x/hw/cdma_hw.c b/drivers/gpu/host1x/hw/cdma_hw.c index 659c1bbfeeba..6b231119193e 100644 --- a/drivers/gpu/host1x/hw/cdma_hw.c +++ b/drivers/gpu/host1x/hw/cdma_hw.c @@ -30,7 +30,7 @@ */ static void push_buffer_init(struct push_buffer *pb) { - *(u32 *)(pb->mapped + pb->size_bytes) = host1x_opcode_restart(0); + *(u32 *)(pb->mapped + pb->size) = host1x_opcode_restart(0); } /* @@ -55,8 +55,8 @@ static void cdma_timeout_cpu_incr(struct host1x_cdma *cdma, u32 getptr, *(p++) = HOST1X_OPCODE_NOP; *(p++) = HOST1X_OPCODE_NOP; dev_dbg(host1x->dev, "%s: NOP at %pad+%#x\n", __func__, - &pb->phys, getptr); - getptr = (getptr + 8) & (pb->size_bytes - 1); + &pb->dma, getptr); + getptr = (getptr + 8) & (pb->size - 1); } wmb(); @@ -78,10 +78,9 @@ static void cdma_start(struct host1x_cdma *cdma) HOST1X_CHANNEL_DMACTRL); /* set base, put and end pointer */ - host1x_ch_writel(ch, cdma->push_buffer.phys, HOST1X_CHANNEL_DMASTART); + host1x_ch_writel(ch, cdma->push_buffer.dma, HOST1X_CHANNEL_DMASTART); host1x_ch_writel(ch, cdma->push_buffer.pos, HOST1X_CHANNEL_DMAPUT); - host1x_ch_writel(ch, cdma->push_buffer.phys + - cdma->push_buffer.size_bytes + 4, + host1x_ch_writel(ch, cdma->push_buffer.dma + cdma->push_buffer.size + 4, HOST1X_CHANNEL_DMAEND); /* reset GET */ @@ -115,9 +114,8 @@ static void cdma_timeout_restart(struct host1x_cdma *cdma, u32 getptr) HOST1X_CHANNEL_DMACTRL); /* set base, end pointer (all of memory) */ - host1x_ch_writel(ch, cdma->push_buffer.phys, HOST1X_CHANNEL_DMASTART); - host1x_ch_writel(ch, cdma->push_buffer.phys + - cdma->push_buffer.size_bytes, + host1x_ch_writel(ch, cdma->push_buffer.dma, HOST1X_CHANNEL_DMASTART); + host1x_ch_writel(ch, cdma->push_buffer.dma + cdma->push_buffer.size, HOST1X_CHANNEL_DMAEND); /* set GET, by loading the value in PUT (then reset GET) */ diff --git a/drivers/gpu/host1x/job.c b/drivers/gpu/host1x/job.c index 92c3df933303..5f5f8ee6143d 100644 --- a/drivers/gpu/host1x/job.c +++ b/drivers/gpu/host1x/job.c @@ -174,9 +174,10 @@ static int do_waitchks(struct host1x_job *job, struct host1x *host, return 0; } -static unsigned int pin_job(struct host1x_job *job) +static unsigned int pin_job(struct host1x *host, struct host1x_job *job) { unsigned int i; + int err; job->num_unpins = 0; @@ -186,12 +187,16 @@ static unsigned int pin_job(struct host1x_job *job) dma_addr_t phys_addr; reloc->target.bo = host1x_bo_get(reloc->target.bo); - if (!reloc->target.bo) + if (!reloc->target.bo) { + err = -EINVAL; goto unpin; + } phys_addr = host1x_bo_pin(reloc->target.bo, &sgt); - if (!phys_addr) + if (!phys_addr) { + err = -EINVAL; goto unpin; + } job->addr_phys[job->num_unpins] = phys_addr; job->unpins[job->num_unpins].bo = reloc->target.bo; @@ -201,28 +206,67 @@ static unsigned int pin_job(struct host1x_job *job) for (i = 0; i < job->num_gathers; i++) { struct host1x_job_gather *g = &job->gathers[i]; + size_t gather_size = 0; + struct scatterlist *sg; struct sg_table *sgt; dma_addr_t phys_addr; + unsigned long shift; + struct iova *alloc; + unsigned int j; g->bo = host1x_bo_get(g->bo); - if (!g->bo) + if (!g->bo) { + err = -EINVAL; goto unpin; + } phys_addr = host1x_bo_pin(g->bo, &sgt); - if (!phys_addr) + if (!phys_addr) { + err = -EINVAL; goto unpin; + } + + if (!IS_ENABLED(CONFIG_TEGRA_HOST1X_FIREWALL) && host->domain) { + for_each_sg(sgt->sgl, sg, sgt->nents, j) + gather_size += sg->length; + gather_size = iova_align(&host->iova, gather_size); + + shift = iova_shift(&host->iova); + alloc = alloc_iova(&host->iova, gather_size >> shift, + host->iova_end >> shift, true); + if (!alloc) { + err = -ENOMEM; + goto unpin; + } + + err = iommu_map_sg(host->domain, + iova_dma_addr(&host->iova, alloc), + sgt->sgl, sgt->nents, IOMMU_READ); + if (err == 0) { + __free_iova(&host->iova, alloc); + err = -EINVAL; + goto unpin; + } + + job->addr_phys[job->num_unpins] = + iova_dma_addr(&host->iova, alloc); + job->unpins[job->num_unpins].size = gather_size; + } else { + job->addr_phys[job->num_unpins] = phys_addr; + } + + job->gather_addr_phys[i] = job->addr_phys[job->num_unpins]; - job->addr_phys[job->num_unpins] = phys_addr; job->unpins[job->num_unpins].bo = g->bo; job->unpins[job->num_unpins].sgt = sgt; job->num_unpins++; } - return job->num_unpins; + return 0; unpin: host1x_job_unpin(job); - return 0; + return err; } static int do_relocs(struct host1x_job *job, struct host1x_bo *cmdbuf) @@ -525,8 +569,8 @@ int host1x_job_pin(struct host1x_job *job, struct device *dev) host1x_syncpt_load(host->syncpt + i); /* pin memory */ - err = pin_job(job); - if (!err) + err = pin_job(host, job); + if (err) goto out; /* patch gathers */ @@ -572,11 +616,19 @@ EXPORT_SYMBOL(host1x_job_pin); void host1x_job_unpin(struct host1x_job *job) { + struct host1x *host = dev_get_drvdata(job->channel->dev->parent); unsigned int i; for (i = 0; i < job->num_unpins; i++) { struct host1x_job_unpin_data *unpin = &job->unpins[i]; + if (!IS_ENABLED(CONFIG_TEGRA_HOST1X_FIREWALL) && host->domain) { + iommu_unmap(host->domain, job->addr_phys[i], + unpin->size); + free_iova(&host->iova, + iova_pfn(&host->iova, job->addr_phys[i])); + } + host1x_bo_unpin(unpin->bo, unpin->sgt); host1x_bo_put(unpin->bo); } diff --git a/drivers/gpu/host1x/job.h b/drivers/gpu/host1x/job.h index 8b3c15df0660..878239c476d2 100644 --- a/drivers/gpu/host1x/job.h +++ b/drivers/gpu/host1x/job.h @@ -44,6 +44,7 @@ struct host1x_waitchk { struct host1x_job_unpin_data { struct host1x_bo *bo; struct sg_table *sgt; + size_t size; }; /* diff --git a/drivers/gpu/host1x/syncpt.c b/drivers/gpu/host1x/syncpt.c index 25c11a85050b..0ac026cdc30c 100644 --- a/drivers/gpu/host1x/syncpt.c +++ b/drivers/gpu/host1x/syncpt.c @@ -484,7 +484,7 @@ unsigned int host1x_syncpt_nb_mlocks(struct host1x *host) struct host1x_syncpt *host1x_syncpt_get(struct host1x *host, unsigned int id) { - if (host->info->nb_pts < id) + if (id >= host->info->nb_pts) return NULL; return host->syncpt + id; |