diff options
155 files changed, 5356 insertions, 2363 deletions
diff --git a/drivers/char/agp/ali-agp.c b/drivers/char/agp/ali-agp.c index dcbbb4ea3cc1..89527bae4602 100644 --- a/drivers/char/agp/ali-agp.c +++ b/drivers/char/agp/ali-agp.c @@ -381,7 +381,7 @@ static void agp_ali_remove(struct pci_dev *pdev) agp_put_bridge(bridge); } -static struct pci_device_id agp_ali_pci_table[] = { +static const struct pci_device_id agp_ali_pci_table[] = { { .class = (PCI_CLASS_BRIDGE_HOST << 8), .class_mask = ~0, diff --git a/drivers/char/agp/amd-k7-agp.c b/drivers/char/agp/amd-k7-agp.c index 5fbd333e4c6d..b450544dcaf0 100644 --- a/drivers/char/agp/amd-k7-agp.c +++ b/drivers/char/agp/amd-k7-agp.c @@ -21,7 +21,7 @@ #define AMD_TLBFLUSH 0x0c /* In mmio region (32-bit register) */ #define AMD_CACHEENTRY 0x10 /* In mmio region (32-bit register) */ -static struct pci_device_id agp_amdk7_pci_table[]; +static const struct pci_device_id agp_amdk7_pci_table[]; struct amd_page_map { unsigned long *real; @@ -508,7 +508,7 @@ static int agp_amdk7_resume(struct pci_dev *pdev) #endif /* CONFIG_PM */ /* must be the same order as name table above */ -static struct pci_device_id agp_amdk7_pci_table[] = { +static const struct pci_device_id agp_amdk7_pci_table[] = { { .class = (PCI_CLASS_BRIDGE_HOST << 8), .class_mask = ~0, diff --git a/drivers/char/agp/amd64-agp.c b/drivers/char/agp/amd64-agp.c index c99cd19d9147..e50c29c97ca7 100644 --- a/drivers/char/agp/amd64-agp.c +++ b/drivers/char/agp/amd64-agp.c @@ -610,7 +610,7 @@ static int agp_amd64_resume(struct pci_dev *pdev) #endif /* CONFIG_PM */ -static struct pci_device_id agp_amd64_pci_table[] = { +static const struct pci_device_id agp_amd64_pci_table[] = { { .class = (PCI_CLASS_BRIDGE_HOST << 8), .class_mask = ~0, diff --git a/drivers/char/agp/ati-agp.c b/drivers/char/agp/ati-agp.c index 0b5ec7af2414..88b4cbee4dac 100644 --- a/drivers/char/agp/ati-agp.c +++ b/drivers/char/agp/ati-agp.c @@ -540,7 +540,7 @@ static void agp_ati_remove(struct pci_dev *pdev) agp_put_bridge(bridge); } -static struct pci_device_id agp_ati_pci_table[] = { +static const struct pci_device_id agp_ati_pci_table[] = { { .class = (PCI_CLASS_BRIDGE_HOST << 8), .class_mask = ~0, diff --git a/drivers/char/agp/efficeon-agp.c b/drivers/char/agp/efficeon-agp.c index 533cb6d229b8..7f88490b5479 100644 --- a/drivers/char/agp/efficeon-agp.c +++ b/drivers/char/agp/efficeon-agp.c @@ -427,7 +427,7 @@ static int agp_efficeon_resume(struct pci_dev *pdev) } #endif -static struct pci_device_id agp_efficeon_pci_table[] = { +static const struct pci_device_id agp_efficeon_pci_table[] = { { .class = (PCI_CLASS_BRIDGE_HOST << 8), .class_mask = ~0, diff --git a/drivers/char/agp/intel-agp.c b/drivers/char/agp/intel-agp.c index 0a21daed5b62..9e4f27a6cb5a 100644 --- a/drivers/char/agp/intel-agp.c +++ b/drivers/char/agp/intel-agp.c @@ -828,7 +828,7 @@ static int agp_intel_resume(struct pci_dev *pdev) } #endif -static struct pci_device_id agp_intel_pci_table[] = { +static const struct pci_device_id agp_intel_pci_table[] = { #define ID(x) \ { \ .class = (PCI_CLASS_BRIDGE_HOST << 8), \ diff --git a/drivers/char/agp/nvidia-agp.c b/drivers/char/agp/nvidia-agp.c index 6c8d39cb566e..828b34445203 100644 --- a/drivers/char/agp/nvidia-agp.c +++ b/drivers/char/agp/nvidia-agp.c @@ -420,7 +420,7 @@ static int agp_nvidia_resume(struct pci_dev *pdev) #endif -static struct pci_device_id agp_nvidia_pci_table[] = { +static const struct pci_device_id agp_nvidia_pci_table[] = { { .class = (PCI_CLASS_BRIDGE_HOST << 8), .class_mask = ~0, diff --git a/drivers/char/agp/sis-agp.c b/drivers/char/agp/sis-agp.c index 2c74038da459..14909fc5d767 100644 --- a/drivers/char/agp/sis-agp.c +++ b/drivers/char/agp/sis-agp.c @@ -237,7 +237,7 @@ static int agp_sis_resume(struct pci_dev *pdev) #endif /* CONFIG_PM */ -static struct pci_device_id agp_sis_pci_table[] = { +static const struct pci_device_id agp_sis_pci_table[] = { { .class = (PCI_CLASS_BRIDGE_HOST << 8), .class_mask = ~0, diff --git a/drivers/char/agp/uninorth-agp.c b/drivers/char/agp/uninorth-agp.c index fdced547ad59..c381c8e396fc 100644 --- a/drivers/char/agp/uninorth-agp.c +++ b/drivers/char/agp/uninorth-agp.c @@ -679,7 +679,7 @@ static void agp_uninorth_remove(struct pci_dev *pdev) agp_put_bridge(bridge); } -static struct pci_device_id agp_uninorth_pci_table[] = { +static const struct pci_device_id agp_uninorth_pci_table[] = { { .class = (PCI_CLASS_BRIDGE_HOST << 8), .class_mask = ~0, diff --git a/drivers/gpu/drm/amd/amdgpu/Makefile b/drivers/gpu/drm/amd/amdgpu/Makefile index faea6349228f..658bac0cdc5e 100644 --- a/drivers/gpu/drm/amd/amdgpu/Makefile +++ b/drivers/gpu/drm/amd/amdgpu/Makefile @@ -25,7 +25,7 @@ amdgpu-y += amdgpu_device.o amdgpu_kms.o \ amdgpu_prime.o amdgpu_vm.o amdgpu_ib.o amdgpu_pll.o \ amdgpu_ucode.o amdgpu_bo_list.o amdgpu_ctx.o amdgpu_sync.o \ amdgpu_gtt_mgr.o amdgpu_vram_mgr.o amdgpu_virt.o amdgpu_atomfirmware.o \ - amdgpu_queue_mgr.o + amdgpu_queue_mgr.o amdgpu_vf_error.o # add asic specific block amdgpu-$(CONFIG_DRM_AMDGPU_CIK)+= cik.o cik_ih.o kv_smc.o kv_dpm.o \ diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h index ff7bf1a9f967..51d1364cf185 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h @@ -68,13 +68,16 @@ #include "gpu_scheduler.h" #include "amdgpu_virt.h" +#include "amdgpu_gart.h" /* * Modules parameters. */ extern int amdgpu_modeset; extern int amdgpu_vram_limit; -extern int amdgpu_gart_size; +extern int amdgpu_vis_vram_limit; +extern unsigned amdgpu_gart_size; +extern int amdgpu_gtt_size; extern int amdgpu_moverate; extern int amdgpu_benchmarking; extern int amdgpu_testing; @@ -104,6 +107,7 @@ extern unsigned amdgpu_pcie_gen_cap; extern unsigned amdgpu_pcie_lane_cap; extern unsigned amdgpu_cg_mask; extern unsigned amdgpu_pg_mask; +extern unsigned amdgpu_sdma_phase_quantum; extern char *amdgpu_disable_cu; extern char *amdgpu_virtual_display; extern unsigned amdgpu_pp_feature_mask; @@ -532,49 +536,6 @@ int amdgpu_fence_slab_init(void); void amdgpu_fence_slab_fini(void); /* - * GART structures, functions & helpers - */ -struct amdgpu_mc; - -#define AMDGPU_GPU_PAGE_SIZE 4096 -#define AMDGPU_GPU_PAGE_MASK (AMDGPU_GPU_PAGE_SIZE - 1) -#define AMDGPU_GPU_PAGE_SHIFT 12 -#define AMDGPU_GPU_PAGE_ALIGN(a) (((a) + AMDGPU_GPU_PAGE_MASK) & ~AMDGPU_GPU_PAGE_MASK) - -struct amdgpu_gart { - dma_addr_t table_addr; - struct amdgpu_bo *robj; - void *ptr; - unsigned num_gpu_pages; - unsigned num_cpu_pages; - unsigned table_size; -#ifdef CONFIG_DRM_AMDGPU_GART_DEBUGFS - struct page **pages; -#endif - bool ready; - - /* Asic default pte flags */ - uint64_t gart_pte_flags; - - const struct amdgpu_gart_funcs *gart_funcs; -}; - -int amdgpu_gart_table_ram_alloc(struct amdgpu_device *adev); -void amdgpu_gart_table_ram_free(struct amdgpu_device *adev); -int amdgpu_gart_table_vram_alloc(struct amdgpu_device *adev); -void amdgpu_gart_table_vram_free(struct amdgpu_device *adev); -int amdgpu_gart_table_vram_pin(struct amdgpu_device *adev); -void amdgpu_gart_table_vram_unpin(struct amdgpu_device *adev); -int amdgpu_gart_init(struct amdgpu_device *adev); -void amdgpu_gart_fini(struct amdgpu_device *adev); -int amdgpu_gart_unbind(struct amdgpu_device *adev, uint64_t offset, - int pages); -int amdgpu_gart_bind(struct amdgpu_device *adev, uint64_t offset, - int pages, struct page **pagelist, - dma_addr_t *dma_addr, uint64_t flags); -int amdgpu_ttm_recover_gart(struct amdgpu_device *adev); - -/* * VMHUB structures, functions & helpers */ struct amdgpu_vmhub { @@ -598,22 +559,20 @@ struct amdgpu_mc { * about vram size near mc fb location */ u64 mc_vram_size; u64 visible_vram_size; - u64 gtt_size; - u64 gtt_start; - u64 gtt_end; + u64 gart_size; + u64 gart_start; + u64 gart_end; u64 vram_start; u64 vram_end; unsigned vram_width; u64 real_vram_size; int vram_mtrr; - u64 gtt_base_align; u64 mc_mask; const struct firmware *fw; /* MC firmware */ uint32_t fw_version; struct amdgpu_irq_src vm_fault; uint32_t vram_type; uint32_t srbm_soft_reset; - struct amdgpu_mode_mc_save save; bool prt_warning; uint64_t stolen_size; /* apertures */ @@ -1159,7 +1118,9 @@ struct amdgpu_cs_parser { struct list_head validated; struct dma_fence *fence; uint64_t bytes_moved_threshold; + uint64_t bytes_moved_vis_threshold; uint64_t bytes_moved; + uint64_t bytes_moved_vis; struct amdgpu_bo_list_entry *evictable; /* user fence */ @@ -1231,7 +1192,9 @@ struct amdgpu_wb { int amdgpu_wb_get(struct amdgpu_device *adev, u32 *wb); void amdgpu_wb_free(struct amdgpu_device *adev, u32 wb); int amdgpu_wb_get_64bit(struct amdgpu_device *adev, u32 *wb); +int amdgpu_wb_get_256Bit(struct amdgpu_device *adev, u32 *wb); void amdgpu_wb_free_64bit(struct amdgpu_device *adev, u32 wb); +void amdgpu_wb_free_256bit(struct amdgpu_device *adev, u32 wb); void amdgpu_get_pcie_info(struct amdgpu_device *adev); @@ -1557,6 +1520,10 @@ struct amdgpu_device { spinlock_t gc_cac_idx_lock; amdgpu_rreg_t gc_cac_rreg; amdgpu_wreg_t gc_cac_wreg; + /* protects concurrent se_cac register access */ + spinlock_t se_cac_idx_lock; + amdgpu_rreg_t se_cac_rreg; + amdgpu_wreg_t se_cac_wreg; /* protects concurrent ENDPOINT (audio) register access */ spinlock_t audio_endpt_idx_lock; amdgpu_block_rreg_t audio_endpt_rreg; @@ -1593,6 +1560,7 @@ struct amdgpu_device { spinlock_t lock; s64 last_update_us; s64 accum_us; /* accumulated microseconds */ + s64 accum_us_vis; /* for visible VRAM */ u32 log2_max_MBps; } mm_stats; @@ -1687,6 +1655,8 @@ struct amdgpu_device { bool has_hw_reset; u8 reset_magic[AMDGPU_RESET_MAGIC_NUM]; + /* record last mm index being written through WREG32*/ + unsigned long last_mm_index; }; static inline struct amdgpu_device *amdgpu_ttm_adev(struct ttm_bo_device *bdev) @@ -1742,6 +1712,8 @@ void amdgpu_mm_wdoorbell64(struct amdgpu_device *adev, u32 index, u64 v); #define WREG32_DIDT(reg, v) adev->didt_wreg(adev, (reg), (v)) #define RREG32_GC_CAC(reg) adev->gc_cac_rreg(adev, (reg)) #define WREG32_GC_CAC(reg, v) adev->gc_cac_wreg(adev, (reg), (v)) +#define RREG32_SE_CAC(reg) adev->se_cac_rreg(adev, (reg)) +#define WREG32_SE_CAC(reg, v) adev->se_cac_wreg(adev, (reg), (v)) #define RREG32_AUDIO_ENDPT(block, reg) adev->audio_endpt_rreg(adev, (block), (reg)) #define WREG32_AUDIO_ENDPT(block, reg, v) adev->audio_endpt_wreg(adev, (block), (reg), (v)) #define WREG32_P(reg, val, mask) \ @@ -1792,50 +1764,6 @@ void amdgpu_mm_wdoorbell64(struct amdgpu_device *adev, u32 index, u64 v); #define RBIOS16(i) (RBIOS8(i) | (RBIOS8((i)+1) << 8)) #define RBIOS32(i) ((RBIOS16(i)) | (RBIOS16((i)+2) << 16)) -/* - * RING helpers. - */ -static inline void amdgpu_ring_write(struct amdgpu_ring *ring, uint32_t v) -{ - if (ring->count_dw <= 0) - DRM_ERROR("amdgpu: writing more dwords to the ring than expected!\n"); - ring->ring[ring->wptr++ & ring->buf_mask] = v; - ring->wptr &= ring->ptr_mask; - ring->count_dw--; -} - -static inline void amdgpu_ring_write_multiple(struct amdgpu_ring *ring, void *src, int count_dw) -{ - unsigned occupied, chunk1, chunk2; - void *dst; - - if (unlikely(ring->count_dw < count_dw)) { - DRM_ERROR("amdgpu: writing more dwords to the ring than expected!\n"); - return; - } - - occupied = ring->wptr & ring->buf_mask; - dst = (void *)&ring->ring[occupied]; - chunk1 = ring->buf_mask + 1 - occupied; - chunk1 = (chunk1 >= count_dw) ? count_dw: chunk1; - chunk2 = count_dw - chunk1; - chunk1 <<= 2; - chunk2 <<= 2; - - if (chunk1) - memcpy(dst, src, chunk1); - - if (chunk2) { - src += chunk1; - dst = (void *)ring->ring; - memcpy(dst, src, chunk2); - } - - ring->wptr += count_dw; - ring->wptr &= ring->ptr_mask; - ring->count_dw -= count_dw; -} - static inline struct amdgpu_sdma_instance * amdgpu_get_sdma_instance(struct amdgpu_ring *ring) { @@ -1898,7 +1826,6 @@ amdgpu_get_sdma_instance(struct amdgpu_ring *ring) #define amdgpu_ih_get_wptr(adev) (adev)->irq.ih_funcs->get_wptr((adev)) #define amdgpu_ih_decode_iv(adev, iv) (adev)->irq.ih_funcs->decode_iv((adev), (iv)) #define amdgpu_ih_set_rptr(adev) (adev)->irq.ih_funcs->set_rptr((adev)) -#define amdgpu_display_set_vga_render_state(adev, r) (adev)->mode_info.funcs->set_vga_render_state((adev), (r)) #define amdgpu_display_vblank_get_counter(adev, crtc) (adev)->mode_info.funcs->vblank_get_counter((adev), (crtc)) #define amdgpu_display_vblank_wait(adev, crtc) (adev)->mode_info.funcs->vblank_wait((adev), (crtc)) #define amdgpu_display_backlight_set_level(adev, e, l) (adev)->mode_info.funcs->backlight_set_level((e), (l)) @@ -1911,8 +1838,6 @@ amdgpu_get_sdma_instance(struct amdgpu_ring *ring) #define amdgpu_display_page_flip_get_scanoutpos(adev, crtc, vbl, pos) (adev)->mode_info.funcs->page_flip_get_scanoutpos((adev), (crtc), (vbl), (pos)) #define amdgpu_display_add_encoder(adev, e, s, c) (adev)->mode_info.funcs->add_encoder((adev), (e), (s), (c)) #define amdgpu_display_add_connector(adev, ci, sd, ct, ib, coi, h, r) (adev)->mode_info.funcs->add_connector((adev), (ci), (sd), (ct), (ib), (coi), (h), (r)) -#define amdgpu_display_stop_mc_access(adev, s) (adev)->mode_info.funcs->stop_mc_access((adev), (s)) -#define amdgpu_display_resume_mc_access(adev, s) (adev)->mode_info.funcs->resume_mc_access((adev), (s)) #define amdgpu_emit_copy_buffer(adev, ib, s, d, b) (adev)->mman.buffer_funcs->emit_copy_buffer((ib), (s), (d), (b)) #define amdgpu_emit_fill_buffer(adev, ib, s, d, b) (adev)->mman.buffer_funcs->emit_fill_buffer((ib), (s), (d), (b)) #define amdgpu_gfx_get_gpu_clock_counter(adev) (adev)->gfx.funcs->get_gpu_clock_counter((adev)) @@ -1927,7 +1852,8 @@ void amdgpu_pci_config_reset(struct amdgpu_device *adev); bool amdgpu_need_post(struct amdgpu_device *adev); void amdgpu_update_display_priority(struct amdgpu_device *adev); -void amdgpu_cs_report_moved_bytes(struct amdgpu_device *adev, u64 num_bytes); +void amdgpu_cs_report_moved_bytes(struct amdgpu_device *adev, u64 num_bytes, + u64 num_vis_bytes); void amdgpu_ttm_placement_from_domain(struct amdgpu_bo *abo, u32 domain); bool amdgpu_ttm_bo_is_amdgpu_bo(struct ttm_buffer_object *bo); int amdgpu_ttm_tt_get_user_pages(struct ttm_tt *ttm, struct page **pages); @@ -1943,7 +1869,7 @@ bool amdgpu_ttm_tt_is_readonly(struct ttm_tt *ttm); uint64_t amdgpu_ttm_tt_pte_flags(struct amdgpu_device *adev, struct ttm_tt *ttm, struct ttm_mem_reg *mem); void amdgpu_vram_location(struct amdgpu_device *adev, struct amdgpu_mc *mc, u64 base); -void amdgpu_gtt_location(struct amdgpu_device *adev, struct amdgpu_mc *mc); +void amdgpu_gart_location(struct amdgpu_device *adev, struct amdgpu_mc *mc); void amdgpu_ttm_set_active_vram_size(struct amdgpu_device *adev, u64 size); int amdgpu_ttm_init(struct amdgpu_device *adev); void amdgpu_ttm_fini(struct amdgpu_device *adev); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_atombios.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_atombios.c index 1e8e1123ddf4..ce443586a0c7 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_atombios.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_atombios.c @@ -1686,7 +1686,7 @@ void amdgpu_atombios_scratch_regs_lock(struct amdgpu_device *adev, bool lock) { uint32_t bios_6_scratch; - bios_6_scratch = RREG32(mmBIOS_SCRATCH_6); + bios_6_scratch = RREG32(adev->bios_scratch_reg_offset + 6); if (lock) { bios_6_scratch |= ATOM_S6_CRITICAL_STATE; @@ -1696,15 +1696,17 @@ void amdgpu_atombios_scratch_regs_lock(struct amdgpu_device *adev, bool lock) bios_6_scratch |= ATOM_S6_ACC_MODE; } - WREG32(mmBIOS_SCRATCH_6, bios_6_scratch); + WREG32(adev->bios_scratch_reg_offset + 6, bios_6_scratch); } void amdgpu_atombios_scratch_regs_init(struct amdgpu_device *adev) { uint32_t bios_2_scratch, bios_6_scratch; - bios_2_scratch = RREG32(mmBIOS_SCRATCH_2); - bios_6_scratch = RREG32(mmBIOS_SCRATCH_6); + adev->bios_scratch_reg_offset = mmBIOS_SCRATCH_0; + + bios_2_scratch = RREG32(adev->bios_scratch_reg_offset + 2); + bios_6_scratch = RREG32(adev->bios_scratch_reg_offset + 6); /* let the bios control the backlight */ bios_2_scratch &= ~ATOM_S2_VRI_BRIGHT_ENABLE; @@ -1715,8 +1717,8 @@ void amdgpu_atombios_scratch_regs_init(struct amdgpu_device *adev) /* clear the vbios dpms state */ bios_2_scratch &= ~ATOM_S2_DEVICE_DPMS_STATE; - WREG32(mmBIOS_SCRATCH_2, bios_2_scratch); - WREG32(mmBIOS_SCRATCH_6, bios_6_scratch); + WREG32(adev->bios_scratch_reg_offset + 2, bios_2_scratch); + WREG32(adev->bios_scratch_reg_offset + 6, bios_6_scratch); } void amdgpu_atombios_scratch_regs_save(struct amdgpu_device *adev) @@ -1724,7 +1726,7 @@ void amdgpu_atombios_scratch_regs_save(struct amdgpu_device *adev) int i; for (i = 0; i < AMDGPU_BIOS_NUM_SCRATCH; i++) - adev->bios_scratch[i] = RREG32(mmBIOS_SCRATCH_0 + i); + adev->bios_scratch[i] = RREG32(adev->bios_scratch_reg_offset + i); } void amdgpu_atombios_scratch_regs_restore(struct amdgpu_device *adev) @@ -1738,20 +1740,30 @@ void amdgpu_atombios_scratch_regs_restore(struct amdgpu_device *adev) 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]); + WREG32(adev->bios_scratch_reg_offset + i, adev->bios_scratch[i]); } void amdgpu_atombios_scratch_regs_engine_hung(struct amdgpu_device *adev, bool hung) { - u32 tmp = RREG32(mmBIOS_SCRATCH_3); + 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(mmBIOS_SCRATCH_3, tmp); + WREG32(adev->bios_scratch_reg_offset + 3, tmp); +} + +bool amdgpu_atombios_scratch_need_asic_init(struct amdgpu_device *adev) +{ + u32 tmp = RREG32(adev->bios_scratch_reg_offset + 7); + + if (tmp & ATOM_S7_ASIC_INIT_COMPLETE_MASK) + return false; + else + return true; } /* Atom needs data in little endian format diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_atombios.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_atombios.h index 38d0fe32e5cd..b0d5d1d7fdba 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_atombios.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_atombios.h @@ -200,6 +200,7 @@ void amdgpu_atombios_scratch_regs_save(struct amdgpu_device *adev); void amdgpu_atombios_scratch_regs_restore(struct amdgpu_device *adev); void amdgpu_atombios_scratch_regs_engine_hung(struct amdgpu_device *adev, bool hung); +bool amdgpu_atombios_scratch_need_asic_init(struct amdgpu_device *adev); void amdgpu_atombios_copy_swap(u8 *dst, u8 *src, u8 num_bytes, bool to_le); int amdgpu_atombios_get_max_vddc(struct amdgpu_device *adev, u8 voltage_type, diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.c index 4bdda56fccee..f9ffe8ef0cd6 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.c @@ -66,41 +66,6 @@ void amdgpu_atomfirmware_scratch_regs_init(struct amdgpu_device *adev) } } -void amdgpu_atomfirmware_scratch_regs_save(struct amdgpu_device *adev) -{ - int i; - - for (i = 0; i < AMDGPU_BIOS_NUM_SCRATCH; i++) - adev->bios_scratch[i] = RREG32(adev->bios_scratch_reg_offset + i); -} - -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; @@ -130,3 +95,129 @@ int amdgpu_atomfirmware_allocate_fb_scratch(struct amdgpu_device *adev) ctx->scratch_size_bytes = usage_bytes; return 0; } + +union igp_info { + struct atom_integrated_system_info_v1_11 v11; +}; + +/* + * Return vram width from integrated system info table, if available, + * or 0 if not. + */ +int amdgpu_atomfirmware_get_vram_width(struct amdgpu_device *adev) +{ + struct amdgpu_mode_info *mode_info = &adev->mode_info; + int index = get_index_into_master_table(atom_master_list_of_data_tables_v2_1, + integratedsysteminfo); + u16 data_offset, size; + union igp_info *igp_info; + u8 frev, crev; + + /* get any igp specific overrides */ + if (amdgpu_atom_parse_data_header(mode_info->atom_context, index, &size, + &frev, &crev, &data_offset)) { + igp_info = (union igp_info *) + (mode_info->atom_context->bios + data_offset); + switch (crev) { + case 11: + return igp_info->v11.umachannelnumber * 64; + default: + return 0; + } + } + + return 0; +} + +union firmware_info { + struct atom_firmware_info_v3_1 v31; +}; + +union smu_info { + struct atom_smu_info_v3_1 v31; +}; + +union umc_info { + struct atom_umc_info_v3_1 v31; +}; + +int amdgpu_atomfirmware_get_clock_info(struct amdgpu_device *adev) +{ + struct amdgpu_mode_info *mode_info = &adev->mode_info; + struct amdgpu_pll *spll = &adev->clock.spll; + struct amdgpu_pll *mpll = &adev->clock.mpll; + uint8_t frev, crev; + uint16_t data_offset; + int ret = -EINVAL, index; + + index = get_index_into_master_table(atom_master_list_of_data_tables_v2_1, + firmwareinfo); + if (amdgpu_atom_parse_data_header(mode_info->atom_context, index, NULL, + &frev, &crev, &data_offset)) { + union firmware_info *firmware_info = + (union firmware_info *)(mode_info->atom_context->bios + + data_offset); + + adev->clock.default_sclk = + le32_to_cpu(firmware_info->v31.bootup_sclk_in10khz); + adev->clock.default_mclk = + le32_to_cpu(firmware_info->v31.bootup_mclk_in10khz); + + adev->pm.current_sclk = adev->clock.default_sclk; + adev->pm.current_mclk = adev->clock.default_mclk; + + /* not technically a clock, but... */ + adev->mode_info.firmware_flags = + le32_to_cpu(firmware_info->v31.firmware_capability); + + ret = 0; + } + + index = get_index_into_master_table(atom_master_list_of_data_tables_v2_1, + smu_info); + if (amdgpu_atom_parse_data_header(mode_info->atom_context, index, NULL, + &frev, &crev, &data_offset)) { + union smu_info *smu_info = + (union smu_info *)(mode_info->atom_context->bios + + data_offset); + + /* system clock */ + spll->reference_freq = le32_to_cpu(smu_info->v31.core_refclk_10khz); + + spll->reference_div = 0; + spll->min_post_div = 1; + spll->max_post_div = 1; + spll->min_ref_div = 2; + spll->max_ref_div = 0xff; + spll->min_feedback_div = 4; + spll->max_feedback_div = 0xff; + spll->best_vco = 0; + + ret = 0; + } + + index = get_index_into_master_table(atom_master_list_of_data_tables_v2_1, + umc_info); + if (amdgpu_atom_parse_data_header(mode_info->atom_context, index, NULL, + &frev, &crev, &data_offset)) { + union umc_info *umc_info = + (union umc_info *)(mode_info->atom_context->bios + + data_offset); + + /* memory clock */ + mpll->reference_freq = le32_to_cpu(umc_info->v31.mem_refclk_10khz); + + mpll->reference_div = 0; + mpll->min_post_div = 1; + mpll->max_post_div = 1; + mpll->min_ref_div = 2; + mpll->max_ref_div = 0xff; + mpll->min_feedback_div = 4; + mpll->max_feedback_div = 0xff; + mpll->best_vco = 0; + + ret = 0; + } + + return ret; +} diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.h index a2c3ebe22c71..288b97e54347 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.h @@ -26,10 +26,8 @@ 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); +int amdgpu_atomfirmware_get_vram_width(struct amdgpu_device *adev); +int amdgpu_atomfirmware_get_clock_info(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 1beae5b930d0..2fb299afc12b 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_benchmark.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_benchmark.c @@ -40,7 +40,7 @@ static int amdgpu_benchmark_do_move(struct amdgpu_device *adev, unsigned size, for (i = 0; i < n; i++) { struct amdgpu_ring *ring = adev->mman.buffer_funcs_ring; r = amdgpu_copy_buffer(ring, saddr, daddr, size, NULL, &fence, - false); + false, false); if (r) goto exit_do_move; r = dma_fence_wait(fence, false); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_bios.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_bios.c index 365e735f6647..c21adf60a7f2 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_bios.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_bios.c @@ -86,19 +86,6 @@ static bool check_atom_bios(uint8_t *bios, size_t size) return false; } -static bool is_atom_fw(uint8_t *bios) -{ - uint16_t bios_header_start = bios[0x48] | (bios[0x49] << 8); - uint8_t frev = bios[bios_header_start + 2]; - uint8_t crev = bios[bios_header_start + 3]; - - if ((frev < 3) || - ((frev == 3) && (crev < 3))) - return false; - - return true; -} - /* If you boot an IGP board with a discrete card as the primary, * the IGP rom is not accessible via the rom bar as the IGP rom is * part of the system bios. On boot, the system bios puts a @@ -117,7 +104,7 @@ static bool igp_read_bios_from_vram(struct amdgpu_device *adev) adev->bios = NULL; vram_base = pci_resource_start(adev->pdev, 0); - bios = ioremap(vram_base, size); + bios = ioremap_wc(vram_base, size); if (!bios) { return false; } @@ -455,6 +442,6 @@ bool amdgpu_get_bios(struct amdgpu_device *adev) return false; success: - adev->is_atom_fw = is_atom_fw(adev->bios); + adev->is_atom_fw = (adev->asic_type >= CHIP_VEGA10) ? true : false; return true; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_bo_list.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_bo_list.c index f621ee115c98..d324e1c24028 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_bo_list.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_bo_list.c @@ -83,7 +83,7 @@ static int amdgpu_bo_list_create(struct amdgpu_device *adev, r = idr_alloc(&fpriv->bo_list_handles, list, 1, 0, GFP_KERNEL); mutex_unlock(&fpriv->bo_list_lock); if (r < 0) { - kfree(list); + amdgpu_bo_list_free(list); return r; } *id = r; @@ -198,12 +198,16 @@ amdgpu_bo_list_get(struct amdgpu_fpriv *fpriv, int id) result = idr_find(&fpriv->bo_list_handles, id); if (result) { - if (kref_get_unless_zero(&result->refcount)) + if (kref_get_unless_zero(&result->refcount)) { + rcu_read_unlock(); mutex_lock(&result->lock); - else + } else { + rcu_read_unlock(); result = NULL; + } + } else { + rcu_read_unlock(); } - rcu_read_unlock(); return result; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c index c0a806280257..a99e0bca6812 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c @@ -240,6 +240,8 @@ static uint32_t amdgpu_cgs_read_ind_register(struct cgs_device *cgs_device, return RREG32_DIDT(index); case CGS_IND_REG_GC_CAC: return RREG32_GC_CAC(index); + case CGS_IND_REG_SE_CAC: + return RREG32_SE_CAC(index); case CGS_IND_REG__AUDIO_ENDPT: DRM_ERROR("audio endpt register access not implemented.\n"); return 0; @@ -266,6 +268,8 @@ static void amdgpu_cgs_write_ind_register(struct cgs_device *cgs_device, return WREG32_DIDT(index, value); case CGS_IND_REG_GC_CAC: return WREG32_GC_CAC(index, value); + case CGS_IND_REG_SE_CAC: + return WREG32_SE_CAC(index, value); case CGS_IND_REG__AUDIO_ENDPT: DRM_ERROR("audio endpt register access not implemented.\n"); return; @@ -610,6 +614,17 @@ static int amdgpu_cgs_enter_safe_mode(struct cgs_device *cgs_device, return 0; } +static void amdgpu_cgs_lock_grbm_idx(struct cgs_device *cgs_device, + bool lock) +{ + CGS_FUNC_ADEV; + + if (lock) + mutex_lock(&adev->grbm_idx_mutex); + else + mutex_unlock(&adev->grbm_idx_mutex); +} + static int amdgpu_cgs_get_firmware_info(struct cgs_device *cgs_device, enum cgs_ucode_id type, struct cgs_firmware_info *info) @@ -719,7 +734,13 @@ static int amdgpu_cgs_get_firmware_info(struct cgs_device *cgs_device, strcpy(fw_name, "amdgpu/polaris12_smc.bin"); break; case CHIP_VEGA10: - strcpy(fw_name, "amdgpu/vega10_smc.bin"); + if ((adev->pdev->device == 0x687f) && + ((adev->pdev->revision == 0xc0) || + (adev->pdev->revision == 0xc1) || + (adev->pdev->revision == 0xc3))) + strcpy(fw_name, "amdgpu/vega10_acg_smc.bin"); + else + strcpy(fw_name, "amdgpu/vega10_smc.bin"); break; default: DRM_ERROR("SMC firmware not supported\n"); @@ -1117,6 +1138,7 @@ static const struct cgs_ops amdgpu_cgs_ops = { .query_system_info = amdgpu_cgs_query_system_info, .is_virtualization_enabled = amdgpu_cgs_is_virtualization_enabled, .enter_safe_mode = amdgpu_cgs_enter_safe_mode, + .lock_grbm_idx = amdgpu_cgs_lock_grbm_idx, }; static const struct cgs_os_ops amdgpu_cgs_os_ops = { diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c index 5599c01b265d..33789510e663 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c @@ -223,10 +223,11 @@ static s64 bytes_to_us(struct amdgpu_device *adev, u64 bytes) * ticks. The accumulated microseconds (us) are converted to bytes and * returned. */ -static u64 amdgpu_cs_get_threshold_for_moves(struct amdgpu_device *adev) +static void amdgpu_cs_get_threshold_for_moves(struct amdgpu_device *adev, + u64 *max_bytes, + u64 *max_vis_bytes) { s64 time_us, increment_us; - u64 max_bytes; u64 free_vram, total_vram, used_vram; /* Allow a maximum of 200 accumulated ms. This is basically per-IB @@ -238,8 +239,11 @@ static u64 amdgpu_cs_get_threshold_for_moves(struct amdgpu_device *adev) */ const s64 us_upper_bound = 200000; - if (!adev->mm_stats.log2_max_MBps) - return 0; + if (!adev->mm_stats.log2_max_MBps) { + *max_bytes = 0; + *max_vis_bytes = 0; + return; + } total_vram = adev->mc.real_vram_size - adev->vram_pin_size; used_vram = atomic64_read(&adev->vram_usage); @@ -280,23 +284,45 @@ static u64 amdgpu_cs_get_threshold_for_moves(struct amdgpu_device *adev) adev->mm_stats.accum_us = max(min_us, adev->mm_stats.accum_us); } - /* This returns 0 if the driver is in debt to disallow (optional) + /* This is set to 0 if the driver is in debt to disallow (optional) * buffer moves. */ - max_bytes = us_to_bytes(adev, adev->mm_stats.accum_us); + *max_bytes = us_to_bytes(adev, adev->mm_stats.accum_us); + + /* Do the same for visible VRAM if half of it is free */ + if (adev->mc.visible_vram_size < adev->mc.real_vram_size) { + u64 total_vis_vram = adev->mc.visible_vram_size; + u64 used_vis_vram = atomic64_read(&adev->vram_vis_usage); + + if (used_vis_vram < total_vis_vram) { + u64 free_vis_vram = total_vis_vram - used_vis_vram; + adev->mm_stats.accum_us_vis = min(adev->mm_stats.accum_us_vis + + increment_us, us_upper_bound); + + if (free_vis_vram >= total_vis_vram / 2) + adev->mm_stats.accum_us_vis = + max(bytes_to_us(adev, free_vis_vram / 2), + adev->mm_stats.accum_us_vis); + } + + *max_vis_bytes = us_to_bytes(adev, adev->mm_stats.accum_us_vis); + } else { + *max_vis_bytes = 0; + } spin_unlock(&adev->mm_stats.lock); - return max_bytes; } /* Report how many bytes have really been moved for the last command * submission. This can result in a debt that can stop buffer migrations * temporarily. */ -void amdgpu_cs_report_moved_bytes(struct amdgpu_device *adev, u64 num_bytes) +void amdgpu_cs_report_moved_bytes(struct amdgpu_device *adev, u64 num_bytes, + u64 num_vis_bytes) { spin_lock(&adev->mm_stats.lock); adev->mm_stats.accum_us -= bytes_to_us(adev, num_bytes); + adev->mm_stats.accum_us_vis -= bytes_to_us(adev, num_vis_bytes); spin_unlock(&adev->mm_stats.lock); } @@ -304,7 +330,7 @@ static int amdgpu_cs_bo_validate(struct amdgpu_cs_parser *p, struct amdgpu_bo *bo) { struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev); - u64 initial_bytes_moved; + u64 initial_bytes_moved, bytes_moved; uint32_t domain; int r; @@ -314,17 +340,35 @@ static int amdgpu_cs_bo_validate(struct amdgpu_cs_parser *p, /* Don't move this buffer if we have depleted our allowance * to move it. Don't move anything if the threshold is zero. */ - if (p->bytes_moved < p->bytes_moved_threshold) - domain = bo->prefered_domains; - else + if (p->bytes_moved < p->bytes_moved_threshold) { + if (adev->mc.visible_vram_size < adev->mc.real_vram_size && + (bo->flags & AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED)) { + /* And don't move a CPU_ACCESS_REQUIRED BO to limited + * visible VRAM if we've depleted our allowance to do + * that. + */ + if (p->bytes_moved_vis < p->bytes_moved_vis_threshold) + domain = bo->prefered_domains; + else + domain = bo->allowed_domains; + } else { + domain = bo->prefered_domains; + } + } else { domain = bo->allowed_domains; + } retry: amdgpu_ttm_placement_from_domain(bo, domain); initial_bytes_moved = atomic64_read(&adev->num_bytes_moved); r = ttm_bo_validate(&bo->tbo, &bo->placement, true, false); - p->bytes_moved += atomic64_read(&adev->num_bytes_moved) - - initial_bytes_moved; + bytes_moved = atomic64_read(&adev->num_bytes_moved) - + initial_bytes_moved; + p->bytes_moved += bytes_moved; + if (adev->mc.visible_vram_size < adev->mc.real_vram_size && + bo->tbo.mem.mem_type == TTM_PL_VRAM && + bo->tbo.mem.start < adev->mc.visible_vram_size >> PAGE_SHIFT) + p->bytes_moved_vis += bytes_moved; if (unlikely(r == -ENOMEM) && domain != bo->allowed_domains) { domain = bo->allowed_domains; @@ -350,7 +394,8 @@ static bool amdgpu_cs_try_evict(struct amdgpu_cs_parser *p, struct amdgpu_bo_list_entry *candidate = p->evictable; struct amdgpu_bo *bo = candidate->robj; struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev); - u64 initial_bytes_moved; + u64 initial_bytes_moved, bytes_moved; + bool update_bytes_moved_vis; uint32_t other; /* If we reached our current BO we can forget it */ @@ -370,10 +415,17 @@ static bool amdgpu_cs_try_evict(struct amdgpu_cs_parser *p, /* Good we can try to move this BO somewhere else */ amdgpu_ttm_placement_from_domain(bo, other); + update_bytes_moved_vis = + adev->mc.visible_vram_size < adev->mc.real_vram_size && + bo->tbo.mem.mem_type == TTM_PL_VRAM && + bo->tbo.mem.start < adev->mc.visible_vram_size >> PAGE_SHIFT; initial_bytes_moved = atomic64_read(&adev->num_bytes_moved); r = ttm_bo_validate(&bo->tbo, &bo->placement, true, false); - p->bytes_moved += atomic64_read(&adev->num_bytes_moved) - + bytes_moved = atomic64_read(&adev->num_bytes_moved) - initial_bytes_moved; + p->bytes_moved += bytes_moved; + if (update_bytes_moved_vis) + p->bytes_moved_vis += bytes_moved; if (unlikely(r)) break; @@ -554,8 +606,10 @@ static int amdgpu_cs_parser_bos(struct amdgpu_cs_parser *p, list_splice(&need_pages, &p->validated); } - p->bytes_moved_threshold = amdgpu_cs_get_threshold_for_moves(p->adev); + amdgpu_cs_get_threshold_for_moves(p->adev, &p->bytes_moved_threshold, + &p->bytes_moved_vis_threshold); p->bytes_moved = 0; + p->bytes_moved_vis = 0; p->evictable = list_last_entry(&p->validated, struct amdgpu_bo_list_entry, tv.head); @@ -579,8 +633,8 @@ static int amdgpu_cs_parser_bos(struct amdgpu_cs_parser *p, goto error_validate; } - amdgpu_cs_report_moved_bytes(p->adev, p->bytes_moved); - + amdgpu_cs_report_moved_bytes(p->adev, p->bytes_moved, + p->bytes_moved_vis); fpriv->vm.last_eviction_counter = atomic64_read(&p->adev->num_evictions); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c index 4a8fc15467cf..6279956e92a4 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c @@ -53,6 +53,9 @@ #include "bif/bif_4_1_d.h" #include <linux/pci.h> #include <linux/firmware.h> +#include "amdgpu_vf_error.h" + +#include "amdgpu_amdkfd.h" MODULE_FIRMWARE("amdgpu/vega10_gpu_info.bin"); MODULE_FIRMWARE("amdgpu/raven_gpu_info.bin"); @@ -128,6 +131,10 @@ void amdgpu_mm_wreg(struct amdgpu_device *adev, uint32_t reg, uint32_t v, { trace_amdgpu_mm_wreg(adev->pdev->device, reg, v); + if (adev->asic_type >= CHIP_VEGA10 && reg == 0) { + adev->last_mm_index = v; + } + if (!(acc_flags & AMDGPU_REGS_NO_KIQ) && amdgpu_sriov_runtime(adev)) { BUG_ON(in_interrupt()); return amdgpu_virt_kiq_wreg(adev, reg, v); @@ -143,6 +150,10 @@ void amdgpu_mm_wreg(struct amdgpu_device *adev, uint32_t reg, uint32_t v, writel(v, ((void __iomem *)adev->rmmio) + (mmMM_DATA * 4)); spin_unlock_irqrestore(&adev->mmio_idx_lock, flags); } + + if (adev->asic_type >= CHIP_VEGA10 && reg == 1 && adev->last_mm_index == 0x5702C) { + udelay(500); + } } u32 amdgpu_io_rreg(struct amdgpu_device *adev, u32 reg) @@ -157,6 +168,9 @@ u32 amdgpu_io_rreg(struct amdgpu_device *adev, u32 reg) void amdgpu_io_wreg(struct amdgpu_device *adev, u32 reg, u32 v) { + if (adev->asic_type >= CHIP_VEGA10 && reg == 0) { + adev->last_mm_index = v; + } if ((reg * 4) < adev->rio_mem_size) iowrite32(v, adev->rio_mem + (reg * 4)); @@ -164,6 +178,10 @@ void amdgpu_io_wreg(struct amdgpu_device *adev, u32 reg, u32 v) iowrite32((reg * 4), adev->rio_mem + (mmMM_INDEX * 4)); iowrite32(v, adev->rio_mem + (mmMM_DATA * 4)); } + + if (adev->asic_type >= CHIP_VEGA10 && reg == 1 && adev->last_mm_index == 0x5702C) { + udelay(500); + } } /** @@ -584,6 +602,21 @@ int amdgpu_wb_get_64bit(struct amdgpu_device *adev, u32 *wb) } } +int amdgpu_wb_get_256Bit(struct amdgpu_device *adev, u32 *wb) +{ + int i = 0; + unsigned long offset = bitmap_find_next_zero_area_off(adev->wb.used, + adev->wb.num_wb, 0, 8, 63, 0); + if ((offset + 7) < adev->wb.num_wb) { + for (i = 0; i < 8; i++) + __set_bit(offset + i, adev->wb.used); + *wb = offset; + return 0; + } else { + return -EINVAL; + } +} + /** * amdgpu_wb_free - Free a wb entry * @@ -615,6 +648,23 @@ void amdgpu_wb_free_64bit(struct amdgpu_device *adev, u32 wb) } /** + * amdgpu_wb_free_256bit - Free a wb entry + * + * @adev: amdgpu_device pointer + * @wb: wb index + * + * Free a wb slot allocated for use by the driver (all asics) + */ +void amdgpu_wb_free_256bit(struct amdgpu_device *adev, u32 wb) +{ + int i = 0; + + if ((wb + 7) < adev->wb.num_wb) + for (i = 0; i < 8; i++) + __clear_bit(wb + i, adev->wb.used); +} + +/** * amdgpu_vram_location - try to find VRAM location * @adev: amdgpu device structure holding all necessary informations * @mc: memory controller structure holding memory informations @@ -665,7 +715,7 @@ void amdgpu_vram_location(struct amdgpu_device *adev, struct amdgpu_mc *mc, u64 } /** - * amdgpu_gtt_location - try to find GTT location + * amdgpu_gart_location - try to find GTT location * @adev: amdgpu device structure holding all necessary informations * @mc: memory controller structure holding memory informations * @@ -676,28 +726,28 @@ void amdgpu_vram_location(struct amdgpu_device *adev, struct amdgpu_mc *mc, u64 * * FIXME: when reducing GTT size align new size on power of 2. */ -void amdgpu_gtt_location(struct amdgpu_device *adev, struct amdgpu_mc *mc) +void amdgpu_gart_location(struct amdgpu_device *adev, struct amdgpu_mc *mc) { u64 size_af, size_bf; - size_af = ((adev->mc.mc_mask - mc->vram_end) + mc->gtt_base_align) & ~mc->gtt_base_align; - size_bf = mc->vram_start & ~mc->gtt_base_align; + size_af = adev->mc.mc_mask - mc->vram_end; + size_bf = mc->vram_start; if (size_bf > size_af) { - if (mc->gtt_size > size_bf) { + if (mc->gart_size > size_bf) { dev_warn(adev->dev, "limiting GTT\n"); - mc->gtt_size = size_bf; + mc->gart_size = size_bf; } - mc->gtt_start = 0; + mc->gart_start = 0; } else { - if (mc->gtt_size > size_af) { + if (mc->gart_size > size_af) { dev_warn(adev->dev, "limiting GTT\n"); - mc->gtt_size = size_af; + mc->gart_size = size_af; } - mc->gtt_start = (mc->vram_end + 1 + mc->gtt_base_align) & ~mc->gtt_base_align; + mc->gart_start = mc->vram_end + 1; } - mc->gtt_end = mc->gtt_start + mc->gtt_size - 1; + mc->gart_end = mc->gart_start + mc->gart_size - 1; dev_info(adev->dev, "GTT: %lluM 0x%016llX - 0x%016llX\n", - mc->gtt_size >> 20, mc->gtt_start, mc->gtt_end); + mc->gart_size >> 20, mc->gart_start, mc->gart_end); } /* @@ -720,7 +770,12 @@ bool amdgpu_need_post(struct amdgpu_device *adev) adev->has_hw_reset = false; return true; } - /* then check MEM_SIZE, in case the crtcs are off */ + + /* bios scratch used on CIK+ */ + if (adev->asic_type >= CHIP_BONAIRE) + return amdgpu_atombios_scratch_need_asic_init(adev); + + /* check MEM_SIZE for older asics */ reg = amdgpu_asic_get_config_memsize(adev); if ((reg != 0) && (reg != 0xffffffff)) @@ -1031,19 +1086,6 @@ static unsigned int amdgpu_vga_set_decode(void *cookie, bool state) return VGA_RSRC_NORMAL_IO | VGA_RSRC_NORMAL_MEM; } -/** - * amdgpu_check_pot_argument - check that argument is a power of two - * - * @arg: value to check - * - * Validates that a certain argument is a power of two (all asics). - * Returns true if argument is valid. - */ -static bool amdgpu_check_pot_argument(int arg) -{ - return (arg & (arg - 1)) == 0; -} - static void amdgpu_check_block_size(struct amdgpu_device *adev) { /* defines number of bits in page table versus page directory, @@ -1077,7 +1119,7 @@ static void amdgpu_check_vm_size(struct amdgpu_device *adev) if (amdgpu_vm_size == -1) return; - if (!amdgpu_check_pot_argument(amdgpu_vm_size)) { + if (!is_power_of_2(amdgpu_vm_size)) { dev_warn(adev->dev, "VM size (%d) must be a power of 2\n", amdgpu_vm_size); goto def_value; @@ -1118,19 +1160,24 @@ static void amdgpu_check_arguments(struct amdgpu_device *adev) dev_warn(adev->dev, "sched jobs (%d) must be at least 4\n", amdgpu_sched_jobs); amdgpu_sched_jobs = 4; - } else if (!amdgpu_check_pot_argument(amdgpu_sched_jobs)){ + } else if (!is_power_of_2(amdgpu_sched_jobs)){ dev_warn(adev->dev, "sched jobs (%d) must be a power of 2\n", amdgpu_sched_jobs); amdgpu_sched_jobs = roundup_pow_of_two(amdgpu_sched_jobs); } - if (amdgpu_gart_size != -1) { + if (amdgpu_gart_size < 32) { + /* gart size must be greater or equal to 32M */ + dev_warn(adev->dev, "gart size (%d) too small\n", + amdgpu_gart_size); + amdgpu_gart_size = 32; + } + + if (amdgpu_gtt_size != -1 && amdgpu_gtt_size < 32) { /* gtt size must be greater or equal to 32M */ - if (amdgpu_gart_size < 32) { - dev_warn(adev->dev, "gart size (%d) too small\n", - amdgpu_gart_size); - amdgpu_gart_size = -1; - } + dev_warn(adev->dev, "gtt size (%d) too small\n", + amdgpu_gtt_size); + amdgpu_gtt_size = -1; } amdgpu_check_vm_size(adev); @@ -1138,7 +1185,7 @@ static void amdgpu_check_arguments(struct amdgpu_device *adev) amdgpu_check_block_size(adev); if (amdgpu_vram_page_split != -1 && (amdgpu_vram_page_split < 16 || - !amdgpu_check_pot_argument(amdgpu_vram_page_split))) { + !is_power_of_2(amdgpu_vram_page_split))) { dev_warn(adev->dev, "invalid VRAM page split (%d)\n", amdgpu_vram_page_split); amdgpu_vram_page_split = 1024; @@ -2019,7 +2066,7 @@ int amdgpu_device_init(struct amdgpu_device *adev, adev->flags = flags; adev->asic_type = flags & AMD_ASIC_MASK; adev->usec_timeout = AMDGPU_MAX_USEC_TIMEOUT; - adev->mc.gtt_size = 512 * 1024 * 1024; + adev->mc.gart_size = 512 * 1024 * 1024; adev->accel_working = false; adev->num_rings = 0; adev->mman.buffer_funcs = NULL; @@ -2068,6 +2115,7 @@ int amdgpu_device_init(struct amdgpu_device *adev, spin_lock_init(&adev->uvd_ctx_idx_lock); spin_lock_init(&adev->didt_idx_lock); spin_lock_init(&adev->gc_cac_idx_lock); + spin_lock_init(&adev->se_cac_idx_lock); spin_lock_init(&adev->audio_endpt_idx_lock); spin_lock_init(&adev->mm_stats.lock); @@ -2143,6 +2191,7 @@ int amdgpu_device_init(struct amdgpu_device *adev, r = amdgpu_atombios_init(adev); if (r) { dev_err(adev->dev, "amdgpu_atombios_init failed\n"); + amdgpu_vf_error_put(AMDGIM_ERROR_VF_ATOMBIOS_INIT_FAIL, 0, 0); goto failed; } @@ -2153,6 +2202,7 @@ int amdgpu_device_init(struct amdgpu_device *adev, if (amdgpu_vpost_needed(adev)) { if (!adev->bios) { dev_err(adev->dev, "no vBIOS found\n"); + amdgpu_vf_error_put(AMDGIM_ERROR_VF_NO_VBIOS, 0, 0); r = -EINVAL; goto failed; } @@ -2160,18 +2210,28 @@ int amdgpu_device_init(struct amdgpu_device *adev, r = amdgpu_atom_asic_init(adev->mode_info.atom_context); if (r) { dev_err(adev->dev, "gpu post error!\n"); + amdgpu_vf_error_put(AMDGIM_ERROR_VF_GPU_POST_ERROR, 0, 0); goto failed; } } else { DRM_INFO("GPU post is not needed\n"); } - if (!adev->is_atom_fw) { + if (adev->is_atom_fw) { + /* Initialize clocks */ + r = amdgpu_atomfirmware_get_clock_info(adev); + if (r) { + dev_err(adev->dev, "amdgpu_atomfirmware_get_clock_info failed\n"); + amdgpu_vf_error_put(AMDGIM_ERROR_VF_ATOMBIOS_GET_CLOCK_FAIL, 0, 0); + goto failed; + } + } else { /* Initialize clocks */ r = amdgpu_atombios_get_clock_info(adev); if (r) { dev_err(adev->dev, "amdgpu_atombios_get_clock_info failed\n"); - return r; + amdgpu_vf_error_put(AMDGIM_ERROR_VF_ATOMBIOS_GET_CLOCK_FAIL, 0, 0); + goto failed; } /* init i2c buses */ amdgpu_atombios_i2c_init(adev); @@ -2181,6 +2241,7 @@ int amdgpu_device_init(struct amdgpu_device *adev, r = amdgpu_fence_driver_init(adev); if (r) { dev_err(adev->dev, "amdgpu_fence_driver_init failed\n"); + amdgpu_vf_error_put(AMDGIM_ERROR_VF_FENCE_INIT_FAIL, 0, 0); goto failed; } @@ -2190,6 +2251,7 @@ int amdgpu_device_init(struct amdgpu_device *adev, r = amdgpu_init(adev); if (r) { dev_err(adev->dev, "amdgpu_init failed\n"); + amdgpu_vf_error_put(AMDGIM_ERROR_VF_AMDGPU_INIT_FAIL, 0, 0); amdgpu_fini(adev); goto failed; } @@ -2209,6 +2271,7 @@ int amdgpu_device_init(struct amdgpu_device *adev, r = amdgpu_ib_pool_init(adev); if (r) { dev_err(adev->dev, "IB initialization failed (%d).\n", r); + amdgpu_vf_error_put(AMDGIM_ERROR_VF_IB_INIT_FAIL, 0, r); goto failed; } @@ -2253,12 +2316,14 @@ int amdgpu_device_init(struct amdgpu_device *adev, r = amdgpu_late_init(adev); if (r) { dev_err(adev->dev, "amdgpu_late_init failed\n"); + amdgpu_vf_error_put(AMDGIM_ERROR_VF_AMDGPU_LATE_INIT_FAIL, 0, r); goto failed; } return 0; failed: + amdgpu_vf_error_trans_all(adev); if (runtime) vga_switcheroo_fini_domain_pm_ops(adev->dev); return r; @@ -2351,6 +2416,8 @@ int amdgpu_device_suspend(struct drm_device *dev, bool suspend, bool fbcon) } drm_modeset_unlock_all(dev); + amdgpu_amdkfd_suspend(adev); + /* unpin the front buffers and cursors */ list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) { struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc); @@ -2392,10 +2459,7 @@ int amdgpu_device_suspend(struct drm_device *dev, bool suspend, bool fbcon) */ amdgpu_bo_evict_vram(adev); - if (adev->is_atom_fw) - amdgpu_atomfirmware_scratch_regs_save(adev); - else - amdgpu_atombios_scratch_regs_save(adev); + amdgpu_atombios_scratch_regs_save(adev); pci_save_state(dev->pdev); if (suspend) { /* Shut down the device */ @@ -2444,10 +2508,7 @@ int amdgpu_device_resume(struct drm_device *dev, bool resume, bool fbcon) if (r) goto unlock; } - if (adev->is_atom_fw) - amdgpu_atomfirmware_scratch_regs_restore(adev); - else - amdgpu_atombios_scratch_regs_restore(adev); + amdgpu_atombios_scratch_regs_restore(adev); /* post card */ if (amdgpu_need_post(adev)) { @@ -2490,6 +2551,9 @@ int amdgpu_device_resume(struct drm_device *dev, bool resume, bool fbcon) } } } + r = amdgpu_amdkfd_resume(adev); + if (r) + return r; /* blat the mode back in */ if (fbcon) { @@ -2860,21 +2924,9 @@ int amdgpu_gpu_reset(struct amdgpu_device *adev) r = amdgpu_suspend(adev); retry: - /* Disable fb access */ - if (adev->mode_info.num_crtc) { - struct amdgpu_mode_mc_save save; - amdgpu_display_stop_mc_access(adev, &save); - amdgpu_wait_for_idle(adev, AMD_IP_BLOCK_TYPE_GMC); - } - if (adev->is_atom_fw) - amdgpu_atomfirmware_scratch_regs_save(adev); - else - amdgpu_atombios_scratch_regs_save(adev); + amdgpu_atombios_scratch_regs_save(adev); r = amdgpu_asic_reset(adev); - if (adev->is_atom_fw) - amdgpu_atomfirmware_scratch_regs_restore(adev); - else - amdgpu_atombios_scratch_regs_restore(adev); + amdgpu_atombios_scratch_regs_restore(adev); /* post card */ amdgpu_atom_asic_init(adev->mode_info.atom_context); @@ -2952,6 +3004,7 @@ out: } } else { dev_err(adev->dev, "asic resume failed (%d).\n", r); + amdgpu_vf_error_put(AMDGIM_ERROR_VF_ASIC_RESUME_FAIL, 0, r); for (i = 0; i < AMDGPU_MAX_RINGS; ++i) { if (adev->rings[i] && adev->rings[i]->sched.thread) { kthread_unpark(adev->rings[i]->sched.thread); @@ -2962,12 +3015,16 @@ out: drm_helper_resume_force_mode(adev->ddev); ttm_bo_unlock_delayed_workqueue(&adev->mman.bdev, resched); - if (r) + if (r) { /* bad news, how to tell it to userspace ? */ dev_info(adev->dev, "GPU reset failed\n"); - else + amdgpu_vf_error_put(AMDGIM_ERROR_VF_GPU_RESET_FAIL, 0, r); + } + else { dev_info(adev->dev, "GPU reset successed!\n"); + } + amdgpu_vf_error_trans_all(adev); return r; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c index 3c57102de634..5e9ce8a29669 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c @@ -74,7 +74,9 @@ #define KMS_DRIVER_PATCHLEVEL 0 int amdgpu_vram_limit = 0; -int amdgpu_gart_size = -1; /* auto */ +int amdgpu_vis_vram_limit = 0; +unsigned amdgpu_gart_size = 256; +int amdgpu_gtt_size = -1; /* auto */ int amdgpu_moverate = -1; /* auto */ int amdgpu_benchmarking = 0; int amdgpu_testing = 0; @@ -106,6 +108,7 @@ unsigned amdgpu_pcie_gen_cap = 0; unsigned amdgpu_pcie_lane_cap = 0; unsigned amdgpu_cg_mask = 0xffffffff; unsigned amdgpu_pg_mask = 0xffffffff; +unsigned amdgpu_sdma_phase_quantum = 32; char *amdgpu_disable_cu = NULL; char *amdgpu_virtual_display = NULL; unsigned amdgpu_pp_feature_mask = 0xffffffff; @@ -120,8 +123,14 @@ int amdgpu_lbpw = -1; MODULE_PARM_DESC(vramlimit, "Restrict VRAM for testing, in megabytes"); module_param_named(vramlimit, amdgpu_vram_limit, int, 0600); -MODULE_PARM_DESC(gartsize, "Size of PCIE/IGP gart to setup in megabytes (32, 64, etc., -1 = auto)"); -module_param_named(gartsize, amdgpu_gart_size, int, 0600); +MODULE_PARM_DESC(vis_vramlimit, "Restrict visible VRAM for testing, in megabytes"); +module_param_named(vis_vramlimit, amdgpu_vis_vram_limit, int, 0444); + +MODULE_PARM_DESC(gartsize, "Size of PCIE/IGP gart to setup in megabytes (32, 64, etc.)"); +module_param_named(gartsize, amdgpu_gart_size, uint, 0600); + +MODULE_PARM_DESC(gttsize, "Size of the GTT domain in megabytes (-1 = auto)"); +module_param_named(gttsize, amdgpu_gtt_size, int, 0600); MODULE_PARM_DESC(moverate, "Maximum buffer migration rate in MB/s. (32, 64, etc., -1=auto, 0=1=disabled)"); module_param_named(moverate, amdgpu_moverate, int, 0600); @@ -186,7 +195,7 @@ module_param_named(vm_debug, amdgpu_vm_debug, int, 0644); MODULE_PARM_DESC(vm_update_mode, "VM update using CPU (0 = never (default except for large BAR(LB)), 1 = Graphics only, 2 = Compute only (default for LB), 3 = Both"); module_param_named(vm_update_mode, amdgpu_vm_update_mode, int, 0444); -MODULE_PARM_DESC(vram_page_split, "Number of pages after we split VRAM allocations (default 1024, -1 = disable)"); +MODULE_PARM_DESC(vram_page_split, "Number of pages after we split VRAM allocations (default 512, -1 = disable)"); module_param_named(vram_page_split, amdgpu_vram_page_split, int, 0444); MODULE_PARM_DESC(exp_hw_support, "experimental hw support (1 = enable, 0 = disable (default))"); @@ -199,7 +208,7 @@ MODULE_PARM_DESC(sched_hw_submission, "the max number of HW submissions (default module_param_named(sched_hw_submission, amdgpu_sched_hw_submission, int, 0444); MODULE_PARM_DESC(ppfeaturemask, "all power features enabled (default))"); -module_param_named(ppfeaturemask, amdgpu_pp_feature_mask, int, 0444); +module_param_named(ppfeaturemask, amdgpu_pp_feature_mask, uint, 0444); MODULE_PARM_DESC(no_evict, "Support pinning request from user space (1 = enable, 0 = disable (default))"); module_param_named(no_evict, amdgpu_no_evict, int, 0444); @@ -219,6 +228,9 @@ module_param_named(cg_mask, amdgpu_cg_mask, uint, 0444); MODULE_PARM_DESC(pg_mask, "Powergating flags mask (0 = disable power gating)"); module_param_named(pg_mask, amdgpu_pg_mask, uint, 0444); +MODULE_PARM_DESC(sdma_phase_quantum, "SDMA context switch phase quantum (x 1K GPU clock cycles, 0 = no change (default 32))"); +module_param_named(sdma_phase_quantum, amdgpu_sdma_phase_quantum, uint, 0444); + MODULE_PARM_DESC(disable_cu, "Disable CUs (se.sh.cu,...)"); module_param_named(disable_cu, amdgpu_disable_cu, charp, 0444); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.c index a57abc1a25fb..5cc4987cd887 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.c @@ -55,6 +55,19 @@ /* * Common GART table functions. */ + +/** + * amdgpu_gart_set_defaults - set the default gart_size + * + * @adev: amdgpu_device pointer + * + * Set the default gart_size based on parameters and available VRAM. + */ +void amdgpu_gart_set_defaults(struct amdgpu_device *adev) +{ + adev->mc.gart_size = (uint64_t)amdgpu_gart_size << 20; +} + /** * amdgpu_gart_table_ram_alloc - allocate system ram for gart page table * @@ -263,6 +276,41 @@ int amdgpu_gart_unbind(struct amdgpu_device *adev, uint64_t offset, } /** + * amdgpu_gart_map - map dma_addresses into GART entries + * + * @adev: amdgpu_device pointer + * @offset: offset into the GPU's gart aperture + * @pages: number of pages to bind + * @dma_addr: DMA addresses of pages + * + * Map the dma_addresses into GART entries (all asics). + * Returns 0 for success, -EINVAL for failure. + */ +int amdgpu_gart_map(struct amdgpu_device *adev, uint64_t offset, + int pages, dma_addr_t *dma_addr, uint64_t flags, + void *dst) +{ + uint64_t page_base; + unsigned i, j, t; + + if (!adev->gart.ready) { + WARN(1, "trying to bind memory to uninitialized GART !\n"); + return -EINVAL; + } + + t = offset / AMDGPU_GPU_PAGE_SIZE; + + for (i = 0; i < pages; i++) { + page_base = dma_addr[i]; + for (j = 0; j < (PAGE_SIZE / AMDGPU_GPU_PAGE_SIZE); j++, t++) { + amdgpu_gart_set_pte_pde(adev, dst, t, page_base, flags); + page_base += AMDGPU_GPU_PAGE_SIZE; + } + } + return 0; +} + +/** * amdgpu_gart_bind - bind pages into the gart page table * * @adev: amdgpu_device pointer @@ -279,31 +327,30 @@ int amdgpu_gart_bind(struct amdgpu_device *adev, uint64_t offset, int pages, struct page **pagelist, dma_addr_t *dma_addr, uint64_t flags) { - unsigned t; - unsigned p; - uint64_t page_base; - int i, j; +#ifdef CONFIG_DRM_AMDGPU_GART_DEBUGFS + unsigned i,t,p; +#endif + int r; if (!adev->gart.ready) { WARN(1, "trying to bind memory to uninitialized GART !\n"); return -EINVAL; } +#ifdef CONFIG_DRM_AMDGPU_GART_DEBUGFS t = offset / AMDGPU_GPU_PAGE_SIZE; p = t / (PAGE_SIZE / AMDGPU_GPU_PAGE_SIZE); - - for (i = 0; i < pages; i++, p++) { -#ifdef CONFIG_DRM_AMDGPU_GART_DEBUGFS + for (i = 0; i < pages; i++, p++) adev->gart.pages[p] = pagelist[i]; #endif - if (adev->gart.ptr) { - page_base = dma_addr[i]; - for (j = 0; j < (PAGE_SIZE / AMDGPU_GPU_PAGE_SIZE); j++, t++) { - amdgpu_gart_set_pte_pde(adev, adev->gart.ptr, t, page_base, flags); - page_base += AMDGPU_GPU_PAGE_SIZE; - } - } + + if (adev->gart.ptr) { + r = amdgpu_gart_map(adev, offset, pages, dma_addr, flags, + adev->gart.ptr); + if (r) + return r; } + mb(); amdgpu_gart_flush_gpu_tlb(adev, 0); return 0; @@ -333,8 +380,8 @@ int amdgpu_gart_init(struct amdgpu_device *adev) if (r) return r; /* Compute table size */ - adev->gart.num_cpu_pages = adev->mc.gtt_size / PAGE_SIZE; - adev->gart.num_gpu_pages = adev->mc.gtt_size / AMDGPU_GPU_PAGE_SIZE; + adev->gart.num_cpu_pages = adev->mc.gart_size / PAGE_SIZE; + adev->gart.num_gpu_pages = adev->mc.gart_size / AMDGPU_GPU_PAGE_SIZE; DRM_INFO("GART: num cpu pages %u, num gpu pages %u\n", adev->gart.num_cpu_pages, adev->gart.num_gpu_pages); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.h new file mode 100644 index 000000000000..d4cce6936200 --- /dev/null +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.h @@ -0,0 +1,77 @@ +/* + * Copyright 2017 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + */ + +#ifndef __AMDGPU_GART_H__ +#define __AMDGPU_GART_H__ + +#include <linux/types.h> + +/* + * GART structures, functions & helpers + */ +struct amdgpu_device; +struct amdgpu_bo; +struct amdgpu_gart_funcs; + +#define AMDGPU_GPU_PAGE_SIZE 4096 +#define AMDGPU_GPU_PAGE_MASK (AMDGPU_GPU_PAGE_SIZE - 1) +#define AMDGPU_GPU_PAGE_SHIFT 12 +#define AMDGPU_GPU_PAGE_ALIGN(a) (((a) + AMDGPU_GPU_PAGE_MASK) & ~AMDGPU_GPU_PAGE_MASK) + +struct amdgpu_gart { + dma_addr_t table_addr; + struct amdgpu_bo *robj; + void *ptr; + unsigned num_gpu_pages; + unsigned num_cpu_pages; + unsigned table_size; +#ifdef CONFIG_DRM_AMDGPU_GART_DEBUGFS + struct page **pages; +#endif + bool ready; + + /* Asic default pte flags */ + uint64_t gart_pte_flags; + + const struct amdgpu_gart_funcs *gart_funcs; +}; + +void amdgpu_gart_set_defaults(struct amdgpu_device *adev); +int amdgpu_gart_table_ram_alloc(struct amdgpu_device *adev); +void amdgpu_gart_table_ram_free(struct amdgpu_device *adev); +int amdgpu_gart_table_vram_alloc(struct amdgpu_device *adev); +void amdgpu_gart_table_vram_free(struct amdgpu_device *adev); +int amdgpu_gart_table_vram_pin(struct amdgpu_device *adev); +void amdgpu_gart_table_vram_unpin(struct amdgpu_device *adev); +int amdgpu_gart_init(struct amdgpu_device *adev); +void amdgpu_gart_fini(struct amdgpu_device *adev); +int amdgpu_gart_unbind(struct amdgpu_device *adev, uint64_t offset, + int pages); +int amdgpu_gart_map(struct amdgpu_device *adev, uint64_t offset, + int pages, dma_addr_t *dma_addr, uint64_t flags, + void *dst); +int amdgpu_gart_bind(struct amdgpu_device *adev, uint64_t offset, + int pages, struct page **pagelist, + dma_addr_t *dma_addr, uint64_t flags); + +#endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c index 621f739103a6..917ac5e074a0 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c @@ -49,7 +49,6 @@ int amdgpu_gem_object_create(struct amdgpu_device *adev, unsigned long size, struct drm_gem_object **obj) { struct amdgpu_bo *robj; - unsigned long max_size; int r; *obj = NULL; @@ -58,17 +57,6 @@ int amdgpu_gem_object_create(struct amdgpu_device *adev, unsigned long size, alignment = PAGE_SIZE; } - if (!(initial_domain & (AMDGPU_GEM_DOMAIN_GDS | AMDGPU_GEM_DOMAIN_GWS | AMDGPU_GEM_DOMAIN_OA))) { - /* Maximum bo size is the unpinned gtt size since we use the gtt to - * handle vram to system pool migrations. - */ - max_size = adev->mc.gtt_size - adev->gart_pin_size; - if (size > max_size) { - DRM_DEBUG("Allocation size %ldMb bigger than %ldMb limit\n", - size >> 20, max_size >> 20); - return -ENOMEM; - } - } retry: r = amdgpu_bo_create(adev, size, alignment, kernel, initial_domain, flags, NULL, NULL, &robj); @@ -784,6 +772,7 @@ static int amdgpu_debugfs_gem_bo_info(int id, void *ptr, void *data) unsigned domain; const char *placement; unsigned pin_count; + uint64_t offset; domain = amdgpu_mem_type_to_domain(bo->tbo.mem.mem_type); switch (domain) { @@ -798,9 +787,12 @@ static int amdgpu_debugfs_gem_bo_info(int id, void *ptr, void *data) placement = " CPU"; break; } - seq_printf(m, "\t0x%08x: %12ld byte %s @ 0x%010Lx", - id, amdgpu_bo_size(bo), placement, - amdgpu_bo_gpu_offset(bo)); + seq_printf(m, "\t0x%08x: %12ld byte %s", + id, amdgpu_bo_size(bo), placement); + + offset = ACCESS_ONCE(bo->tbo.mem.start); + if (offset != AMDGPU_BO_INVALID_OFFSET) + seq_printf(m, " @ 0x%010Lx", offset); pin_count = ACCESS_ONCE(bo->pin_count); if (pin_count) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c index e26108aad3fe..4f6c68fc1dd9 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c @@ -125,7 +125,8 @@ void amdgpu_gfx_compute_queue_acquire(struct amdgpu_device *adev) if (mec >= adev->gfx.mec.num_mec) break; - if (adev->gfx.mec.num_mec > 1) { + /* FIXME: spreading the queues across pipes causes perf regressions */ + if (0) { /* policy: amdgpu owns the first two queues of the first MEC */ if (mec == 0 && queue < 2) set_bit(i, adev->gfx.mec.queue_bitmap); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gtt_mgr.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gtt_mgr.c index f7d22c44034d..5e6b90c6794f 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gtt_mgr.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gtt_mgr.c @@ -42,13 +42,17 @@ struct amdgpu_gtt_mgr { static int amdgpu_gtt_mgr_init(struct ttm_mem_type_manager *man, unsigned long p_size) { + struct amdgpu_device *adev = amdgpu_ttm_adev(man->bdev); struct amdgpu_gtt_mgr *mgr; + uint64_t start, size; mgr = kzalloc(sizeof(*mgr), GFP_KERNEL); if (!mgr) return -ENOMEM; - drm_mm_init(&mgr->mm, 0, p_size); + start = AMDGPU_GTT_MAX_TRANSFER_SIZE * AMDGPU_GTT_NUM_TRANSFER_WINDOWS; + size = (adev->mc.gart_size >> PAGE_SHIFT) - start; + drm_mm_init(&mgr->mm, start, size); spin_lock_init(&mgr->lock); mgr->available = p_size; man->priv = mgr; @@ -81,6 +85,20 @@ static int amdgpu_gtt_mgr_fini(struct ttm_mem_type_manager *man) } /** + * amdgpu_gtt_mgr_is_allocated - Check if mem has address space + * + * @mem: the mem object to check + * + * Check if a mem object has already address space allocated. + */ +bool amdgpu_gtt_mgr_is_allocated(struct ttm_mem_reg *mem) +{ + struct drm_mm_node *node = mem->mm_node; + + return (node->start != AMDGPU_BO_INVALID_OFFSET); +} + +/** * amdgpu_gtt_mgr_alloc - allocate new ranges * * @man: TTM memory type manager @@ -95,13 +113,14 @@ int amdgpu_gtt_mgr_alloc(struct ttm_mem_type_manager *man, const struct ttm_place *place, struct ttm_mem_reg *mem) { + struct amdgpu_device *adev = amdgpu_ttm_adev(man->bdev); struct amdgpu_gtt_mgr *mgr = man->priv; struct drm_mm_node *node = mem->mm_node; enum drm_mm_insert_mode mode; unsigned long fpfn, lpfn; int r; - if (node->start != AMDGPU_BO_INVALID_OFFSET) + if (amdgpu_gtt_mgr_is_allocated(mem)) return 0; if (place) @@ -112,7 +131,7 @@ int amdgpu_gtt_mgr_alloc(struct ttm_mem_type_manager *man, if (place && place->lpfn) lpfn = place->lpfn; else - lpfn = man->size; + lpfn = adev->gart.num_cpu_pages; mode = DRM_MM_INSERT_BEST; if (place && place->flags & TTM_PL_FLAG_TOPDOWN) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c index f774b3f497d2..659997bfff30 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c @@ -130,6 +130,7 @@ int amdgpu_ib_schedule(struct amdgpu_ring *ring, unsigned num_ibs, unsigned i; int r = 0; + bool need_pipe_sync = false; if (num_ibs == 0) return -EINVAL; @@ -165,15 +166,15 @@ int amdgpu_ib_schedule(struct amdgpu_ring *ring, unsigned num_ibs, if (ring->funcs->emit_pipeline_sync && job && ((tmp = amdgpu_sync_get_fence(&job->sched_sync)) || amdgpu_vm_need_pipeline_sync(ring, job))) { - amdgpu_ring_emit_pipeline_sync(ring); + need_pipe_sync = true; dma_fence_put(tmp); } if (ring->funcs->insert_start) ring->funcs->insert_start(ring); - if (vm) { - r = amdgpu_vm_flush(ring, job); + if (job) { + r = amdgpu_vm_flush(ring, job, need_pipe_sync); if (r) { amdgpu_ring_undo(ring); return r; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_irq.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_irq.c index 2480273c1dca..4bdd851f56d0 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_irq.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_irq.c @@ -220,6 +220,10 @@ int amdgpu_irq_init(struct amdgpu_device *adev) int r = 0; spin_lock_init(&adev->irq.lock); + + /* Disable vblank irqs aggressively for power-saving */ + adev->ddev->vblank_disable_immediate = true; + r = drm_vblank_init(adev->ddev, adev->mode_info.num_crtc); if (r) { return r; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c index 3d641e10e6b6..4510627ae83e 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c @@ -81,6 +81,8 @@ int amdgpu_job_alloc_with_ib(struct amdgpu_device *adev, unsigned size, r = amdgpu_ib_get(adev, NULL, size, &(*job)->ibs[0]); if (r) kfree(*job); + else + (*job)->vm_pd_addr = adev->gart.table_addr; return r; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c index b0b23101d1c8..09f833255ba1 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c @@ -485,7 +485,8 @@ static int amdgpu_info_ioctl(struct drm_device *dev, void *data, struct drm_file vram_gtt.vram_size -= adev->vram_pin_size; vram_gtt.vram_cpu_accessible_size = adev->mc.visible_vram_size; vram_gtt.vram_cpu_accessible_size -= (adev->vram_pin_size - adev->invisible_pin_size); - vram_gtt.gtt_size = adev->mc.gtt_size; + vram_gtt.gtt_size = adev->mman.bdev.man[TTM_PL_TT].size; + vram_gtt.gtt_size *= PAGE_SIZE; vram_gtt.gtt_size -= adev->gart_pin_size; return copy_to_user(out, &vram_gtt, min((size_t)size, sizeof(vram_gtt))) ? -EFAULT : 0; @@ -510,9 +511,10 @@ static int amdgpu_info_ioctl(struct drm_device *dev, void *data, struct drm_file mem.cpu_accessible_vram.max_allocation = mem.cpu_accessible_vram.usable_heap_size * 3 / 4; - mem.gtt.total_heap_size = adev->mc.gtt_size; - mem.gtt.usable_heap_size = - adev->mc.gtt_size - adev->gart_pin_size; + mem.gtt.total_heap_size = adev->mman.bdev.man[TTM_PL_TT].size; + mem.gtt.total_heap_size *= PAGE_SIZE; + mem.gtt.usable_heap_size = mem.gtt.total_heap_size + - adev->gart_pin_size; mem.gtt.heap_usage = atomic64_read(&adev->gtt_usage); mem.gtt.max_allocation = mem.gtt.usable_heap_size * 3 / 4; @@ -571,8 +573,8 @@ static int amdgpu_info_ioctl(struct drm_device *dev, void *data, struct drm_file dev_info.max_engine_clock = amdgpu_dpm_get_sclk(adev, false) * 10; dev_info.max_memory_clock = amdgpu_dpm_get_mclk(adev, false) * 10; } else { - dev_info.max_engine_clock = adev->pm.default_sclk * 10; - dev_info.max_memory_clock = adev->pm.default_mclk * 10; + dev_info.max_engine_clock = adev->clock.default_sclk * 10; + dev_info.max_memory_clock = adev->clock.default_mclk * 10; } dev_info.enabled_rb_pipes_mask = adev->gfx.config.backend_enable_mask; dev_info.num_rb_pipes = adev->gfx.config.max_backends_per_se * @@ -587,8 +589,9 @@ static int amdgpu_info_ioctl(struct drm_device *dev, void *data, struct drm_file dev_info.virtual_address_offset = AMDGPU_VA_RESERVED_SIZE; dev_info.virtual_address_max = (uint64_t)adev->vm_manager.max_pfn * AMDGPU_GPU_PAGE_SIZE; dev_info.virtual_address_alignment = max((int)PAGE_SIZE, AMDGPU_GPU_PAGE_SIZE); - dev_info.pte_fragment_size = (1 << AMDGPU_LOG2_PAGES_PER_FRAG) * - AMDGPU_GPU_PAGE_SIZE; + dev_info.pte_fragment_size = + (1 << AMDGPU_LOG2_PAGES_PER_FRAG(adev)) * + AMDGPU_GPU_PAGE_SIZE; dev_info.gart_page_size = AMDGPU_GPU_PAGE_SIZE; dev_info.cu_active_number = adev->gfx.cu_info.number; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h index 39f7eda6091e..2af2678ddaf6 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h @@ -257,15 +257,7 @@ struct amdgpu_audio { int num_pins; }; -struct amdgpu_mode_mc_save { - u32 vga_render_control; - u32 vga_hdp_control; - bool crtc_enabled[AMDGPU_MAX_CRTCS]; -}; - struct amdgpu_display_funcs { - /* vga render */ - void (*set_vga_render_state)(struct amdgpu_device *adev, bool render); /* display watermarks */ void (*bandwidth_update)(struct amdgpu_device *adev); /* get frame count */ @@ -300,10 +292,6 @@ struct amdgpu_display_funcs { uint16_t connector_object_id, struct amdgpu_hpd *hpd, struct amdgpu_router *router); - void (*stop_mc_access)(struct amdgpu_device *adev, - struct amdgpu_mode_mc_save *save); - void (*resume_mc_access)(struct amdgpu_device *adev, - struct amdgpu_mode_mc_save *save); }; struct amdgpu_mode_info { diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c index 8ee69652be8c..3ec43cf9ad78 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c @@ -93,6 +93,7 @@ static void amdgpu_ttm_bo_destroy(struct ttm_buffer_object *tbo) bo = container_of(tbo, struct amdgpu_bo, tbo); + amdgpu_bo_kunmap(bo); amdgpu_update_memory_usage(adev, &bo->tbo.mem, NULL); drm_gem_object_release(&bo->gem_base); @@ -322,7 +323,7 @@ int amdgpu_bo_create_restricted(struct amdgpu_device *adev, struct amdgpu_bo *bo; enum ttm_bo_type type; unsigned long page_align; - u64 initial_bytes_moved; + u64 initial_bytes_moved, bytes_moved; size_t acc_size; int r; @@ -398,8 +399,14 @@ int amdgpu_bo_create_restricted(struct amdgpu_device *adev, r = ttm_bo_init_reserved(&adev->mman.bdev, &bo->tbo, size, type, &bo->placement, page_align, !kernel, NULL, acc_size, sg, resv, &amdgpu_ttm_bo_destroy); - amdgpu_cs_report_moved_bytes(adev, - atomic64_read(&adev->num_bytes_moved) - initial_bytes_moved); + bytes_moved = atomic64_read(&adev->num_bytes_moved) - + initial_bytes_moved; + if (adev->mc.visible_vram_size < adev->mc.real_vram_size && + bo->tbo.mem.mem_type == TTM_PL_VRAM && + bo->tbo.mem.start < adev->mc.visible_vram_size >> PAGE_SHIFT) + amdgpu_cs_report_moved_bytes(adev, bytes_moved, bytes_moved); + else + amdgpu_cs_report_moved_bytes(adev, bytes_moved, 0); if (unlikely(r != 0)) return r; @@ -426,6 +433,10 @@ int amdgpu_bo_create_restricted(struct amdgpu_device *adev, trace_amdgpu_bo_create(bo); + /* Treat CPU_ACCESS_REQUIRED only as a hint if given by UMD */ + if (type == ttm_bo_type_device) + bo->flags &= ~AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED; + return 0; fail_unreserve: @@ -535,7 +546,7 @@ int amdgpu_bo_backup_to_shadow(struct amdgpu_device *adev, r = amdgpu_copy_buffer(ring, bo_addr, shadow_addr, amdgpu_bo_size(bo), resv, fence, - direct); + direct, false); if (!r) amdgpu_bo_fence(bo, *fence, true); @@ -588,7 +599,7 @@ int amdgpu_bo_restore_from_shadow(struct amdgpu_device *adev, r = amdgpu_copy_buffer(ring, shadow_addr, bo_addr, amdgpu_bo_size(bo), resv, fence, - direct); + direct, false); if (!r) amdgpu_bo_fence(bo, *fence, true); @@ -724,15 +735,16 @@ int amdgpu_bo_pin_restricted(struct amdgpu_bo *bo, u32 domain, dev_err(adev->dev, "%p pin failed\n", bo); goto error; } - r = amdgpu_ttm_bind(&bo->tbo, &bo->tbo.mem); - if (unlikely(r)) { - dev_err(adev->dev, "%p bind failed\n", bo); - goto error; - } bo->pin_count = 1; - if (gpu_addr != NULL) + if (gpu_addr != NULL) { + r = amdgpu_ttm_bind(&bo->tbo, &bo->tbo.mem); + if (unlikely(r)) { + dev_err(adev->dev, "%p bind failed\n", bo); + goto error; + } *gpu_addr = amdgpu_bo_gpu_offset(bo); + } if (domain == AMDGPU_GEM_DOMAIN_VRAM) { adev->vram_pin_size += amdgpu_bo_size(bo); if (bo->flags & AMDGPU_GEM_CREATE_NO_CPU_ACCESS) @@ -921,6 +933,8 @@ void amdgpu_bo_move_notify(struct ttm_buffer_object *bo, abo = container_of(bo, struct amdgpu_bo, tbo); amdgpu_vm_bo_invalidate(adev, abo); + amdgpu_bo_kunmap(abo); + /* remember the eviction */ if (evict) atomic64_inc(&adev->num_evictions); @@ -939,19 +953,22 @@ int amdgpu_bo_fault_reserve_notify(struct ttm_buffer_object *bo) { struct amdgpu_device *adev = amdgpu_ttm_adev(bo->bdev); struct amdgpu_bo *abo; - unsigned long offset, size, lpfn; - int i, r; + unsigned long offset, size; + int r; if (!amdgpu_ttm_bo_is_amdgpu_bo(bo)) return 0; abo = container_of(bo, struct amdgpu_bo, tbo); + + /* Remember that this BO was accessed by the CPU */ + abo->flags |= AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED; + if (bo->mem.mem_type != TTM_PL_VRAM) return 0; size = bo->mem.num_pages << PAGE_SHIFT; offset = bo->mem.start << PAGE_SHIFT; - /* TODO: figure out how to map scattered VRAM to the CPU */ if ((offset + size) <= adev->mc.visible_vram_size) return 0; @@ -961,26 +978,21 @@ int amdgpu_bo_fault_reserve_notify(struct ttm_buffer_object *bo) /* hurrah the memory is not visible ! */ atomic64_inc(&adev->num_vram_cpu_page_faults); - amdgpu_ttm_placement_from_domain(abo, AMDGPU_GEM_DOMAIN_VRAM); - lpfn = adev->mc.visible_vram_size >> PAGE_SHIFT; - for (i = 0; i < abo->placement.num_placement; i++) { - /* Force into visible VRAM */ - if ((abo->placements[i].flags & TTM_PL_FLAG_VRAM) && - (!abo->placements[i].lpfn || - abo->placements[i].lpfn > lpfn)) - abo->placements[i].lpfn = lpfn; - } + amdgpu_ttm_placement_from_domain(abo, AMDGPU_GEM_DOMAIN_VRAM | + AMDGPU_GEM_DOMAIN_GTT); + + /* Avoid costly evictions; only set GTT as a busy placement */ + abo->placement.num_busy_placement = 1; + abo->placement.busy_placement = &abo->placements[1]; + r = ttm_bo_validate(bo, &abo->placement, false, false); - if (unlikely(r == -ENOMEM)) { - amdgpu_ttm_placement_from_domain(abo, AMDGPU_GEM_DOMAIN_GTT); - return ttm_bo_validate(bo, &abo->placement, false, false); - } else if (unlikely(r != 0)) { + if (unlikely(r != 0)) return r; - } offset = bo->mem.start << PAGE_SHIFT; /* this should never happen */ - if ((offset + size) > adev->mc.visible_vram_size) + if (bo->mem.mem_type == TTM_PL_VRAM && + (offset + size) > adev->mc.visible_vram_size) return -EINVAL; return 0; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h index 382485115b06..833b172a2c2a 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h @@ -120,7 +120,11 @@ static inline u64 amdgpu_bo_mmap_offset(struct amdgpu_bo *bo) */ static inline bool amdgpu_bo_gpu_accessible(struct amdgpu_bo *bo) { - return bo->tbo.mem.mem_type != TTM_PL_SYSTEM; + switch (bo->tbo.mem.mem_type) { + case TTM_PL_TT: return amdgpu_ttm_is_bound(bo->tbo.ttm); + case TTM_PL_VRAM: return true; + default: return false; + } } int amdgpu_bo_create(struct amdgpu_device *adev, diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c index 4083be61b328..8c2204c7b384 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c @@ -63,8 +63,13 @@ static int psp_sw_init(void *handle) psp->smu_reload_quirk = psp_v3_1_smu_reload_quirk; break; case CHIP_RAVEN: +#if 0 + psp->init_microcode = psp_v10_0_init_microcode; +#endif psp->prep_cmd_buf = psp_v10_0_prep_cmd_buf; psp->ring_init = psp_v10_0_ring_init; + psp->ring_create = psp_v10_0_ring_create; + psp->ring_destroy = psp_v10_0_ring_destroy; psp->cmd_submit = psp_v10_0_cmd_submit; psp->compare_sram_data = psp_v10_0_compare_sram_data; break; @@ -95,9 +100,8 @@ int psp_wait_for(struct psp_context *psp, uint32_t reg_index, int i; struct amdgpu_device *adev = psp->adev; - val = RREG32(reg_index); - for (i = 0; i < adev->usec_timeout; i++) { + val = RREG32(reg_index); if (check_changed) { if (val != reg_val) return 0; @@ -118,33 +122,18 @@ psp_cmd_submit_buf(struct psp_context *psp, int index) { int ret; - struct amdgpu_bo *cmd_buf_bo; - uint64_t cmd_buf_mc_addr; - struct psp_gfx_cmd_resp *cmd_buf_mem; - struct amdgpu_device *adev = psp->adev; - - ret = amdgpu_bo_create_kernel(adev, PSP_CMD_BUFFER_SIZE, PAGE_SIZE, - AMDGPU_GEM_DOMAIN_VRAM, - &cmd_buf_bo, &cmd_buf_mc_addr, - (void **)&cmd_buf_mem); - if (ret) - return ret; - memset(cmd_buf_mem, 0, PSP_CMD_BUFFER_SIZE); + memset(psp->cmd_buf_mem, 0, PSP_CMD_BUFFER_SIZE); - memcpy(cmd_buf_mem, cmd, sizeof(struct psp_gfx_cmd_resp)); + memcpy(psp->cmd_buf_mem, cmd, sizeof(struct psp_gfx_cmd_resp)); - ret = psp_cmd_submit(psp, ucode, cmd_buf_mc_addr, + ret = psp_cmd_submit(psp, ucode, psp->cmd_buf_mc_addr, fence_mc_addr, index); while (*((unsigned int *)psp->fence_buf) != index) { msleep(1); } - amdgpu_bo_free_kernel(&cmd_buf_bo, - &cmd_buf_mc_addr, - (void **)&cmd_buf_mem); - return ret; } @@ -352,13 +341,20 @@ static int psp_load_fw(struct amdgpu_device *adev) &psp->fence_buf_mc_addr, &psp->fence_buf); if (ret) + goto failed_mem2; + + ret = amdgpu_bo_create_kernel(adev, PSP_CMD_BUFFER_SIZE, PAGE_SIZE, + AMDGPU_GEM_DOMAIN_VRAM, + &psp->cmd_buf_bo, &psp->cmd_buf_mc_addr, + (void **)&psp->cmd_buf_mem); + 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; + goto failed_mem; ret = psp_tmr_init(psp); if (ret) @@ -379,9 +375,13 @@ static int psp_load_fw(struct amdgpu_device *adev) return 0; failed_mem: + amdgpu_bo_free_kernel(&psp->cmd_buf_bo, + &psp->cmd_buf_mc_addr, + (void **)&psp->cmd_buf_mem); +failed_mem1: amdgpu_bo_free_kernel(&psp->fence_buf_bo, &psp->fence_buf_mc_addr, &psp->fence_buf); -failed_mem1: +failed_mem2: amdgpu_bo_free_kernel(&psp->fw_pri_bo, &psp->fw_pri_mc_addr, &psp->fw_pri_buf); failed: @@ -435,16 +435,15 @@ static int psp_hw_fini(void *handle) 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); + amdgpu_bo_free_kernel(&psp->tmr_bo, &psp->tmr_mc_addr, &psp->tmr_buf); + amdgpu_bo_free_kernel(&psp->fw_pri_bo, + &psp->fw_pri_mc_addr, &psp->fw_pri_buf); + amdgpu_bo_free_kernel(&psp->fence_buf_bo, + &psp->fence_buf_mc_addr, &psp->fence_buf); + amdgpu_bo_free_kernel(&psp->asd_shared_bo, &psp->asd_shared_mc_addr, + &psp->asd_shared_buf); + amdgpu_bo_free_kernel(&psp->cmd_buf_bo, &psp->cmd_buf_mc_addr, + (void **)&psp->cmd_buf_mem); kfree(psp->cmd); psp->cmd = NULL; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.h index 1a1c8b469f93..538fa9dbfb21 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.h @@ -108,6 +108,11 @@ struct psp_context struct amdgpu_bo *fence_buf_bo; uint64_t fence_buf_mc_addr; void *fence_buf; + + /* cmd buffer */ + struct amdgpu_bo *cmd_buf_bo; + uint64_t cmd_buf_mc_addr; + struct psp_gfx_cmd_resp *cmd_buf_mem; }; struct amdgpu_psp_funcs { diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c index 75165e07b1cd..15b7149d1204 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c @@ -212,10 +212,19 @@ int amdgpu_ring_init(struct amdgpu_device *adev, struct amdgpu_ring *ring, } - r = amdgpu_wb_get(adev, &ring->fence_offs); - if (r) { - dev_err(adev->dev, "(%d) ring fence_offs wb alloc failed\n", r); - return r; + if (amdgpu_sriov_vf(adev) && ring->funcs->type == AMDGPU_RING_TYPE_GFX) { + r = amdgpu_wb_get_256Bit(adev, &ring->fence_offs); + if (r) { + dev_err(adev->dev, "(%d) ring fence_offs wb alloc failed\n", r); + return r; + } + + } else { + r = amdgpu_wb_get(adev, &ring->fence_offs); + if (r) { + dev_err(adev->dev, "(%d) ring fence_offs wb alloc failed\n", r); + return r; + } } r = amdgpu_wb_get(adev, &ring->cond_exe_offs); @@ -278,17 +287,18 @@ void amdgpu_ring_fini(struct amdgpu_ring *ring) ring->ready = false; if (ring->funcs->support_64bit_ptrs) { - amdgpu_wb_free_64bit(ring->adev, ring->cond_exe_offs); - amdgpu_wb_free_64bit(ring->adev, ring->fence_offs); amdgpu_wb_free_64bit(ring->adev, ring->rptr_offs); amdgpu_wb_free_64bit(ring->adev, ring->wptr_offs); } else { - amdgpu_wb_free(ring->adev, ring->cond_exe_offs); - amdgpu_wb_free(ring->adev, ring->fence_offs); amdgpu_wb_free(ring->adev, ring->rptr_offs); amdgpu_wb_free(ring->adev, ring->wptr_offs); } + amdgpu_wb_free(ring->adev, ring->cond_exe_offs); + if (amdgpu_sriov_vf(ring->adev) && ring->funcs->type == AMDGPU_RING_TYPE_GFX) + amdgpu_wb_free_256bit(ring->adev, ring->fence_offs); + else + amdgpu_wb_free(ring->adev, ring->fence_offs); amdgpu_bo_free_kernel(&ring->ring_obj, &ring->gpu_addr, diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h index bc8dec992f73..322d25299a00 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h @@ -212,4 +212,44 @@ static inline void amdgpu_ring_clear_ring(struct amdgpu_ring *ring) } +static inline void amdgpu_ring_write(struct amdgpu_ring *ring, uint32_t v) +{ + if (ring->count_dw <= 0) + DRM_ERROR("amdgpu: writing more dwords to the ring than expected!\n"); + ring->ring[ring->wptr++ & ring->buf_mask] = v; + ring->wptr &= ring->ptr_mask; + ring->count_dw--; +} + +static inline void amdgpu_ring_write_multiple(struct amdgpu_ring *ring, + void *src, int count_dw) +{ + unsigned occupied, chunk1, chunk2; + void *dst; + + if (unlikely(ring->count_dw < count_dw)) + DRM_ERROR("amdgpu: writing more dwords to the ring than expected!\n"); + + occupied = ring->wptr & ring->buf_mask; + dst = (void *)&ring->ring[occupied]; + chunk1 = ring->buf_mask + 1 - occupied; + chunk1 = (chunk1 >= count_dw) ? count_dw: chunk1; + chunk2 = count_dw - chunk1; + chunk1 <<= 2; + chunk2 <<= 2; + + if (chunk1) + memcpy(dst, src, chunk1); + + if (chunk2) { + src += chunk1; + dst = (void *)ring->ring; + memcpy(dst, src, chunk2); + } + + ring->wptr += count_dw; + ring->wptr &= ring->ptr_mask; + ring->count_dw -= count_dw; +} + #endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_test.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_test.c index 15510dadde01..3c4d7574d704 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_test.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_test.c @@ -33,7 +33,7 @@ static void amdgpu_do_test_moves(struct amdgpu_device *adev) struct amdgpu_ring *ring = adev->mman.buffer_funcs_ring; struct amdgpu_bo *vram_obj = NULL; struct amdgpu_bo **gtt_obj = NULL; - uint64_t gtt_addr, vram_addr; + uint64_t gart_addr, vram_addr; unsigned n, size; int i, r; @@ -42,7 +42,7 @@ static void amdgpu_do_test_moves(struct amdgpu_device *adev) /* Number of tests = * (Total GTT - IB pool - writeback page - ring buffers) / test size */ - n = adev->mc.gtt_size - AMDGPU_IB_POOL_SIZE*64*1024; + n = adev->mc.gart_size - AMDGPU_IB_POOL_SIZE*64*1024; for (i = 0; i < AMDGPU_MAX_RINGS; ++i) if (adev->rings[i]) n -= adev->rings[i]->ring_size; @@ -76,7 +76,7 @@ static void amdgpu_do_test_moves(struct amdgpu_device *adev) } for (i = 0; i < n; i++) { void *gtt_map, *vram_map; - void **gtt_start, **gtt_end; + void **gart_start, **gart_end; void **vram_start, **vram_end; struct dma_fence *fence = NULL; @@ -91,7 +91,7 @@ static void amdgpu_do_test_moves(struct amdgpu_device *adev) r = amdgpu_bo_reserve(gtt_obj[i], false); if (unlikely(r != 0)) goto out_lclean_unref; - r = amdgpu_bo_pin(gtt_obj[i], AMDGPU_GEM_DOMAIN_GTT, >t_addr); + r = amdgpu_bo_pin(gtt_obj[i], AMDGPU_GEM_DOMAIN_GTT, &gart_addr); if (r) { DRM_ERROR("Failed to pin GTT object %d\n", i); goto out_lclean_unres; @@ -103,15 +103,15 @@ static void amdgpu_do_test_moves(struct amdgpu_device *adev) goto out_lclean_unpin; } - for (gtt_start = gtt_map, gtt_end = gtt_map + size; - gtt_start < gtt_end; - gtt_start++) - *gtt_start = gtt_start; + for (gart_start = gtt_map, gart_end = gtt_map + size; + gart_start < gart_end; + gart_start++) + *gart_start = gart_start; amdgpu_bo_kunmap(gtt_obj[i]); - r = amdgpu_copy_buffer(ring, gtt_addr, vram_addr, - size, NULL, &fence, false); + r = amdgpu_copy_buffer(ring, gart_addr, vram_addr, + size, NULL, &fence, false, false); if (r) { DRM_ERROR("Failed GTT->VRAM copy %d\n", i); @@ -132,21 +132,21 @@ static void amdgpu_do_test_moves(struct amdgpu_device *adev) goto out_lclean_unpin; } - for (gtt_start = gtt_map, gtt_end = gtt_map + size, + for (gart_start = gtt_map, gart_end = gtt_map + size, vram_start = vram_map, vram_end = vram_map + size; vram_start < vram_end; - gtt_start++, vram_start++) { - if (*vram_start != gtt_start) { + gart_start++, vram_start++) { + if (*vram_start != gart_start) { DRM_ERROR("Incorrect GTT->VRAM copy %d: Got 0x%p, " "expected 0x%p (GTT/VRAM offset " "0x%16llx/0x%16llx)\n", - i, *vram_start, gtt_start, + i, *vram_start, gart_start, (unsigned long long) - (gtt_addr - adev->mc.gtt_start + - (void*)gtt_start - gtt_map), + (gart_addr - adev->mc.gart_start + + (void*)gart_start - gtt_map), (unsigned long long) (vram_addr - adev->mc.vram_start + - (void*)gtt_start - gtt_map)); + (void*)gart_start - gtt_map)); amdgpu_bo_kunmap(vram_obj); goto out_lclean_unpin; } @@ -155,8 +155,8 @@ static void amdgpu_do_test_moves(struct amdgpu_device *adev) amdgpu_bo_kunmap(vram_obj); - r = amdgpu_copy_buffer(ring, vram_addr, gtt_addr, - size, NULL, &fence, false); + r = amdgpu_copy_buffer(ring, vram_addr, gart_addr, + size, NULL, &fence, false, false); if (r) { DRM_ERROR("Failed VRAM->GTT copy %d\n", i); @@ -177,20 +177,20 @@ static void amdgpu_do_test_moves(struct amdgpu_device *adev) goto out_lclean_unpin; } - for (gtt_start = gtt_map, gtt_end = gtt_map + size, + for (gart_start = gtt_map, gart_end = gtt_map + size, vram_start = vram_map, vram_end = vram_map + size; - gtt_start < gtt_end; - gtt_start++, vram_start++) { - if (*gtt_start != vram_start) { + gart_start < gart_end; + gart_start++, vram_start++) { + if (*gart_start != vram_start) { DRM_ERROR("Incorrect VRAM->GTT copy %d: Got 0x%p, " "expected 0x%p (VRAM/GTT offset " "0x%16llx/0x%16llx)\n", - i, *gtt_start, vram_start, + i, *gart_start, vram_start, (unsigned long long) (vram_addr - adev->mc.vram_start + (void*)vram_start - vram_map), (unsigned long long) - (gtt_addr - adev->mc.gtt_start + + (gart_addr - adev->mc.gart_start + (void*)vram_start - vram_map)); amdgpu_bo_kunmap(gtt_obj[i]); goto out_lclean_unpin; @@ -200,7 +200,7 @@ static void amdgpu_do_test_moves(struct amdgpu_device *adev) amdgpu_bo_kunmap(gtt_obj[i]); DRM_INFO("Tested GTT->VRAM and VRAM->GTT copy for GTT offset 0x%llx\n", - gtt_addr - adev->mc.gtt_start); + gart_addr - adev->mc.gart_start); continue; out_lclean_unpin: diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h index 8601904e670a..509f7a63d40c 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h @@ -224,7 +224,7 @@ TRACE_EVENT(amdgpu_vm_bo_map, __field(long, start) __field(long, last) __field(u64, offset) - __field(u32, flags) + __field(u64, flags) ), TP_fast_assign( @@ -234,7 +234,7 @@ TRACE_EVENT(amdgpu_vm_bo_map, __entry->offset = mapping->offset; __entry->flags = mapping->flags; ), - TP_printk("bo=%p, start=%lx, last=%lx, offset=%010llx, flags=%08x", + TP_printk("bo=%p, start=%lx, last=%lx, offset=%010llx, flags=%llx", __entry->bo, __entry->start, __entry->last, __entry->offset, __entry->flags) ); @@ -248,7 +248,7 @@ TRACE_EVENT(amdgpu_vm_bo_unmap, __field(long, start) __field(long, last) __field(u64, offset) - __field(u32, flags) + __field(u64, flags) ), TP_fast_assign( @@ -258,7 +258,7 @@ TRACE_EVENT(amdgpu_vm_bo_unmap, __entry->offset = mapping->offset; __entry->flags = mapping->flags; ), - TP_printk("bo=%p, start=%lx, last=%lx, offset=%010llx, flags=%08x", + TP_printk("bo=%p, start=%lx, last=%lx, offset=%010llx, flags=%llx", __entry->bo, __entry->start, __entry->last, __entry->offset, __entry->flags) ); @@ -269,7 +269,7 @@ DECLARE_EVENT_CLASS(amdgpu_vm_mapping, TP_STRUCT__entry( __field(u64, soffset) __field(u64, eoffset) - __field(u32, flags) + __field(u64, flags) ), TP_fast_assign( @@ -277,7 +277,7 @@ DECLARE_EVENT_CLASS(amdgpu_vm_mapping, __entry->eoffset = mapping->last + 1; __entry->flags = mapping->flags; ), - TP_printk("soffs=%010llx, eoffs=%010llx, flags=%08x", + TP_printk("soffs=%010llx, eoffs=%010llx, flags=%llx", __entry->soffset, __entry->eoffset, __entry->flags) ); @@ -293,14 +293,14 @@ DEFINE_EVENT(amdgpu_vm_mapping, amdgpu_vm_bo_mapping, TRACE_EVENT(amdgpu_vm_set_ptes, TP_PROTO(uint64_t pe, uint64_t addr, unsigned count, - uint32_t incr, uint32_t flags), + uint32_t incr, uint64_t flags), TP_ARGS(pe, addr, count, incr, flags), TP_STRUCT__entry( __field(u64, pe) __field(u64, addr) __field(u32, count) __field(u32, incr) - __field(u32, flags) + __field(u64, flags) ), TP_fast_assign( @@ -310,7 +310,7 @@ TRACE_EVENT(amdgpu_vm_set_ptes, __entry->incr = incr; __entry->flags = flags; ), - TP_printk("pe=%010Lx, addr=%010Lx, incr=%u, flags=%08x, count=%u", + TP_printk("pe=%010Lx, addr=%010Lx, incr=%u, flags=%llx, count=%u", __entry->pe, __entry->addr, __entry->incr, __entry->flags, __entry->count) ); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c index c9b131b13ef7..e6f9a54c959d 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c @@ -47,10 +47,15 @@ #define DRM_FILE_PAGE_OFFSET (0x100000000ULL >> PAGE_SHIFT) +static int amdgpu_map_buffer(struct ttm_buffer_object *bo, + struct ttm_mem_reg *mem, unsigned num_pages, + uint64_t offset, unsigned window, + struct amdgpu_ring *ring, + uint64_t *addr); + static int amdgpu_ttm_debugfs_init(struct amdgpu_device *adev); static void amdgpu_ttm_debugfs_fini(struct amdgpu_device *adev); - /* * Global memory. */ @@ -97,6 +102,8 @@ static int amdgpu_ttm_global_init(struct amdgpu_device *adev) goto error_bo; } + mutex_init(&adev->mman.gtt_window_lock); + ring = adev->mman.buffer_funcs_ring; rq = &ring->sched.sched_rq[AMD_SCHED_PRIORITY_KERNEL]; r = amd_sched_entity_init(&ring->sched, &adev->mman.entity, @@ -123,6 +130,7 @@ static void amdgpu_ttm_global_fini(struct amdgpu_device *adev) if (adev->mman.mem_global_referenced) { amd_sched_entity_fini(adev->mman.entity.sched, &adev->mman.entity); + mutex_destroy(&adev->mman.gtt_window_lock); drm_global_item_unref(&adev->mman.bo_global_ref.ref); drm_global_item_unref(&adev->mman.mem_global_ref); adev->mman.mem_global_referenced = false; @@ -150,7 +158,7 @@ static int amdgpu_init_mem_type(struct ttm_bo_device *bdev, uint32_t type, break; case TTM_PL_TT: man->func = &amdgpu_gtt_mgr_func; - man->gpu_offset = adev->mc.gtt_start; + man->gpu_offset = adev->mc.gart_start; man->available_caching = TTM_PL_MASK_CACHING; man->default_caching = TTM_PL_FLAG_CACHED; man->flags = TTM_MEMTYPE_FLAG_MAPPABLE | TTM_MEMTYPE_FLAG_CMA; @@ -186,12 +194,11 @@ static void amdgpu_evict_flags(struct ttm_buffer_object *bo, { struct amdgpu_device *adev = amdgpu_ttm_adev(bo->bdev); struct amdgpu_bo *abo; - static struct ttm_place placements = { + static const struct ttm_place placements = { .fpfn = 0, .lpfn = 0, .flags = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM }; - unsigned i; if (!amdgpu_ttm_bo_is_amdgpu_bo(bo)) { placement->placement = &placements; @@ -207,22 +214,36 @@ static void amdgpu_evict_flags(struct ttm_buffer_object *bo, adev->mman.buffer_funcs_ring && adev->mman.buffer_funcs_ring->ready == false) { amdgpu_ttm_placement_from_domain(abo, AMDGPU_GEM_DOMAIN_CPU); + } else if (adev->mc.visible_vram_size < adev->mc.real_vram_size && + !(abo->flags & AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED)) { + unsigned fpfn = adev->mc.visible_vram_size >> PAGE_SHIFT; + struct drm_mm_node *node = bo->mem.mm_node; + unsigned long pages_left; + + for (pages_left = bo->mem.num_pages; + pages_left; + pages_left -= node->size, node++) { + if (node->start < fpfn) + break; + } + + if (!pages_left) + goto gtt; + + /* Try evicting to the CPU inaccessible part of VRAM + * first, but only set GTT as busy placement, so this + * BO will be evicted to GTT rather than causing other + * BOs to be evicted from VRAM + */ + amdgpu_ttm_placement_from_domain(abo, AMDGPU_GEM_DOMAIN_VRAM | + AMDGPU_GEM_DOMAIN_GTT); + abo->placements[0].fpfn = fpfn; + abo->placements[0].lpfn = 0; + abo->placement.busy_placement = &abo->placements[1]; + abo->placement.num_busy_placement = 1; } else { +gtt: amdgpu_ttm_placement_from_domain(abo, AMDGPU_GEM_DOMAIN_GTT); - for (i = 0; i < abo->placement.num_placement; ++i) { - if (!(abo->placements[i].flags & - TTM_PL_FLAG_TT)) - continue; - - if (abo->placements[i].lpfn) - continue; - - /* set an upper limit to force directly - * allocating address space for the BO. - */ - abo->placements[i].lpfn = - adev->mc.gtt_size >> PAGE_SHIFT; - } } break; case TTM_PL_TT: @@ -252,29 +273,18 @@ static void amdgpu_move_null(struct ttm_buffer_object *bo, new_mem->mm_node = NULL; } -static int amdgpu_mm_node_addr(struct ttm_buffer_object *bo, - struct drm_mm_node *mm_node, - struct ttm_mem_reg *mem, - uint64_t *addr) +static uint64_t amdgpu_mm_node_addr(struct ttm_buffer_object *bo, + struct drm_mm_node *mm_node, + struct ttm_mem_reg *mem) { - int r; - - switch (mem->mem_type) { - case TTM_PL_TT: - r = amdgpu_ttm_bind(bo, mem); - if (r) - return r; + uint64_t addr = 0; - case TTM_PL_VRAM: - *addr = mm_node->start << PAGE_SHIFT; - *addr += bo->bdev->man[mem->mem_type].gpu_offset; - break; - default: - DRM_ERROR("Unknown placement %d\n", mem->mem_type); - return -EINVAL; + if (mem->mem_type != TTM_PL_TT || + amdgpu_gtt_mgr_is_allocated(mem)) { + addr = mm_node->start << PAGE_SHIFT; + addr += bo->bdev->man[mem->mem_type].gpu_offset; } - - return 0; + return addr; } static int amdgpu_move_blit(struct ttm_buffer_object *bo, @@ -299,26 +309,40 @@ static int amdgpu_move_blit(struct ttm_buffer_object *bo, } old_mm = old_mem->mm_node; - r = amdgpu_mm_node_addr(bo, old_mm, old_mem, &old_start); - if (r) - return r; old_size = old_mm->size; - + old_start = amdgpu_mm_node_addr(bo, old_mm, old_mem); new_mm = new_mem->mm_node; - r = amdgpu_mm_node_addr(bo, new_mm, new_mem, &new_start); - if (r) - return r; new_size = new_mm->size; + new_start = amdgpu_mm_node_addr(bo, new_mm, new_mem); num_pages = new_mem->num_pages; + mutex_lock(&adev->mman.gtt_window_lock); while (num_pages) { - unsigned long cur_pages = min(old_size, new_size); + unsigned long cur_pages = min(min(old_size, new_size), + (u64)AMDGPU_GTT_MAX_TRANSFER_SIZE); + uint64_t from = old_start, to = new_start; struct dma_fence *next; - r = amdgpu_copy_buffer(ring, old_start, new_start, + if (old_mem->mem_type == TTM_PL_TT && + !amdgpu_gtt_mgr_is_allocated(old_mem)) { + r = amdgpu_map_buffer(bo, old_mem, cur_pages, + old_start, 0, ring, &from); + if (r) + goto error; + } + + if (new_mem->mem_type == TTM_PL_TT && + !amdgpu_gtt_mgr_is_allocated(new_mem)) { + r = amdgpu_map_buffer(bo, new_mem, cur_pages, + new_start, 1, ring, &to); + if (r) + goto error; + } + + r = amdgpu_copy_buffer(ring, from, to, cur_pages * PAGE_SIZE, - bo->resv, &next, false); + bo->resv, &next, false, true); if (r) goto error; @@ -331,10 +355,7 @@ static int amdgpu_move_blit(struct ttm_buffer_object *bo, old_size -= cur_pages; if (!old_size) { - r = amdgpu_mm_node_addr(bo, ++old_mm, old_mem, - &old_start); - if (r) - goto error; + old_start = amdgpu_mm_node_addr(bo, ++old_mm, old_mem); old_size = old_mm->size; } else { old_start += cur_pages * PAGE_SIZE; @@ -342,22 +363,21 @@ static int amdgpu_move_blit(struct ttm_buffer_object *bo, new_size -= cur_pages; if (!new_size) { - r = amdgpu_mm_node_addr(bo, ++new_mm, new_mem, - &new_start); - if (r) - goto error; - + new_start = amdgpu_mm_node_addr(bo, ++new_mm, new_mem); new_size = new_mm->size; } else { new_start += cur_pages * PAGE_SIZE; } } + mutex_unlock(&adev->mman.gtt_window_lock); r = ttm_bo_pipeline_move(bo, fence, evict, new_mem); dma_fence_put(fence); return r; error: + mutex_unlock(&adev->mman.gtt_window_lock); + if (fence) dma_fence_wait(fence, false); dma_fence_put(fence); @@ -384,7 +404,7 @@ static int amdgpu_move_vram_ram(struct ttm_buffer_object *bo, placement.num_busy_placement = 1; placement.busy_placement = &placements; placements.fpfn = 0; - placements.lpfn = adev->mc.gtt_size >> PAGE_SHIFT; + placements.lpfn = 0; placements.flags = TTM_PL_MASK_CACHING | TTM_PL_FLAG_TT; r = ttm_bo_mem_space(bo, &placement, &tmp_mem, interruptible, no_wait_gpu); @@ -431,7 +451,7 @@ static int amdgpu_move_ram_vram(struct ttm_buffer_object *bo, placement.num_busy_placement = 1; placement.busy_placement = &placements; placements.fpfn = 0; - placements.lpfn = adev->mc.gtt_size >> PAGE_SHIFT; + placements.lpfn = 0; placements.flags = TTM_PL_MASK_CACHING | TTM_PL_FLAG_TT; r = ttm_bo_mem_space(bo, &placement, &tmp_mem, interruptible, no_wait_gpu); @@ -507,6 +527,15 @@ memcpy: } } + if (bo->type == ttm_bo_type_device && + new_mem->mem_type == TTM_PL_VRAM && + old_mem->mem_type != TTM_PL_VRAM) { + /* amdgpu_bo_fault_reserve_notify will re-set this if the CPU + * accesses the BO after it's moved. + */ + abo->flags &= ~AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED; + } + /* update statistics */ atomic64_add((u64)bo->num_pages << PAGE_SHIFT, &adev->num_bytes_moved); return 0; @@ -695,6 +724,31 @@ static void amdgpu_ttm_tt_unpin_userptr(struct ttm_tt *ttm) sg_free_table(ttm->sg); } +static int amdgpu_ttm_do_bind(struct ttm_tt *ttm, struct ttm_mem_reg *mem) +{ + struct amdgpu_ttm_tt *gtt = (void *)ttm; + uint64_t flags; + int r; + + spin_lock(>t->adev->gtt_list_lock); + flags = amdgpu_ttm_tt_pte_flags(gtt->adev, ttm, mem); + gtt->offset = (u64)mem->start << PAGE_SHIFT; + r = amdgpu_gart_bind(gtt->adev, gtt->offset, ttm->num_pages, + ttm->pages, gtt->ttm.dma_address, flags); + + if (r) { + DRM_ERROR("failed to bind %lu pages at 0x%08llX\n", + ttm->num_pages, gtt->offset); + goto error_gart_bind; + } + + list_add_tail(>t->list, >t->adev->gtt_list); +error_gart_bind: + spin_unlock(>t->adev->gtt_list_lock); + return r; + +} + static int amdgpu_ttm_backend_bind(struct ttm_tt *ttm, struct ttm_mem_reg *bo_mem) { @@ -718,7 +772,10 @@ static int amdgpu_ttm_backend_bind(struct ttm_tt *ttm, bo_mem->mem_type == AMDGPU_PL_OA) return -EINVAL; - return 0; + if (amdgpu_gtt_mgr_is_allocated(bo_mem)) + r = amdgpu_ttm_do_bind(ttm, bo_mem); + + return r; } bool amdgpu_ttm_is_bound(struct ttm_tt *ttm) @@ -731,8 +788,6 @@ bool amdgpu_ttm_is_bound(struct ttm_tt *ttm) int amdgpu_ttm_bind(struct ttm_buffer_object *bo, struct ttm_mem_reg *bo_mem) { struct ttm_tt *ttm = bo->ttm; - struct amdgpu_ttm_tt *gtt = (void *)bo->ttm; - uint64_t flags; int r; if (!ttm || amdgpu_ttm_is_bound(ttm)) @@ -745,22 +800,7 @@ int amdgpu_ttm_bind(struct ttm_buffer_object *bo, struct ttm_mem_reg *bo_mem) return r; } - spin_lock(>t->adev->gtt_list_lock); - flags = amdgpu_ttm_tt_pte_flags(gtt->adev, ttm, bo_mem); - gtt->offset = (u64)bo_mem->start << PAGE_SHIFT; - r = amdgpu_gart_bind(gtt->adev, gtt->offset, ttm->num_pages, - ttm->pages, gtt->ttm.dma_address, flags); - - if (r) { - DRM_ERROR("failed to bind %lu pages at 0x%08llX\n", - ttm->num_pages, gtt->offset); - goto error_gart_bind; - } - - list_add_tail(>t->list, >t->adev->gtt_list); -error_gart_bind: - spin_unlock(>t->adev->gtt_list_lock); - return r; + return amdgpu_ttm_do_bind(ttm, bo_mem); } int amdgpu_ttm_recover_gart(struct amdgpu_device *adev) @@ -1075,6 +1115,67 @@ static bool amdgpu_ttm_bo_eviction_valuable(struct ttm_buffer_object *bo, return ttm_bo_eviction_valuable(bo, place); } +static int amdgpu_ttm_access_memory(struct ttm_buffer_object *bo, + unsigned long offset, + void *buf, int len, int write) +{ + struct amdgpu_bo *abo = container_of(bo, struct amdgpu_bo, tbo); + struct amdgpu_device *adev = amdgpu_ttm_adev(abo->tbo.bdev); + struct drm_mm_node *nodes = abo->tbo.mem.mm_node; + uint32_t value = 0; + int ret = 0; + uint64_t pos; + unsigned long flags; + + if (bo->mem.mem_type != TTM_PL_VRAM) + return -EIO; + + while (offset >= (nodes->size << PAGE_SHIFT)) { + offset -= nodes->size << PAGE_SHIFT; + ++nodes; + } + pos = (nodes->start << PAGE_SHIFT) + offset; + + while (len && pos < adev->mc.mc_vram_size) { + uint64_t aligned_pos = pos & ~(uint64_t)3; + uint32_t bytes = 4 - (pos & 3); + uint32_t shift = (pos & 3) * 8; + uint32_t mask = 0xffffffff << shift; + + if (len < bytes) { + mask &= 0xffffffff >> (bytes - len) * 8; + bytes = len; + } + + spin_lock_irqsave(&adev->mmio_idx_lock, flags); + WREG32(mmMM_INDEX, ((uint32_t)aligned_pos) | 0x80000000); + WREG32(mmMM_INDEX_HI, aligned_pos >> 31); + if (!write || mask != 0xffffffff) + value = RREG32(mmMM_DATA); + if (write) { + value &= ~mask; + value |= (*(uint32_t *)buf << shift) & mask; + WREG32(mmMM_DATA, value); + } + spin_unlock_irqrestore(&adev->mmio_idx_lock, flags); + if (!write) { + value = (value & mask) >> shift; + memcpy(buf, &value, bytes); + } + + ret += bytes; + buf = (uint8_t *)buf + bytes; + pos += bytes; + len -= bytes; + if (pos >= (nodes->start + nodes->size) << PAGE_SHIFT) { + ++nodes; + pos = (nodes->start << PAGE_SHIFT); + } + } + + return ret; +} + static struct ttm_bo_driver amdgpu_bo_driver = { .ttm_tt_create = &amdgpu_ttm_tt_create, .ttm_tt_populate = &amdgpu_ttm_tt_populate, @@ -1090,11 +1191,14 @@ static struct ttm_bo_driver amdgpu_bo_driver = { .io_mem_reserve = &amdgpu_ttm_io_mem_reserve, .io_mem_free = &amdgpu_ttm_io_mem_free, .io_mem_pfn = amdgpu_ttm_io_mem_pfn, + .access_memory = &amdgpu_ttm_access_memory }; int amdgpu_ttm_init(struct amdgpu_device *adev) { + uint64_t gtt_size; int r; + u64 vis_vram_limit; r = amdgpu_ttm_global_init(adev); if (r) { @@ -1118,6 +1222,13 @@ int amdgpu_ttm_init(struct amdgpu_device *adev) DRM_ERROR("Failed initializing VRAM heap.\n"); return r; } + + /* Reduce size of CPU-visible VRAM if requested */ + vis_vram_limit = (u64)amdgpu_vis_vram_limit * 1024 * 1024; + if (amdgpu_vis_vram_limit > 0 && + vis_vram_limit <= adev->mc.visible_vram_size) + adev->mc.visible_vram_size = vis_vram_limit; + /* Change the size here instead of the init above so only lpfn is affected */ amdgpu_ttm_set_active_vram_size(adev, adev->mc.visible_vram_size); @@ -1140,14 +1251,19 @@ int amdgpu_ttm_init(struct amdgpu_device *adev) } DRM_INFO("amdgpu: %uM of VRAM memory ready\n", (unsigned) (adev->mc.real_vram_size / (1024 * 1024))); - r = ttm_bo_init_mm(&adev->mman.bdev, TTM_PL_TT, - adev->mc.gtt_size >> PAGE_SHIFT); + + if (amdgpu_gtt_size == -1) + gtt_size = max((AMDGPU_DEFAULT_GTT_SIZE_MB << 20), + adev->mc.mc_vram_size); + else + gtt_size = (uint64_t)amdgpu_gtt_size << 20; + r = ttm_bo_init_mm(&adev->mman.bdev, TTM_PL_TT, gtt_size >> PAGE_SHIFT); if (r) { DRM_ERROR("Failed initializing GTT heap.\n"); return r; } DRM_INFO("amdgpu: %uM of GTT memory ready.\n", - (unsigned)(adev->mc.gtt_size / (1024 * 1024))); + (unsigned)(gtt_size / (1024 * 1024))); adev->gds.mem.total_size = adev->gds.mem.total_size << AMDGPU_GDS_SHIFT; adev->gds.mem.gfx_partition_size = adev->gds.mem.gfx_partition_size << AMDGPU_GDS_SHIFT; @@ -1256,12 +1372,77 @@ int amdgpu_mmap(struct file *filp, struct vm_area_struct *vma) return ttm_bo_mmap(filp, vma, &adev->mman.bdev); } -int amdgpu_copy_buffer(struct amdgpu_ring *ring, - uint64_t src_offset, - uint64_t dst_offset, - uint32_t byte_count, +static int amdgpu_map_buffer(struct ttm_buffer_object *bo, + struct ttm_mem_reg *mem, unsigned num_pages, + uint64_t offset, unsigned window, + struct amdgpu_ring *ring, + uint64_t *addr) +{ + struct amdgpu_ttm_tt *gtt = (void *)bo->ttm; + struct amdgpu_device *adev = ring->adev; + struct ttm_tt *ttm = bo->ttm; + struct amdgpu_job *job; + unsigned num_dw, num_bytes; + dma_addr_t *dma_address; + struct dma_fence *fence; + uint64_t src_addr, dst_addr; + uint64_t flags; + int r; + + BUG_ON(adev->mman.buffer_funcs->copy_max_bytes < + AMDGPU_GTT_MAX_TRANSFER_SIZE * 8); + + *addr = adev->mc.gart_start; + *addr += (u64)window * AMDGPU_GTT_MAX_TRANSFER_SIZE * + AMDGPU_GPU_PAGE_SIZE; + + num_dw = adev->mman.buffer_funcs->copy_num_dw; + while (num_dw & 0x7) + num_dw++; + + num_bytes = num_pages * 8; + + r = amdgpu_job_alloc_with_ib(adev, num_dw * 4 + num_bytes, &job); + if (r) + return r; + + src_addr = num_dw * 4; + src_addr += job->ibs[0].gpu_addr; + + dst_addr = adev->gart.table_addr; + dst_addr += window * AMDGPU_GTT_MAX_TRANSFER_SIZE * 8; + amdgpu_emit_copy_buffer(adev, &job->ibs[0], src_addr, + dst_addr, num_bytes); + + amdgpu_ring_pad_ib(ring, &job->ibs[0]); + WARN_ON(job->ibs[0].length_dw > num_dw); + + dma_address = >t->ttm.dma_address[offset >> PAGE_SHIFT]; + flags = amdgpu_ttm_tt_pte_flags(adev, ttm, mem); + r = amdgpu_gart_map(adev, 0, num_pages, dma_address, flags, + &job->ibs[0].ptr[num_dw]); + if (r) + goto error_free; + + r = amdgpu_job_submit(job, ring, &adev->mman.entity, + AMDGPU_FENCE_OWNER_UNDEFINED, &fence); + if (r) + goto error_free; + + dma_fence_put(fence); + + return r; + +error_free: + amdgpu_job_free(job); + return r; +} + +int amdgpu_copy_buffer(struct amdgpu_ring *ring, uint64_t src_offset, + uint64_t dst_offset, uint32_t byte_count, struct reservation_object *resv, - struct dma_fence **fence, bool direct_submit) + struct dma_fence **fence, bool direct_submit, + bool vm_needs_flush) { struct amdgpu_device *adev = ring->adev; struct amdgpu_job *job; @@ -1283,6 +1464,7 @@ int amdgpu_copy_buffer(struct amdgpu_ring *ring, if (r) return r; + job->vm_needs_flush = vm_needs_flush; if (resv) { r = amdgpu_sync_resv(adev, &job->sync, resv, AMDGPU_FENCE_OWNER_UNDEFINED); @@ -1347,6 +1529,12 @@ int amdgpu_fill_buffer(struct amdgpu_bo *bo, return -EINVAL; } + if (bo->tbo.mem.mem_type == TTM_PL_TT) { + r = amdgpu_ttm_bind(&bo->tbo, &bo->tbo.mem); + if (r) + return r; + } + num_pages = bo->tbo.num_pages; mm_node = bo->tbo.mem.mm_node; num_loops = 0; @@ -1382,11 +1570,7 @@ int amdgpu_fill_buffer(struct amdgpu_bo *bo, uint32_t byte_count = mm_node->size << PAGE_SHIFT; uint64_t dst_addr; - r = amdgpu_mm_node_addr(&bo->tbo, mm_node, - &bo->tbo.mem, &dst_addr); - if (r) - return r; - + dst_addr = amdgpu_mm_node_addr(&bo->tbo, mm_node, &bo->tbo.mem); while (byte_count) { uint32_t cur_size_in_bytes = min(byte_count, max_bytes); @@ -1574,7 +1758,7 @@ static int amdgpu_ttm_debugfs_init(struct amdgpu_device *adev) adev, &amdgpu_ttm_gtt_fops); if (IS_ERR(ent)) return PTR_ERR(ent); - i_size_write(ent->d_inode, adev->mc.gtt_size); + i_size_write(ent->d_inode, adev->mc.gart_size); adev->mman.gtt = ent; #endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h index 6bdede8ff12b..f137c2458ee8 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h @@ -34,6 +34,9 @@ #define AMDGPU_PL_FLAG_GWS (TTM_PL_FLAG_PRIV << 1) #define AMDGPU_PL_FLAG_OA (TTM_PL_FLAG_PRIV << 2) +#define AMDGPU_GTT_MAX_TRANSFER_SIZE 512 +#define AMDGPU_GTT_NUM_TRANSFER_WINDOWS 2 + struct amdgpu_mman { struct ttm_bo_global_ref bo_global_ref; struct drm_global_reference mem_global_ref; @@ -49,6 +52,8 @@ struct amdgpu_mman { /* buffer handling */ const struct amdgpu_buffer_funcs *buffer_funcs; struct amdgpu_ring *buffer_funcs_ring; + + struct mutex gtt_window_lock; /* Scheduler entity for buffer moves */ struct amd_sched_entity entity; }; @@ -56,17 +61,17 @@ struct amdgpu_mman { extern const struct ttm_mem_type_manager_func amdgpu_gtt_mgr_func; extern const struct ttm_mem_type_manager_func amdgpu_vram_mgr_func; +bool amdgpu_gtt_mgr_is_allocated(struct ttm_mem_reg *mem); int amdgpu_gtt_mgr_alloc(struct ttm_mem_type_manager *man, struct ttm_buffer_object *tbo, const struct ttm_place *place, struct ttm_mem_reg *mem); -int amdgpu_copy_buffer(struct amdgpu_ring *ring, - uint64_t src_offset, - uint64_t dst_offset, - uint32_t byte_count, +int amdgpu_copy_buffer(struct amdgpu_ring *ring, uint64_t src_offset, + uint64_t dst_offset, uint32_t byte_count, struct reservation_object *resv, - struct dma_fence **fence, bool direct_submit); + struct dma_fence **fence, bool direct_submit, + bool vm_needs_flush); int amdgpu_fill_buffer(struct amdgpu_bo *bo, uint32_t src_data, struct reservation_object *resv, @@ -75,5 +80,6 @@ int amdgpu_fill_buffer(struct amdgpu_bo *bo, int amdgpu_mmap(struct file *filp, struct vm_area_struct *vma); bool amdgpu_ttm_is_bound(struct ttm_tt *ttm); int amdgpu_ttm_bind(struct ttm_buffer_object *bo, struct ttm_mem_reg *bo_mem); +int amdgpu_ttm_recover_gart(struct amdgpu_device *adev); #endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c index 4f50eeb65855..fcfb9d4f7477 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c @@ -275,14 +275,10 @@ amdgpu_ucode_get_load_type(struct amdgpu_device *adev, int load_type) else return AMDGPU_FW_LOAD_PSP; case CHIP_RAVEN: -#if 0 - if (!load_type) + if (load_type != 2) return AMDGPU_FW_LOAD_DIRECT; else return AMDGPU_FW_LOAD_PSP; -#else - return AMDGPU_FW_LOAD_DIRECT; -#endif default: DRM_ERROR("Unknow firmware load type\n"); } @@ -377,6 +373,11 @@ int amdgpu_ucode_init_bo(struct amdgpu_device *adev) struct amdgpu_firmware_info *ucode = NULL; const struct common_firmware_header *header = NULL; + if (!adev->firmware.fw_size) { + dev_warn(adev->dev, "No ip firmware need to load\n"); + return 0; + } + err = amdgpu_bo_create(adev, adev->firmware.fw_size, PAGE_SIZE, true, amdgpu_sriov_vf(adev) ? AMDGPU_GEM_DOMAIN_VRAM : AMDGPU_GEM_DOMAIN_GTT, AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS, @@ -459,6 +460,9 @@ int amdgpu_ucode_fini_bo(struct amdgpu_device *adev) int i; struct amdgpu_firmware_info *ucode = NULL; + if (!adev->firmware.fw_size) + return 0; + for (i = 0; i < adev->firmware.max_ucodes; i++) { ucode = &adev->firmware.ucode[i]; if (ucode->fw) { diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vf_error.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vf_error.c new file mode 100644 index 000000000000..45ac91861965 --- /dev/null +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vf_error.c @@ -0,0 +1,85 @@ +/* + * Copyright 2017 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + */ + +#include "amdgpu.h" +#include "amdgpu_vf_error.h" +#include "mxgpu_ai.h" + +#define AMDGPU_VF_ERROR_ENTRY_SIZE 16 + +/* struct error_entry - amdgpu VF error information. */ +struct amdgpu_vf_error_buffer { + int read_count; + int write_count; + uint16_t code[AMDGPU_VF_ERROR_ENTRY_SIZE]; + uint16_t flags[AMDGPU_VF_ERROR_ENTRY_SIZE]; + uint64_t data[AMDGPU_VF_ERROR_ENTRY_SIZE]; +}; + +struct amdgpu_vf_error_buffer admgpu_vf_errors; + + +void amdgpu_vf_error_put(uint16_t sub_error_code, uint16_t error_flags, uint64_t error_data) +{ + int index; + uint16_t error_code = AMDGIM_ERROR_CODE(AMDGIM_ERROR_CATEGORY_VF, sub_error_code); + + index = admgpu_vf_errors.write_count % AMDGPU_VF_ERROR_ENTRY_SIZE; + admgpu_vf_errors.code [index] = error_code; + admgpu_vf_errors.flags [index] = error_flags; + admgpu_vf_errors.data [index] = error_data; + admgpu_vf_errors.write_count ++; +} + + +void amdgpu_vf_error_trans_all(struct amdgpu_device *adev) +{ + /* u32 pf2vf_flags = 0; */ + u32 data1, data2, data3; + int index; + + if ((NULL == adev) || (!amdgpu_sriov_vf(adev)) || (!adev->virt.ops) || (!adev->virt.ops->trans_msg)) { + return; + } +/* + TODO: Enable these code when pv2vf_info is merged + AMDGPU_FW_VRAM_PF2VF_READ (adev, feature_flags, &pf2vf_flags); + if (!(pf2vf_flags & AMDGIM_FEATURE_ERROR_LOG_COLLECT)) { + return; + } +*/ + /* The errors are overlay of array, correct read_count as full. */ + if (admgpu_vf_errors.write_count - admgpu_vf_errors.read_count > AMDGPU_VF_ERROR_ENTRY_SIZE) { + admgpu_vf_errors.read_count = admgpu_vf_errors.write_count - AMDGPU_VF_ERROR_ENTRY_SIZE; + } + + while (admgpu_vf_errors.read_count < admgpu_vf_errors.write_count) { + index =admgpu_vf_errors.read_count % AMDGPU_VF_ERROR_ENTRY_SIZE; + data1 = AMDGIM_ERROR_CODE_FLAGS_TO_MAILBOX (admgpu_vf_errors.code[index], admgpu_vf_errors.flags[index]); + data2 = admgpu_vf_errors.data[index] & 0xFFFFFFFF; + data3 = (admgpu_vf_errors.data[index] >> 32) & 0xFFFFFFFF; + + adev->virt.ops->trans_msg(adev, IDH_LOG_VF_ERROR, data1, data2, data3); + admgpu_vf_errors.read_count ++; + } +} diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vf_error.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_vf_error.h new file mode 100644 index 000000000000..2a3278ec76ba --- /dev/null +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vf_error.h @@ -0,0 +1,62 @@ +/* + * Copyright 2017 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + */ + +#ifndef __VF_ERROR_H__ +#define __VF_ERROR_H__ + +#define AMDGIM_ERROR_CODE_FLAGS_TO_MAILBOX(c,f) (((c & 0xFFFF) << 16) | (f & 0xFFFF)) +#define AMDGIM_ERROR_CODE(t,c) (((t&0xF)<<12)|(c&0xFFF)) + +/* Please keep enum same as AMD GIM driver */ +enum AMDGIM_ERROR_VF { + AMDGIM_ERROR_VF_ATOMBIOS_INIT_FAIL = 0, + AMDGIM_ERROR_VF_NO_VBIOS, + AMDGIM_ERROR_VF_GPU_POST_ERROR, + AMDGIM_ERROR_VF_ATOMBIOS_GET_CLOCK_FAIL, + AMDGIM_ERROR_VF_FENCE_INIT_FAIL, + + AMDGIM_ERROR_VF_AMDGPU_INIT_FAIL, + AMDGIM_ERROR_VF_IB_INIT_FAIL, + AMDGIM_ERROR_VF_AMDGPU_LATE_INIT_FAIL, + AMDGIM_ERROR_VF_ASIC_RESUME_FAIL, + AMDGIM_ERROR_VF_GPU_RESET_FAIL, + + AMDGIM_ERROR_VF_TEST, + AMDGIM_ERROR_VF_MAX +}; + +enum AMDGIM_ERROR_CATEGORY { + AMDGIM_ERROR_CATEGORY_NON_USED = 0, + AMDGIM_ERROR_CATEGORY_GIM, + AMDGIM_ERROR_CATEGORY_PF, + AMDGIM_ERROR_CATEGORY_VF, + AMDGIM_ERROR_CATEGORY_VBIOS, + AMDGIM_ERROR_CATEGORY_MONITOR, + + AMDGIM_ERROR_CATEGORY_MAX +}; + +void amdgpu_vf_error_put(uint16_t sub_error_code, uint16_t error_flags, uint64_t error_data); +void amdgpu_vf_error_trans_all (struct amdgpu_device *adev); + +#endif /* __VF_ERROR_H__ */ diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.h index 9e1062edb76e..e5b1baf387c1 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.h @@ -43,6 +43,7 @@ struct amdgpu_virt_ops { int (*req_full_gpu)(struct amdgpu_device *adev, bool init); int (*rel_full_gpu)(struct amdgpu_device *adev, bool init); int (*reset_gpu)(struct amdgpu_device *adev); + void (*trans_msg)(struct amdgpu_device *adev, u32 req, u32 data1, u32 data2, u32 data3); }; /* GPU virtualization */ diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c index 5795f81369f0..250c8e80e646 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c @@ -77,8 +77,6 @@ struct amdgpu_pte_update_params { void (*func)(struct amdgpu_pte_update_params *params, uint64_t pe, uint64_t addr, unsigned count, uint32_t incr, uint64_t flags); - /* indicate update pt or its shadow */ - bool shadow; /* The next two are used during VM update by CPU * DMA addresses to use for mapping * Kernel pointer of PD/PT BO that needs to be updated @@ -161,11 +159,17 @@ void amdgpu_vm_get_pd_bo(struct amdgpu_vm *vm, */ static int amdgpu_vm_validate_level(struct amdgpu_vm_pt *parent, int (*validate)(void *, struct amdgpu_bo *), - void *param) + void *param, bool use_cpu_for_update) { unsigned i; int r; + if (use_cpu_for_update) { + r = amdgpu_bo_kmap(parent->bo, NULL); + if (r) + return r; + } + if (!parent->entries) return 0; @@ -183,7 +187,8 @@ static int amdgpu_vm_validate_level(struct amdgpu_vm_pt *parent, * Recurse into the sub directory. This is harmless because we * have only a maximum of 5 layers. */ - r = amdgpu_vm_validate_level(entry, validate, param); + r = amdgpu_vm_validate_level(entry, validate, param, + use_cpu_for_update); if (r) return r; } @@ -214,7 +219,8 @@ int amdgpu_vm_validate_pt_bos(struct amdgpu_device *adev, struct amdgpu_vm *vm, if (num_evictions == vm->last_eviction_counter) return 0; - return amdgpu_vm_validate_level(&vm->root, validate, param); + return amdgpu_vm_validate_level(&vm->root, validate, param, + vm->use_cpu_for_update); } /** @@ -331,6 +337,14 @@ static int amdgpu_vm_alloc_levels(struct amdgpu_device *adev, if (r) return r; + if (vm->use_cpu_for_update) { + r = amdgpu_bo_kmap(pt, NULL); + if (r) { + amdgpu_bo_unref(&pt); + return r; + } + } + /* Keep a reference to the root directory to avoid * freeing them up in the wrong order. */ @@ -338,6 +352,7 @@ static int amdgpu_vm_alloc_levels(struct amdgpu_device *adev, entry->bo = pt; entry->addr = 0; + entry->huge_page = false; } if (level < adev->vm_manager.num_level) { @@ -424,7 +439,7 @@ static int amdgpu_vm_grab_reserved_vmid_locked(struct amdgpu_vm *vm, struct dma_fence *updates = sync->last_vm_update; int r = 0; struct dma_fence *flushed, *tmp; - bool needs_flush = false; + bool needs_flush = vm->use_cpu_for_update; flushed = id->flushed_updates; if ((amdgpu_vm_had_gpu_reset(adev, id)) || @@ -545,11 +560,11 @@ int amdgpu_vm_grab_id(struct amdgpu_vm *vm, struct amdgpu_ring *ring, } kfree(fences); - job->vm_needs_flush = false; + job->vm_needs_flush = vm->use_cpu_for_update; /* Check if we can use a VMID already assigned to this VM */ list_for_each_entry_reverse(id, &id_mgr->ids_lru, list) { struct dma_fence *flushed; - bool needs_flush = false; + bool needs_flush = vm->use_cpu_for_update; /* Check all the prerequisites to using this VMID */ if (amdgpu_vm_had_gpu_reset(adev, id)) @@ -745,7 +760,7 @@ static bool amdgpu_vm_is_large_bar(struct amdgpu_device *adev) * * Emit a VM flush when it is necessary. */ -int amdgpu_vm_flush(struct amdgpu_ring *ring, struct amdgpu_job *job) +int amdgpu_vm_flush(struct amdgpu_ring *ring, struct amdgpu_job *job, bool need_pipe_sync) { struct amdgpu_device *adev = ring->adev; unsigned vmhub = ring->funcs->vmhub; @@ -767,12 +782,15 @@ int amdgpu_vm_flush(struct amdgpu_ring *ring, struct amdgpu_job *job) vm_flush_needed = true; } - if (!vm_flush_needed && !gds_switch_needed) + if (!vm_flush_needed && !gds_switch_needed && !need_pipe_sync) return 0; if (ring->funcs->init_cond_exec) patch_offset = amdgpu_ring_init_cond_exec(ring); + if (need_pipe_sync) + amdgpu_ring_emit_pipeline_sync(ring); + if (ring->funcs->emit_vm_flush && vm_flush_needed) { struct dma_fence *fence; @@ -981,6 +999,8 @@ static void amdgpu_vm_cpu_set_ptes(struct amdgpu_pte_update_params *params, unsigned int i; uint64_t value; + trace_amdgpu_vm_set_ptes(pe, addr, count, incr, flags); + for (i = 0; i < count; i++) { value = params->pages_addr ? amdgpu_vm_map_gart(params->pages_addr, addr) : @@ -989,19 +1009,16 @@ static void amdgpu_vm_cpu_set_ptes(struct amdgpu_pte_update_params *params, i, value, flags); addr += incr; } - - /* Flush HDP */ - mb(); - amdgpu_gart_flush_gpu_tlb(params->adev, 0); } -static int amdgpu_vm_bo_wait(struct amdgpu_device *adev, struct amdgpu_bo *bo) +static int amdgpu_vm_wait_pd(struct amdgpu_device *adev, struct amdgpu_vm *vm, + void *owner) { struct amdgpu_sync sync; int r; amdgpu_sync_create(&sync); - amdgpu_sync_resv(adev, &sync, bo->tbo.resv, AMDGPU_FENCE_OWNER_VM); + amdgpu_sync_resv(adev, &sync, vm->root.bo->tbo.resv, owner); r = amdgpu_sync_wait(&sync, true); amdgpu_sync_free(&sync); @@ -1042,16 +1059,12 @@ static int amdgpu_vm_update_level(struct amdgpu_device *adev, params.adev = adev; shadow = parent->bo->shadow; - WARN_ON(vm->use_cpu_for_update && shadow); - if (vm->use_cpu_for_update && !shadow) { - r = amdgpu_bo_kmap(parent->bo, (void **)&pd_addr); - if (r) - return r; - r = amdgpu_vm_bo_wait(adev, parent->bo); - if (unlikely(r)) { - amdgpu_bo_kunmap(parent->bo); + if (vm->use_cpu_for_update) { + pd_addr = (unsigned long)parent->bo->kptr; + r = amdgpu_vm_wait_pd(adev, vm, AMDGPU_FENCE_OWNER_VM); + if (unlikely(r)) return r; - } + params.func = amdgpu_vm_cpu_set_ptes; } else { if (shadow) { @@ -1105,7 +1118,8 @@ static int amdgpu_vm_update_level(struct amdgpu_device *adev, pt = amdgpu_bo_gpu_offset(bo); pt = amdgpu_gart_get_vm_pde(adev, pt); - if (parent->entries[pt_idx].addr == pt) + if (parent->entries[pt_idx].addr == pt || + parent->entries[pt_idx].huge_page) continue; parent->entries[pt_idx].addr = pt; @@ -1146,28 +1160,29 @@ static int amdgpu_vm_update_level(struct amdgpu_device *adev, count, incr, AMDGPU_PTE_VALID); } - if (params.func == amdgpu_vm_cpu_set_ptes) - amdgpu_bo_kunmap(parent->bo); - else if (params.ib->length_dw == 0) { - amdgpu_job_free(job); - } else { - amdgpu_ring_pad_ib(ring, params.ib); - amdgpu_sync_resv(adev, &job->sync, parent->bo->tbo.resv, - AMDGPU_FENCE_OWNER_VM); - if (shadow) - amdgpu_sync_resv(adev, &job->sync, shadow->tbo.resv, + if (!vm->use_cpu_for_update) { + if (params.ib->length_dw == 0) { + amdgpu_job_free(job); + } else { + amdgpu_ring_pad_ib(ring, params.ib); + amdgpu_sync_resv(adev, &job->sync, parent->bo->tbo.resv, AMDGPU_FENCE_OWNER_VM); + if (shadow) + amdgpu_sync_resv(adev, &job->sync, + shadow->tbo.resv, + AMDGPU_FENCE_OWNER_VM); + + WARN_ON(params.ib->length_dw > ndw); + r = amdgpu_job_submit(job, ring, &vm->entity, + AMDGPU_FENCE_OWNER_VM, &fence); + if (r) + goto error_free; - WARN_ON(params.ib->length_dw > ndw); - r = amdgpu_job_submit(job, ring, &vm->entity, - AMDGPU_FENCE_OWNER_VM, &fence); - if (r) - goto error_free; - - amdgpu_bo_fence(parent->bo, fence, true); - dma_fence_put(vm->last_dir_update); - vm->last_dir_update = dma_fence_get(fence); - dma_fence_put(fence); + amdgpu_bo_fence(parent->bo, fence, true); + dma_fence_put(vm->last_dir_update); + vm->last_dir_update = dma_fence_get(fence); + dma_fence_put(fence); + } } /* * Recurse into the subdirectories. This recursion is harmless because @@ -1235,33 +1250,105 @@ int amdgpu_vm_update_directories(struct amdgpu_device *adev, if (r) amdgpu_vm_invalidate_level(&vm->root); + if (vm->use_cpu_for_update) { + /* Flush HDP */ + mb(); + amdgpu_gart_flush_gpu_tlb(adev, 0); + } + return r; } /** - * amdgpu_vm_find_pt - find the page table for an address + * amdgpu_vm_find_entry - find the entry for an address * * @p: see amdgpu_pte_update_params definition * @addr: virtual address in question + * @entry: resulting entry or NULL + * @parent: parent entry * - * Find the page table BO for a virtual address, return NULL when none found. + * Find the vm_pt entry and it's parent for the given address. */ -static struct amdgpu_bo *amdgpu_vm_get_pt(struct amdgpu_pte_update_params *p, - uint64_t addr) +void amdgpu_vm_get_entry(struct amdgpu_pte_update_params *p, uint64_t addr, + struct amdgpu_vm_pt **entry, + struct amdgpu_vm_pt **parent) { - struct amdgpu_vm_pt *entry = &p->vm->root; unsigned idx, level = p->adev->vm_manager.num_level; - while (entry->entries) { + *parent = NULL; + *entry = &p->vm->root; + while ((*entry)->entries) { idx = addr >> (p->adev->vm_manager.block_size * level--); - idx %= amdgpu_bo_size(entry->bo) / 8; - entry = &entry->entries[idx]; + idx %= amdgpu_bo_size((*entry)->bo) / 8; + *parent = *entry; + *entry = &(*entry)->entries[idx]; } if (level) - return NULL; + *entry = NULL; +} + +/** + * amdgpu_vm_handle_huge_pages - handle updating the PD with huge pages + * + * @p: see amdgpu_pte_update_params definition + * @entry: vm_pt entry to check + * @parent: parent entry + * @nptes: number of PTEs updated with this operation + * @dst: destination address where the PTEs should point to + * @flags: access flags fro the PTEs + * + * Check if we can update the PD with a huge page. + */ +static int amdgpu_vm_handle_huge_pages(struct amdgpu_pte_update_params *p, + struct amdgpu_vm_pt *entry, + struct amdgpu_vm_pt *parent, + unsigned nptes, uint64_t dst, + uint64_t flags) +{ + bool use_cpu_update = (p->func == amdgpu_vm_cpu_set_ptes); + uint64_t pd_addr, pde; + int r; - return entry->bo; + /* In the case of a mixed PT the PDE must point to it*/ + if (p->adev->asic_type < CHIP_VEGA10 || + nptes != AMDGPU_VM_PTE_COUNT(p->adev) || + p->func == amdgpu_vm_do_copy_ptes || + !(flags & AMDGPU_PTE_VALID)) { + + dst = amdgpu_bo_gpu_offset(entry->bo); + dst = amdgpu_gart_get_vm_pde(p->adev, dst); + flags = AMDGPU_PTE_VALID; + } else { + flags |= AMDGPU_PDE_PTE; + } + + if (entry->addr == dst && + entry->huge_page == !!(flags & AMDGPU_PDE_PTE)) + return 0; + + entry->addr = dst; + entry->huge_page = !!(flags & AMDGPU_PDE_PTE); + + if (use_cpu_update) { + r = amdgpu_bo_kmap(parent->bo, (void *)&pd_addr); + if (r) + return r; + + pde = pd_addr + (entry - parent->entries) * 8; + amdgpu_vm_cpu_set_ptes(p, pde, dst, 1, 0, flags); + } else { + if (parent->bo->shadow) { + pd_addr = amdgpu_bo_gpu_offset(parent->bo->shadow); + pde = pd_addr + (entry - parent->entries) * 8; + amdgpu_vm_do_set_ptes(p, pde, dst, 1, 0, flags); + } + pd_addr = amdgpu_bo_gpu_offset(parent->bo); + pde = pd_addr + (entry - parent->entries) * 8; + amdgpu_vm_do_set_ptes(p, pde, dst, 1, 0, flags); + } + + return 0; } /** @@ -1287,49 +1374,47 @@ static int amdgpu_vm_update_ptes(struct amdgpu_pte_update_params *params, uint64_t addr, pe_start; struct amdgpu_bo *pt; unsigned nptes; - int r; bool use_cpu_update = (params->func == amdgpu_vm_cpu_set_ptes); - + int r; /* walk over the address space and update the page tables */ - for (addr = start; addr < end; addr += nptes) { - pt = amdgpu_vm_get_pt(params, addr); - if (!pt) { - pr_err("PT not found, aborting update_ptes\n"); - return -EINVAL; - } + for (addr = start; addr < end; addr += nptes, + dst += nptes * AMDGPU_GPU_PAGE_SIZE) { + struct amdgpu_vm_pt *entry, *parent; - if (params->shadow) { - if (WARN_ONCE(use_cpu_update, - "CPU VM update doesn't suuport shadow pages")) - return 0; - - if (!pt->shadow) - return 0; - pt = pt->shadow; - } + amdgpu_vm_get_entry(params, addr, &entry, &parent); + if (!entry) + return -ENOENT; if ((addr & ~mask) == (end & ~mask)) nptes = end - addr; else nptes = AMDGPU_VM_PTE_COUNT(adev) - (addr & mask); + r = amdgpu_vm_handle_huge_pages(params, entry, parent, + nptes, dst, flags); + if (r) + return r; + + if (entry->huge_page) + continue; + + pt = entry->bo; if (use_cpu_update) { - r = amdgpu_bo_kmap(pt, (void *)&pe_start); - if (r) - return r; - } else + pe_start = (unsigned long)pt->kptr; + } else { + if (pt->shadow) { + pe_start = amdgpu_bo_gpu_offset(pt->shadow); + pe_start += (addr & mask) * 8; + params->func(params, pe_start, dst, nptes, + AMDGPU_GPU_PAGE_SIZE, flags); + } pe_start = amdgpu_bo_gpu_offset(pt); + } pe_start += (addr & mask) * 8; - params->func(params, pe_start, dst, nptes, AMDGPU_GPU_PAGE_SIZE, flags); - - dst += nptes * AMDGPU_GPU_PAGE_SIZE; - - if (use_cpu_update) - amdgpu_bo_kunmap(pt); } return 0; @@ -1372,8 +1457,9 @@ static int amdgpu_vm_frag_ptes(struct amdgpu_pte_update_params *params, */ /* SI and newer are optimized for 64KB */ - uint64_t frag_flags = AMDGPU_PTE_FRAG(AMDGPU_LOG2_PAGES_PER_FRAG); - uint64_t frag_align = 1 << AMDGPU_LOG2_PAGES_PER_FRAG; + unsigned pages_per_frag = AMDGPU_LOG2_PAGES_PER_FRAG(params->adev); + uint64_t frag_flags = AMDGPU_PTE_FRAG(pages_per_frag); + uint64_t frag_align = 1 << pages_per_frag; uint64_t frag_start = ALIGN(start, frag_align); uint64_t frag_end = end & ~(frag_align - 1); @@ -1445,6 +1531,10 @@ static int amdgpu_vm_bo_update_mapping(struct amdgpu_device *adev, params.vm = vm; params.src = src; + /* sync to everything on unmapping */ + if (!(flags & AMDGPU_PTE_VALID)) + owner = AMDGPU_FENCE_OWNER_UNDEFINED; + if (vm->use_cpu_for_update) { /* params.src is used as flag to indicate system Memory */ if (pages_addr) @@ -1453,23 +1543,18 @@ static int amdgpu_vm_bo_update_mapping(struct amdgpu_device *adev, /* Wait for PT BOs to be free. PTs share the same resv. object * as the root PD BO */ - r = amdgpu_vm_bo_wait(adev, vm->root.bo); + r = amdgpu_vm_wait_pd(adev, vm, owner); if (unlikely(r)) return r; params.func = amdgpu_vm_cpu_set_ptes; params.pages_addr = pages_addr; - params.shadow = false; return amdgpu_vm_frag_ptes(¶ms, start, last + 1, addr, flags); } ring = container_of(vm->entity.sched, struct amdgpu_ring, sched); - /* sync to everything on unmapping */ - if (!(flags & AMDGPU_PTE_VALID)) - owner = AMDGPU_FENCE_OWNER_UNDEFINED; - nptes = last - start + 1; /* @@ -1481,6 +1566,9 @@ static int amdgpu_vm_bo_update_mapping(struct amdgpu_device *adev, /* padding, etc. */ ndw = 64; + /* one PDE write for each huge page */ + ndw += ((nptes >> adev->vm_manager.block_size) + 1) * 6; + if (src) { /* only copy commands needed */ ndw += ncmds * 7; @@ -1542,11 +1630,6 @@ static int amdgpu_vm_bo_update_mapping(struct amdgpu_device *adev, if (r) goto error_free; - params.shadow = true; - r = amdgpu_vm_frag_ptes(¶ms, start, last + 1, addr, flags); - if (r) - goto error_free; - params.shadow = false; r = amdgpu_vm_frag_ptes(¶ms, start, last + 1, addr, flags); if (r) goto error_free; @@ -1565,6 +1648,7 @@ static int amdgpu_vm_bo_update_mapping(struct amdgpu_device *adev, error_free: amdgpu_job_free(job); + amdgpu_vm_invalidate_level(&vm->root); return r; } @@ -1752,6 +1836,12 @@ int amdgpu_vm_bo_update(struct amdgpu_device *adev, list_add(&bo_va->vm_status, &vm->cleared); spin_unlock(&vm->status_lock); + if (vm->use_cpu_for_update) { + /* Flush HDP */ + mb(); + amdgpu_gart_flush_gpu_tlb(adev, 0); + } + return 0; } @@ -2457,6 +2547,13 @@ int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm, goto error_free_root; vm->last_eviction_counter = atomic64_read(&adev->num_evictions); + + if (vm->use_cpu_for_update) { + r = amdgpu_bo_kmap(vm->root.bo, NULL); + if (r) + goto error_free_root; + } + amdgpu_bo_unreserve(vm->root.bo); return 0; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h index 936f158bc5ec..34d9174ebff2 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h @@ -51,7 +51,9 @@ struct amdgpu_bo_list_entry; #define AMDGPU_VM_PTB_ALIGN_SIZE 32768 /* LOG2 number of continuous pages for the fragment field */ -#define AMDGPU_LOG2_PAGES_PER_FRAG 4 +#define AMDGPU_LOG2_PAGES_PER_FRAG(adev) \ + ((adev)->asic_type < CHIP_VEGA10 ? 4 : \ + (adev)->vm_manager.block_size) #define AMDGPU_PTE_VALID (1ULL << 0) #define AMDGPU_PTE_SYSTEM (1ULL << 1) @@ -68,6 +70,9 @@ struct amdgpu_bo_list_entry; /* TILED for VEGA10, reserved for older ASICs */ #define AMDGPU_PTE_PRT (1ULL << 51) +/* PDE is handled as PTE for VEGA10 */ +#define AMDGPU_PDE_PTE (1ULL << 54) + /* VEGA10 only */ #define AMDGPU_PTE_MTYPE(a) ((uint64_t)a << 57) #define AMDGPU_PTE_MTYPE_MASK AMDGPU_PTE_MTYPE(3ULL) @@ -98,6 +103,7 @@ struct amdgpu_bo_list_entry; struct amdgpu_vm_pt { struct amdgpu_bo *bo; uint64_t addr; + bool huge_page; /* array of page tables, one for each directory entry */ struct amdgpu_vm_pt *entries; @@ -222,7 +228,7 @@ int amdgpu_vm_alloc_pts(struct amdgpu_device *adev, 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); +int amdgpu_vm_flush(struct amdgpu_ring *ring, struct amdgpu_job *job, bool need_pipe_sync); void amdgpu_vm_reset_id(struct amdgpu_device *adev, unsigned vmhub, unsigned vmid); void amdgpu_vm_reset_all_ids(struct amdgpu_device *adev); diff --git a/drivers/gpu/drm/amd/amdgpu/cik.c b/drivers/gpu/drm/amd/amdgpu/cik.c index 37a499ab30eb..567c4a5cf90c 100644 --- a/drivers/gpu/drm/amd/amdgpu/cik.c +++ b/drivers/gpu/drm/amd/amdgpu/cik.c @@ -1824,21 +1824,14 @@ static int cik_common_suspend(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; - amdgpu_amdkfd_suspend(adev); - return cik_common_hw_fini(adev); } static int cik_common_resume(void *handle) { - int r; struct amdgpu_device *adev = (struct amdgpu_device *)handle; - r = cik_common_hw_init(adev); - if (r) - return r; - - return amdgpu_amdkfd_resume(adev); + return cik_common_hw_init(adev); } static bool cik_common_is_idle(void *handle) diff --git a/drivers/gpu/drm/amd/amdgpu/cik_sdma.c b/drivers/gpu/drm/amd/amdgpu/cik_sdma.c index c216e16826c9..f508f4d01e4a 100644 --- a/drivers/gpu/drm/amd/amdgpu/cik_sdma.c +++ b/drivers/gpu/drm/amd/amdgpu/cik_sdma.c @@ -342,6 +342,63 @@ static void cik_sdma_rlc_stop(struct amdgpu_device *adev) } /** + * cik_ctx_switch_enable - stop the async dma engines context switch + * + * @adev: amdgpu_device pointer + * @enable: enable/disable the DMA MEs context switch. + * + * Halt or unhalt the async dma engines context switch (VI). + */ +static void cik_ctx_switch_enable(struct amdgpu_device *adev, bool enable) +{ + u32 f32_cntl, phase_quantum = 0; + int i; + + if (amdgpu_sdma_phase_quantum) { + unsigned value = amdgpu_sdma_phase_quantum; + unsigned unit = 0; + + while (value > (SDMA0_PHASE0_QUANTUM__VALUE_MASK >> + SDMA0_PHASE0_QUANTUM__VALUE__SHIFT)) { + value = (value + 1) >> 1; + unit++; + } + if (unit > (SDMA0_PHASE0_QUANTUM__UNIT_MASK >> + SDMA0_PHASE0_QUANTUM__UNIT__SHIFT)) { + value = (SDMA0_PHASE0_QUANTUM__VALUE_MASK >> + SDMA0_PHASE0_QUANTUM__VALUE__SHIFT); + unit = (SDMA0_PHASE0_QUANTUM__UNIT_MASK >> + SDMA0_PHASE0_QUANTUM__UNIT__SHIFT); + WARN_ONCE(1, + "clamping sdma_phase_quantum to %uK clock cycles\n", + value << unit); + } + phase_quantum = + value << SDMA0_PHASE0_QUANTUM__VALUE__SHIFT | + unit << SDMA0_PHASE0_QUANTUM__UNIT__SHIFT; + } + + for (i = 0; i < adev->sdma.num_instances; i++) { + f32_cntl = RREG32(mmSDMA0_CNTL + sdma_offsets[i]); + if (enable) { + f32_cntl = REG_SET_FIELD(f32_cntl, SDMA0_CNTL, + AUTO_CTXSW_ENABLE, 1); + if (amdgpu_sdma_phase_quantum) { + WREG32(mmSDMA0_PHASE0_QUANTUM + sdma_offsets[i], + phase_quantum); + WREG32(mmSDMA0_PHASE1_QUANTUM + sdma_offsets[i], + phase_quantum); + } + } else { + f32_cntl = REG_SET_FIELD(f32_cntl, SDMA0_CNTL, + AUTO_CTXSW_ENABLE, 0); + } + + WREG32(mmSDMA0_CNTL + sdma_offsets[i], f32_cntl); + } +} + +/** * cik_sdma_enable - stop the async dma engines * * @adev: amdgpu_device pointer @@ -537,6 +594,8 @@ static int cik_sdma_start(struct amdgpu_device *adev) /* halt the engine before programing */ cik_sdma_enable(adev, false); + /* enable sdma ring preemption */ + cik_ctx_switch_enable(adev, true); /* start the gfx rings and rlc compute queues */ r = cik_sdma_gfx_resume(adev); @@ -984,6 +1043,7 @@ static int cik_sdma_hw_fini(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; + cik_ctx_switch_enable(adev, false); cik_sdma_enable(adev, false); return 0; diff --git a/drivers/gpu/drm/amd/amdgpu/clearstate_gfx9.h b/drivers/gpu/drm/amd/amdgpu/clearstate_gfx9.h index 18fd01f3e4b2..003a131bad47 100644 --- a/drivers/gpu/drm/amd/amdgpu/clearstate_gfx9.h +++ b/drivers/gpu/drm/amd/amdgpu/clearstate_gfx9.h @@ -1,24 +1,25 @@ - /* -*************************************************************************************************** -* -* Trade secret of Advanced Micro Devices, Inc. -* Copyright (c) 2010 Advanced Micro Devices, Inc. (unpublished) -* -* All rights reserved. This notice is intended as a precaution against inadvertent publication and -* does not imply publication or any waiver of confidentiality. The year included in the foregoing -* notice is the year of creation of the work. -* -*************************************************************************************************** -*/ -/** -*************************************************************************************************** -* @brief gfx9 Clearstate Definitions -*************************************************************************************************** -* -* Do not edit! This is a machine-generated file! -* -*/ + * Copyright 2017 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + */ static const unsigned int gfx9_SECT_CONTEXT_def_1[] = { diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c index 162d761e2f4e..490e84944851 100644 --- a/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c +++ b/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c @@ -484,134 +484,6 @@ static bool dce_v10_0_is_display_hung(struct amdgpu_device *adev) return true; } -static void dce_v10_0_stop_mc_access(struct amdgpu_device *adev, - struct amdgpu_mode_mc_save *save) -{ - u32 crtc_enabled, tmp; - int i; - - save->vga_render_control = RREG32(mmVGA_RENDER_CONTROL); - save->vga_hdp_control = RREG32(mmVGA_HDP_CONTROL); - - /* disable VGA render */ - tmp = RREG32(mmVGA_RENDER_CONTROL); - tmp = REG_SET_FIELD(tmp, VGA_RENDER_CONTROL, VGA_VSTATUS_CNTL, 0); - WREG32(mmVGA_RENDER_CONTROL, tmp); - - /* blank the display controllers */ - for (i = 0; i < adev->mode_info.num_crtc; i++) { - crtc_enabled = REG_GET_FIELD(RREG32(mmCRTC_CONTROL + crtc_offsets[i]), - CRTC_CONTROL, CRTC_MASTER_EN); - if (crtc_enabled) { -#if 0 - u32 frame_count; - int j; - - save->crtc_enabled[i] = true; - tmp = RREG32(mmCRTC_BLANK_CONTROL + crtc_offsets[i]); - if (REG_GET_FIELD(tmp, CRTC_BLANK_CONTROL, CRTC_BLANK_DATA_EN) == 0) { - amdgpu_display_vblank_wait(adev, i); - WREG32(mmCRTC_UPDATE_LOCK + crtc_offsets[i], 1); - tmp = REG_SET_FIELD(tmp, CRTC_BLANK_CONTROL, CRTC_BLANK_DATA_EN, 1); - WREG32(mmCRTC_BLANK_CONTROL + crtc_offsets[i], tmp); - WREG32(mmCRTC_UPDATE_LOCK + crtc_offsets[i], 0); - } - /* wait for the next frame */ - frame_count = amdgpu_display_vblank_get_counter(adev, i); - for (j = 0; j < adev->usec_timeout; j++) { - if (amdgpu_display_vblank_get_counter(adev, i) != frame_count) - break; - udelay(1); - } - tmp = RREG32(mmGRPH_UPDATE + crtc_offsets[i]); - if (REG_GET_FIELD(tmp, GRPH_UPDATE, GRPH_UPDATE_LOCK) == 0) { - tmp = REG_SET_FIELD(tmp, GRPH_UPDATE, GRPH_UPDATE_LOCK, 1); - WREG32(mmGRPH_UPDATE + crtc_offsets[i], tmp); - } - tmp = RREG32(mmMASTER_UPDATE_LOCK + crtc_offsets[i]); - if (REG_GET_FIELD(tmp, MASTER_UPDATE_LOCK, MASTER_UPDATE_LOCK) == 0) { - tmp = REG_SET_FIELD(tmp, MASTER_UPDATE_LOCK, MASTER_UPDATE_LOCK, 1); - WREG32(mmMASTER_UPDATE_LOCK + crtc_offsets[i], tmp); - } -#else - /* XXX this is a hack to avoid strange behavior with EFI on certain systems */ - WREG32(mmCRTC_UPDATE_LOCK + crtc_offsets[i], 1); - tmp = RREG32(mmCRTC_CONTROL + crtc_offsets[i]); - tmp = REG_SET_FIELD(tmp, CRTC_CONTROL, CRTC_MASTER_EN, 0); - WREG32(mmCRTC_CONTROL + crtc_offsets[i], tmp); - WREG32(mmCRTC_UPDATE_LOCK + crtc_offsets[i], 0); - save->crtc_enabled[i] = false; - /* ***** */ -#endif - } else { - save->crtc_enabled[i] = false; - } - } -} - -static void dce_v10_0_resume_mc_access(struct amdgpu_device *adev, - struct amdgpu_mode_mc_save *save) -{ - u32 tmp, frame_count; - int i, j; - - /* update crtc base addresses */ - for (i = 0; i < adev->mode_info.num_crtc; i++) { - WREG32(mmGRPH_PRIMARY_SURFACE_ADDRESS_HIGH + crtc_offsets[i], - upper_32_bits(adev->mc.vram_start)); - WREG32(mmGRPH_SECONDARY_SURFACE_ADDRESS_HIGH + crtc_offsets[i], - upper_32_bits(adev->mc.vram_start)); - WREG32(mmGRPH_PRIMARY_SURFACE_ADDRESS + crtc_offsets[i], - (u32)adev->mc.vram_start); - WREG32(mmGRPH_SECONDARY_SURFACE_ADDRESS + crtc_offsets[i], - (u32)adev->mc.vram_start); - - if (save->crtc_enabled[i]) { - tmp = RREG32(mmMASTER_UPDATE_MODE + crtc_offsets[i]); - if (REG_GET_FIELD(tmp, MASTER_UPDATE_MODE, MASTER_UPDATE_MODE) != 0) { - tmp = REG_SET_FIELD(tmp, MASTER_UPDATE_MODE, MASTER_UPDATE_MODE, 0); - WREG32(mmMASTER_UPDATE_MODE + crtc_offsets[i], tmp); - } - tmp = RREG32(mmGRPH_UPDATE + crtc_offsets[i]); - if (REG_GET_FIELD(tmp, GRPH_UPDATE, GRPH_UPDATE_LOCK)) { - tmp = REG_SET_FIELD(tmp, GRPH_UPDATE, GRPH_UPDATE_LOCK, 0); - WREG32(mmGRPH_UPDATE + crtc_offsets[i], tmp); - } - tmp = RREG32(mmMASTER_UPDATE_LOCK + crtc_offsets[i]); - if (REG_GET_FIELD(tmp, MASTER_UPDATE_LOCK, MASTER_UPDATE_LOCK)) { - tmp = REG_SET_FIELD(tmp, MASTER_UPDATE_LOCK, MASTER_UPDATE_LOCK, 0); - WREG32(mmMASTER_UPDATE_LOCK + crtc_offsets[i], tmp); - } - for (j = 0; j < adev->usec_timeout; j++) { - tmp = RREG32(mmGRPH_UPDATE + crtc_offsets[i]); - if (REG_GET_FIELD(tmp, GRPH_UPDATE, GRPH_SURFACE_UPDATE_PENDING) == 0) - break; - udelay(1); - } - tmp = RREG32(mmCRTC_BLANK_CONTROL + crtc_offsets[i]); - tmp = REG_SET_FIELD(tmp, CRTC_BLANK_CONTROL, CRTC_BLANK_DATA_EN, 0); - WREG32(mmCRTC_UPDATE_LOCK + crtc_offsets[i], 1); - WREG32(mmCRTC_BLANK_CONTROL + crtc_offsets[i], tmp); - WREG32(mmCRTC_UPDATE_LOCK + crtc_offsets[i], 0); - /* wait for the next frame */ - frame_count = amdgpu_display_vblank_get_counter(adev, i); - for (j = 0; j < adev->usec_timeout; j++) { - if (amdgpu_display_vblank_get_counter(adev, i) != frame_count) - break; - udelay(1); - } - } - } - - WREG32(mmVGA_MEMORY_BASE_ADDRESS_HIGH, upper_32_bits(adev->mc.vram_start)); - WREG32(mmVGA_MEMORY_BASE_ADDRESS, lower_32_bits(adev->mc.vram_start)); - - /* Unlock vga access */ - WREG32(mmVGA_HDP_CONTROL, save->vga_hdp_control); - mdelay(1); - WREG32(mmVGA_RENDER_CONTROL, save->vga_render_control); -} - static void dce_v10_0_set_vga_render_state(struct amdgpu_device *adev, bool render) { @@ -3012,6 +2884,8 @@ static int dce_v10_0_hw_init(void *handle) dce_v10_0_init_golden_registers(adev); + /* disable vga render */ + dce_v10_0_set_vga_render_state(adev, false); /* init dig PHYs, disp eng pll */ amdgpu_atombios_encoder_init_dig(adev); amdgpu_atombios_crtc_set_disp_eng_pll(adev, adev->clock.default_dispclk); @@ -3724,7 +3598,6 @@ static void dce_v10_0_encoder_add(struct amdgpu_device *adev, } static const struct amdgpu_display_funcs dce_v10_0_display_funcs = { - .set_vga_render_state = &dce_v10_0_set_vga_render_state, .bandwidth_update = &dce_v10_0_bandwidth_update, .vblank_get_counter = &dce_v10_0_vblank_get_counter, .vblank_wait = &dce_v10_0_vblank_wait, @@ -3737,8 +3610,6 @@ static const struct amdgpu_display_funcs dce_v10_0_display_funcs = { .page_flip_get_scanoutpos = &dce_v10_0_crtc_get_scanoutpos, .add_encoder = &dce_v10_0_encoder_add, .add_connector = &amdgpu_connector_add, - .stop_mc_access = &dce_v10_0_stop_mc_access, - .resume_mc_access = &dce_v10_0_resume_mc_access, }; static void dce_v10_0_set_display_funcs(struct amdgpu_device *adev) diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c index f420e5815edb..921c6f772f11 100644 --- a/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c +++ b/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c @@ -499,79 +499,6 @@ static bool dce_v11_0_is_display_hung(struct amdgpu_device *adev) return true; } -static void dce_v11_0_stop_mc_access(struct amdgpu_device *adev, - struct amdgpu_mode_mc_save *save) -{ - u32 crtc_enabled, tmp; - int i; - - save->vga_render_control = RREG32(mmVGA_RENDER_CONTROL); - save->vga_hdp_control = RREG32(mmVGA_HDP_CONTROL); - - /* disable VGA render */ - tmp = RREG32(mmVGA_RENDER_CONTROL); - tmp = REG_SET_FIELD(tmp, VGA_RENDER_CONTROL, VGA_VSTATUS_CNTL, 0); - WREG32(mmVGA_RENDER_CONTROL, tmp); - - /* blank the display controllers */ - for (i = 0; i < adev->mode_info.num_crtc; i++) { - crtc_enabled = REG_GET_FIELD(RREG32(mmCRTC_CONTROL + crtc_offsets[i]), - CRTC_CONTROL, CRTC_MASTER_EN); - if (crtc_enabled) { -#if 1 - save->crtc_enabled[i] = true; - tmp = RREG32(mmCRTC_BLANK_CONTROL + crtc_offsets[i]); - if (REG_GET_FIELD(tmp, CRTC_BLANK_CONTROL, CRTC_BLANK_DATA_EN) == 0) { - /*it is correct only for RGB ; black is 0*/ - WREG32(mmCRTC_BLANK_DATA_COLOR + crtc_offsets[i], 0); - tmp = REG_SET_FIELD(tmp, CRTC_BLANK_CONTROL, CRTC_BLANK_DATA_EN, 1); - WREG32(mmCRTC_BLANK_CONTROL + crtc_offsets[i], tmp); - } -#else - /* XXX this is a hack to avoid strange behavior with EFI on certain systems */ - WREG32(mmCRTC_UPDATE_LOCK + crtc_offsets[i], 1); - tmp = RREG32(mmCRTC_CONTROL + crtc_offsets[i]); - tmp = REG_SET_FIELD(tmp, CRTC_CONTROL, CRTC_MASTER_EN, 0); - WREG32(mmCRTC_CONTROL + crtc_offsets[i], tmp); - WREG32(mmCRTC_UPDATE_LOCK + crtc_offsets[i], 0); - save->crtc_enabled[i] = false; - /* ***** */ -#endif - } else { - save->crtc_enabled[i] = false; - } - } -} - -static void dce_v11_0_resume_mc_access(struct amdgpu_device *adev, - struct amdgpu_mode_mc_save *save) -{ - u32 tmp; - int i; - - /* update crtc base addresses */ - for (i = 0; i < adev->mode_info.num_crtc; i++) { - WREG32(mmGRPH_PRIMARY_SURFACE_ADDRESS_HIGH + crtc_offsets[i], - upper_32_bits(adev->mc.vram_start)); - WREG32(mmGRPH_PRIMARY_SURFACE_ADDRESS + crtc_offsets[i], - (u32)adev->mc.vram_start); - - if (save->crtc_enabled[i]) { - tmp = RREG32(mmCRTC_BLANK_CONTROL + crtc_offsets[i]); - tmp = REG_SET_FIELD(tmp, CRTC_BLANK_CONTROL, CRTC_BLANK_DATA_EN, 0); - WREG32(mmCRTC_BLANK_CONTROL + crtc_offsets[i], tmp); - } - } - - WREG32(mmVGA_MEMORY_BASE_ADDRESS_HIGH, upper_32_bits(adev->mc.vram_start)); - WREG32(mmVGA_MEMORY_BASE_ADDRESS, lower_32_bits(adev->mc.vram_start)); - - /* Unlock vga access */ - WREG32(mmVGA_HDP_CONTROL, save->vga_hdp_control); - mdelay(1); - WREG32(mmVGA_RENDER_CONTROL, save->vga_render_control); -} - static void dce_v11_0_set_vga_render_state(struct amdgpu_device *adev, bool render) { @@ -3073,6 +3000,8 @@ static int dce_v11_0_hw_init(void *handle) dce_v11_0_init_golden_registers(adev); + /* disable vga render */ + dce_v11_0_set_vga_render_state(adev, false); /* init dig PHYs, disp eng pll */ amdgpu_atombios_crtc_powergate_init(adev); amdgpu_atombios_encoder_init_dig(adev); @@ -3793,7 +3722,6 @@ static void dce_v11_0_encoder_add(struct amdgpu_device *adev, } static const struct amdgpu_display_funcs dce_v11_0_display_funcs = { - .set_vga_render_state = &dce_v11_0_set_vga_render_state, .bandwidth_update = &dce_v11_0_bandwidth_update, .vblank_get_counter = &dce_v11_0_vblank_get_counter, .vblank_wait = &dce_v11_0_vblank_wait, @@ -3806,8 +3734,6 @@ static const struct amdgpu_display_funcs dce_v11_0_display_funcs = { .page_flip_get_scanoutpos = &dce_v11_0_crtc_get_scanoutpos, .add_encoder = &dce_v11_0_encoder_add, .add_connector = &amdgpu_connector_add, - .stop_mc_access = &dce_v11_0_stop_mc_access, - .resume_mc_access = &dce_v11_0_resume_mc_access, }; static void dce_v11_0_set_display_funcs(struct amdgpu_device *adev) diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c index f3c422e93ed1..bcd9521237f4 100644 --- a/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c +++ b/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c @@ -392,117 +392,6 @@ static u32 dce_v6_0_hpd_get_gpio_reg(struct amdgpu_device *adev) return mmDC_GPIO_HPD_A; } -static u32 evergreen_get_vblank_counter(struct amdgpu_device* adev, int crtc) -{ - if (crtc >= adev->mode_info.num_crtc) - return 0; - else - return RREG32(mmCRTC_STATUS_FRAME_COUNT + crtc_offsets[crtc]); -} - -static void dce_v6_0_stop_mc_access(struct amdgpu_device *adev, - struct amdgpu_mode_mc_save *save) -{ - u32 crtc_enabled, tmp, frame_count; - int i, j; - - save->vga_render_control = RREG32(mmVGA_RENDER_CONTROL); - save->vga_hdp_control = RREG32(mmVGA_HDP_CONTROL); - - /* disable VGA render */ - WREG32(mmVGA_RENDER_CONTROL, 0); - - /* blank the display controllers */ - for (i = 0; i < adev->mode_info.num_crtc; i++) { - crtc_enabled = RREG32(mmCRTC_CONTROL + crtc_offsets[i]) & CRTC_CONTROL__CRTC_MASTER_EN_MASK; - if (crtc_enabled) { - save->crtc_enabled[i] = true; - tmp = RREG32(mmCRTC_BLANK_CONTROL + crtc_offsets[i]); - - if (!(tmp & CRTC_BLANK_CONTROL__CRTC_BLANK_DATA_EN_MASK)) { - dce_v6_0_vblank_wait(adev, i); - WREG32(mmCRTC_UPDATE_LOCK + crtc_offsets[i], 1); - tmp |= CRTC_BLANK_CONTROL__CRTC_BLANK_DATA_EN_MASK; - WREG32(mmCRTC_BLANK_CONTROL + crtc_offsets[i], tmp); - WREG32(mmCRTC_UPDATE_LOCK + crtc_offsets[i], 0); - } - /* wait for the next frame */ - frame_count = evergreen_get_vblank_counter(adev, i); - for (j = 0; j < adev->usec_timeout; j++) { - if (evergreen_get_vblank_counter(adev, i) != frame_count) - break; - udelay(1); - } - - /* XXX this is a hack to avoid strange behavior with EFI on certain systems */ - WREG32(mmCRTC_UPDATE_LOCK + crtc_offsets[i], 1); - tmp = RREG32(mmCRTC_CONTROL + crtc_offsets[i]); - tmp &= ~CRTC_CONTROL__CRTC_MASTER_EN_MASK; - WREG32(mmCRTC_CONTROL + crtc_offsets[i], tmp); - WREG32(mmCRTC_UPDATE_LOCK + crtc_offsets[i], 0); - save->crtc_enabled[i] = false; - /* ***** */ - } else { - save->crtc_enabled[i] = false; - } - } -} - -static void dce_v6_0_resume_mc_access(struct amdgpu_device *adev, - struct amdgpu_mode_mc_save *save) -{ - u32 tmp; - int i, j; - - /* update crtc base addresses */ - for (i = 0; i < adev->mode_info.num_crtc; i++) { - WREG32(mmGRPH_PRIMARY_SURFACE_ADDRESS_HIGH + crtc_offsets[i], - upper_32_bits(adev->mc.vram_start)); - WREG32(mmGRPH_SECONDARY_SURFACE_ADDRESS_HIGH + crtc_offsets[i], - upper_32_bits(adev->mc.vram_start)); - WREG32(mmGRPH_PRIMARY_SURFACE_ADDRESS + crtc_offsets[i], - (u32)adev->mc.vram_start); - WREG32(mmGRPH_SECONDARY_SURFACE_ADDRESS + crtc_offsets[i], - (u32)adev->mc.vram_start); - } - - WREG32(mmVGA_MEMORY_BASE_ADDRESS_HIGH, upper_32_bits(adev->mc.vram_start)); - WREG32(mmVGA_MEMORY_BASE_ADDRESS, (u32)adev->mc.vram_start); - - /* unlock regs and wait for update */ - for (i = 0; i < adev->mode_info.num_crtc; i++) { - if (save->crtc_enabled[i]) { - tmp = RREG32(mmMASTER_UPDATE_MODE + crtc_offsets[i]); - if ((tmp & 0x7) != 0) { - tmp &= ~0x7; - WREG32(mmMASTER_UPDATE_MODE + crtc_offsets[i], tmp); - } - tmp = RREG32(mmGRPH_UPDATE + crtc_offsets[i]); - if (tmp & GRPH_UPDATE__GRPH_UPDATE_LOCK_MASK) { - tmp &= ~GRPH_UPDATE__GRPH_UPDATE_LOCK_MASK; - WREG32(mmGRPH_UPDATE + crtc_offsets[i], tmp); - } - tmp = RREG32(mmMASTER_UPDATE_LOCK + crtc_offsets[i]); - if (tmp & 1) { - tmp &= ~1; - WREG32(mmMASTER_UPDATE_LOCK + crtc_offsets[i], tmp); - } - for (j = 0; j < adev->usec_timeout; j++) { - tmp = RREG32(mmGRPH_UPDATE + crtc_offsets[i]); - if ((tmp & GRPH_UPDATE__GRPH_SURFACE_UPDATE_PENDING_MASK) == 0) - break; - udelay(1); - } - } - } - - /* Unlock vga access */ - WREG32(mmVGA_HDP_CONTROL, save->vga_hdp_control); - mdelay(1); - WREG32(mmVGA_RENDER_CONTROL, save->vga_render_control); - -} - static void dce_v6_0_set_vga_render_state(struct amdgpu_device *adev, bool render) { @@ -2860,6 +2749,8 @@ static int dce_v6_0_hw_init(void *handle) int i; struct amdgpu_device *adev = (struct amdgpu_device *)handle; + /* disable vga render */ + dce_v6_0_set_vga_render_state(adev, false); /* init dig PHYs, disp eng pll */ amdgpu_atombios_encoder_init_dig(adev); amdgpu_atombios_crtc_set_disp_eng_pll(adev, adev->clock.default_dispclk); @@ -3512,7 +3403,6 @@ static void dce_v6_0_encoder_add(struct amdgpu_device *adev, } static const struct amdgpu_display_funcs dce_v6_0_display_funcs = { - .set_vga_render_state = &dce_v6_0_set_vga_render_state, .bandwidth_update = &dce_v6_0_bandwidth_update, .vblank_get_counter = &dce_v6_0_vblank_get_counter, .vblank_wait = &dce_v6_0_vblank_wait, @@ -3525,8 +3415,6 @@ static const struct amdgpu_display_funcs dce_v6_0_display_funcs = { .page_flip_get_scanoutpos = &dce_v6_0_crtc_get_scanoutpos, .add_encoder = &dce_v6_0_encoder_add, .add_connector = &amdgpu_connector_add, - .stop_mc_access = &dce_v6_0_stop_mc_access, - .resume_mc_access = &dce_v6_0_resume_mc_access, }; static void dce_v6_0_set_display_funcs(struct amdgpu_device *adev) diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c index b19aa39f647d..609438fe8584 100644 --- a/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c +++ b/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c @@ -419,81 +419,6 @@ static bool dce_v8_0_is_display_hung(struct amdgpu_device *adev) return true; } -static void dce_v8_0_stop_mc_access(struct amdgpu_device *adev, - struct amdgpu_mode_mc_save *save) -{ - u32 crtc_enabled, tmp; - int i; - - save->vga_render_control = RREG32(mmVGA_RENDER_CONTROL); - save->vga_hdp_control = RREG32(mmVGA_HDP_CONTROL); - - /* disable VGA render */ - tmp = RREG32(mmVGA_RENDER_CONTROL); - tmp = REG_SET_FIELD(tmp, VGA_RENDER_CONTROL, VGA_VSTATUS_CNTL, 0); - WREG32(mmVGA_RENDER_CONTROL, tmp); - - /* blank the display controllers */ - for (i = 0; i < adev->mode_info.num_crtc; i++) { - crtc_enabled = REG_GET_FIELD(RREG32(mmCRTC_CONTROL + crtc_offsets[i]), - CRTC_CONTROL, CRTC_MASTER_EN); - if (crtc_enabled) { -#if 1 - save->crtc_enabled[i] = true; - tmp = RREG32(mmCRTC_BLANK_CONTROL + crtc_offsets[i]); - if (REG_GET_FIELD(tmp, CRTC_BLANK_CONTROL, CRTC_BLANK_DATA_EN) == 0) { - /*it is correct only for RGB ; black is 0*/ - WREG32(mmCRTC_BLANK_DATA_COLOR + crtc_offsets[i], 0); - tmp = REG_SET_FIELD(tmp, CRTC_BLANK_CONTROL, CRTC_BLANK_DATA_EN, 1); - WREG32(mmCRTC_BLANK_CONTROL + crtc_offsets[i], tmp); - } - mdelay(20); -#else - /* XXX this is a hack to avoid strange behavior with EFI on certain systems */ - WREG32(mmCRTC_UPDATE_LOCK + crtc_offsets[i], 1); - tmp = RREG32(mmCRTC_CONTROL + crtc_offsets[i]); - tmp = REG_SET_FIELD(tmp, CRTC_CONTROL, CRTC_MASTER_EN, 0); - WREG32(mmCRTC_CONTROL + crtc_offsets[i], tmp); - WREG32(mmCRTC_UPDATE_LOCK + crtc_offsets[i], 0); - save->crtc_enabled[i] = false; - /* ***** */ -#endif - } else { - save->crtc_enabled[i] = false; - } - } -} - -static void dce_v8_0_resume_mc_access(struct amdgpu_device *adev, - struct amdgpu_mode_mc_save *save) -{ - u32 tmp; - int i; - - /* update crtc base addresses */ - for (i = 0; i < adev->mode_info.num_crtc; i++) { - WREG32(mmGRPH_PRIMARY_SURFACE_ADDRESS_HIGH + crtc_offsets[i], - upper_32_bits(adev->mc.vram_start)); - WREG32(mmGRPH_PRIMARY_SURFACE_ADDRESS + crtc_offsets[i], - (u32)adev->mc.vram_start); - - if (save->crtc_enabled[i]) { - tmp = RREG32(mmCRTC_BLANK_CONTROL + crtc_offsets[i]); - tmp = REG_SET_FIELD(tmp, CRTC_BLANK_CONTROL, CRTC_BLANK_DATA_EN, 0); - WREG32(mmCRTC_BLANK_CONTROL + crtc_offsets[i], tmp); - } - mdelay(20); - } - - WREG32(mmVGA_MEMORY_BASE_ADDRESS_HIGH, upper_32_bits(adev->mc.vram_start)); - WREG32(mmVGA_MEMORY_BASE_ADDRESS, lower_32_bits(adev->mc.vram_start)); - - /* Unlock vga access */ - WREG32(mmVGA_HDP_CONTROL, save->vga_hdp_control); - mdelay(1); - WREG32(mmVGA_RENDER_CONTROL, save->vga_render_control); -} - static void dce_v8_0_set_vga_render_state(struct amdgpu_device *adev, bool render) { @@ -2857,6 +2782,8 @@ static int dce_v8_0_hw_init(void *handle) int i; struct amdgpu_device *adev = (struct amdgpu_device *)handle; + /* disable vga render */ + dce_v8_0_set_vga_render_state(adev, false); /* init dig PHYs, disp eng pll */ amdgpu_atombios_encoder_init_dig(adev); amdgpu_atombios_crtc_set_disp_eng_pll(adev, adev->clock.default_dispclk); @@ -3561,7 +3488,6 @@ static void dce_v8_0_encoder_add(struct amdgpu_device *adev, } static const struct amdgpu_display_funcs dce_v8_0_display_funcs = { - .set_vga_render_state = &dce_v8_0_set_vga_render_state, .bandwidth_update = &dce_v8_0_bandwidth_update, .vblank_get_counter = &dce_v8_0_vblank_get_counter, .vblank_wait = &dce_v8_0_vblank_wait, @@ -3574,8 +3500,6 @@ static const struct amdgpu_display_funcs dce_v8_0_display_funcs = { .page_flip_get_scanoutpos = &dce_v8_0_crtc_get_scanoutpos, .add_encoder = &dce_v8_0_encoder_add, .add_connector = &amdgpu_connector_add, - .stop_mc_access = &dce_v8_0_stop_mc_access, - .resume_mc_access = &dce_v8_0_resume_mc_access, }; static void dce_v8_0_set_display_funcs(struct amdgpu_device *adev) diff --git a/drivers/gpu/drm/amd/amdgpu/dce_virtual.c b/drivers/gpu/drm/amd/amdgpu/dce_virtual.c index ecf34bc77a63..5ed919e45351 100644 --- a/drivers/gpu/drm/amd/amdgpu/dce_virtual.c +++ b/drivers/gpu/drm/amd/amdgpu/dce_virtual.c @@ -95,62 +95,6 @@ static u32 dce_virtual_hpd_get_gpio_reg(struct amdgpu_device *adev) return 0; } -static void dce_virtual_stop_mc_access(struct amdgpu_device *adev, - struct amdgpu_mode_mc_save *save) -{ - switch (adev->asic_type) { -#ifdef CONFIG_DRM_AMDGPU_SI - case CHIP_TAHITI: - case CHIP_PITCAIRN: - case CHIP_VERDE: - case CHIP_OLAND: - dce_v6_0_disable_dce(adev); - break; -#endif -#ifdef CONFIG_DRM_AMDGPU_CIK - case CHIP_BONAIRE: - case CHIP_HAWAII: - case CHIP_KAVERI: - case CHIP_KABINI: - case CHIP_MULLINS: - dce_v8_0_disable_dce(adev); - break; -#endif - case CHIP_FIJI: - case CHIP_TONGA: - dce_v10_0_disable_dce(adev); - break; - case CHIP_CARRIZO: - case CHIP_STONEY: - case CHIP_POLARIS10: - case CHIP_POLARIS11: - case CHIP_POLARIS12: - dce_v11_0_disable_dce(adev); - break; - case CHIP_TOPAZ: -#ifdef CONFIG_DRM_AMDGPU_SI - case CHIP_HAINAN: -#endif - /* no DCE */ - return; - default: - DRM_ERROR("Virtual display unsupported ASIC type: 0x%X\n", adev->asic_type); - } - - return; -} -static void dce_virtual_resume_mc_access(struct amdgpu_device *adev, - struct amdgpu_mode_mc_save *save) -{ - return; -} - -static void dce_virtual_set_vga_render_state(struct amdgpu_device *adev, - bool render) -{ - return; -} - /** * dce_virtual_bandwidth_update - program display watermarks * @@ -499,6 +443,45 @@ static int dce_virtual_sw_fini(void *handle) static int dce_virtual_hw_init(void *handle) { + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + switch (adev->asic_type) { +#ifdef CONFIG_DRM_AMDGPU_SI + case CHIP_TAHITI: + case CHIP_PITCAIRN: + case CHIP_VERDE: + case CHIP_OLAND: + dce_v6_0_disable_dce(adev); + break; +#endif +#ifdef CONFIG_DRM_AMDGPU_CIK + case CHIP_BONAIRE: + case CHIP_HAWAII: + case CHIP_KAVERI: + case CHIP_KABINI: + case CHIP_MULLINS: + dce_v8_0_disable_dce(adev); + break; +#endif + case CHIP_FIJI: + case CHIP_TONGA: + dce_v10_0_disable_dce(adev); + break; + case CHIP_CARRIZO: + case CHIP_STONEY: + case CHIP_POLARIS11: + case CHIP_POLARIS10: + dce_v11_0_disable_dce(adev); + break; + case CHIP_TOPAZ: +#ifdef CONFIG_DRM_AMDGPU_SI + case CHIP_HAINAN: +#endif + /* no DCE */ + break; + default: + DRM_ERROR("Virtual display unsupported ASIC type: 0x%X\n", adev->asic_type); + } return 0; } @@ -654,7 +637,6 @@ static int dce_virtual_connector_encoder_init(struct amdgpu_device *adev, } static const struct amdgpu_display_funcs dce_virtual_display_funcs = { - .set_vga_render_state = &dce_virtual_set_vga_render_state, .bandwidth_update = &dce_virtual_bandwidth_update, .vblank_get_counter = &dce_virtual_vblank_get_counter, .vblank_wait = &dce_virtual_vblank_wait, @@ -667,8 +649,6 @@ static const struct amdgpu_display_funcs dce_virtual_display_funcs = { .page_flip_get_scanoutpos = &dce_virtual_crtc_get_scanoutpos, .add_encoder = NULL, .add_connector = NULL, - .stop_mc_access = &dce_virtual_stop_mc_access, - .resume_mc_access = &dce_virtual_resume_mc_access, }; static void dce_virtual_set_display_funcs(struct amdgpu_device *adev) @@ -786,7 +766,7 @@ static const struct amdgpu_irq_src_funcs dce_virtual_crtc_irq_funcs = { static void dce_virtual_set_irq_funcs(struct amdgpu_device *adev) { - adev->crtc_irq.num_types = AMDGPU_CRTC_IRQ_LAST; + adev->crtc_irq.num_types = AMDGPU_CRTC_IRQ_VBLANK6 + 1; adev->crtc_irq.funcs = &dce_virtual_crtc_irq_funcs; } diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c index 5173ca1fd159..4ac85f47f287 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c @@ -1573,7 +1573,7 @@ static void gfx_v6_0_gpu_init(struct amdgpu_device *adev) static void gfx_v6_0_scratch_init(struct amdgpu_device *adev) { - adev->gfx.scratch.num_reg = 7; + adev->gfx.scratch.num_reg = 8; adev->gfx.scratch.reg_base = mmSCRATCH_REG0; adev->gfx.scratch.free_mask = (1u << adev->gfx.scratch.num_reg) - 1; } diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c index 37b45e4403d1..17b7c6934b0a 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c @@ -2021,7 +2021,7 @@ static void gfx_v7_0_gpu_init(struct amdgpu_device *adev) */ static void gfx_v7_0_scratch_init(struct amdgpu_device *adev) { - adev->gfx.scratch.num_reg = 7; + adev->gfx.scratch.num_reg = 8; adev->gfx.scratch.reg_base = mmSCRATCH_REG0; adev->gfx.scratch.free_mask = (1u << adev->gfx.scratch.num_reg) - 1; } diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c index aa5a50f5eac8..05436b8730b4 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c @@ -193,8 +193,8 @@ static const u32 tonga_golden_common_all[] = mmGB_ADDR_CONFIG, 0xffffffff, 0x22011003, mmSPI_RESOURCE_RESERVE_CU_0, 0xffffffff, 0x00000800, mmSPI_RESOURCE_RESERVE_CU_1, 0xffffffff, 0x00000800, - mmSPI_RESOURCE_RESERVE_EN_CU_0, 0xffffffff, 0x00007FBF, - mmSPI_RESOURCE_RESERVE_EN_CU_1, 0xffffffff, 0x00007FAF + mmSPI_RESOURCE_RESERVE_EN_CU_0, 0xffffffff, 0x00FF7FBF, + mmSPI_RESOURCE_RESERVE_EN_CU_1, 0xffffffff, 0x00FF7FAF }; static const u32 tonga_mgcg_cgcg_init[] = @@ -303,8 +303,8 @@ static const u32 polaris11_golden_common_all[] = mmGB_ADDR_CONFIG, 0xffffffff, 0x22011002, mmSPI_RESOURCE_RESERVE_CU_0, 0xffffffff, 0x00000800, mmSPI_RESOURCE_RESERVE_CU_1, 0xffffffff, 0x00000800, - mmSPI_RESOURCE_RESERVE_EN_CU_0, 0xffffffff, 0x00007FBF, - mmSPI_RESOURCE_RESERVE_EN_CU_1, 0xffffffff, 0x00007FAF, + mmSPI_RESOURCE_RESERVE_EN_CU_0, 0xffffffff, 0x00FF7FBF, + mmSPI_RESOURCE_RESERVE_EN_CU_1, 0xffffffff, 0x00FF7FAF, }; static const u32 golden_settings_polaris10_a11[] = @@ -336,8 +336,8 @@ static const u32 polaris10_golden_common_all[] = mmGB_ADDR_CONFIG, 0xffffffff, 0x22011003, mmSPI_RESOURCE_RESERVE_CU_0, 0xffffffff, 0x00000800, mmSPI_RESOURCE_RESERVE_CU_1, 0xffffffff, 0x00000800, - mmSPI_RESOURCE_RESERVE_EN_CU_0, 0xffffffff, 0x00007FBF, - mmSPI_RESOURCE_RESERVE_EN_CU_1, 0xffffffff, 0x00007FAF, + mmSPI_RESOURCE_RESERVE_EN_CU_0, 0xffffffff, 0x00FF7FBF, + mmSPI_RESOURCE_RESERVE_EN_CU_1, 0xffffffff, 0x00FF7FAF, }; static const u32 fiji_golden_common_all[] = @@ -348,8 +348,8 @@ static const u32 fiji_golden_common_all[] = mmGB_ADDR_CONFIG, 0xffffffff, 0x22011003, mmSPI_RESOURCE_RESERVE_CU_0, 0xffffffff, 0x00000800, mmSPI_RESOURCE_RESERVE_CU_1, 0xffffffff, 0x00000800, - mmSPI_RESOURCE_RESERVE_EN_CU_0, 0xffffffff, 0x00007FBF, - mmSPI_RESOURCE_RESERVE_EN_CU_1, 0xffffffff, 0x00007FAF, + mmSPI_RESOURCE_RESERVE_EN_CU_0, 0xffffffff, 0x00FF7FBF, + mmSPI_RESOURCE_RESERVE_EN_CU_1, 0xffffffff, 0x00FF7FAF, mmGRBM_GFX_INDEX, 0xffffffff, 0xe0000000, mmSPI_CONFIG_CNTL_1, 0x0000000f, 0x00000009, }; @@ -436,8 +436,8 @@ static const u32 iceland_golden_common_all[] = mmGB_ADDR_CONFIG, 0xffffffff, 0x22010001, mmSPI_RESOURCE_RESERVE_CU_0, 0xffffffff, 0x00000800, mmSPI_RESOURCE_RESERVE_CU_1, 0xffffffff, 0x00000800, - mmSPI_RESOURCE_RESERVE_EN_CU_0, 0xffffffff, 0x00007FBF, - mmSPI_RESOURCE_RESERVE_EN_CU_1, 0xffffffff, 0x00007FAF + mmSPI_RESOURCE_RESERVE_EN_CU_0, 0xffffffff, 0x00FF7FBF, + mmSPI_RESOURCE_RESERVE_EN_CU_1, 0xffffffff, 0x00FF7FAF }; static const u32 iceland_mgcg_cgcg_init[] = @@ -532,8 +532,8 @@ static const u32 cz_golden_common_all[] = mmGB_ADDR_CONFIG, 0xffffffff, 0x22010001, mmSPI_RESOURCE_RESERVE_CU_0, 0xffffffff, 0x00000800, mmSPI_RESOURCE_RESERVE_CU_1, 0xffffffff, 0x00000800, - mmSPI_RESOURCE_RESERVE_EN_CU_0, 0xffffffff, 0x00007FBF, - mmSPI_RESOURCE_RESERVE_EN_CU_1, 0xffffffff, 0x00007FAF + mmSPI_RESOURCE_RESERVE_EN_CU_0, 0xffffffff, 0x00FF7FBF, + mmSPI_RESOURCE_RESERVE_EN_CU_1, 0xffffffff, 0x00FF7FAF }; static const u32 cz_mgcg_cgcg_init[] = @@ -637,8 +637,8 @@ static const u32 stoney_golden_common_all[] = mmGB_ADDR_CONFIG, 0xffffffff, 0x12010001, mmSPI_RESOURCE_RESERVE_CU_0, 0xffffffff, 0x00000800, mmSPI_RESOURCE_RESERVE_CU_1, 0xffffffff, 0x00000800, - mmSPI_RESOURCE_RESERVE_EN_CU_0, 0xffffffff, 0x00007FBF, - mmSPI_RESOURCE_RESERVE_EN_CU_1, 0xffffffff, 0x00007FAF, + mmSPI_RESOURCE_RESERVE_EN_CU_0, 0xffffffff, 0x00FF7FBF, + mmSPI_RESOURCE_RESERVE_EN_CU_1, 0xffffffff, 0x00FF7FAF, }; static const u32 stoney_mgcg_cgcg_init[] = @@ -750,7 +750,7 @@ static void gfx_v8_0_init_golden_registers(struct amdgpu_device *adev) static void gfx_v8_0_scratch_init(struct amdgpu_device *adev) { - adev->gfx.scratch.num_reg = 7; + adev->gfx.scratch.num_reg = 8; adev->gfx.scratch.reg_base = mmSCRATCH_REG0; adev->gfx.scratch.free_mask = (1u << adev->gfx.scratch.num_reg) - 1; } @@ -4564,7 +4564,7 @@ static int gfx_v8_0_kiq_kcq_enable(struct amdgpu_device *adev) /* This situation may be hit in the future if a new HW * generation exposes more than 64 queues. If so, the * definition of queue_mask needs updating */ - if (WARN_ON(i > (sizeof(queue_mask)*8))) { + if (WARN_ON(i >= (sizeof(queue_mask)*8))) { DRM_ERROR("Invalid KCQ enabled: %d\n", i); break; } diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c index 3a0b69b09ed6..435db6f5efcf 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c @@ -211,7 +211,7 @@ static void gfx_v9_0_init_golden_registers(struct amdgpu_device *adev) static void gfx_v9_0_scratch_init(struct amdgpu_device *adev) { - adev->gfx.scratch.num_reg = 7; + adev->gfx.scratch.num_reg = 8; adev->gfx.scratch.reg_base = SOC15_REG_OFFSET(GC, 0, mmSCRATCH_REG0); adev->gfx.scratch.free_mask = (1u << adev->gfx.scratch.num_reg) - 1; } @@ -1475,21 +1475,23 @@ static void gfx_v9_0_tiling_mode_table_init(struct amdgpu_device *adev) static void gfx_v9_0_select_se_sh(struct amdgpu_device *adev, u32 se_num, u32 sh_num, u32 instance) { - u32 data = REG_SET_FIELD(0, GRBM_GFX_INDEX, INSTANCE_BROADCAST_WRITES, 1); + u32 data; - if ((se_num == 0xffffffff) && (sh_num == 0xffffffff)) { - data = REG_SET_FIELD(data, GRBM_GFX_INDEX, SH_BROADCAST_WRITES, 1); - data = REG_SET_FIELD(data, GRBM_GFX_INDEX, SE_BROADCAST_WRITES, 1); - } else if (se_num == 0xffffffff) { - data = REG_SET_FIELD(data, GRBM_GFX_INDEX, SH_INDEX, sh_num); + if (instance == 0xffffffff) + data = REG_SET_FIELD(0, GRBM_GFX_INDEX, INSTANCE_BROADCAST_WRITES, 1); + else + data = REG_SET_FIELD(0, GRBM_GFX_INDEX, INSTANCE_INDEX, instance); + + if (se_num == 0xffffffff) data = REG_SET_FIELD(data, GRBM_GFX_INDEX, SE_BROADCAST_WRITES, 1); - } else if (sh_num == 0xffffffff) { - data = REG_SET_FIELD(data, GRBM_GFX_INDEX, SH_BROADCAST_WRITES, 1); + else data = REG_SET_FIELD(data, GRBM_GFX_INDEX, SE_INDEX, se_num); - } else { + + if (sh_num == 0xffffffff) + data = REG_SET_FIELD(data, GRBM_GFX_INDEX, SH_BROADCAST_WRITES, 1); + else 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(GC, 0, mmGRBM_GFX_INDEX, data); } @@ -2425,7 +2427,7 @@ static int gfx_v9_0_kiq_kcq_enable(struct amdgpu_device *adev) /* This situation may be hit in the future if a new HW * generation exposes more than 64 queues. If so, the * definition of queue_mask needs updating */ - if (WARN_ON(i > (sizeof(queue_mask)*8))) { + if (WARN_ON(i >= (sizeof(queue_mask)*8))) { DRM_ERROR("Invalid KCQ enabled: %d\n", i); break; } diff --git a/drivers/gpu/drm/amd/amdgpu/gfxhub_v1_0.c b/drivers/gpu/drm/amd/amdgpu/gfxhub_v1_0.c index a42f483767e7..408723ef157c 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfxhub_v1_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfxhub_v1_0.c @@ -58,14 +58,14 @@ static void gfxhub_v1_0_init_gart_aperture_regs(struct amdgpu_device *adev) gfxhub_v1_0_init_gart_pt_regs(adev); WREG32_SOC15(GC, 0, mmVM_CONTEXT0_PAGE_TABLE_START_ADDR_LO32, - (u32)(adev->mc.gtt_start >> 12)); + (u32)(adev->mc.gart_start >> 12)); WREG32_SOC15(GC, 0, mmVM_CONTEXT0_PAGE_TABLE_START_ADDR_HI32, - (u32)(adev->mc.gtt_start >> 44)); + (u32)(adev->mc.gart_start >> 44)); WREG32_SOC15(GC, 0, mmVM_CONTEXT0_PAGE_TABLE_END_ADDR_LO32, - (u32)(adev->mc.gtt_end >> 12)); + (u32)(adev->mc.gart_end >> 12)); WREG32_SOC15(GC, 0, mmVM_CONTEXT0_PAGE_TABLE_END_ADDR_HI32, - (u32)(adev->mc.gtt_end >> 44)); + (u32)(adev->mc.gart_end >> 44)); } static void gfxhub_v1_0_init_system_aperture_regs(struct amdgpu_device *adev) @@ -129,7 +129,7 @@ static void gfxhub_v1_0_init_cache_regs(struct amdgpu_device *adev) /* Setup L2 cache */ tmp = RREG32_SOC15(GC, 0, mmVM_L2_CNTL); tmp = REG_SET_FIELD(tmp, VM_L2_CNTL, ENABLE_L2_CACHE, 1); - tmp = REG_SET_FIELD(tmp, VM_L2_CNTL, ENABLE_L2_FRAGMENT_PROCESSING, 0); + tmp = REG_SET_FIELD(tmp, VM_L2_CNTL, ENABLE_L2_FRAGMENT_PROCESSING, 1); /* XXX for emulation, Refer to closed source code.*/ tmp = REG_SET_FIELD(tmp, VM_L2_CNTL, L2_PDE0_CACHE_TAG_GENERATION_MODE, 0); @@ -144,6 +144,8 @@ static void gfxhub_v1_0_init_cache_regs(struct amdgpu_device *adev) WREG32_SOC15(GC, 0, mmVM_L2_CNTL2, tmp); tmp = mmVM_L2_CNTL3_DEFAULT; + tmp = REG_SET_FIELD(tmp, VM_L2_CNTL3, BANK_SELECT, 12); + tmp = REG_SET_FIELD(tmp, VM_L2_CNTL3, L2_CACHE_BIGK_FRAGMENT_SIZE, 9); WREG32_SOC15(GC, 0, mmVM_L2_CNTL3, tmp); tmp = mmVM_L2_CNTL4_DEFAULT; @@ -206,6 +208,9 @@ static void gfxhub_v1_0_setup_vmid_config(struct amdgpu_device *adev) tmp = REG_SET_FIELD(tmp, VM_CONTEXT1_CNTL, PAGE_TABLE_BLOCK_SIZE, adev->vm_manager.block_size - 9); + /* Send no-retry XNACK on fault to suppress VM fault storm. */ + tmp = REG_SET_FIELD(tmp, VM_CONTEXT1_CNTL, + RETRY_PERMISSION_OR_INVALID_PAGE_FAULT, 0); WREG32_SOC15_OFFSET(GC, 0, mmVM_CONTEXT1_CNTL, i, tmp); WREG32_SOC15_OFFSET(GC, 0, mmVM_CONTEXT1_PAGE_TABLE_START_ADDR_LO32, i*2, 0); WREG32_SOC15_OFFSET(GC, 0, mmVM_CONTEXT1_PAGE_TABLE_START_ADDR_HI32, i*2, 0); diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c index d0214d942bfc..93c45f26b7c8 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c @@ -66,14 +66,10 @@ static const u32 crtc_offsets[6] = SI_CRTC5_REGISTER_OFFSET }; -static void gmc_v6_0_mc_stop(struct amdgpu_device *adev, - struct amdgpu_mode_mc_save *save) +static void gmc_v6_0_mc_stop(struct amdgpu_device *adev) { u32 blackout; - if (adev->mode_info.num_crtc) - amdgpu_display_stop_mc_access(adev, save); - gmc_v6_0_wait_for_idle((void *)adev); blackout = RREG32(mmMC_SHARED_BLACKOUT_CNTL); @@ -90,8 +86,7 @@ static void gmc_v6_0_mc_stop(struct amdgpu_device *adev, } -static void gmc_v6_0_mc_resume(struct amdgpu_device *adev, - struct amdgpu_mode_mc_save *save) +static void gmc_v6_0_mc_resume(struct amdgpu_device *adev) { u32 tmp; @@ -103,10 +98,6 @@ static void gmc_v6_0_mc_resume(struct amdgpu_device *adev, tmp = REG_SET_FIELD(0, BIF_FB_EN, FB_READ_EN, 1); tmp = REG_SET_FIELD(tmp, BIF_FB_EN, FB_WRITE_EN, 1); WREG32(mmBIF_FB_EN, tmp); - - if (adev->mode_info.num_crtc) - amdgpu_display_resume_mc_access(adev, save); - } static int gmc_v6_0_init_microcode(struct amdgpu_device *adev) @@ -228,20 +219,20 @@ static int gmc_v6_0_mc_load_microcode(struct amdgpu_device *adev) static void gmc_v6_0_vram_gtt_location(struct amdgpu_device *adev, struct amdgpu_mc *mc) { + u64 base = RREG32(mmMC_VM_FB_LOCATION) & 0xFFFF; + base <<= 24; + if (mc->mc_vram_size > 0xFFC0000000ULL) { dev_warn(adev->dev, "limiting VRAM\n"); mc->real_vram_size = 0xFFC0000000ULL; mc->mc_vram_size = 0xFFC0000000ULL; } - amdgpu_vram_location(adev, &adev->mc, 0); - adev->mc.gtt_base_align = 0; - amdgpu_gtt_location(adev, mc); + amdgpu_vram_location(adev, &adev->mc, base); + amdgpu_gart_location(adev, mc); } static void gmc_v6_0_mc_program(struct amdgpu_device *adev) { - struct amdgpu_mode_mc_save save; - u32 tmp; int i, j; /* Initialize HDP */ @@ -254,16 +245,23 @@ static void gmc_v6_0_mc_program(struct amdgpu_device *adev) } WREG32(mmHDP_REG_COHERENCY_FLUSH_CNTL, 0); - if (adev->mode_info.num_crtc) - amdgpu_display_set_vga_render_state(adev, false); - - gmc_v6_0_mc_stop(adev, &save); - if (gmc_v6_0_wait_for_idle((void *)adev)) { dev_warn(adev->dev, "Wait for MC idle timedout !\n"); } - WREG32(mmVGA_HDP_CONTROL, VGA_HDP_CONTROL__VGA_MEMORY_DISABLE_MASK); + if (adev->mode_info.num_crtc) { + u32 tmp; + + /* Lockout access through VGA aperture*/ + tmp = RREG32(mmVGA_HDP_CONTROL); + tmp |= VGA_HDP_CONTROL__VGA_MEMORY_DISABLE_MASK; + WREG32(mmVGA_HDP_CONTROL, tmp); + + /* disable VGA render */ + tmp = RREG32(mmVGA_RENDER_CONTROL); + tmp &= ~VGA_VSTATUS_CNTL; + WREG32(mmVGA_RENDER_CONTROL, tmp); + } /* Update configuration */ WREG32(mmMC_VM_SYSTEM_APERTURE_LOW_ADDR, adev->mc.vram_start >> 12); @@ -271,13 +269,6 @@ static void gmc_v6_0_mc_program(struct amdgpu_device *adev) adev->mc.vram_end >> 12); WREG32(mmMC_VM_SYSTEM_APERTURE_DEFAULT_ADDR, adev->vram_scratch.gpu_addr >> 12); - tmp = ((adev->mc.vram_end >> 24) & 0xFFFF) << 16; - tmp |= ((adev->mc.vram_start >> 24) & 0xFFFF); - WREG32(mmMC_VM_FB_LOCATION, tmp); - /* XXX double check these! */ - WREG32(mmHDP_NONSURFACE_BASE, (adev->mc.vram_start >> 8)); - WREG32(mmHDP_NONSURFACE_INFO, (2 << 7) | (1 << 30)); - WREG32(mmHDP_NONSURFACE_SIZE, 0x3FFFFFFF); WREG32(mmMC_VM_AGP_BASE, 0); WREG32(mmMC_VM_AGP_TOP, 0x0FFFFFFF); WREG32(mmMC_VM_AGP_BOT, 0x0FFFFFFF); @@ -285,7 +276,6 @@ static void gmc_v6_0_mc_program(struct amdgpu_device *adev) if (gmc_v6_0_wait_for_idle((void *)adev)) { dev_warn(adev->dev, "Wait for MC idle timedout !\n"); } - gmc_v6_0_mc_resume(adev, &save); } static int gmc_v6_0_mc_init(struct amdgpu_device *adev) @@ -342,15 +332,7 @@ static int gmc_v6_0_mc_init(struct amdgpu_device *adev) adev->mc.real_vram_size = RREG32(mmCONFIG_MEMSIZE) * 1024ULL * 1024ULL; adev->mc.visible_vram_size = adev->mc.aper_size; - /* unless the user had overridden it, set the gart - * size equal to the 1024 or vram, whichever is larger. - */ - if (amdgpu_gart_size == -1) - adev->mc.gtt_size = max((AMDGPU_DEFAULT_GTT_SIZE_MB << 20), - adev->mc.mc_vram_size); - else - adev->mc.gtt_size = (uint64_t)amdgpu_gart_size << 20; - + amdgpu_gart_set_defaults(adev); gmc_v6_0_vram_gtt_location(adev, &adev->mc); return 0; @@ -511,8 +493,8 @@ static int gmc_v6_0_gart_enable(struct amdgpu_device *adev) (4UL << VM_L2_CNTL3__BANK_SELECT__SHIFT) | (4UL << VM_L2_CNTL3__L2_CACHE_BIGK_FRAGMENT_SIZE__SHIFT)); /* setup context0 */ - WREG32(mmVM_CONTEXT0_PAGE_TABLE_START_ADDR, adev->mc.gtt_start >> 12); - WREG32(mmVM_CONTEXT0_PAGE_TABLE_END_ADDR, adev->mc.gtt_end >> 12); + WREG32(mmVM_CONTEXT0_PAGE_TABLE_START_ADDR, adev->mc.gart_start >> 12); + WREG32(mmVM_CONTEXT0_PAGE_TABLE_END_ADDR, adev->mc.gart_end >> 12); WREG32(mmVM_CONTEXT0_PAGE_TABLE_BASE_ADDR, adev->gart.table_addr >> 12); WREG32(mmVM_CONTEXT0_PROTECTION_FAULT_DEFAULT_ADDR, (u32)(adev->dummy_page.addr >> 12)); @@ -559,7 +541,7 @@ static int gmc_v6_0_gart_enable(struct amdgpu_device *adev) gmc_v6_0_gart_flush_gpu_tlb(adev, 0); dev_info(adev->dev, "PCIE GART of %uM enabled (table at 0x%016llX).\n", - (unsigned)(adev->mc.gtt_size >> 20), + (unsigned)(adev->mc.gart_size >> 20), (unsigned long long)adev->gart.table_addr); adev->gart.ready = true; return 0; @@ -987,7 +969,6 @@ static int gmc_v6_0_wait_for_idle(void *handle) static int gmc_v6_0_soft_reset(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; - struct amdgpu_mode_mc_save save; u32 srbm_soft_reset = 0; u32 tmp = RREG32(mmSRBM_STATUS); @@ -1003,7 +984,7 @@ static int gmc_v6_0_soft_reset(void *handle) } if (srbm_soft_reset) { - gmc_v6_0_mc_stop(adev, &save); + gmc_v6_0_mc_stop(adev); if (gmc_v6_0_wait_for_idle(adev)) { dev_warn(adev->dev, "Wait for GMC idle timed out !\n"); } @@ -1023,7 +1004,7 @@ static int gmc_v6_0_soft_reset(void *handle) udelay(50); - gmc_v6_0_mc_resume(adev, &save); + gmc_v6_0_mc_resume(adev); udelay(50); } diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c index 7e9ea53edf8b..4a9e84062874 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c @@ -37,6 +37,9 @@ #include "oss/oss_2_0_d.h" #include "oss/oss_2_0_sh_mask.h" +#include "dce/dce_8_0_d.h" +#include "dce/dce_8_0_sh_mask.h" + #include "amdgpu_atombios.h" static void gmc_v7_0_set_gart_funcs(struct amdgpu_device *adev); @@ -76,14 +79,10 @@ static void gmc_v7_0_init_golden_registers(struct amdgpu_device *adev) } } -static void gmc_v7_0_mc_stop(struct amdgpu_device *adev, - struct amdgpu_mode_mc_save *save) +static void gmc_v7_0_mc_stop(struct amdgpu_device *adev) { u32 blackout; - if (adev->mode_info.num_crtc) - amdgpu_display_stop_mc_access(adev, save); - gmc_v7_0_wait_for_idle((void *)adev); blackout = RREG32(mmMC_SHARED_BLACKOUT_CNTL); @@ -99,8 +98,7 @@ static void gmc_v7_0_mc_stop(struct amdgpu_device *adev, udelay(100); } -static void gmc_v7_0_mc_resume(struct amdgpu_device *adev, - struct amdgpu_mode_mc_save *save) +static void gmc_v7_0_mc_resume(struct amdgpu_device *adev) { u32 tmp; @@ -112,9 +110,6 @@ static void gmc_v7_0_mc_resume(struct amdgpu_device *adev, tmp = REG_SET_FIELD(0, BIF_FB_EN, FB_READ_EN, 1); tmp = REG_SET_FIELD(tmp, BIF_FB_EN, FB_WRITE_EN, 1); WREG32(mmBIF_FB_EN, tmp); - - if (adev->mode_info.num_crtc) - amdgpu_display_resume_mc_access(adev, save); } /** @@ -242,15 +237,17 @@ static int gmc_v7_0_mc_load_microcode(struct amdgpu_device *adev) static void gmc_v7_0_vram_gtt_location(struct amdgpu_device *adev, struct amdgpu_mc *mc) { + u64 base = RREG32(mmMC_VM_FB_LOCATION) & 0xFFFF; + base <<= 24; + if (mc->mc_vram_size > 0xFFC0000000ULL) { /* leave room for at least 1024M GTT */ dev_warn(adev->dev, "limiting VRAM\n"); mc->real_vram_size = 0xFFC0000000ULL; mc->mc_vram_size = 0xFFC0000000ULL; } - amdgpu_vram_location(adev, &adev->mc, 0); - adev->mc.gtt_base_align = 0; - amdgpu_gtt_location(adev, mc); + amdgpu_vram_location(adev, &adev->mc, base); + amdgpu_gart_location(adev, mc); } /** @@ -263,7 +260,6 @@ static void gmc_v7_0_vram_gtt_location(struct amdgpu_device *adev, */ static void gmc_v7_0_mc_program(struct amdgpu_device *adev) { - struct amdgpu_mode_mc_save save; u32 tmp; int i, j; @@ -277,13 +273,20 @@ static void gmc_v7_0_mc_program(struct amdgpu_device *adev) } WREG32(mmHDP_REG_COHERENCY_FLUSH_CNTL, 0); - if (adev->mode_info.num_crtc) - amdgpu_display_set_vga_render_state(adev, false); - - gmc_v7_0_mc_stop(adev, &save); if (gmc_v7_0_wait_for_idle((void *)adev)) { dev_warn(adev->dev, "Wait for MC idle timedout !\n"); } + if (adev->mode_info.num_crtc) { + /* Lockout access through VGA aperture*/ + tmp = RREG32(mmVGA_HDP_CONTROL); + tmp = REG_SET_FIELD(tmp, VGA_HDP_CONTROL, VGA_MEMORY_DISABLE, 1); + WREG32(mmVGA_HDP_CONTROL, tmp); + + /* disable VGA render */ + tmp = RREG32(mmVGA_RENDER_CONTROL); + tmp = REG_SET_FIELD(tmp, VGA_RENDER_CONTROL, VGA_VSTATUS_CNTL, 0); + WREG32(mmVGA_RENDER_CONTROL, tmp); + } /* Update configuration */ WREG32(mmMC_VM_SYSTEM_APERTURE_LOW_ADDR, adev->mc.vram_start >> 12); @@ -291,20 +294,12 @@ static void gmc_v7_0_mc_program(struct amdgpu_device *adev) adev->mc.vram_end >> 12); WREG32(mmMC_VM_SYSTEM_APERTURE_DEFAULT_ADDR, adev->vram_scratch.gpu_addr >> 12); - tmp = ((adev->mc.vram_end >> 24) & 0xFFFF) << 16; - tmp |= ((adev->mc.vram_start >> 24) & 0xFFFF); - WREG32(mmMC_VM_FB_LOCATION, tmp); - /* XXX double check these! */ - WREG32(mmHDP_NONSURFACE_BASE, (adev->mc.vram_start >> 8)); - WREG32(mmHDP_NONSURFACE_INFO, (2 << 7) | (1 << 30)); - WREG32(mmHDP_NONSURFACE_SIZE, 0x3FFFFFFF); WREG32(mmMC_VM_AGP_BASE, 0); WREG32(mmMC_VM_AGP_TOP, 0x0FFFFFFF); WREG32(mmMC_VM_AGP_BOT, 0x0FFFFFFF); if (gmc_v7_0_wait_for_idle((void *)adev)) { dev_warn(adev->dev, "Wait for MC idle timedout !\n"); } - gmc_v7_0_mc_resume(adev, &save); WREG32(mmBIF_FB_EN, BIF_FB_EN__FB_READ_EN_MASK | BIF_FB_EN__FB_WRITE_EN_MASK); @@ -391,15 +386,7 @@ static int gmc_v7_0_mc_init(struct amdgpu_device *adev) if (adev->mc.visible_vram_size > adev->mc.real_vram_size) adev->mc.visible_vram_size = adev->mc.real_vram_size; - /* unless the user had overridden it, set the gart - * size equal to the 1024 or vram, whichever is larger. - */ - if (amdgpu_gart_size == -1) - adev->mc.gtt_size = max((AMDGPU_DEFAULT_GTT_SIZE_MB << 20), - adev->mc.mc_vram_size); - else - adev->mc.gtt_size = (uint64_t)amdgpu_gart_size << 20; - + amdgpu_gart_set_defaults(adev); gmc_v7_0_vram_gtt_location(adev, &adev->mc); return 0; @@ -611,8 +598,8 @@ static int gmc_v7_0_gart_enable(struct amdgpu_device *adev) tmp = REG_SET_FIELD(tmp, VM_L2_CNTL3, L2_CACHE_BIGK_FRAGMENT_SIZE, 4); WREG32(mmVM_L2_CNTL3, tmp); /* setup context0 */ - WREG32(mmVM_CONTEXT0_PAGE_TABLE_START_ADDR, adev->mc.gtt_start >> 12); - WREG32(mmVM_CONTEXT0_PAGE_TABLE_END_ADDR, adev->mc.gtt_end >> 12); + WREG32(mmVM_CONTEXT0_PAGE_TABLE_START_ADDR, adev->mc.gart_start >> 12); + WREG32(mmVM_CONTEXT0_PAGE_TABLE_END_ADDR, adev->mc.gart_end >> 12); WREG32(mmVM_CONTEXT0_PAGE_TABLE_BASE_ADDR, adev->gart.table_addr >> 12); WREG32(mmVM_CONTEXT0_PROTECTION_FAULT_DEFAULT_ADDR, (u32)(adev->dummy_page.addr >> 12)); @@ -666,7 +653,7 @@ static int gmc_v7_0_gart_enable(struct amdgpu_device *adev) gmc_v7_0_gart_flush_gpu_tlb(adev, 0); DRM_INFO("PCIE GART of %uM enabled (table at 0x%016llX).\n", - (unsigned)(adev->mc.gtt_size >> 20), + (unsigned)(adev->mc.gart_size >> 20), (unsigned long long)adev->gart.table_addr); adev->gart.ready = true; return 0; @@ -1138,7 +1125,6 @@ static int gmc_v7_0_wait_for_idle(void *handle) static int gmc_v7_0_soft_reset(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; - struct amdgpu_mode_mc_save save; u32 srbm_soft_reset = 0; u32 tmp = RREG32(mmSRBM_STATUS); @@ -1154,7 +1140,7 @@ static int gmc_v7_0_soft_reset(void *handle) } if (srbm_soft_reset) { - gmc_v7_0_mc_stop(adev, &save); + gmc_v7_0_mc_stop(adev); if (gmc_v7_0_wait_for_idle((void *)adev)) { dev_warn(adev->dev, "Wait for GMC idle timed out !\n"); } @@ -1175,7 +1161,7 @@ static int gmc_v7_0_soft_reset(void *handle) /* Wait a little for things to settle down */ udelay(50); - gmc_v7_0_mc_resume(adev, &save); + gmc_v7_0_mc_resume(adev); udelay(50); } diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c index cc9f88057cd5..85c937b5e40b 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c @@ -35,6 +35,9 @@ #include "oss/oss_3_0_d.h" #include "oss/oss_3_0_sh_mask.h" +#include "dce/dce_10_0_d.h" +#include "dce/dce_10_0_sh_mask.h" + #include "vid.h" #include "vi.h" @@ -161,14 +164,10 @@ static void gmc_v8_0_init_golden_registers(struct amdgpu_device *adev) } } -static void gmc_v8_0_mc_stop(struct amdgpu_device *adev, - struct amdgpu_mode_mc_save *save) +static void gmc_v8_0_mc_stop(struct amdgpu_device *adev) { u32 blackout; - if (adev->mode_info.num_crtc) - amdgpu_display_stop_mc_access(adev, save); - gmc_v8_0_wait_for_idle(adev); blackout = RREG32(mmMC_SHARED_BLACKOUT_CNTL); @@ -184,8 +183,7 @@ static void gmc_v8_0_mc_stop(struct amdgpu_device *adev, udelay(100); } -static void gmc_v8_0_mc_resume(struct amdgpu_device *adev, - struct amdgpu_mode_mc_save *save) +static void gmc_v8_0_mc_resume(struct amdgpu_device *adev) { u32 tmp; @@ -197,9 +195,6 @@ static void gmc_v8_0_mc_resume(struct amdgpu_device *adev, tmp = REG_SET_FIELD(0, BIF_FB_EN, FB_READ_EN, 1); tmp = REG_SET_FIELD(tmp, BIF_FB_EN, FB_WRITE_EN, 1); WREG32(mmBIF_FB_EN, tmp); - - if (adev->mode_info.num_crtc) - amdgpu_display_resume_mc_access(adev, save); } /** @@ -404,15 +399,20 @@ static int gmc_v8_0_polaris_mc_load_microcode(struct amdgpu_device *adev) static void gmc_v8_0_vram_gtt_location(struct amdgpu_device *adev, struct amdgpu_mc *mc) { + u64 base = 0; + + if (!amdgpu_sriov_vf(adev)) + base = RREG32(mmMC_VM_FB_LOCATION) & 0xFFFF; + base <<= 24; + if (mc->mc_vram_size > 0xFFC0000000ULL) { /* leave room for at least 1024M GTT */ dev_warn(adev->dev, "limiting VRAM\n"); mc->real_vram_size = 0xFFC0000000ULL; mc->mc_vram_size = 0xFFC0000000ULL; } - amdgpu_vram_location(adev, &adev->mc, 0); - adev->mc.gtt_base_align = 0; - amdgpu_gtt_location(adev, mc); + amdgpu_vram_location(adev, &adev->mc, base); + amdgpu_gart_location(adev, mc); } /** @@ -425,7 +425,6 @@ static void gmc_v8_0_vram_gtt_location(struct amdgpu_device *adev, */ static void gmc_v8_0_mc_program(struct amdgpu_device *adev) { - struct amdgpu_mode_mc_save save; u32 tmp; int i, j; @@ -439,13 +438,20 @@ static void gmc_v8_0_mc_program(struct amdgpu_device *adev) } WREG32(mmHDP_REG_COHERENCY_FLUSH_CNTL, 0); - if (adev->mode_info.num_crtc) - amdgpu_display_set_vga_render_state(adev, false); - - gmc_v8_0_mc_stop(adev, &save); if (gmc_v8_0_wait_for_idle((void *)adev)) { dev_warn(adev->dev, "Wait for MC idle timedout !\n"); } + if (adev->mode_info.num_crtc) { + /* Lockout access through VGA aperture*/ + tmp = RREG32(mmVGA_HDP_CONTROL); + tmp = REG_SET_FIELD(tmp, VGA_HDP_CONTROL, VGA_MEMORY_DISABLE, 1); + WREG32(mmVGA_HDP_CONTROL, tmp); + + /* disable VGA render */ + tmp = RREG32(mmVGA_RENDER_CONTROL); + tmp = REG_SET_FIELD(tmp, VGA_RENDER_CONTROL, VGA_VSTATUS_CNTL, 0); + WREG32(mmVGA_RENDER_CONTROL, tmp); + } /* Update configuration */ WREG32(mmMC_VM_SYSTEM_APERTURE_LOW_ADDR, adev->mc.vram_start >> 12); @@ -453,20 +459,23 @@ static void gmc_v8_0_mc_program(struct amdgpu_device *adev) adev->mc.vram_end >> 12); WREG32(mmMC_VM_SYSTEM_APERTURE_DEFAULT_ADDR, adev->vram_scratch.gpu_addr >> 12); - tmp = ((adev->mc.vram_end >> 24) & 0xFFFF) << 16; - tmp |= ((adev->mc.vram_start >> 24) & 0xFFFF); - WREG32(mmMC_VM_FB_LOCATION, tmp); - /* XXX double check these! */ - WREG32(mmHDP_NONSURFACE_BASE, (adev->mc.vram_start >> 8)); - WREG32(mmHDP_NONSURFACE_INFO, (2 << 7) | (1 << 30)); - WREG32(mmHDP_NONSURFACE_SIZE, 0x3FFFFFFF); + + if (amdgpu_sriov_vf(adev)) { + tmp = ((adev->mc.vram_end >> 24) & 0xFFFF) << 16; + tmp |= ((adev->mc.vram_start >> 24) & 0xFFFF); + WREG32(mmMC_VM_FB_LOCATION, tmp); + /* XXX double check these! */ + WREG32(mmHDP_NONSURFACE_BASE, (adev->mc.vram_start >> 8)); + WREG32(mmHDP_NONSURFACE_INFO, (2 << 7) | (1 << 30)); + WREG32(mmHDP_NONSURFACE_SIZE, 0x3FFFFFFF); + } + WREG32(mmMC_VM_AGP_BASE, 0); WREG32(mmMC_VM_AGP_TOP, 0x0FFFFFFF); WREG32(mmMC_VM_AGP_BOT, 0x0FFFFFFF); if (gmc_v8_0_wait_for_idle((void *)adev)) { dev_warn(adev->dev, "Wait for MC idle timedout !\n"); } - gmc_v8_0_mc_resume(adev, &save); WREG32(mmBIF_FB_EN, BIF_FB_EN__FB_READ_EN_MASK | BIF_FB_EN__FB_WRITE_EN_MASK); @@ -553,15 +562,7 @@ static int gmc_v8_0_mc_init(struct amdgpu_device *adev) if (adev->mc.visible_vram_size > adev->mc.real_vram_size) adev->mc.visible_vram_size = adev->mc.real_vram_size; - /* unless the user had overridden it, set the gart - * size equal to the 1024 or vram, whichever is larger. - */ - if (amdgpu_gart_size == -1) - adev->mc.gtt_size = max((AMDGPU_DEFAULT_GTT_SIZE_MB << 20), - adev->mc.mc_vram_size); - else - adev->mc.gtt_size = (uint64_t)amdgpu_gart_size << 20; - + amdgpu_gart_set_defaults(adev); gmc_v8_0_vram_gtt_location(adev, &adev->mc); return 0; @@ -813,8 +814,8 @@ static int gmc_v8_0_gart_enable(struct amdgpu_device *adev) tmp = REG_SET_FIELD(tmp, VM_L2_CNTL4, VMC_TAP_CONTEXT1_PTE_REQUEST_SNOOP, 0); WREG32(mmVM_L2_CNTL4, tmp); /* setup context0 */ - WREG32(mmVM_CONTEXT0_PAGE_TABLE_START_ADDR, adev->mc.gtt_start >> 12); - WREG32(mmVM_CONTEXT0_PAGE_TABLE_END_ADDR, adev->mc.gtt_end >> 12); + WREG32(mmVM_CONTEXT0_PAGE_TABLE_START_ADDR, adev->mc.gart_start >> 12); + WREG32(mmVM_CONTEXT0_PAGE_TABLE_END_ADDR, adev->mc.gart_end >> 12); WREG32(mmVM_CONTEXT0_PAGE_TABLE_BASE_ADDR, adev->gart.table_addr >> 12); WREG32(mmVM_CONTEXT0_PROTECTION_FAULT_DEFAULT_ADDR, (u32)(adev->dummy_page.addr >> 12)); @@ -869,7 +870,7 @@ static int gmc_v8_0_gart_enable(struct amdgpu_device *adev) gmc_v8_0_gart_flush_gpu_tlb(adev, 0); DRM_INFO("PCIE GART of %uM enabled (table at 0x%016llX).\n", - (unsigned)(adev->mc.gtt_size >> 20), + (unsigned)(adev->mc.gart_size >> 20), (unsigned long long)adev->gart.table_addr); adev->gart.ready = true; return 0; @@ -1260,7 +1261,7 @@ static int gmc_v8_0_pre_soft_reset(void *handle) if (!adev->mc.srbm_soft_reset) return 0; - gmc_v8_0_mc_stop(adev, &adev->mc.save); + gmc_v8_0_mc_stop(adev); if (gmc_v8_0_wait_for_idle(adev)) { dev_warn(adev->dev, "Wait for GMC idle timed out !\n"); } @@ -1306,7 +1307,7 @@ static int gmc_v8_0_post_soft_reset(void *handle) if (!adev->mc.srbm_soft_reset) return 0; - gmc_v8_0_mc_resume(adev, &adev->mc.save); + gmc_v8_0_mc_resume(adev); return 0; } diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c index 175ba5f9691c..c22899a08106 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c @@ -23,11 +23,14 @@ #include <linux/firmware.h> #include "amdgpu.h" #include "gmc_v9_0.h" +#include "amdgpu_atomfirmware.h" #include "vega10/soc15ip.h" #include "vega10/HDP/hdp_4_0_offset.h" #include "vega10/HDP/hdp_4_0_sh_mask.h" #include "vega10/GC/gc_9_0_sh_mask.h" +#include "vega10/DC/dce_12_0_offset.h" +#include "vega10/DC/dce_12_0_sh_mask.h" #include "vega10/vega10_enum.h" #include "soc15_common.h" @@ -419,8 +422,7 @@ static void gmc_v9_0_vram_gtt_location(struct amdgpu_device *adev, if (!amdgpu_sriov_vf(adev)) base = mmhub_v1_0_get_fb_location(adev); amdgpu_vram_location(adev, &adev->mc, base); - adev->mc.gtt_base_align = 0; - amdgpu_gtt_location(adev, mc); + amdgpu_gart_location(adev, mc); /* base offset of vram pages */ if (adev->flags & AMD_IS_APU) adev->vm_manager.vram_base_offset = gfxhub_v1_0_get_mc_fb_offset(adev); @@ -442,43 +444,46 @@ static int gmc_v9_0_mc_init(struct amdgpu_device *adev) u32 tmp; int chansize, numchan; - /* hbm memory channel size */ - chansize = 128; - - tmp = RREG32_SOC15(DF, 0, mmDF_CS_AON0_DramBaseAddress0); - tmp &= DF_CS_AON0_DramBaseAddress0__IntLvNumChan_MASK; - tmp >>= DF_CS_AON0_DramBaseAddress0__IntLvNumChan__SHIFT; - switch (tmp) { - case 0: - default: - numchan = 1; - break; - case 1: - numchan = 2; - break; - case 2: - numchan = 0; - break; - case 3: - numchan = 4; - break; - case 4: - numchan = 0; - break; - case 5: - numchan = 8; - break; - case 6: - numchan = 0; - break; - case 7: - numchan = 16; - break; - case 8: - numchan = 2; - break; + adev->mc.vram_width = amdgpu_atomfirmware_get_vram_width(adev); + if (!adev->mc.vram_width) { + /* hbm memory channel size */ + chansize = 128; + + tmp = RREG32_SOC15(DF, 0, mmDF_CS_AON0_DramBaseAddress0); + tmp &= DF_CS_AON0_DramBaseAddress0__IntLvNumChan_MASK; + tmp >>= DF_CS_AON0_DramBaseAddress0__IntLvNumChan__SHIFT; + switch (tmp) { + case 0: + default: + numchan = 1; + break; + case 1: + numchan = 2; + break; + case 2: + numchan = 0; + break; + case 3: + numchan = 4; + break; + case 4: + numchan = 0; + break; + case 5: + numchan = 8; + break; + case 6: + numchan = 0; + break; + case 7: + numchan = 16; + break; + case 8: + numchan = 2; + break; + } + adev->mc.vram_width = numchan * chansize; } - adev->mc.vram_width = numchan * chansize; /* Could aper size report 0 ? */ adev->mc.aper_base = pci_resource_start(adev->pdev, 0); @@ -494,15 +499,7 @@ static int gmc_v9_0_mc_init(struct amdgpu_device *adev) if (adev->mc.visible_vram_size > adev->mc.real_vram_size) adev->mc.visible_vram_size = adev->mc.real_vram_size; - /* unless the user had overridden it, set the gart - * size equal to the 1024 or vram, whichever is larger. - */ - if (amdgpu_gart_size == -1) - adev->mc.gtt_size = max((AMDGPU_DEFAULT_GTT_SIZE_MB << 20), - adev->mc.mc_vram_size); - else - adev->mc.gtt_size = (uint64_t)amdgpu_gart_size << 20; - + amdgpu_gart_set_defaults(adev); gmc_v9_0_vram_gtt_location(adev, &adev->mc); return 0; @@ -537,10 +534,20 @@ static int gmc_v9_0_sw_init(void *handle) spin_lock_init(&adev->mc.invalidate_lock); - if (adev->flags & AMD_IS_APU) { + switch (adev->asic_type) { + case CHIP_RAVEN: adev->mc.vram_type = AMDGPU_VRAM_TYPE_UNKNOWN; - amdgpu_vm_adjust_size(adev, 64); - } else { + if (adev->rev_id == 0x0 || adev->rev_id == 0x1) { + adev->vm_manager.vm_size = 1U << 18; + adev->vm_manager.block_size = 9; + adev->vm_manager.num_level = 3; + } else { + /* vm_size is 64GB for legacy 2-level page support*/ + amdgpu_vm_adjust_size(adev, 64); + adev->vm_manager.num_level = 1; + } + break; + case CHIP_VEGA10: /* XXX Don't know how to get VRAM type yet. */ adev->mc.vram_type = AMDGPU_VRAM_TYPE_HBM; /* @@ -550,11 +557,16 @@ static int gmc_v9_0_sw_init(void *handle) */ adev->vm_manager.vm_size = 1U << 18; adev->vm_manager.block_size = 9; - DRM_INFO("vm size is %llu GB, block size is %u-bit\n", - adev->vm_manager.vm_size, - adev->vm_manager.block_size); + adev->vm_manager.num_level = 3; + break; + default: + break; } + DRM_INFO("vm size is %llu GB, block size is %u-bit\n", + adev->vm_manager.vm_size, + adev->vm_manager.block_size); + /* This interrupt is VMC page fault.*/ r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_VMC, 0, &adev->mc.vm_fault); @@ -619,11 +631,6 @@ static int gmc_v9_0_sw_init(void *handle) 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) - adev->vm_manager.num_level = 1; - else - adev->vm_manager.num_level = 3; amdgpu_vm_manager_init(adev); return 0; @@ -731,7 +738,7 @@ static int gmc_v9_0_gart_enable(struct amdgpu_device *adev) gmc_v9_0_gart_flush_gpu_tlb(adev, 0); DRM_INFO("PCIE GART of %uM enabled (table at 0x%016llX).\n", - (unsigned)(adev->mc.gtt_size >> 20), + (unsigned)(adev->mc.gart_size >> 20), (unsigned long long)adev->gart.table_addr); adev->gart.ready = true; return 0; @@ -745,6 +752,20 @@ static int gmc_v9_0_hw_init(void *handle) /* The sequence of these two function calls matters.*/ gmc_v9_0_init_golden_registers(adev); + if (adev->mode_info.num_crtc) { + u32 tmp; + + /* Lockout access through VGA aperture*/ + tmp = RREG32_SOC15(DCE, 0, mmVGA_HDP_CONTROL); + tmp = REG_SET_FIELD(tmp, VGA_HDP_CONTROL, VGA_MEMORY_DISABLE, 1); + WREG32_SOC15(DCE, 0, mmVGA_HDP_CONTROL, tmp); + + /* disable VGA render */ + tmp = RREG32_SOC15(DCE, 0, mmVGA_RENDER_CONTROL); + tmp = REG_SET_FIELD(tmp, VGA_RENDER_CONTROL, VGA_VSTATUS_CNTL, 0); + WREG32_SOC15(DCE, 0, mmVGA_RENDER_CONTROL, tmp); + } + r = gmc_v9_0_gart_enable(adev); return r; diff --git a/drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.c b/drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.c index 9804318f3488..ad8def3cc343 100644 --- a/drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.c +++ b/drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.c @@ -69,14 +69,14 @@ static void mmhub_v1_0_init_gart_aperture_regs(struct amdgpu_device *adev) mmhub_v1_0_init_gart_pt_regs(adev); WREG32_SOC15(MMHUB, 0, mmVM_CONTEXT0_PAGE_TABLE_START_ADDR_LO32, - (u32)(adev->mc.gtt_start >> 12)); + (u32)(adev->mc.gart_start >> 12)); WREG32_SOC15(MMHUB, 0, mmVM_CONTEXT0_PAGE_TABLE_START_ADDR_HI32, - (u32)(adev->mc.gtt_start >> 44)); + (u32)(adev->mc.gart_start >> 44)); WREG32_SOC15(MMHUB, 0, mmVM_CONTEXT0_PAGE_TABLE_END_ADDR_LO32, - (u32)(adev->mc.gtt_end >> 12)); + (u32)(adev->mc.gart_end >> 12)); WREG32_SOC15(MMHUB, 0, mmVM_CONTEXT0_PAGE_TABLE_END_ADDR_HI32, - (u32)(adev->mc.gtt_end >> 44)); + (u32)(adev->mc.gart_end >> 44)); } static void mmhub_v1_0_init_system_aperture_regs(struct amdgpu_device *adev) @@ -143,7 +143,7 @@ static void mmhub_v1_0_init_cache_regs(struct amdgpu_device *adev) /* Setup L2 cache */ tmp = RREG32_SOC15(MMHUB, 0, mmVM_L2_CNTL); tmp = REG_SET_FIELD(tmp, VM_L2_CNTL, ENABLE_L2_CACHE, 1); - tmp = REG_SET_FIELD(tmp, VM_L2_CNTL, ENABLE_L2_FRAGMENT_PROCESSING, 0); + tmp = REG_SET_FIELD(tmp, VM_L2_CNTL, ENABLE_L2_FRAGMENT_PROCESSING, 1); /* XXX for emulation, Refer to closed source code.*/ tmp = REG_SET_FIELD(tmp, VM_L2_CNTL, L2_PDE0_CACHE_TAG_GENERATION_MODE, 0); @@ -158,6 +158,8 @@ static void mmhub_v1_0_init_cache_regs(struct amdgpu_device *adev) WREG32_SOC15(MMHUB, 0, mmVM_L2_CNTL2, tmp); tmp = mmVM_L2_CNTL3_DEFAULT; + tmp = REG_SET_FIELD(tmp, VM_L2_CNTL3, BANK_SELECT, 12); + tmp = REG_SET_FIELD(tmp, VM_L2_CNTL3, L2_CACHE_BIGK_FRAGMENT_SIZE, 9); WREG32_SOC15(MMHUB, 0, mmVM_L2_CNTL3, tmp); tmp = mmVM_L2_CNTL4_DEFAULT; @@ -222,6 +224,9 @@ static void mmhub_v1_0_setup_vmid_config(struct amdgpu_device *adev) tmp = REG_SET_FIELD(tmp, VM_CONTEXT1_CNTL, PAGE_TABLE_BLOCK_SIZE, adev->vm_manager.block_size - 9); + /* Send no-retry XNACK on fault to suppress VM fault storm. */ + tmp = REG_SET_FIELD(tmp, VM_CONTEXT1_CNTL, + RETRY_PERMISSION_OR_INVALID_PAGE_FAULT, 0); WREG32_SOC15_OFFSET(MMHUB, 0, mmVM_CONTEXT1_CNTL, i, tmp); WREG32_SOC15_OFFSET(MMHUB, 0, mmVM_CONTEXT1_PAGE_TABLE_START_ADDR_LO32, i*2, 0); WREG32_SOC15_OFFSET(MMHUB, 0, mmVM_CONTEXT1_PAGE_TABLE_START_ADDR_HI32, i*2, 0); @@ -245,28 +250,28 @@ static void mmhub_v1_0_program_invalidation(struct amdgpu_device *adev) } struct pctl_data { - uint32_t index; - uint32_t data; + uint32_t index; + uint32_t data; }; -const struct pctl_data pctl0_data[] = { - {0x0, 0x7a640}, - {0x9, 0x2a64a}, - {0xd, 0x2a680}, - {0x11, 0x6a684}, - {0x19, 0xea68e}, - {0x29, 0xa69e}, - {0x2b, 0x34a6c0}, - {0x61, 0x83a707}, - {0xe6, 0x8a7a4}, - {0xf0, 0x1a7b8}, - {0xf3, 0xfa7cc}, - {0x104, 0x17a7dd}, - {0x11d, 0xa7dc}, - {0x11f, 0x12a7f5}, - {0x133, 0xa808}, - {0x135, 0x12a810}, - {0x149, 0x7a82c} +static const struct pctl_data pctl0_data[] = { + {0x0, 0x7a640}, + {0x9, 0x2a64a}, + {0xd, 0x2a680}, + {0x11, 0x6a684}, + {0x19, 0xea68e}, + {0x29, 0xa69e}, + {0x2b, 0x34a6c0}, + {0x61, 0x83a707}, + {0xe6, 0x8a7a4}, + {0xf0, 0x1a7b8}, + {0xf3, 0xfa7cc}, + {0x104, 0x17a7dd}, + {0x11d, 0xa7dc}, + {0x11f, 0x12a7f5}, + {0x133, 0xa808}, + {0x135, 0x12a810}, + {0x149, 0x7a82c} }; #define PCTL0_DATA_LEN (sizeof(pctl0_data)/sizeof(pctl0_data[0])) @@ -274,32 +279,39 @@ const struct pctl_data pctl0_data[] = { #define PCTL0_STCTRL_REG_SAVE_RANGE0_BASE 0xa640 #define PCTL0_STCTRL_REG_SAVE_RANGE0_LIMIT 0xa833 -const struct pctl_data pctl1_data[] = { - {0x0, 0x39a000}, - {0x3b, 0x44a040}, - {0x81, 0x2a08d}, - {0x85, 0x6ba094}, - {0xf2, 0x18a100}, - {0x10c, 0x4a132}, - {0x112, 0xca141}, - {0x120, 0x2fa158}, - {0x151, 0x17a1d0}, - {0x16a, 0x1a1e9}, - {0x16d, 0x13a1ec}, - {0x182, 0x7a201}, - {0x18b, 0x3a20a}, - {0x190, 0x7a580}, - {0x199, 0xa590}, - {0x19b, 0x4a594}, - {0x1a1, 0x1a59c}, - {0x1a4, 0x7a82c}, - {0x1ad, 0xfa7cc}, - {0x1be, 0x17a7dd}, - {0x1d7, 0x12a810} +static const struct pctl_data pctl1_data[] = { + {0x0, 0x39a000}, + {0x3b, 0x44a040}, + {0x81, 0x2a08d}, + {0x85, 0x6ba094}, + {0xf2, 0x18a100}, + {0x10c, 0x4a132}, + {0x112, 0xca141}, + {0x120, 0x2fa158}, + {0x151, 0x17a1d0}, + {0x16a, 0x1a1e9}, + {0x16d, 0x13a1ec}, + {0x182, 0x7a201}, + {0x18b, 0x3a20a}, + {0x190, 0x7a580}, + {0x199, 0xa590}, + {0x19b, 0x4a594}, + {0x1a1, 0x1a59c}, + {0x1a4, 0x7a82c}, + {0x1ad, 0xfa7cc}, + {0x1be, 0x17a7dd}, + {0x1d7, 0x12a810}, + {0x1eb, 0x4000a7e1}, + {0x1ec, 0x5000a7f5}, + {0x1ed, 0x4000a7e2}, + {0x1ee, 0x5000a7dc}, + {0x1ef, 0x4000a7e3}, + {0x1f0, 0x5000a7f6}, + {0x1f1, 0x5000a7e4} }; #define PCTL1_DATA_LEN (sizeof(pctl1_data)/sizeof(pctl1_data[0])) -#define PCTL1_RENG_EXEC_END_PTR 0x1ea +#define PCTL1_RENG_EXEC_END_PTR 0x1f1 #define PCTL1_STCTRL_REG_SAVE_RANGE0_BASE 0xa000 #define PCTL1_STCTRL_REG_SAVE_RANGE0_LIMIT 0xa20d #define PCTL1_STCTRL_REG_SAVE_RANGE1_BASE 0xa580 diff --git a/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c b/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c index bde3ca3c21c1..2812d88a8bdd 100644 --- a/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c +++ b/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c @@ -72,21 +72,6 @@ static void xgpu_ai_mailbox_set_valid(struct amdgpu_device *adev, bool val) reg); } -static void xgpu_ai_mailbox_trans_msg(struct amdgpu_device *adev, - enum idh_request req) -{ - u32 reg; - - reg = RREG32_NO_KIQ(SOC15_REG_OFFSET(NBIO, 0, - mmBIF_BX_PF0_MAILBOX_MSGBUF_TRN_DW0)); - reg = REG_SET_FIELD(reg, BIF_BX_PF0_MAILBOX_MSGBUF_TRN_DW0, - MSGBUF_DATA, req); - WREG32_NO_KIQ(SOC15_REG_OFFSET(NBIO, 0, mmBIF_BX_PF0_MAILBOX_MSGBUF_TRN_DW0), - reg); - - xgpu_ai_mailbox_set_valid(adev, true); -} - static int xgpu_ai_mailbox_rcv_msg(struct amdgpu_device *adev, enum idh_event event) { @@ -154,13 +139,25 @@ static int xgpu_ai_poll_msg(struct amdgpu_device *adev, enum idh_event event) return r; } - -static int xgpu_ai_send_access_requests(struct amdgpu_device *adev, - enum idh_request req) -{ +static void xgpu_ai_mailbox_trans_msg (struct amdgpu_device *adev, + enum idh_request req, u32 data1, u32 data2, u32 data3) { + u32 reg; int r; - xgpu_ai_mailbox_trans_msg(adev, req); + reg = RREG32_NO_KIQ(SOC15_REG_OFFSET(NBIO, 0, + mmBIF_BX_PF0_MAILBOX_MSGBUF_TRN_DW0)); + reg = REG_SET_FIELD(reg, BIF_BX_PF0_MAILBOX_MSGBUF_TRN_DW0, + MSGBUF_DATA, req); + WREG32_NO_KIQ(SOC15_REG_OFFSET(NBIO, 0, mmBIF_BX_PF0_MAILBOX_MSGBUF_TRN_DW0), + reg); + WREG32_NO_KIQ(SOC15_REG_OFFSET(NBIO, 0, mmBIF_BX_PF0_MAILBOX_MSGBUF_TRN_DW1), + data1); + WREG32_NO_KIQ(SOC15_REG_OFFSET(NBIO, 0, mmBIF_BX_PF0_MAILBOX_MSGBUF_TRN_DW2), + data2); + WREG32_NO_KIQ(SOC15_REG_OFFSET(NBIO, 0, mmBIF_BX_PF0_MAILBOX_MSGBUF_TRN_DW3), + data3); + + xgpu_ai_mailbox_set_valid(adev, true); /* start to poll ack */ r = xgpu_ai_poll_ack(adev); @@ -168,6 +165,14 @@ static int xgpu_ai_send_access_requests(struct amdgpu_device *adev, pr_err("Doesn't get ack from pf, continue\n"); xgpu_ai_mailbox_set_valid(adev, false); +} + +static int xgpu_ai_send_access_requests(struct amdgpu_device *adev, + enum idh_request req) +{ + int r; + + xgpu_ai_mailbox_trans_msg(adev, req, 0, 0, 0); /* start to check msg if request is idh_req_gpu_init_access */ if (req == IDH_REQ_GPU_INIT_ACCESS || @@ -342,4 +347,5 @@ const struct amdgpu_virt_ops xgpu_ai_virt_ops = { .req_full_gpu = xgpu_ai_request_full_gpu_access, .rel_full_gpu = xgpu_ai_release_full_gpu_access, .reset_gpu = xgpu_ai_request_reset, + .trans_msg = xgpu_ai_mailbox_trans_msg, }; diff --git a/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.h b/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.h index 9aefc44d2c34..1e91b9a1c591 100644 --- a/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.h +++ b/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.h @@ -31,7 +31,9 @@ enum idh_request { IDH_REL_GPU_INIT_ACCESS, IDH_REQ_GPU_FINI_ACCESS, IDH_REL_GPU_FINI_ACCESS, - IDH_REQ_GPU_RESET_ACCESS + IDH_REQ_GPU_RESET_ACCESS, + + IDH_LOG_VF_ERROR = 200, }; enum idh_event { diff --git a/drivers/gpu/drm/amd/amdgpu/mxgpu_vi.c b/drivers/gpu/drm/amd/amdgpu/mxgpu_vi.c index 171a658135b5..c25a831f94ec 100644 --- a/drivers/gpu/drm/amd/amdgpu/mxgpu_vi.c +++ b/drivers/gpu/drm/amd/amdgpu/mxgpu_vi.c @@ -613,4 +613,5 @@ const struct amdgpu_virt_ops xgpu_vi_virt_ops = { .req_full_gpu = xgpu_vi_request_full_gpu_access, .rel_full_gpu = xgpu_vi_release_full_gpu_access, .reset_gpu = xgpu_vi_request_reset, + .trans_msg = NULL, /* Does not need to trans VF errors to host. */ }; diff --git a/drivers/gpu/drm/amd/amdgpu/mxgpu_vi.h b/drivers/gpu/drm/amd/amdgpu/mxgpu_vi.h index 2db741131bc6..c791d73d2d54 100644 --- a/drivers/gpu/drm/amd/amdgpu/mxgpu_vi.h +++ b/drivers/gpu/drm/amd/amdgpu/mxgpu_vi.h @@ -32,7 +32,9 @@ enum idh_request { IDH_REL_GPU_INIT_ACCESS, IDH_REQ_GPU_FINI_ACCESS, IDH_REL_GPU_FINI_ACCESS, - IDH_REQ_GPU_RESET_ACCESS + IDH_REQ_GPU_RESET_ACCESS, + + IDH_LOG_VF_ERROR = 200, }; /* VI mailbox messages data */ diff --git a/drivers/gpu/drm/amd/amdgpu/nbio_v6_1.c b/drivers/gpu/drm/amd/amdgpu/nbio_v6_1.c index 1e272f785def..045988b18bc3 100644 --- a/drivers/gpu/drm/amd/amdgpu/nbio_v6_1.c +++ b/drivers/gpu/drm/amd/amdgpu/nbio_v6_1.c @@ -32,6 +32,7 @@ #define smnCPM_CONTROL 0x11180460 #define smnPCIE_CNTL2 0x11180070 +#define smnPCIE_CONFIG_CNTL 0x11180044 u32 nbio_v6_1_get_rev_id(struct amdgpu_device *adev) { @@ -67,7 +68,7 @@ void nbio_v6_1_mc_access_enable(struct amdgpu_device *adev, bool enable) void nbio_v6_1_hdp_flush(struct amdgpu_device *adev) { - WREG32_SOC15(NBIO, 0, mmBIF_BX_PF0_HDP_MEM_COHERENCY_FLUSH_CNTL, 0); + WREG32_SOC15_NO_KIQ(NBIO, 0, mmBIF_BX_PF0_HDP_MEM_COHERENCY_FLUSH_CNTL, 0); } u32 nbio_v6_1_get_memsize(struct amdgpu_device *adev) @@ -256,3 +257,15 @@ void nbio_v6_1_detect_hw_virt(struct amdgpu_device *adev) adev->virt.caps |= AMDGPU_PASSTHROUGH_MODE; } } + +void nbio_v6_1_init_registers(struct amdgpu_device *adev) +{ + uint32_t def, data; + + def = data = RREG32_PCIE(smnPCIE_CONFIG_CNTL); + data = REG_SET_FIELD(data, PCIE_CONFIG_CNTL, CI_SWUS_MAX_READ_REQUEST_SIZE_MODE, 1); + data = REG_SET_FIELD(data, PCIE_CONFIG_CNTL, CI_SWUS_MAX_READ_REQUEST_SIZE_PRIV, 1); + + if (def != data) + WREG32_PCIE(smnPCIE_CONFIG_CNTL, data); +} diff --git a/drivers/gpu/drm/amd/amdgpu/nbio_v6_1.h b/drivers/gpu/drm/amd/amdgpu/nbio_v6_1.h index f6f8bc045518..686e4b4d296a 100644 --- a/drivers/gpu/drm/amd/amdgpu/nbio_v6_1.h +++ b/drivers/gpu/drm/amd/amdgpu/nbio_v6_1.h @@ -50,5 +50,6 @@ void nbio_v6_1_update_medium_grain_clock_gating(struct amdgpu_device *adev, bool void nbio_v6_1_update_medium_grain_light_sleep(struct amdgpu_device *adev, bool enable); void nbio_v6_1_get_clockgating_state(struct amdgpu_device *adev, u32 *flags); void nbio_v6_1_detect_hw_virt(struct amdgpu_device *adev); +void nbio_v6_1_init_registers(struct amdgpu_device *adev); #endif diff --git a/drivers/gpu/drm/amd/amdgpu/nbio_v7_0.c b/drivers/gpu/drm/amd/amdgpu/nbio_v7_0.c index aa04632523fa..11b70d601922 100644 --- a/drivers/gpu/drm/amd/amdgpu/nbio_v7_0.c +++ b/drivers/gpu/drm/amd/amdgpu/nbio_v7_0.c @@ -65,7 +65,7 @@ void nbio_v7_0_mc_access_enable(struct amdgpu_device *adev, bool enable) void nbio_v7_0_hdp_flush(struct amdgpu_device *adev) { - WREG32_SOC15(NBIO, 0, mmHDP_MEM_COHERENCY_FLUSH_CNTL, 0); + WREG32_SOC15_NO_KIQ(NBIO, 0, mmHDP_MEM_COHERENCY_FLUSH_CNTL, 0); } u32 nbio_v7_0_get_memsize(struct amdgpu_device *adev) diff --git a/drivers/gpu/drm/amd/amdgpu/psp_v10_0.c b/drivers/gpu/drm/amd/amdgpu/psp_v10_0.c index 2258323a3c26..f7cf994b1da2 100644 --- a/drivers/gpu/drm/amd/amdgpu/psp_v10_0.c +++ b/drivers/gpu/drm/amd/amdgpu/psp_v10_0.c @@ -86,6 +86,52 @@ psp_v10_0_get_fw_type(struct amdgpu_firmware_info *ucode, enum psp_gfx_fw_type * return 0; } +int psp_v10_0_init_microcode(struct psp_context *psp) +{ + struct amdgpu_device *adev = psp->adev; + const char *chip_name; + char fw_name[30]; + int err = 0; + const struct psp_firmware_header_v1_0 *hdr; + + DRM_DEBUG("\n"); + + switch (adev->asic_type) { + case CHIP_RAVEN: + chip_name = "raven"; + break; + default: BUG(); + } + + snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_asd.bin", chip_name); + err = request_firmware(&adev->psp.asd_fw, fw_name, adev->dev); + if (err) + goto out; + + err = amdgpu_ucode_validate(adev->psp.asd_fw); + if (err) + goto out; + + hdr = (const struct psp_firmware_header_v1_0 *)adev->psp.asd_fw->data; + adev->psp.asd_fw_version = le32_to_cpu(hdr->header.ucode_version); + adev->psp.asd_feature_version = le32_to_cpu(hdr->ucode_feature_version); + adev->psp.asd_ucode_size = le32_to_cpu(hdr->header.ucode_size_bytes); + adev->psp.asd_start_addr = (uint8_t *)hdr + + le32_to_cpu(hdr->header.ucode_array_offset_bytes); + + return 0; +out: + if (err) { + dev_err(adev->dev, + "psp v10.0: Failed to load firmware \"%s\"\n", + fw_name); + release_firmware(adev->psp.asd_fw); + adev->psp.asd_fw = NULL; + } + + return err; +} + int psp_v10_0_prep_cmd_buf(struct amdgpu_firmware_info *ucode, struct psp_gfx_cmd_resp *cmd) { int ret; @@ -110,7 +156,6 @@ int psp_v10_0_prep_cmd_buf(struct amdgpu_firmware_info *ucode, struct psp_gfx_cm int psp_v10_0_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; @@ -130,6 +175,16 @@ int psp_v10_0_ring_init(struct psp_context *psp, enum psp_ring_type ring_type) return ret; } + return 0; +} + +int psp_v10_0_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(MP0, 0, mmMP0_SMN_C2PMSG_69, psp_ring_reg); @@ -143,13 +198,42 @@ int psp_v10_0_ring_init(struct psp_context *psp, enum psp_ring_type ring_type) psp_ring_reg = ring_type; psp_ring_reg = psp_ring_reg << 16; WREG32_SOC15(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 */ - psp_ring_reg = 0; - while ((psp_ring_reg & 0x80000000) == 0) { - psp_ring_reg = RREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_64); - } + ret = psp_wait_for(psp, SOC15_REG_OFFSET(MP0, 0, mmMP0_SMN_C2PMSG_64), + 0x80000000, 0x8000FFFF, false); - return 0; + return ret; +} + +int psp_v10_0_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(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); + + amdgpu_bo_free_kernel(&adev->firmware.rbuf, + &ring->ring_mem_mc_addr, + (void **)&ring->ring_mem); + + return ret; } int psp_v10_0_cmd_submit(struct psp_context *psp, diff --git a/drivers/gpu/drm/amd/amdgpu/psp_v10_0.h b/drivers/gpu/drm/amd/amdgpu/psp_v10_0.h index 2022b7b7151e..e76cde2f01f9 100644 --- a/drivers/gpu/drm/amd/amdgpu/psp_v10_0.h +++ b/drivers/gpu/drm/amd/amdgpu/psp_v10_0.h @@ -27,10 +27,15 @@ #include "amdgpu_psp.h" +extern int psp_v10_0_init_microcode(struct psp_context *psp); extern int psp_v10_0_prep_cmd_buf(struct amdgpu_firmware_info *ucode, struct psp_gfx_cmd_resp *cmd); extern int psp_v10_0_ring_init(struct psp_context *psp, enum psp_ring_type ring_type); +extern int psp_v10_0_ring_create(struct psp_context *psp, + enum psp_ring_type ring_type); +extern int psp_v10_0_ring_destroy(struct psp_context *psp, + enum psp_ring_type ring_type); extern int psp_v10_0_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.c b/drivers/gpu/drm/amd/amdgpu/psp_v3_1.c index c98d77d0c8f8..2a535a4b8d5b 100644 --- a/drivers/gpu/drm/amd/amdgpu/psp_v3_1.c +++ b/drivers/gpu/drm/amd/amdgpu/psp_v3_1.c @@ -237,11 +237,9 @@ int psp_v3_1_bootloader_load_sos(struct psp_context *psp) /* there might be handshake issue with hardware which needs delay */ mdelay(20); -#if 0 ret = psp_wait_for(psp, SOC15_REG_OFFSET(MP0, 0, mmMP0_SMN_C2PMSG_81), RREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_81), 0, true); -#endif return ret; } @@ -341,10 +339,10 @@ int psp_v3_1_ring_destroy(struct psp_context *psp, enum psp_ring_type ring_type) 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); + amdgpu_bo_free_kernel(&adev->firmware.rbuf, + &ring->ring_mem_mc_addr, + (void **)&ring->ring_mem); + return ret; } diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c b/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c index 1d766ae98dc8..b1de44f22824 100644 --- a/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c +++ b/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c @@ -551,17 +551,53 @@ static void sdma_v3_0_rlc_stop(struct amdgpu_device *adev) */ static void sdma_v3_0_ctx_switch_enable(struct amdgpu_device *adev, bool enable) { - u32 f32_cntl; + u32 f32_cntl, phase_quantum = 0; int i; + if (amdgpu_sdma_phase_quantum) { + unsigned value = amdgpu_sdma_phase_quantum; + unsigned unit = 0; + + while (value > (SDMA0_PHASE0_QUANTUM__VALUE_MASK >> + SDMA0_PHASE0_QUANTUM__VALUE__SHIFT)) { + value = (value + 1) >> 1; + unit++; + } + if (unit > (SDMA0_PHASE0_QUANTUM__UNIT_MASK >> + SDMA0_PHASE0_QUANTUM__UNIT__SHIFT)) { + value = (SDMA0_PHASE0_QUANTUM__VALUE_MASK >> + SDMA0_PHASE0_QUANTUM__VALUE__SHIFT); + unit = (SDMA0_PHASE0_QUANTUM__UNIT_MASK >> + SDMA0_PHASE0_QUANTUM__UNIT__SHIFT); + WARN_ONCE(1, + "clamping sdma_phase_quantum to %uK clock cycles\n", + value << unit); + } + phase_quantum = + value << SDMA0_PHASE0_QUANTUM__VALUE__SHIFT | + unit << SDMA0_PHASE0_QUANTUM__UNIT__SHIFT; + } + for (i = 0; i < adev->sdma.num_instances; i++) { f32_cntl = RREG32(mmSDMA0_CNTL + sdma_offsets[i]); - if (enable) + if (enable) { f32_cntl = REG_SET_FIELD(f32_cntl, SDMA0_CNTL, AUTO_CTXSW_ENABLE, 1); - else + f32_cntl = REG_SET_FIELD(f32_cntl, SDMA0_CNTL, + ATC_L1_ENABLE, 1); + if (amdgpu_sdma_phase_quantum) { + WREG32(mmSDMA0_PHASE0_QUANTUM + sdma_offsets[i], + phase_quantum); + WREG32(mmSDMA0_PHASE1_QUANTUM + sdma_offsets[i], + phase_quantum); + } + } else { f32_cntl = REG_SET_FIELD(f32_cntl, SDMA0_CNTL, AUTO_CTXSW_ENABLE, 0); + f32_cntl = REG_SET_FIELD(f32_cntl, SDMA0_CNTL, + ATC_L1_ENABLE, 1); + } + WREG32(mmSDMA0_CNTL + sdma_offsets[i], f32_cntl); } } diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c b/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c index 4a65697ccc94..591f3e7fb508 100644 --- a/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c +++ b/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c @@ -493,13 +493,45 @@ static void sdma_v4_0_rlc_stop(struct amdgpu_device *adev) */ static void sdma_v4_0_ctx_switch_enable(struct amdgpu_device *adev, bool enable) { - u32 f32_cntl; + u32 f32_cntl, phase_quantum = 0; int i; + if (amdgpu_sdma_phase_quantum) { + unsigned value = amdgpu_sdma_phase_quantum; + unsigned unit = 0; + + while (value > (SDMA0_PHASE0_QUANTUM__VALUE_MASK >> + SDMA0_PHASE0_QUANTUM__VALUE__SHIFT)) { + value = (value + 1) >> 1; + unit++; + } + if (unit > (SDMA0_PHASE0_QUANTUM__UNIT_MASK >> + SDMA0_PHASE0_QUANTUM__UNIT__SHIFT)) { + value = (SDMA0_PHASE0_QUANTUM__VALUE_MASK >> + SDMA0_PHASE0_QUANTUM__VALUE__SHIFT); + unit = (SDMA0_PHASE0_QUANTUM__UNIT_MASK >> + SDMA0_PHASE0_QUANTUM__UNIT__SHIFT); + WARN_ONCE(1, + "clamping sdma_phase_quantum to %uK clock cycles\n", + value << unit); + } + phase_quantum = + value << SDMA0_PHASE0_QUANTUM__VALUE__SHIFT | + unit << SDMA0_PHASE0_QUANTUM__UNIT__SHIFT; + } + for (i = 0; i < adev->sdma.num_instances; i++) { f32_cntl = RREG32(sdma_v4_0_get_reg_offset(i, mmSDMA0_CNTL)); f32_cntl = REG_SET_FIELD(f32_cntl, SDMA0_CNTL, AUTO_CTXSW_ENABLE, enable ? 1 : 0); + if (enable && amdgpu_sdma_phase_quantum) { + WREG32(sdma_v4_0_get_reg_offset(i, mmSDMA0_PHASE0_QUANTUM), + phase_quantum); + WREG32(sdma_v4_0_get_reg_offset(i, mmSDMA0_PHASE1_QUANTUM), + phase_quantum); + WREG32(sdma_v4_0_get_reg_offset(i, mmSDMA0_PHASE2_QUANTUM), + phase_quantum); + } WREG32(sdma_v4_0_get_reg_offset(i, mmSDMA0_CNTL), f32_cntl); } diff --git a/drivers/gpu/drm/amd/amdgpu/si.c b/drivers/gpu/drm/amd/amdgpu/si.c index f45fb0f022b3..812a24dd1204 100644 --- a/drivers/gpu/drm/amd/amdgpu/si.c +++ b/drivers/gpu/drm/amd/amdgpu/si.c @@ -1150,6 +1150,33 @@ static bool si_read_disabled_bios(struct amdgpu_device *adev) return r; } +#define mmROM_INDEX 0x2A +#define mmROM_DATA 0x2B + +static bool si_read_bios_from_rom(struct amdgpu_device *adev, + u8 *bios, u32 length_bytes) +{ + u32 *dw_ptr; + u32 i, length_dw; + + if (bios == NULL) + return false; + if (length_bytes == 0) + return false; + /* APU vbios image is part of sbios image */ + if (adev->flags & AMD_IS_APU) + return false; + + dw_ptr = (u32 *)bios; + length_dw = ALIGN(length_bytes, 4) / 4; + /* set rom index to 0 */ + WREG32(mmROM_INDEX, 0); + for (i = 0; i < length_dw; i++) + dw_ptr[i] = RREG32(mmROM_DATA); + + return true; +} + //xxx: not implemented static int si_asic_reset(struct amdgpu_device *adev) { @@ -1206,6 +1233,7 @@ static void si_detect_hw_virtualization(struct amdgpu_device *adev) static const struct amdgpu_asic_funcs si_asic_funcs = { .read_disabled_bios = &si_read_disabled_bios, + .read_bios_from_rom = &si_read_bios_from_rom, .read_register = &si_read_register, .reset = &si_asic_reset, .set_vga_state = &si_vga_set_state, diff --git a/drivers/gpu/drm/amd/amdgpu/si_dpm.c b/drivers/gpu/drm/amd/amdgpu/si_dpm.c index a7ad8390981c..d63873f3f574 100644 --- a/drivers/gpu/drm/amd/amdgpu/si_dpm.c +++ b/drivers/gpu/drm/amd/amdgpu/si_dpm.c @@ -2055,6 +2055,7 @@ static void si_initialize_powertune_defaults(struct amdgpu_device *adev) case 0x682C: si_pi->cac_weights = cac_weights_cape_verde_pro; si_pi->dte_data = dte_data_sun_xt; + update_dte_from_pl2 = true; break; case 0x6825: case 0x6827: diff --git a/drivers/gpu/drm/amd/amdgpu/soc15.c b/drivers/gpu/drm/amd/amdgpu/soc15.c index a7341d88a320..f2c3a49f73a0 100644 --- a/drivers/gpu/drm/amd/amdgpu/soc15.c +++ b/drivers/gpu/drm/amd/amdgpu/soc15.c @@ -25,7 +25,7 @@ #include <linux/module.h> #include <drm/drmP.h> #include "amdgpu.h" -#include "amdgpu_atomfirmware.h" +#include "amdgpu_atombios.h" #include "amdgpu_ih.h" #include "amdgpu_uvd.h" #include "amdgpu_vce.h" @@ -62,8 +62,6 @@ #include "dce_virtual.h" #include "mxgpu_ai.h" -MODULE_FIRMWARE("amdgpu/vega10_smc.bin"); - #define mmFabricConfigAccessControl 0x0410 #define mmFabricConfigAccessControl_BASE_IDX 0 #define mmFabricConfigAccessControl_DEFAULT 0x00000000 @@ -198,6 +196,50 @@ static void soc15_didt_wreg(struct amdgpu_device *adev, u32 reg, u32 v) spin_unlock_irqrestore(&adev->didt_idx_lock, flags); } +static u32 soc15_gc_cac_rreg(struct amdgpu_device *adev, u32 reg) +{ + unsigned long flags; + u32 r; + + spin_lock_irqsave(&adev->gc_cac_idx_lock, flags); + WREG32_SOC15(GC, 0, mmGC_CAC_IND_INDEX, (reg)); + r = RREG32_SOC15(GC, 0, mmGC_CAC_IND_DATA); + spin_unlock_irqrestore(&adev->gc_cac_idx_lock, flags); + return r; +} + +static void soc15_gc_cac_wreg(struct amdgpu_device *adev, u32 reg, u32 v) +{ + unsigned long flags; + + spin_lock_irqsave(&adev->gc_cac_idx_lock, flags); + WREG32_SOC15(GC, 0, mmGC_CAC_IND_INDEX, (reg)); + WREG32_SOC15(GC, 0, mmGC_CAC_IND_DATA, (v)); + spin_unlock_irqrestore(&adev->gc_cac_idx_lock, flags); +} + +static u32 soc15_se_cac_rreg(struct amdgpu_device *adev, u32 reg) +{ + unsigned long flags; + u32 r; + + spin_lock_irqsave(&adev->se_cac_idx_lock, flags); + WREG32_SOC15(GC, 0, mmSE_CAC_IND_INDEX, (reg)); + r = RREG32_SOC15(GC, 0, mmSE_CAC_IND_DATA); + spin_unlock_irqrestore(&adev->se_cac_idx_lock, flags); + return r; +} + +static void soc15_se_cac_wreg(struct amdgpu_device *adev, u32 reg, u32 v) +{ + unsigned long flags; + + spin_lock_irqsave(&adev->se_cac_idx_lock, flags); + WREG32_SOC15(GC, 0, mmSE_CAC_IND_INDEX, (reg)); + WREG32_SOC15(GC, 0, mmSE_CAC_IND_DATA, (v)); + spin_unlock_irqrestore(&adev->se_cac_idx_lock, flags); +} + static u32 soc15_get_config_memsize(struct amdgpu_device *adev) { if (adev->flags & AMD_IS_APU) @@ -392,11 +434,11 @@ static void soc15_gpu_pci_config_reset(struct amdgpu_device *adev) static int soc15_asic_reset(struct amdgpu_device *adev) { - amdgpu_atomfirmware_scratch_regs_engine_hung(adev, true); + amdgpu_atombios_scratch_regs_engine_hung(adev, true); soc15_gpu_pci_config_reset(adev); - amdgpu_atomfirmware_scratch_regs_engine_hung(adev, false); + amdgpu_atombios_scratch_regs_engine_hung(adev, false); return 0; } @@ -524,13 +566,6 @@ static uint32_t soc15_get_rev_id(struct amdgpu_device *adev) return nbio_v6_1_get_rev_id(adev); } - -int gmc_v9_0_mc_wait_for_idle(struct amdgpu_device *adev) -{ - /* to be implemented in MC IP*/ - return 0; -} - static const struct amdgpu_asic_funcs soc15_asic_funcs = { .read_disabled_bios = &soc15_read_disabled_bios, @@ -557,6 +592,10 @@ static int soc15_common_early_init(void *handle) adev->uvd_ctx_wreg = &soc15_uvd_ctx_wreg; adev->didt_rreg = &soc15_didt_rreg; adev->didt_wreg = &soc15_didt_wreg; + adev->gc_cac_rreg = &soc15_gc_cac_rreg; + adev->gc_cac_wreg = &soc15_gc_cac_wreg; + adev->se_cac_rreg = &soc15_se_cac_rreg; + adev->se_cac_wreg = &soc15_se_cac_wreg; adev->asic_funcs = &soc15_asic_funcs; @@ -681,6 +720,9 @@ static int soc15_common_hw_init(void *handle) soc15_pcie_gen3_enable(adev); /* enable aspm */ soc15_program_aspm(adev); + /* setup nbio registers */ + if (!(adev->flags & AMD_IS_APU)) + nbio_v6_1_init_registers(adev); /* enable the doorbell aperture */ soc15_enable_doorbell_aperture(adev, true); diff --git a/drivers/gpu/drm/amd/amdgpu/soc15_common.h b/drivers/gpu/drm/amd/amdgpu/soc15_common.h index e2d330eed952..7a8e4e28abb2 100644 --- a/drivers/gpu/drm/amd/amdgpu/soc15_common.h +++ b/drivers/gpu/drm/amd/amdgpu/soc15_common.h @@ -77,6 +77,13 @@ struct nbio_pcie_index_data { (3 == reg##_BASE_IDX ? ip##_BASE__INST##inst##_SEG3 + reg : \ (ip##_BASE__INST##inst##_SEG4 + reg))))), value) +#define WREG32_SOC15_NO_KIQ(ip, inst, reg, value) \ + WREG32_NO_KIQ( (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) + #define WREG32_SOC15_OFFSET(ip, inst, reg, offset, value) \ WREG32( (0 == reg##_BASE_IDX ? ip##_BASE__INST##inst##_SEG0 + reg : \ (1 == reg##_BASE_IDX ? ip##_BASE__INST##inst##_SEG1 + reg : \ diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c index 602769ced3bd..42de22bbe14c 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c @@ -664,7 +664,7 @@ static int set_sched_resources(struct device_queue_manager *dqm) /* This situation may be hit in the future if a new HW * generation exposes more than 64 queues. If so, the * definition of res.queue_mask needs updating */ - if (WARN_ON(i > (sizeof(res.queue_mask)*8))) { + if (WARN_ON(i >= (sizeof(res.queue_mask)*8))) { pr_err("Invalid queue enabled by amdgpu: %d\n", i); break; } diff --git a/drivers/gpu/drm/amd/include/atomfirmware.h b/drivers/gpu/drm/amd/include/atomfirmware.h index 0021a1c63356..837296db9628 100644 --- a/drivers/gpu/drm/amd/include/atomfirmware.h +++ b/drivers/gpu/drm/amd/include/atomfirmware.h @@ -1233,6 +1233,69 @@ struct atom_asic_profiling_info_v4_1 uint32_t phyclk2gfxclk_c; }; +struct atom_asic_profiling_info_v4_2 { + struct atom_common_table_header table_header; + uint32_t maxvddc; + uint32_t minvddc; + uint32_t avfs_meannsigma_acontant0; + uint32_t avfs_meannsigma_acontant1; + uint32_t avfs_meannsigma_acontant2; + uint16_t avfs_meannsigma_dc_tol_sigma; + uint16_t avfs_meannsigma_platform_mean; + uint16_t avfs_meannsigma_platform_sigma; + uint32_t gb_vdroop_table_cksoff_a0; + uint32_t gb_vdroop_table_cksoff_a1; + uint32_t gb_vdroop_table_cksoff_a2; + uint32_t gb_vdroop_table_ckson_a0; + uint32_t gb_vdroop_table_ckson_a1; + uint32_t gb_vdroop_table_ckson_a2; + uint32_t avfsgb_fuse_table_cksoff_m1; + uint32_t avfsgb_fuse_table_cksoff_m2; + uint32_t avfsgb_fuse_table_cksoff_b; + uint32_t avfsgb_fuse_table_ckson_m1; + uint32_t avfsgb_fuse_table_ckson_m2; + uint32_t avfsgb_fuse_table_ckson_b; + uint16_t max_voltage_0_25mv; + uint8_t enable_gb_vdroop_table_cksoff; + uint8_t enable_gb_vdroop_table_ckson; + uint8_t enable_gb_fuse_table_cksoff; + uint8_t enable_gb_fuse_table_ckson; + uint16_t psm_age_comfactor; + uint8_t enable_apply_avfs_cksoff_voltage; + uint8_t reserved; + uint32_t dispclk2gfxclk_a; + uint32_t dispclk2gfxclk_b; + uint32_t dispclk2gfxclk_c; + uint32_t pixclk2gfxclk_a; + uint32_t pixclk2gfxclk_b; + uint32_t pixclk2gfxclk_c; + uint32_t dcefclk2gfxclk_a; + uint32_t dcefclk2gfxclk_b; + uint32_t dcefclk2gfxclk_c; + uint32_t phyclk2gfxclk_a; + uint32_t phyclk2gfxclk_b; + uint32_t phyclk2gfxclk_c; + uint32_t acg_gb_vdroop_table_a0; + uint32_t acg_gb_vdroop_table_a1; + uint32_t acg_gb_vdroop_table_a2; + uint32_t acg_avfsgb_fuse_table_m1; + uint32_t acg_avfsgb_fuse_table_m2; + uint32_t acg_avfsgb_fuse_table_b; + uint8_t enable_acg_gb_vdroop_table; + uint8_t enable_acg_gb_fuse_table; + uint32_t acg_dispclk2gfxclk_a; + uint32_t acg_dispclk2gfxclk_b; + uint32_t acg_dispclk2gfxclk_c; + uint32_t acg_pixclk2gfxclk_a; + uint32_t acg_pixclk2gfxclk_b; + uint32_t acg_pixclk2gfxclk_c; + uint32_t acg_dcefclk2gfxclk_a; + uint32_t acg_dcefclk2gfxclk_b; + uint32_t acg_dcefclk2gfxclk_c; + uint32_t acg_phyclk2gfxclk_a; + uint32_t acg_phyclk2gfxclk_b; + uint32_t acg_phyclk2gfxclk_c; +}; /* *************************************************************************** diff --git a/drivers/gpu/drm/amd/include/cgs_common.h b/drivers/gpu/drm/amd/include/cgs_common.h index 0a94f749e3c0..0214f63f52fc 100644 --- a/drivers/gpu/drm/amd/include/cgs_common.h +++ b/drivers/gpu/drm/amd/include/cgs_common.h @@ -50,6 +50,7 @@ enum cgs_ind_reg { CGS_IND_REG__UVD_CTX, CGS_IND_REG__DIDT, CGS_IND_REG_GC_CAC, + CGS_IND_REG_SE_CAC, CGS_IND_REG__AUDIO_ENDPT }; @@ -406,6 +407,8 @@ typedef int (*cgs_is_virtualization_enabled_t)(void *cgs_device); typedef int (*cgs_enter_safe_mode)(struct cgs_device *cgs_device, bool en); +typedef void (*cgs_lock_grbm_idx)(struct cgs_device *cgs_device, bool lock); + struct cgs_ops { /* memory management calls (similar to KFD interface) */ cgs_alloc_gpu_mem_t alloc_gpu_mem; @@ -441,6 +444,7 @@ struct cgs_ops { cgs_query_system_info query_system_info; cgs_is_virtualization_enabled_t is_virtualization_enabled; cgs_enter_safe_mode enter_safe_mode; + cgs_lock_grbm_idx lock_grbm_idx; }; struct cgs_os_ops; /* To be define in OS-specific CGS header */ @@ -517,4 +521,6 @@ struct cgs_device #define cgs_enter_safe_mode(cgs_device, en) \ CGS_CALL(enter_safe_mode, cgs_device, en) +#define cgs_lock_grbm_idx(cgs_device, lock) \ + CGS_CALL(lock_grbm_idx, cgs_device, lock) #endif /* _CGS_COMMON_H */ diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomfwctrl.c b/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomfwctrl.c index 720d5006ff62..cd33eb179db2 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomfwctrl.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomfwctrl.c @@ -276,7 +276,10 @@ int pp_atomfwctrl_get_avfs_information(struct pp_hwmgr *hwmgr, struct pp_atomfwctrl_avfs_parameters *param) { uint16_t idx; + uint8_t format_revision, content_revision; + struct atom_asic_profiling_info_v4_1 *profile; + struct atom_asic_profiling_info_v4_2 *profile_v4_2; idx = GetIndexIntoMasterDataTable(asic_profiling_info); profile = (struct atom_asic_profiling_info_v4_1 *) @@ -286,76 +289,172 @@ int pp_atomfwctrl_get_avfs_information(struct pp_hwmgr *hwmgr, if (!profile) return -1; - param->ulMaxVddc = le32_to_cpu(profile->maxvddc); - param->ulMinVddc = le32_to_cpu(profile->minvddc); - param->ulMeanNsigmaAcontant0 = - le32_to_cpu(profile->avfs_meannsigma_acontant0); - param->ulMeanNsigmaAcontant1 = - le32_to_cpu(profile->avfs_meannsigma_acontant1); - param->ulMeanNsigmaAcontant2 = - le32_to_cpu(profile->avfs_meannsigma_acontant2); - param->usMeanNsigmaDcTolSigma = - le16_to_cpu(profile->avfs_meannsigma_dc_tol_sigma); - param->usMeanNsigmaPlatformMean = - le16_to_cpu(profile->avfs_meannsigma_platform_mean); - param->usMeanNsigmaPlatformSigma = - le16_to_cpu(profile->avfs_meannsigma_platform_sigma); - param->ulGbVdroopTableCksoffA0 = - le32_to_cpu(profile->gb_vdroop_table_cksoff_a0); - param->ulGbVdroopTableCksoffA1 = - le32_to_cpu(profile->gb_vdroop_table_cksoff_a1); - param->ulGbVdroopTableCksoffA2 = - le32_to_cpu(profile->gb_vdroop_table_cksoff_a2); - param->ulGbVdroopTableCksonA0 = - le32_to_cpu(profile->gb_vdroop_table_ckson_a0); - param->ulGbVdroopTableCksonA1 = - le32_to_cpu(profile->gb_vdroop_table_ckson_a1); - param->ulGbVdroopTableCksonA2 = - le32_to_cpu(profile->gb_vdroop_table_ckson_a2); - param->ulGbFuseTableCksoffM1 = - le32_to_cpu(profile->avfsgb_fuse_table_cksoff_m1); - param->ulGbFuseTableCksoffM2 = - le32_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->ulGbFuseTableCksonM2 = - le32_to_cpu(profile->avfsgb_fuse_table_ckson_m2); - param->ulGbFuseTableCksonB = - le32_to_cpu(profile->avfsgb_fuse_table_ckson_b); - - param->ucEnableGbVdroopTableCkson = - profile->enable_gb_vdroop_table_ckson; - param->ucEnableGbFuseTableCkson = - profile->enable_gb_fuse_table_ckson; - param->usPsmAgeComfactor = - le16_to_cpu(profile->psm_age_comfactor); - - param->ulDispclk2GfxclkM1 = - le32_to_cpu(profile->dispclk2gfxclk_a); - param->ulDispclk2GfxclkM2 = - le32_to_cpu(profile->dispclk2gfxclk_b); - param->ulDispclk2GfxclkB = - le32_to_cpu(profile->dispclk2gfxclk_c); - param->ulDcefclk2GfxclkM1 = - le32_to_cpu(profile->dcefclk2gfxclk_a); - param->ulDcefclk2GfxclkM2 = - le32_to_cpu(profile->dcefclk2gfxclk_b); - param->ulDcefclk2GfxclkB = - le32_to_cpu(profile->dcefclk2gfxclk_c); - param->ulPixelclk2GfxclkM1 = - le32_to_cpu(profile->pixclk2gfxclk_a); - param->ulPixelclk2GfxclkM2 = - le32_to_cpu(profile->pixclk2gfxclk_b); - param->ulPixelclk2GfxclkB = - le32_to_cpu(profile->pixclk2gfxclk_c); - param->ulPhyclk2GfxclkM1 = - le32_to_cpu(profile->phyclk2gfxclk_a); - param->ulPhyclk2GfxclkM2 = - le32_to_cpu(profile->phyclk2gfxclk_b); - param->ulPhyclk2GfxclkB = - le32_to_cpu(profile->phyclk2gfxclk_c); + format_revision = ((struct atom_common_table_header *)profile)->format_revision; + content_revision = ((struct atom_common_table_header *)profile)->content_revision; + + if (format_revision == 4 && content_revision == 1) { + param->ulMaxVddc = le32_to_cpu(profile->maxvddc); + param->ulMinVddc = le32_to_cpu(profile->minvddc); + param->ulMeanNsigmaAcontant0 = + le32_to_cpu(profile->avfs_meannsigma_acontant0); + param->ulMeanNsigmaAcontant1 = + le32_to_cpu(profile->avfs_meannsigma_acontant1); + param->ulMeanNsigmaAcontant2 = + le32_to_cpu(profile->avfs_meannsigma_acontant2); + param->usMeanNsigmaDcTolSigma = + le16_to_cpu(profile->avfs_meannsigma_dc_tol_sigma); + param->usMeanNsigmaPlatformMean = + le16_to_cpu(profile->avfs_meannsigma_platform_mean); + param->usMeanNsigmaPlatformSigma = + le16_to_cpu(profile->avfs_meannsigma_platform_sigma); + param->ulGbVdroopTableCksoffA0 = + le32_to_cpu(profile->gb_vdroop_table_cksoff_a0); + param->ulGbVdroopTableCksoffA1 = + le32_to_cpu(profile->gb_vdroop_table_cksoff_a1); + param->ulGbVdroopTableCksoffA2 = + le32_to_cpu(profile->gb_vdroop_table_cksoff_a2); + param->ulGbVdroopTableCksonA0 = + le32_to_cpu(profile->gb_vdroop_table_ckson_a0); + param->ulGbVdroopTableCksonA1 = + le32_to_cpu(profile->gb_vdroop_table_ckson_a1); + param->ulGbVdroopTableCksonA2 = + le32_to_cpu(profile->gb_vdroop_table_ckson_a2); + param->ulGbFuseTableCksoffM1 = + le32_to_cpu(profile->avfsgb_fuse_table_cksoff_m1); + param->ulGbFuseTableCksoffM2 = + le32_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->ulGbFuseTableCksonM2 = + le32_to_cpu(profile->avfsgb_fuse_table_ckson_m2); + param->ulGbFuseTableCksonB = + le32_to_cpu(profile->avfsgb_fuse_table_ckson_b); + + param->ucEnableGbVdroopTableCkson = + profile->enable_gb_vdroop_table_ckson; + param->ucEnableGbFuseTableCkson = + profile->enable_gb_fuse_table_ckson; + param->usPsmAgeComfactor = + le16_to_cpu(profile->psm_age_comfactor); + + param->ulDispclk2GfxclkM1 = + le32_to_cpu(profile->dispclk2gfxclk_a); + param->ulDispclk2GfxclkM2 = + le32_to_cpu(profile->dispclk2gfxclk_b); + param->ulDispclk2GfxclkB = + le32_to_cpu(profile->dispclk2gfxclk_c); + param->ulDcefclk2GfxclkM1 = + le32_to_cpu(profile->dcefclk2gfxclk_a); + param->ulDcefclk2GfxclkM2 = + le32_to_cpu(profile->dcefclk2gfxclk_b); + param->ulDcefclk2GfxclkB = + le32_to_cpu(profile->dcefclk2gfxclk_c); + param->ulPixelclk2GfxclkM1 = + le32_to_cpu(profile->pixclk2gfxclk_a); + param->ulPixelclk2GfxclkM2 = + le32_to_cpu(profile->pixclk2gfxclk_b); + param->ulPixelclk2GfxclkB = + le32_to_cpu(profile->pixclk2gfxclk_c); + param->ulPhyclk2GfxclkM1 = + le32_to_cpu(profile->phyclk2gfxclk_a); + param->ulPhyclk2GfxclkM2 = + le32_to_cpu(profile->phyclk2gfxclk_b); + param->ulPhyclk2GfxclkB = + le32_to_cpu(profile->phyclk2gfxclk_c); + param->ulAcgGbVdroopTableA0 = 0; + param->ulAcgGbVdroopTableA1 = 0; + param->ulAcgGbVdroopTableA2 = 0; + param->ulAcgGbFuseTableM1 = 0; + param->ulAcgGbFuseTableM2 = 0; + param->ulAcgGbFuseTableB = 0; + param->ucAcgEnableGbVdroopTable = 0; + param->ucAcgEnableGbFuseTable = 0; + } else if (format_revision == 4 && content_revision == 2) { + profile_v4_2 = (struct atom_asic_profiling_info_v4_2 *)profile; + param->ulMaxVddc = le32_to_cpu(profile_v4_2->maxvddc); + param->ulMinVddc = le32_to_cpu(profile_v4_2->minvddc); + param->ulMeanNsigmaAcontant0 = + le32_to_cpu(profile_v4_2->avfs_meannsigma_acontant0); + param->ulMeanNsigmaAcontant1 = + le32_to_cpu(profile_v4_2->avfs_meannsigma_acontant1); + param->ulMeanNsigmaAcontant2 = + le32_to_cpu(profile_v4_2->avfs_meannsigma_acontant2); + param->usMeanNsigmaDcTolSigma = + le16_to_cpu(profile_v4_2->avfs_meannsigma_dc_tol_sigma); + param->usMeanNsigmaPlatformMean = + le16_to_cpu(profile_v4_2->avfs_meannsigma_platform_mean); + param->usMeanNsigmaPlatformSigma = + le16_to_cpu(profile_v4_2->avfs_meannsigma_platform_sigma); + param->ulGbVdroopTableCksoffA0 = + le32_to_cpu(profile_v4_2->gb_vdroop_table_cksoff_a0); + param->ulGbVdroopTableCksoffA1 = + le32_to_cpu(profile_v4_2->gb_vdroop_table_cksoff_a1); + param->ulGbVdroopTableCksoffA2 = + le32_to_cpu(profile_v4_2->gb_vdroop_table_cksoff_a2); + param->ulGbVdroopTableCksonA0 = + le32_to_cpu(profile_v4_2->gb_vdroop_table_ckson_a0); + param->ulGbVdroopTableCksonA1 = + le32_to_cpu(profile_v4_2->gb_vdroop_table_ckson_a1); + param->ulGbVdroopTableCksonA2 = + le32_to_cpu(profile_v4_2->gb_vdroop_table_ckson_a2); + param->ulGbFuseTableCksoffM1 = + le32_to_cpu(profile_v4_2->avfsgb_fuse_table_cksoff_m1); + param->ulGbFuseTableCksoffM2 = + le32_to_cpu(profile_v4_2->avfsgb_fuse_table_cksoff_m2); + param->ulGbFuseTableCksoffB = + le32_to_cpu(profile_v4_2->avfsgb_fuse_table_cksoff_b); + param->ulGbFuseTableCksonM1 = + le32_to_cpu(profile_v4_2->avfsgb_fuse_table_ckson_m1); + param->ulGbFuseTableCksonM2 = + le32_to_cpu(profile_v4_2->avfsgb_fuse_table_ckson_m2); + param->ulGbFuseTableCksonB = + le32_to_cpu(profile_v4_2->avfsgb_fuse_table_ckson_b); + + param->ucEnableGbVdroopTableCkson = + profile_v4_2->enable_gb_vdroop_table_ckson; + param->ucEnableGbFuseTableCkson = + profile_v4_2->enable_gb_fuse_table_ckson; + param->usPsmAgeComfactor = + le16_to_cpu(profile_v4_2->psm_age_comfactor); + + param->ulDispclk2GfxclkM1 = + le32_to_cpu(profile_v4_2->dispclk2gfxclk_a); + param->ulDispclk2GfxclkM2 = + le32_to_cpu(profile_v4_2->dispclk2gfxclk_b); + param->ulDispclk2GfxclkB = + le32_to_cpu(profile_v4_2->dispclk2gfxclk_c); + param->ulDcefclk2GfxclkM1 = + le32_to_cpu(profile_v4_2->dcefclk2gfxclk_a); + param->ulDcefclk2GfxclkM2 = + le32_to_cpu(profile_v4_2->dcefclk2gfxclk_b); + param->ulDcefclk2GfxclkB = + le32_to_cpu(profile_v4_2->dcefclk2gfxclk_c); + param->ulPixelclk2GfxclkM1 = + le32_to_cpu(profile_v4_2->pixclk2gfxclk_a); + param->ulPixelclk2GfxclkM2 = + le32_to_cpu(profile_v4_2->pixclk2gfxclk_b); + param->ulPixelclk2GfxclkB = + le32_to_cpu(profile_v4_2->pixclk2gfxclk_c); + param->ulPhyclk2GfxclkM1 = + le32_to_cpu(profile->phyclk2gfxclk_a); + param->ulPhyclk2GfxclkM2 = + le32_to_cpu(profile_v4_2->phyclk2gfxclk_b); + param->ulPhyclk2GfxclkB = + le32_to_cpu(profile_v4_2->phyclk2gfxclk_c); + param->ulAcgGbVdroopTableA0 = le32_to_cpu(profile_v4_2->acg_gb_vdroop_table_a0); + param->ulAcgGbVdroopTableA1 = le32_to_cpu(profile_v4_2->acg_gb_vdroop_table_a1); + param->ulAcgGbVdroopTableA2 = le32_to_cpu(profile_v4_2->acg_gb_vdroop_table_a2); + param->ulAcgGbFuseTableM1 = le32_to_cpu(profile_v4_2->acg_avfsgb_fuse_table_m1); + param->ulAcgGbFuseTableM2 = le32_to_cpu(profile_v4_2->acg_avfsgb_fuse_table_m2); + param->ulAcgGbFuseTableB = le32_to_cpu(profile_v4_2->acg_avfsgb_fuse_table_b); + param->ucAcgEnableGbVdroopTable = le32_to_cpu(profile_v4_2->enable_acg_gb_vdroop_table); + param->ucAcgEnableGbFuseTable = le32_to_cpu(profile_v4_2->enable_acg_gb_fuse_table); + } else { + pr_info("Invalid VBIOS AVFS ProfilingInfo Revision!\n"); + return -EINVAL; + } return 0; } diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomfwctrl.h b/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomfwctrl.h index 81908b5cfd5f..8e6b1f0ddebc 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomfwctrl.h +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomfwctrl.h @@ -109,6 +109,14 @@ struct pp_atomfwctrl_avfs_parameters { uint32_t ulPhyclk2GfxclkM1; uint32_t ulPhyclk2GfxclkM2; uint32_t ulPhyclk2GfxclkB; + uint32_t ulAcgGbVdroopTableA0; + uint32_t ulAcgGbVdroopTableA1; + uint32_t ulAcgGbVdroopTableA2; + uint32_t ulAcgGbFuseTableM1; + uint32_t ulAcgGbFuseTableM2; + uint32_t ulAcgGbFuseTableB; + uint32_t ucAcgEnableGbVdroopTable; + uint32_t ucAcgEnableGbFuseTable; }; struct pp_atomfwctrl_gpio_parameters { diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c index 1f01020ce3a9..f01cda93f178 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c @@ -4630,6 +4630,15 @@ static int smu7_set_power_profile_state(struct pp_hwmgr *hwmgr, static int smu7_avfs_control(struct pp_hwmgr *hwmgr, bool enable) { + struct pp_smumgr *smumgr = (struct pp_smumgr *)(hwmgr->smumgr); + struct smu7_smumgr *smu_data = (struct smu7_smumgr *)(smumgr->backend); + + if (smu_data == NULL) + return -EINVAL; + + if (smu_data->avfs.avfs_btc_status == AVFS_BTC_NOTSUPPORTED) + return 0; + if (enable) { if (!PHM_READ_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, FEATURE_STATUS, AVS_ON)) diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c index d6f097f44b6c..01ff5054041b 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c @@ -78,6 +78,8 @@ uint32_t channel_number[] = {1, 2, 0, 4, 0, 8, 0, 16, 2}; #define DF_CS_AON0_DramBaseAddress0__IntLvNumChan_MASK 0x000000F0L #define DF_CS_AON0_DramBaseAddress0__IntLvAddrSel_MASK 0x00000700L #define DF_CS_AON0_DramBaseAddress0__DramBaseAddr_MASK 0xFFFFF000L +static int vega10_force_clock_level(struct pp_hwmgr *hwmgr, + enum pp_clock_type type, uint32_t mask); const ULONG PhwVega10_Magic = (ULONG)(PHM_VIslands_Magic); @@ -146,6 +148,19 @@ static void vega10_set_default_registry_data(struct pp_hwmgr *hwmgr) data->registry_data.vr1hot_enabled = 1; data->registry_data.regulator_hot_gpio_support = 1; + data->registry_data.didt_support = 1; + if (data->registry_data.didt_support) { + data->registry_data.didt_mode = 6; + data->registry_data.sq_ramping_support = 1; + data->registry_data.db_ramping_support = 0; + data->registry_data.td_ramping_support = 0; + data->registry_data.tcp_ramping_support = 0; + data->registry_data.dbr_ramping_support = 0; + data->registry_data.edc_didt_support = 1; + data->registry_data.gc_didt_support = 0; + data->registry_data.psm_didt_support = 0; + } + data->display_voltage_mode = PPVEGA10_VEGA10DISPLAYVOLTAGEMODE_DFLT; data->dcef_clk_quad_eqn_a = PPREGKEY_VEGA10QUADRATICEQUATION_DFLT; data->dcef_clk_quad_eqn_b = PPREGKEY_VEGA10QUADRATICEQUATION_DFLT; @@ -223,6 +238,8 @@ static int vega10_set_features_platform_caps(struct pp_hwmgr *hwmgr) phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_PowerContainment); phm_cap_unset(hwmgr->platform_descriptor.platformCaps, + PHM_PlatformCaps_DiDtSupport); + phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_SQRamping); phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_DBRamping); @@ -230,6 +247,34 @@ static int vega10_set_features_platform_caps(struct pp_hwmgr *hwmgr) PHM_PlatformCaps_TDRamping); phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_TCPRamping); + phm_cap_unset(hwmgr->platform_descriptor.platformCaps, + PHM_PlatformCaps_DBRRamping); + phm_cap_unset(hwmgr->platform_descriptor.platformCaps, + PHM_PlatformCaps_DiDtEDCEnable); + phm_cap_unset(hwmgr->platform_descriptor.platformCaps, + PHM_PlatformCaps_GCEDC); + phm_cap_unset(hwmgr->platform_descriptor.platformCaps, + PHM_PlatformCaps_PSM); + + if (data->registry_data.didt_support) { + phm_cap_set(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_DiDtSupport); + if (data->registry_data.sq_ramping_support) + phm_cap_set(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_SQRamping); + if (data->registry_data.db_ramping_support) + phm_cap_set(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_DBRamping); + if (data->registry_data.td_ramping_support) + phm_cap_set(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_TDRamping); + if (data->registry_data.tcp_ramping_support) + phm_cap_set(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_TCPRamping); + if (data->registry_data.dbr_ramping_support) + phm_cap_set(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_DBRRamping); + if (data->registry_data.edc_didt_support) + phm_cap_set(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_DiDtEDCEnable); + if (data->registry_data.gc_didt_support) + phm_cap_set(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_GCEDC); + if (data->registry_data.psm_didt_support) + phm_cap_set(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_PSM); + } if (data->registry_data.power_containment_support) phm_cap_set(hwmgr->platform_descriptor.platformCaps, @@ -321,8 +366,8 @@ static void vega10_init_dpm_defaults(struct pp_hwmgr *hwmgr) FEATURE_LED_DISPLAY_BIT; data->smu_features[GNLD_FAN_CONTROL].smu_feature_id = FEATURE_FAN_CONTROL_BIT; - data->smu_features[GNLD_VOLTAGE_CONTROLLER].smu_feature_id = - FEATURE_VOLTAGE_CONTROLLER_BIT; + data->smu_features[GNLD_ACG].smu_feature_id = FEATURE_ACG_BIT; + data->smu_features[GNLD_DIDT].smu_feature_id = FEATURE_GFX_EDC_BIT; if (!data->registry_data.prefetcher_dpm_key_disabled) data->smu_features[GNLD_DPM_PREFETCHER].supported = true; @@ -386,6 +431,15 @@ static void vega10_init_dpm_defaults(struct pp_hwmgr *hwmgr) if (data->registry_data.vr0hot_enabled) data->smu_features[GNLD_VR0HOT].supported = true; + smum_send_msg_to_smc(hwmgr->smumgr, PPSMC_MSG_GetSmuVersion); + vega10_read_arg_from_smc(hwmgr->smumgr, &(data->smu_version)); + /* ACG firmware has major version 5 */ + if ((data->smu_version & 0xff000000) == 0x5000000) + data->smu_features[GNLD_ACG].supported = true; + + if (data->registry_data.didt_support) + data->smu_features[GNLD_DIDT].supported = true; + } #ifdef PPLIB_VEGA10_EVV_SUPPORT @@ -2128,15 +2182,9 @@ static int vega10_populate_avfs_parameters(struct pp_hwmgr *hwmgr) pp_table->AvfsGbCksOff.m2_shift = 12; 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); - } + for (i = 0; i < dep_table->count; i++) + pp_table->StaticVoltageOffsetVid[i] = + convert_to_vid((uint8_t)(dep_table->entries[i].sclk_offset)); if ((PPREGKEY_VEGA10QUADRATICEQUATION_DFLT != data->disp_clk_quad_eqn_a) && @@ -2228,6 +2276,21 @@ static int vega10_populate_avfs_parameters(struct pp_hwmgr *hwmgr) pp_table->DisplayClock2Gfxclk[DSPCLK_PHYCLK].m1_shift = 24; pp_table->DisplayClock2Gfxclk[DSPCLK_PHYCLK].m2_shift = 12; pp_table->DisplayClock2Gfxclk[DSPCLK_PHYCLK].b_shift = 12; + + pp_table->AcgBtcGbVdroopTable.a0 = avfs_params.ulAcgGbVdroopTableA0; + pp_table->AcgBtcGbVdroopTable.a0_shift = 20; + pp_table->AcgBtcGbVdroopTable.a1 = avfs_params.ulAcgGbVdroopTableA1; + pp_table->AcgBtcGbVdroopTable.a1_shift = 20; + pp_table->AcgBtcGbVdroopTable.a2 = avfs_params.ulAcgGbVdroopTableA2; + pp_table->AcgBtcGbVdroopTable.a2_shift = 20; + + pp_table->AcgAvfsGb.m1 = avfs_params.ulAcgGbFuseTableM1; + pp_table->AcgAvfsGb.m2 = avfs_params.ulAcgGbFuseTableM2; + pp_table->AcgAvfsGb.b = avfs_params.ulAcgGbFuseTableB; + pp_table->AcgAvfsGb.m1_shift = 0; + pp_table->AcgAvfsGb.m2_shift = 0; + pp_table->AcgAvfsGb.b_shift = 0; + } else { data->smu_features[GNLD_AVFS].supported = false; } @@ -2236,6 +2299,55 @@ static int vega10_populate_avfs_parameters(struct pp_hwmgr *hwmgr) return 0; } +static int vega10_acg_enable(struct pp_hwmgr *hwmgr) +{ + struct vega10_hwmgr *data = + (struct vega10_hwmgr *)(hwmgr->backend); + uint32_t agc_btc_response; + + if (data->smu_features[GNLD_ACG].supported) { + if (0 == vega10_enable_smc_features(hwmgr->smumgr, true, + data->smu_features[GNLD_DPM_PREFETCHER].smu_feature_bitmap)) + data->smu_features[GNLD_DPM_PREFETCHER].enabled = true; + + smum_send_msg_to_smc(hwmgr->smumgr, PPSMC_MSG_InitializeAcg); + + smum_send_msg_to_smc(hwmgr->smumgr, PPSMC_MSG_RunAcgBtc); + vega10_read_arg_from_smc(hwmgr->smumgr, &agc_btc_response);; + + if (1 == agc_btc_response) { + if (1 == data->acg_loop_state) + smum_send_msg_to_smc(hwmgr->smumgr, PPSMC_MSG_RunAcgInClosedLoop); + else if (2 == data->acg_loop_state) + smum_send_msg_to_smc(hwmgr->smumgr, PPSMC_MSG_RunAcgInOpenLoop); + if (0 == vega10_enable_smc_features(hwmgr->smumgr, true, + data->smu_features[GNLD_ACG].smu_feature_bitmap)) + data->smu_features[GNLD_ACG].enabled = true; + } else { + pr_info("[ACG_Enable] ACG BTC Returned Failed Status!\n"); + data->smu_features[GNLD_ACG].enabled = false; + } + } + + return 0; +} + +static int vega10_acg_disable(struct pp_hwmgr *hwmgr) +{ + struct vega10_hwmgr *data = + (struct vega10_hwmgr *)(hwmgr->backend); + + if (data->smu_features[GNLD_ACG].supported) { + if (data->smu_features[GNLD_ACG].enabled) { + if (0 == vega10_enable_smc_features(hwmgr->smumgr, false, + data->smu_features[GNLD_ACG].smu_feature_bitmap)) + data->smu_features[GNLD_ACG].enabled = false; + } + } + + return 0; +} + static int vega10_populate_gpio_parameters(struct pp_hwmgr *hwmgr) { struct vega10_hwmgr *data = @@ -2506,7 +2618,7 @@ static int vega10_init_smc_table(struct pp_hwmgr *hwmgr) result = vega10_avfs_enable(hwmgr, true); PP_ASSERT_WITH_CODE(!result, "Attempt to enable AVFS feature Failed!", return result); - + vega10_acg_enable(hwmgr); vega10_save_default_power_profile(hwmgr); return 0; @@ -2838,6 +2950,11 @@ static int vega10_enable_dpm_tasks(struct pp_hwmgr *hwmgr) PP_ASSERT_WITH_CODE(!tmp_result, "Failed to start DPM!", result = tmp_result); + /* enable didt, do not abort if failed didt */ + tmp_result = vega10_enable_didt_config(hwmgr); + PP_ASSERT(!tmp_result, + "Failed to enable didt config!"); + tmp_result = vega10_enable_power_containment(hwmgr); PP_ASSERT_WITH_CODE(!tmp_result, "Failed to enable power containment!", @@ -4103,34 +4220,30 @@ static int vega10_unforce_dpm_levels(struct pp_hwmgr *hwmgr) return 0; } -static int vega10_dpm_force_dpm_level(struct pp_hwmgr *hwmgr, - enum amd_dpm_forced_level level) +static int vega10_get_profiling_clk_mask(struct pp_hwmgr *hwmgr, enum amd_dpm_forced_level level, + uint32_t *sclk_mask, uint32_t *mclk_mask, uint32_t *soc_mask) { - int ret = 0; + struct phm_ppt_v2_information *table_info = + (struct phm_ppt_v2_information *)(hwmgr->pptable); - switch (level) { - case AMD_DPM_FORCED_LEVEL_HIGH: - ret = vega10_force_dpm_highest(hwmgr); - if (ret) - return ret; - break; - case AMD_DPM_FORCED_LEVEL_LOW: - ret = vega10_force_dpm_lowest(hwmgr); - if (ret) - return ret; - break; - case AMD_DPM_FORCED_LEVEL_AUTO: - ret = vega10_unforce_dpm_levels(hwmgr); - if (ret) - return ret; - break; - default: - break; + if (table_info->vdd_dep_on_sclk->count > VEGA10_UMD_PSTATE_GFXCLK_LEVEL && + table_info->vdd_dep_on_socclk->count > VEGA10_UMD_PSTATE_SOCCLK_LEVEL && + table_info->vdd_dep_on_mclk->count > VEGA10_UMD_PSTATE_MCLK_LEVEL) { + *sclk_mask = VEGA10_UMD_PSTATE_GFXCLK_LEVEL; + *soc_mask = VEGA10_UMD_PSTATE_SOCCLK_LEVEL; + *mclk_mask = VEGA10_UMD_PSTATE_MCLK_LEVEL; } - hwmgr->dpm_level = level; - - return ret; + if (level == AMD_DPM_FORCED_LEVEL_PROFILE_MIN_SCLK) { + *sclk_mask = 0; + } else if (level == AMD_DPM_FORCED_LEVEL_PROFILE_MIN_MCLK) { + *mclk_mask = 0; + } else if (level == AMD_DPM_FORCED_LEVEL_PROFILE_PEAK) { + *sclk_mask = table_info->vdd_dep_on_sclk->count - 1; + *soc_mask = table_info->vdd_dep_on_socclk->count - 1; + *mclk_mask = table_info->vdd_dep_on_mclk->count - 1; + } + return 0; } static int vega10_set_fan_control_mode(struct pp_hwmgr *hwmgr, uint32_t mode) @@ -4157,6 +4270,86 @@ static int vega10_set_fan_control_mode(struct pp_hwmgr *hwmgr, uint32_t mode) return result; } +static int vega10_dpm_force_dpm_level(struct pp_hwmgr *hwmgr, + enum amd_dpm_forced_level level) +{ + int ret = 0; + uint32_t sclk_mask = 0; + uint32_t mclk_mask = 0; + uint32_t soc_mask = 0; + uint32_t profile_mode_mask = AMD_DPM_FORCED_LEVEL_PROFILE_STANDARD | + AMD_DPM_FORCED_LEVEL_PROFILE_MIN_SCLK | + AMD_DPM_FORCED_LEVEL_PROFILE_MIN_MCLK | + AMD_DPM_FORCED_LEVEL_PROFILE_PEAK; + + if (level == hwmgr->dpm_level) + return ret; + + if (!(hwmgr->dpm_level & profile_mode_mask)) { + /* enter profile mode, save current level, disable gfx cg*/ + if (level & profile_mode_mask) { + hwmgr->saved_dpm_level = hwmgr->dpm_level; + cgs_set_clockgating_state(hwmgr->device, + AMD_IP_BLOCK_TYPE_GFX, + AMD_CG_STATE_UNGATE); + } + } else { + /* exit profile mode, restore level, enable gfx cg*/ + if (!(level & profile_mode_mask)) { + if (level == AMD_DPM_FORCED_LEVEL_PROFILE_EXIT) + level = hwmgr->saved_dpm_level; + cgs_set_clockgating_state(hwmgr->device, + AMD_IP_BLOCK_TYPE_GFX, + AMD_CG_STATE_GATE); + } + } + + switch (level) { + case AMD_DPM_FORCED_LEVEL_HIGH: + ret = vega10_force_dpm_highest(hwmgr); + if (ret) + return ret; + hwmgr->dpm_level = level; + break; + case AMD_DPM_FORCED_LEVEL_LOW: + ret = vega10_force_dpm_lowest(hwmgr); + if (ret) + return ret; + hwmgr->dpm_level = level; + break; + case AMD_DPM_FORCED_LEVEL_AUTO: + ret = vega10_unforce_dpm_levels(hwmgr); + if (ret) + return ret; + hwmgr->dpm_level = level; + break; + case AMD_DPM_FORCED_LEVEL_PROFILE_STANDARD: + case AMD_DPM_FORCED_LEVEL_PROFILE_MIN_SCLK: + case AMD_DPM_FORCED_LEVEL_PROFILE_MIN_MCLK: + case AMD_DPM_FORCED_LEVEL_PROFILE_PEAK: + ret = vega10_get_profiling_clk_mask(hwmgr, level, &sclk_mask, &mclk_mask, &soc_mask); + if (ret) + return ret; + hwmgr->dpm_level = level; + vega10_force_clock_level(hwmgr, PP_SCLK, 1<<sclk_mask); + vega10_force_clock_level(hwmgr, PP_MCLK, 1<<mclk_mask); + break; + case AMD_DPM_FORCED_LEVEL_MANUAL: + hwmgr->dpm_level = level; + break; + case AMD_DPM_FORCED_LEVEL_PROFILE_EXIT: + default: + break; + } + + if (level == AMD_DPM_FORCED_LEVEL_PROFILE_PEAK && hwmgr->saved_dpm_level != AMD_DPM_FORCED_LEVEL_PROFILE_PEAK) + vega10_set_fan_control_mode(hwmgr, AMD_FAN_CTRL_NONE); + else if (level != AMD_DPM_FORCED_LEVEL_PROFILE_PEAK && hwmgr->saved_dpm_level == AMD_DPM_FORCED_LEVEL_PROFILE_PEAK) + vega10_set_fan_control_mode(hwmgr, AMD_FAN_CTRL_AUTO); + + return 0; +} + static int vega10_get_fan_control_mode(struct pp_hwmgr *hwmgr) { struct vega10_hwmgr *data = (struct vega10_hwmgr *)(hwmgr->backend); @@ -4402,7 +4595,9 @@ static int vega10_force_clock_level(struct pp_hwmgr *hwmgr, struct vega10_hwmgr *data = (struct vega10_hwmgr *)(hwmgr->backend); int i; - if (hwmgr->dpm_level != AMD_DPM_FORCED_LEVEL_MANUAL) + if (hwmgr->dpm_level & (AMD_DPM_FORCED_LEVEL_AUTO | + AMD_DPM_FORCED_LEVEL_LOW | + AMD_DPM_FORCED_LEVEL_HIGH)) return -EINVAL; switch (type) { @@ -4667,6 +4862,10 @@ static int vega10_disable_dpm_tasks(struct pp_hwmgr *hwmgr) PP_ASSERT_WITH_CODE((tmp_result == 0), "Failed to disable power containment!", result = tmp_result); + tmp_result = vega10_disable_didt_config(hwmgr); + PP_ASSERT_WITH_CODE((tmp_result == 0), + "Failed to disable didt config!", result = tmp_result); + tmp_result = vega10_avfs_enable(hwmgr, false); PP_ASSERT_WITH_CODE((tmp_result == 0), "Failed to disable AVFS!", result = tmp_result); @@ -4683,6 +4882,9 @@ static int vega10_disable_dpm_tasks(struct pp_hwmgr *hwmgr) PP_ASSERT_WITH_CODE((tmp_result == 0), "Failed to disable ulv!", result = tmp_result); + tmp_result = vega10_acg_disable(hwmgr); + PP_ASSERT_WITH_CODE((tmp_result == 0), + "Failed to disable acg!", result = tmp_result); return result; } diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.h b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.h index 6e5c5b99593b..676cd7735883 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.h +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.h @@ -64,7 +64,9 @@ enum { GNLD_FW_CTF, GNLD_LED_DISPLAY, GNLD_FAN_CONTROL, - GNLD_VOLTAGE_CONTROLLER, + GNLD_FEATURE_FAST_PPT_BIT, + GNLD_DIDT, + GNLD_ACG, GNLD_FEATURES_MAX }; @@ -230,7 +232,9 @@ struct vega10_registry_data { uint8_t cac_support; uint8_t clock_stretcher_support; uint8_t db_ramping_support; + uint8_t didt_mode; uint8_t didt_support; + uint8_t edc_didt_support; uint8_t dynamic_state_patching_support; uint8_t enable_pkg_pwr_tracking_feature; uint8_t enable_tdc_limit_feature; @@ -263,6 +267,9 @@ struct vega10_registry_data { uint8_t tcp_ramping_support; uint8_t tdc_support; uint8_t td_ramping_support; + uint8_t dbr_ramping_support; + uint8_t gc_didt_support; + uint8_t psm_didt_support; uint8_t thermal_out_gpio_support; uint8_t thermal_support; uint8_t fw_ctf_enabled; @@ -381,6 +388,8 @@ struct vega10_hwmgr { struct vega10_smc_state_table smc_state_table; uint32_t config_telemetry; + uint32_t smu_version; + uint32_t acg_loop_state; }; #define VEGA10_DPM2_NEAR_TDP_DEC 10 @@ -425,6 +434,10 @@ struct vega10_hwmgr { #define PPVEGA10_VEGA10UCLKCLKAVERAGEALPHA_DFLT 25 /* 10% * 255 = 25 */ #define PPVEGA10_VEGA10GFXACTIVITYAVERAGEALPHA_DFLT 25 /* 10% * 255 = 25 */ +#define VEGA10_UMD_PSTATE_GFXCLK_LEVEL 0x3 +#define VEGA10_UMD_PSTATE_SOCCLK_LEVEL 0x3 +#define VEGA10_UMD_PSTATE_MCLK_LEVEL 0x2 + extern int tonga_initializa_dynamic_state_adjustment_rule_settings(struct pp_hwmgr *hwmgr); extern int tonga_hwmgr_backend_fini(struct pp_hwmgr *hwmgr); extern int tonga_get_mc_microcode_version (struct pp_hwmgr *hwmgr); diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_powertune.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_powertune.c index 3f72268e99bb..fbafc849ea71 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_powertune.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_powertune.c @@ -26,7 +26,1298 @@ #include "vega10_powertune.h" #include "vega10_smumgr.h" #include "vega10_ppsmc.h" +#include "vega10_inc.h" #include "pp_debug.h" +#include "pp_soc15.h" + +static const struct vega10_didt_config_reg SEDiDtTuningCtrlConfig_Vega10[] = +{ +/* --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- + * Offset Mask Shift Value + * --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- + */ + /* DIDT_SQ */ + { ixDIDT_SQ_TUNING_CTRL, DIDT_SQ_TUNING_CTRL__MAX_POWER_DELTA_HI_MASK, DIDT_SQ_TUNING_CTRL__MAX_POWER_DELTA_HI__SHIFT, 0x3853 }, + { ixDIDT_SQ_TUNING_CTRL, DIDT_SQ_TUNING_CTRL__MAX_POWER_DELTA_LO_MASK, DIDT_SQ_TUNING_CTRL__MAX_POWER_DELTA_LO__SHIFT, 0x3153 }, + + /* DIDT_TD */ + { ixDIDT_TD_TUNING_CTRL, DIDT_TD_TUNING_CTRL__MAX_POWER_DELTA_HI_MASK, DIDT_TD_TUNING_CTRL__MAX_POWER_DELTA_HI__SHIFT, 0x0dde }, + { ixDIDT_TD_TUNING_CTRL, DIDT_TD_TUNING_CTRL__MAX_POWER_DELTA_LO_MASK, DIDT_TD_TUNING_CTRL__MAX_POWER_DELTA_LO__SHIFT, 0x0dde }, + + /* DIDT_TCP */ + { ixDIDT_TCP_TUNING_CTRL, DIDT_TCP_TUNING_CTRL__MAX_POWER_DELTA_HI_MASK, DIDT_TCP_TUNING_CTRL__MAX_POWER_DELTA_HI__SHIFT, 0x3dde }, + { ixDIDT_TCP_TUNING_CTRL, DIDT_TCP_TUNING_CTRL__MAX_POWER_DELTA_LO_MASK, DIDT_TCP_TUNING_CTRL__MAX_POWER_DELTA_LO__SHIFT, 0x3dde }, + + /* DIDT_DB */ + { ixDIDT_DB_TUNING_CTRL, DIDT_DB_TUNING_CTRL__MAX_POWER_DELTA_HI_MASK, DIDT_DB_TUNING_CTRL__MAX_POWER_DELTA_HI__SHIFT, 0x3dde }, + { ixDIDT_DB_TUNING_CTRL, DIDT_DB_TUNING_CTRL__MAX_POWER_DELTA_LO_MASK, DIDT_DB_TUNING_CTRL__MAX_POWER_DELTA_LO__SHIFT, 0x3dde }, + + { 0xFFFFFFFF } /* End of list */ +}; + +static const struct vega10_didt_config_reg SEDiDtCtrl3Config_vega10[] = +{ +/* --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- + * Offset Mask Shift Value + * --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- + */ + /*DIDT_SQ_CTRL3 */ + { ixDIDT_SQ_CTRL3, DIDT_SQ_CTRL3__GC_DIDT_ENABLE_MASK, DIDT_SQ_CTRL3__GC_DIDT_ENABLE__SHIFT, 0x0000 }, + { ixDIDT_SQ_CTRL3, DIDT_SQ_CTRL3__GC_DIDT_CLK_EN_OVERRIDE_MASK, DIDT_SQ_CTRL3__GC_DIDT_CLK_EN_OVERRIDE__SHIFT, 0x0000 }, + { ixDIDT_SQ_CTRL3, DIDT_SQ_CTRL3__THROTTLE_POLICY_MASK, DIDT_SQ_CTRL3__THROTTLE_POLICY__SHIFT, 0x0003 }, + { ixDIDT_SQ_CTRL3, DIDT_SQ_CTRL3__DIDT_TRIGGER_THROTTLE_LOWBIT_MASK, DIDT_SQ_CTRL3__DIDT_TRIGGER_THROTTLE_LOWBIT__SHIFT, 0x0000 }, + { ixDIDT_SQ_CTRL3, DIDT_SQ_CTRL3__DIDT_POWER_LEVEL_LOWBIT_MASK, DIDT_SQ_CTRL3__DIDT_POWER_LEVEL_LOWBIT__SHIFT, 0x0000 }, + { ixDIDT_SQ_CTRL3, DIDT_SQ_CTRL3__DIDT_STALL_PATTERN_BIT_NUMS_MASK, DIDT_SQ_CTRL3__DIDT_STALL_PATTERN_BIT_NUMS__SHIFT, 0x0003 }, + { ixDIDT_SQ_CTRL3, DIDT_SQ_CTRL3__GC_DIDT_LEVEL_COMB_EN_MASK, DIDT_SQ_CTRL3__GC_DIDT_LEVEL_COMB_EN__SHIFT, 0x0000 }, + { ixDIDT_SQ_CTRL3, DIDT_SQ_CTRL3__SE_DIDT_LEVEL_COMB_EN_MASK, DIDT_SQ_CTRL3__SE_DIDT_LEVEL_COMB_EN__SHIFT, 0x0000 }, + { ixDIDT_SQ_CTRL3, DIDT_SQ_CTRL3__QUALIFY_STALL_EN_MASK, DIDT_SQ_CTRL3__QUALIFY_STALL_EN__SHIFT, 0x0000 }, + { ixDIDT_SQ_CTRL3, DIDT_SQ_CTRL3__DIDT_STALL_SEL_MASK, DIDT_SQ_CTRL3__DIDT_STALL_SEL__SHIFT, 0x0000 }, + { ixDIDT_SQ_CTRL3, DIDT_SQ_CTRL3__DIDT_FORCE_STALL_MASK, DIDT_SQ_CTRL3__DIDT_FORCE_STALL__SHIFT, 0x0000 }, + { ixDIDT_SQ_CTRL3, DIDT_SQ_CTRL3__DIDT_STALL_DELAY_EN_MASK, DIDT_SQ_CTRL3__DIDT_STALL_DELAY_EN__SHIFT, 0x0000 }, + + /*DIDT_TCP_CTRL3 */ + { ixDIDT_TCP_CTRL3, DIDT_TCP_CTRL3__GC_DIDT_ENABLE_MASK, DIDT_TCP_CTRL3__GC_DIDT_ENABLE__SHIFT, 0x0000 }, + { ixDIDT_TCP_CTRL3, DIDT_TCP_CTRL3__GC_DIDT_CLK_EN_OVERRIDE_MASK, DIDT_TCP_CTRL3__GC_DIDT_CLK_EN_OVERRIDE__SHIFT, 0x0000 }, + { ixDIDT_TCP_CTRL3, DIDT_TCP_CTRL3__THROTTLE_POLICY_MASK, DIDT_TCP_CTRL3__THROTTLE_POLICY__SHIFT, 0x0003 }, + { ixDIDT_TCP_CTRL3, DIDT_TCP_CTRL3__DIDT_TRIGGER_THROTTLE_LOWBIT_MASK, DIDT_TCP_CTRL3__DIDT_TRIGGER_THROTTLE_LOWBIT__SHIFT, 0x0000 }, + { ixDIDT_TCP_CTRL3, DIDT_TCP_CTRL3__DIDT_POWER_LEVEL_LOWBIT_MASK, DIDT_TCP_CTRL3__DIDT_POWER_LEVEL_LOWBIT__SHIFT, 0x0000 }, + { ixDIDT_TCP_CTRL3, DIDT_TCP_CTRL3__DIDT_STALL_PATTERN_BIT_NUMS_MASK, DIDT_TCP_CTRL3__DIDT_STALL_PATTERN_BIT_NUMS__SHIFT, 0x0003 }, + { ixDIDT_TCP_CTRL3, DIDT_TCP_CTRL3__GC_DIDT_LEVEL_COMB_EN_MASK, DIDT_TCP_CTRL3__GC_DIDT_LEVEL_COMB_EN__SHIFT, 0x0000 }, + { ixDIDT_TCP_CTRL3, DIDT_TCP_CTRL3__SE_DIDT_LEVEL_COMB_EN_MASK, DIDT_TCP_CTRL3__SE_DIDT_LEVEL_COMB_EN__SHIFT, 0x0000 }, + { ixDIDT_TCP_CTRL3, DIDT_TCP_CTRL3__QUALIFY_STALL_EN_MASK, DIDT_TCP_CTRL3__QUALIFY_STALL_EN__SHIFT, 0x0000 }, + { ixDIDT_TCP_CTRL3, DIDT_TCP_CTRL3__DIDT_STALL_SEL_MASK, DIDT_TCP_CTRL3__DIDT_STALL_SEL__SHIFT, 0x0000 }, + { ixDIDT_TCP_CTRL3, DIDT_TCP_CTRL3__DIDT_FORCE_STALL_MASK, DIDT_TCP_CTRL3__DIDT_FORCE_STALL__SHIFT, 0x0000 }, + { ixDIDT_TCP_CTRL3, DIDT_TCP_CTRL3__DIDT_STALL_DELAY_EN_MASK, DIDT_TCP_CTRL3__DIDT_STALL_DELAY_EN__SHIFT, 0x0000 }, + + /*DIDT_TD_CTRL3 */ + { ixDIDT_TD_CTRL3, DIDT_TD_CTRL3__GC_DIDT_ENABLE_MASK, DIDT_TD_CTRL3__GC_DIDT_ENABLE__SHIFT, 0x0000 }, + { ixDIDT_TD_CTRL3, DIDT_TD_CTRL3__GC_DIDT_CLK_EN_OVERRIDE_MASK, DIDT_TD_CTRL3__GC_DIDT_CLK_EN_OVERRIDE__SHIFT, 0x0000 }, + { ixDIDT_TD_CTRL3, DIDT_TD_CTRL3__THROTTLE_POLICY_MASK, DIDT_TD_CTRL3__THROTTLE_POLICY__SHIFT, 0x0003 }, + { ixDIDT_TD_CTRL3, DIDT_TD_CTRL3__DIDT_TRIGGER_THROTTLE_LOWBIT_MASK, DIDT_TD_CTRL3__DIDT_TRIGGER_THROTTLE_LOWBIT__SHIFT, 0x0000 }, + { ixDIDT_TD_CTRL3, DIDT_TD_CTRL3__DIDT_POWER_LEVEL_LOWBIT_MASK, DIDT_TD_CTRL3__DIDT_POWER_LEVEL_LOWBIT__SHIFT, 0x0000 }, + { ixDIDT_TD_CTRL3, DIDT_TD_CTRL3__DIDT_STALL_PATTERN_BIT_NUMS_MASK, DIDT_TD_CTRL3__DIDT_STALL_PATTERN_BIT_NUMS__SHIFT, 0x0003 }, + { ixDIDT_TD_CTRL3, DIDT_TD_CTRL3__GC_DIDT_LEVEL_COMB_EN_MASK, DIDT_TD_CTRL3__GC_DIDT_LEVEL_COMB_EN__SHIFT, 0x0000 }, + { ixDIDT_TD_CTRL3, DIDT_TD_CTRL3__SE_DIDT_LEVEL_COMB_EN_MASK, DIDT_TD_CTRL3__SE_DIDT_LEVEL_COMB_EN__SHIFT, 0x0000 }, + { ixDIDT_TD_CTRL3, DIDT_TD_CTRL3__QUALIFY_STALL_EN_MASK, DIDT_TD_CTRL3__QUALIFY_STALL_EN__SHIFT, 0x0000 }, + { ixDIDT_TD_CTRL3, DIDT_TD_CTRL3__DIDT_STALL_SEL_MASK, DIDT_TD_CTRL3__DIDT_STALL_SEL__SHIFT, 0x0000 }, + { ixDIDT_TD_CTRL3, DIDT_TD_CTRL3__DIDT_FORCE_STALL_MASK, DIDT_TD_CTRL3__DIDT_FORCE_STALL__SHIFT, 0x0000 }, + { ixDIDT_TD_CTRL3, DIDT_TD_CTRL3__DIDT_STALL_DELAY_EN_MASK, DIDT_TD_CTRL3__DIDT_STALL_DELAY_EN__SHIFT, 0x0000 }, + + /*DIDT_DB_CTRL3 */ + { ixDIDT_DB_CTRL3, DIDT_DB_CTRL3__GC_DIDT_ENABLE_MASK, DIDT_DB_CTRL3__GC_DIDT_ENABLE__SHIFT, 0x0000 }, + { ixDIDT_DB_CTRL3, DIDT_DB_CTRL3__GC_DIDT_CLK_EN_OVERRIDE_MASK, DIDT_DB_CTRL3__GC_DIDT_CLK_EN_OVERRIDE__SHIFT, 0x0000 }, + { ixDIDT_DB_CTRL3, DIDT_DB_CTRL3__THROTTLE_POLICY_MASK, DIDT_DB_CTRL3__THROTTLE_POLICY__SHIFT, 0x0003 }, + { ixDIDT_DB_CTRL3, DIDT_DB_CTRL3__DIDT_TRIGGER_THROTTLE_LOWBIT_MASK, DIDT_DB_CTRL3__DIDT_TRIGGER_THROTTLE_LOWBIT__SHIFT, 0x0000 }, + { ixDIDT_DB_CTRL3, DIDT_DB_CTRL3__DIDT_POWER_LEVEL_LOWBIT_MASK, DIDT_DB_CTRL3__DIDT_POWER_LEVEL_LOWBIT__SHIFT, 0x0000 }, + { ixDIDT_DB_CTRL3, DIDT_DB_CTRL3__DIDT_STALL_PATTERN_BIT_NUMS_MASK, DIDT_DB_CTRL3__DIDT_STALL_PATTERN_BIT_NUMS__SHIFT, 0x0003 }, + { ixDIDT_DB_CTRL3, DIDT_DB_CTRL3__GC_DIDT_LEVEL_COMB_EN_MASK, DIDT_DB_CTRL3__GC_DIDT_LEVEL_COMB_EN__SHIFT, 0x0000 }, + { ixDIDT_DB_CTRL3, DIDT_DB_CTRL3__SE_DIDT_LEVEL_COMB_EN_MASK, DIDT_DB_CTRL3__SE_DIDT_LEVEL_COMB_EN__SHIFT, 0x0000 }, + { ixDIDT_DB_CTRL3, DIDT_DB_CTRL3__QUALIFY_STALL_EN_MASK, DIDT_DB_CTRL3__QUALIFY_STALL_EN__SHIFT, 0x0000 }, + { ixDIDT_DB_CTRL3, DIDT_DB_CTRL3__DIDT_STALL_SEL_MASK, DIDT_DB_CTRL3__DIDT_STALL_SEL__SHIFT, 0x0000 }, + { ixDIDT_DB_CTRL3, DIDT_DB_CTRL3__DIDT_FORCE_STALL_MASK, DIDT_DB_CTRL3__DIDT_FORCE_STALL__SHIFT, 0x0000 }, + { ixDIDT_DB_CTRL3, DIDT_DB_CTRL3__DIDT_STALL_DELAY_EN_MASK, DIDT_DB_CTRL3__DIDT_STALL_DELAY_EN__SHIFT, 0x0000 }, + + { 0xFFFFFFFF } /* End of list */ +}; + +static const struct vega10_didt_config_reg SEDiDtCtrl2Config_Vega10[] = +{ +/* --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- + * Offset Mask Shift Value + * --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- + */ + /* DIDT_SQ */ + { ixDIDT_SQ_CTRL2, DIDT_SQ_CTRL2__MAX_POWER_DELTA_MASK, DIDT_SQ_CTRL2__MAX_POWER_DELTA__SHIFT, 0x3853 }, + { ixDIDT_SQ_CTRL2, DIDT_SQ_CTRL2__SHORT_TERM_INTERVAL_SIZE_MASK, DIDT_SQ_CTRL2__SHORT_TERM_INTERVAL_SIZE__SHIFT, 0x00c0 }, + { ixDIDT_SQ_CTRL2, DIDT_SQ_CTRL2__LONG_TERM_INTERVAL_RATIO_MASK, DIDT_SQ_CTRL2__LONG_TERM_INTERVAL_RATIO__SHIFT, 0x0000 }, + + /* DIDT_TD */ + { ixDIDT_TD_CTRL2, DIDT_TD_CTRL2__MAX_POWER_DELTA_MASK, DIDT_TD_CTRL2__MAX_POWER_DELTA__SHIFT, 0x3fff }, + { ixDIDT_TD_CTRL2, DIDT_TD_CTRL2__SHORT_TERM_INTERVAL_SIZE_MASK, DIDT_TD_CTRL2__SHORT_TERM_INTERVAL_SIZE__SHIFT, 0x00c0 }, + { ixDIDT_TD_CTRL2, DIDT_TD_CTRL2__LONG_TERM_INTERVAL_RATIO_MASK, DIDT_TD_CTRL2__LONG_TERM_INTERVAL_RATIO__SHIFT, 0x0001 }, + + /* DIDT_TCP */ + { ixDIDT_TCP_CTRL2, DIDT_TCP_CTRL2__MAX_POWER_DELTA_MASK, DIDT_TCP_CTRL2__MAX_POWER_DELTA__SHIFT, 0x3dde }, + { ixDIDT_TCP_CTRL2, DIDT_TCP_CTRL2__SHORT_TERM_INTERVAL_SIZE_MASK, DIDT_TCP_CTRL2__SHORT_TERM_INTERVAL_SIZE__SHIFT, 0x00c0 }, + { ixDIDT_TCP_CTRL2, DIDT_TCP_CTRL2__LONG_TERM_INTERVAL_RATIO_MASK, DIDT_TCP_CTRL2__LONG_TERM_INTERVAL_RATIO__SHIFT, 0x0001 }, + + /* DIDT_DB */ + { ixDIDT_DB_CTRL2, DIDT_DB_CTRL2__MAX_POWER_DELTA_MASK, DIDT_DB_CTRL2__MAX_POWER_DELTA__SHIFT, 0x3dde }, + { ixDIDT_DB_CTRL2, DIDT_DB_CTRL2__SHORT_TERM_INTERVAL_SIZE_MASK, DIDT_DB_CTRL2__SHORT_TERM_INTERVAL_SIZE__SHIFT, 0x00c0 }, + { ixDIDT_DB_CTRL2, DIDT_DB_CTRL2__LONG_TERM_INTERVAL_RATIO_MASK, DIDT_DB_CTRL2__LONG_TERM_INTERVAL_RATIO__SHIFT, 0x0001 }, + + { 0xFFFFFFFF } /* End of list */ +}; + +static const struct vega10_didt_config_reg SEDiDtCtrl1Config_Vega10[] = +{ +/* --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- + * Offset Mask Shift Value + * --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- + */ + /* DIDT_SQ */ + { ixDIDT_SQ_CTRL1, DIDT_SQ_CTRL1__MIN_POWER_MASK, DIDT_SQ_CTRL1__MIN_POWER__SHIFT, 0x0000 }, + { ixDIDT_SQ_CTRL1, DIDT_SQ_CTRL1__MAX_POWER_MASK, DIDT_SQ_CTRL1__MAX_POWER__SHIFT, 0xffff }, + /* DIDT_TD */ + { ixDIDT_TD_CTRL1, DIDT_TD_CTRL1__MIN_POWER_MASK, DIDT_TD_CTRL1__MIN_POWER__SHIFT, 0x0000 }, + { ixDIDT_TD_CTRL1, DIDT_TD_CTRL1__MAX_POWER_MASK, DIDT_TD_CTRL1__MAX_POWER__SHIFT, 0xffff }, + /* DIDT_TCP */ + { ixDIDT_TCP_CTRL1, DIDT_TCP_CTRL1__MIN_POWER_MASK, DIDT_TCP_CTRL1__MIN_POWER__SHIFT, 0x0000 }, + { ixDIDT_TCP_CTRL1, DIDT_TCP_CTRL1__MAX_POWER_MASK, DIDT_TCP_CTRL1__MAX_POWER__SHIFT, 0xffff }, + /* DIDT_DB */ + { ixDIDT_DB_CTRL1, DIDT_DB_CTRL1__MIN_POWER_MASK, DIDT_DB_CTRL1__MIN_POWER__SHIFT, 0x0000 }, + { ixDIDT_DB_CTRL1, DIDT_DB_CTRL1__MAX_POWER_MASK, DIDT_DB_CTRL1__MAX_POWER__SHIFT, 0xffff }, + + { 0xFFFFFFFF } /* End of list */ +}; + + +static const struct vega10_didt_config_reg SEDiDtWeightConfig_Vega10[] = +{ +/* --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- + * Offset Mask Shift Value + * --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- + */ + /* DIDT_SQ */ + { ixDIDT_SQ_WEIGHT0_3, 0xFFFFFFFF, 0, 0x2B363B1A }, + { ixDIDT_SQ_WEIGHT4_7, 0xFFFFFFFF, 0, 0x270B2432 }, + { ixDIDT_SQ_WEIGHT8_11, 0xFFFFFFFF, 0, 0x00000018 }, + + /* DIDT_TD */ + { ixDIDT_TD_WEIGHT0_3, 0xFFFFFFFF, 0, 0x2B1D220F }, + { ixDIDT_TD_WEIGHT4_7, 0xFFFFFFFF, 0, 0x00007558 }, + { ixDIDT_TD_WEIGHT8_11, 0xFFFFFFFF, 0, 0x00000000 }, + + /* DIDT_TCP */ + { ixDIDT_TCP_WEIGHT0_3, 0xFFFFFFFF, 0, 0x5ACE160D }, + { ixDIDT_TCP_WEIGHT4_7, 0xFFFFFFFF, 0, 0x00000000 }, + { ixDIDT_TCP_WEIGHT8_11, 0xFFFFFFFF, 0, 0x00000000 }, + + /* DIDT_DB */ + { ixDIDT_DB_WEIGHT0_3, 0xFFFFFFFF, 0, 0x0E152A0F }, + { ixDIDT_DB_WEIGHT4_7, 0xFFFFFFFF, 0, 0x09061813 }, + { ixDIDT_DB_WEIGHT8_11, 0xFFFFFFFF, 0, 0x00000013 }, + + { 0xFFFFFFFF } /* End of list */ +}; + +static const struct vega10_didt_config_reg SEDiDtCtrl0Config_Vega10[] = +{ +/* --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- + * Offset Mask Shift Value + * --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- + */ + /* DIDT_SQ */ + { ixDIDT_SQ_CTRL0, DIDT_SQ_CTRL0__DIDT_CTRL_EN_MASK, DIDT_SQ_CTRL0__DIDT_CTRL_EN__SHIFT, 0x0000 }, + { ixDIDT_SQ_CTRL0, DIDT_SQ_CTRL0__PHASE_OFFSET_MASK, DIDT_SQ_CTRL0__PHASE_OFFSET__SHIFT, 0x0000 }, + { ixDIDT_SQ_CTRL0, DIDT_SQ_CTRL0__DIDT_CTRL_RST_MASK, DIDT_SQ_CTRL0__DIDT_CTRL_RST__SHIFT, 0x0000 }, + { ixDIDT_SQ_CTRL0, DIDT_SQ_CTRL0__DIDT_CLK_EN_OVERRIDE_MASK, DIDT_SQ_CTRL0__DIDT_CLK_EN_OVERRIDE__SHIFT, 0x0000 }, + { ixDIDT_SQ_CTRL0, DIDT_SQ_CTRL0__DIDT_STALL_CTRL_EN_MASK, DIDT_SQ_CTRL0__DIDT_STALL_CTRL_EN__SHIFT, 0x0001 }, + { ixDIDT_SQ_CTRL0, DIDT_SQ_CTRL0__DIDT_TUNING_CTRL_EN_MASK, DIDT_SQ_CTRL0__DIDT_TUNING_CTRL_EN__SHIFT, 0x0001 }, + { ixDIDT_SQ_CTRL0, DIDT_SQ_CTRL0__DIDT_STALL_AUTO_RELEASE_EN_MASK, DIDT_SQ_CTRL0__DIDT_STALL_AUTO_RELEASE_EN__SHIFT, 0x0001 }, + { ixDIDT_SQ_CTRL0, DIDT_SQ_CTRL0__DIDT_HI_POWER_THRESHOLD_MASK, DIDT_SQ_CTRL0__DIDT_HI_POWER_THRESHOLD__SHIFT, 0xffff }, + { ixDIDT_SQ_CTRL0, DIDT_SQ_CTRL0__DIDT_AUTO_MPD_EN_MASK, DIDT_SQ_CTRL0__DIDT_AUTO_MPD_EN__SHIFT, 0x0000 }, + { ixDIDT_SQ_CTRL0, DIDT_SQ_CTRL0__DIDT_STALL_EVENT_EN_MASK, DIDT_SQ_CTRL0__DIDT_STALL_EVENT_EN__SHIFT, 0x0000 }, + { ixDIDT_SQ_CTRL0, DIDT_SQ_CTRL0__DIDT_STALL_EVENT_COUNTER_CLEAR_MASK, DIDT_SQ_CTRL0__DIDT_STALL_EVENT_COUNTER_CLEAR__SHIFT, 0x0000 }, + /* DIDT_TD */ + { ixDIDT_TD_CTRL0, DIDT_TD_CTRL0__DIDT_CTRL_EN_MASK, DIDT_TD_CTRL0__DIDT_CTRL_EN__SHIFT, 0x0000 }, + { ixDIDT_TD_CTRL0, DIDT_TD_CTRL0__PHASE_OFFSET_MASK, DIDT_TD_CTRL0__PHASE_OFFSET__SHIFT, 0x0000 }, + { ixDIDT_TD_CTRL0, DIDT_TD_CTRL0__DIDT_CTRL_RST_MASK, DIDT_TD_CTRL0__DIDT_CTRL_RST__SHIFT, 0x0000 }, + { ixDIDT_TD_CTRL0, DIDT_TD_CTRL0__DIDT_CLK_EN_OVERRIDE_MASK, DIDT_TD_CTRL0__DIDT_CLK_EN_OVERRIDE__SHIFT, 0x0000 }, + { ixDIDT_TD_CTRL0, DIDT_TD_CTRL0__DIDT_STALL_CTRL_EN_MASK, DIDT_TD_CTRL0__DIDT_STALL_CTRL_EN__SHIFT, 0x0001 }, + { ixDIDT_TD_CTRL0, DIDT_TD_CTRL0__DIDT_TUNING_CTRL_EN_MASK, DIDT_TD_CTRL0__DIDT_TUNING_CTRL_EN__SHIFT, 0x0001 }, + { ixDIDT_TD_CTRL0, DIDT_TD_CTRL0__DIDT_STALL_AUTO_RELEASE_EN_MASK, DIDT_TD_CTRL0__DIDT_STALL_AUTO_RELEASE_EN__SHIFT, 0x0001 }, + { ixDIDT_TD_CTRL0, DIDT_TD_CTRL0__DIDT_HI_POWER_THRESHOLD_MASK, DIDT_TD_CTRL0__DIDT_HI_POWER_THRESHOLD__SHIFT, 0xffff }, + { ixDIDT_TD_CTRL0, DIDT_TD_CTRL0__DIDT_AUTO_MPD_EN_MASK, DIDT_TD_CTRL0__DIDT_AUTO_MPD_EN__SHIFT, 0x0000 }, + { ixDIDT_TD_CTRL0, DIDT_TD_CTRL0__DIDT_STALL_EVENT_EN_MASK, DIDT_TD_CTRL0__DIDT_STALL_EVENT_EN__SHIFT, 0x0000 }, + { ixDIDT_TD_CTRL0, DIDT_TD_CTRL0__DIDT_STALL_EVENT_COUNTER_CLEAR_MASK, DIDT_TD_CTRL0__DIDT_STALL_EVENT_COUNTER_CLEAR__SHIFT, 0x0000 }, + /* DIDT_TCP */ + { ixDIDT_TCP_CTRL0, DIDT_TCP_CTRL0__DIDT_CTRL_EN_MASK, DIDT_TCP_CTRL0__DIDT_CTRL_EN__SHIFT, 0x0000 }, + { ixDIDT_TCP_CTRL0, DIDT_TCP_CTRL0__PHASE_OFFSET_MASK, DIDT_TCP_CTRL0__PHASE_OFFSET__SHIFT, 0x0000 }, + { ixDIDT_TCP_CTRL0, DIDT_TCP_CTRL0__DIDT_CTRL_RST_MASK, DIDT_TCP_CTRL0__DIDT_CTRL_RST__SHIFT, 0x0000 }, + { ixDIDT_TCP_CTRL0, DIDT_TCP_CTRL0__DIDT_CLK_EN_OVERRIDE_MASK, DIDT_TCP_CTRL0__DIDT_CLK_EN_OVERRIDE__SHIFT, 0x0000 }, + { ixDIDT_TCP_CTRL0, DIDT_TCP_CTRL0__DIDT_STALL_CTRL_EN_MASK, DIDT_TCP_CTRL0__DIDT_STALL_CTRL_EN__SHIFT, 0x0001 }, + { ixDIDT_TCP_CTRL0, DIDT_TCP_CTRL0__DIDT_TUNING_CTRL_EN_MASK, DIDT_TCP_CTRL0__DIDT_TUNING_CTRL_EN__SHIFT, 0x0001 }, + { ixDIDT_TCP_CTRL0, DIDT_TCP_CTRL0__DIDT_STALL_AUTO_RELEASE_EN_MASK, DIDT_TCP_CTRL0__DIDT_STALL_AUTO_RELEASE_EN__SHIFT, 0x0001 }, + { ixDIDT_TCP_CTRL0, DIDT_TCP_CTRL0__DIDT_HI_POWER_THRESHOLD_MASK, DIDT_TCP_CTRL0__DIDT_HI_POWER_THRESHOLD__SHIFT, 0xffff }, + { ixDIDT_TCP_CTRL0, DIDT_TCP_CTRL0__DIDT_AUTO_MPD_EN_MASK, DIDT_TCP_CTRL0__DIDT_AUTO_MPD_EN__SHIFT, 0x0000 }, + { ixDIDT_TCP_CTRL0, DIDT_TCP_CTRL0__DIDT_STALL_EVENT_EN_MASK, DIDT_TCP_CTRL0__DIDT_STALL_EVENT_EN__SHIFT, 0x0000 }, + { ixDIDT_TCP_CTRL0, DIDT_TCP_CTRL0__DIDT_STALL_EVENT_COUNTER_CLEAR_MASK, DIDT_TCP_CTRL0__DIDT_STALL_EVENT_COUNTER_CLEAR__SHIFT, 0x0000 }, + /* DIDT_DB */ + { ixDIDT_DB_CTRL0, DIDT_DB_CTRL0__DIDT_CTRL_EN_MASK, DIDT_DB_CTRL0__DIDT_CTRL_EN__SHIFT, 0x0000 }, + { ixDIDT_DB_CTRL0, DIDT_DB_CTRL0__PHASE_OFFSET_MASK, DIDT_DB_CTRL0__PHASE_OFFSET__SHIFT, 0x0000 }, + { ixDIDT_DB_CTRL0, DIDT_DB_CTRL0__DIDT_CTRL_RST_MASK, DIDT_DB_CTRL0__DIDT_CTRL_RST__SHIFT, 0x0000 }, + { ixDIDT_DB_CTRL0, DIDT_DB_CTRL0__DIDT_CLK_EN_OVERRIDE_MASK, DIDT_DB_CTRL0__DIDT_CLK_EN_OVERRIDE__SHIFT, 0x0000 }, + { ixDIDT_DB_CTRL0, DIDT_DB_CTRL0__DIDT_STALL_CTRL_EN_MASK, DIDT_DB_CTRL0__DIDT_STALL_CTRL_EN__SHIFT, 0x0001 }, + { ixDIDT_DB_CTRL0, DIDT_DB_CTRL0__DIDT_TUNING_CTRL_EN_MASK, DIDT_DB_CTRL0__DIDT_TUNING_CTRL_EN__SHIFT, 0x0001 }, + { ixDIDT_DB_CTRL0, DIDT_DB_CTRL0__DIDT_STALL_AUTO_RELEASE_EN_MASK, DIDT_DB_CTRL0__DIDT_STALL_AUTO_RELEASE_EN__SHIFT, 0x0001 }, + { ixDIDT_DB_CTRL0, DIDT_DB_CTRL0__DIDT_HI_POWER_THRESHOLD_MASK, DIDT_DB_CTRL0__DIDT_HI_POWER_THRESHOLD__SHIFT, 0xffff }, + { ixDIDT_DB_CTRL0, DIDT_DB_CTRL0__DIDT_AUTO_MPD_EN_MASK, DIDT_DB_CTRL0__DIDT_AUTO_MPD_EN__SHIFT, 0x0000 }, + { ixDIDT_DB_CTRL0, DIDT_DB_CTRL0__DIDT_STALL_EVENT_EN_MASK, DIDT_DB_CTRL0__DIDT_STALL_EVENT_EN__SHIFT, 0x0000 }, + { ixDIDT_DB_CTRL0, DIDT_DB_CTRL0__DIDT_STALL_EVENT_COUNTER_CLEAR_MASK, DIDT_DB_CTRL0__DIDT_STALL_EVENT_COUNTER_CLEAR__SHIFT, 0x0000 }, + + { 0xFFFFFFFF } /* End of list */ +}; + + +static const struct vega10_didt_config_reg SEDiDtStallCtrlConfig_vega10[] = +{ +/* --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- + * Offset Mask Shift Value + * --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- + */ + /* DIDT_SQ */ + { ixDIDT_SQ_STALL_CTRL, DIDT_SQ_STALL_CTRL__DIDT_STALL_DELAY_HI_MASK, DIDT_SQ_STALL_CTRL__DIDT_STALL_DELAY_HI__SHIFT, 0x0004 }, + { ixDIDT_SQ_STALL_CTRL, DIDT_SQ_STALL_CTRL__DIDT_STALL_DELAY_LO_MASK, DIDT_SQ_STALL_CTRL__DIDT_STALL_DELAY_LO__SHIFT, 0x0004 }, + { ixDIDT_SQ_STALL_CTRL, DIDT_SQ_STALL_CTRL__DIDT_MAX_STALLS_ALLOWED_HI_MASK, DIDT_SQ_STALL_CTRL__DIDT_MAX_STALLS_ALLOWED_HI__SHIFT, 0x000a }, + { ixDIDT_SQ_STALL_CTRL, DIDT_SQ_STALL_CTRL__DIDT_MAX_STALLS_ALLOWED_LO_MASK, DIDT_SQ_STALL_CTRL__DIDT_MAX_STALLS_ALLOWED_LO__SHIFT, 0x000a }, + + /* DIDT_TD */ + { ixDIDT_TD_STALL_CTRL, DIDT_TD_STALL_CTRL__DIDT_STALL_DELAY_HI_MASK, DIDT_TD_STALL_CTRL__DIDT_STALL_DELAY_HI__SHIFT, 0x0001 }, + { ixDIDT_TD_STALL_CTRL, DIDT_TD_STALL_CTRL__DIDT_STALL_DELAY_LO_MASK, DIDT_TD_STALL_CTRL__DIDT_STALL_DELAY_LO__SHIFT, 0x0001 }, + { ixDIDT_TD_STALL_CTRL, DIDT_TD_STALL_CTRL__DIDT_MAX_STALLS_ALLOWED_HI_MASK, DIDT_TD_STALL_CTRL__DIDT_MAX_STALLS_ALLOWED_HI__SHIFT, 0x000a }, + { ixDIDT_TD_STALL_CTRL, DIDT_TD_STALL_CTRL__DIDT_MAX_STALLS_ALLOWED_LO_MASK, DIDT_TD_STALL_CTRL__DIDT_MAX_STALLS_ALLOWED_LO__SHIFT, 0x000a }, + + /* DIDT_TCP */ + { ixDIDT_TCP_STALL_CTRL, DIDT_TCP_STALL_CTRL__DIDT_STALL_DELAY_HI_MASK, DIDT_TCP_STALL_CTRL__DIDT_STALL_DELAY_HI__SHIFT, 0x0001 }, + { ixDIDT_TCP_STALL_CTRL, DIDT_TCP_STALL_CTRL__DIDT_STALL_DELAY_LO_MASK, DIDT_TCP_STALL_CTRL__DIDT_STALL_DELAY_LO__SHIFT, 0x0001 }, + { ixDIDT_TCP_STALL_CTRL, DIDT_TCP_STALL_CTRL__DIDT_MAX_STALLS_ALLOWED_HI_MASK, DIDT_TCP_STALL_CTRL__DIDT_MAX_STALLS_ALLOWED_HI__SHIFT, 0x000a }, + { ixDIDT_TCP_STALL_CTRL, DIDT_TCP_STALL_CTRL__DIDT_MAX_STALLS_ALLOWED_LO_MASK, DIDT_TCP_STALL_CTRL__DIDT_MAX_STALLS_ALLOWED_LO__SHIFT, 0x000a }, + + /* DIDT_DB */ + { ixDIDT_DB_STALL_CTRL, DIDT_DB_STALL_CTRL__DIDT_STALL_DELAY_HI_MASK, DIDT_DB_STALL_CTRL__DIDT_STALL_DELAY_HI__SHIFT, 0x0004 }, + { ixDIDT_DB_STALL_CTRL, DIDT_DB_STALL_CTRL__DIDT_STALL_DELAY_LO_MASK, DIDT_DB_STALL_CTRL__DIDT_STALL_DELAY_LO__SHIFT, 0x0004 }, + { ixDIDT_DB_STALL_CTRL, DIDT_DB_STALL_CTRL__DIDT_MAX_STALLS_ALLOWED_HI_MASK, DIDT_DB_STALL_CTRL__DIDT_MAX_STALLS_ALLOWED_HI__SHIFT, 0x000a }, + { ixDIDT_DB_STALL_CTRL, DIDT_DB_STALL_CTRL__DIDT_MAX_STALLS_ALLOWED_LO_MASK, DIDT_DB_STALL_CTRL__DIDT_MAX_STALLS_ALLOWED_LO__SHIFT, 0x000a }, + + { 0xFFFFFFFF } /* End of list */ +}; + +static const struct vega10_didt_config_reg SEDiDtStallPatternConfig_vega10[] = +{ +/* --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- + * Offset Mask Shift Value + * --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- + */ + /* DIDT_SQ_STALL_PATTERN_1_2 */ + { ixDIDT_SQ_STALL_PATTERN_1_2, DIDT_SQ_STALL_PATTERN_1_2__DIDT_STALL_PATTERN_1_MASK, DIDT_SQ_STALL_PATTERN_1_2__DIDT_STALL_PATTERN_1__SHIFT, 0x0001 }, + { ixDIDT_SQ_STALL_PATTERN_1_2, DIDT_SQ_STALL_PATTERN_1_2__DIDT_STALL_PATTERN_2_MASK, DIDT_SQ_STALL_PATTERN_1_2__DIDT_STALL_PATTERN_2__SHIFT, 0x0001 }, + + /* DIDT_SQ_STALL_PATTERN_3_4 */ + { ixDIDT_SQ_STALL_PATTERN_3_4, DIDT_SQ_STALL_PATTERN_3_4__DIDT_STALL_PATTERN_3_MASK, DIDT_SQ_STALL_PATTERN_3_4__DIDT_STALL_PATTERN_3__SHIFT, 0x0001 }, + { ixDIDT_SQ_STALL_PATTERN_3_4, DIDT_SQ_STALL_PATTERN_3_4__DIDT_STALL_PATTERN_4_MASK, DIDT_SQ_STALL_PATTERN_3_4__DIDT_STALL_PATTERN_4__SHIFT, 0x0001 }, + + /* DIDT_SQ_STALL_PATTERN_5_6 */ + { ixDIDT_SQ_STALL_PATTERN_5_6, DIDT_SQ_STALL_PATTERN_5_6__DIDT_STALL_PATTERN_5_MASK, DIDT_SQ_STALL_PATTERN_5_6__DIDT_STALL_PATTERN_5__SHIFT, 0x0000 }, + { ixDIDT_SQ_STALL_PATTERN_5_6, DIDT_SQ_STALL_PATTERN_5_6__DIDT_STALL_PATTERN_6_MASK, DIDT_SQ_STALL_PATTERN_5_6__DIDT_STALL_PATTERN_6__SHIFT, 0x0000 }, + + /* DIDT_SQ_STALL_PATTERN_7 */ + { ixDIDT_SQ_STALL_PATTERN_7, DIDT_SQ_STALL_PATTERN_7__DIDT_STALL_PATTERN_7_MASK, DIDT_SQ_STALL_PATTERN_7__DIDT_STALL_PATTERN_7__SHIFT, 0x0000 }, + + /* DIDT_TCP_STALL_PATTERN_1_2 */ + { ixDIDT_TCP_STALL_PATTERN_1_2, DIDT_TCP_STALL_PATTERN_1_2__DIDT_STALL_PATTERN_1_MASK, DIDT_TCP_STALL_PATTERN_1_2__DIDT_STALL_PATTERN_1__SHIFT, 0x0001 }, + { ixDIDT_TCP_STALL_PATTERN_1_2, DIDT_TCP_STALL_PATTERN_1_2__DIDT_STALL_PATTERN_2_MASK, DIDT_TCP_STALL_PATTERN_1_2__DIDT_STALL_PATTERN_2__SHIFT, 0x0001 }, + + /* DIDT_TCP_STALL_PATTERN_3_4 */ + { ixDIDT_TCP_STALL_PATTERN_3_4, DIDT_TCP_STALL_PATTERN_3_4__DIDT_STALL_PATTERN_3_MASK, DIDT_TCP_STALL_PATTERN_3_4__DIDT_STALL_PATTERN_3__SHIFT, 0x0001 }, + { ixDIDT_TCP_STALL_PATTERN_3_4, DIDT_TCP_STALL_PATTERN_3_4__DIDT_STALL_PATTERN_4_MASK, DIDT_TCP_STALL_PATTERN_3_4__DIDT_STALL_PATTERN_4__SHIFT, 0x0001 }, + + /* DIDT_TCP_STALL_PATTERN_5_6 */ + { ixDIDT_TCP_STALL_PATTERN_5_6, DIDT_TCP_STALL_PATTERN_5_6__DIDT_STALL_PATTERN_5_MASK, DIDT_TCP_STALL_PATTERN_5_6__DIDT_STALL_PATTERN_5__SHIFT, 0x0000 }, + { ixDIDT_TCP_STALL_PATTERN_5_6, DIDT_TCP_STALL_PATTERN_5_6__DIDT_STALL_PATTERN_6_MASK, DIDT_TCP_STALL_PATTERN_5_6__DIDT_STALL_PATTERN_6__SHIFT, 0x0000 }, + + /* DIDT_TCP_STALL_PATTERN_7 */ + { ixDIDT_TCP_STALL_PATTERN_7, DIDT_TCP_STALL_PATTERN_7__DIDT_STALL_PATTERN_7_MASK, DIDT_TCP_STALL_PATTERN_7__DIDT_STALL_PATTERN_7__SHIFT, 0x0000 }, + + /* DIDT_TD_STALL_PATTERN_1_2 */ + { ixDIDT_TD_STALL_PATTERN_1_2, DIDT_TD_STALL_PATTERN_1_2__DIDT_STALL_PATTERN_1_MASK, DIDT_TD_STALL_PATTERN_1_2__DIDT_STALL_PATTERN_1__SHIFT, 0x0001 }, + { ixDIDT_TD_STALL_PATTERN_1_2, DIDT_TD_STALL_PATTERN_1_2__DIDT_STALL_PATTERN_2_MASK, DIDT_TD_STALL_PATTERN_1_2__DIDT_STALL_PATTERN_2__SHIFT, 0x0001 }, + + /* DIDT_TD_STALL_PATTERN_3_4 */ + { ixDIDT_TD_STALL_PATTERN_3_4, DIDT_TD_STALL_PATTERN_3_4__DIDT_STALL_PATTERN_3_MASK, DIDT_TD_STALL_PATTERN_3_4__DIDT_STALL_PATTERN_3__SHIFT, 0x0001 }, + { ixDIDT_TD_STALL_PATTERN_3_4, DIDT_TD_STALL_PATTERN_3_4__DIDT_STALL_PATTERN_4_MASK, DIDT_TD_STALL_PATTERN_3_4__DIDT_STALL_PATTERN_4__SHIFT, 0x0001 }, + + /* DIDT_TD_STALL_PATTERN_5_6 */ + { ixDIDT_TD_STALL_PATTERN_5_6, DIDT_TD_STALL_PATTERN_5_6__DIDT_STALL_PATTERN_5_MASK, DIDT_TD_STALL_PATTERN_5_6__DIDT_STALL_PATTERN_5__SHIFT, 0x0000 }, + { ixDIDT_TD_STALL_PATTERN_5_6, DIDT_TD_STALL_PATTERN_5_6__DIDT_STALL_PATTERN_6_MASK, DIDT_TD_STALL_PATTERN_5_6__DIDT_STALL_PATTERN_6__SHIFT, 0x0000 }, + + /* DIDT_TD_STALL_PATTERN_7 */ + { ixDIDT_TD_STALL_PATTERN_7, DIDT_TD_STALL_PATTERN_7__DIDT_STALL_PATTERN_7_MASK, DIDT_TD_STALL_PATTERN_7__DIDT_STALL_PATTERN_7__SHIFT, 0x0000 }, + + /* DIDT_DB_STALL_PATTERN_1_2 */ + { ixDIDT_DB_STALL_PATTERN_1_2, DIDT_DB_STALL_PATTERN_1_2__DIDT_STALL_PATTERN_1_MASK, DIDT_DB_STALL_PATTERN_1_2__DIDT_STALL_PATTERN_1__SHIFT, 0x0001 }, + { ixDIDT_DB_STALL_PATTERN_1_2, DIDT_DB_STALL_PATTERN_1_2__DIDT_STALL_PATTERN_2_MASK, DIDT_DB_STALL_PATTERN_1_2__DIDT_STALL_PATTERN_2__SHIFT, 0x0001 }, + + /* DIDT_DB_STALL_PATTERN_3_4 */ + { ixDIDT_DB_STALL_PATTERN_3_4, DIDT_DB_STALL_PATTERN_3_4__DIDT_STALL_PATTERN_3_MASK, DIDT_DB_STALL_PATTERN_3_4__DIDT_STALL_PATTERN_3__SHIFT, 0x0001 }, + { ixDIDT_DB_STALL_PATTERN_3_4, DIDT_DB_STALL_PATTERN_3_4__DIDT_STALL_PATTERN_4_MASK, DIDT_DB_STALL_PATTERN_3_4__DIDT_STALL_PATTERN_4__SHIFT, 0x0001 }, + + /* DIDT_DB_STALL_PATTERN_5_6 */ + { ixDIDT_DB_STALL_PATTERN_5_6, DIDT_DB_STALL_PATTERN_5_6__DIDT_STALL_PATTERN_5_MASK, DIDT_DB_STALL_PATTERN_5_6__DIDT_STALL_PATTERN_5__SHIFT, 0x0000 }, + { ixDIDT_DB_STALL_PATTERN_5_6, DIDT_DB_STALL_PATTERN_5_6__DIDT_STALL_PATTERN_6_MASK, DIDT_DB_STALL_PATTERN_5_6__DIDT_STALL_PATTERN_6__SHIFT, 0x0000 }, + + /* DIDT_DB_STALL_PATTERN_7 */ + { ixDIDT_DB_STALL_PATTERN_7, DIDT_DB_STALL_PATTERN_7__DIDT_STALL_PATTERN_7_MASK, DIDT_DB_STALL_PATTERN_7__DIDT_STALL_PATTERN_7__SHIFT, 0x0000 }, + + { 0xFFFFFFFF } /* End of list */ +}; + +static const struct vega10_didt_config_reg SELCacConfig_Vega10[] = +{ +/* --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- + * Offset Mask Shift Value + * --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- + */ + /* SQ */ + { ixSE_CAC_CNTL, 0xFFFFFFFF, 0, 0x00060021 }, + { ixSE_CAC_CNTL, 0xFFFFFFFF, 0, 0x00860021 }, + { ixSE_CAC_CNTL, 0xFFFFFFFF, 0, 0x01060021 }, + { ixSE_CAC_CNTL, 0xFFFFFFFF, 0, 0x01860021 }, + { ixSE_CAC_CNTL, 0xFFFFFFFF, 0, 0x02060021 }, + { ixSE_CAC_CNTL, 0xFFFFFFFF, 0, 0x02860021 }, + { ixSE_CAC_CNTL, 0xFFFFFFFF, 0, 0x03060021 }, + { ixSE_CAC_CNTL, 0xFFFFFFFF, 0, 0x03860021 }, + { ixSE_CAC_CNTL, 0xFFFFFFFF, 0, 0x04060021 }, + /* TD */ + { ixSE_CAC_CNTL, 0xFFFFFFFF, 0, 0x000E0020 }, + { ixSE_CAC_CNTL, 0xFFFFFFFF, 0, 0x008E0020 }, + { ixSE_CAC_CNTL, 0xFFFFFFFF, 0, 0x010E0020 }, + { ixSE_CAC_CNTL, 0xFFFFFFFF, 0, 0x018E0020 }, + { ixSE_CAC_CNTL, 0xFFFFFFFF, 0, 0x020E0020 }, + { ixSE_CAC_CNTL, 0xFFFFFFFF, 0, 0x028E0020 }, + /* TCP */ + { ixSE_CAC_CNTL, 0xFFFFFFFF, 0, 0x001c0020 }, + { ixSE_CAC_CNTL, 0xFFFFFFFF, 0, 0x009c0020 }, + { ixSE_CAC_CNTL, 0xFFFFFFFF, 0, 0x011c0020 }, + { ixSE_CAC_CNTL, 0xFFFFFFFF, 0, 0x019c0020 }, + { ixSE_CAC_CNTL, 0xFFFFFFFF, 0, 0x021c0020 }, + /* DB */ + { ixSE_CAC_CNTL, 0xFFFFFFFF, 0, 0x00200008 }, + { ixSE_CAC_CNTL, 0xFFFFFFFF, 0, 0x00820008 }, + { ixSE_CAC_CNTL, 0xFFFFFFFF, 0, 0x01020008 }, + { ixSE_CAC_CNTL, 0xFFFFFFFF, 0, 0x01820008 }, + + { 0xFFFFFFFF } /* End of list */ +}; + + +static const struct vega10_didt_config_reg SEEDCStallPatternConfig_Vega10[] = +{ +/* --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- + * Offset Mask Shift Value + * --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- + */ + /* SQ */ + { ixDIDT_SQ_EDC_STALL_PATTERN_1_2, 0xFFFFFFFF, 0, 0x00030001 }, + { ixDIDT_SQ_EDC_STALL_PATTERN_3_4, 0xFFFFFFFF, 0, 0x000F0007 }, + { ixDIDT_SQ_EDC_STALL_PATTERN_5_6, 0xFFFFFFFF, 0, 0x003F001F }, + { ixDIDT_SQ_EDC_STALL_PATTERN_7, 0xFFFFFFFF, 0, 0x0000007F }, + /* TD */ + { ixDIDT_TD_EDC_STALL_PATTERN_1_2, 0xFFFFFFFF, 0, 0x00000000 }, + { ixDIDT_TD_EDC_STALL_PATTERN_3_4, 0xFFFFFFFF, 0, 0x00000000 }, + { ixDIDT_TD_EDC_STALL_PATTERN_5_6, 0xFFFFFFFF, 0, 0x00000000 }, + { ixDIDT_TD_EDC_STALL_PATTERN_7, 0xFFFFFFFF, 0, 0x00000000 }, + /* TCP */ + { ixDIDT_TCP_EDC_STALL_PATTERN_1_2, 0xFFFFFFFF, 0, 0x00000000 }, + { ixDIDT_TCP_EDC_STALL_PATTERN_3_4, 0xFFFFFFFF, 0, 0x00000000 }, + { ixDIDT_TCP_EDC_STALL_PATTERN_5_6, 0xFFFFFFFF, 0, 0x00000000 }, + { ixDIDT_TCP_EDC_STALL_PATTERN_7, 0xFFFFFFFF, 0, 0x00000000 }, + /* DB */ + { ixDIDT_DB_EDC_STALL_PATTERN_1_2, 0xFFFFFFFF, 0, 0x00000000 }, + { ixDIDT_DB_EDC_STALL_PATTERN_3_4, 0xFFFFFFFF, 0, 0x00000000 }, + { ixDIDT_DB_EDC_STALL_PATTERN_5_6, 0xFFFFFFFF, 0, 0x00000000 }, + { ixDIDT_DB_EDC_STALL_PATTERN_7, 0xFFFFFFFF, 0, 0x00000000 }, + + { 0xFFFFFFFF } /* End of list */ +}; + +static const struct vega10_didt_config_reg SEEDCForceStallPatternConfig_Vega10[] = +{ +/* --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- + * Offset Mask Shift Value + * --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- + */ + /* SQ */ + { ixDIDT_SQ_EDC_STALL_PATTERN_1_2, 0xFFFFFFFF, 0, 0x00000015 }, + { ixDIDT_SQ_EDC_STALL_PATTERN_3_4, 0xFFFFFFFF, 0, 0x00000000 }, + { ixDIDT_SQ_EDC_STALL_PATTERN_5_6, 0xFFFFFFFF, 0, 0x00000000 }, + { ixDIDT_SQ_EDC_STALL_PATTERN_7, 0xFFFFFFFF, 0, 0x00000000 }, + /* TD */ + { ixDIDT_TD_EDC_STALL_PATTERN_1_2, 0xFFFFFFFF, 0, 0x00000015 }, + { ixDIDT_TD_EDC_STALL_PATTERN_3_4, 0xFFFFFFFF, 0, 0x00000000 }, + { ixDIDT_TD_EDC_STALL_PATTERN_5_6, 0xFFFFFFFF, 0, 0x00000000 }, + { ixDIDT_TD_EDC_STALL_PATTERN_7, 0xFFFFFFFF, 0, 0x00000000 }, + + { 0xFFFFFFFF } /* End of list */ +}; + +static const struct vega10_didt_config_reg SEEDCStallDelayConfig_Vega10[] = +{ +/* --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- + * Offset Mask Shift Value + * --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- + */ + /* SQ */ + { ixDIDT_SQ_EDC_STALL_DELAY_1, 0xFFFFFFFF, 0, 0x00000000 }, + { ixDIDT_SQ_EDC_STALL_DELAY_2, 0xFFFFFFFF, 0, 0x00000000 }, + { ixDIDT_SQ_EDC_STALL_DELAY_3, 0xFFFFFFFF, 0, 0x00000000 }, + { ixDIDT_SQ_EDC_STALL_DELAY_4, 0xFFFFFFFF, 0, 0x00000000 }, + /* TD */ + { ixDIDT_TD_EDC_STALL_DELAY_1, 0xFFFFFFFF, 0, 0x00000000 }, + { ixDIDT_TD_EDC_STALL_DELAY_2, 0xFFFFFFFF, 0, 0x00000000 }, + { ixDIDT_TD_EDC_STALL_DELAY_3, 0xFFFFFFFF, 0, 0x00000000 }, + { ixDIDT_TD_EDC_STALL_DELAY_4, 0xFFFFFFFF, 0, 0x00000000 }, + /* TCP */ + { ixDIDT_TCP_EDC_STALL_DELAY_1, 0xFFFFFFFF, 0, 0x00000000 }, + { ixDIDT_TCP_EDC_STALL_DELAY_2, 0xFFFFFFFF, 0, 0x00000000 }, + { ixDIDT_TCP_EDC_STALL_DELAY_3, 0xFFFFFFFF, 0, 0x00000000 }, + { ixDIDT_TCP_EDC_STALL_DELAY_4, 0xFFFFFFFF, 0, 0x00000000 }, + /* DB */ + { ixDIDT_DB_EDC_STALL_DELAY_1, 0xFFFFFFFF, 0, 0x00000000 }, + + { 0xFFFFFFFF } /* End of list */ +}; + +static const struct vega10_didt_config_reg SEEDCThresholdConfig_Vega10[] = +{ +/* --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- + * Offset Mask Shift Value + * --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- + */ + { ixDIDT_SQ_EDC_THRESHOLD, 0xFFFFFFFF, 0, 0x0000010E }, + { ixDIDT_TD_EDC_THRESHOLD, 0xFFFFFFFF, 0, 0xFFFFFFFF }, + { ixDIDT_TCP_EDC_THRESHOLD, 0xFFFFFFFF, 0, 0xFFFFFFFF }, + { ixDIDT_DB_EDC_THRESHOLD, 0xFFFFFFFF, 0, 0xFFFFFFFF }, + + { 0xFFFFFFFF } /* End of list */ +}; + +static const struct vega10_didt_config_reg SEEDCCtrlResetConfig_Vega10[] = +{ +/* --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- + * Offset Mask Shift Value + * --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- + */ + /* SQ */ + { ixDIDT_SQ_EDC_CTRL, DIDT_SQ_EDC_CTRL__EDC_EN_MASK, DIDT_SQ_EDC_CTRL__EDC_EN__SHIFT, 0x0000 }, + { ixDIDT_SQ_EDC_CTRL, DIDT_SQ_EDC_CTRL__EDC_SW_RST_MASK, DIDT_SQ_EDC_CTRL__EDC_SW_RST__SHIFT, 0x0001 }, + { ixDIDT_SQ_EDC_CTRL, DIDT_SQ_EDC_CTRL__EDC_CLK_EN_OVERRIDE_MASK, DIDT_SQ_EDC_CTRL__EDC_CLK_EN_OVERRIDE__SHIFT, 0x0000 }, + { ixDIDT_SQ_EDC_CTRL, DIDT_SQ_EDC_CTRL__EDC_FORCE_STALL_MASK, DIDT_SQ_EDC_CTRL__EDC_FORCE_STALL__SHIFT, 0x0000 }, + { ixDIDT_SQ_EDC_CTRL, DIDT_SQ_EDC_CTRL__EDC_TRIGGER_THROTTLE_LOWBIT_MASK, DIDT_SQ_EDC_CTRL__EDC_TRIGGER_THROTTLE_LOWBIT__SHIFT, 0x0000 }, + { ixDIDT_SQ_EDC_CTRL, DIDT_SQ_EDC_CTRL__EDC_STALL_PATTERN_BIT_NUMS_MASK, DIDT_SQ_EDC_CTRL__EDC_STALL_PATTERN_BIT_NUMS__SHIFT, 0x0000 }, + { ixDIDT_SQ_EDC_CTRL, DIDT_SQ_EDC_CTRL__EDC_ALLOW_WRITE_PWRDELTA_MASK, DIDT_SQ_EDC_CTRL__EDC_ALLOW_WRITE_PWRDELTA__SHIFT, 0x0000 }, + { ixDIDT_SQ_EDC_CTRL, DIDT_SQ_EDC_CTRL__GC_EDC_EN_MASK, DIDT_SQ_EDC_CTRL__GC_EDC_EN__SHIFT, 0x0000 }, + { ixDIDT_SQ_EDC_CTRL, DIDT_SQ_EDC_CTRL__GC_EDC_STALL_POLICY_MASK, DIDT_SQ_EDC_CTRL__GC_EDC_STALL_POLICY__SHIFT, 0x0000 }, + { ixDIDT_SQ_EDC_CTRL, DIDT_SQ_EDC_CTRL__GC_EDC_LEVEL_COMB_EN_MASK, DIDT_SQ_EDC_CTRL__GC_EDC_LEVEL_COMB_EN__SHIFT, 0x0000 }, + { ixDIDT_SQ_EDC_CTRL, DIDT_SQ_EDC_CTRL__SE_EDC_LEVEL_COMB_EN_MASK, DIDT_SQ_EDC_CTRL__SE_EDC_LEVEL_COMB_EN__SHIFT, 0x0000 }, + + { 0xFFFFFFFF } /* End of list */ +}; + +static const struct vega10_didt_config_reg SEEDCCtrlConfig_Vega10[] = +{ +/* --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- + * Offset Mask Shift Value + * --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- + */ + /* SQ */ + { ixDIDT_SQ_EDC_CTRL, DIDT_SQ_EDC_CTRL__EDC_EN_MASK, DIDT_SQ_EDC_CTRL__EDC_EN__SHIFT, 0x0001 }, + { ixDIDT_SQ_EDC_CTRL, DIDT_SQ_EDC_CTRL__EDC_SW_RST_MASK, DIDT_SQ_EDC_CTRL__EDC_SW_RST__SHIFT, 0x0000 }, + { ixDIDT_SQ_EDC_CTRL, DIDT_SQ_EDC_CTRL__EDC_CLK_EN_OVERRIDE_MASK, DIDT_SQ_EDC_CTRL__EDC_CLK_EN_OVERRIDE__SHIFT, 0x0000 }, + { ixDIDT_SQ_EDC_CTRL, DIDT_SQ_EDC_CTRL__EDC_FORCE_STALL_MASK, DIDT_SQ_EDC_CTRL__EDC_FORCE_STALL__SHIFT, 0x0000 }, + { ixDIDT_SQ_EDC_CTRL, DIDT_SQ_EDC_CTRL__EDC_TRIGGER_THROTTLE_LOWBIT_MASK, DIDT_SQ_EDC_CTRL__EDC_TRIGGER_THROTTLE_LOWBIT__SHIFT, 0x0004 }, + { ixDIDT_SQ_EDC_CTRL, DIDT_SQ_EDC_CTRL__EDC_STALL_PATTERN_BIT_NUMS_MASK, DIDT_SQ_EDC_CTRL__EDC_STALL_PATTERN_BIT_NUMS__SHIFT, 0x0006 }, + { ixDIDT_SQ_EDC_CTRL, DIDT_SQ_EDC_CTRL__EDC_ALLOW_WRITE_PWRDELTA_MASK, DIDT_SQ_EDC_CTRL__EDC_ALLOW_WRITE_PWRDELTA__SHIFT, 0x0000 }, + { ixDIDT_SQ_EDC_CTRL, DIDT_SQ_EDC_CTRL__GC_EDC_EN_MASK, DIDT_SQ_EDC_CTRL__GC_EDC_EN__SHIFT, 0x0000 }, + { ixDIDT_SQ_EDC_CTRL, DIDT_SQ_EDC_CTRL__GC_EDC_STALL_POLICY_MASK, DIDT_SQ_EDC_CTRL__GC_EDC_STALL_POLICY__SHIFT, 0x0000 }, + { ixDIDT_SQ_EDC_CTRL, DIDT_SQ_EDC_CTRL__GC_EDC_LEVEL_COMB_EN_MASK, DIDT_SQ_EDC_CTRL__GC_EDC_LEVEL_COMB_EN__SHIFT, 0x0001 }, + { ixDIDT_SQ_EDC_CTRL, DIDT_SQ_EDC_CTRL__SE_EDC_LEVEL_COMB_EN_MASK, DIDT_SQ_EDC_CTRL__SE_EDC_LEVEL_COMB_EN__SHIFT, 0x0000 }, + + { 0xFFFFFFFF } /* End of list */ +}; + +static const struct vega10_didt_config_reg SEEDCCtrlForceStallConfig_Vega10[] = +{ +/* --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- + * Offset Mask Shift Value + * --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- + */ + /* SQ */ + { ixDIDT_SQ_EDC_CTRL, DIDT_SQ_EDC_CTRL__EDC_EN_MASK, DIDT_SQ_EDC_CTRL__EDC_EN__SHIFT, 0x0001 }, + { ixDIDT_SQ_EDC_CTRL, DIDT_SQ_EDC_CTRL__EDC_SW_RST_MASK, DIDT_SQ_EDC_CTRL__EDC_SW_RST__SHIFT, 0x0000 }, + { ixDIDT_SQ_EDC_CTRL, DIDT_SQ_EDC_CTRL__EDC_CLK_EN_OVERRIDE_MASK, DIDT_SQ_EDC_CTRL__EDC_CLK_EN_OVERRIDE__SHIFT, 0x0000 }, + { ixDIDT_SQ_EDC_CTRL, DIDT_SQ_EDC_CTRL__EDC_FORCE_STALL_MASK, DIDT_SQ_EDC_CTRL__EDC_FORCE_STALL__SHIFT, 0x0001 }, + { ixDIDT_SQ_EDC_CTRL, DIDT_SQ_EDC_CTRL__EDC_TRIGGER_THROTTLE_LOWBIT_MASK, DIDT_SQ_EDC_CTRL__EDC_TRIGGER_THROTTLE_LOWBIT__SHIFT, 0x0001 }, + { ixDIDT_SQ_EDC_CTRL, DIDT_SQ_EDC_CTRL__EDC_STALL_PATTERN_BIT_NUMS_MASK, DIDT_SQ_EDC_CTRL__EDC_STALL_PATTERN_BIT_NUMS__SHIFT, 0x000C }, + { ixDIDT_SQ_EDC_CTRL, DIDT_SQ_EDC_CTRL__EDC_ALLOW_WRITE_PWRDELTA_MASK, DIDT_SQ_EDC_CTRL__EDC_ALLOW_WRITE_PWRDELTA__SHIFT, 0x0000 }, + { ixDIDT_SQ_EDC_CTRL, DIDT_SQ_EDC_CTRL__GC_EDC_EN_MASK, DIDT_SQ_EDC_CTRL__GC_EDC_EN__SHIFT, 0x0000 }, + { ixDIDT_SQ_EDC_CTRL, DIDT_SQ_EDC_CTRL__GC_EDC_STALL_POLICY_MASK, DIDT_SQ_EDC_CTRL__GC_EDC_STALL_POLICY__SHIFT, 0x0000 }, + { ixDIDT_SQ_EDC_CTRL, DIDT_SQ_EDC_CTRL__GC_EDC_LEVEL_COMB_EN_MASK, DIDT_SQ_EDC_CTRL__GC_EDC_LEVEL_COMB_EN__SHIFT, 0x0000 }, + { ixDIDT_SQ_EDC_CTRL, DIDT_SQ_EDC_CTRL__SE_EDC_LEVEL_COMB_EN_MASK, DIDT_SQ_EDC_CTRL__SE_EDC_LEVEL_COMB_EN__SHIFT, 0x0001 }, + + /* TD */ + { ixDIDT_TD_EDC_CTRL, DIDT_TD_EDC_CTRL__EDC_EN_MASK, DIDT_TD_EDC_CTRL__EDC_EN__SHIFT, 0x0001 }, + { ixDIDT_TD_EDC_CTRL, DIDT_TD_EDC_CTRL__EDC_SW_RST_MASK, DIDT_TD_EDC_CTRL__EDC_SW_RST__SHIFT, 0x0000 }, + { ixDIDT_TD_EDC_CTRL, DIDT_TD_EDC_CTRL__EDC_CLK_EN_OVERRIDE_MASK, DIDT_TD_EDC_CTRL__EDC_CLK_EN_OVERRIDE__SHIFT, 0x0000 }, + { ixDIDT_TD_EDC_CTRL, DIDT_TD_EDC_CTRL__EDC_FORCE_STALL_MASK, DIDT_TD_EDC_CTRL__EDC_FORCE_STALL__SHIFT, 0x0001 }, + { ixDIDT_TD_EDC_CTRL, DIDT_TD_EDC_CTRL__EDC_TRIGGER_THROTTLE_LOWBIT_MASK, DIDT_TD_EDC_CTRL__EDC_TRIGGER_THROTTLE_LOWBIT__SHIFT, 0x0001 }, + { ixDIDT_TD_EDC_CTRL, DIDT_TD_EDC_CTRL__EDC_STALL_PATTERN_BIT_NUMS_MASK, DIDT_TD_EDC_CTRL__EDC_STALL_PATTERN_BIT_NUMS__SHIFT, 0x000E }, + { ixDIDT_TD_EDC_CTRL, DIDT_TD_EDC_CTRL__EDC_ALLOW_WRITE_PWRDELTA_MASK, DIDT_TD_EDC_CTRL__EDC_ALLOW_WRITE_PWRDELTA__SHIFT, 0x0000 }, + { ixDIDT_TD_EDC_CTRL, DIDT_TD_EDC_CTRL__GC_EDC_EN_MASK, DIDT_TD_EDC_CTRL__GC_EDC_EN__SHIFT, 0x0000 }, + { ixDIDT_TD_EDC_CTRL, DIDT_TD_EDC_CTRL__GC_EDC_STALL_POLICY_MASK, DIDT_TD_EDC_CTRL__GC_EDC_STALL_POLICY__SHIFT, 0x0000 }, + { ixDIDT_TD_EDC_CTRL, DIDT_TD_EDC_CTRL__GC_EDC_LEVEL_COMB_EN_MASK, DIDT_TD_EDC_CTRL__GC_EDC_LEVEL_COMB_EN__SHIFT, 0x0000 }, + { ixDIDT_TD_EDC_CTRL, DIDT_TD_EDC_CTRL__SE_EDC_LEVEL_COMB_EN_MASK, DIDT_TD_EDC_CTRL__SE_EDC_LEVEL_COMB_EN__SHIFT, 0x0001 }, + + { 0xFFFFFFFF } /* End of list */ +}; + +static const struct vega10_didt_config_reg GCDiDtDroopCtrlConfig_vega10[] = +{ +/* --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- + * Offset Mask Shift Value + * --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- + */ + { mmGC_DIDT_DROOP_CTRL, GC_DIDT_DROOP_CTRL__DIDT_DROOP_LEVEL_EN_MASK, GC_DIDT_DROOP_CTRL__DIDT_DROOP_LEVEL_EN__SHIFT, 0x0000 }, + { mmGC_DIDT_DROOP_CTRL, GC_DIDT_DROOP_CTRL__DIDT_DROOP_THRESHOLD_MASK, GC_DIDT_DROOP_CTRL__DIDT_DROOP_THRESHOLD__SHIFT, 0x0000 }, + { mmGC_DIDT_DROOP_CTRL, GC_DIDT_DROOP_CTRL__DIDT_DROOP_LEVEL_INDEX_MASK, GC_DIDT_DROOP_CTRL__DIDT_DROOP_LEVEL_INDEX__SHIFT, 0x0000 }, + { mmGC_DIDT_DROOP_CTRL, GC_DIDT_DROOP_CTRL__DIDT_LEVEL_SEL_MASK, GC_DIDT_DROOP_CTRL__DIDT_LEVEL_SEL__SHIFT, 0x0000 }, + { mmGC_DIDT_DROOP_CTRL, GC_DIDT_DROOP_CTRL__DIDT_DROOP_LEVEL_OVERFLOW_MASK, GC_DIDT_DROOP_CTRL__DIDT_DROOP_LEVEL_OVERFLOW__SHIFT, 0x0000 }, + + { 0xFFFFFFFF } /* End of list */ +}; + +static const struct vega10_didt_config_reg GCDiDtCtrl0Config_vega10[] = +{ +/* --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- + * Offset Mask Shift Value + * --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- + */ + { mmGC_DIDT_CTRL0, GC_DIDT_CTRL0__DIDT_CTRL_EN_MASK, GC_DIDT_CTRL0__DIDT_CTRL_EN__SHIFT, 0x0000 }, + { mmGC_DIDT_CTRL0, GC_DIDT_CTRL0__PHASE_OFFSET_MASK, GC_DIDT_CTRL0__PHASE_OFFSET__SHIFT, 0x0000 }, + { mmGC_DIDT_CTRL0, GC_DIDT_CTRL0__DIDT_SW_RST_MASK, GC_DIDT_CTRL0__DIDT_SW_RST__SHIFT, 0x0000 }, + { mmGC_DIDT_CTRL0, GC_DIDT_CTRL0__DIDT_CLK_EN_OVERRIDE_MASK, GC_DIDT_CTRL0__DIDT_CLK_EN_OVERRIDE__SHIFT, 0x0000 }, + { mmGC_DIDT_CTRL0, GC_DIDT_CTRL0__DIDT_TRIGGER_THROTTLE_LOWBIT_MASK, GC_DIDT_CTRL0__DIDT_TRIGGER_THROTTLE_LOWBIT__SHIFT, 0x0000 }, + { 0xFFFFFFFF } /* End of list */ +}; + + +static const struct vega10_didt_config_reg PSMSEEDCStallPatternConfig_Vega10[] = +{ +/* --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- + * Offset Mask Shift Value + * --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- + */ + /* SQ EDC STALL PATTERNs */ + { ixDIDT_SQ_EDC_STALL_PATTERN_1_2, DIDT_SQ_EDC_STALL_PATTERN_1_2__EDC_STALL_PATTERN_1_MASK, DIDT_SQ_EDC_STALL_PATTERN_1_2__EDC_STALL_PATTERN_1__SHIFT, 0x0101 }, + { ixDIDT_SQ_EDC_STALL_PATTERN_1_2, DIDT_SQ_EDC_STALL_PATTERN_1_2__EDC_STALL_PATTERN_2_MASK, DIDT_SQ_EDC_STALL_PATTERN_1_2__EDC_STALL_PATTERN_2__SHIFT, 0x0101 }, + { ixDIDT_SQ_EDC_STALL_PATTERN_3_4, DIDT_SQ_EDC_STALL_PATTERN_3_4__EDC_STALL_PATTERN_3_MASK, DIDT_SQ_EDC_STALL_PATTERN_3_4__EDC_STALL_PATTERN_3__SHIFT, 0x1111 }, + { ixDIDT_SQ_EDC_STALL_PATTERN_3_4, DIDT_SQ_EDC_STALL_PATTERN_3_4__EDC_STALL_PATTERN_4_MASK, DIDT_SQ_EDC_STALL_PATTERN_3_4__EDC_STALL_PATTERN_4__SHIFT, 0x1111 }, + + { ixDIDT_SQ_EDC_STALL_PATTERN_5_6, DIDT_SQ_EDC_STALL_PATTERN_5_6__EDC_STALL_PATTERN_5_MASK, DIDT_SQ_EDC_STALL_PATTERN_5_6__EDC_STALL_PATTERN_5__SHIFT, 0x1515 }, + { ixDIDT_SQ_EDC_STALL_PATTERN_5_6, DIDT_SQ_EDC_STALL_PATTERN_5_6__EDC_STALL_PATTERN_6_MASK, DIDT_SQ_EDC_STALL_PATTERN_5_6__EDC_STALL_PATTERN_6__SHIFT, 0x1515 }, + + { ixDIDT_SQ_EDC_STALL_PATTERN_7, DIDT_SQ_EDC_STALL_PATTERN_7__EDC_STALL_PATTERN_7_MASK, DIDT_SQ_EDC_STALL_PATTERN_7__EDC_STALL_PATTERN_7__SHIFT, 0x5555 }, + + { 0xFFFFFFFF } /* End of list */ +}; + +static const struct vega10_didt_config_reg PSMSEEDCStallDelayConfig_Vega10[] = +{ +/* --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- + * Offset Mask Shift Value + * --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- + */ + /* SQ EDC STALL DELAYs */ + { ixDIDT_SQ_EDC_STALL_DELAY_1, DIDT_SQ_EDC_STALL_DELAY_1__EDC_STALL_DELAY_SQ0_MASK, DIDT_SQ_EDC_STALL_DELAY_1__EDC_STALL_DELAY_SQ0__SHIFT, 0x0000 }, + { ixDIDT_SQ_EDC_STALL_DELAY_1, DIDT_SQ_EDC_STALL_DELAY_1__EDC_STALL_DELAY_SQ1_MASK, DIDT_SQ_EDC_STALL_DELAY_1__EDC_STALL_DELAY_SQ1__SHIFT, 0x0000 }, + { ixDIDT_SQ_EDC_STALL_DELAY_1, DIDT_SQ_EDC_STALL_DELAY_1__EDC_STALL_DELAY_SQ2_MASK, DIDT_SQ_EDC_STALL_DELAY_1__EDC_STALL_DELAY_SQ2__SHIFT, 0x0000 }, + { ixDIDT_SQ_EDC_STALL_DELAY_1, DIDT_SQ_EDC_STALL_DELAY_1__EDC_STALL_DELAY_SQ3_MASK, DIDT_SQ_EDC_STALL_DELAY_1__EDC_STALL_DELAY_SQ3__SHIFT, 0x0000 }, + + { ixDIDT_SQ_EDC_STALL_DELAY_2, DIDT_SQ_EDC_STALL_DELAY_2__EDC_STALL_DELAY_SQ4_MASK, DIDT_SQ_EDC_STALL_DELAY_2__EDC_STALL_DELAY_SQ4__SHIFT, 0x0000 }, + { ixDIDT_SQ_EDC_STALL_DELAY_2, DIDT_SQ_EDC_STALL_DELAY_2__EDC_STALL_DELAY_SQ5_MASK, DIDT_SQ_EDC_STALL_DELAY_2__EDC_STALL_DELAY_SQ5__SHIFT, 0x0000 }, + { ixDIDT_SQ_EDC_STALL_DELAY_2, DIDT_SQ_EDC_STALL_DELAY_2__EDC_STALL_DELAY_SQ6_MASK, DIDT_SQ_EDC_STALL_DELAY_2__EDC_STALL_DELAY_SQ6__SHIFT, 0x0000 }, + { ixDIDT_SQ_EDC_STALL_DELAY_2, DIDT_SQ_EDC_STALL_DELAY_2__EDC_STALL_DELAY_SQ7_MASK, DIDT_SQ_EDC_STALL_DELAY_2__EDC_STALL_DELAY_SQ7__SHIFT, 0x0000 }, + + { ixDIDT_SQ_EDC_STALL_DELAY_3, DIDT_SQ_EDC_STALL_DELAY_3__EDC_STALL_DELAY_SQ8_MASK, DIDT_SQ_EDC_STALL_DELAY_3__EDC_STALL_DELAY_SQ8__SHIFT, 0x0000 }, + { ixDIDT_SQ_EDC_STALL_DELAY_3, DIDT_SQ_EDC_STALL_DELAY_3__EDC_STALL_DELAY_SQ9_MASK, DIDT_SQ_EDC_STALL_DELAY_3__EDC_STALL_DELAY_SQ9__SHIFT, 0x0000 }, + { ixDIDT_SQ_EDC_STALL_DELAY_3, DIDT_SQ_EDC_STALL_DELAY_3__EDC_STALL_DELAY_SQ10_MASK, DIDT_SQ_EDC_STALL_DELAY_3__EDC_STALL_DELAY_SQ10__SHIFT, 0x0000 }, + { ixDIDT_SQ_EDC_STALL_DELAY_3, DIDT_SQ_EDC_STALL_DELAY_3__EDC_STALL_DELAY_SQ11_MASK, DIDT_SQ_EDC_STALL_DELAY_3__EDC_STALL_DELAY_SQ11__SHIFT, 0x0000 }, + + { ixDIDT_SQ_EDC_STALL_DELAY_4, DIDT_SQ_EDC_STALL_DELAY_4__EDC_STALL_DELAY_SQ12_MASK, DIDT_SQ_EDC_STALL_DELAY_4__EDC_STALL_DELAY_SQ12__SHIFT, 0x0000 }, + { ixDIDT_SQ_EDC_STALL_DELAY_4, DIDT_SQ_EDC_STALL_DELAY_4__EDC_STALL_DELAY_SQ12_MASK, DIDT_SQ_EDC_STALL_DELAY_4__EDC_STALL_DELAY_SQ13__SHIFT, 0x0000 }, + { ixDIDT_SQ_EDC_STALL_DELAY_4, DIDT_SQ_EDC_STALL_DELAY_4__EDC_STALL_DELAY_SQ14_MASK, DIDT_SQ_EDC_STALL_DELAY_4__EDC_STALL_DELAY_SQ14__SHIFT, 0x0000 }, + { ixDIDT_SQ_EDC_STALL_DELAY_4, DIDT_SQ_EDC_STALL_DELAY_4__EDC_STALL_DELAY_SQ15_MASK, DIDT_SQ_EDC_STALL_DELAY_4__EDC_STALL_DELAY_SQ15__SHIFT, 0x0000 }, + + { 0xFFFFFFFF } /* End of list */ +}; + +static const struct vega10_didt_config_reg PSMSEEDCThresholdConfig_Vega10[] = +{ +/* --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- + * Offset Mask Shift Value + * --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- + */ + /* SQ EDC THRESHOLD */ + { ixDIDT_SQ_EDC_THRESHOLD, DIDT_SQ_EDC_THRESHOLD__EDC_THRESHOLD_MASK, DIDT_SQ_EDC_THRESHOLD__EDC_THRESHOLD__SHIFT, 0x0000 }, + + { 0xFFFFFFFF } /* End of list */ +}; + +static const struct vega10_didt_config_reg PSMSEEDCCtrlResetConfig_Vega10[] = +{ +/* --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- + * Offset Mask Shift Value + * --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- + */ + /* SQ EDC CTRL */ + { ixDIDT_SQ_EDC_CTRL, DIDT_SQ_EDC_CTRL__EDC_EN_MASK, DIDT_SQ_EDC_CTRL__EDC_EN__SHIFT, 0x0000 }, + { ixDIDT_SQ_EDC_CTRL, DIDT_SQ_EDC_CTRL__EDC_SW_RST_MASK, DIDT_SQ_EDC_CTRL__EDC_SW_RST__SHIFT, 0x0001 }, + { ixDIDT_SQ_EDC_CTRL, DIDT_SQ_EDC_CTRL__EDC_CLK_EN_OVERRIDE_MASK, DIDT_SQ_EDC_CTRL__EDC_CLK_EN_OVERRIDE__SHIFT, 0x0000 }, + { ixDIDT_SQ_EDC_CTRL, DIDT_SQ_EDC_CTRL__EDC_FORCE_STALL_MASK, DIDT_SQ_EDC_CTRL__EDC_FORCE_STALL__SHIFT, 0x0000 }, + { ixDIDT_SQ_EDC_CTRL, DIDT_SQ_EDC_CTRL__EDC_TRIGGER_THROTTLE_LOWBIT_MASK, DIDT_SQ_EDC_CTRL__EDC_TRIGGER_THROTTLE_LOWBIT__SHIFT, 0x0000 }, + { ixDIDT_SQ_EDC_CTRL, DIDT_SQ_EDC_CTRL__EDC_STALL_PATTERN_BIT_NUMS_MASK, DIDT_SQ_EDC_CTRL__EDC_STALL_PATTERN_BIT_NUMS__SHIFT, 0x0000 }, + { ixDIDT_SQ_EDC_CTRL, DIDT_SQ_EDC_CTRL__EDC_ALLOW_WRITE_PWRDELTA_MASK, DIDT_SQ_EDC_CTRL__EDC_ALLOW_WRITE_PWRDELTA__SHIFT, 0x0000 }, + { ixDIDT_SQ_EDC_CTRL, DIDT_SQ_EDC_CTRL__GC_EDC_EN_MASK, DIDT_SQ_EDC_CTRL__GC_EDC_EN__SHIFT, 0x0000 }, + { ixDIDT_SQ_EDC_CTRL, DIDT_SQ_EDC_CTRL__GC_EDC_STALL_POLICY_MASK, DIDT_SQ_EDC_CTRL__GC_EDC_STALL_POLICY__SHIFT, 0x0000 }, + { ixDIDT_SQ_EDC_CTRL, DIDT_SQ_EDC_CTRL__GC_EDC_LEVEL_COMB_EN_MASK, DIDT_SQ_EDC_CTRL__GC_EDC_LEVEL_COMB_EN__SHIFT, 0x0000 }, + { ixDIDT_SQ_EDC_CTRL, DIDT_SQ_EDC_CTRL__SE_EDC_LEVEL_COMB_EN_MASK, DIDT_SQ_EDC_CTRL__SE_EDC_LEVEL_COMB_EN__SHIFT, 0x0000 }, + + { 0xFFFFFFFF } /* End of list */ +}; + +static const struct vega10_didt_config_reg PSMSEEDCCtrlConfig_Vega10[] = +{ +/* --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- + * Offset Mask Shift Value + * --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- + */ + /* SQ EDC CTRL */ + { ixDIDT_SQ_EDC_CTRL, DIDT_SQ_EDC_CTRL__EDC_EN_MASK, DIDT_SQ_EDC_CTRL__EDC_EN__SHIFT, 0x0001 }, + { ixDIDT_SQ_EDC_CTRL, DIDT_SQ_EDC_CTRL__EDC_SW_RST_MASK, DIDT_SQ_EDC_CTRL__EDC_SW_RST__SHIFT, 0x0000 }, + { ixDIDT_SQ_EDC_CTRL, DIDT_SQ_EDC_CTRL__EDC_CLK_EN_OVERRIDE_MASK, DIDT_SQ_EDC_CTRL__EDC_CLK_EN_OVERRIDE__SHIFT, 0x0000 }, + { ixDIDT_SQ_EDC_CTRL, DIDT_SQ_EDC_CTRL__EDC_FORCE_STALL_MASK, DIDT_SQ_EDC_CTRL__EDC_FORCE_STALL__SHIFT, 0x0000 }, + { ixDIDT_SQ_EDC_CTRL, DIDT_SQ_EDC_CTRL__EDC_TRIGGER_THROTTLE_LOWBIT_MASK, DIDT_SQ_EDC_CTRL__EDC_TRIGGER_THROTTLE_LOWBIT__SHIFT, 0x0000 }, + { ixDIDT_SQ_EDC_CTRL, DIDT_SQ_EDC_CTRL__EDC_STALL_PATTERN_BIT_NUMS_MASK, DIDT_SQ_EDC_CTRL__EDC_STALL_PATTERN_BIT_NUMS__SHIFT, 0x000E }, + { ixDIDT_SQ_EDC_CTRL, DIDT_SQ_EDC_CTRL__EDC_ALLOW_WRITE_PWRDELTA_MASK, DIDT_SQ_EDC_CTRL__EDC_ALLOW_WRITE_PWRDELTA__SHIFT, 0x0000 }, + { ixDIDT_SQ_EDC_CTRL, DIDT_SQ_EDC_CTRL__GC_EDC_EN_MASK, DIDT_SQ_EDC_CTRL__GC_EDC_EN__SHIFT, 0x0001 }, + { ixDIDT_SQ_EDC_CTRL, DIDT_SQ_EDC_CTRL__GC_EDC_STALL_POLICY_MASK, DIDT_SQ_EDC_CTRL__GC_EDC_STALL_POLICY__SHIFT, 0x0003 }, + { ixDIDT_SQ_EDC_CTRL, DIDT_SQ_EDC_CTRL__GC_EDC_LEVEL_COMB_EN_MASK, DIDT_SQ_EDC_CTRL__GC_EDC_LEVEL_COMB_EN__SHIFT, 0x0001 }, + { ixDIDT_SQ_EDC_CTRL, DIDT_SQ_EDC_CTRL__SE_EDC_LEVEL_COMB_EN_MASK, DIDT_SQ_EDC_CTRL__SE_EDC_LEVEL_COMB_EN__SHIFT, 0x0000 }, + + { 0xFFFFFFFF } /* End of list */ +}; + +static const struct vega10_didt_config_reg PSMGCEDCThresholdConfig_vega10[] = +{ +/* --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- + * Offset Mask Shift Value + * --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- + */ + { mmGC_EDC_THRESHOLD, GC_EDC_THRESHOLD__EDC_THRESHOLD_MASK, GC_EDC_THRESHOLD__EDC_THRESHOLD__SHIFT, 0x0000000 }, + + { 0xFFFFFFFF } /* End of list */ +}; + +static const struct vega10_didt_config_reg PSMGCEDCDroopCtrlConfig_vega10[] = +{ +/* --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- + * Offset Mask Shift Value + * --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- + */ + { mmGC_EDC_DROOP_CTRL, GC_EDC_DROOP_CTRL__EDC_DROOP_LEVEL_EN_MASK, GC_EDC_DROOP_CTRL__EDC_DROOP_LEVEL_EN__SHIFT, 0x0001 }, + { mmGC_EDC_DROOP_CTRL, GC_EDC_DROOP_CTRL__EDC_DROOP_THRESHOLD_MASK, GC_EDC_DROOP_CTRL__EDC_DROOP_THRESHOLD__SHIFT, 0x0384 }, + { mmGC_EDC_DROOP_CTRL, GC_EDC_DROOP_CTRL__EDC_DROOP_LEVEL_INDEX_MASK, GC_EDC_DROOP_CTRL__EDC_DROOP_LEVEL_INDEX__SHIFT, 0x0001 }, + { mmGC_EDC_DROOP_CTRL, GC_EDC_DROOP_CTRL__AVG_PSM_SEL_MASK, GC_EDC_DROOP_CTRL__AVG_PSM_SEL__SHIFT, 0x0001 }, + { mmGC_EDC_DROOP_CTRL, GC_EDC_DROOP_CTRL__EDC_LEVEL_SEL_MASK, GC_EDC_DROOP_CTRL__EDC_LEVEL_SEL__SHIFT, 0x0001 }, + + { 0xFFFFFFFF } /* End of list */ +}; + +static const struct vega10_didt_config_reg PSMGCEDCCtrlResetConfig_vega10[] = +{ +/* --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- + * Offset Mask Shift Value + * --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- + */ + { mmGC_EDC_CTRL, GC_EDC_CTRL__EDC_EN_MASK, GC_EDC_CTRL__EDC_EN__SHIFT, 0x0000 }, + { mmGC_EDC_CTRL, GC_EDC_CTRL__EDC_SW_RST_MASK, GC_EDC_CTRL__EDC_SW_RST__SHIFT, 0x0001 }, + { mmGC_EDC_CTRL, GC_EDC_CTRL__EDC_CLK_EN_OVERRIDE_MASK, GC_EDC_CTRL__EDC_CLK_EN_OVERRIDE__SHIFT, 0x0000 }, + { mmGC_EDC_CTRL, GC_EDC_CTRL__EDC_FORCE_STALL_MASK, GC_EDC_CTRL__EDC_FORCE_STALL__SHIFT, 0x0000 }, + { mmGC_EDC_CTRL, GC_EDC_CTRL__EDC_TRIGGER_THROTTLE_LOWBIT_MASK, GC_EDC_CTRL__EDC_TRIGGER_THROTTLE_LOWBIT__SHIFT, 0x0000 }, + { mmGC_EDC_CTRL, GC_EDC_CTRL__EDC_ALLOW_WRITE_PWRDELTA_MASK, GC_EDC_CTRL__EDC_ALLOW_WRITE_PWRDELTA__SHIFT, 0x0000 }, + + { 0xFFFFFFFF } /* End of list */ +}; + +static const struct vega10_didt_config_reg PSMGCEDCCtrlConfig_vega10[] = +{ +/* --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- + * Offset Mask Shift Value + * --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- + */ + { mmGC_EDC_CTRL, GC_EDC_CTRL__EDC_EN_MASK, GC_EDC_CTRL__EDC_EN__SHIFT, 0x0001 }, + { mmGC_EDC_CTRL, GC_EDC_CTRL__EDC_SW_RST_MASK, GC_EDC_CTRL__EDC_SW_RST__SHIFT, 0x0000 }, + { mmGC_EDC_CTRL, GC_EDC_CTRL__EDC_CLK_EN_OVERRIDE_MASK, GC_EDC_CTRL__EDC_CLK_EN_OVERRIDE__SHIFT, 0x0000 }, + { mmGC_EDC_CTRL, GC_EDC_CTRL__EDC_FORCE_STALL_MASK, GC_EDC_CTRL__EDC_FORCE_STALL__SHIFT, 0x0000 }, + { mmGC_EDC_CTRL, GC_EDC_CTRL__EDC_TRIGGER_THROTTLE_LOWBIT_MASK, GC_EDC_CTRL__EDC_TRIGGER_THROTTLE_LOWBIT__SHIFT, 0x0000 }, + { mmGC_EDC_CTRL, GC_EDC_CTRL__EDC_ALLOW_WRITE_PWRDELTA_MASK, GC_EDC_CTRL__EDC_ALLOW_WRITE_PWRDELTA__SHIFT, 0x0000 }, + + { 0xFFFFFFFF } /* End of list */ +}; + +static const struct vega10_didt_config_reg AvfsPSMResetConfig_vega10[]= +{ +/* --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- + * Offset Mask Shift Value + * --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- + */ + { 0x16A02, 0xFFFFFFFF, 0x0, 0x0000005F }, + { 0x16A05, 0xFFFFFFFF, 0x0, 0x00000001 }, + { 0x16A06, 0x00000001, 0x0, 0x02000000 }, + { 0x16A01, 0xFFFFFFFF, 0x0, 0x00003027 }, + + { 0xFFFFFFFF } /* End of list */ +}; + +static const struct vega10_didt_config_reg AvfsPSMInitConfig_vega10[] = +{ +/* --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- + * Offset Mask Shift Value + * --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- + */ + { 0x16A05, 0xFFFFFFFF, 0x18, 0x00000001 }, + { 0x16A05, 0xFFFFFFFF, 0x8, 0x00000003 }, + { 0x16A05, 0xFFFFFFFF, 0xa, 0x00000006 }, + { 0x16A05, 0xFFFFFFFF, 0x7, 0x00000000 }, + { 0x16A06, 0xFFFFFFFF, 0x18, 0x00000001 }, + { 0x16A06, 0xFFFFFFFF, 0x19, 0x00000001 }, + { 0x16A01, 0xFFFFFFFF, 0x0, 0x00003027 }, + + { 0xFFFFFFFF } /* End of list */ +}; + +static int vega10_program_didt_config_registers(struct pp_hwmgr *hwmgr, const struct vega10_didt_config_reg *config_regs, enum vega10_didt_config_reg_type reg_type) +{ + uint32_t data; + + PP_ASSERT_WITH_CODE((config_regs != NULL), "[vega10_program_didt_config_registers] Invalid config register table!", return -EINVAL); + + while (config_regs->offset != 0xFFFFFFFF) { + switch (reg_type) { + case VEGA10_CONFIGREG_DIDT: + data = cgs_read_ind_register(hwmgr->device, CGS_IND_REG__DIDT, config_regs->offset); + data &= ~config_regs->mask; + data |= ((config_regs->value << config_regs->shift) & config_regs->mask); + cgs_write_ind_register(hwmgr->device, CGS_IND_REG__DIDT, config_regs->offset, data); + break; + case VEGA10_CONFIGREG_GCCAC: + data = cgs_read_ind_register(hwmgr->device, CGS_IND_REG_GC_CAC, config_regs->offset); + data &= ~config_regs->mask; + data |= ((config_regs->value << config_regs->shift) & config_regs->mask); + cgs_write_ind_register(hwmgr->device, CGS_IND_REG_GC_CAC, config_regs->offset, data); + break; + case VEGA10_CONFIGREG_SECAC: + data = cgs_read_ind_register(hwmgr->device, CGS_IND_REG_SE_CAC, config_regs->offset); + data &= ~config_regs->mask; + data |= ((config_regs->value << config_regs->shift) & config_regs->mask); + cgs_write_ind_register(hwmgr->device, CGS_IND_REG_SE_CAC, config_regs->offset, data); + break; + default: + return -EINVAL; + } + + config_regs++; + } + + return 0; +} + +static int vega10_program_gc_didt_config_registers(struct pp_hwmgr *hwmgr, const struct vega10_didt_config_reg *config_regs) +{ + uint32_t data; + + while (config_regs->offset != 0xFFFFFFFF) { + data = cgs_read_register(hwmgr->device, config_regs->offset); + data &= ~config_regs->mask; + data |= ((config_regs->value << config_regs->shift) & config_regs->mask); + cgs_write_register(hwmgr->device, config_regs->offset, data); + config_regs++; + } + + return 0; +} + +static void vega10_didt_set_mask(struct pp_hwmgr *hwmgr, const bool enable) +{ + uint32_t data; + int result; + uint32_t en = (enable ? 1 : 0); + uint32_t didt_block_info = SQ_IR_MASK | TCP_IR_MASK | TD_PCC_MASK; + + if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_SQRamping)) { + data = cgs_read_ind_register(hwmgr->device, CGS_IND_REG__DIDT, ixDIDT_SQ_CTRL0); + data &= ~DIDT_SQ_CTRL0__DIDT_CTRL_EN_MASK; + data |= ((en << DIDT_SQ_CTRL0__DIDT_CTRL_EN__SHIFT) & DIDT_SQ_CTRL0__DIDT_CTRL_EN_MASK); + cgs_write_ind_register(hwmgr->device, CGS_IND_REG__DIDT, ixDIDT_SQ_CTRL0, data); + didt_block_info &= ~SQ_Enable_MASK; + didt_block_info |= en << SQ_Enable_SHIFT; + } + + if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_DBRamping)) { + data = cgs_read_ind_register(hwmgr->device, CGS_IND_REG__DIDT, ixDIDT_DB_CTRL0); + data &= ~DIDT_DB_CTRL0__DIDT_CTRL_EN_MASK; + data |= ((en << DIDT_DB_CTRL0__DIDT_CTRL_EN__SHIFT) & DIDT_DB_CTRL0__DIDT_CTRL_EN_MASK); + cgs_write_ind_register(hwmgr->device, CGS_IND_REG__DIDT, ixDIDT_DB_CTRL0, data); + didt_block_info &= ~DB_Enable_MASK; + didt_block_info |= en << DB_Enable_SHIFT; + } + + if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_TDRamping)) { + data = cgs_read_ind_register(hwmgr->device, CGS_IND_REG__DIDT, ixDIDT_TD_CTRL0); + data &= ~DIDT_TD_CTRL0__DIDT_CTRL_EN_MASK; + data |= ((en << DIDT_TD_CTRL0__DIDT_CTRL_EN__SHIFT) & DIDT_TD_CTRL0__DIDT_CTRL_EN_MASK); + cgs_write_ind_register(hwmgr->device, CGS_IND_REG__DIDT, ixDIDT_TD_CTRL0, data); + didt_block_info &= ~TD_Enable_MASK; + didt_block_info |= en << TD_Enable_SHIFT; + } + + if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_TCPRamping)) { + data = cgs_read_ind_register(hwmgr->device, CGS_IND_REG__DIDT, ixDIDT_TCP_CTRL0); + data &= ~DIDT_TCP_CTRL0__DIDT_CTRL_EN_MASK; + data |= ((en << DIDT_TCP_CTRL0__DIDT_CTRL_EN__SHIFT) & DIDT_TCP_CTRL0__DIDT_CTRL_EN_MASK); + cgs_write_ind_register(hwmgr->device, CGS_IND_REG__DIDT, ixDIDT_TCP_CTRL0, data); + didt_block_info &= ~TCP_Enable_MASK; + didt_block_info |= en << TCP_Enable_SHIFT; + } + + if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_DBRRamping)) { + data = cgs_read_ind_register(hwmgr->device, CGS_IND_REG__DIDT, ixDIDT_DBR_CTRL0); + data &= ~DIDT_DBR_CTRL0__DIDT_CTRL_EN_MASK; + data |= ((en << DIDT_DBR_CTRL0__DIDT_CTRL_EN__SHIFT) & DIDT_DBR_CTRL0__DIDT_CTRL_EN_MASK); + cgs_write_ind_register(hwmgr->device, CGS_IND_REG__DIDT, ixDIDT_DBR_CTRL0, data); + } + + if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_DiDtEDCEnable)) { + if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_SQRamping)) { + data = cgs_read_ind_register(hwmgr->device, CGS_IND_REG__DIDT, ixDIDT_SQ_EDC_CTRL); + data &= ~DIDT_SQ_EDC_CTRL__EDC_EN_MASK; + data |= ((en << DIDT_SQ_EDC_CTRL__EDC_EN__SHIFT) & DIDT_SQ_EDC_CTRL__EDC_EN_MASK); + data &= ~DIDT_SQ_EDC_CTRL__EDC_SW_RST_MASK; + data |= ((~en << DIDT_SQ_EDC_CTRL__EDC_SW_RST__SHIFT) & DIDT_SQ_EDC_CTRL__EDC_SW_RST_MASK); + cgs_write_ind_register(hwmgr->device, CGS_IND_REG__DIDT, ixDIDT_SQ_EDC_CTRL, data); + } + + if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_DBRamping)) { + data = cgs_read_ind_register(hwmgr->device, CGS_IND_REG__DIDT, ixDIDT_DB_EDC_CTRL); + data &= ~DIDT_DB_EDC_CTRL__EDC_EN_MASK; + data |= ((en << DIDT_DB_EDC_CTRL__EDC_EN__SHIFT) & DIDT_DB_EDC_CTRL__EDC_EN_MASK); + data &= ~DIDT_DB_EDC_CTRL__EDC_SW_RST_MASK; + data |= ((~en << DIDT_DB_EDC_CTRL__EDC_SW_RST__SHIFT) & DIDT_DB_EDC_CTRL__EDC_SW_RST_MASK); + cgs_write_ind_register(hwmgr->device, CGS_IND_REG__DIDT, ixDIDT_DB_EDC_CTRL, data); + } + + if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_TDRamping)) { + data = cgs_read_ind_register(hwmgr->device, CGS_IND_REG__DIDT, ixDIDT_TD_EDC_CTRL); + data &= ~DIDT_TD_EDC_CTRL__EDC_EN_MASK; + data |= ((en << DIDT_TD_EDC_CTRL__EDC_EN__SHIFT) & DIDT_TD_EDC_CTRL__EDC_EN_MASK); + data &= ~DIDT_TD_EDC_CTRL__EDC_SW_RST_MASK; + data |= ((~en << DIDT_TD_EDC_CTRL__EDC_SW_RST__SHIFT) & DIDT_TD_EDC_CTRL__EDC_SW_RST_MASK); + cgs_write_ind_register(hwmgr->device, CGS_IND_REG__DIDT, ixDIDT_TD_EDC_CTRL, data); + } + + if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_TCPRamping)) { + data = cgs_read_ind_register(hwmgr->device, CGS_IND_REG__DIDT, ixDIDT_TCP_EDC_CTRL); + data &= ~DIDT_TCP_EDC_CTRL__EDC_EN_MASK; + data |= ((en << DIDT_TCP_EDC_CTRL__EDC_EN__SHIFT) & DIDT_TCP_EDC_CTRL__EDC_EN_MASK); + data &= ~DIDT_TCP_EDC_CTRL__EDC_SW_RST_MASK; + data |= ((~en << DIDT_TCP_EDC_CTRL__EDC_SW_RST__SHIFT) & DIDT_TCP_EDC_CTRL__EDC_SW_RST_MASK); + cgs_write_ind_register(hwmgr->device, CGS_IND_REG__DIDT, ixDIDT_TCP_EDC_CTRL, data); + } + + if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_DBRRamping)) { + data = cgs_read_ind_register(hwmgr->device, CGS_IND_REG__DIDT, ixDIDT_DBR_EDC_CTRL); + data &= ~DIDT_DBR_EDC_CTRL__EDC_EN_MASK; + data |= ((en << DIDT_DBR_EDC_CTRL__EDC_EN__SHIFT) & DIDT_DBR_EDC_CTRL__EDC_EN_MASK); + data &= ~DIDT_DBR_EDC_CTRL__EDC_SW_RST_MASK; + data |= ((~en << DIDT_DBR_EDC_CTRL__EDC_SW_RST__SHIFT) & DIDT_DBR_EDC_CTRL__EDC_SW_RST_MASK); + cgs_write_ind_register(hwmgr->device, CGS_IND_REG__DIDT, ixDIDT_DBR_EDC_CTRL, data); + } + } + + if (enable) { + /* For Vega10, SMC does not support any mask yet. */ + result = smum_send_msg_to_smc_with_parameter(hwmgr->smumgr, PPSMC_MSG_ConfigureGfxDidt, didt_block_info); + PP_ASSERT((0 == result), "[EnableDiDtConfig] SMC Configure Gfx Didt Failed!"); + } +} + +static int vega10_enable_cac_driving_se_didt_config(struct pp_hwmgr *hwmgr) +{ + int result; + uint32_t num_se = 0, count, data; + struct cgs_system_info sys_info = {0}; + uint32_t reg; + + sys_info.size = sizeof(struct cgs_system_info); + sys_info.info_id = CGS_SYSTEM_INFO_GFX_SE_INFO; + if (cgs_query_system_info(hwmgr->device, &sys_info) == 0) + num_se = sys_info.value; + + cgs_enter_safe_mode(hwmgr->device, true); + + cgs_lock_grbm_idx(hwmgr->device, true); + reg = soc15_get_register_offset(GC_HWID, 0, mmGRBM_GFX_INDEX_BASE_IDX, mmGRBM_GFX_INDEX); + for (count = 0; count < num_se; count++) { + data = GRBM_GFX_INDEX__INSTANCE_BROADCAST_WRITES_MASK | GRBM_GFX_INDEX__SH_BROADCAST_WRITES_MASK | ( count << GRBM_GFX_INDEX__SE_INDEX__SHIFT); + cgs_write_register(hwmgr->device, reg, data); + + result = vega10_program_didt_config_registers(hwmgr, SEDiDtStallCtrlConfig_vega10, VEGA10_CONFIGREG_DIDT); + result |= vega10_program_didt_config_registers(hwmgr, SEDiDtStallPatternConfig_vega10, VEGA10_CONFIGREG_DIDT); + result |= vega10_program_didt_config_registers(hwmgr, SEDiDtWeightConfig_Vega10, VEGA10_CONFIGREG_DIDT); + result |= vega10_program_didt_config_registers(hwmgr, SEDiDtCtrl1Config_Vega10, VEGA10_CONFIGREG_DIDT); + result |= vega10_program_didt_config_registers(hwmgr, SEDiDtCtrl2Config_Vega10, VEGA10_CONFIGREG_DIDT); + result |= vega10_program_didt_config_registers(hwmgr, SEDiDtCtrl3Config_vega10, VEGA10_CONFIGREG_DIDT); + result |= vega10_program_didt_config_registers(hwmgr, SEDiDtTuningCtrlConfig_Vega10, VEGA10_CONFIGREG_DIDT); + result |= vega10_program_didt_config_registers(hwmgr, SELCacConfig_Vega10, VEGA10_CONFIGREG_SECAC); + result |= vega10_program_didt_config_registers(hwmgr, SEDiDtCtrl0Config_Vega10, VEGA10_CONFIGREG_DIDT); + + if (0 != result) + break; + } + cgs_write_register(hwmgr->device, reg, 0xE0000000); + cgs_lock_grbm_idx(hwmgr->device, false); + + vega10_didt_set_mask(hwmgr, true); + + cgs_enter_safe_mode(hwmgr->device, false); + + return 0; +} + +static int vega10_disable_cac_driving_se_didt_config(struct pp_hwmgr *hwmgr) +{ + cgs_enter_safe_mode(hwmgr->device, true); + + vega10_didt_set_mask(hwmgr, false); + + cgs_enter_safe_mode(hwmgr->device, false); + + return 0; +} + +static int vega10_enable_psm_gc_didt_config(struct pp_hwmgr *hwmgr) +{ + int result; + uint32_t num_se = 0, count, data; + struct cgs_system_info sys_info = {0}; + uint32_t reg; + + sys_info.size = sizeof(struct cgs_system_info); + sys_info.info_id = CGS_SYSTEM_INFO_GFX_SE_INFO; + if (cgs_query_system_info(hwmgr->device, &sys_info) == 0) + num_se = sys_info.value; + + cgs_enter_safe_mode(hwmgr->device, true); + + cgs_lock_grbm_idx(hwmgr->device, true); + reg = soc15_get_register_offset(GC_HWID, 0, mmGRBM_GFX_INDEX_BASE_IDX, mmGRBM_GFX_INDEX); + for (count = 0; count < num_se; count++) { + data = GRBM_GFX_INDEX__INSTANCE_BROADCAST_WRITES_MASK | GRBM_GFX_INDEX__SH_BROADCAST_WRITES_MASK | ( count << GRBM_GFX_INDEX__SE_INDEX__SHIFT); + cgs_write_register(hwmgr->device, reg, data); + + result = vega10_program_didt_config_registers(hwmgr, SEDiDtStallCtrlConfig_vega10, VEGA10_CONFIGREG_DIDT); + result |= vega10_program_didt_config_registers(hwmgr, SEDiDtStallPatternConfig_vega10, VEGA10_CONFIGREG_DIDT); + result |= vega10_program_didt_config_registers(hwmgr, SEDiDtCtrl3Config_vega10, VEGA10_CONFIGREG_DIDT); + result |= vega10_program_didt_config_registers(hwmgr, SEDiDtCtrl0Config_Vega10, VEGA10_CONFIGREG_DIDT); + if (0 != result) + break; + } + cgs_write_register(hwmgr->device, reg, 0xE0000000); + cgs_lock_grbm_idx(hwmgr->device, false); + + vega10_didt_set_mask(hwmgr, true); + + cgs_enter_safe_mode(hwmgr->device, false); + + vega10_program_gc_didt_config_registers(hwmgr, GCDiDtDroopCtrlConfig_vega10); + if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_GCEDC)) + vega10_program_gc_didt_config_registers(hwmgr, GCDiDtCtrl0Config_vega10); + + if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_PSM)) + vega10_program_gc_didt_config_registers(hwmgr, AvfsPSMInitConfig_vega10); + + return 0; +} + +static int vega10_disable_psm_gc_didt_config(struct pp_hwmgr *hwmgr) +{ + uint32_t data; + + cgs_enter_safe_mode(hwmgr->device, true); + + vega10_didt_set_mask(hwmgr, false); + + cgs_enter_safe_mode(hwmgr->device, false); + + if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_GCEDC)) { + data = 0x00000000; + cgs_write_register(hwmgr->device, mmGC_DIDT_CTRL0, data); + } + + if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_PSM)) + vega10_program_gc_didt_config_registers(hwmgr, AvfsPSMResetConfig_vega10); + + return 0; +} + +static int vega10_enable_se_edc_config(struct pp_hwmgr *hwmgr) +{ + int result; + uint32_t num_se = 0, count, data; + struct cgs_system_info sys_info = {0}; + uint32_t reg; + + sys_info.size = sizeof(struct cgs_system_info); + sys_info.info_id = CGS_SYSTEM_INFO_GFX_SE_INFO; + if (cgs_query_system_info(hwmgr->device, &sys_info) == 0) + num_se = sys_info.value; + + cgs_enter_safe_mode(hwmgr->device, true); + + cgs_lock_grbm_idx(hwmgr->device, true); + reg = soc15_get_register_offset(GC_HWID, 0, mmGRBM_GFX_INDEX_BASE_IDX, mmGRBM_GFX_INDEX); + for (count = 0; count < num_se; count++) { + data = GRBM_GFX_INDEX__INSTANCE_BROADCAST_WRITES_MASK | GRBM_GFX_INDEX__SH_BROADCAST_WRITES_MASK | ( count << GRBM_GFX_INDEX__SE_INDEX__SHIFT); + cgs_write_register(hwmgr->device, reg, data); + result = vega10_program_didt_config_registers(hwmgr, SEDiDtWeightConfig_Vega10, VEGA10_CONFIGREG_DIDT); + result |= vega10_program_didt_config_registers(hwmgr, SEEDCStallPatternConfig_Vega10, VEGA10_CONFIGREG_DIDT); + result |= vega10_program_didt_config_registers(hwmgr, SEEDCStallDelayConfig_Vega10, VEGA10_CONFIGREG_DIDT); + result |= vega10_program_didt_config_registers(hwmgr, SEEDCThresholdConfig_Vega10, VEGA10_CONFIGREG_DIDT); + result |= vega10_program_didt_config_registers(hwmgr, SEEDCCtrlResetConfig_Vega10, VEGA10_CONFIGREG_DIDT); + result |= vega10_program_didt_config_registers(hwmgr, SEEDCCtrlConfig_Vega10, VEGA10_CONFIGREG_DIDT); + + if (0 != result) + break; + } + cgs_write_register(hwmgr->device, reg, 0xE0000000); + cgs_lock_grbm_idx(hwmgr->device, false); + + vega10_didt_set_mask(hwmgr, true); + + cgs_enter_safe_mode(hwmgr->device, false); + + return 0; +} + +static int vega10_disable_se_edc_config(struct pp_hwmgr *hwmgr) +{ + cgs_enter_safe_mode(hwmgr->device, true); + + vega10_didt_set_mask(hwmgr, false); + + cgs_enter_safe_mode(hwmgr->device, false); + + return 0; +} + +static int vega10_enable_psm_gc_edc_config(struct pp_hwmgr *hwmgr) +{ + int result; + uint32_t num_se = 0; + uint32_t count, data; + struct cgs_system_info sys_info = {0}; + uint32_t reg; + + sys_info.size = sizeof(struct cgs_system_info); + sys_info.info_id = CGS_SYSTEM_INFO_GFX_SE_INFO; + if (cgs_query_system_info(hwmgr->device, &sys_info) == 0) + num_se = sys_info.value; + + cgs_enter_safe_mode(hwmgr->device, true); + + vega10_program_gc_didt_config_registers(hwmgr, AvfsPSMResetConfig_vega10); + + cgs_lock_grbm_idx(hwmgr->device, true); + reg = soc15_get_register_offset(GC_HWID, 0, mmGRBM_GFX_INDEX_BASE_IDX, mmGRBM_GFX_INDEX); + for (count = 0; count < num_se; count++) { + data = GRBM_GFX_INDEX__INSTANCE_BROADCAST_WRITES_MASK | GRBM_GFX_INDEX__SH_BROADCAST_WRITES_MASK | ( count << GRBM_GFX_INDEX__SE_INDEX__SHIFT); + cgs_write_register(hwmgr->device, reg, data); + result |= vega10_program_didt_config_registers(hwmgr, PSMSEEDCStallPatternConfig_Vega10, VEGA10_CONFIGREG_DIDT); + result |= vega10_program_didt_config_registers(hwmgr, PSMSEEDCStallDelayConfig_Vega10, VEGA10_CONFIGREG_DIDT); + result |= vega10_program_didt_config_registers(hwmgr, PSMSEEDCCtrlResetConfig_Vega10, VEGA10_CONFIGREG_DIDT); + result |= vega10_program_didt_config_registers(hwmgr, PSMSEEDCCtrlConfig_Vega10, VEGA10_CONFIGREG_DIDT); + + if (0 != result) + break; + } + cgs_write_register(hwmgr->device, reg, 0xE0000000); + cgs_lock_grbm_idx(hwmgr->device, false); + + vega10_didt_set_mask(hwmgr, true); + + cgs_enter_safe_mode(hwmgr->device, false); + + vega10_program_gc_didt_config_registers(hwmgr, PSMGCEDCDroopCtrlConfig_vega10); + + if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_GCEDC)) { + vega10_program_gc_didt_config_registers(hwmgr, PSMGCEDCCtrlResetConfig_vega10); + vega10_program_gc_didt_config_registers(hwmgr, PSMGCEDCCtrlConfig_vega10); + } + + if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_PSM)) + vega10_program_gc_didt_config_registers(hwmgr, AvfsPSMInitConfig_vega10); + + return 0; +} + +static int vega10_disable_psm_gc_edc_config(struct pp_hwmgr *hwmgr) +{ + uint32_t data; + + cgs_enter_safe_mode(hwmgr->device, true); + + vega10_didt_set_mask(hwmgr, false); + + cgs_enter_safe_mode(hwmgr->device, false); + + if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_GCEDC)) { + data = 0x00000000; + cgs_write_register(hwmgr->device, mmGC_EDC_CTRL, data); + } + + if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_PSM)) + vega10_program_gc_didt_config_registers(hwmgr, AvfsPSMResetConfig_vega10); + + return 0; +} + +static int vega10_enable_se_edc_force_stall_config(struct pp_hwmgr *hwmgr) +{ + uint32_t reg; + int result; + + cgs_enter_safe_mode(hwmgr->device, true); + + cgs_lock_grbm_idx(hwmgr->device, true); + reg = soc15_get_register_offset(GC_HWID, 0, mmGRBM_GFX_INDEX_BASE_IDX, mmGRBM_GFX_INDEX); + cgs_write_register(hwmgr->device, reg, 0xE0000000); + cgs_lock_grbm_idx(hwmgr->device, false); + + result = vega10_program_didt_config_registers(hwmgr, SEEDCForceStallPatternConfig_Vega10, VEGA10_CONFIGREG_DIDT); + result |= vega10_program_didt_config_registers(hwmgr, SEEDCCtrlForceStallConfig_Vega10, VEGA10_CONFIGREG_DIDT); + if (0 != result) + return result; + + vega10_didt_set_mask(hwmgr, true); + + cgs_enter_safe_mode(hwmgr->device, false); + + return 0; +} + +static int vega10_disable_se_edc_force_stall_config(struct pp_hwmgr *hwmgr) +{ + int result; + + result = vega10_disable_se_edc_config(hwmgr); + PP_ASSERT_WITH_CODE((0 == result), "[DisableDiDtConfig] Pre DIDT disable clock gating failed!", return result); + + return 0; +} + +int vega10_enable_didt_config(struct pp_hwmgr *hwmgr) +{ + int result = 0; + struct vega10_hwmgr *data = (struct vega10_hwmgr *)(hwmgr->backend); + + if (data->smu_features[GNLD_DIDT].supported) { + if (data->smu_features[GNLD_DIDT].enabled) + PP_DBG_LOG("[EnableDiDtConfig] Feature DiDt Already enabled!\n"); + + switch (data->registry_data.didt_mode) { + case 0: + result = vega10_enable_cac_driving_se_didt_config(hwmgr); + PP_ASSERT_WITH_CODE((0 == result), "[EnableDiDt] Attempt to enable DiDt Mode 0 Failed!", return result); + break; + case 2: + result = vega10_enable_psm_gc_didt_config(hwmgr); + PP_ASSERT_WITH_CODE((0 == result), "[EnableDiDt] Attempt to enable DiDt Mode 2 Failed!", return result); + break; + case 3: + result = vega10_enable_se_edc_config(hwmgr); + PP_ASSERT_WITH_CODE((0 == result), "[EnableDiDt] Attempt to enable DiDt Mode 3 Failed!", return result); + break; + case 1: + case 4: + case 5: + result = vega10_enable_psm_gc_edc_config(hwmgr); + PP_ASSERT_WITH_CODE((0 == result), "[EnableDiDt] Attempt to enable DiDt Mode 5 Failed!", return result); + break; + case 6: + result = vega10_enable_se_edc_force_stall_config(hwmgr); + PP_ASSERT_WITH_CODE((0 == result), "[EnableDiDt] Attempt to enable DiDt Mode 6 Failed!", return result); + break; + default: + result = -EINVAL; + break; + } + + if (0 == result) { + PP_ASSERT_WITH_CODE((!vega10_enable_smc_features(hwmgr->smumgr, true, data->smu_features[GNLD_DIDT].smu_feature_bitmap)), + "[EnableDiDtConfig] Attempt to Enable DiDt feature Failed!", return result); + data->smu_features[GNLD_DIDT].enabled = true; + } + } + + return result; +} + +int vega10_disable_didt_config(struct pp_hwmgr *hwmgr) +{ + int result = 0; + struct vega10_hwmgr *data = (struct vega10_hwmgr *)(hwmgr->backend); + + if (data->smu_features[GNLD_DIDT].supported) { + if (!data->smu_features[GNLD_DIDT].enabled) + PP_DBG_LOG("[DisableDiDtConfig] Feature DiDt Already Disabled!\n"); + + switch (data->registry_data.didt_mode) { + case 0: + result = vega10_disable_cac_driving_se_didt_config(hwmgr); + PP_ASSERT_WITH_CODE((0 == result), "[DisableDiDt] Attempt to disable DiDt Mode 0 Failed!", return result); + break; + case 2: + result = vega10_disable_psm_gc_didt_config(hwmgr); + PP_ASSERT_WITH_CODE((0 == result), "[DisableDiDt] Attempt to disable DiDt Mode 2 Failed!", return result); + break; + case 3: + result = vega10_disable_se_edc_config(hwmgr); + PP_ASSERT_WITH_CODE((0 == result), "[DisableDiDt] Attempt to disable DiDt Mode 3 Failed!", return result); + break; + case 1: + case 4: + case 5: + result = vega10_disable_psm_gc_edc_config(hwmgr); + PP_ASSERT_WITH_CODE((0 == result), "[DisableDiDt] Attempt to disable DiDt Mode 5 Failed!", return result); + break; + case 6: + result = vega10_disable_se_edc_force_stall_config(hwmgr); + PP_ASSERT_WITH_CODE((0 == result), "[DisableDiDt] Attempt to disable DiDt Mode 6 Failed!", return result); + break; + default: + result = -EINVAL; + break; + } + + if (0 == result) { + PP_ASSERT_WITH_CODE((0 != vega10_enable_smc_features(hwmgr->smumgr, false, data->smu_features[GNLD_DIDT].smu_feature_bitmap)), + "[DisableDiDtConfig] Attempt to Disable DiDt feature Failed!", return result); + data->smu_features[GNLD_DIDT].enabled = false; + } + } + + return result; +} void vega10_initialize_power_tune_defaults(struct pp_hwmgr *hwmgr) { diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_powertune.h b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_powertune.h index 9ecaa27c0bb5..b95771ab89cd 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_powertune.h +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_powertune.h @@ -31,6 +31,12 @@ enum vega10_pt_config_reg_type { VEGA10_CONFIGREG_MAX }; +enum vega10_didt_config_reg_type { + VEGA10_CONFIGREG_DIDT = 0, + VEGA10_CONFIGREG_GCCAC, + VEGA10_CONFIGREG_SECAC +}; + /* PowerContainment Features */ #define POWERCONTAINMENT_FEATURE_DTE 0x00000001 #define POWERCONTAINMENT_FEATURE_TDCLimit 0x00000002 @@ -44,6 +50,13 @@ struct vega10_pt_config_reg { enum vega10_pt_config_reg_type type; }; +struct vega10_didt_config_reg { + uint32_t offset; + uint32_t mask; + uint32_t shift; + uint32_t value; +}; + struct vega10_pt_defaults { uint8_t SviLoadLineEn; uint8_t SviLoadLineVddC; @@ -62,5 +75,8 @@ 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); +int vega10_enable_didt_config(struct pp_hwmgr *hwmgr); +int vega10_disable_didt_config(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 1623644ea49a..e343df190375 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_processpptables.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_processpptables.c @@ -31,6 +31,8 @@ #include "cgs_common.h" #include "vega10_pptable.h" +#define NUM_DSPCLK_LEVELS 8 + static void set_hw_cap(struct pp_hwmgr *hwmgr, bool enable, enum phm_platform_caps cap) { @@ -644,11 +646,11 @@ static int get_gfxclk_voltage_dependency_table( return 0; } -static int get_dcefclk_voltage_dependency_table( +static int get_pix_clk_voltage_dependency_table( struct pp_hwmgr *hwmgr, struct phm_ppt_v1_clock_voltage_dependency_table **pp_vega10_clk_dep_table, - const ATOM_Vega10_DCEFCLK_Dependency_Table *clk_dep_table) + const ATOM_Vega10_PIXCLK_Dependency_Table *clk_dep_table) { uint32_t table_size, i; struct phm_ppt_v1_clock_voltage_dependency_table @@ -681,6 +683,76 @@ static int get_dcefclk_voltage_dependency_table( return 0; } +static int get_dcefclk_voltage_dependency_table( + struct pp_hwmgr *hwmgr, + struct phm_ppt_v1_clock_voltage_dependency_table + **pp_vega10_clk_dep_table, + const ATOM_Vega10_DCEFCLK_Dependency_Table *clk_dep_table) +{ + uint32_t table_size, i; + uint8_t num_entries; + struct phm_ppt_v1_clock_voltage_dependency_table + *clk_table; + struct cgs_system_info sys_info = {0}; + uint32_t dev_id; + uint32_t rev_id; + + PP_ASSERT_WITH_CODE((clk_dep_table->ucNumEntries != 0), + "Invalid PowerPlay Table!", return -1); + +/* + * workaround needed to add another DPM level for pioneer cards + * as VBIOS is locked down. + * This DPM level was added to support 3DPM monitors @ 4K120Hz + * + */ + sys_info.size = sizeof(struct cgs_system_info); + sys_info.info_id = CGS_SYSTEM_INFO_PCIE_DEV; + cgs_query_system_info(hwmgr->device, &sys_info); + dev_id = (uint32_t)sys_info.value; + + sys_info.size = sizeof(struct cgs_system_info); + sys_info.info_id = CGS_SYSTEM_INFO_PCIE_REV; + cgs_query_system_info(hwmgr->device, &sys_info); + rev_id = (uint32_t)sys_info.value; + + if (dev_id == 0x6863 && rev_id == 0 && + clk_dep_table->entries[clk_dep_table->ucNumEntries - 1].ulClk < 90000) + num_entries = clk_dep_table->ucNumEntries + 1 > NUM_DSPCLK_LEVELS ? + NUM_DSPCLK_LEVELS : clk_dep_table->ucNumEntries + 1; + else + num_entries = clk_dep_table->ucNumEntries; + + + table_size = sizeof(uint32_t) + + sizeof(phm_ppt_v1_clock_voltage_dependency_record) * + num_entries; + + clk_table = (struct phm_ppt_v1_clock_voltage_dependency_table *) + kzalloc(table_size, GFP_KERNEL); + + if (!clk_table) + return -ENOMEM; + + clk_table->count = (uint32_t)num_entries; + + for (i = 0; i < clk_dep_table->ucNumEntries; i++) { + clk_table->entries[i].vddInd = + clk_dep_table->entries[i].ucVddInd; + clk_table->entries[i].clk = + le32_to_cpu(clk_dep_table->entries[i].ulClk); + } + + if (i < num_entries) { + clk_table->entries[i].vddInd = clk_dep_table->entries[i-1].ucVddInd; + clk_table->entries[i].clk = 90000; + } + + *pp_vega10_clk_dep_table = clk_table; + + return 0; +} + static int get_pcie_table(struct pp_hwmgr *hwmgr, struct phm_ppt_v1_pcie_table **vega10_pcie_table, const Vega10_PPTable_Generic_SubTable_Header *table) @@ -862,21 +934,21 @@ static int init_powerplay_extended_tables( gfxclk_dep_table); if (!result && powerplay_table->usPixclkDependencyTableOffset) - result = get_dcefclk_voltage_dependency_table(hwmgr, + result = get_pix_clk_voltage_dependency_table(hwmgr, &pp_table_info->vdd_dep_on_pixclk, - (const ATOM_Vega10_DCEFCLK_Dependency_Table*) + (const ATOM_Vega10_PIXCLK_Dependency_Table*) pixclk_dep_table); if (!result && powerplay_table->usPhyClkDependencyTableOffset) - result = get_dcefclk_voltage_dependency_table(hwmgr, + result = get_pix_clk_voltage_dependency_table(hwmgr, &pp_table_info->vdd_dep_on_phyclk, - (const ATOM_Vega10_DCEFCLK_Dependency_Table *) + (const ATOM_Vega10_PIXCLK_Dependency_Table *) phyclk_dep_table); if (!result && powerplay_table->usDispClkDependencyTableOffset) - result = get_dcefclk_voltage_dependency_table(hwmgr, + result = get_pix_clk_voltage_dependency_table(hwmgr, &pp_table_info->vdd_dep_on_dispclk, - (const ATOM_Vega10_DCEFCLK_Dependency_Table *) + (const ATOM_Vega10_PIXCLK_Dependency_Table *) dispclk_dep_table); if (!result && powerplay_table->usDcefclkDependencyTableOffset) diff --git a/drivers/gpu/drm/amd/powerplay/inc/hardwaremanager.h b/drivers/gpu/drm/amd/powerplay/inc/hardwaremanager.h index a1ebe1014492..a4c8b09b6f14 100644 --- a/drivers/gpu/drm/amd/powerplay/inc/hardwaremanager.h +++ b/drivers/gpu/drm/amd/powerplay/inc/hardwaremanager.h @@ -164,9 +164,14 @@ enum phm_platform_caps { PHM_PlatformCaps_EnablePlatformPowerManagement, /* indicates that Platform Power Management feature is supported */ PHM_PlatformCaps_SurpriseRemoval, /* indicates that surprise removal feature is requested */ PHM_PlatformCaps_NewCACVoltage, /* indicates new CAC voltage table support */ + PHM_PlatformCaps_DiDtSupport, /* for dI/dT feature */ PHM_PlatformCaps_DBRamping, /* for dI/dT feature */ PHM_PlatformCaps_TDRamping, /* for dI/dT feature */ PHM_PlatformCaps_TCPRamping, /* for dI/dT feature */ + PHM_PlatformCaps_DBRRamping, /* for dI/dT feature */ + PHM_PlatformCaps_DiDtEDCEnable, /* for dI/dT feature */ + PHM_PlatformCaps_GCEDC, /* for dI/dT feature */ + PHM_PlatformCaps_PSM, /* for dI/dT feature */ PHM_PlatformCaps_EnableSMU7ThermalManagement, /* SMC will manage thermal events */ PHM_PlatformCaps_FPS, /* FPS support */ PHM_PlatformCaps_ACP, /* ACP support */ diff --git a/drivers/gpu/drm/amd/powerplay/inc/pp_debug.h b/drivers/gpu/drm/amd/powerplay/inc/pp_debug.h index f3f9ebb631a5..822cd8b5bf90 100644 --- a/drivers/gpu/drm/amd/powerplay/inc/pp_debug.h +++ b/drivers/gpu/drm/amd/powerplay/inc/pp_debug.h @@ -42,6 +42,12 @@ } \ } while (0) +#define PP_ASSERT(cond, msg) \ + do { \ + if (!(cond)) { \ + pr_warn("%s\n", msg); \ + } \ + } while (0) #define PP_DBG_LOG(fmt, ...) \ do { \ diff --git a/drivers/gpu/drm/amd/powerplay/inc/pp_soc15.h b/drivers/gpu/drm/amd/powerplay/inc/pp_soc15.h index 227d999b6bd1..a511611ec7e0 100644 --- a/drivers/gpu/drm/amd/powerplay/inc/pp_soc15.h +++ b/drivers/gpu/drm/amd/powerplay/inc/pp_soc15.h @@ -41,6 +41,8 @@ inline static uint32_t soc15_get_register_offset( reg = MP1_BASE.instance[inst].segment[segment] + offset; else if (hw_id == DF_HWID) reg = DF_BASE.instance[inst].segment[segment] + offset; + else if (hw_id == GC_HWID) + reg = GC_BASE.instance[inst].segment[segment] + offset; return reg; } diff --git a/drivers/gpu/drm/amd/powerplay/inc/smu9.h b/drivers/gpu/drm/amd/powerplay/inc/smu9.h index 9ef2490c7c2e..550ed675027a 100644 --- a/drivers/gpu/drm/amd/powerplay/inc/smu9.h +++ b/drivers/gpu/drm/amd/powerplay/inc/smu9.h @@ -55,9 +55,9 @@ #define FEATURE_FW_CTF_BIT 23 #define FEATURE_LED_DISPLAY_BIT 24 #define FEATURE_FAN_CONTROL_BIT 25 -#define FEATURE_VOLTAGE_CONTROLLER_BIT 26 -#define FEATURE_SPARE_27_BIT 27 -#define FEATURE_SPARE_28_BIT 28 +#define FEATURE_FAST_PPT_BIT 26 +#define FEATURE_GFX_EDC_BIT 27 +#define FEATURE_ACG_BIT 28 #define FEATURE_SPARE_29_BIT 29 #define FEATURE_SPARE_30_BIT 30 #define FEATURE_SPARE_31_BIT 31 @@ -90,9 +90,10 @@ #define FFEATURE_FW_CTF_MASK (1 << FEATURE_FW_CTF_BIT ) #define FFEATURE_LED_DISPLAY_MASK (1 << FEATURE_LED_DISPLAY_BIT ) #define FFEATURE_FAN_CONTROL_MASK (1 << FEATURE_FAN_CONTROL_BIT ) -#define FFEATURE_VOLTAGE_CONTROLLER_MASK (1 << FEATURE_VOLTAGE_CONTROLLER_BIT ) -#define FFEATURE_SPARE_27_MASK (1 << FEATURE_SPARE_27_BIT ) -#define FFEATURE_SPARE_28_MASK (1 << FEATURE_SPARE_28_BIT ) + +#define FEATURE_FAST_PPT_MASK (1 << FAST_PPT_BIT ) +#define FEATURE_GFX_EDC_MASK (1 << FEATURE_GFX_EDC_BIT ) +#define FEATURE_ACG_MASK (1 << FEATURE_ACG_BIT ) #define FFEATURE_SPARE_29_MASK (1 << FEATURE_SPARE_29_BIT ) #define FFEATURE_SPARE_30_MASK (1 << FEATURE_SPARE_30_BIT ) #define FFEATURE_SPARE_31_MASK (1 << FEATURE_SPARE_31_BIT ) 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 532186b6f941..f6d6c61f796a 100644 --- a/drivers/gpu/drm/amd/powerplay/inc/smu9_driver_if.h +++ b/drivers/gpu/drm/amd/powerplay/inc/smu9_driver_if.h @@ -312,7 +312,10 @@ typedef struct { PllSetting_t GfxBoostState; - uint32_t Reserved[14]; + uint8_t AcgEnable[NUM_GFXCLK_DPM_LEVELS]; + GbVdroopTable_t AcgBtcGbVdroopTable; + QuadraticInt_t AcgAvfsGb; + uint32_t Reserved[4]; /* Padding - ignore */ uint32_t MmHubPadding[7]; /* SMU internal use */ diff --git a/drivers/gpu/drm/amd/powerplay/inc/smumgr.h b/drivers/gpu/drm/amd/powerplay/inc/smumgr.h index 976e942ec694..5d61cc9d4554 100644 --- a/drivers/gpu/drm/amd/powerplay/inc/smumgr.h +++ b/drivers/gpu/drm/amd/powerplay/inc/smumgr.h @@ -131,6 +131,7 @@ struct pp_smumgr_func { bool (*is_dpm_running)(struct pp_hwmgr *hwmgr); int (*populate_requested_graphic_levels)(struct pp_hwmgr *hwmgr, struct amd_pp_profile *request); + bool (*is_hw_avfs_present)(struct pp_smumgr *smumgr); }; struct pp_smumgr { @@ -202,6 +203,8 @@ extern bool smum_is_dpm_running(struct pp_hwmgr *hwmgr); extern int smum_populate_requested_graphic_levels(struct pp_hwmgr *hwmgr, struct amd_pp_profile *request); +extern bool smum_is_hw_avfs_present(struct pp_smumgr *smumgr); + #define SMUM_FIELD_SHIFT(reg, field) reg##__##field##__SHIFT #define SMUM_FIELD_MASK(reg, field) reg##__##field##_MASK diff --git a/drivers/gpu/drm/amd/powerplay/inc/vega10_ppsmc.h b/drivers/gpu/drm/amd/powerplay/inc/vega10_ppsmc.h index b4af9e85dfa5..cb070ebc7de1 100644 --- a/drivers/gpu/drm/amd/powerplay/inc/vega10_ppsmc.h +++ b/drivers/gpu/drm/amd/powerplay/inc/vega10_ppsmc.h @@ -124,6 +124,10 @@ typedef uint16_t PPSMC_Result; #define PPSMC_MSG_NumOfDisplays 0x56 #define PPSMC_MSG_ReadSerialNumTop32 0x58 #define PPSMC_MSG_ReadSerialNumBottom32 0x59 +#define PPSMC_MSG_RunAcgBtc 0x5C +#define PPSMC_MSG_RunAcgInClosedLoop 0x5D +#define PPSMC_MSG_RunAcgInOpenLoop 0x5E +#define PPSMC_MSG_InitializeAcg 0x5F #define PPSMC_MSG_GetCurrPkgPwr 0x61 #define PPSMC_Message_Count 0x62 diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/fiji_smc.c b/drivers/gpu/drm/amd/powerplay/smumgr/fiji_smc.c index 6a320b27aefd..8712f093d6d9 100644 --- a/drivers/gpu/drm/amd/powerplay/smumgr/fiji_smc.c +++ b/drivers/gpu/drm/amd/powerplay/smumgr/fiji_smc.c @@ -2129,6 +2129,25 @@ int fiji_thermal_setup_fan_table(struct pp_hwmgr *hwmgr) return 0; } + +int fiji_thermal_avfs_enable(struct pp_hwmgr *hwmgr) +{ + int ret; + struct pp_smumgr *smumgr = (struct pp_smumgr *)(hwmgr->smumgr); + struct smu7_smumgr *smu_data = (struct smu7_smumgr *)(smumgr->backend); + + if (smu_data->avfs.avfs_btc_status != AVFS_BTC_ENABLEAVFS) + return 0; + + ret = smum_send_msg_to_smc(smumgr, PPSMC_MSG_EnableAvfs); + + if (!ret) + /* If this param is not changed, this function could fire unnecessarily */ + smu_data->avfs.avfs_btc_status = AVFS_BTC_COMPLETED_PREVIOUSLY; + + return ret; +} + static int fiji_program_mem_timing_parameters(struct pp_hwmgr *hwmgr) { struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend); diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/fiji_smc.h b/drivers/gpu/drm/amd/powerplay/smumgr/fiji_smc.h index 0e9e1f2d7238..d9c72d992e30 100644 --- a/drivers/gpu/drm/amd/powerplay/smumgr/fiji_smc.h +++ b/drivers/gpu/drm/amd/powerplay/smumgr/fiji_smc.h @@ -48,5 +48,6 @@ int fiji_initialize_mc_reg_table(struct pp_hwmgr *hwmgr); bool fiji_is_dpm_running(struct pp_hwmgr *hwmgr); int fiji_populate_requested_graphic_levels(struct pp_hwmgr *hwmgr, struct amd_pp_profile *request); +int fiji_thermal_avfs_enable(struct pp_hwmgr *hwmgr); #endif diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/fiji_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/fiji_smumgr.c index a1cb78552cf6..6ae948fc524f 100644 --- a/drivers/gpu/drm/amd/powerplay/smumgr/fiji_smumgr.c +++ b/drivers/gpu/drm/amd/powerplay/smumgr/fiji_smumgr.c @@ -161,56 +161,47 @@ static int fiji_start_smu_in_non_protection_mode(struct pp_smumgr *smumgr) static int fiji_setup_pwr_virus(struct pp_smumgr *smumgr) { - int i, result = -1; + int i; + int result = -EINVAL; uint32_t reg, data; - const PWR_Command_Table *virus = PwrVirusTable; - struct fiji_smumgr *priv = (struct fiji_smumgr *)(smumgr->backend); - priv->avfs.AvfsBtcStatus = AVFS_LOAD_VIRUS; - for (i = 0; (i < PWR_VIRUS_TABLE_SIZE); i++) { - switch (virus->command) { + const PWR_Command_Table *pvirus = PwrVirusTable; + struct smu7_smumgr *smu_data = (struct smu7_smumgr *)(smumgr->backend); + + for (i = 0; i < PWR_VIRUS_TABLE_SIZE; i++) { + switch (pvirus->command) { case PwrCmdWrite: - reg = virus->reg; - data = virus->data; + reg = pvirus->reg; + data = pvirus->data; cgs_write_register(smumgr->device, reg, data); break; + case PwrCmdEnd: - priv->avfs.AvfsBtcStatus = AVFS_BTC_VIRUS_LOADED; result = 0; break; + default: - pr_err("Table Exit with Invalid Command!"); - priv->avfs.AvfsBtcStatus = AVFS_BTC_VIRUS_FAIL; - result = -1; + pr_info("Table Exit with Invalid Command!"); + smu_data->avfs.avfs_btc_status = AVFS_BTC_VIRUS_FAIL; + result = -EINVAL; break; } - virus++; + pvirus++; } + return result; } static int fiji_start_avfs_btc(struct pp_smumgr *smumgr) { int result = 0; - struct fiji_smumgr *priv = (struct fiji_smumgr *)(smumgr->backend); + struct smu7_smumgr *smu_data = (struct smu7_smumgr *)(smumgr->backend); - priv->avfs.AvfsBtcStatus = AVFS_BTC_STARTED; - if (priv->avfs.AvfsBtcParam) { - if (!smum_send_msg_to_smc_with_parameter(smumgr, - PPSMC_MSG_PerformBtc, priv->avfs.AvfsBtcParam)) { - if (!smum_send_msg_to_smc(smumgr, PPSMC_MSG_EnableAvfs)) { - priv->avfs.AvfsBtcStatus = AVFS_BTC_COMPLETED_UNSAVED; - result = 0; - } else { - pr_err("[AVFS][fiji_start_avfs_btc] Attempt" - " to Enable AVFS Failed!"); - smum_send_msg_to_smc(smumgr, PPSMC_MSG_DisableAvfs); - result = -1; - } - } else { - pr_err("[AVFS][fiji_start_avfs_btc] " - "PerformBTC SMU msg failed"); - result = -1; + if (0 != smu_data->avfs.avfs_btc_param) { + if (0 != smu7_send_msg_to_smc_with_parameter(smumgr, + PPSMC_MSG_PerformBtc, smu_data->avfs.avfs_btc_param)) { + pr_info("[AVFS][Fiji_PerformBtc] PerformBTC SMU msg failed"); + result = -EINVAL; } } /* Soft-Reset to reset the engine before loading uCode */ @@ -224,42 +215,6 @@ static int fiji_start_avfs_btc(struct pp_smumgr *smumgr) return result; } -static int fiji_setup_pm_fuse_for_avfs(struct pp_smumgr *smumgr) -{ - int result = 0; - uint32_t table_start; - uint32_t charz_freq_addr, inversion_voltage_addr, charz_freq; - uint16_t inversion_voltage; - - charz_freq = 0x30750000; /* In 10KHz units 0x00007530 Actual value */ - inversion_voltage = 0x1A04; /* mV Q14.2 0x41A Actual value */ - - PP_ASSERT_WITH_CODE(0 == smu7_read_smc_sram_dword(smumgr, - SMU7_FIRMWARE_HEADER_LOCATION + offsetof(SMU73_Firmware_Header, - PmFuseTable), &table_start, 0x40000), - "[AVFS][Fiji_SetupGfxLvlStruct] SMU could not communicate " - "starting address of PmFuse structure", - return -1;); - - charz_freq_addr = table_start + - offsetof(struct SMU73_Discrete_PmFuses, PsmCharzFreq); - inversion_voltage_addr = table_start + - offsetof(struct SMU73_Discrete_PmFuses, InversionVoltage); - - result = smu7_copy_bytes_to_smc(smumgr, charz_freq_addr, - (uint8_t *)(&charz_freq), sizeof(charz_freq), 0x40000); - PP_ASSERT_WITH_CODE(0 == result, - "[AVFS][fiji_setup_pm_fuse_for_avfs] charz_freq could not " - "be populated.", return -1;); - - result = smu7_copy_bytes_to_smc(smumgr, inversion_voltage_addr, - (uint8_t *)(&inversion_voltage), sizeof(inversion_voltage), 0x40000); - PP_ASSERT_WITH_CODE(0 == result, "[AVFS][fiji_setup_pm_fuse_for_avfs] " - "charz_freq could not be populated.", return -1;); - - return result; -} - static int fiji_setup_graphics_level_structure(struct pp_smumgr *smumgr) { int32_t vr_config; @@ -298,93 +253,41 @@ static int fiji_setup_graphics_level_structure(struct pp_smumgr *smumgr) return 0; } -/* Work in Progress */ -static int fiji_restore_vft_table(struct pp_smumgr *smumgr) -{ - struct fiji_smumgr *priv = (struct fiji_smumgr *)(smumgr->backend); - - if (AVFS_BTC_COMPLETED_SAVED == priv->avfs.AvfsBtcStatus) { - priv->avfs.AvfsBtcStatus = AVFS_BTC_COMPLETED_RESTORED; - return 0; - } else - return -EINVAL; -} - -/* Work in Progress */ -static int fiji_save_vft_table(struct pp_smumgr *smumgr) -{ - struct fiji_smumgr *priv = (struct fiji_smumgr *)(smumgr->backend); - - if (AVFS_BTC_COMPLETED_SAVED == priv->avfs.AvfsBtcStatus) { - priv->avfs.AvfsBtcStatus = AVFS_BTC_COMPLETED_RESTORED; - return 0; - } else - return -EINVAL; -} - static int fiji_avfs_event_mgr(struct pp_smumgr *smumgr, bool smu_started) { - struct fiji_smumgr *priv = (struct fiji_smumgr *)(smumgr->backend); + struct smu7_smumgr *smu_data = (struct smu7_smumgr *)(smumgr->backend); - switch (priv->avfs.AvfsBtcStatus) { - case AVFS_BTC_COMPLETED_SAVED: /*S3 State - Pre SMU Start */ - priv->avfs.AvfsBtcStatus = AVFS_BTC_RESTOREVFT_FAILED; - PP_ASSERT_WITH_CODE(0 == fiji_restore_vft_table(smumgr), - "[AVFS][fiji_avfs_event_mgr] Could not Copy Graphics " - "Level table over to SMU", - return -1;); - priv->avfs.AvfsBtcStatus = AVFS_BTC_COMPLETED_RESTORED; - break; - case AVFS_BTC_COMPLETED_RESTORED: /*S3 State - Post SMU Start*/ - priv->avfs.AvfsBtcStatus = AVFS_BTC_SMUMSG_ERROR; - PP_ASSERT_WITH_CODE(0 == smum_send_msg_to_smc(smumgr, - 0x666), - "[AVFS][fiji_avfs_event_mgr] SMU did not respond " - "correctly to VftTableIsValid Msg", - return -1;); - priv->avfs.AvfsBtcStatus = AVFS_BTC_SMUMSG_ERROR; - PP_ASSERT_WITH_CODE(0 == smum_send_msg_to_smc(smumgr, - PPSMC_MSG_EnableAvfs), - "[AVFS][fiji_avfs_event_mgr] SMU did not respond " - "correctly to EnableAvfs Message Msg", - return -1;); - priv->avfs.AvfsBtcStatus = AVFS_BTC_COMPLETED_SAVED; + switch (smu_data->avfs.avfs_btc_status) { + case AVFS_BTC_COMPLETED_PREVIOUSLY: break; + case AVFS_BTC_BOOT: /*Cold Boot State - Post SMU Start*/ if (!smu_started) break; - priv->avfs.AvfsBtcStatus = AVFS_BTC_FAILED; - PP_ASSERT_WITH_CODE(0 == fiji_setup_pm_fuse_for_avfs(smumgr), - "[AVFS][fiji_avfs_event_mgr] Failure at " - "fiji_setup_pm_fuse_for_avfs", - return -1;); - priv->avfs.AvfsBtcStatus = AVFS_BTC_DPMTABLESETUP_FAILED; + smu_data->avfs.avfs_btc_status = AVFS_BTC_FAILED; PP_ASSERT_WITH_CODE(0 == fiji_setup_graphics_level_structure(smumgr), "[AVFS][fiji_avfs_event_mgr] Could not Copy Graphics Level" " table over to SMU", - return -1;); - priv->avfs.AvfsBtcStatus = AVFS_BTC_VIRUS_FAIL; + return -EINVAL;); + smu_data->avfs.avfs_btc_status = AVFS_BTC_VIRUS_FAIL; PP_ASSERT_WITH_CODE(0 == fiji_setup_pwr_virus(smumgr), "[AVFS][fiji_avfs_event_mgr] Could not setup " "Pwr Virus for AVFS ", - return -1;); - priv->avfs.AvfsBtcStatus = AVFS_BTC_FAILED; + return -EINVAL;); + smu_data->avfs.avfs_btc_status = AVFS_BTC_FAILED; PP_ASSERT_WITH_CODE(0 == fiji_start_avfs_btc(smumgr), "[AVFS][fiji_avfs_event_mgr] Failure at " "fiji_start_avfs_btc. AVFS Disabled", - return -1;); - priv->avfs.AvfsBtcStatus = AVFS_BTC_SAVEVFT_FAILED; - PP_ASSERT_WITH_CODE(0 == fiji_save_vft_table(smumgr), - "[AVFS][fiji_avfs_event_mgr] Could not save VFT Table", - return -1;); - priv->avfs.AvfsBtcStatus = AVFS_BTC_COMPLETED_SAVED; + return -EINVAL;); + + smu_data->avfs.avfs_btc_status = AVFS_BTC_ENABLEAVFS; break; case AVFS_BTC_DISABLED: /* Do nothing */ - break; case AVFS_BTC_NOTSUPPORTED: /* Do nothing */ + case AVFS_BTC_ENABLEAVFS: break; default: - pr_err("[AVFS] Something is broken. See log!"); + pr_err("AVFS failed status is %x !\n", smu_data->avfs.avfs_btc_status); break; } return 0; @@ -477,19 +380,6 @@ static int fiji_smu_init(struct pp_smumgr *smumgr) if (smu7_init(smumgr)) return -EINVAL; - fiji_priv->avfs.AvfsBtcStatus = AVFS_BTC_BOOT; - if (fiji_is_hw_avfs_present(smumgr)) - /* AVFS Parameter - * 0 - BTC DC disabled, BTC AC disabled - * 1 - BTC DC enabled, BTC AC disabled - * 2 - BTC DC disabled, BTC AC enabled - * 3 - BTC DC enabled, BTC AC enabled - * Default is 0 - BTC DC disabled, BTC AC disabled - */ - fiji_priv->avfs.AvfsBtcParam = 0; - else - fiji_priv->avfs.AvfsBtcStatus = AVFS_BTC_NOTSUPPORTED; - for (i = 0; i < SMU73_MAX_LEVELS_GRAPHICS; i++) fiji_priv->activity_target[i] = 30; @@ -514,10 +404,12 @@ const struct pp_smumgr_func fiji_smu_funcs = { .init_smc_table = fiji_init_smc_table, .update_sclk_threshold = fiji_update_sclk_threshold, .thermal_setup_fan_table = fiji_thermal_setup_fan_table, + .thermal_avfs_enable = fiji_thermal_avfs_enable, .populate_all_graphic_levels = fiji_populate_all_graphic_levels, .populate_all_memory_levels = fiji_populate_all_memory_levels, .get_mac_definition = fiji_get_mac_definition, .initialize_mc_reg_table = fiji_initialize_mc_reg_table, .is_dpm_running = fiji_is_dpm_running, .populate_requested_graphic_levels = fiji_populate_requested_graphic_levels, + .is_hw_avfs_present = fiji_is_hw_avfs_present, }; diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/fiji_smumgr.h b/drivers/gpu/drm/amd/powerplay/smumgr/fiji_smumgr.h index adcbdfb209be..175bf9f8ef9c 100644 --- a/drivers/gpu/drm/amd/powerplay/smumgr/fiji_smumgr.h +++ b/drivers/gpu/drm/amd/powerplay/smumgr/fiji_smumgr.h @@ -28,17 +28,8 @@ #include "smu7_smumgr.h" - -struct fiji_smu_avfs { - enum AVFS_BTC_STATUS AvfsBtcStatus; - uint32_t AvfsBtcParam; -}; - - struct fiji_smumgr { struct smu7_smumgr smu7_data; - - struct fiji_smu_avfs avfs; struct SMU73_Discrete_DpmTable smc_state_table; struct SMU73_Discrete_Ulv ulv_setting; struct SMU73_Discrete_PmFuses power_tune_table; @@ -47,7 +38,5 @@ struct fiji_smumgr { }; - - #endif diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smc.c b/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smc.c index f68e759e8be2..99a00bd39256 100644 --- a/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smc.c +++ b/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smc.c @@ -1498,7 +1498,7 @@ static int polaris10_populate_avfs_parameters(struct pp_hwmgr *hwmgr) table_info->vdd_dep_on_sclk; - if (smu_data->avfs.avfs_btc_status == AVFS_BTC_NOTSUPPORTED) + if (((struct smu7_smumgr *)smu_data)->avfs.avfs_btc_status == AVFS_BTC_NOTSUPPORTED) return result; result = atomctrl_get_avfs_information(hwmgr, &avfs_params); @@ -1889,7 +1889,7 @@ int polaris10_thermal_avfs_enable(struct pp_hwmgr *hwmgr) { int ret; struct pp_smumgr *smumgr = (struct pp_smumgr *)(hwmgr->smumgr); - struct polaris10_smumgr *smu_data = (struct polaris10_smumgr *)(smumgr->backend); + struct smu7_smumgr *smu_data = (struct smu7_smumgr *)(smumgr->backend); struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend); if (smu_data->avfs.avfs_btc_status == AVFS_BTC_NOTSUPPORTED) diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smumgr.c index 9616cedc139c..75f43dadc56b 100644 --- a/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smumgr.c +++ b/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smumgr.c @@ -60,16 +60,14 @@ static const SMU74_Discrete_GraphicsLevel avfs_graphics_level_polaris10[8] = { static const SMU74_Discrete_MemoryLevel avfs_memory_level_polaris10 = { 0x100ea446, 0, 0x30750000, 0x01, 0x01, 0x01, 0x00, 0x00, 0x64, 0x00, 0x00, 0x1f00, 0x00, 0x00}; - static int polaris10_setup_pwr_virus(struct pp_smumgr *smumgr) { int i; - int result = -1; + int result = -EINVAL; uint32_t reg, data; const PWR_Command_Table *pvirus = pwr_virus_table; - struct polaris10_smumgr *smu_data = (struct polaris10_smumgr *)(smumgr->backend); - + struct smu7_smumgr *smu_data = (struct smu7_smumgr *)(smumgr->backend); for (i = 0; i < PWR_VIRUS_TABLE_SIZE; i++) { switch (pvirus->command) { @@ -86,7 +84,7 @@ static int polaris10_setup_pwr_virus(struct pp_smumgr *smumgr) default: pr_info("Table Exit with Invalid Command!"); smu_data->avfs.avfs_btc_status = AVFS_BTC_VIRUS_FAIL; - result = -1; + result = -EINVAL; break; } pvirus++; @@ -98,7 +96,7 @@ static int polaris10_setup_pwr_virus(struct pp_smumgr *smumgr) static int polaris10_perform_btc(struct pp_smumgr *smumgr) { int result = 0; - struct polaris10_smumgr *smu_data = (struct polaris10_smumgr *)(smumgr->backend); + struct smu7_smumgr *smu_data = (struct smu7_smumgr *)(smumgr->backend); if (0 != smu_data->avfs.avfs_btc_param) { if (0 != smu7_send_msg_to_smc_with_parameter(smumgr, PPSMC_MSG_PerformBtc, smu_data->avfs.avfs_btc_param)) { @@ -172,10 +170,11 @@ static int polaris10_setup_graphics_level_structure(struct pp_smumgr *smumgr) return 0; } + static int polaris10_avfs_event_mgr(struct pp_smumgr *smumgr, bool SMU_VFT_INTACT) { - struct polaris10_smumgr *smu_data = (struct polaris10_smumgr *)(smumgr->backend); + struct smu7_smumgr *smu_data = (struct smu7_smumgr *)(smumgr->backend); switch (smu_data->avfs.avfs_btc_status) { case AVFS_BTC_COMPLETED_PREVIOUSLY: @@ -185,30 +184,31 @@ polaris10_avfs_event_mgr(struct pp_smumgr *smumgr, bool SMU_VFT_INTACT) smu_data->avfs.avfs_btc_status = AVFS_BTC_DPMTABLESETUP_FAILED; PP_ASSERT_WITH_CODE(0 == polaris10_setup_graphics_level_structure(smumgr), - "[AVFS][Polaris10_AVFSEventMgr] Could not Copy Graphics Level table over to SMU", - return -1); + "[AVFS][Polaris10_AVFSEventMgr] Could not Copy Graphics Level table over to SMU", + return -EINVAL); if (smu_data->avfs.avfs_btc_param > 1) { pr_info("[AVFS][Polaris10_AVFSEventMgr] AC BTC has not been successfully verified on Fiji. There may be in this setting."); smu_data->avfs.avfs_btc_status = AVFS_BTC_VIRUS_FAIL; - PP_ASSERT_WITH_CODE(-1 == polaris10_setup_pwr_virus(smumgr), + PP_ASSERT_WITH_CODE(0 == polaris10_setup_pwr_virus(smumgr), "[AVFS][Polaris10_AVFSEventMgr] Could not setup Pwr Virus for AVFS ", - return -1); + return -EINVAL); } smu_data->avfs.avfs_btc_status = AVFS_BTC_FAILED; PP_ASSERT_WITH_CODE(0 == polaris10_perform_btc(smumgr), "[AVFS][Polaris10_AVFSEventMgr] Failure at SmuPolaris10_PerformBTC. AVFS Disabled", - return -1); - + return -EINVAL); + smu_data->avfs.avfs_btc_status = AVFS_BTC_ENABLEAVFS; break; case AVFS_BTC_DISABLED: + case AVFS_BTC_ENABLEAVFS: case AVFS_BTC_NOTSUPPORTED: break; default: - pr_info("[AVFS] Something is broken. See log!"); + pr_err("AVFS failed status is %x!\n", smu_data->avfs.avfs_btc_status); break; } @@ -376,11 +376,6 @@ static int polaris10_smu_init(struct pp_smumgr *smumgr) if (smu7_init(smumgr)) return -EINVAL; - if (polaris10_is_hw_avfs_present(smumgr)) - smu_data->avfs.avfs_btc_status = AVFS_BTC_BOOT; - else - smu_data->avfs.avfs_btc_status = AVFS_BTC_NOTSUPPORTED; - for (i = 0; i < SMU74_MAX_LEVELS_GRAPHICS; i++) smu_data->activity_target[i] = PPPOLARIS10_TARGETACTIVITY_DFLT; @@ -410,4 +405,5 @@ const struct pp_smumgr_func polaris10_smu_funcs = { .get_mac_definition = polaris10_get_mac_definition, .is_dpm_running = polaris10_is_dpm_running, .populate_requested_graphic_levels = polaris10_populate_requested_graphic_levels, + .is_hw_avfs_present = polaris10_is_hw_avfs_present, }; diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smumgr.h b/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smumgr.h index 49ebf1d5a53c..5e19c24b0561 100644 --- a/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smumgr.h +++ b/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smumgr.h @@ -32,11 +32,6 @@ #define SMC_RAM_END 0x40000 -struct polaris10_avfs { - enum AVFS_BTC_STATUS avfs_btc_status; - uint32_t avfs_btc_param; -}; - struct polaris10_pt_defaults { uint8_t SviLoadLineEn; uint8_t SviLoadLineVddC; @@ -51,8 +46,6 @@ struct polaris10_pt_defaults { uint16_t BAPMTI_RC[SMU74_DTE_ITERATIONS * SMU74_DTE_SOURCES * SMU74_DTE_SINKS]; }; - - struct polaris10_range_table { uint32_t trans_lower_frequency; /* in 10khz */ uint32_t trans_upper_frequency; @@ -61,14 +54,13 @@ struct polaris10_range_table { struct polaris10_smumgr { struct smu7_smumgr smu7_data; uint8_t protected_mode; - struct polaris10_avfs avfs; SMU74_Discrete_DpmTable smc_state_table; struct SMU74_Discrete_Ulv ulv_setting; struct SMU74_Discrete_PmFuses power_tune_table; struct polaris10_range_table range_table[NUM_SCLK_RANGE]; const struct polaris10_pt_defaults *power_tune_defaults; - uint32_t activity_target[SMU74_MAX_LEVELS_GRAPHICS]; - uint32_t bif_sclk_table[SMU74_MAX_LEVELS_LINK]; + uint32_t activity_target[SMU74_MAX_LEVELS_GRAPHICS]; + uint32_t bif_sclk_table[SMU74_MAX_LEVELS_LINK]; }; diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.c index 35ac27681415..76347ff6d655 100644 --- a/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.c +++ b/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.c @@ -540,7 +540,6 @@ int smu7_upload_smu_firmware_image(struct pp_smumgr *smumgr) return result; } - int smu7_init(struct pp_smumgr *smumgr) { struct smu7_smumgr *smu_data; @@ -596,6 +595,11 @@ int smu7_init(struct pp_smumgr *smumgr) (cgs_handle_t)smu_data->smu_buffer.handle); return -EINVAL); + if (smum_is_hw_avfs_present(smumgr)) + smu_data->avfs.avfs_btc_status = AVFS_BTC_BOOT; + else + smu_data->avfs.avfs_btc_status = AVFS_BTC_NOTSUPPORTED; + return 0; } diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.h b/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.h index 919be435b49c..ee5e32d2921e 100644 --- a/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.h +++ b/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.h @@ -37,6 +37,11 @@ struct smu7_buffer_entry { unsigned long handle; }; +struct smu7_avfs { + enum AVFS_BTC_STATUS avfs_btc_status; + uint32_t avfs_btc_param; +}; + struct smu7_smumgr { uint8_t *header; uint8_t *mec_image; @@ -50,7 +55,8 @@ struct smu7_smumgr { uint32_t arb_table_start; uint32_t ulv_setting_starts; uint8_t security_hard_key; - uint32_t acpi_optimization; + uint32_t acpi_optimization; + struct smu7_avfs avfs; }; diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/smumgr.c index bcc61ffd13cb..3bdf6478de7f 100644 --- a/drivers/gpu/drm/amd/powerplay/smumgr/smumgr.c +++ b/drivers/gpu/drm/amd/powerplay/smumgr/smumgr.c @@ -43,7 +43,8 @@ MODULE_FIRMWARE("amdgpu/polaris11_smc.bin"); MODULE_FIRMWARE("amdgpu/polaris11_smc_sk.bin"); MODULE_FIRMWARE("amdgpu/polaris11_k_smc.bin"); MODULE_FIRMWARE("amdgpu/polaris12_smc.bin"); - +MODULE_FIRMWARE("amdgpu/vega10_smc.bin"); +MODULE_FIRMWARE("amdgpu/vega10_acg_smc.bin"); int smum_early_init(struct pp_instance *handle) { @@ -403,3 +404,11 @@ int smum_populate_requested_graphic_levels(struct pp_hwmgr *hwmgr, return 0; } + +bool smum_is_hw_avfs_present(struct pp_smumgr *smumgr) +{ + if (smumgr->smumgr_funcs->is_hw_avfs_present) + return smumgr->smumgr_funcs->is_hw_avfs_present(smumgr); + + return false; +} diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/vega10_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/vega10_smumgr.c index 269678443862..408514c965a0 100644 --- a/drivers/gpu/drm/amd/powerplay/smumgr/vega10_smumgr.c +++ b/drivers/gpu/drm/amd/powerplay/smumgr/vega10_smumgr.c @@ -356,6 +356,9 @@ int vega10_set_tools_address(struct pp_smumgr *smumgr) static int vega10_verify_smc_interface(struct pp_smumgr *smumgr) { uint32_t smc_driver_if_version; + struct cgs_system_info sys_info = {0}; + uint32_t dev_id; + uint32_t rev_id; PP_ASSERT_WITH_CODE(!vega10_send_msg_to_smc(smumgr, PPSMC_MSG_GetDriverIfVersion), @@ -363,12 +366,27 @@ static int vega10_verify_smc_interface(struct pp_smumgr *smumgr) 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; + sys_info.size = sizeof(struct cgs_system_info); + sys_info.info_id = CGS_SYSTEM_INFO_PCIE_DEV; + cgs_query_system_info(smumgr->device, &sys_info); + dev_id = (uint32_t)sys_info.value; + + sys_info.size = sizeof(struct cgs_system_info); + sys_info.info_id = CGS_SYSTEM_INFO_PCIE_REV; + cgs_query_system_info(smumgr->device, &sys_info); + rev_id = (uint32_t)sys_info.value; + + if (!((dev_id == 0x687f) && + ((rev_id == 0xc0) || + (rev_id == 0xc1) || + (rev_id == 0xc3)))) { + 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; diff --git a/drivers/gpu/drm/amd/scheduler/gpu_sched_trace.h b/drivers/gpu/drm/amd/scheduler/gpu_sched_trace.h index dbd4fd3a810b..8bd38102b58e 100644 --- a/drivers/gpu/drm/amd/scheduler/gpu_sched_trace.h +++ b/drivers/gpu/drm/amd/scheduler/gpu_sched_trace.h @@ -16,16 +16,16 @@ TRACE_EVENT(amd_sched_job, TP_ARGS(sched_job), TP_STRUCT__entry( __field(struct amd_sched_entity *, entity) - __field(struct amd_sched_job *, sched_job) __field(struct dma_fence *, fence) __field(const char *, name) + __field(uint64_t, id) __field(u32, job_count) __field(int, hw_job_count) ), TP_fast_assign( __entry->entity = sched_job->s_entity; - __entry->sched_job = sched_job; + __entry->id = sched_job->id; __entry->fence = &sched_job->s_fence->finished; __entry->name = sched_job->sched->name; __entry->job_count = kfifo_len( @@ -33,8 +33,9 @@ TRACE_EVENT(amd_sched_job, __entry->hw_job_count = atomic_read( &sched_job->sched->hw_rq_count); ), - TP_printk("entity=%p, sched job=%p, fence=%p, ring=%s, job count:%u, hw job count:%d", - __entry->entity, __entry->sched_job, __entry->fence, __entry->name, + TP_printk("entity=%p, id=%llu, fence=%p, ring=%s, job count:%u, hw job count:%d", + __entry->entity, __entry->id, + __entry->fence, __entry->name, __entry->job_count, __entry->hw_job_count) ); diff --git a/drivers/gpu/drm/arc/arcpgu_crtc.c b/drivers/gpu/drm/arc/arcpgu_crtc.c index 799416651f2f..16903dc7fe0d 100644 --- a/drivers/gpu/drm/arc/arcpgu_crtc.c +++ b/drivers/gpu/drm/arc/arcpgu_crtc.c @@ -69,12 +69,13 @@ static enum drm_mode_status arc_pgu_crtc_mode_valid(struct drm_crtc *crtc, { struct arcpgu_drm_private *arcpgu = crtc_to_arcpgu_priv(crtc); long rate, clk_rate = mode->clock * 1000; + long diff = clk_rate / 200; /* +-0.5% allowed by HDMI spec */ rate = clk_round_rate(arcpgu->clk, clk_rate); - if (rate != clk_rate) - return MODE_NOCLOCK; + if ((max(rate, clk_rate) - min(rate, clk_rate) < diff) && (rate > 0)) + return MODE_OK; - return MODE_OK; + return MODE_NOCLOCK; } static void arc_pgu_crtc_mode_set_nofb(struct drm_crtc *crtc) diff --git a/drivers/gpu/drm/arc/arcpgu_drv.c b/drivers/gpu/drm/arc/arcpgu_drv.c index 415b7dae0960..e3c13aa202b8 100644 --- a/drivers/gpu/drm/arc/arcpgu_drv.c +++ b/drivers/gpu/drm/arc/arcpgu_drv.c @@ -48,29 +48,7 @@ static void arcpgu_setup_mode_config(struct drm_device *drm) drm->mode_config.funcs = &arcpgu_drm_modecfg_funcs; } -static int arcpgu_gem_mmap(struct file *filp, struct vm_area_struct *vma) -{ - int ret; - - ret = drm_gem_mmap(filp, vma); - if (ret) - return ret; - - vma->vm_page_prot = pgprot_noncached(vm_get_page_prot(vma->vm_flags)); - return 0; -} - -static const struct file_operations arcpgu_drm_ops = { - .owner = THIS_MODULE, - .open = drm_open, - .release = drm_release, - .unlocked_ioctl = drm_ioctl, - .compat_ioctl = drm_compat_ioctl, - .poll = drm_poll, - .read = drm_read, - .llseek = no_llseek, - .mmap = arcpgu_gem_mmap, -}; +DEFINE_DRM_GEM_CMA_FOPS(arcpgu_drm_ops); static void arcpgu_lastclose(struct drm_device *drm) { @@ -142,7 +120,7 @@ static int arcpgu_load(struct drm_device *drm) return -ENODEV; } - platform_set_drvdata(pdev, arcpgu); + platform_set_drvdata(pdev, drm); return 0; } @@ -160,11 +138,37 @@ static int arcpgu_unload(struct drm_device *drm) return 0; } +#ifdef CONFIG_DEBUG_FS +static int arcpgu_show_pxlclock(struct seq_file *m, void *arg) +{ + struct drm_info_node *node = (struct drm_info_node *)m->private; + struct drm_device *drm = node->minor->dev; + struct arcpgu_drm_private *arcpgu = drm->dev_private; + unsigned long clkrate = clk_get_rate(arcpgu->clk); + unsigned long mode_clock = arcpgu->crtc.mode.crtc_clock * 1000; + + seq_printf(m, "hw : %lu\n", clkrate); + seq_printf(m, "mode: %lu\n", mode_clock); + return 0; +} + +static struct drm_info_list arcpgu_debugfs_list[] = { + { "clocks", arcpgu_show_pxlclock, 0 }, + { "fb", drm_fb_cma_debugfs_show, 0 }, +}; + +static int arcpgu_debugfs_init(struct drm_minor *minor) +{ + return drm_debugfs_create_files(arcpgu_debugfs_list, + ARRAY_SIZE(arcpgu_debugfs_list), minor->debugfs_root, minor); +} +#endif + static struct drm_driver arcpgu_drm_driver = { .driver_features = DRIVER_MODESET | DRIVER_GEM | DRIVER_PRIME | DRIVER_ATOMIC, .lastclose = arcpgu_lastclose, - .name = "drm-arcpgu", + .name = "arcpgu", .desc = "ARC PGU Controller", .date = "20160219", .major = 1, @@ -183,6 +187,9 @@ static struct drm_driver arcpgu_drm_driver = { .gem_prime_vmap = drm_gem_cma_prime_vmap, .gem_prime_vunmap = drm_gem_cma_prime_vunmap, .gem_prime_mmap = drm_gem_cma_prime_mmap, +#ifdef CONFIG_DEBUG_FS + .debugfs_init = arcpgu_debugfs_init, +#endif }; static int arcpgu_probe(struct platform_device *pdev) diff --git a/drivers/gpu/drm/ast/ast_dp501.c b/drivers/gpu/drm/ast/ast_dp501.c index 76f07f38b941..749646ae365f 100644 --- a/drivers/gpu/drm/ast/ast_dp501.c +++ b/drivers/gpu/drm/ast/ast_dp501.c @@ -4,16 +4,11 @@ #include "ast_drv.h" MODULE_FIRMWARE("ast_dp501_fw.bin"); -int ast_load_dp501_microcode(struct drm_device *dev) +static int ast_load_dp501_microcode(struct drm_device *dev) { struct ast_private *ast = dev->dev_private; - static char *fw_name = "ast_dp501_fw.bin"; - int err; - err = request_firmware(&ast->dp501_fw, fw_name, dev->dev); - if (err) - return err; - return 0; + return request_firmware(&ast->dp501_fw, "ast_dp501_fw.bin", dev->dev); } static void send_ack(struct ast_private *ast) @@ -187,7 +182,7 @@ bool ast_backup_fw(struct drm_device *dev, u8 *addr, u32 size) return false; } -bool ast_launch_m68k(struct drm_device *dev) +static bool ast_launch_m68k(struct drm_device *dev) { struct ast_private *ast = dev->dev_private; u32 i, data, len = 0; @@ -201,7 +196,11 @@ bool ast_launch_m68k(struct drm_device *dev) if (ast->dp501_fw_addr) { fw_addr = ast->dp501_fw_addr; len = 32*1024; - } else if (ast->dp501_fw) { + } else { + if (!ast->dp501_fw && + ast_load_dp501_microcode(dev) < 0) + return false; + fw_addr = (u8 *)ast->dp501_fw->data; len = ast->dp501_fw->size; } @@ -432,3 +431,11 @@ void ast_init_3rdtx(struct drm_device *dev) } } } + +void ast_release_firmware(struct drm_device *dev) +{ + struct ast_private *ast = dev->dev_private; + + release_firmware(ast->dp501_fw); + ast->dp501_fw = NULL; +} diff --git a/drivers/gpu/drm/ast/ast_drv.h b/drivers/gpu/drm/ast/ast_drv.h index 569a1484d523..e6c4cd3dc50e 100644 --- a/drivers/gpu/drm/ast/ast_drv.h +++ b/drivers/gpu/drm/ast/ast_drv.h @@ -400,11 +400,10 @@ void ast_post_gpu(struct drm_device *dev); u32 ast_mindwm(struct ast_private *ast, u32 r); void ast_moutdwm(struct ast_private *ast, u32 r, u32 v); /* ast dp501 */ -int ast_load_dp501_microcode(struct drm_device *dev); void ast_set_dp501_video_output(struct drm_device *dev, u8 mode); -bool ast_launch_m68k(struct drm_device *dev); bool ast_backup_fw(struct drm_device *dev, u8 *addr, u32 size); bool ast_dp501_read_edid(struct drm_device *dev, u8 *ediddata); u8 ast_get_dp501_max_clk(struct drm_device *dev); void ast_init_3rdtx(struct drm_device *dev); +void ast_release_firmware(struct drm_device *dev); #endif diff --git a/drivers/gpu/drm/ast/ast_main.c b/drivers/gpu/drm/ast/ast_main.c index 262c2c0e43b4..9a44cdec3bca 100644 --- a/drivers/gpu/drm/ast/ast_main.c +++ b/drivers/gpu/drm/ast/ast_main.c @@ -389,7 +389,7 @@ static void ast_user_framebuffer_destroy(struct drm_framebuffer *fb) drm_gem_object_unreference_unlocked(ast_fb->obj); drm_framebuffer_cleanup(fb); - kfree(fb); + kfree(ast_fb); } static const struct drm_framebuffer_funcs ast_fb_funcs = { @@ -576,6 +576,7 @@ void ast_driver_unload(struct drm_device *dev) { struct ast_private *ast = dev->dev_private; + ast_release_firmware(dev); kfree(ast->dp501_fw_addr); ast_mode_fini(dev); ast_fbdev_fini(dev); diff --git a/drivers/gpu/drm/ast/ast_mode.c b/drivers/gpu/drm/ast/ast_mode.c index 724c16bb6a62..43245229f437 100644 --- a/drivers/gpu/drm/ast/ast_mode.c +++ b/drivers/gpu/drm/ast/ast_mode.c @@ -616,7 +616,23 @@ static int ast_crtc_mode_set(struct drm_crtc *crtc, static void ast_crtc_disable(struct drm_crtc *crtc) { + int ret; + + DRM_DEBUG_KMS("\n"); + ast_crtc_dpms(crtc, DRM_MODE_DPMS_OFF); + if (crtc->primary->fb) { + struct ast_framebuffer *ast_fb = to_ast_framebuffer(crtc->primary->fb); + struct drm_gem_object *obj = ast_fb->obj; + struct ast_bo *bo = gem_to_ast_bo(obj); + + ret = ast_bo_reserve(bo, false); + if (ret) + return; + ast_bo_push_sysram(bo); + ast_bo_unreserve(bo); + } + crtc->primary->fb = NULL; } static void ast_crtc_prepare(struct drm_crtc *crtc) diff --git a/drivers/gpu/drm/ast/ast_ttm.c b/drivers/gpu/drm/ast/ast_ttm.c index 58084985e6cf..696a15dc2f3f 100644 --- a/drivers/gpu/drm/ast/ast_ttm.c +++ b/drivers/gpu/drm/ast/ast_ttm.c @@ -323,10 +323,8 @@ int ast_bo_create(struct drm_device *dev, int size, int align, return -ENOMEM; ret = drm_gem_object_init(dev, &astbo->gem, size); - if (ret) { - kfree(astbo); - return ret; - } + if (ret) + goto error; astbo->bo.bdev = &ast->ttm.bdev; @@ -340,10 +338,13 @@ int ast_bo_create(struct drm_device *dev, int size, int align, align >> PAGE_SHIFT, false, NULL, acc_size, NULL, NULL, ast_bo_ttm_destroy); if (ret) - return ret; + goto error; *pastbo = astbo; return 0; +error: + kfree(astbo); + return ret; } static inline u64 ast_bo_gpu_offset(struct ast_bo *bo) @@ -376,7 +377,7 @@ int ast_bo_pin(struct ast_bo *bo, u32 pl_flag, u64 *gpu_addr) int ast_bo_unpin(struct ast_bo *bo) { - int i, ret; + int i; if (!bo->pin_count) { DRM_ERROR("unpin bad %p\n", bo); return 0; @@ -387,11 +388,7 @@ int ast_bo_unpin(struct ast_bo *bo) for (i = 0; i < bo->placement.num_placement ; i++) bo->placements[i].flags &= ~TTM_PL_FLAG_NO_EVICT; - ret = ttm_bo_validate(&bo->bo, &bo->placement, false, false); - if (ret) - return ret; - - return 0; + return ttm_bo_validate(&bo->bo, &bo->placement, false, false); } int ast_bo_push_sysram(struct ast_bo *bo) diff --git a/drivers/gpu/drm/radeon/radeon_display.c b/drivers/gpu/drm/radeon/radeon_display.c index 8b7d7a0d3ca8..ee274c6e374d 100644 --- a/drivers/gpu/drm/radeon/radeon_display.c +++ b/drivers/gpu/drm/radeon/radeon_display.c @@ -1373,12 +1373,12 @@ static const struct drm_mode_config_funcs radeon_mode_funcs = { .output_poll_changed = radeon_output_poll_changed }; -static struct drm_prop_enum_list radeon_tmds_pll_enum_list[] = +static const struct drm_prop_enum_list radeon_tmds_pll_enum_list[] = { { 0, "driver" }, { 1, "bios" }, }; -static struct drm_prop_enum_list radeon_tv_std_enum_list[] = +static const struct drm_prop_enum_list radeon_tv_std_enum_list[] = { { TV_STD_NTSC, "ntsc" }, { TV_STD_PAL, "pal" }, { TV_STD_PAL_M, "pal-m" }, @@ -1389,25 +1389,25 @@ static struct drm_prop_enum_list radeon_tv_std_enum_list[] = { TV_STD_SECAM, "secam" }, }; -static struct drm_prop_enum_list radeon_underscan_enum_list[] = +static const struct drm_prop_enum_list radeon_underscan_enum_list[] = { { UNDERSCAN_OFF, "off" }, { UNDERSCAN_ON, "on" }, { UNDERSCAN_AUTO, "auto" }, }; -static struct drm_prop_enum_list radeon_audio_enum_list[] = +static const struct drm_prop_enum_list radeon_audio_enum_list[] = { { RADEON_AUDIO_DISABLE, "off" }, { RADEON_AUDIO_ENABLE, "on" }, { RADEON_AUDIO_AUTO, "auto" }, }; /* XXX support different dither options? spatial, temporal, both, etc. */ -static struct drm_prop_enum_list radeon_dither_enum_list[] = +static const struct drm_prop_enum_list radeon_dither_enum_list[] = { { RADEON_FMT_DITHER_DISABLE, "off" }, { RADEON_FMT_DITHER_ENABLE, "on" }, }; -static struct drm_prop_enum_list radeon_output_csc_enum_list[] = +static const struct drm_prop_enum_list radeon_output_csc_enum_list[] = { { RADEON_OUTPUT_CSC_BYPASS, "bypass" }, { RADEON_OUTPUT_CSC_TVRGB, "tvrgb" }, { RADEON_OUTPUT_CSC_YCBCR601, "ycbcr601" }, diff --git a/drivers/gpu/drm/radeon/radeon_fb.c b/drivers/gpu/drm/radeon/radeon_fb.c index 4b8f072d480a..af6ee7d9b465 100644 --- a/drivers/gpu/drm/radeon/radeon_fb.c +++ b/drivers/gpu/drm/radeon/radeon_fb.c @@ -344,9 +344,12 @@ int radeon_fbdev_init(struct radeon_device *rdev) if (list_empty(&rdev->ddev->mode_config.connector_list)) return 0; - /* select 8 bpp console on RN50 or 16MB cards */ - if (ASIC_IS_RN50(rdev) || rdev->mc.real_vram_size <= (32*1024*1024)) + /* select 8 bpp console on 8MB cards, or 16 bpp on RN50 or 32MB */ + if (rdev->mc.real_vram_size <= (8*1024*1024)) bpp_sel = 8; + else if (ASIC_IS_RN50(rdev) || + rdev->mc.real_vram_size <= (32*1024*1024)) + bpp_sel = 16; rfbdev = kzalloc(sizeof(struct radeon_fbdev), GFP_KERNEL); if (!rfbdev) diff --git a/drivers/gpu/drm/radeon/radeon_irq_kms.c b/drivers/gpu/drm/radeon/radeon_irq_kms.c index fff0d11b0600..afaf10db47cc 100644 --- a/drivers/gpu/drm/radeon/radeon_irq_kms.c +++ b/drivers/gpu/drm/radeon/radeon_irq_kms.c @@ -283,6 +283,10 @@ int radeon_irq_kms_init(struct radeon_device *rdev) int r = 0; spin_lock_init(&rdev->irq.lock); + + /* Disable vblank irqs aggressively for power-saving */ + rdev->ddev->vblank_disable_immediate = true; + r = drm_vblank_init(rdev->ddev, rdev->num_crtc); if (r) { return r; diff --git a/drivers/gpu/drm/radeon/radeon_ttm.c b/drivers/gpu/drm/radeon/radeon_ttm.c index faa021396da3..2804b4a15896 100644 --- a/drivers/gpu/drm/radeon/radeon_ttm.c +++ b/drivers/gpu/drm/radeon/radeon_ttm.c @@ -178,7 +178,7 @@ static int radeon_init_mem_type(struct ttm_bo_device *bdev, uint32_t type, static void radeon_evict_flags(struct ttm_buffer_object *bo, struct ttm_placement *placement) { - static struct ttm_place placements = { + static const struct ttm_place placements = { .fpfn = 0, .lpfn = 0, .flags = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM diff --git a/drivers/gpu/drm/radeon/vce_v2_0.c b/drivers/gpu/drm/radeon/vce_v2_0.c index fce214482e72..b0a43b68776d 100644 --- a/drivers/gpu/drm/radeon/vce_v2_0.c +++ b/drivers/gpu/drm/radeon/vce_v2_0.c @@ -104,6 +104,10 @@ static void vce_v2_0_disable_cg(struct radeon_device *rdev) WREG32(VCE_CGTT_CLK_OVERRIDE, 7); } +/* + * Local variable sw_cg is used for debugging purposes, in case we + * ran into problems with dynamic clock gating. Don't remove it. + */ void vce_v2_0_enable_mgcg(struct radeon_device *rdev, bool enable) { bool sw_cg = false; diff --git a/drivers/gpu/drm/rcar-du/rcar_du_crtc.c b/drivers/gpu/drm/rcar-du/rcar_du_crtc.c index f131fc68cc46..301ea1a8018e 100644 --- a/drivers/gpu/drm/rcar-du/rcar_du_crtc.c +++ b/drivers/gpu/drm/rcar-du/rcar_du_crtc.c @@ -13,6 +13,7 @@ #include <linux/clk.h> #include <linux/mutex.h> +#include <linux/sys_soc.h> #include <drm/drmP.h> #include <drm/drm_atomic.h> @@ -129,10 +130,8 @@ static void rcar_du_dpll_divider(struct rcar_du_crtc *rcrtc, for (fdpll = 1; fdpll < 32; fdpll++) { unsigned long output; - /* 1/2 (FRQSEL=1) for duty rate 50% */ output = input * (n + 1) / (m + 1) - / (fdpll + 1) / 2; - + / (fdpll + 1); if (output >= 400000000) continue; @@ -158,6 +157,11 @@ done: best_diff); } +static const struct soc_device_attribute rcar_du_r8a7795_es1[] = { + { .soc_id = "r8a7795", .revision = "ES1.*" }, + { /* sentinel */ } +}; + static void rcar_du_crtc_set_display_timing(struct rcar_du_crtc *rcrtc) { const struct drm_display_mode *mode = &rcrtc->crtc.state->adjusted_mode; @@ -168,7 +172,8 @@ static void rcar_du_crtc_set_display_timing(struct rcar_du_crtc *rcrtc) u32 escr; u32 div; - /* Compute the clock divisor and select the internal or external dot + /* + * Compute the clock divisor and select the internal or external dot * clock based on the requested frequency. */ clk = clk_get_rate(rcrtc->clock); @@ -185,7 +190,20 @@ static void rcar_du_crtc_set_display_timing(struct rcar_du_crtc *rcrtc) extclk = clk_get_rate(rcrtc->extclock); if (rcdu->info->dpll_ch & (1 << rcrtc->index)) { - rcar_du_dpll_divider(rcrtc, &dpll, extclk, mode_clock); + unsigned long target = mode_clock; + + /* + * The H3 ES1.x exhibits dot clock duty cycle stability + * issues. We can work around them by configuring the + * DPLL to twice the desired frequency, coupled with a + * /2 post-divider. This isn't needed on other SoCs and + * breaks HDMI output on M3-W for a currently unknown + * reason, so restrict the workaround to H3 ES1.x. + */ + if (soc_device_match(rcar_du_r8a7795_es1)) + target *= 2; + + rcar_du_dpll_divider(rcrtc, &dpll, extclk, target); extclk = dpll.output; } @@ -197,8 +215,6 @@ static void rcar_du_crtc_set_display_timing(struct rcar_du_crtc *rcrtc) if (abs((long)extrate - (long)mode_clock) < abs((long)rate - (long)mode_clock)) { - dev_dbg(rcrtc->group->dev->dev, - "crtc%u: using external clock\n", rcrtc->index); if (rcdu->info->dpll_ch & (1 << rcrtc->index)) { u32 dpllcr = DPLLCR_CODE | DPLLCR_CLKE @@ -215,12 +231,14 @@ static void rcar_du_crtc_set_display_timing(struct rcar_du_crtc *rcrtc) rcar_du_group_write(rcrtc->group, DPLLCR, dpllcr); - - escr = ESCR_DCLKSEL_DCLKIN | 1; - } else { - escr = ESCR_DCLKSEL_DCLKIN | extdiv; } + + escr = ESCR_DCLKSEL_DCLKIN | extdiv; } + + dev_dbg(rcrtc->group->dev->dev, + "mode clock %lu extrate %lu rate %lu ESCR 0x%08x\n", + mode_clock, extrate, rate, escr); } rcar_du_group_write(rcrtc->group, rcrtc->index % 2 ? ESCR2 : ESCR, @@ -261,12 +279,14 @@ void rcar_du_crtc_route_output(struct drm_crtc *crtc, struct rcar_du_crtc *rcrtc = to_rcar_crtc(crtc); struct rcar_du_device *rcdu = rcrtc->group->dev; - /* Store the route from the CRTC output to the DU output. The DU will be + /* + * Store the route from the CRTC output to the DU output. The DU will be * configured when starting the CRTC. */ rcrtc->outputs |= BIT(output); - /* Store RGB routing to DPAD0, the hardware will be configured when + /* + * Store RGB routing to DPAD0, the hardware will be configured when * starting the CRTC. */ if (output == RCAR_DU_OUTPUT_DPAD0) @@ -342,7 +362,8 @@ static void rcar_du_crtc_update_planes(struct rcar_du_crtc *rcrtc) } } - /* Update the planes to display timing and dot clock generator + /* + * Update the planes to display timing and dot clock generator * associations. * * Updating the DPTSR register requires restarting the CRTC group, @@ -431,14 +452,8 @@ static void rcar_du_crtc_wait_page_flip(struct rcar_du_crtc *rcrtc) * Start/Stop and Suspend/Resume */ -static void rcar_du_crtc_start(struct rcar_du_crtc *rcrtc) +static void rcar_du_crtc_setup(struct rcar_du_crtc *rcrtc) { - struct drm_crtc *crtc = &rcrtc->crtc; - bool interlaced; - - if (rcrtc->started) - return; - /* Set display off and background to black */ rcar_du_crtc_write(rcrtc, DOOR, DOOR_RGB(0, 0, 0)); rcar_du_crtc_write(rcrtc, BPOR, BPOR_RGB(0, 0, 0)); @@ -450,7 +465,20 @@ static void rcar_du_crtc_start(struct rcar_du_crtc *rcrtc) /* Start with all planes disabled. */ rcar_du_group_write(rcrtc->group, rcrtc->index % 2 ? DS2PR : DS1PR, 0); - /* Select master sync mode. This enables display operation in master + /* Enable the VSP compositor. */ + if (rcar_du_has(rcrtc->group->dev, RCAR_DU_FEATURE_VSP1_SOURCE)) + rcar_du_vsp_enable(rcrtc); + + /* Turn vertical blanking interrupt reporting on. */ + drm_crtc_vblank_on(&rcrtc->crtc); +} + +static void rcar_du_crtc_start(struct rcar_du_crtc *rcrtc) +{ + bool interlaced; + + /* + * Select master sync mode. This enables display operation in master * sync mode (with the HSYNC and VSYNC signals configured as outputs and * actively driven). */ @@ -460,38 +488,56 @@ static void rcar_du_crtc_start(struct rcar_du_crtc *rcrtc) DSYSR_TVM_MASTER); rcar_du_group_start_stop(rcrtc->group, true); +} - /* Enable the VSP compositor. */ - if (rcar_du_has(rcrtc->group->dev, RCAR_DU_FEATURE_VSP1_SOURCE)) - rcar_du_vsp_enable(rcrtc); +static void rcar_du_crtc_disable_planes(struct rcar_du_crtc *rcrtc) +{ + struct rcar_du_device *rcdu = rcrtc->group->dev; + struct drm_crtc *crtc = &rcrtc->crtc; + u32 status; - /* Turn vertical blanking interrupt reporting back on. */ - drm_crtc_vblank_on(crtc); + /* Make sure vblank interrupts are enabled. */ + drm_crtc_vblank_get(crtc); - rcrtc->started = true; + /* + * Disable planes and calculate how many vertical blanking interrupts we + * have to wait for. If a vertical blanking interrupt has been triggered + * but not processed yet, we don't know whether it occurred before or + * after the planes got disabled. We thus have to wait for two vblank + * interrupts in that case. + */ + spin_lock_irq(&rcrtc->vblank_lock); + rcar_du_group_write(rcrtc->group, rcrtc->index % 2 ? DS2PR : DS1PR, 0); + status = rcar_du_crtc_read(rcrtc, DSSR); + rcrtc->vblank_count = status & DSSR_VBK ? 2 : 1; + spin_unlock_irq(&rcrtc->vblank_lock); + + if (!wait_event_timeout(rcrtc->vblank_wait, rcrtc->vblank_count == 0, + msecs_to_jiffies(100))) + dev_warn(rcdu->dev, "vertical blanking timeout\n"); + + drm_crtc_vblank_put(crtc); } static void rcar_du_crtc_stop(struct rcar_du_crtc *rcrtc) { struct drm_crtc *crtc = &rcrtc->crtc; - if (!rcrtc->started) - return; - - /* Disable all planes and wait for the change to take effect. This is - * required as the DSnPR registers are updated on vblank, and no vblank - * will occur once the CRTC is stopped. Disabling planes when starting - * the CRTC thus wouldn't be enough as it would start scanning out - * immediately from old frame buffers until the next vblank. + /* + * Disable all planes and wait for the change to take effect. This is + * required as the plane enable registers are updated on vblank, and no + * vblank will occur once the CRTC is stopped. Disabling planes when + * starting the CRTC thus wouldn't be enough as it would start scanning + * out immediately from old frame buffers until the next vblank. * * This increases the CRTC stop delay, especially when multiple CRTCs * are stopped in one operation as we now wait for one vblank per CRTC. * Whether this can be improved needs to be researched. */ - rcar_du_group_write(rcrtc->group, rcrtc->index % 2 ? DS2PR : DS1PR, 0); - drm_crtc_wait_one_vblank(crtc); + rcar_du_crtc_disable_planes(rcrtc); - /* Disable vertical blanking interrupt reporting. We first need to wait + /* + * Disable vertical blanking interrupt reporting. We first need to wait * for page flip completion before stopping the CRTC as userspace * expects page flips to eventually complete. */ @@ -502,14 +548,13 @@ static void rcar_du_crtc_stop(struct rcar_du_crtc *rcrtc) if (rcar_du_has(rcrtc->group->dev, RCAR_DU_FEATURE_VSP1_SOURCE)) rcar_du_vsp_disable(rcrtc); - /* Select switch sync mode. This stops display operation and configures + /* + * Select switch sync mode. This stops display operation and configures * the HSYNC and VSYNC signals as inputs. */ rcar_du_crtc_clr_set(rcrtc, DSYSR, DSYSR_TVM_MASK, DSYSR_TVM_SWITCH); rcar_du_group_start_stop(rcrtc->group, false); - - rcrtc->started = false; } void rcar_du_crtc_suspend(struct rcar_du_crtc *rcrtc) @@ -529,12 +574,10 @@ void rcar_du_crtc_resume(struct rcar_du_crtc *rcrtc) return; rcar_du_crtc_get(rcrtc); - rcar_du_crtc_start(rcrtc); + rcar_du_crtc_setup(rcrtc); /* Commit the planes state. */ - if (rcar_du_has(rcrtc->group->dev, RCAR_DU_FEATURE_VSP1_SOURCE)) { - rcar_du_vsp_enable(rcrtc); - } else { + if (!rcar_du_has(rcrtc->group->dev, RCAR_DU_FEATURE_VSP1_SOURCE)) { for (i = 0; i < rcrtc->group->num_planes; ++i) { struct rcar_du_plane *plane = &rcrtc->group->planes[i]; @@ -546,6 +589,7 @@ void rcar_du_crtc_resume(struct rcar_du_crtc *rcrtc) } rcar_du_crtc_update_planes(rcrtc); + rcar_du_crtc_start(rcrtc); } /* ----------------------------------------------------------------------------- @@ -557,7 +601,16 @@ static void rcar_du_crtc_atomic_enable(struct drm_crtc *crtc, { struct rcar_du_crtc *rcrtc = to_rcar_crtc(crtc); - rcar_du_crtc_get(rcrtc); + /* + * If the CRTC has already been setup by the .atomic_begin() handler we + * can skip the setup stage. + */ + if (!rcrtc->initialized) { + rcar_du_crtc_get(rcrtc); + rcar_du_crtc_setup(rcrtc); + rcrtc->initialized = true; + } + rcar_du_crtc_start(rcrtc); } @@ -576,6 +629,7 @@ static void rcar_du_crtc_atomic_disable(struct drm_crtc *crtc, } spin_unlock_irq(&crtc->dev->event_lock); + rcrtc->initialized = false; rcrtc->outputs = 0; } @@ -584,6 +638,19 @@ static void rcar_du_crtc_atomic_begin(struct drm_crtc *crtc, { struct rcar_du_crtc *rcrtc = to_rcar_crtc(crtc); + WARN_ON(!crtc->state->enable); + + /* + * If a mode set is in progress we can be called with the CRTC disabled. + * We then need to first setup the CRTC in order to configure planes. + * The .atomic_enable() handler will notice and skip the CRTC setup. + */ + if (!rcrtc->initialized) { + rcar_du_crtc_get(rcrtc); + rcar_du_crtc_setup(rcrtc); + rcrtc->initialized = true; + } + if (rcar_du_has(rcrtc->group->dev, RCAR_DU_FEATURE_VSP1_SOURCE)) rcar_du_vsp_atomic_begin(rcrtc); } @@ -623,6 +690,7 @@ static int rcar_du_crtc_enable_vblank(struct drm_crtc *crtc) rcar_du_crtc_write(rcrtc, DSRCR, DSRCR_VBCL); rcar_du_crtc_set(rcrtc, DIER, DIER_VBE); + rcrtc->vblank_enable = true; return 0; } @@ -632,6 +700,7 @@ static void rcar_du_crtc_disable_vblank(struct drm_crtc *crtc) struct rcar_du_crtc *rcrtc = to_rcar_crtc(crtc); rcar_du_crtc_clr(rcrtc, DIER, DIER_VBE); + rcrtc->vblank_enable = false; } static const struct drm_crtc_funcs crtc_funcs = { @@ -656,14 +725,30 @@ static irqreturn_t rcar_du_crtc_irq(int irq, void *arg) irqreturn_t ret = IRQ_NONE; u32 status; + spin_lock(&rcrtc->vblank_lock); + status = rcar_du_crtc_read(rcrtc, DSSR); rcar_du_crtc_write(rcrtc, DSRCR, status & DSRCR_MASK); - if (status & DSSR_FRM) { - drm_crtc_handle_vblank(&rcrtc->crtc); + if (status & DSSR_VBK) { + /* + * Wake up the vblank wait if the counter reaches 0. This must + * be protected by the vblank_lock to avoid races in + * rcar_du_crtc_disable_planes(). + */ + if (rcrtc->vblank_count) { + if (--rcrtc->vblank_count == 0) + wake_up(&rcrtc->vblank_wait); + } + } - if (rcdu->info->gen < 3) + spin_unlock(&rcrtc->vblank_lock); + + if (status & DSSR_VBK) { + if (rcdu->info->gen < 3) { + drm_crtc_handle_vblank(&rcrtc->crtc); rcar_du_crtc_finish_page_flip(rcrtc); + } ret = IRQ_HANDLED; } @@ -717,13 +802,15 @@ int rcar_du_crtc_create(struct rcar_du_group *rgrp, unsigned int index) } init_waitqueue_head(&rcrtc->flip_wait); + init_waitqueue_head(&rcrtc->vblank_wait); + spin_lock_init(&rcrtc->vblank_lock); rcrtc->group = rgrp; rcrtc->mmio_offset = mmio_offsets[index]; rcrtc->index = index; if (rcar_du_has(rcdu, RCAR_DU_FEATURE_VSP1_SOURCE)) - primary = &rcrtc->vsp->planes[0].plane; + primary = &rcrtc->vsp->planes[rcrtc->vsp_pipe].plane; else primary = &rgrp->planes[index % 2].plane; diff --git a/drivers/gpu/drm/rcar-du/rcar_du_crtc.h b/drivers/gpu/drm/rcar-du/rcar_du_crtc.h index b199ed5adf36..fdc2bf99bda1 100644 --- a/drivers/gpu/drm/rcar-du/rcar_du_crtc.h +++ b/drivers/gpu/drm/rcar-du/rcar_du_crtc.h @@ -15,6 +15,7 @@ #define __RCAR_DU_CRTC_H__ #include <linux/mutex.h> +#include <linux/spinlock.h> #include <linux/wait.h> #include <drm/drmP.h> @@ -30,11 +31,17 @@ struct rcar_du_vsp; * @extclock: external pixel dot clock (optional) * @mmio_offset: offset of the CRTC registers in the DU MMIO block * @index: CRTC software and hardware index - * @started: whether the CRTC has been started and is running + * @initialized: whether the CRTC has been initialized and clocks enabled + * @vblank_enable: whether vblank events are enabled on this CRTC * @event: event to post when the pending page flip completes * @flip_wait: wait queue used to signal page flip completion + * @vblank_lock: protects vblank_wait and vblank_count + * @vblank_wait: wait queue used to signal vertical blanking + * @vblank_count: number of vertical blanking interrupts to wait for * @outputs: bitmask of the outputs (enum rcar_du_output) driven by this CRTC * @group: CRTC group this CRTC belongs to + * @vsp: VSP feeding video to this CRTC + * @vsp_pipe: index of the VSP pipeline feeding video to this CRTC */ struct rcar_du_crtc { struct drm_crtc crtc; @@ -43,15 +50,21 @@ struct rcar_du_crtc { struct clk *extclock; unsigned int mmio_offset; unsigned int index; - bool started; + bool initialized; + bool vblank_enable; struct drm_pending_vblank_event *event; wait_queue_head_t flip_wait; + spinlock_t vblank_lock; + wait_queue_head_t vblank_wait; + unsigned int vblank_count; + unsigned int outputs; struct rcar_du_group *group; struct rcar_du_vsp *vsp; + unsigned int vsp_pipe; }; #define to_rcar_crtc(c) container_of(c, struct rcar_du_crtc, crtc) diff --git a/drivers/gpu/drm/rcar-du/rcar_du_drv.c b/drivers/gpu/drm/rcar-du/rcar_du_drv.c index c09cf847a657..d2f29e6b1112 100644 --- a/drivers/gpu/drm/rcar-du/rcar_du_drv.c +++ b/drivers/gpu/drm/rcar-du/rcar_du_drv.c @@ -39,7 +39,8 @@ static const struct rcar_du_device_info rcar_du_r8a7779_info = { .features = 0, .num_crtcs = 2, .routes = { - /* R8A7779 has two RGB outputs and one (currently unsupported) + /* + * R8A7779 has two RGB outputs and one (currently unsupported) * TCON output. */ [RCAR_DU_OUTPUT_DPAD0] = { @@ -61,7 +62,8 @@ static const struct rcar_du_device_info rcar_du_r8a7790_info = { .quirks = RCAR_DU_QUIRK_ALIGN_128B | RCAR_DU_QUIRK_LVDS_LANES, .num_crtcs = 3, .routes = { - /* R8A7790 has one RGB output, two LVDS outputs and one + /* + * R8A7790 has one RGB output, two LVDS outputs and one * (currently unsupported) TCON output. */ [RCAR_DU_OUTPUT_DPAD0] = { @@ -87,7 +89,8 @@ static const struct rcar_du_device_info rcar_du_r8a7791_info = { | RCAR_DU_FEATURE_EXT_CTRL_REGS, .num_crtcs = 2, .routes = { - /* R8A779[13] has one RGB output, one LVDS output and one + /* + * R8A779[13] has one RGB output, one LVDS output and one * (currently unsupported) TCON output. */ [RCAR_DU_OUTPUT_DPAD0] = { @@ -127,7 +130,8 @@ static const struct rcar_du_device_info rcar_du_r8a7794_info = { | RCAR_DU_FEATURE_EXT_CTRL_REGS, .num_crtcs = 2, .routes = { - /* R8A7794 has two RGB outputs and one (currently unsupported) + /* + * R8A7794 has two RGB outputs and one (currently unsupported) * TCON output. */ [RCAR_DU_OUTPUT_DPAD0] = { @@ -149,7 +153,8 @@ static const struct rcar_du_device_info rcar_du_r8a7795_info = { | RCAR_DU_FEATURE_VSP1_SOURCE, .num_crtcs = 4, .routes = { - /* R8A7795 has one RGB output, two HDMI outputs and one + /* + * R8A7795 has one RGB output, two HDMI outputs and one * LVDS output. */ [RCAR_DU_OUTPUT_DPAD0] = { @@ -180,19 +185,25 @@ static const struct rcar_du_device_info rcar_du_r8a7796_info = { | RCAR_DU_FEATURE_VSP1_SOURCE, .num_crtcs = 3, .routes = { - /* R8A7796 has one RGB output, one LVDS output and one - * (currently unsupported) HDMI output. + /* + * R8A7796 has one RGB output, one LVDS output and one HDMI + * output. */ [RCAR_DU_OUTPUT_DPAD0] = { .possible_crtcs = BIT(2), .port = 0, }, + [RCAR_DU_OUTPUT_HDMI0] = { + .possible_crtcs = BIT(1), + .port = 1, + }, [RCAR_DU_OUTPUT_LVDS0] = { .possible_crtcs = BIT(0), .port = 2, }, }, .num_lvds = 1, + .dpll_ch = BIT(1), }; static const struct of_device_id rcar_du_of_table[] = { @@ -339,7 +350,8 @@ static int rcar_du_probe(struct platform_device *pdev) ddev->irq_enabled = 1; - /* Register the DRM device with the core and the connectors with + /* + * Register the DRM device with the core and the connectors with * sysfs. */ ret = drm_dev_register(ddev, 0); diff --git a/drivers/gpu/drm/rcar-du/rcar_du_group.c b/drivers/gpu/drm/rcar-du/rcar_du_group.c index 64738fca96d0..2f37ea901873 100644 --- a/drivers/gpu/drm/rcar-du/rcar_du_group.c +++ b/drivers/gpu/drm/rcar-du/rcar_du_group.c @@ -64,7 +64,8 @@ static void rcar_du_group_setup_defr8(struct rcar_du_group *rgrp) if (rcdu->info->gen < 3) { defr8 |= DEFR8_DEFE8; - /* On Gen2 the DEFR8 register for the first group also controls + /* + * On Gen2 the DEFR8 register for the first group also controls * RGB output routing to DPAD0 and VSPD1 routing to DU0/1/2 for * DU instances that support it. */ @@ -75,7 +76,8 @@ static void rcar_du_group_setup_defr8(struct rcar_du_group *rgrp) defr8 |= DEFR8_VSCS; } } else { - /* On Gen3 VSPD routing can't be configured, but DPAD routing + /* + * On Gen3 VSPD routing can't be configured, but DPAD routing * needs to be set despite having a single option available. */ u32 crtc = ffs(possible_crtcs) - 1; @@ -124,7 +126,8 @@ static void rcar_du_group_setup(struct rcar_du_group *rgrp) if (rcdu->info->gen >= 3) rcar_du_group_write(rgrp, DEFR10, DEFR10_CODE | DEFR10_DEFE10); - /* Use DS1PR and DS2PR to configure planes priorities and connects the + /* + * Use DS1PR and DS2PR to configure planes priorities and connects the * superposition 0 to DU0 pins. DU1 pins will be configured dynamically. */ rcar_du_group_write(rgrp, DORCR, DORCR_PG1D_DS1 | DORCR_DPRS); @@ -177,7 +180,8 @@ static void __rcar_du_group_start_stop(struct rcar_du_group *rgrp, bool start) void rcar_du_group_start_stop(struct rcar_du_group *rgrp, bool start) { - /* Many of the configuration bits are only updated when the display + /* + * Many of the configuration bits are only updated when the display * reset (DRES) bit in DSYSR is set to 1, disabling *both* CRTCs. Some * of those bits could be pre-configured, but others (especially the * bits related to plane assignment to display timing controllers) need @@ -208,23 +212,32 @@ void rcar_du_group_restart(struct rcar_du_group *rgrp) int rcar_du_set_dpad0_vsp1_routing(struct rcar_du_device *rcdu) { + struct rcar_du_group *rgrp; + struct rcar_du_crtc *crtc; + unsigned int index; int ret; if (!rcar_du_has(rcdu, RCAR_DU_FEATURE_EXT_CTRL_REGS)) return 0; - /* RGB output routing to DPAD0 and VSP1D routing to DU0/1/2 are - * configured in the DEFR8 register of the first group. As this function - * can be called with the DU0 and DU1 CRTCs disabled, we need to enable - * the first group clock before accessing the register. + /* + * RGB output routing to DPAD0 and VSP1D routing to DU0/1/2 are + * configured in the DEFR8 register of the first group on Gen2 and the + * last group on Gen3. As this function can be called with the DU + * channels of the corresponding CRTCs disabled, we need to enable the + * group clock before accessing the register. */ - ret = clk_prepare_enable(rcdu->crtcs[0].clock); + index = rcdu->info->gen < 3 ? 0 : DIV_ROUND_UP(rcdu->num_crtcs, 2) - 1; + rgrp = &rcdu->groups[index]; + crtc = &rcdu->crtcs[index * 2]; + + ret = clk_prepare_enable(crtc->clock); if (ret < 0) return ret; - rcar_du_group_setup_defr8(&rcdu->groups[0]); + rcar_du_group_setup_defr8(rgrp); - clk_disable_unprepare(rcdu->crtcs[0].clock); + clk_disable_unprepare(crtc->clock); return 0; } @@ -236,7 +249,8 @@ int rcar_du_group_set_routing(struct rcar_du_group *rgrp) dorcr &= ~(DORCR_PG2T | DORCR_DK2S | DORCR_PG2D_MASK); - /* Set the DPAD1 pins sources. Select CRTC 0 if explicitly requested and + /* + * Set the DPAD1 pins sources. Select CRTC 0 if explicitly requested and * CRTC 1 in all other cases to avoid cloning CRTC 0 to DPAD0 and DPAD1 * by default. */ diff --git a/drivers/gpu/drm/rcar-du/rcar_du_kms.c b/drivers/gpu/drm/rcar-du/rcar_du_kms.c index 5d681ea53be6..7278b9703c15 100644 --- a/drivers/gpu/drm/rcar-du/rcar_du_kms.c +++ b/drivers/gpu/drm/rcar-du/rcar_du_kms.c @@ -96,7 +96,8 @@ static const struct rcar_du_format_info rcar_du_format_infos[] = { .pnmr = PnMR_SPIM_TP_OFF | PnMR_DDDF_YC, .edf = PnDDCR4_EDF_NONE, }, - /* The following formats are not supported on Gen2 and thus have no + /* + * The following formats are not supported on Gen2 and thus have no * associated .pnmr or .edf settings. */ { @@ -153,7 +154,8 @@ int rcar_du_dumb_create(struct drm_file *file, struct drm_device *dev, unsigned int min_pitch = DIV_ROUND_UP(args->width * args->bpp, 8); unsigned int align; - /* The R8A7779 DU requires a 16 pixels pitch alignment as documented, + /* + * The R8A7779 DU requires a 16 pixels pitch alignment as documented, * but the R8A7790 DU seems to require a 128 bytes pitch alignment. */ if (rcar_du_needs(rcdu, RCAR_DU_QUIRK_ALIGN_128B)) @@ -255,12 +257,12 @@ static void rcar_du_atomic_commit_tail(struct drm_atomic_state *old_state) /* Apply the atomic update. */ drm_atomic_helper_commit_modeset_disables(dev, old_state); - drm_atomic_helper_commit_modeset_enables(dev, old_state); drm_atomic_helper_commit_planes(dev, old_state, DRM_PLANE_COMMIT_ACTIVE_ONLY); + drm_atomic_helper_commit_modeset_enables(dev, old_state); drm_atomic_helper_commit_hw_done(old_state); - drm_atomic_helper_wait_for_vblanks(dev, old_state); + drm_atomic_helper_wait_for_flip_done(dev, old_state); drm_atomic_helper_cleanup_planes(dev, old_state); } @@ -309,7 +311,7 @@ static int rcar_du_encoders_init_one(struct rcar_du_device *rcdu, return -ENODEV; } - entity_ep_node = of_parse_phandle(ep->local_node, "remote-endpoint", 0); + entity_ep_node = of_graph_get_remote_endpoint(ep->local_node); for_each_endpoint_of_node(entity, ep_node) { if (ep_node == entity_ep_node) @@ -419,7 +421,8 @@ static int rcar_du_properties_init(struct rcar_du_device *rcdu) if (rcdu->props.alpha == NULL) return -ENOMEM; - /* The color key is expressed as an RGB888 triplet stored in a 32-bit + /* + * The color key is expressed as an RGB888 triplet stored in a 32-bit * integer in XRGB8888 format. Bit 24 is used as a flag to disable (0) * or enable source color keying (1). */ @@ -432,6 +435,81 @@ static int rcar_du_properties_init(struct rcar_du_device *rcdu) return 0; } +static int rcar_du_vsps_init(struct rcar_du_device *rcdu) +{ + const struct device_node *np = rcdu->dev->of_node; + struct of_phandle_args args; + struct { + struct device_node *np; + unsigned int crtcs_mask; + } vsps[RCAR_DU_MAX_VSPS] = { { 0, }, }; + unsigned int vsps_count = 0; + unsigned int cells; + unsigned int i; + int ret; + + /* + * First parse the DT vsps property to populate the list of VSPs. Each + * entry contains a pointer to the VSP DT node and a bitmask of the + * connected DU CRTCs. + */ + cells = of_property_count_u32_elems(np, "vsps") / rcdu->num_crtcs - 1; + if (cells > 1) + return -EINVAL; + + for (i = 0; i < rcdu->num_crtcs; ++i) { + unsigned int j; + + ret = of_parse_phandle_with_fixed_args(np, "vsps", cells, i, + &args); + if (ret < 0) + goto error; + + /* + * Add the VSP to the list or update the corresponding existing + * entry if the VSP has already been added. + */ + for (j = 0; j < vsps_count; ++j) { + if (vsps[j].np == args.np) + break; + } + + if (j < vsps_count) + of_node_put(args.np); + else + vsps[vsps_count++].np = args.np; + + vsps[j].crtcs_mask |= BIT(i); + + /* Store the VSP pointer and pipe index in the CRTC. */ + rcdu->crtcs[i].vsp = &rcdu->vsps[j]; + rcdu->crtcs[i].vsp_pipe = cells >= 1 ? args.args[0] : 0; + } + + /* + * Then initialize all the VSPs from the node pointers and CRTCs bitmask + * computed previously. + */ + for (i = 0; i < vsps_count; ++i) { + struct rcar_du_vsp *vsp = &rcdu->vsps[i]; + + vsp->index = i; + vsp->dev = rcdu; + + ret = rcar_du_vsp_init(vsp, vsps[i].np, vsps[i].crtcs_mask); + if (ret < 0) + goto error; + } + + return 0; + +error: + for (i = 0; i < ARRAY_SIZE(vsps); ++i) + of_node_put(vsps[i].np); + + return ret; +} + int rcar_du_modeset_init(struct rcar_du_device *rcdu) { static const unsigned int mmio_offsets[] = { @@ -461,7 +539,8 @@ int rcar_du_modeset_init(struct rcar_du_device *rcdu) if (ret < 0) return ret; - /* Initialize vertical blanking interrupts handling. Start with vblank + /* + * Initialize vertical blanking interrupts handling. Start with vblank * disabled for all CRTCs. */ ret = drm_vblank_init(dev, (1 << rcdu->info->num_crtcs) - 1); @@ -481,7 +560,8 @@ int rcar_du_modeset_init(struct rcar_du_device *rcdu) rgrp->index = i; rgrp->num_crtcs = min(rcdu->num_crtcs - 2 * i, 2U); - /* If we have more than one CRTCs in this group pre-associate + /* + * If we have more than one CRTCs in this group pre-associate * the low-order planes with CRTC 0 and the high-order planes * with CRTC 1 to minimize flicker occurring when the * association is changed. @@ -499,17 +579,9 @@ int rcar_du_modeset_init(struct rcar_du_device *rcdu) /* Initialize the compositors. */ if (rcar_du_has(rcdu, RCAR_DU_FEATURE_VSP1_SOURCE)) { - for (i = 0; i < rcdu->num_crtcs; ++i) { - struct rcar_du_vsp *vsp = &rcdu->vsps[i]; - - vsp->index = i; - vsp->dev = rcdu; - rcdu->crtcs[i].vsp = vsp; - - ret = rcar_du_vsp_init(vsp); - if (ret < 0) - return ret; - } + ret = rcar_du_vsps_init(rcdu); + if (ret < 0) + return ret; } /* Create the CRTCs. */ @@ -537,7 +609,8 @@ int rcar_du_modeset_init(struct rcar_du_device *rcdu) num_encoders = ret; - /* Set the possible CRTCs and possible clones. There's always at least + /* + * Set the possible CRTCs and possible clones. There's always at least * one way for all encoders to clone each other, set all bits in the * possible clones field. */ diff --git a/drivers/gpu/drm/rcar-du/rcar_du_lvdsenc.c b/drivers/gpu/drm/rcar-du/rcar_du_lvdsenc.c index 1661f6201210..12d22f3db1af 100644 --- a/drivers/gpu/drm/rcar-du/rcar_du_lvdsenc.c +++ b/drivers/gpu/drm/rcar-du/rcar_du_lvdsenc.c @@ -59,7 +59,8 @@ static void rcar_du_lvdsenc_start_gen2(struct rcar_du_lvdsenc *lvds, rcar_lvds_write(lvds, LVDPLLCR, pllcr); - /* Select the input, hardcode mode 0, enable LVDS operation and turn + /* + * Select the input, hardcode mode 0, enable LVDS operation and turn * bias circuitry on. */ lvdcr0 = (lvds->mode << LVDCR0_LVMD_SHIFT) | LVDCR0_BEN | LVDCR0_LVEN; @@ -73,7 +74,8 @@ static void rcar_du_lvdsenc_start_gen2(struct rcar_du_lvdsenc *lvds, LVDCR1_CHSTBY_GEN2(1) | LVDCR1_CHSTBY_GEN2(0) | LVDCR1_CLKSTBY_GEN2); - /* Turn the PLL on, wait for the startup delay, and turn the output + /* + * Turn the PLL on, wait for the startup delay, and turn the output * on. */ lvdcr0 |= LVDCR0_PLLON; @@ -140,7 +142,8 @@ static int rcar_du_lvdsenc_start(struct rcar_du_lvdsenc *lvds, if (ret < 0) return ret; - /* Hardcode the channels and control signals routing for now. + /* + * Hardcode the channels and control signals routing for now. * * HSYNC -> CTRL0 * VSYNC -> CTRL1 @@ -202,7 +205,8 @@ void rcar_du_lvdsenc_atomic_check(struct rcar_du_lvdsenc *lvds, { struct rcar_du_device *rcdu = lvds->dev; - /* The internal LVDS encoder has a restricted clock frequency operating + /* + * The internal LVDS encoder has a restricted clock frequency operating * range (30MHz to 150MHz on Gen2, 25.175MHz to 148.5MHz on Gen3). Clamp * the clock accordingly. */ diff --git a/drivers/gpu/drm/rcar-du/rcar_du_plane.c b/drivers/gpu/drm/rcar-du/rcar_du_plane.c index 5a7967498628..61833cc1c699 100644 --- a/drivers/gpu/drm/rcar-du/rcar_du_plane.c +++ b/drivers/gpu/drm/rcar-du/rcar_du_plane.c @@ -50,23 +50,21 @@ * automatically when the core swaps the old and new states. */ -static bool rcar_du_plane_needs_realloc(struct rcar_du_plane *plane, - struct rcar_du_plane_state *new_state) +static bool rcar_du_plane_needs_realloc( + const struct rcar_du_plane_state *old_state, + const struct rcar_du_plane_state *new_state) { - struct rcar_du_plane_state *cur_state; - - cur_state = to_rcar_plane_state(plane->plane.state); - - /* Lowering the number of planes doesn't strictly require reallocation + /* + * Lowering the number of planes doesn't strictly require reallocation * as the extra hardware plane will be freed when committing, but doing * so could lead to more fragmentation. */ - if (!cur_state->format || - cur_state->format->planes != new_state->format->planes) + if (!old_state->format || + old_state->format->planes != new_state->format->planes) return true; /* Reallocate hardware planes if the source has changed. */ - if (cur_state->source != new_state->source) + if (old_state->source != new_state->source) return true; return false; @@ -141,37 +139,43 @@ int rcar_du_atomic_check_planes(struct drm_device *dev, unsigned int groups = 0; unsigned int i; struct drm_plane *drm_plane; - struct drm_plane_state *drm_plane_state; + struct drm_plane_state *old_drm_plane_state; + struct drm_plane_state *new_drm_plane_state; /* Check if hardware planes need to be reallocated. */ - for_each_plane_in_state(state, drm_plane, drm_plane_state, i) { - struct rcar_du_plane_state *plane_state; + for_each_oldnew_plane_in_state(state, drm_plane, old_drm_plane_state, + new_drm_plane_state, i) { + struct rcar_du_plane_state *old_plane_state; + struct rcar_du_plane_state *new_plane_state; struct rcar_du_plane *plane; unsigned int index; plane = to_rcar_plane(drm_plane); - plane_state = to_rcar_plane_state(drm_plane_state); + old_plane_state = to_rcar_plane_state(old_drm_plane_state); + new_plane_state = to_rcar_plane_state(new_drm_plane_state); dev_dbg(rcdu->dev, "%s: checking plane (%u,%tu)\n", __func__, plane->group->index, plane - plane->group->planes); - /* If the plane is being disabled we don't need to go through + /* + * If the plane is being disabled we don't need to go through * the full reallocation procedure. Just mark the hardware * plane(s) as freed. */ - if (!plane_state->format) { + if (!new_plane_state->format) { dev_dbg(rcdu->dev, "%s: plane is being disabled\n", __func__); index = plane - plane->group->planes; group_freed_planes[plane->group->index] |= 1 << index; - plane_state->hwindex = -1; + new_plane_state->hwindex = -1; continue; } - /* If the plane needs to be reallocated mark it as such, and + /* + * If the plane needs to be reallocated mark it as such, and * mark the hardware plane(s) as free. */ - if (rcar_du_plane_needs_realloc(plane, plane_state)) { + if (rcar_du_plane_needs_realloc(old_plane_state, new_plane_state)) { dev_dbg(rcdu->dev, "%s: plane needs reallocation\n", __func__); groups |= 1 << plane->group->index; @@ -179,14 +183,15 @@ int rcar_du_atomic_check_planes(struct drm_device *dev, index = plane - plane->group->planes; group_freed_planes[plane->group->index] |= 1 << index; - plane_state->hwindex = -1; + new_plane_state->hwindex = -1; } } if (!needs_realloc) return 0; - /* Grab all plane states for the groups that need reallocation to ensure + /* + * Grab all plane states for the groups that need reallocation to ensure * locking and avoid racy updates. This serializes the update operation, * but there's not much we can do about it as that's the hardware * design. @@ -204,14 +209,15 @@ int rcar_du_atomic_check_planes(struct drm_device *dev, for (i = 0; i < group->num_planes; ++i) { struct rcar_du_plane *plane = &group->planes[i]; - struct rcar_du_plane_state *plane_state; + struct rcar_du_plane_state *new_plane_state; struct drm_plane_state *s; s = drm_atomic_get_plane_state(state, &plane->plane); if (IS_ERR(s)) return PTR_ERR(s); - /* If the plane has been freed in the above loop its + /* + * If the plane has been freed in the above loop its * hardware planes must not be added to the used planes * bitmask. However, the current state doesn't reflect * the free state yet, as we've modified the new state @@ -226,16 +232,16 @@ int rcar_du_atomic_check_planes(struct drm_device *dev, continue; } - plane_state = to_rcar_plane_state(plane->plane.state); - used_planes |= rcar_du_plane_hwmask(plane_state); + new_plane_state = to_rcar_plane_state(s); + used_planes |= rcar_du_plane_hwmask(new_plane_state); dev_dbg(rcdu->dev, "%s: plane (%u,%tu) uses %u hwplanes (index %d)\n", __func__, plane->group->index, plane - plane->group->planes, - plane_state->format ? - plane_state->format->planes : 0, - plane_state->hwindex); + new_plane_state->format ? + new_plane_state->format->planes : 0, + new_plane_state->hwindex); } group_free_planes[index] = 0xff & ~used_planes; @@ -246,40 +252,45 @@ int rcar_du_atomic_check_planes(struct drm_device *dev, } /* Reallocate hardware planes for each plane that needs it. */ - for_each_plane_in_state(state, drm_plane, drm_plane_state, i) { - struct rcar_du_plane_state *plane_state; + for_each_oldnew_plane_in_state(state, drm_plane, old_drm_plane_state, + new_drm_plane_state, i) { + struct rcar_du_plane_state *old_plane_state; + struct rcar_du_plane_state *new_plane_state; struct rcar_du_plane *plane; unsigned int crtc_planes; unsigned int free; int idx; plane = to_rcar_plane(drm_plane); - plane_state = to_rcar_plane_state(drm_plane_state); + old_plane_state = to_rcar_plane_state(old_drm_plane_state); + new_plane_state = to_rcar_plane_state(new_drm_plane_state); dev_dbg(rcdu->dev, "%s: allocating plane (%u,%tu)\n", __func__, plane->group->index, plane - plane->group->planes); - /* Skip planes that are being disabled or don't need to be + /* + * Skip planes that are being disabled or don't need to be * reallocated. */ - if (!plane_state->format || - !rcar_du_plane_needs_realloc(plane, plane_state)) + if (!new_plane_state->format || + !rcar_du_plane_needs_realloc(old_plane_state, new_plane_state)) continue; - /* Try to allocate the plane from the free planes currently + /* + * Try to allocate the plane from the free planes currently * associated with the target CRTC to avoid restarting the CRTC * group and thus minimize flicker. If it fails fall back to * allocating from all free planes. */ - crtc_planes = to_rcar_crtc(plane_state->state.crtc)->index % 2 + crtc_planes = to_rcar_crtc(new_plane_state->state.crtc)->index % 2 ? plane->group->dptsr_planes : ~plane->group->dptsr_planes; free = group_free_planes[plane->group->index]; - idx = rcar_du_plane_hwalloc(plane, plane_state, + idx = rcar_du_plane_hwalloc(plane, new_plane_state, free & crtc_planes); if (idx < 0) - idx = rcar_du_plane_hwalloc(plane, plane_state, + idx = rcar_du_plane_hwalloc(plane, new_plane_state, free); if (idx < 0) { dev_dbg(rcdu->dev, "%s: no available hardware plane\n", @@ -288,12 +299,12 @@ int rcar_du_atomic_check_planes(struct drm_device *dev, } dev_dbg(rcdu->dev, "%s: allocated %u hwplanes (index %u)\n", - __func__, plane_state->format->planes, idx); + __func__, new_plane_state->format->planes, idx); - plane_state->hwindex = idx; + new_plane_state->hwindex = idx; group_free_planes[plane->group->index] &= - ~rcar_du_plane_hwmask(plane_state); + ~rcar_du_plane_hwmask(new_plane_state); dev_dbg(rcdu->dev, "%s: group %u free planes mask 0x%02x\n", __func__, plane->group->index, @@ -351,14 +362,16 @@ static void rcar_du_plane_setup_scanout(struct rcar_du_group *rgrp, dma[1] = 0; } - /* Memory pitch (expressed in pixels). Must be doubled for interlaced + /* + * Memory pitch (expressed in pixels). Must be doubled for interlaced * operation with 32bpp formats. */ rcar_du_plane_write(rgrp, index, PnMWR, (interlaced && state->format->bpp == 32) ? pitch * 2 : pitch); - /* The Y position is expressed in raster line units and must be doubled + /* + * The Y position is expressed in raster line units and must be doubled * for 32bpp formats, according to the R8A7790 datasheet. No mention of * doubling the Y position is found in the R8A7779 datasheet, but the * rule seems to apply there as well. @@ -396,7 +409,8 @@ static void rcar_du_plane_setup_mode(struct rcar_du_group *rgrp, u32 colorkey; u32 pnmr; - /* The PnALPHAR register controls alpha-blending in 16bpp formats + /* + * The PnALPHAR register controls alpha-blending in 16bpp formats * (ARGB1555 and XRGB1555). * * For ARGB, set the alpha value to 0, and enable alpha-blending when @@ -413,7 +427,8 @@ static void rcar_du_plane_setup_mode(struct rcar_du_group *rgrp, pnmr = PnMR_BM_MD | state->format->pnmr; - /* Disable color keying when requested. YUV formats have the + /* + * Disable color keying when requested. YUV formats have the * PnMR_SPIM_TP_OFF bit set in their pnmr field, disabling color keying * automatically. */ @@ -457,7 +472,8 @@ static void rcar_du_plane_setup_format_gen2(struct rcar_du_group *rgrp, u32 ddcr2 = PnDDCR2_CODE; u32 ddcr4; - /* Data format + /* + * Data format * * The data format is selected by the DDDF field in PnMR and the EDF * field in DDCR4. @@ -589,7 +605,8 @@ static void rcar_du_plane_atomic_update(struct drm_plane *plane, rcar_du_plane_setup(rplane); - /* Check whether the source has changed from memory to live source or + /* + * Check whether the source has changed from memory to live source or * from live source to memory. The source has been configured by the * VSPS bit in the PnDDCR4 register. Although the datasheet states that * the bit is updated during vertical blanking, it seems that updates @@ -725,7 +742,8 @@ int rcar_du_planes_init(struct rcar_du_group *rgrp) unsigned int i; int ret; - /* Create one primary plane per CRTC in this group and seven overlay + /* + * Create one primary plane per CRTC in this group and seven overlay * planes. */ rgrp->num_planes = rgrp->num_crtcs + 7; diff --git a/drivers/gpu/drm/rcar-du/rcar_du_plane.h b/drivers/gpu/drm/rcar-du/rcar_du_plane.h index 8b91dd3a46e4..f62e09f195de 100644 --- a/drivers/gpu/drm/rcar-du/rcar_du_plane.h +++ b/drivers/gpu/drm/rcar-du/rcar_du_plane.h @@ -20,7 +20,8 @@ struct rcar_du_format_info; struct rcar_du_group; -/* The RCAR DU has 8 hardware planes, shared between primary and overlay planes. +/* + * The RCAR DU has 8 hardware planes, shared between primary and overlay planes. * As using overlay planes requires at least one of the CRTCs being enabled, no * more than 7 overlay planes can be available. We thus create 1 primary plane * per CRTC and 7 overlay planes, for a total of up to 9 KMS planes. diff --git a/drivers/gpu/drm/rcar-du/rcar_du_vsp.c b/drivers/gpu/drm/rcar-du/rcar_du_vsp.c index 4e46e479e961..2c96147bc444 100644 --- a/drivers/gpu/drm/rcar-du/rcar_du_vsp.c +++ b/drivers/gpu/drm/rcar-du/rcar_du_vsp.c @@ -19,6 +19,7 @@ #include <drm/drm_gem_cma_helper.h> #include <drm/drm_plane_helper.h> +#include <linux/bitops.h> #include <linux/dma-mapping.h> #include <linux/of_platform.h> #include <linux/scatterlist.h> @@ -30,11 +31,15 @@ #include "rcar_du_kms.h" #include "rcar_du_vsp.h" -static void rcar_du_vsp_complete(void *private) +static void rcar_du_vsp_complete(void *private, bool completed) { struct rcar_du_crtc *crtc = private; - rcar_du_crtc_finish_page_flip(crtc); + if (crtc->vblank_enable) + drm_crtc_handle_vblank(&crtc->crtc); + + if (completed) + rcar_du_crtc_finish_page_flip(crtc); } void rcar_du_vsp_enable(struct rcar_du_crtc *crtc) @@ -73,7 +78,8 @@ void rcar_du_vsp_enable(struct rcar_du_crtc *crtc) __rcar_du_plane_setup(crtc->group, &state); - /* Ensure that the plane source configuration takes effect by requesting + /* + * Ensure that the plane source configuration takes effect by requesting * a restart of the group. See rcar_du_plane_atomic_update() for a more * detailed explanation. * @@ -81,22 +87,22 @@ void rcar_du_vsp_enable(struct rcar_du_crtc *crtc) */ crtc->group->need_restart = true; - vsp1_du_setup_lif(crtc->vsp->vsp, &cfg); + vsp1_du_setup_lif(crtc->vsp->vsp, crtc->vsp_pipe, &cfg); } void rcar_du_vsp_disable(struct rcar_du_crtc *crtc) { - vsp1_du_setup_lif(crtc->vsp->vsp, NULL); + vsp1_du_setup_lif(crtc->vsp->vsp, crtc->vsp_pipe, NULL); } void rcar_du_vsp_atomic_begin(struct rcar_du_crtc *crtc) { - vsp1_du_atomic_begin(crtc->vsp->vsp); + vsp1_du_atomic_begin(crtc->vsp->vsp, crtc->vsp_pipe); } void rcar_du_vsp_atomic_flush(struct rcar_du_crtc *crtc) { - vsp1_du_atomic_flush(crtc->vsp->vsp); + vsp1_du_atomic_flush(crtc->vsp->vsp, crtc->vsp_pipe); } /* Keep the two tables in sync. */ @@ -162,6 +168,7 @@ static void rcar_du_vsp_plane_setup(struct rcar_du_vsp_plane *plane) { struct rcar_du_vsp_plane_state *state = to_rcar_vsp_plane_state(plane->plane.state); + struct rcar_du_crtc *crtc = to_rcar_crtc(state->state.crtc); struct drm_framebuffer *fb = plane->plane.state->fb; struct vsp1_du_atomic_config cfg = { .pixelformat = 0, @@ -192,7 +199,8 @@ static void rcar_du_vsp_plane_setup(struct rcar_du_vsp_plane *plane) } } - vsp1_du_atomic_update(plane->vsp->vsp, plane->index, &cfg); + vsp1_du_atomic_update(plane->vsp->vsp, crtc->vsp_pipe, + plane->index, &cfg); } static int rcar_du_vsp_plane_prepare_fb(struct drm_plane *plane, @@ -288,11 +296,13 @@ static void rcar_du_vsp_plane_atomic_update(struct drm_plane *plane, struct drm_plane_state *old_state) { struct rcar_du_vsp_plane *rplane = to_rcar_vsp_plane(plane); + struct rcar_du_crtc *crtc = to_rcar_crtc(old_state->crtc); if (plane->state->crtc) rcar_du_vsp_plane_setup(rplane); else - vsp1_du_atomic_update(rplane->vsp->vsp, rplane->index, NULL); + vsp1_du_atomic_update(rplane->vsp->vsp, crtc->vsp_pipe, + rplane->index, NULL); } static const struct drm_plane_helper_funcs rcar_du_vsp_plane_helper_funcs = { @@ -390,23 +400,17 @@ static const struct drm_plane_funcs rcar_du_vsp_plane_funcs = { .atomic_get_property = rcar_du_vsp_plane_atomic_get_property, }; -int rcar_du_vsp_init(struct rcar_du_vsp *vsp) +int rcar_du_vsp_init(struct rcar_du_vsp *vsp, struct device_node *np, + unsigned int crtcs) { struct rcar_du_device *rcdu = vsp->dev; struct platform_device *pdev; - struct device_node *np; + unsigned int num_crtcs = hweight32(crtcs); unsigned int i; int ret; /* Find the VSP device and initialize it. */ - np = of_parse_phandle(rcdu->dev->of_node, "vsps", vsp->index); - if (!np) { - dev_err(rcdu->dev, "vsps node not found\n"); - return -ENXIO; - } - pdev = of_find_device_by_node(np); - of_node_put(np); if (!pdev) return -ENXIO; @@ -416,7 +420,8 @@ int rcar_du_vsp_init(struct rcar_du_vsp *vsp) if (ret < 0) return ret; - /* The VSP2D (Gen3) has 5 RPFs, but the VSP1D (Gen2) is limited to + /* + * The VSP2D (Gen3) has 5 RPFs, but the VSP1D (Gen2) is limited to * 4 RPFs. */ vsp->num_planes = rcdu->info->gen >= 3 ? 5 : 4; @@ -427,15 +432,15 @@ int rcar_du_vsp_init(struct rcar_du_vsp *vsp) return -ENOMEM; for (i = 0; i < vsp->num_planes; ++i) { - enum drm_plane_type type = i ? DRM_PLANE_TYPE_OVERLAY - : DRM_PLANE_TYPE_PRIMARY; + enum drm_plane_type type = i < num_crtcs + ? DRM_PLANE_TYPE_PRIMARY + : DRM_PLANE_TYPE_OVERLAY; struct rcar_du_vsp_plane *plane = &vsp->planes[i]; plane->vsp = vsp; plane->index = i; - ret = drm_universal_plane_init(rcdu->ddev, &plane->plane, - 1 << vsp->index, + ret = drm_universal_plane_init(rcdu->ddev, &plane->plane, crtcs, &rcar_du_vsp_plane_funcs, formats_kms, ARRAY_SIZE(formats_kms), diff --git a/drivers/gpu/drm/rcar-du/rcar_du_vsp.h b/drivers/gpu/drm/rcar-du/rcar_du_vsp.h index 8861661590ff..f876c512163c 100644 --- a/drivers/gpu/drm/rcar-du/rcar_du_vsp.h +++ b/drivers/gpu/drm/rcar-du/rcar_du_vsp.h @@ -64,13 +64,19 @@ to_rcar_vsp_plane_state(struct drm_plane_state *state) } #ifdef CONFIG_DRM_RCAR_VSP -int rcar_du_vsp_init(struct rcar_du_vsp *vsp); +int rcar_du_vsp_init(struct rcar_du_vsp *vsp, struct device_node *np, + unsigned int crtcs); void rcar_du_vsp_enable(struct rcar_du_crtc *crtc); void rcar_du_vsp_disable(struct rcar_du_crtc *crtc); void rcar_du_vsp_atomic_begin(struct rcar_du_crtc *crtc); void rcar_du_vsp_atomic_flush(struct rcar_du_crtc *crtc); #else -static inline int rcar_du_vsp_init(struct rcar_du_vsp *vsp) { return -ENXIO; }; +static inline int rcar_du_vsp_init(struct rcar_du_vsp *vsp, + struct device_node *np, + unsigned int crtcs) +{ + return -ENXIO; +} static inline void rcar_du_vsp_enable(struct rcar_du_crtc *crtc) { }; static inline void rcar_du_vsp_disable(struct rcar_du_crtc *crtc) { }; static inline void rcar_du_vsp_atomic_begin(struct rcar_du_crtc *crtc) { }; diff --git a/drivers/gpu/drm/rcar-du/rcar_dw_hdmi.c b/drivers/gpu/drm/rcar-du/rcar_dw_hdmi.c index 7539626b8ebd..dc85b53d58ef 100644 --- a/drivers/gpu/drm/rcar-du/rcar_dw_hdmi.c +++ b/drivers/gpu/drm/rcar-du/rcar_dw_hdmi.c @@ -45,7 +45,7 @@ static int rcar_hdmi_phy_configure(struct dw_hdmi *hdmi, { const struct rcar_hdmi_phy_params *params = rcar_hdmi_phy_params; - for (; params && params->mpixelclock != ~0UL; ++params) { + for (; params->mpixelclock != ~0UL; ++params) { if (mpixelclock <= params->mpixelclock) break; } diff --git a/drivers/gpu/drm/ttm/ttm_bo_vm.c b/drivers/gpu/drm/ttm/ttm_bo_vm.c index b442d12f2f7d..a01e5c90fd87 100644 --- a/drivers/gpu/drm/ttm/ttm_bo_vm.c +++ b/drivers/gpu/drm/ttm/ttm_bo_vm.c @@ -294,10 +294,87 @@ static void ttm_bo_vm_close(struct vm_area_struct *vma) vma->vm_private_data = NULL; } +static int ttm_bo_vm_access_kmap(struct ttm_buffer_object *bo, + unsigned long offset, + void *buf, int len, int write) +{ + unsigned long page = offset >> PAGE_SHIFT; + unsigned long bytes_left = len; + int ret; + + /* Copy a page at a time, that way no extra virtual address + * mapping is needed + */ + offset -= page << PAGE_SHIFT; + do { + unsigned long bytes = min(bytes_left, PAGE_SIZE - offset); + struct ttm_bo_kmap_obj map; + void *ptr; + bool is_iomem; + + ret = ttm_bo_kmap(bo, page, 1, &map); + if (ret) + return ret; + + ptr = (uint8_t *)ttm_kmap_obj_virtual(&map, &is_iomem) + offset; + WARN_ON_ONCE(is_iomem); + if (write) + memcpy(ptr, buf, bytes); + else + memcpy(buf, ptr, bytes); + ttm_bo_kunmap(&map); + + page++; + bytes_left -= bytes; + offset = 0; + } while (bytes_left); + + return len; +} + +static int ttm_bo_vm_access(struct vm_area_struct *vma, unsigned long addr, + void *buf, int len, int write) +{ + unsigned long offset = (addr) - vma->vm_start; + struct ttm_buffer_object *bo = vma->vm_private_data; + int ret; + + if (len < 1 || (offset + len) >> PAGE_SHIFT > bo->num_pages) + return -EIO; + + ret = ttm_bo_reserve(bo, true, false, NULL); + if (ret) + return ret; + + switch (bo->mem.mem_type) { + case TTM_PL_SYSTEM: + if (unlikely(bo->ttm->page_flags & TTM_PAGE_FLAG_SWAPPED)) { + ret = ttm_tt_swapin(bo->ttm); + if (unlikely(ret != 0)) + return ret; + } + /* fall through */ + case TTM_PL_TT: + ret = ttm_bo_vm_access_kmap(bo, offset, buf, len, write); + break; + default: + if (bo->bdev->driver->access_memory) + ret = bo->bdev->driver->access_memory( + bo, offset, buf, len, write); + else + ret = -EIO; + } + + ttm_bo_unreserve(bo); + + return ret; +} + static const struct vm_operations_struct ttm_bo_vm_ops = { .fault = ttm_bo_vm_fault, .open = ttm_bo_vm_open, - .close = ttm_bo_vm_close + .close = ttm_bo_vm_close, + .access = ttm_bo_vm_access }; static struct ttm_buffer_object *ttm_bo_vm_lookup(struct ttm_bo_device *bdev, diff --git a/drivers/media/platform/vsp1/vsp1.h b/drivers/media/platform/vsp1/vsp1.h index 847963b6e9eb..78ef838416b3 100644 --- a/drivers/media/platform/vsp1/vsp1.h +++ b/drivers/media/platform/vsp1/vsp1.h @@ -41,11 +41,11 @@ struct vsp1_rwpf; struct vsp1_sru; struct vsp1_uds; +#define VSP1_MAX_LIF 2 #define VSP1_MAX_RPF 5 #define VSP1_MAX_UDS 3 #define VSP1_MAX_WPF 4 -#define VSP1_HAS_LIF (1 << 0) #define VSP1_HAS_LUT (1 << 1) #define VSP1_HAS_SRU (1 << 2) #define VSP1_HAS_BRU (1 << 3) @@ -54,12 +54,14 @@ struct vsp1_uds; #define VSP1_HAS_WPF_HFLIP (1 << 6) #define VSP1_HAS_HGO (1 << 7) #define VSP1_HAS_HGT (1 << 8) +#define VSP1_HAS_BRS (1 << 9) struct vsp1_device_info { u32 version; const char *model; unsigned int gen; unsigned int features; + unsigned int lif_count; unsigned int rpf_count; unsigned int uds_count; unsigned int wpf_count; @@ -76,13 +78,14 @@ struct vsp1_device { struct rcar_fcp_device *fcp; struct device *bus_master; + struct vsp1_bru *brs; struct vsp1_bru *bru; struct vsp1_clu *clu; struct vsp1_hgo *hgo; struct vsp1_hgt *hgt; struct vsp1_hsit *hsi; struct vsp1_hsit *hst; - struct vsp1_lif *lif; + struct vsp1_lif *lif[VSP1_MAX_LIF]; struct vsp1_lut *lut; struct vsp1_rwpf *rpf[VSP1_MAX_RPF]; struct vsp1_sru *sru; diff --git a/drivers/media/platform/vsp1/vsp1_bru.c b/drivers/media/platform/vsp1/vsp1_bru.c index 85362c5ef57a..e8fd2ae3b3eb 100644 --- a/drivers/media/platform/vsp1/vsp1_bru.c +++ b/drivers/media/platform/vsp1/vsp1_bru.c @@ -33,7 +33,7 @@ static inline void vsp1_bru_write(struct vsp1_bru *bru, struct vsp1_dl_list *dl, u32 reg, u32 data) { - vsp1_dl_list_write(dl, reg, data); + vsp1_dl_list_write(dl, bru->base + reg, data); } /* ----------------------------------------------------------------------------- @@ -332,11 +332,14 @@ static void bru_configure(struct vsp1_entity *entity, /* * Route BRU input 1 as SRC input to the ROP unit and configure the ROP * unit with a NOP operation to make BRU input 1 available as the - * Blend/ROP unit B SRC input. + * Blend/ROP unit B SRC input. Only needed for BRU, the BRS has no ROP + * unit. */ - vsp1_bru_write(bru, dl, VI6_BRU_ROP, VI6_BRU_ROP_DSTSEL_BRUIN(1) | - VI6_BRU_ROP_CROP(VI6_ROP_NOP) | - VI6_BRU_ROP_AROP(VI6_ROP_NOP)); + if (entity->type == VSP1_ENTITY_BRU) + vsp1_bru_write(bru, dl, VI6_BRU_ROP, + VI6_BRU_ROP_DSTSEL_BRUIN(1) | + VI6_BRU_ROP_CROP(VI6_ROP_NOP) | + VI6_BRU_ROP_AROP(VI6_ROP_NOP)); for (i = 0; i < bru->entity.source_pad; ++i) { bool premultiplied = false; @@ -366,12 +369,13 @@ static void bru_configure(struct vsp1_entity *entity, ctrl |= VI6_BRU_CTRL_DSTSEL_VRPF; /* - * Route BRU inputs 0 to 3 as SRC inputs to Blend/ROP units A to - * D in that order. The Blend/ROP unit B SRC is hardwired to the - * ROP unit output, the corresponding register bits must be set - * to 0. + * Route inputs 0 to 3 as SRC inputs to Blend/ROP units A to D + * in that order. In the BRU the Blend/ROP unit B SRC is + * hardwired to the ROP unit output, the corresponding register + * bits must be set to 0. The BRS has no ROP unit and doesn't + * need any special processing. */ - if (i != 1) + if (!(entity->type == VSP1_ENTITY_BRU && i == 1)) ctrl |= VI6_BRU_CTRL_SRCSEL_BRUIN(i); vsp1_bru_write(bru, dl, VI6_BRU_CTRL(i), ctrl); @@ -407,20 +411,31 @@ static const struct vsp1_entity_operations bru_entity_ops = { * Initialization and Cleanup */ -struct vsp1_bru *vsp1_bru_create(struct vsp1_device *vsp1) +struct vsp1_bru *vsp1_bru_create(struct vsp1_device *vsp1, + enum vsp1_entity_type type) { struct vsp1_bru *bru; + unsigned int num_pads; + const char *name; int ret; bru = devm_kzalloc(vsp1->dev, sizeof(*bru), GFP_KERNEL); if (bru == NULL) return ERR_PTR(-ENOMEM); + bru->base = type == VSP1_ENTITY_BRU ? VI6_BRU_BASE : VI6_BRS_BASE; bru->entity.ops = &bru_entity_ops; - bru->entity.type = VSP1_ENTITY_BRU; + bru->entity.type = type; + + if (type == VSP1_ENTITY_BRU) { + num_pads = vsp1->info->num_bru_inputs + 1; + name = "bru"; + } else { + num_pads = 3; + name = "brs"; + } - ret = vsp1_entity_init(vsp1, &bru->entity, "bru", - vsp1->info->num_bru_inputs + 1, &bru_ops, + ret = vsp1_entity_init(vsp1, &bru->entity, name, num_pads, &bru_ops, MEDIA_ENT_F_PROC_VIDEO_COMPOSER); if (ret < 0) return ERR_PTR(ret); @@ -435,7 +450,7 @@ struct vsp1_bru *vsp1_bru_create(struct vsp1_device *vsp1) bru->entity.subdev.ctrl_handler = &bru->ctrls; if (bru->ctrls.error) { - dev_err(vsp1->dev, "bru: failed to initialize controls\n"); + dev_err(vsp1->dev, "%s: failed to initialize controls\n", name); ret = bru->ctrls.error; vsp1_entity_destroy(&bru->entity); return ERR_PTR(ret); diff --git a/drivers/media/platform/vsp1/vsp1_bru.h b/drivers/media/platform/vsp1/vsp1_bru.h index 828a3fcadea8..c98ed96d8de6 100644 --- a/drivers/media/platform/vsp1/vsp1_bru.h +++ b/drivers/media/platform/vsp1/vsp1_bru.h @@ -26,6 +26,7 @@ struct vsp1_rwpf; struct vsp1_bru { struct vsp1_entity entity; + unsigned int base; struct v4l2_ctrl_handler ctrls; @@ -41,6 +42,7 @@ static inline struct vsp1_bru *to_bru(struct v4l2_subdev *subdev) return container_of(subdev, struct vsp1_bru, entity.subdev); } -struct vsp1_bru *vsp1_bru_create(struct vsp1_device *vsp1); +struct vsp1_bru *vsp1_bru_create(struct vsp1_device *vsp1, + enum vsp1_entity_type type); #endif /* __VSP1_BRU_H__ */ diff --git a/drivers/media/platform/vsp1/vsp1_dl.c b/drivers/media/platform/vsp1/vsp1_dl.c index aaf17b13fd78..8b5cbb6b7a70 100644 --- a/drivers/media/platform/vsp1/vsp1_dl.c +++ b/drivers/media/platform/vsp1/vsp1_dl.c @@ -95,6 +95,7 @@ enum vsp1_dl_mode { * struct vsp1_dl_manager - Display List manager * @index: index of the related WPF * @mode: display list operation mode (header or headerless) + * @singleshot: execute the display list in single-shot mode * @vsp1: the VSP1 device * @lock: protects the free, active, queued, pending and gc_fragments lists * @free: array of all free display lists @@ -107,6 +108,7 @@ enum vsp1_dl_mode { struct vsp1_dl_manager { unsigned int index; enum vsp1_dl_mode mode; + bool singleshot; struct vsp1_device *vsp1; spinlock_t lock; @@ -437,6 +439,7 @@ int vsp1_dl_list_add_chain(struct vsp1_dl_list *head, static void vsp1_dl_list_fill_header(struct vsp1_dl_list *dl, bool is_last) { + struct vsp1_dl_manager *dlm = dl->dlm; struct vsp1_dl_header_list *hdr = dl->header->lists; struct vsp1_dl_body *dlb; unsigned int num_lists = 0; @@ -461,106 +464,152 @@ static void vsp1_dl_list_fill_header(struct vsp1_dl_list *dl, bool is_last) dl->header->num_lists = num_lists; - /* - * If this display list's chain is not empty, we are on a list, where - * the next item in the list is the display list entity which should be - * automatically queued by the hardware. - */ if (!list_empty(&dl->chain) && !is_last) { + /* + * If this display list's chain is not empty, we are on a list, + * and the next item is the display list that we must queue for + * automatic processing by the hardware. + */ struct vsp1_dl_list *next = list_next_entry(dl, chain); dl->header->next_header = next->dma; dl->header->flags = VSP1_DLH_AUTO_START; + } else if (!dlm->singleshot) { + /* + * if the display list manager works in continuous mode, the VSP + * should loop over the display list continuously until + * instructed to do otherwise. + */ + dl->header->next_header = dl->dma; + dl->header->flags = VSP1_DLH_INT_ENABLE | VSP1_DLH_AUTO_START; } else { + /* + * Otherwise, in mem-to-mem mode, we work in single-shot mode + * and the next display list must not be started automatically. + */ dl->header->flags = VSP1_DLH_INT_ENABLE; } } -void vsp1_dl_list_commit(struct vsp1_dl_list *dl) +static bool vsp1_dl_list_hw_update_pending(struct vsp1_dl_manager *dlm) { - struct vsp1_dl_manager *dlm = dl->dlm; struct vsp1_device *vsp1 = dlm->vsp1; - unsigned long flags; - bool update; - spin_lock_irqsave(&dlm->lock, flags); + if (!dlm->queued) + return false; - if (dl->dlm->mode == VSP1_DL_MODE_HEADER) { - struct vsp1_dl_list *dl_child; + /* + * Check whether the VSP1 has taken the update. In headerless mode the + * hardware indicates this by clearing the UPD bit in the DL_BODY_SIZE + * register, and in header mode by clearing the UPDHDR bit in the CMD + * register. + */ + if (dlm->mode == VSP1_DL_MODE_HEADERLESS) + return !!(vsp1_read(vsp1, VI6_DL_BODY_SIZE) + & VI6_DL_BODY_SIZE_UPD); + else + return !!(vsp1_read(vsp1, VI6_CMD(dlm->index) & VI6_CMD_UPDHDR)); +} +static void vsp1_dl_list_hw_enqueue(struct vsp1_dl_list *dl) +{ + struct vsp1_dl_manager *dlm = dl->dlm; + struct vsp1_device *vsp1 = dlm->vsp1; + + if (dlm->mode == VSP1_DL_MODE_HEADERLESS) { /* - * In header mode the caller guarantees that the hardware is - * idle at this point. + * In headerless mode, program the hardware directly with the + * display list body address and size and set the UPD bit. The + * bit will be cleared by the hardware when the display list + * processing starts. */ - - /* Fill the header for the head and chained display lists. */ - vsp1_dl_list_fill_header(dl, list_empty(&dl->chain)); - - list_for_each_entry(dl_child, &dl->chain, chain) { - bool last = list_is_last(&dl_child->chain, &dl->chain); - - vsp1_dl_list_fill_header(dl_child, last); - } - + vsp1_write(vsp1, VI6_DL_HDR_ADDR(0), dl->body0.dma); + vsp1_write(vsp1, VI6_DL_BODY_SIZE, VI6_DL_BODY_SIZE_UPD | + (dl->body0.num_entries * sizeof(*dl->header->lists))); + } else { /* - * Commit the head display list to hardware. Chained headers - * will auto-start. + * In header mode, program the display list header address. If + * the hardware is idle (single-shot mode or first frame in + * continuous mode) it will then be started independently. If + * the hardware is operating, the VI6_DL_HDR_REF_ADDR register + * will be updated with the display list address. */ vsp1_write(vsp1, VI6_DL_HDR_ADDR(dlm->index), dl->dma); - - dlm->active = dl; - goto done; } +} + +static void vsp1_dl_list_commit_continuous(struct vsp1_dl_list *dl) +{ + struct vsp1_dl_manager *dlm = dl->dlm; /* - * Once the UPD bit has been set the hardware can start processing the - * display list at any time and we can't touch the address and size - * registers. In that case mark the update as pending, it will be - * queued up to the hardware by the frame end interrupt handler. + * If a previous display list has been queued to the hardware but not + * processed yet, the VSP can start processing it at any time. In that + * case we can't replace the queued list by the new one, as we could + * race with the hardware. We thus mark the update as pending, it will + * be queued up to the hardware by the frame end interrupt handler. */ - update = !!(vsp1_read(vsp1, VI6_DL_BODY_SIZE) & VI6_DL_BODY_SIZE_UPD); - if (update) { + if (vsp1_dl_list_hw_update_pending(dlm)) { __vsp1_dl_list_put(dlm->pending); dlm->pending = dl; - goto done; + return; } /* - * Program the hardware with the display list body address and size. - * The UPD bit will be cleared by the device when the display list is - * processed. + * Pass the new display list to the hardware and mark it as queued. It + * will become active when the hardware starts processing it. */ - vsp1_write(vsp1, VI6_DL_HDR_ADDR(0), dl->body0.dma); - vsp1_write(vsp1, VI6_DL_BODY_SIZE, VI6_DL_BODY_SIZE_UPD | - (dl->body0.num_entries * sizeof(*dl->header->lists))); + vsp1_dl_list_hw_enqueue(dl); __vsp1_dl_list_put(dlm->queued); dlm->queued = dl; - -done: - spin_unlock_irqrestore(&dlm->lock, flags); } -/* ----------------------------------------------------------------------------- - * Display List Manager - */ - -/* Interrupt Handling */ -void vsp1_dlm_irq_display_start(struct vsp1_dl_manager *dlm) +static void vsp1_dl_list_commit_singleshot(struct vsp1_dl_list *dl) { - spin_lock(&dlm->lock); + struct vsp1_dl_manager *dlm = dl->dlm; /* - * The display start interrupt signals the end of the display list - * processing by the device. The active display list, if any, won't be - * accessed anymore and can be reused. + * When working in single-shot mode, the caller guarantees that the + * hardware is idle at this point. Just commit the head display list + * to hardware. Chained lists will be started automatically. */ - __vsp1_dl_list_put(dlm->active); - dlm->active = NULL; + vsp1_dl_list_hw_enqueue(dl); - spin_unlock(&dlm->lock); + dlm->active = dl; +} + +void vsp1_dl_list_commit(struct vsp1_dl_list *dl) +{ + struct vsp1_dl_manager *dlm = dl->dlm; + struct vsp1_dl_list *dl_child; + unsigned long flags; + + if (dlm->mode == VSP1_DL_MODE_HEADER) { + /* Fill the header for the head and chained display lists. */ + vsp1_dl_list_fill_header(dl, list_empty(&dl->chain)); + + list_for_each_entry(dl_child, &dl->chain, chain) { + bool last = list_is_last(&dl_child->chain, &dl->chain); + + vsp1_dl_list_fill_header(dl_child, last); + } + } + + spin_lock_irqsave(&dlm->lock, flags); + + if (dlm->singleshot) + vsp1_dl_list_commit_singleshot(dl); + else + vsp1_dl_list_commit_continuous(dl); + + spin_unlock_irqrestore(&dlm->lock, flags); } +/* ----------------------------------------------------------------------------- + * Display List Manager + */ + /** * vsp1_dlm_irq_frame_end - Display list handler for the frame end interrupt * @dlm: the display list manager @@ -572,31 +621,28 @@ void vsp1_dlm_irq_display_start(struct vsp1_dl_manager *dlm) */ bool vsp1_dlm_irq_frame_end(struct vsp1_dl_manager *dlm) { - struct vsp1_device *vsp1 = dlm->vsp1; bool completed = false; spin_lock(&dlm->lock); - __vsp1_dl_list_put(dlm->active); - dlm->active = NULL; - /* - * Header mode is used for mem-to-mem pipelines only. We don't need to - * perform any operation as there can't be any new display list queued - * in that case. + * The mem-to-mem pipelines work in single-shot mode. No new display + * list can be queued, we don't have to do anything. */ - if (dlm->mode == VSP1_DL_MODE_HEADER) { + if (dlm->singleshot) { + __vsp1_dl_list_put(dlm->active); + dlm->active = NULL; completed = true; goto done; } /* - * The UPD bit set indicates that the commit operation raced with the - * interrupt and occurred after the frame end event and UPD clear but - * before interrupt processing. The hardware hasn't taken the update - * into account yet, we'll thus skip one frame and retry. + * If the commit operation raced with the interrupt and occurred after + * the frame end event but before interrupt processing, the hardware + * hasn't taken the update into account yet. We have to skip one frame + * and retry. */ - if (vsp1_read(vsp1, VI6_DL_BODY_SIZE) & VI6_DL_BODY_SIZE_UPD) + if (vsp1_dl_list_hw_update_pending(dlm)) goto done; /* @@ -604,24 +650,20 @@ bool vsp1_dlm_irq_frame_end(struct vsp1_dl_manager *dlm) * frame end interrupt. The display list thus becomes active. */ if (dlm->queued) { + __vsp1_dl_list_put(dlm->active); dlm->active = dlm->queued; dlm->queued = NULL; completed = true; } /* - * Now that the UPD bit has been cleared we can queue the next display - * list to the hardware if one has been prepared. + * Now that the VSP has started processing the queued display list, we + * can queue the pending display list to the hardware if one has been + * prepared. */ if (dlm->pending) { - struct vsp1_dl_list *dl = dlm->pending; - - vsp1_write(vsp1, VI6_DL_HDR_ADDR(0), dl->body0.dma); - vsp1_write(vsp1, VI6_DL_BODY_SIZE, VI6_DL_BODY_SIZE_UPD | - (dl->body0.num_entries * - sizeof(*dl->header->lists))); - - dlm->queued = dl; + vsp1_dl_list_hw_enqueue(dlm->pending); + dlm->queued = dlm->pending; dlm->pending = NULL; } @@ -714,6 +756,7 @@ struct vsp1_dl_manager *vsp1_dlm_create(struct vsp1_device *vsp1, dlm->index = index; dlm->mode = index == 0 && !vsp1->info->uapi ? VSP1_DL_MODE_HEADERLESS : VSP1_DL_MODE_HEADER; + dlm->singleshot = vsp1->info->uapi; dlm->vsp1 = vsp1; spin_lock_init(&dlm->lock); diff --git a/drivers/media/platform/vsp1/vsp1_dl.h b/drivers/media/platform/vsp1/vsp1_dl.h index 6ec1380a10af..ee3508172f0a 100644 --- a/drivers/media/platform/vsp1/vsp1_dl.h +++ b/drivers/media/platform/vsp1/vsp1_dl.h @@ -27,7 +27,6 @@ struct vsp1_dl_manager *vsp1_dlm_create(struct vsp1_device *vsp1, unsigned int prealloc); void vsp1_dlm_destroy(struct vsp1_dl_manager *dlm); void vsp1_dlm_reset(struct vsp1_dl_manager *dlm); -void vsp1_dlm_irq_display_start(struct vsp1_dl_manager *dlm); bool vsp1_dlm_irq_frame_end(struct vsp1_dl_manager *dlm); struct vsp1_dl_list *vsp1_dl_list_get(struct vsp1_dl_manager *dlm); diff --git a/drivers/media/platform/vsp1/vsp1_drm.c b/drivers/media/platform/vsp1/vsp1_drm.c index 9377aafa8996..4dfbeac8f42c 100644 --- a/drivers/media/platform/vsp1/vsp1_drm.c +++ b/drivers/media/platform/vsp1/vsp1_drm.c @@ -32,17 +32,13 @@ * Interrupt Handling */ -void vsp1_drm_display_start(struct vsp1_device *vsp1) +static void vsp1_du_pipeline_frame_end(struct vsp1_pipeline *pipe, + bool completed) { - vsp1_dlm_irq_display_start(vsp1->drm->pipe.output->dlm); -} - -static void vsp1_du_pipeline_frame_end(struct vsp1_pipeline *pipe) -{ - struct vsp1_drm *drm = to_vsp1_drm(pipe); + struct vsp1_drm_pipeline *drm_pipe = to_vsp1_drm_pipeline(pipe); - if (drm->du_complete) - drm->du_complete(drm->du_private); + if (drm_pipe->du_complete) + drm_pipe->du_complete(drm_pipe->du_private, completed); } /* ----------------------------------------------------------------------------- @@ -63,29 +59,44 @@ EXPORT_SYMBOL_GPL(vsp1_du_init); /** * vsp1_du_setup_lif - Setup the output part of the VSP pipeline * @dev: the VSP device + * @pipe_index: the DRM pipeline index * @cfg: the LIF configuration * * Configure the output part of VSP DRM pipeline for the given frame @cfg.width - * and @cfg.height. This sets up formats on the BRU source pad, the WPF0 sink - * and source pads, and the LIF sink pad. + * and @cfg.height. This sets up formats on the blend unit (BRU or BRS) source + * pad, the WPF sink and source pads, and the LIF sink pad. + * + * The @pipe_index argument selects which DRM pipeline to setup. The number of + * available pipelines depend on the VSP instance. * - * As the media bus code on the BRU source pad is conditioned by the - * configuration of the BRU sink 0 pad, we also set up the formats on all BRU + * As the media bus code on the blend unit source pad is conditioned by the + * configuration of its sink 0 pad, we also set up the formats on all blend unit * sinks, even if the configuration will be overwritten later by - * vsp1_du_setup_rpf(). This ensures that the BRU configuration is set to a well - * defined state. + * vsp1_du_setup_rpf(). This ensures that the blend unit configuration is set to + * a well defined state. * * Return 0 on success or a negative error code on failure. */ -int vsp1_du_setup_lif(struct device *dev, const struct vsp1_du_lif_config *cfg) +int vsp1_du_setup_lif(struct device *dev, unsigned int pipe_index, + const struct vsp1_du_lif_config *cfg) { struct vsp1_device *vsp1 = dev_get_drvdata(dev); - struct vsp1_pipeline *pipe = &vsp1->drm->pipe; - struct vsp1_bru *bru = vsp1->bru; + struct vsp1_drm_pipeline *drm_pipe; + struct vsp1_pipeline *pipe; + struct vsp1_bru *bru; struct v4l2_subdev_format format; + const char *bru_name; unsigned int i; int ret; + if (pipe_index >= vsp1->info->lif_count) + return -EINVAL; + + drm_pipe = &vsp1->drm->pipe[pipe_index]; + pipe = &drm_pipe->pipe; + bru = to_bru(&pipe->bru->subdev); + bru_name = pipe->bru->type == VSP1_ENTITY_BRU ? "BRU" : "BRS"; + if (!cfg) { /* * NULL configuration means the CRTC is being disabled, stop @@ -97,14 +108,25 @@ int vsp1_du_setup_lif(struct device *dev, const struct vsp1_du_lif_config *cfg) media_pipeline_stop(&pipe->output->entity.subdev.entity); - for (i = 0; i < bru->entity.source_pad; ++i) { - vsp1->drm->inputs[i].enabled = false; - bru->inputs[i].rpf = NULL; + for (i = 0; i < ARRAY_SIZE(pipe->inputs); ++i) { + struct vsp1_rwpf *rpf = pipe->inputs[i]; + + if (!rpf) + continue; + + /* + * Remove the RPF from the pipe and the list of BRU + * inputs. + */ + WARN_ON(list_empty(&rpf->entity.list_pipe)); + list_del_init(&rpf->entity.list_pipe); pipe->inputs[i] = NULL; + + bru->inputs[rpf->bru_input].rpf = NULL; } + drm_pipe->du_complete = NULL; pipe->num_inputs = 0; - vsp1->drm->du_complete = NULL; vsp1_dlm_reset(pipe->output->dlm); vsp1_device_put(vsp1); @@ -114,8 +136,8 @@ int vsp1_du_setup_lif(struct device *dev, const struct vsp1_du_lif_config *cfg) return 0; } - dev_dbg(vsp1->dev, "%s: configuring LIF with format %ux%u\n", - __func__, cfg->width, cfg->height); + dev_dbg(vsp1->dev, "%s: configuring LIF%u with format %ux%u\n", + __func__, pipe_index, cfg->width, cfg->height); /* * Configure the format at the BRU sinks and propagate it through the @@ -124,7 +146,7 @@ int vsp1_du_setup_lif(struct device *dev, const struct vsp1_du_lif_config *cfg) memset(&format, 0, sizeof(format)); format.which = V4L2_SUBDEV_FORMAT_ACTIVE; - for (i = 0; i < bru->entity.source_pad; ++i) { + for (i = 0; i < pipe->bru->source_pad; ++i) { format.pad = i; format.format.width = cfg->width; @@ -132,60 +154,60 @@ int vsp1_du_setup_lif(struct device *dev, const struct vsp1_du_lif_config *cfg) format.format.code = MEDIA_BUS_FMT_ARGB8888_1X32; format.format.field = V4L2_FIELD_NONE; - ret = v4l2_subdev_call(&bru->entity.subdev, pad, + ret = v4l2_subdev_call(&pipe->bru->subdev, pad, set_fmt, NULL, &format); if (ret < 0) return ret; - dev_dbg(vsp1->dev, "%s: set format %ux%u (%x) on BRU pad %u\n", + dev_dbg(vsp1->dev, "%s: set format %ux%u (%x) on %s pad %u\n", __func__, format.format.width, format.format.height, - format.format.code, i); + format.format.code, bru_name, i); } - format.pad = bru->entity.source_pad; + format.pad = pipe->bru->source_pad; format.format.width = cfg->width; format.format.height = cfg->height; format.format.code = MEDIA_BUS_FMT_ARGB8888_1X32; format.format.field = V4L2_FIELD_NONE; - ret = v4l2_subdev_call(&bru->entity.subdev, pad, set_fmt, NULL, + ret = v4l2_subdev_call(&pipe->bru->subdev, pad, set_fmt, NULL, &format); if (ret < 0) return ret; - dev_dbg(vsp1->dev, "%s: set format %ux%u (%x) on BRU pad %u\n", + dev_dbg(vsp1->dev, "%s: set format %ux%u (%x) on %s pad %u\n", __func__, format.format.width, format.format.height, - format.format.code, i); + format.format.code, bru_name, i); format.pad = RWPF_PAD_SINK; - ret = v4l2_subdev_call(&vsp1->wpf[0]->entity.subdev, pad, set_fmt, NULL, + ret = v4l2_subdev_call(&pipe->output->entity.subdev, pad, set_fmt, NULL, &format); if (ret < 0) return ret; - dev_dbg(vsp1->dev, "%s: set format %ux%u (%x) on WPF0 sink\n", + dev_dbg(vsp1->dev, "%s: set format %ux%u (%x) on WPF%u sink\n", __func__, format.format.width, format.format.height, - format.format.code); + format.format.code, pipe->output->entity.index); format.pad = RWPF_PAD_SOURCE; - ret = v4l2_subdev_call(&vsp1->wpf[0]->entity.subdev, pad, get_fmt, NULL, + ret = v4l2_subdev_call(&pipe->output->entity.subdev, pad, get_fmt, NULL, &format); if (ret < 0) return ret; - dev_dbg(vsp1->dev, "%s: got format %ux%u (%x) on WPF0 source\n", + dev_dbg(vsp1->dev, "%s: got format %ux%u (%x) on WPF%u source\n", __func__, format.format.width, format.format.height, - format.format.code); + format.format.code, pipe->output->entity.index); format.pad = LIF_PAD_SINK; - ret = v4l2_subdev_call(&vsp1->lif->entity.subdev, pad, set_fmt, NULL, + ret = v4l2_subdev_call(&pipe->lif->subdev, pad, set_fmt, NULL, &format); if (ret < 0) return ret; - dev_dbg(vsp1->dev, "%s: set format %ux%u (%x) on LIF sink\n", + dev_dbg(vsp1->dev, "%s: set format %ux%u (%x) on LIF%u sink\n", __func__, format.format.width, format.format.height, - format.format.code); + format.format.code, pipe_index); /* * Verify that the format at the output of the pipeline matches the @@ -213,8 +235,8 @@ int vsp1_du_setup_lif(struct device *dev, const struct vsp1_du_lif_config *cfg) * Register a callback to allow us to notify the DRM driver of frame * completion events. */ - vsp1->drm->du_complete = cfg->callback; - vsp1->drm->du_private = cfg->callback_data; + drm_pipe->du_complete = cfg->callback; + drm_pipe->du_private = cfg->callback_data; ret = media_pipeline_start(&pipe->output->entity.subdev.entity, &pipe->pipe); @@ -224,6 +246,10 @@ int vsp1_du_setup_lif(struct device *dev, const struct vsp1_du_lif_config *cfg) return ret; } + /* Disable the display interrupts. */ + vsp1_write(vsp1, VI6_DISP_IRQ_STA, 0); + vsp1_write(vsp1, VI6_DISP_IRQ_ENB, 0); + dev_dbg(vsp1->dev, "%s: pipeline enabled\n", __func__); return 0; @@ -233,19 +259,21 @@ EXPORT_SYMBOL_GPL(vsp1_du_setup_lif); /** * vsp1_du_atomic_begin - Prepare for an atomic update * @dev: the VSP device + * @pipe_index: the DRM pipeline index */ -void vsp1_du_atomic_begin(struct device *dev) +void vsp1_du_atomic_begin(struct device *dev, unsigned int pipe_index) { struct vsp1_device *vsp1 = dev_get_drvdata(dev); - struct vsp1_pipeline *pipe = &vsp1->drm->pipe; + struct vsp1_drm_pipeline *drm_pipe = &vsp1->drm->pipe[pipe_index]; - vsp1->drm->num_inputs = pipe->num_inputs; + drm_pipe->enabled = drm_pipe->pipe.num_inputs != 0; } EXPORT_SYMBOL_GPL(vsp1_du_atomic_begin); /** * vsp1_du_atomic_update - Setup one RPF input of the VSP pipeline * @dev: the VSP device + * @pipe_index: the DRM pipeline index * @rpf_index: index of the RPF to setup (0-based) * @cfg: the RPF configuration * @@ -272,10 +300,12 @@ EXPORT_SYMBOL_GPL(vsp1_du_atomic_begin); * * Return 0 on success or a negative error code on failure. */ -int vsp1_du_atomic_update(struct device *dev, unsigned int rpf_index, +int vsp1_du_atomic_update(struct device *dev, unsigned int pipe_index, + unsigned int rpf_index, const struct vsp1_du_atomic_config *cfg) { struct vsp1_device *vsp1 = dev_get_drvdata(dev); + struct vsp1_drm_pipeline *drm_pipe = &vsp1->drm->pipe[pipe_index]; const struct vsp1_format_info *fmtinfo; struct vsp1_rwpf *rpf; @@ -288,7 +318,12 @@ int vsp1_du_atomic_update(struct device *dev, unsigned int rpf_index, dev_dbg(vsp1->dev, "%s: RPF%u: disable requested\n", __func__, rpf_index); - vsp1->drm->inputs[rpf_index].enabled = false; + /* + * Remove the RPF from the pipe's inputs. The atomic flush + * handler will disable the input and remove the entity from the + * pipe's entities list. + */ + drm_pipe->pipe.inputs[rpf_index] = NULL; return 0; } @@ -324,13 +359,15 @@ int vsp1_du_atomic_update(struct device *dev, unsigned int rpf_index, vsp1->drm->inputs[rpf_index].crop = cfg->src; vsp1->drm->inputs[rpf_index].compose = cfg->dst; vsp1->drm->inputs[rpf_index].zpos = cfg->zpos; - vsp1->drm->inputs[rpf_index].enabled = true; + + drm_pipe->pipe.inputs[rpf_index] = rpf; return 0; } EXPORT_SYMBOL_GPL(vsp1_du_atomic_update); static int vsp1_du_setup_rpf_pipe(struct vsp1_device *vsp1, + struct vsp1_pipeline *pipe, struct vsp1_rwpf *rpf, unsigned int bru_input) { struct v4l2_subdev_selection sel; @@ -404,7 +441,7 @@ static int vsp1_du_setup_rpf_pipe(struct vsp1_device *vsp1, /* BRU sink, propagate the format from the RPF source. */ format.pad = bru_input; - ret = v4l2_subdev_call(&vsp1->bru->entity.subdev, pad, set_fmt, NULL, + ret = v4l2_subdev_call(&pipe->bru->subdev, pad, set_fmt, NULL, &format); if (ret < 0) return ret; @@ -417,8 +454,8 @@ static int vsp1_du_setup_rpf_pipe(struct vsp1_device *vsp1, sel.target = V4L2_SEL_TGT_COMPOSE; sel.r = vsp1->drm->inputs[rpf->entity.index].compose; - ret = v4l2_subdev_call(&vsp1->bru->entity.subdev, pad, set_selection, - NULL, &sel); + ret = v4l2_subdev_call(&pipe->bru->subdev, pad, set_selection, NULL, + &sel); if (ret < 0) return ret; @@ -438,18 +475,25 @@ static unsigned int rpf_zpos(struct vsp1_device *vsp1, struct vsp1_rwpf *rpf) /** * vsp1_du_atomic_flush - Commit an atomic update * @dev: the VSP device + * @pipe_index: the DRM pipeline index */ -void vsp1_du_atomic_flush(struct device *dev) +void vsp1_du_atomic_flush(struct device *dev, unsigned int pipe_index) { struct vsp1_device *vsp1 = dev_get_drvdata(dev); - struct vsp1_pipeline *pipe = &vsp1->drm->pipe; + struct vsp1_drm_pipeline *drm_pipe = &vsp1->drm->pipe[pipe_index]; + struct vsp1_pipeline *pipe = &drm_pipe->pipe; struct vsp1_rwpf *inputs[VSP1_MAX_RPF] = { NULL, }; + struct vsp1_bru *bru = to_bru(&pipe->bru->subdev); struct vsp1_entity *entity; + struct vsp1_entity *next; struct vsp1_dl_list *dl; + const char *bru_name; unsigned long flags; unsigned int i; int ret; + bru_name = pipe->bru->type == VSP1_ENTITY_BRU ? "BRU" : "BRS"; + /* Prepare the display list. */ dl = vsp1_dl_list_get(pipe->output->dlm); @@ -460,12 +504,8 @@ void vsp1_du_atomic_flush(struct device *dev) struct vsp1_rwpf *rpf = vsp1->rpf[i]; unsigned int j; - if (!vsp1->drm->inputs[i].enabled) { - pipe->inputs[i] = NULL; + if (!pipe->inputs[i]) continue; - } - - pipe->inputs[i] = rpf; /* Insert the RPF in the sorted RPFs array. */ for (j = pipe->num_inputs++; j > 0; --j) { @@ -478,22 +518,26 @@ void vsp1_du_atomic_flush(struct device *dev) } /* Setup the RPF input pipeline for every enabled input. */ - for (i = 0; i < vsp1->info->num_bru_inputs; ++i) { + for (i = 0; i < pipe->bru->source_pad; ++i) { struct vsp1_rwpf *rpf = inputs[i]; if (!rpf) { - vsp1->bru->inputs[i].rpf = NULL; + bru->inputs[i].rpf = NULL; continue; } - vsp1->bru->inputs[i].rpf = rpf; + if (list_empty(&rpf->entity.list_pipe)) + list_add_tail(&rpf->entity.list_pipe, &pipe->entities); + + bru->inputs[i].rpf = rpf; rpf->bru_input = i; + rpf->entity.sink = pipe->bru; rpf->entity.sink_pad = i; - dev_dbg(vsp1->dev, "%s: connecting RPF.%u to BRU:%u\n", - __func__, rpf->entity.index, i); + dev_dbg(vsp1->dev, "%s: connecting RPF.%u to %s:%u\n", + __func__, rpf->entity.index, bru_name, i); - ret = vsp1_du_setup_rpf_pipe(vsp1, rpf, i); + ret = vsp1_du_setup_rpf_pipe(vsp1, pipe, rpf, i); if (ret < 0) dev_err(vsp1->dev, "%s: failed to setup RPF.%u\n", @@ -501,16 +545,16 @@ void vsp1_du_atomic_flush(struct device *dev) } /* Configure all entities in the pipeline. */ - list_for_each_entry(entity, &pipe->entities, list_pipe) { + list_for_each_entry_safe(entity, next, &pipe->entities, list_pipe) { /* Disconnect unused RPFs from the pipeline. */ - if (entity->type == VSP1_ENTITY_RPF) { - struct vsp1_rwpf *rpf = to_rwpf(&entity->subdev); + if (entity->type == VSP1_ENTITY_RPF && + !pipe->inputs[entity->index]) { + vsp1_dl_list_write(dl, entity->route->reg, + VI6_DPR_NODE_UNUSED); - if (!pipe->inputs[rpf->entity.index]) { - vsp1_dl_list_write(dl, entity->route->reg, - VI6_DPR_NODE_UNUSED); - continue; - } + list_del_init(&entity->list_pipe); + + continue; } vsp1_entity_route_setup(entity, pipe, dl); @@ -528,14 +572,11 @@ void vsp1_du_atomic_flush(struct device *dev) vsp1_dl_list_commit(dl); /* Start or stop the pipeline if needed. */ - if (!vsp1->drm->num_inputs && pipe->num_inputs) { - vsp1_write(vsp1, VI6_DISP_IRQ_STA, 0); - vsp1_write(vsp1, VI6_DISP_IRQ_ENB, VI6_DISP_IRQ_ENB_DSTE); + if (!drm_pipe->enabled && pipe->num_inputs) { spin_lock_irqsave(&pipe->irqlock, flags); vsp1_pipeline_run(pipe); spin_unlock_irqrestore(&pipe->irqlock, flags); - } else if (vsp1->drm->num_inputs && !pipe->num_inputs) { - vsp1_write(vsp1, VI6_DISP_IRQ_ENB, 0); + } else if (drm_pipe->enabled && !pipe->num_inputs) { vsp1_pipeline_stop(pipe); } } @@ -568,83 +609,48 @@ EXPORT_SYMBOL_GPL(vsp1_du_unmap_sg); * Initialization */ -int vsp1_drm_create_links(struct vsp1_device *vsp1) -{ - const u32 flags = MEDIA_LNK_FL_ENABLED | MEDIA_LNK_FL_IMMUTABLE; - unsigned int i; - int ret; - - /* - * VSPD instances require a BRU to perform composition and a LIF to - * output to the DU. - */ - if (!vsp1->bru || !vsp1->lif) - return -ENXIO; - - for (i = 0; i < vsp1->info->rpf_count; ++i) { - struct vsp1_rwpf *rpf = vsp1->rpf[i]; - - ret = media_create_pad_link(&rpf->entity.subdev.entity, - RWPF_PAD_SOURCE, - &vsp1->bru->entity.subdev.entity, - i, flags); - if (ret < 0) - return ret; - - rpf->entity.sink = &vsp1->bru->entity.subdev.entity; - rpf->entity.sink_pad = i; - } - - ret = media_create_pad_link(&vsp1->bru->entity.subdev.entity, - vsp1->bru->entity.source_pad, - &vsp1->wpf[0]->entity.subdev.entity, - RWPF_PAD_SINK, flags); - if (ret < 0) - return ret; - - vsp1->bru->entity.sink = &vsp1->wpf[0]->entity.subdev.entity; - vsp1->bru->entity.sink_pad = RWPF_PAD_SINK; - - ret = media_create_pad_link(&vsp1->wpf[0]->entity.subdev.entity, - RWPF_PAD_SOURCE, - &vsp1->lif->entity.subdev.entity, - LIF_PAD_SINK, flags); - if (ret < 0) - return ret; - - return 0; -} - int vsp1_drm_init(struct vsp1_device *vsp1) { - struct vsp1_pipeline *pipe; unsigned int i; vsp1->drm = devm_kzalloc(vsp1->dev, sizeof(*vsp1->drm), GFP_KERNEL); if (!vsp1->drm) return -ENOMEM; - pipe = &vsp1->drm->pipe; + /* Create one DRM pipeline per LIF. */ + for (i = 0; i < vsp1->info->lif_count; ++i) { + struct vsp1_drm_pipeline *drm_pipe = &vsp1->drm->pipe[i]; + struct vsp1_pipeline *pipe = &drm_pipe->pipe; - vsp1_pipeline_init(pipe); + vsp1_pipeline_init(pipe); - /* The DRM pipeline is static, add entities manually. */ + /* + * The DRM pipeline is static, add entities manually. The first + * pipeline uses the BRU and the second pipeline the BRS. + */ + pipe->bru = i == 0 ? &vsp1->bru->entity : &vsp1->brs->entity; + pipe->lif = &vsp1->lif[i]->entity; + pipe->output = vsp1->wpf[i]; + pipe->output->pipe = pipe; + pipe->frame_end = vsp1_du_pipeline_frame_end; + + pipe->bru->sink = &pipe->output->entity; + pipe->bru->sink_pad = 0; + pipe->output->entity.sink = pipe->lif; + pipe->output->entity.sink_pad = 0; + + list_add_tail(&pipe->bru->list_pipe, &pipe->entities); + list_add_tail(&pipe->lif->list_pipe, &pipe->entities); + list_add_tail(&pipe->output->entity.list_pipe, &pipe->entities); + } + + /* Disable all RPFs initially. */ for (i = 0; i < vsp1->info->rpf_count; ++i) { struct vsp1_rwpf *input = vsp1->rpf[i]; - list_add_tail(&input->entity.list_pipe, &pipe->entities); + INIT_LIST_HEAD(&input->entity.list_pipe); } - list_add_tail(&vsp1->bru->entity.list_pipe, &pipe->entities); - list_add_tail(&vsp1->wpf[0]->entity.list_pipe, &pipe->entities); - list_add_tail(&vsp1->lif->entity.list_pipe, &pipe->entities); - - pipe->bru = &vsp1->bru->entity; - pipe->lif = &vsp1->lif->entity; - pipe->output = vsp1->wpf[0]; - pipe->output->pipe = pipe; - pipe->frame_end = vsp1_du_pipeline_frame_end; - return 0; } diff --git a/drivers/media/platform/vsp1/vsp1_drm.h b/drivers/media/platform/vsp1/vsp1_drm.h index e9f80727ff92..1cd9db785bf7 100644 --- a/drivers/media/platform/vsp1/vsp1_drm.h +++ b/drivers/media/platform/vsp1/vsp1_drm.h @@ -18,38 +18,44 @@ #include "vsp1_pipe.h" /** - * vsp1_drm - State for the API exposed to the DRM driver + * vsp1_drm_pipeline - State for the API exposed to the DRM driver * @pipe: the VSP1 pipeline used for display - * @num_inputs: number of active pipeline inputs at the beginning of an update - * @inputs: source crop rectangle, destination compose rectangle and z-order - * position for every input + * @enabled: pipeline state at the beginning of an update * @du_complete: frame completion callback for the DU driver (optional) * @du_private: data to be passed to the du_complete callback */ -struct vsp1_drm { +struct vsp1_drm_pipeline { struct vsp1_pipeline pipe; - unsigned int num_inputs; + bool enabled; + + /* Frame synchronisation */ + void (*du_complete)(void *, bool); + void *du_private; +}; + +/** + * vsp1_drm - State for the API exposed to the DRM driver + * @pipe: the VSP1 DRM pipeline used for display + * @inputs: source crop rectangle, destination compose rectangle and z-order + * position for every input (indexed by RPF index) + */ +struct vsp1_drm { + struct vsp1_drm_pipeline pipe[VSP1_MAX_LIF]; + struct { - bool enabled; struct v4l2_rect crop; struct v4l2_rect compose; unsigned int zpos; } inputs[VSP1_MAX_RPF]; - - /* Frame synchronisation */ - void (*du_complete)(void *); - void *du_private; }; -static inline struct vsp1_drm *to_vsp1_drm(struct vsp1_pipeline *pipe) +static inline struct vsp1_drm_pipeline * +to_vsp1_drm_pipeline(struct vsp1_pipeline *pipe) { - return container_of(pipe, struct vsp1_drm, pipe); + return container_of(pipe, struct vsp1_drm_pipeline, pipe); } int vsp1_drm_init(struct vsp1_device *vsp1); void vsp1_drm_cleanup(struct vsp1_device *vsp1); -int vsp1_drm_create_links(struct vsp1_device *vsp1); - -void vsp1_drm_display_start(struct vsp1_device *vsp1); #endif /* __VSP1_DRM_H__ */ diff --git a/drivers/media/platform/vsp1/vsp1_drv.c b/drivers/media/platform/vsp1/vsp1_drv.c index 95c26edead85..962e4c304076 100644 --- a/drivers/media/platform/vsp1/vsp1_drv.c +++ b/drivers/media/platform/vsp1/vsp1_drv.c @@ -68,14 +68,6 @@ static irqreturn_t vsp1_irq_handler(int irq, void *data) } } - status = vsp1_read(vsp1, VI6_DISP_IRQ_STA); - vsp1_write(vsp1, VI6_DISP_IRQ_STA, ~status & VI6_DISP_IRQ_STA_DST); - - if (status & VI6_DISP_IRQ_STA_DST) { - vsp1_drm_display_start(vsp1); - ret = IRQ_HANDLED; - } - return ret; } @@ -92,6 +84,10 @@ static irqreturn_t vsp1_irq_handler(int irq, void *data) * * - from a UDS to a UDS (UDS entities can't be chained) * - from an entity to itself (no loops are allowed) + * + * Furthermore, the BRS can't be connected to histogram generators, but no + * special check is currently needed as all VSP instances that include a BRS + * have no histogram generator. */ static int vsp1_create_sink_links(struct vsp1_device *vsp1, struct vsp1_entity *sink) @@ -129,7 +125,7 @@ static int vsp1_create_sink_links(struct vsp1_device *vsp1, return ret; if (flags & MEDIA_LNK_FL_ENABLED) - source->sink = entity; + source->sink = sink; } } @@ -172,10 +168,13 @@ static int vsp1_uapi_create_links(struct vsp1_device *vsp1) return ret; } - if (vsp1->lif) { - ret = media_create_pad_link(&vsp1->wpf[0]->entity.subdev.entity, + for (i = 0; i < vsp1->info->lif_count; ++i) { + if (!vsp1->lif[i]) + continue; + + ret = media_create_pad_link(&vsp1->wpf[i]->entity.subdev.entity, RWPF_PAD_SOURCE, - &vsp1->lif->entity.subdev.entity, + &vsp1->lif[i]->entity.subdev.entity, LIF_PAD_SINK, 0); if (ret < 0) return ret; @@ -269,8 +268,18 @@ static int vsp1_create_entities(struct vsp1_device *vsp1) } /* Instantiate all the entities. */ + if (vsp1->info->features & VSP1_HAS_BRS) { + vsp1->brs = vsp1_bru_create(vsp1, VSP1_ENTITY_BRS); + if (IS_ERR(vsp1->brs)) { + ret = PTR_ERR(vsp1->brs); + goto done; + } + + list_add_tail(&vsp1->brs->entity.list_dev, &vsp1->entities); + } + if (vsp1->info->features & VSP1_HAS_BRU) { - vsp1->bru = vsp1_bru_create(vsp1); + vsp1->bru = vsp1_bru_create(vsp1, VSP1_ENTITY_BRU); if (IS_ERR(vsp1->bru)) { ret = PTR_ERR(vsp1->bru); goto done; @@ -328,18 +337,23 @@ static int vsp1_create_entities(struct vsp1_device *vsp1) } /* - * The LIF is only supported when used in conjunction with the DU, in + * The LIFs are only supported when used in conjunction with the DU, in * which case the userspace API is disabled. If the userspace API is - * enabled skip the LIF, even when present. + * enabled skip the LIFs, even when present. */ - if (vsp1->info->features & VSP1_HAS_LIF && !vsp1->info->uapi) { - vsp1->lif = vsp1_lif_create(vsp1); - if (IS_ERR(vsp1->lif)) { - ret = PTR_ERR(vsp1->lif); - goto done; - } + if (!vsp1->info->uapi) { + for (i = 0; i < vsp1->info->lif_count; ++i) { + struct vsp1_lif *lif; + + lif = vsp1_lif_create(vsp1, i); + if (IS_ERR(lif)) { + ret = PTR_ERR(lif); + goto done; + } - list_add_tail(&vsp1->lif->entity.list_dev, &vsp1->entities); + vsp1->lif[i] = lif; + list_add_tail(&lif->entity.list_dev, &vsp1->entities); + } } if (vsp1->info->features & VSP1_HAS_LUT) { @@ -420,7 +434,6 @@ static int vsp1_create_entities(struct vsp1_device *vsp1) } list_add_tail(&video->list, &vsp1->videos); - wpf->entity.sink = &video->video.entity; } } @@ -432,19 +445,15 @@ static int vsp1_create_entities(struct vsp1_device *vsp1) goto done; } - /* Create links. */ - if (vsp1->info->uapi) - ret = vsp1_uapi_create_links(vsp1); - else - ret = vsp1_drm_create_links(vsp1); - if (ret < 0) - goto done; - /* - * Register subdev nodes if the userspace API is enabled or initialize - * the DRM pipeline otherwise. + * Create links and register subdev nodes if the userspace API is + * enabled or initialize the DRM pipeline otherwise. */ if (vsp1->info->uapi) { + ret = vsp1_uapi_create_links(vsp1); + if (ret < 0) + goto done; + ret = v4l2_device_register_subdev_nodes(&vsp1->v4l2_dev); if (ret < 0) goto done; @@ -515,6 +524,9 @@ static int vsp1_device_init(struct vsp1_device *vsp1) vsp1_write(vsp1, VI6_DPR_HSI_ROUTE, VI6_DPR_NODE_UNUSED); vsp1_write(vsp1, VI6_DPR_BRU_ROUTE, VI6_DPR_NODE_UNUSED); + if (vsp1->info->features & VSP1_HAS_BRS) + vsp1_write(vsp1, VI6_DPR_ILV_BRS_ROUTE, VI6_DPR_NODE_UNUSED); + vsp1_write(vsp1, VI6_DPR_HGO_SMPPT, (7 << VI6_DPR_SMPPT_TGW_SHIFT) | (VI6_DPR_NODE_UNUSED << VI6_DPR_SMPPT_PT_SHIFT)); vsp1_write(vsp1, VI6_DPR_HGT_SMPPT, (7 << VI6_DPR_SMPPT_TGW_SHIFT) | @@ -634,8 +646,8 @@ static const struct vsp1_device_info vsp1_device_infos[] = { .version = VI6_IP_VERSION_MODEL_VSPD_GEN2, .model = "VSP1-D", .gen = 2, - .features = VSP1_HAS_BRU | VSP1_HAS_HGO | VSP1_HAS_LIF - | VSP1_HAS_LUT, + .features = VSP1_HAS_BRU | VSP1_HAS_HGO | VSP1_HAS_LUT, + .lif_count = 1, .rpf_count = 4, .uds_count = 1, .wpf_count = 1, @@ -668,8 +680,8 @@ static const struct vsp1_device_info vsp1_device_infos[] = { .version = VI6_IP_VERSION_MODEL_VSPD_V2H, .model = "VSP1V-D", .gen = 2, - .features = VSP1_HAS_BRU | VSP1_HAS_CLU | VSP1_HAS_LUT - | VSP1_HAS_LIF, + .features = VSP1_HAS_BRU | VSP1_HAS_CLU | VSP1_HAS_LUT, + .lif_count = 1, .rpf_count = 4, .uds_count = 1, .wpf_count = 1, @@ -706,10 +718,37 @@ static const struct vsp1_device_info vsp1_device_infos[] = { .num_bru_inputs = 5, .uapi = true, }, { + .version = VI6_IP_VERSION_MODEL_VSPBS_GEN3, + .model = "VSP2-BS", + .gen = 3, + .features = VSP1_HAS_BRS | VSP1_HAS_WPF_VFLIP, + .rpf_count = 2, + .wpf_count = 1, + .uapi = true, + }, { .version = VI6_IP_VERSION_MODEL_VSPD_GEN3, .model = "VSP2-D", .gen = 3, - .features = VSP1_HAS_BRU | VSP1_HAS_LIF | VSP1_HAS_WPF_VFLIP, + .features = VSP1_HAS_BRU | VSP1_HAS_WPF_VFLIP, + .lif_count = 1, + .rpf_count = 5, + .wpf_count = 2, + .num_bru_inputs = 5, + }, { + .version = VI6_IP_VERSION_MODEL_VSPD_V3, + .model = "VSP2-D", + .gen = 3, + .features = VSP1_HAS_BRS | VSP1_HAS_BRU, + .lif_count = 1, + .rpf_count = 5, + .wpf_count = 1, + .num_bru_inputs = 5, + }, { + .version = VI6_IP_VERSION_MODEL_VSPDL_GEN3, + .model = "VSP2-DL", + .gen = 3, + .features = VSP1_HAS_BRS | VSP1_HAS_BRU, + .lif_count = 2, .rpf_count = 5, .wpf_count = 2, .num_bru_inputs = 5, diff --git a/drivers/media/platform/vsp1/vsp1_entity.c b/drivers/media/platform/vsp1/vsp1_entity.c index 4bdb3b141611..54de15095709 100644 --- a/drivers/media/platform/vsp1/vsp1_entity.c +++ b/drivers/media/platform/vsp1/vsp1_entity.c @@ -24,18 +24,12 @@ #include "vsp1_pipe.h" #include "vsp1_rwpf.h" -static inline struct vsp1_entity * -media_entity_to_vsp1_entity(struct media_entity *entity) -{ - return container_of(entity, struct vsp1_entity, subdev.entity); -} - void vsp1_entity_route_setup(struct vsp1_entity *entity, struct vsp1_pipeline *pipe, struct vsp1_dl_list *dl) { struct vsp1_entity *source; - struct vsp1_entity *sink; + u32 route; if (entity->type == VSP1_ENTITY_HGO) { u32 smppt; @@ -44,7 +38,7 @@ void vsp1_entity_route_setup(struct vsp1_entity *entity, * The HGO is a special case, its routing is configured on the * sink pad. */ - source = media_entity_to_vsp1_entity(entity->sources[0]); + source = entity->sources[0]; smppt = (pipe->output->entity.index << VI6_DPR_SMPPT_TGW_SHIFT) | (source->route->output << VI6_DPR_SMPPT_PT_SHIFT); @@ -57,7 +51,7 @@ void vsp1_entity_route_setup(struct vsp1_entity *entity, * The HGT is a special case, its routing is configured on the * sink pad. */ - source = media_entity_to_vsp1_entity(entity->sources[0]); + source = entity->sources[0]; smppt = (pipe->output->entity.index << VI6_DPR_SMPPT_TGW_SHIFT) | (source->route->output << VI6_DPR_SMPPT_PT_SHIFT); @@ -69,9 +63,14 @@ void vsp1_entity_route_setup(struct vsp1_entity *entity, if (source->route->reg == 0) return; - sink = media_entity_to_vsp1_entity(source->sink); - vsp1_dl_list_write(dl, source->route->reg, - sink->route->inputs[source->sink_pad]); + route = source->sink->route->inputs[source->sink_pad]; + /* + * The ILV and BRS share the same data path route. The extra BRSSEL bit + * selects between the ILV and BRS. + */ + if (source->type == VSP1_ENTITY_BRS) + route |= VI6_DPR_ROUTE_BRSSEL; + vsp1_dl_list_write(dl, source->route->reg, route); } /* ----------------------------------------------------------------------------- @@ -316,6 +315,12 @@ done: * Media Operations */ +static inline struct vsp1_entity * +media_entity_to_vsp1_entity(struct media_entity *entity) +{ + return container_of(entity, struct vsp1_entity, subdev.entity); +} + static int vsp1_entity_link_setup_source(const struct media_pad *source_pad, const struct media_pad *sink_pad, u32 flags) @@ -339,7 +344,7 @@ static int vsp1_entity_link_setup_source(const struct media_pad *source_pad, sink->type != VSP1_ENTITY_HGT) { if (source->sink) return -EBUSY; - source->sink = sink_pad->entity; + source->sink = sink; source->sink_pad = sink_pad->index; } } else { @@ -355,15 +360,17 @@ static int vsp1_entity_link_setup_sink(const struct media_pad *source_pad, u32 flags) { struct vsp1_entity *sink; + struct vsp1_entity *source; sink = media_entity_to_vsp1_entity(sink_pad->entity); + source = media_entity_to_vsp1_entity(source_pad->entity); if (flags & MEDIA_LNK_FL_ENABLED) { /* Fan-in is limited to one. */ if (sink->sources[sink_pad->index]) return -EBUSY; - sink->sources[sink_pad->index] = source_pad->entity; + sink->sources[sink_pad->index] = source; } else { sink->sources[sink_pad->index] = NULL; } @@ -450,6 +457,8 @@ struct media_pad *vsp1_entity_remote_pad(struct media_pad *pad) { VI6_DPR_NODE_WPF(idx) }, VI6_DPR_NODE_WPF(idx) } static const struct vsp1_route vsp1_routes[] = { + { VSP1_ENTITY_BRS, 0, VI6_DPR_ILV_BRS_ROUTE, + { VI6_DPR_NODE_BRS_IN(0), VI6_DPR_NODE_BRS_IN(1) }, 0 }, { VSP1_ENTITY_BRU, 0, VI6_DPR_BRU_ROUTE, { VI6_DPR_NODE_BRU_IN(0), VI6_DPR_NODE_BRU_IN(1), VI6_DPR_NODE_BRU_IN(2), VI6_DPR_NODE_BRU_IN(3), @@ -459,7 +468,8 @@ static const struct vsp1_route vsp1_routes[] = { { VSP1_ENTITY_HGT, 0, 0, { 0, }, 0 }, VSP1_ENTITY_ROUTE(HSI), VSP1_ENTITY_ROUTE(HST), - { VSP1_ENTITY_LIF, 0, 0, { VI6_DPR_NODE_LIF, }, VI6_DPR_NODE_LIF }, + { VSP1_ENTITY_LIF, 0, 0, { 0, }, 0 }, + { VSP1_ENTITY_LIF, 1, 0, { 0, }, 0 }, VSP1_ENTITY_ROUTE(LUT), VSP1_ENTITY_ROUTE_RPF(0), VSP1_ENTITY_ROUTE_RPF(1), diff --git a/drivers/media/platform/vsp1/vsp1_entity.h b/drivers/media/platform/vsp1/vsp1_entity.h index c169a060b6d2..11f8363fa6b0 100644 --- a/drivers/media/platform/vsp1/vsp1_entity.h +++ b/drivers/media/platform/vsp1/vsp1_entity.h @@ -23,6 +23,7 @@ struct vsp1_dl_list; struct vsp1_pipeline; enum vsp1_entity_type { + VSP1_ENTITY_BRS, VSP1_ENTITY_BRU, VSP1_ENTITY_CLU, VSP1_ENTITY_HGO, @@ -104,8 +105,8 @@ struct vsp1_entity { struct media_pad *pads; unsigned int source_pad; - struct media_entity **sources; - struct media_entity *sink; + struct vsp1_entity **sources; + struct vsp1_entity *sink; unsigned int sink_pad; struct v4l2_subdev subdev; diff --git a/drivers/media/platform/vsp1/vsp1_lif.c b/drivers/media/platform/vsp1/vsp1_lif.c index 702487f895b3..e6fa16d7fda8 100644 --- a/drivers/media/platform/vsp1/vsp1_lif.c +++ b/drivers/media/platform/vsp1/vsp1_lif.c @@ -30,7 +30,7 @@ static inline void vsp1_lif_write(struct vsp1_lif *lif, struct vsp1_dl_list *dl, u32 reg, u32 data) { - vsp1_dl_list_write(dl, reg, data); + vsp1_dl_list_write(dl, reg + lif->entity.index * VI6_LIF_OFFSET, data); } /* ----------------------------------------------------------------------------- @@ -165,7 +165,7 @@ static const struct vsp1_entity_operations lif_entity_ops = { * Initialization and Cleanup */ -struct vsp1_lif *vsp1_lif_create(struct vsp1_device *vsp1) +struct vsp1_lif *vsp1_lif_create(struct vsp1_device *vsp1, unsigned int index) { struct vsp1_lif *lif; int ret; @@ -176,6 +176,7 @@ struct vsp1_lif *vsp1_lif_create(struct vsp1_device *vsp1) lif->entity.ops = &lif_entity_ops; lif->entity.type = VSP1_ENTITY_LIF; + lif->entity.index = index; /* * The LIF is never exposed to userspace, but media entity registration diff --git a/drivers/media/platform/vsp1/vsp1_lif.h b/drivers/media/platform/vsp1/vsp1_lif.h index 7b35879028de..3417339379b1 100644 --- a/drivers/media/platform/vsp1/vsp1_lif.h +++ b/drivers/media/platform/vsp1/vsp1_lif.h @@ -32,6 +32,6 @@ static inline struct vsp1_lif *to_lif(struct v4l2_subdev *subdev) return container_of(subdev, struct vsp1_lif, entity.subdev); } -struct vsp1_lif *vsp1_lif_create(struct vsp1_device *vsp1); +struct vsp1_lif *vsp1_lif_create(struct vsp1_device *vsp1, unsigned int index); #endif /* __VSP1_LIF_H__ */ diff --git a/drivers/media/platform/vsp1/vsp1_pipe.c b/drivers/media/platform/vsp1/vsp1_pipe.c index e817623b84e0..4f4b732df84b 100644 --- a/drivers/media/platform/vsp1/vsp1_pipe.c +++ b/drivers/media/platform/vsp1/vsp1_pipe.c @@ -335,16 +335,12 @@ void vsp1_pipeline_frame_end(struct vsp1_pipeline *pipe) if (pipe == NULL) return; + /* + * If the DL commit raced with the frame end interrupt, the commit ends + * up being postponed by one frame. @completed represents whether the + * active frame was finished or postponed. + */ completed = vsp1_dlm_irq_frame_end(pipe->output->dlm); - if (!completed) { - /* - * If the DL commit raced with the frame end interrupt, the - * commit ends up being postponed by one frame. Return - * immediately without calling the pipeline's frame end handler - * or incrementing the sequence number. - */ - return; - } if (pipe->hgo) vsp1_hgo_frame_end(pipe->hgo); @@ -352,8 +348,12 @@ void vsp1_pipeline_frame_end(struct vsp1_pipeline *pipe) if (pipe->hgt) vsp1_hgt_frame_end(pipe->hgt); + /* + * Regardless of frame completion we still need to notify the pipe + * frame_end to account for vblank events. + */ if (pipe->frame_end) - pipe->frame_end(pipe); + pipe->frame_end(pipe, completed); pipe->sequence++; } @@ -373,10 +373,11 @@ void vsp1_pipeline_propagate_alpha(struct vsp1_pipeline *pipe, return; /* - * The BRU background color has a fixed alpha value set to 255, the - * output alpha value is thus always equal to 255. + * The BRU and BRS background color has a fixed alpha value set to 255, + * the output alpha value is thus always equal to 255. */ - if (pipe->uds_input->type == VSP1_ENTITY_BRU) + if (pipe->uds_input->type == VSP1_ENTITY_BRU || + pipe->uds_input->type == VSP1_ENTITY_BRS) alpha = 255; vsp1_uds_set_alpha(pipe->uds, dl, alpha); diff --git a/drivers/media/platform/vsp1/vsp1_pipe.h b/drivers/media/platform/vsp1/vsp1_pipe.h index 91a784a13422..c5d01a365370 100644 --- a/drivers/media/platform/vsp1/vsp1_pipe.h +++ b/drivers/media/platform/vsp1/vsp1_pipe.h @@ -91,7 +91,7 @@ struct vsp1_pipeline { enum vsp1_pipeline_state state; wait_queue_head_t wq; - void (*frame_end)(struct vsp1_pipeline *pipe); + void (*frame_end)(struct vsp1_pipeline *pipe, bool completed); struct mutex lock; struct kref kref; diff --git a/drivers/media/platform/vsp1/vsp1_regs.h b/drivers/media/platform/vsp1/vsp1_regs.h index cd3e32af6e3b..58d0bea963a6 100644 --- a/drivers/media/platform/vsp1/vsp1_regs.h +++ b/drivers/media/platform/vsp1/vsp1_regs.h @@ -18,6 +18,7 @@ */ #define VI6_CMD(n) (0x0000 + (n) * 4) +#define VI6_CMD_UPDHDR (1 << 4) #define VI6_CMD_STRCMD (1 << 0) #define VI6_CLK_DCSWT 0x0018 @@ -238,6 +239,10 @@ #define VI6_WPF_SRCRPF_VIRACT_SUB (1 << 28) #define VI6_WPF_SRCRPF_VIRACT_MST (2 << 28) #define VI6_WPF_SRCRPF_VIRACT_MASK (3 << 28) +#define VI6_WPF_SRCRPF_VIRACT2_DIS (0 << 24) +#define VI6_WPF_SRCRPF_VIRACT2_SUB (1 << 24) +#define VI6_WPF_SRCRPF_VIRACT2_MST (2 << 24) +#define VI6_WPF_SRCRPF_VIRACT2_MASK (3 << 24) #define VI6_WPF_SRCRPF_RPF_ACT_DIS(n) (0 << ((n) * 2)) #define VI6_WPF_SRCRPF_RPF_ACT_SUB(n) (1 << ((n) * 2)) #define VI6_WPF_SRCRPF_RPF_ACT_MST(n) (2 << ((n) * 2)) @@ -321,6 +326,8 @@ #define VI6_DPR_HST_ROUTE 0x2044 #define VI6_DPR_HSI_ROUTE 0x2048 #define VI6_DPR_BRU_ROUTE 0x204c +#define VI6_DPR_ILV_BRS_ROUTE 0x2050 +#define VI6_DPR_ROUTE_BRSSEL (1 << 28) #define VI6_DPR_ROUTE_FXA_MASK (0xff << 16) #define VI6_DPR_ROUTE_FXA_SHIFT 16 #define VI6_DPR_ROUTE_FP_MASK (0x3f << 8) @@ -344,7 +351,8 @@ #define VI6_DPR_NODE_CLU 29 #define VI6_DPR_NODE_HST 30 #define VI6_DPR_NODE_HSI 31 -#define VI6_DPR_NODE_LIF 55 +#define VI6_DPR_NODE_BRS_IN(n) (38 + (n)) +#define VI6_DPR_NODE_LIF 55 /* Gen2 only */ #define VI6_DPR_NODE_WPF(n) (56 + (n)) #define VI6_DPR_NODE_UNUSED 63 @@ -476,7 +484,7 @@ #define VI6_HSI_CTRL_EN (1 << 0) /* ----------------------------------------------------------------------------- - * BRU Control Registers + * BRS and BRU Control Registers */ #define VI6_ROP_NOP 0 @@ -496,7 +504,10 @@ #define VI6_ROP_NAND 14 #define VI6_ROP_SET 15 -#define VI6_BRU_INCTRL 0x2c00 +#define VI6_BRU_BASE 0x2c00 +#define VI6_BRS_BASE 0x3900 + +#define VI6_BRU_INCTRL 0x0000 #define VI6_BRU_INCTRL_NRM (1 << 28) #define VI6_BRU_INCTRL_DnON (1 << (16 + (n))) #define VI6_BRU_INCTRL_DITHn_OFF (0 << ((n) * 4)) @@ -508,19 +519,19 @@ #define VI6_BRU_INCTRL_DITHn_MASK (7 << ((n) * 4)) #define VI6_BRU_INCTRL_DITHn_SHIFT ((n) * 4) -#define VI6_BRU_VIRRPF_SIZE 0x2c04 +#define VI6_BRU_VIRRPF_SIZE 0x0004 #define VI6_BRU_VIRRPF_SIZE_HSIZE_MASK (0x1fff << 16) #define VI6_BRU_VIRRPF_SIZE_HSIZE_SHIFT 16 #define VI6_BRU_VIRRPF_SIZE_VSIZE_MASK (0x1fff << 0) #define VI6_BRU_VIRRPF_SIZE_VSIZE_SHIFT 0 -#define VI6_BRU_VIRRPF_LOC 0x2c08 +#define VI6_BRU_VIRRPF_LOC 0x0008 #define VI6_BRU_VIRRPF_LOC_HCOORD_MASK (0x1fff << 16) #define VI6_BRU_VIRRPF_LOC_HCOORD_SHIFT 16 #define VI6_BRU_VIRRPF_LOC_VCOORD_MASK (0x1fff << 0) #define VI6_BRU_VIRRPF_LOC_VCOORD_SHIFT 0 -#define VI6_BRU_VIRRPF_COL 0x2c0c +#define VI6_BRU_VIRRPF_COL 0x000c #define VI6_BRU_VIRRPF_COL_A_MASK (0xff << 24) #define VI6_BRU_VIRRPF_COL_A_SHIFT 24 #define VI6_BRU_VIRRPF_COL_RCR_MASK (0xff << 16) @@ -530,7 +541,7 @@ #define VI6_BRU_VIRRPF_COL_BCB_MASK (0xff << 0) #define VI6_BRU_VIRRPF_COL_BCB_SHIFT 0 -#define VI6_BRU_CTRL(n) (0x2c10 + (n) * 8 + ((n) <= 3 ? 0 : 4)) +#define VI6_BRU_CTRL(n) (0x0010 + (n) * 8 + ((n) <= 3 ? 0 : 4)) #define VI6_BRU_CTRL_RBC (1 << 31) #define VI6_BRU_CTRL_DSTSEL_BRUIN(n) (((n) <= 3 ? (n) : (n)+1) << 20) #define VI6_BRU_CTRL_DSTSEL_VRPF (4 << 20) @@ -543,7 +554,7 @@ #define VI6_BRU_CTRL_AROP(rop) ((rop) << 0) #define VI6_BRU_CTRL_AROP_MASK (0xf << 0) -#define VI6_BRU_BLD(n) (0x2c14 + (n) * 8 + ((n) <= 3 ? 0 : 4)) +#define VI6_BRU_BLD(n) (0x0014 + (n) * 8 + ((n) <= 3 ? 0 : 4)) #define VI6_BRU_BLD_CBES (1 << 31) #define VI6_BRU_BLD_CCMDX_DST_A (0 << 28) #define VI6_BRU_BLD_CCMDX_255_DST_A (1 << 28) @@ -576,7 +587,7 @@ #define VI6_BRU_BLD_COEFY_MASK (0xff << 0) #define VI6_BRU_BLD_COEFY_SHIFT 0 -#define VI6_BRU_ROP 0x2c30 +#define VI6_BRU_ROP 0x0030 /* Only available on BRU */ #define VI6_BRU_ROP_DSTSEL_BRUIN(n) (((n) <= 3 ? (n) : (n)+1) << 20) #define VI6_BRU_ROP_DSTSEL_VRPF (4 << 20) #define VI6_BRU_ROP_DSTSEL_MASK (7 << 20) @@ -653,6 +664,8 @@ * LIF Control Registers */ +#define VI6_LIF_OFFSET (-0x100) + #define VI6_LIF_CTRL 0x3b00 #define VI6_LIF_CTRL_OBTH_MASK (0x7ff << 16) #define VI6_LIF_CTRL_OBTH_SHIFT 16 @@ -689,9 +702,20 @@ #define VI6_IP_VERSION_MODEL_VSPBD_GEN3 (0x15 << 8) #define VI6_IP_VERSION_MODEL_VSPBC_GEN3 (0x16 << 8) #define VI6_IP_VERSION_MODEL_VSPD_GEN3 (0x17 << 8) +#define VI6_IP_VERSION_MODEL_VSPD_V3 (0x18 << 8) +#define VI6_IP_VERSION_MODEL_VSPDL_GEN3 (0x19 << 8) +#define VI6_IP_VERSION_MODEL_VSPBS_GEN3 (0x1a << 8) #define VI6_IP_VERSION_SOC_MASK (0xff << 0) -#define VI6_IP_VERSION_SOC_H (0x01 << 0) -#define VI6_IP_VERSION_SOC_M (0x02 << 0) +#define VI6_IP_VERSION_SOC_H2 (0x01 << 0) +#define VI6_IP_VERSION_SOC_V2H (0x01 << 0) +#define VI6_IP_VERSION_SOC_V3M (0x01 << 0) +#define VI6_IP_VERSION_SOC_M2 (0x02 << 0) +#define VI6_IP_VERSION_SOC_M3W (0x02 << 0) +#define VI6_IP_VERSION_SOC_V3H (0x02 << 0) +#define VI6_IP_VERSION_SOC_H3 (0x03 << 0) +#define VI6_IP_VERSION_SOC_D3 (0x04 << 0) +#define VI6_IP_VERSION_SOC_M3N (0x04 << 0) +#define VI6_IP_VERSION_SOC_E3 (0x04 << 0) /* ----------------------------------------------------------------------------- * RPF CLUT Registers diff --git a/drivers/media/platform/vsp1/vsp1_video.c b/drivers/media/platform/vsp1/vsp1_video.c index 5af3486afe07..e9f5dcb8fae5 100644 --- a/drivers/media/platform/vsp1/vsp1_video.c +++ b/drivers/media/platform/vsp1/vsp1_video.c @@ -440,13 +440,17 @@ static void vsp1_video_pipeline_run(struct vsp1_pipeline *pipe) vsp1_pipeline_run(pipe); } -static void vsp1_video_pipeline_frame_end(struct vsp1_pipeline *pipe) +static void vsp1_video_pipeline_frame_end(struct vsp1_pipeline *pipe, + bool completed) { struct vsp1_device *vsp1 = pipe->output->entity.vsp1; enum vsp1_pipeline_state state; unsigned long flags; unsigned int i; + /* M2M Pipelines should never call here with an incomplete frame. */ + WARN_ON_ONCE(!completed); + spin_lock_irqsave(&pipe->irqlock, flags); /* Complete buffers on all video nodes. */ @@ -481,7 +485,7 @@ static int vsp1_video_pipeline_build_branch(struct vsp1_pipeline *pipe, struct media_entity_enum ent_enum; struct vsp1_entity *entity; struct media_pad *pad; - bool bru_found = false; + struct vsp1_bru *bru = NULL; int ret; ret = media_entity_enum_init(&ent_enum, &input->entity.vsp1->media_dev); @@ -511,16 +515,20 @@ static int vsp1_video_pipeline_build_branch(struct vsp1_pipeline *pipe, media_entity_to_v4l2_subdev(pad->entity)); /* - * A BRU is present in the pipeline, store the BRU input pad + * A BRU or BRS is present in the pipeline, store its input pad * number in the input RPF for use when configuring the RPF. */ - if (entity->type == VSP1_ENTITY_BRU) { - struct vsp1_bru *bru = to_bru(&entity->subdev); + if (entity->type == VSP1_ENTITY_BRU || + entity->type == VSP1_ENTITY_BRS) { + /* BRU and BRS can't be chained. */ + if (bru) { + ret = -EPIPE; + goto out; + } + bru = to_bru(&entity->subdev); bru->inputs[pad->index].rpf = input; input->bru_input = pad->index; - - bru_found = true; } /* We've reached the WPF, we're done. */ @@ -542,8 +550,7 @@ static int vsp1_video_pipeline_build_branch(struct vsp1_pipeline *pipe, } pipe->uds = entity; - pipe->uds_input = bru_found ? pipe->bru - : &input->entity; + pipe->uds_input = bru ? &bru->entity : &input->entity; } /* Follow the source link, ignoring any HGO or HGT. */ @@ -589,30 +596,42 @@ static int vsp1_video_pipeline_build(struct vsp1_pipeline *pipe, e = to_vsp1_entity(subdev); list_add_tail(&e->list_pipe, &pipe->entities); - if (e->type == VSP1_ENTITY_RPF) { + switch (e->type) { + case VSP1_ENTITY_RPF: rwpf = to_rwpf(subdev); pipe->inputs[rwpf->entity.index] = rwpf; rwpf->video->pipe_index = ++pipe->num_inputs; rwpf->pipe = pipe; - } else if (e->type == VSP1_ENTITY_WPF) { + break; + + case VSP1_ENTITY_WPF: rwpf = to_rwpf(subdev); pipe->output = rwpf; rwpf->video->pipe_index = 0; rwpf->pipe = pipe; - } else if (e->type == VSP1_ENTITY_LIF) { + break; + + case VSP1_ENTITY_LIF: pipe->lif = e; - } else if (e->type == VSP1_ENTITY_BRU) { + break; + + case VSP1_ENTITY_BRU: + case VSP1_ENTITY_BRS: pipe->bru = e; - } else if (e->type == VSP1_ENTITY_HGO) { - struct vsp1_hgo *hgo = to_hgo(subdev); + break; + case VSP1_ENTITY_HGO: pipe->hgo = e; - hgo->histo.pipe = pipe; - } else if (e->type == VSP1_ENTITY_HGT) { - struct vsp1_hgt *hgt = to_hgt(subdev); + to_hgo(subdev)->histo.pipe = pipe; + break; + case VSP1_ENTITY_HGT: pipe->hgt = e; - hgt->histo.pipe = pipe; + to_hgt(subdev)->histo.pipe = pipe; + break; + + default: + break; } } @@ -796,12 +815,14 @@ static int vsp1_video_setup_pipeline(struct vsp1_pipeline *pipe) struct vsp1_uds *uds = to_uds(&pipe->uds->subdev); /* - * If a BRU is present in the pipeline before the UDS, the alpha - * component doesn't need to be scaled as the BRU output alpha - * value is fixed to 255. Otherwise we need to scale the alpha - * component only when available at the input RPF. + * If a BRU or BRS is present in the pipeline before the UDS, + * the alpha component doesn't need to be scaled as the BRU and + * BRS output alpha value is fixed to 255. Otherwise we need to + * scale the alpha component only when available at the input + * RPF. */ - if (pipe->uds_input->type == VSP1_ENTITY_BRU) { + if (pipe->uds_input->type == VSP1_ENTITY_BRU || + pipe->uds_input->type == VSP1_ENTITY_BRS) { uds->scale_alpha = false; } else { struct vsp1_rwpf *rpf = diff --git a/drivers/media/platform/vsp1/vsp1_wpf.c b/drivers/media/platform/vsp1/vsp1_wpf.c index 32df109b119f..b6c902be225b 100644 --- a/drivers/media/platform/vsp1/vsp1_wpf.c +++ b/drivers/media/platform/vsp1/vsp1_wpf.c @@ -453,7 +453,9 @@ static void wpf_configure(struct vsp1_entity *entity, } if (pipe->bru || pipe->num_inputs > 1) - srcrpf |= VI6_WPF_SRCRPF_VIRACT_MST; + srcrpf |= pipe->bru->type == VSP1_ENTITY_BRU + ? VI6_WPF_SRCRPF_VIRACT_MST + : VI6_WPF_SRCRPF_VIRACT2_MST; vsp1_wpf_write(wpf, dl, VI6_WPF_SRCRPF, srcrpf); diff --git a/include/drm/ttm/ttm_bo_driver.h b/include/drm/ttm/ttm_bo_driver.h index 990d529f823c..d30850e07936 100644 --- a/include/drm/ttm/ttm_bo_driver.h +++ b/include/drm/ttm/ttm_bo_driver.h @@ -472,6 +472,23 @@ struct ttm_bo_driver { */ unsigned long (*io_mem_pfn)(struct ttm_buffer_object *bo, unsigned long page_offset); + + /** + * Read/write memory buffers for ptrace access + * + * @bo: the BO to access + * @offset: the offset from the start of the BO + * @buf: pointer to source/destination buffer + * @len: number of bytes to copy + * @write: whether to read (0) from or write (non-0) to BO + * + * If successful, this function should return the number of + * bytes copied, -EIO otherwise. If the number of bytes + * returned is < len, the function may be called again with + * the remainder of the buffer to copy. + */ + int (*access_memory)(struct ttm_buffer_object *bo, unsigned long offset, + void *buf, int len, int write); }; /** diff --git a/include/media/vsp1.h b/include/media/vsp1.h index c837383b2013..68a8abe4fac5 100644 --- a/include/media/vsp1.h +++ b/include/media/vsp1.h @@ -34,11 +34,12 @@ struct vsp1_du_lif_config { unsigned int width; unsigned int height; - void (*callback)(void *); + void (*callback)(void *, bool); void *callback_data; }; -int vsp1_du_setup_lif(struct device *dev, const struct vsp1_du_lif_config *cfg); +int vsp1_du_setup_lif(struct device *dev, unsigned int pipe_index, + const struct vsp1_du_lif_config *cfg); struct vsp1_du_atomic_config { u32 pixelformat; @@ -50,10 +51,11 @@ struct vsp1_du_atomic_config { unsigned int zpos; }; -void vsp1_du_atomic_begin(struct device *dev); -int vsp1_du_atomic_update(struct device *dev, unsigned int rpf, +void vsp1_du_atomic_begin(struct device *dev, unsigned int pipe_index); +int vsp1_du_atomic_update(struct device *dev, unsigned int pipe_index, + unsigned int rpf, const struct vsp1_du_atomic_config *cfg); -void vsp1_du_atomic_flush(struct device *dev); +void vsp1_du_atomic_flush(struct device *dev, unsigned int pipe_index); int vsp1_du_map_sg(struct device *dev, struct sg_table *sgt); void vsp1_du_unmap_sg(struct device *dev, struct sg_table *sgt); |