diff options
Diffstat (limited to 'drivers/gpu')
243 files changed, 6664 insertions, 4182 deletions
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h index 262056778f52..6a8129949333 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h @@ -32,7 +32,7 @@ #include <linux/wait.h> #include <linux/list.h> #include <linux/kref.h> -#include <linux/interval_tree.h> +#include <linux/rbtree.h> #include <linux/hashtable.h> #include <linux/dma-fence.h> @@ -122,14 +122,6 @@ extern int amdgpu_param_buf_per_se; /* max number of IP instances */ #define AMDGPU_MAX_SDMA_INSTANCES 2 -/* max number of VMHUB */ -#define AMDGPU_MAX_VMHUBS 2 -#define AMDGPU_MMHUB 0 -#define AMDGPU_GFXHUB 1 - -/* hardcode that limit for now */ -#define AMDGPU_VA_RESERVED_SIZE (8 << 20) - /* hard reset data */ #define AMDGPU_ASIC_RESET_DATA 0x39d5e86b @@ -312,12 +304,9 @@ struct amdgpu_gart_funcs { /* set pte flags based per asic */ uint64_t (*get_vm_pte_flags)(struct amdgpu_device *adev, uint32_t flags); -}; - -/* provided by the mc block */ -struct amdgpu_mc_funcs { /* adjust mc addr in fb for APU case */ u64 (*adjust_mc_addr)(struct amdgpu_device *adev, u64 addr); + uint32_t (*get_invalidate_req)(unsigned int vm_id); }; /* provided by the ih block */ @@ -379,7 +368,10 @@ struct amdgpu_bo_list_entry { struct amdgpu_bo_va_mapping { struct list_head list; - struct interval_tree_node it; + struct rb_node rb; + uint64_t start; + uint64_t last; + uint64_t __subtree_last; uint64_t offset; uint64_t flags; }; @@ -579,8 +571,6 @@ struct amdgpu_vmhub { uint32_t vm_context0_cntl; uint32_t vm_l2_pro_fault_status; uint32_t vm_l2_pro_fault_cntl; - uint32_t (*get_invalidate_req)(unsigned int vm_id); - uint32_t (*get_vm_protection_bits)(void); }; /* @@ -618,7 +608,6 @@ struct amdgpu_mc { u64 private_aperture_end; /* protects concurrent invalidation */ spinlock_t invalidate_lock; - const struct amdgpu_mc_funcs *mc_funcs; }; /* @@ -1712,6 +1701,12 @@ void amdgpu_mm_wdoorbell64(struct amdgpu_device *adev, u32 index, u64 v); #define WREG32_FIELD(reg, field, val) \ WREG32(mm##reg, (RREG32(mm##reg) & ~REG_FIELD_MASK(reg, field)) | (val) << REG_FIELD_SHIFT(reg, field)) +#define WREG32_FIELD_OFFSET(reg, offset, field, val) \ + WREG32(mm##reg + offset, (RREG32(mm##reg + offset) & ~REG_FIELD_MASK(reg, field)) | (val) << REG_FIELD_SHIFT(reg, field)) + +#define WREG32_FIELD15(ip, idx, reg, field, val) \ + WREG32(SOC15_REG_OFFSET(ip, idx, mm##reg), (RREG32(SOC15_REG_OFFSET(ip, idx, mm##reg)) & ~REG_FIELD_MASK(reg, field)) | (val) << REG_FIELD_SHIFT(reg, field)) + /* * BIOS helpers. */ @@ -1887,12 +1882,14 @@ void amdgpu_unregister_atpx_handler(void); bool amdgpu_has_atpx_dgpu_power_cntl(void); bool amdgpu_is_atpx_hybrid(void); bool amdgpu_atpx_dgpu_req_power_for_displays(void); +bool amdgpu_has_atpx(void); #else static inline void amdgpu_register_atpx_handler(void) {} static inline void amdgpu_unregister_atpx_handler(void) {} static inline bool amdgpu_has_atpx_dgpu_power_cntl(void) { return false; } static inline bool amdgpu_is_atpx_hybrid(void) { return false; } static inline bool amdgpu_atpx_dgpu_req_power_for_displays(void) { return false; } +static inline bool amdgpu_has_atpx(void) { return false; } #endif /* diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_atombios.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_atombios.c index f52b1bf3d3d9..ad4329922f79 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_atombios.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_atombios.c @@ -754,6 +754,35 @@ union igp_info { struct _ATOM_INTEGRATED_SYSTEM_INFO_V1_9 info_9; }; +/* + * Return vram width from integrated system info table, if available, + * or 0 if not. + */ +int amdgpu_atombios_get_vram_width(struct amdgpu_device *adev) +{ + struct amdgpu_mode_info *mode_info = &adev->mode_info; + int index = GetIndexIntoMasterTable(DATA, 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 8: + case 9: + return igp_info->info_8.ucUMAChannelNumber * 64; + default: + return 0; + } + } + + return 0; +} + static void amdgpu_atombios_get_igp_ss_overrides(struct amdgpu_device *adev, struct amdgpu_atom_ss *ss, int id) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_atombios.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_atombios.h index 4e0f488487f3..38d0fe32e5cd 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_atombios.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_atombios.h @@ -148,6 +148,8 @@ int amdgpu_atombios_get_clock_info(struct amdgpu_device *adev); int amdgpu_atombios_get_gfx_info(struct amdgpu_device *adev); +int amdgpu_atombios_get_vram_width(struct amdgpu_device *adev); + bool amdgpu_atombios_get_asic_ss_info(struct amdgpu_device *adev, struct amdgpu_atom_ss *ss, int id, u32 clock); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_bo_list.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_bo_list.c index 0218cea6be4d..a6649874e6ce 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_bo_list.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_bo_list.c @@ -237,7 +237,7 @@ int amdgpu_bo_list_ioctl(struct drm_device *dev, void *data, struct amdgpu_fpriv *fpriv = filp->driver_priv; union drm_amdgpu_bo_list *args = data; uint32_t handle = args->in.list_handle; - const void __user *uptr = (const void*)(long)args->in.bo_info_ptr; + const void __user *uptr = (const void*)(uintptr_t)args->in.bo_info_ptr; struct drm_amdgpu_bo_list_entry *info; struct amdgpu_bo_list *list; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c index 97f661372a1c..ec71b9320561 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c @@ -161,7 +161,7 @@ int amdgpu_cs_parser_init(struct amdgpu_cs_parser *p, void *data) } /* get chunks */ - chunk_array_user = (uint64_t __user *)(unsigned long)(cs->in.chunks); + chunk_array_user = (uint64_t __user *)(uintptr_t)(cs->in.chunks); if (copy_from_user(chunk_array, chunk_array_user, sizeof(uint64_t)*cs->in.num_chunks)) { ret = -EFAULT; @@ -181,7 +181,7 @@ int amdgpu_cs_parser_init(struct amdgpu_cs_parser *p, void *data) struct drm_amdgpu_cs_chunk user_chunk; uint32_t __user *cdata; - chunk_ptr = (void __user *)(unsigned long)chunk_array[i]; + chunk_ptr = (void __user *)(uintptr_t)chunk_array[i]; if (copy_from_user(&user_chunk, chunk_ptr, sizeof(struct drm_amdgpu_cs_chunk))) { ret = -EFAULT; @@ -192,7 +192,7 @@ int amdgpu_cs_parser_init(struct amdgpu_cs_parser *p, void *data) p->chunks[i].length_dw = user_chunk.length_dw; size = p->chunks[i].length_dw; - cdata = (void __user *)(unsigned long)user_chunk.chunk_data; + cdata = (void __user *)(uintptr_t)user_chunk.chunk_data; p->chunks[i].kdata = drm_malloc_ab(size, sizeof(uint32_t)); if (p->chunks[i].kdata == NULL) { @@ -949,7 +949,7 @@ static int amdgpu_cs_ib_fill(struct amdgpu_device *adev, } if ((chunk_ib->va_start + chunk_ib->ib_bytes) > - (m->it.last + 1) * AMDGPU_GPU_PAGE_SIZE) { + (m->last + 1) * AMDGPU_GPU_PAGE_SIZE) { DRM_ERROR("IB va_start+ib_bytes is invalid\n"); return -EINVAL; } @@ -960,7 +960,7 @@ static int amdgpu_cs_ib_fill(struct amdgpu_device *adev, return r; } - offset = ((uint64_t)m->it.start) * AMDGPU_GPU_PAGE_SIZE; + offset = m->start * AMDGPU_GPU_PAGE_SIZE; kptr += chunk_ib->va_start - offset; r = amdgpu_ib_get(adev, vm, chunk_ib->ib_bytes, ib); @@ -1242,6 +1242,7 @@ static int amdgpu_cs_wait_all_fences(struct amdgpu_device *adev, continue; r = dma_fence_wait_timeout(fence, true, timeout); + dma_fence_put(fence); if (r < 0) return r; @@ -1339,7 +1340,7 @@ int amdgpu_cs_wait_fences_ioctl(struct drm_device *dev, void *data, if (fences == NULL) return -ENOMEM; - fences_user = (void __user *)(unsigned long)(wait->in.fences); + fences_user = (void __user *)(uintptr_t)(wait->in.fences); if (copy_from_user(fences, fences_user, sizeof(struct drm_amdgpu_fence) * fence_count)) { r = -EFAULT; @@ -1388,8 +1389,8 @@ amdgpu_cs_find_mapping(struct amdgpu_cs_parser *parser, continue; list_for_each_entry(mapping, &lobj->bo_va->valids, list) { - if (mapping->it.start > addr || - addr > mapping->it.last) + if (mapping->start > addr || + addr > mapping->last) continue; *bo = lobj->bo_va->bo; @@ -1397,8 +1398,8 @@ amdgpu_cs_find_mapping(struct amdgpu_cs_parser *parser, } list_for_each_entry(mapping, &lobj->bo_va->invalids, list) { - if (mapping->it.start > addr || - addr > mapping->it.last) + if (mapping->start > addr || + addr > mapping->last) continue; *bo = lobj->bo_va->bo; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c index 83dda05325b8..483660742f75 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c @@ -1040,43 +1040,60 @@ static bool amdgpu_check_pot_argument(int arg) return (arg & (arg - 1)) == 0; } -static void amdgpu_get_block_size(struct amdgpu_device *adev) -{ - /* from AI, asic starts to support multiple level VMPT */ - if (adev->asic_type >= CHIP_VEGA10) { - if (amdgpu_vm_block_size != 9) - dev_warn(adev->dev, - "Multi-VMPT limits block size to one page!\n"); - amdgpu_vm_block_size = 9; - return; - } +static void amdgpu_check_block_size(struct amdgpu_device *adev) +{ /* defines number of bits in page table versus page directory, * a page is 4KB so we have 12 bits offset, minimum 9 bits in the * page table and the remaining bits are in the page directory */ - if (amdgpu_vm_block_size == -1) { - - /* Total bits covered by PD + PTs */ - unsigned bits = ilog2(amdgpu_vm_size) + 18; - - /* Make sure the PD is 4K in size up to 8GB address space. - Above that split equal between PD and PTs */ - if (amdgpu_vm_size <= 8) - amdgpu_vm_block_size = bits - 9; - else - amdgpu_vm_block_size = (bits + 3) / 2; + if (amdgpu_vm_block_size == -1) + return; - } else if (amdgpu_vm_block_size < 9) { + if (amdgpu_vm_block_size < 9) { dev_warn(adev->dev, "VM page table size (%d) too small\n", amdgpu_vm_block_size); - amdgpu_vm_block_size = 9; + goto def_value; } if (amdgpu_vm_block_size > 24 || (amdgpu_vm_size * 1024) < (1ull << amdgpu_vm_block_size)) { dev_warn(adev->dev, "VM page table size (%d) too large\n", amdgpu_vm_block_size); - amdgpu_vm_block_size = 9; + goto def_value; } + + return; + +def_value: + amdgpu_vm_block_size = -1; +} + +static void amdgpu_check_vm_size(struct amdgpu_device *adev) +{ + if (!amdgpu_check_pot_argument(amdgpu_vm_size)) { + dev_warn(adev->dev, "VM size (%d) must be a power of 2\n", + amdgpu_vm_size); + goto def_value; + } + + if (amdgpu_vm_size < 1) { + dev_warn(adev->dev, "VM size (%d) too small, min is 1GB\n", + amdgpu_vm_size); + goto def_value; + } + + /* + * Max GPUVM size for Cayman, SI, CI VI are 40 bits. + */ + if (amdgpu_vm_size > 1024) { + dev_warn(adev->dev, "VM size (%d) too large, max is 1TB\n", + amdgpu_vm_size); + goto def_value; + } + + return; + +def_value: + amdgpu_vm_size = -1; } /** @@ -1108,28 +1125,9 @@ static void amdgpu_check_arguments(struct amdgpu_device *adev) } } - if (!amdgpu_check_pot_argument(amdgpu_vm_size)) { - dev_warn(adev->dev, "VM size (%d) must be a power of 2\n", - amdgpu_vm_size); - amdgpu_vm_size = 8; - } + amdgpu_check_vm_size(adev); - if (amdgpu_vm_size < 1) { - dev_warn(adev->dev, "VM size (%d) too small, min is 1GB\n", - amdgpu_vm_size); - amdgpu_vm_size = 8; - } - - /* - * Max GPUVM size for Cayman, SI and CI are 40 bits. - */ - if (amdgpu_vm_size > 1024) { - dev_warn(adev->dev, "VM size (%d) too large, max is 1TB\n", - amdgpu_vm_size); - amdgpu_vm_size = 8; - } - - amdgpu_get_block_size(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))) { @@ -2249,9 +2247,10 @@ int amdgpu_device_resume(struct drm_device *dev, bool resume, bool fbcon) } r = amdgpu_resume(adev); - if (r) + if (r) { DRM_ERROR("amdgpu_resume failed (%d).\n", r); - + return r; + } amdgpu_fence_driver_resume(adev); if (resume) { diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c index ce15721cadda..96926a221bd5 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c @@ -614,6 +614,12 @@ amdgpu_user_framebuffer_create(struct drm_device *dev, return ERR_PTR(-ENOENT); } + /* Handle is imported dma-buf, so cannot be migrated to VRAM for scanout */ + if (obj->import_attach) { + DRM_DEBUG_KMS("Cannot create framebuffer from imported dma_buf\n"); + return ERR_PTR(-EINVAL); + } + amdgpu_fb = kzalloc(sizeof(*amdgpu_fb), GFP_KERNEL); if (amdgpu_fb == NULL) { drm_gem_object_unreference_unlocked(obj); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c index 400917fd7486..4e0f7d2d87f1 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c @@ -86,7 +86,7 @@ int amdgpu_runtime_pm = -1; unsigned amdgpu_ip_block_mask = 0xffffffff; int amdgpu_bapm = -1; int amdgpu_deep_color = 0; -int amdgpu_vm_size = 64; +int amdgpu_vm_size = -1; int amdgpu_vm_block_size = -1; int amdgpu_vm_fault_stop = 0; int amdgpu_vm_debug = 0; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c index f85520d4e711..03a9c5cad222 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c @@ -717,7 +717,7 @@ int amdgpu_gem_op_ioctl(struct drm_device *dev, void *data, switch (args->op) { case AMDGPU_GEM_OP_GET_GEM_CREATE_INFO: { struct drm_amdgpu_gem_create_in info; - void __user *out = (void __user *)(long)args->value; + void __user *out = (void __user *)(uintptr_t)args->value; info.bo_size = robj->gem_base.size; info.alignment = robj->tbo.mem.page_alignment << PAGE_SHIFT; @@ -729,6 +729,11 @@ int amdgpu_gem_op_ioctl(struct drm_device *dev, void *data, break; } case AMDGPU_GEM_OP_SET_PLACEMENT: + if (robj->prime_shared_count && (args->value & AMDGPU_GEM_DOMAIN_VRAM)) { + r = -EINVAL; + amdgpu_bo_unreserve(robj); + break; + } if (amdgpu_ttm_tt_get_usermm(robj->tbo.ttm)) { r = -EPERM; amdgpu_bo_unreserve(robj); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_irq.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_irq.c index 13b487235a8b..a6b7e367a860 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_irq.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_irq.c @@ -316,9 +316,10 @@ int amdgpu_irq_add_id(struct amdgpu_device *adev, return -EINVAL; if (!adev->irq.client[client_id].sources) { - adev->irq.client[client_id].sources = kcalloc(AMDGPU_MAX_IRQ_SRC_ID, - sizeof(struct amdgpu_irq_src), - GFP_KERNEL); + adev->irq.client[client_id].sources = + kcalloc(AMDGPU_MAX_IRQ_SRC_ID, + sizeof(struct amdgpu_irq_src *), + GFP_KERNEL); if (!adev->irq.client[client_id].sources) return -ENOMEM; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c index dfb029ab3448..832be632478f 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c @@ -36,12 +36,6 @@ #include <linux/pm_runtime.h> #include "amdgpu_amdkfd.h" -#if defined(CONFIG_VGA_SWITCHEROO) -bool amdgpu_has_atpx(void); -#else -static inline bool amdgpu_has_atpx(void) { return false; } -#endif - /** * amdgpu_driver_unload_kms - Main unload function for KMS. * @@ -243,7 +237,7 @@ static int amdgpu_info_ioctl(struct drm_device *dev, void *data, struct drm_file struct amdgpu_device *adev = dev->dev_private; struct drm_amdgpu_info *info = data; struct amdgpu_mode_info *minfo = &adev->mode_info; - void __user *out = (void __user *)(long)info->return_pointer; + void __user *out = (void __user *)(uintptr_t)info->return_pointer; uint32_t size = info->return_size; struct drm_crtc *crtc; uint32_t ui32 = 0; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c index 7ea3cacf9f9f..38f739fb727b 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c @@ -31,6 +31,7 @@ #include <linux/firmware.h> #include <linux/module.h> #include <linux/mmu_notifier.h> +#include <linux/interval_tree.h> #include <drm/drmP.h> #include <drm/drm.h> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c index 5aac350b007f..cb89fff863c0 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c @@ -122,20 +122,19 @@ static void amdgpu_ttm_placement_init(struct amdgpu_device *adev, if (domain & AMDGPU_GEM_DOMAIN_VRAM) { unsigned visible_pfn = adev->mc.visible_vram_size >> PAGE_SHIFT; - unsigned lpfn = 0; - - /* This forces a reallocation if the flag wasn't set before */ - if (flags & AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS) - lpfn = adev->mc.real_vram_size >> PAGE_SHIFT; places[c].fpfn = 0; - places[c].lpfn = lpfn; + places[c].lpfn = 0; places[c].flags = TTM_PL_FLAG_WC | TTM_PL_FLAG_UNCACHED | TTM_PL_FLAG_VRAM; + if (flags & AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED) places[c].lpfn = visible_pfn; else places[c].flags |= TTM_PL_FLAG_TOPDOWN; + + if (flags & AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS) + places[c].flags |= TTM_PL_FLAG_CONTIGUOUS; c++; } @@ -651,6 +650,10 @@ int amdgpu_bo_pin_restricted(struct amdgpu_bo *bo, u32 domain, if (WARN_ON_ONCE(min_offset > max_offset)) return -EINVAL; + /* A shared bo cannot be migrated to VRAM */ + if (bo->prime_shared_count && (domain == AMDGPU_GEM_DOMAIN_VRAM)) + return -EINVAL; + if (bo->pin_count) { uint32_t mem_type = bo->tbo.mem.mem_type; @@ -928,8 +931,7 @@ int amdgpu_bo_fault_reserve_notify(struct ttm_buffer_object *bo) 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 && - (abo->flags & AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS)) + if ((offset + size) <= adev->mc.visible_vram_size) return 0; /* Can't move a pinned BO to visible VRAM */ @@ -937,7 +939,6 @@ int amdgpu_bo_fault_reserve_notify(struct ttm_buffer_object *bo) return -EINVAL; /* hurrah the memory is not visible ! */ - abo->flags |= AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS; 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++) { diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c index 4731015f6101..ed6e5799016e 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c @@ -130,7 +130,7 @@ psp_cmd_submit_buf(struct psp_context *psp, while (*((unsigned int *)psp->fence_buf) != index) { msleep(1); - }; + } amdgpu_bo_free_kernel(&cmd_buf_bo, &cmd_buf_mc_addr, diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h index a87de18160a8..ee9d0f346d75 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h @@ -226,8 +226,8 @@ TRACE_EVENT(amdgpu_vm_bo_map, TP_fast_assign( __entry->bo = bo_va ? bo_va->bo : NULL; - __entry->start = mapping->it.start; - __entry->last = mapping->it.last; + __entry->start = mapping->start; + __entry->last = mapping->last; __entry->offset = mapping->offset; __entry->flags = mapping->flags; ), @@ -250,8 +250,8 @@ TRACE_EVENT(amdgpu_vm_bo_unmap, TP_fast_assign( __entry->bo = bo_va->bo; - __entry->start = mapping->it.start; - __entry->last = mapping->it.last; + __entry->start = mapping->start; + __entry->last = mapping->last; __entry->offset = mapping->offset; __entry->flags = mapping->flags; ), @@ -270,8 +270,8 @@ DECLARE_EVENT_CLASS(amdgpu_vm_mapping, ), TP_fast_assign( - __entry->soffset = mapping->it.start; - __entry->eoffset = mapping->it.last + 1; + __entry->soffset = mapping->start; + __entry->eoffset = mapping->last + 1; __entry->flags = mapping->flags; ), TP_printk("soffs=%010llx, eoffs=%010llx, flags=%08x", diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c index 244bb9aacf86..35d53a0d9ba6 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c @@ -529,40 +529,12 @@ static int amdgpu_ttm_io_mem_reserve(struct ttm_bo_device *bdev, struct ttm_mem_ case TTM_PL_TT: break; case TTM_PL_VRAM: - if (mem->start == AMDGPU_BO_INVALID_OFFSET) - return -EINVAL; - mem->bus.offset = mem->start << PAGE_SHIFT; /* check if it's visible */ if ((mem->bus.offset + mem->bus.size) > adev->mc.visible_vram_size) return -EINVAL; mem->bus.base = adev->mc.aper_base; mem->bus.is_iomem = true; -#ifdef __alpha__ - /* - * Alpha: use bus.addr to hold the ioremap() return, - * so we can modify bus.base below. - */ - if (mem->placement & TTM_PL_FLAG_WC) - mem->bus.addr = - ioremap_wc(mem->bus.base + mem->bus.offset, - mem->bus.size); - else - mem->bus.addr = - ioremap_nocache(mem->bus.base + mem->bus.offset, - mem->bus.size); - if (!mem->bus.addr) - return -ENOMEM; - - /* - * Alpha: Use just the bus offset plus - * the hose/domain memory base for bus.base. - * It then can be used to build PTEs for VRAM - * access, as done in ttm_bo_vm_fault(). - */ - mem->bus.base = (mem->bus.base & 0x0ffffffffUL) + - adev->ddev->hose->dense_mem_base; -#endif break; default: return -EINVAL; @@ -574,6 +546,18 @@ static void amdgpu_ttm_io_mem_free(struct ttm_bo_device *bdev, struct ttm_mem_re { } +static unsigned long amdgpu_ttm_io_mem_pfn(struct ttm_buffer_object *bo, + unsigned long page_offset) +{ + struct drm_mm_node *mm = bo->mem.mm_node; + uint64_t size = mm->size; + uint64_t offset = page_offset; + + page_offset = do_div(offset, size); + mm += offset; + return (bo->mem.bus.base >> PAGE_SHIFT) + mm->start + page_offset; +} + /* * TTM backend functions. */ @@ -1089,6 +1073,7 @@ static struct ttm_bo_driver amdgpu_bo_driver = { .fault_reserve_notify = &amdgpu_bo_fault_reserve_notify, .io_mem_reserve = &amdgpu_ttm_io_mem_reserve, .io_mem_free = &amdgpu_ttm_io_mem_free, + .io_mem_pfn = amdgpu_ttm_io_mem_pfn, }; int amdgpu_ttm_init(struct amdgpu_device *adev) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c index 0b92dd0c1d70..2ca09f111f08 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c @@ -741,10 +741,10 @@ static int amdgpu_uvd_cs_pass2(struct amdgpu_uvd_cs_ctx *ctx) start = amdgpu_bo_gpu_offset(bo); - end = (mapping->it.last + 1 - mapping->it.start); + end = (mapping->last + 1 - mapping->start); end = end * AMDGPU_GPU_PAGE_SIZE + start; - addr -= ((uint64_t)mapping->it.start) * AMDGPU_GPU_PAGE_SIZE; + addr -= mapping->start * AMDGPU_GPU_PAGE_SIZE; start += addr; amdgpu_set_ib_value(ctx->parser, ctx->ib_idx, ctx->data0, diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c index 0184197eb000..c853400805d1 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c @@ -595,13 +595,13 @@ static int amdgpu_vce_cs_reloc(struct amdgpu_cs_parser *p, uint32_t ib_idx, } if ((addr + (uint64_t)size) > - ((uint64_t)mapping->it.last + 1) * AMDGPU_GPU_PAGE_SIZE) { + (mapping->last + 1) * AMDGPU_GPU_PAGE_SIZE) { DRM_ERROR("BO to small for addr 0x%010Lx %d %d\n", addr, lo, hi); return -EINVAL; } - addr -= ((uint64_t)mapping->it.start) * AMDGPU_GPU_PAGE_SIZE; + addr -= mapping->start * AMDGPU_GPU_PAGE_SIZE; addr += amdgpu_bo_gpu_offset(bo); addr -= ((uint64_t)size) * ((uint64_t)index); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c index ecef35a1fe33..ba8b8ae6234f 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c @@ -122,9 +122,7 @@ uint32_t amdgpu_virt_kiq_rreg(struct amdgpu_device *adev, uint32_t reg) mutex_lock(&adev->virt.lock_kiq); amdgpu_ring_alloc(ring, 32); - amdgpu_ring_emit_hdp_flush(ring); amdgpu_ring_emit_rreg(ring, reg); - amdgpu_ring_emit_hdp_invalidate(ring); amdgpu_fence_emit(ring, &f); amdgpu_ring_commit(ring); mutex_unlock(&adev->virt.lock_kiq); @@ -150,9 +148,7 @@ void amdgpu_virt_kiq_wreg(struct amdgpu_device *adev, uint32_t reg, uint32_t v) mutex_lock(&adev->virt.lock_kiq); amdgpu_ring_alloc(ring, 32); - amdgpu_ring_emit_hdp_flush(ring); amdgpu_ring_emit_wreg(ring, reg, v); - amdgpu_ring_emit_hdp_invalidate(ring); amdgpu_fence_emit(ring, &f); amdgpu_ring_commit(ring); mutex_unlock(&adev->virt.lock_kiq); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c index 0235d7933efd..7ed5302b511a 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c @@ -26,6 +26,7 @@ * Jerome Glisse */ #include <linux/dma-fence-array.h> +#include <linux/interval_tree_generic.h> #include <drm/drmP.h> #include <drm/amdgpu_drm.h> #include "amdgpu.h" @@ -51,6 +52,15 @@ * SI supports 16. */ +#define START(node) ((node)->start) +#define LAST(node) ((node)->last) + +INTERVAL_TREE_DEFINE(struct amdgpu_bo_va_mapping, rb, uint64_t, __subtree_last, + START, LAST, static, amdgpu_vm_it) + +#undef START +#undef LAST + /* Local structure. Encapsulate some VM table update parameters to reduce * the number of function parameters */ @@ -90,13 +100,14 @@ static unsigned amdgpu_vm_num_entries(struct amdgpu_device *adev, if (level == 0) /* For the root directory */ return adev->vm_manager.max_pfn >> - (amdgpu_vm_block_size * adev->vm_manager.num_level); + (adev->vm_manager.block_size * + adev->vm_manager.num_level); else if (level == adev->vm_manager.num_level) /* For the page tables on the leaves */ - return AMDGPU_VM_PTE_COUNT; + return AMDGPU_VM_PTE_COUNT(adev); else /* Everything in between */ - return 1 << amdgpu_vm_block_size; + return 1 << adev->vm_manager.block_size; } /** @@ -261,7 +272,7 @@ static int amdgpu_vm_alloc_levels(struct amdgpu_device *adev, unsigned level) { unsigned shift = (adev->vm_manager.num_level - level) * - amdgpu_vm_block_size; + adev->vm_manager.block_size; unsigned pt_idx, from, to; int r; @@ -365,11 +376,19 @@ int amdgpu_vm_alloc_pts(struct amdgpu_device *adev, return amdgpu_vm_alloc_levels(adev, vm, &vm->root, saddr, eaddr, 0); } -static bool amdgpu_vm_is_gpu_reset(struct amdgpu_device *adev, - struct amdgpu_vm_id *id) +/** + * amdgpu_vm_had_gpu_reset - check if reset occured since last use + * + * @adev: amdgpu_device pointer + * @id: VMID structure + * + * Check if GPU reset occured since last use of the VMID. + */ +static bool amdgpu_vm_had_gpu_reset(struct amdgpu_device *adev, + struct amdgpu_vm_id *id) { return id->current_gpu_reset_count != - atomic_read(&adev->gpu_reset_counter) ? true : false; + atomic_read(&adev->gpu_reset_counter); } /** @@ -455,7 +474,7 @@ int amdgpu_vm_grab_id(struct amdgpu_vm *vm, struct amdgpu_ring *ring, /* Check all the prerequisites to using this VMID */ if (!id) continue; - if (amdgpu_vm_is_gpu_reset(adev, id)) + if (amdgpu_vm_had_gpu_reset(adev, id)) continue; if (atomic64_read(&id->owner) != vm->client_id) @@ -483,7 +502,6 @@ int amdgpu_vm_grab_id(struct amdgpu_vm *vm, struct amdgpu_ring *ring, if (r) goto error; - id->current_gpu_reset_count = atomic_read(&adev->gpu_reset_counter); list_move_tail(&id->list, &adev->vm_manager.ids_lru); vm->ids[ring->idx] = id; @@ -504,9 +522,6 @@ int amdgpu_vm_grab_id(struct amdgpu_vm *vm, struct amdgpu_ring *ring, if (r) goto error; - dma_fence_put(id->first); - id->first = dma_fence_get(fence); - dma_fence_put(id->last_flush); id->last_flush = NULL; @@ -557,8 +572,8 @@ static u64 amdgpu_vm_adjust_mc_addr(struct amdgpu_device *adev, u64 mc_addr) { u64 addr = mc_addr; - if (adev->mc.mc_funcs && adev->mc.mc_funcs->adjust_mc_addr) - addr = adev->mc.mc_funcs->adjust_mc_addr(adev, addr); + if (adev->gart.gart_funcs->adjust_mc_addr) + addr = adev->gart.gart_funcs->adjust_mc_addr(adev, addr); return addr; } @@ -583,60 +598,62 @@ int amdgpu_vm_flush(struct amdgpu_ring *ring, struct amdgpu_job *job) id->gws_size != job->gws_size || id->oa_base != job->oa_base || id->oa_size != job->oa_size); + bool vm_flush_needed = job->vm_needs_flush || + amdgpu_vm_ring_has_compute_vm_bug(ring); + unsigned patch_offset = 0; int r; - if (job->vm_needs_flush || gds_switch_needed || - amdgpu_vm_is_gpu_reset(adev, id) || - amdgpu_vm_ring_has_compute_vm_bug(ring)) { - unsigned patch_offset = 0; + if (amdgpu_vm_had_gpu_reset(adev, id)) { + gds_switch_needed = true; + vm_flush_needed = true; + } - if (ring->funcs->init_cond_exec) - patch_offset = amdgpu_ring_init_cond_exec(ring); + if (!vm_flush_needed && !gds_switch_needed) + return 0; - if (ring->funcs->emit_pipeline_sync && - (job->vm_needs_flush || gds_switch_needed || - amdgpu_vm_ring_has_compute_vm_bug(ring))) - amdgpu_ring_emit_pipeline_sync(ring); + if (ring->funcs->init_cond_exec) + patch_offset = amdgpu_ring_init_cond_exec(ring); - if (ring->funcs->emit_vm_flush && (job->vm_needs_flush || - amdgpu_vm_is_gpu_reset(adev, id))) { - struct dma_fence *fence; - u64 pd_addr = amdgpu_vm_adjust_mc_addr(adev, job->vm_pd_addr); + if (ring->funcs->emit_pipeline_sync) + amdgpu_ring_emit_pipeline_sync(ring); - trace_amdgpu_vm_flush(pd_addr, ring->idx, job->vm_id); - amdgpu_ring_emit_vm_flush(ring, job->vm_id, pd_addr); + if (ring->funcs->emit_vm_flush && vm_flush_needed) { + u64 pd_addr = amdgpu_vm_adjust_mc_addr(adev, job->vm_pd_addr); + struct dma_fence *fence; - r = amdgpu_fence_emit(ring, &fence); - if (r) - return r; + trace_amdgpu_vm_flush(pd_addr, ring->idx, job->vm_id); + amdgpu_ring_emit_vm_flush(ring, job->vm_id, pd_addr); - mutex_lock(&adev->vm_manager.lock); - dma_fence_put(id->last_flush); - id->last_flush = fence; - mutex_unlock(&adev->vm_manager.lock); - } + r = amdgpu_fence_emit(ring, &fence); + if (r) + return r; - if (gds_switch_needed) { - id->gds_base = job->gds_base; - id->gds_size = job->gds_size; - id->gws_base = job->gws_base; - id->gws_size = job->gws_size; - id->oa_base = job->oa_base; - id->oa_size = job->oa_size; - amdgpu_ring_emit_gds_switch(ring, job->vm_id, - job->gds_base, job->gds_size, - job->gws_base, job->gws_size, - job->oa_base, job->oa_size); - } + mutex_lock(&adev->vm_manager.lock); + dma_fence_put(id->last_flush); + id->last_flush = fence; + mutex_unlock(&adev->vm_manager.lock); + } - if (ring->funcs->patch_cond_exec) - amdgpu_ring_patch_cond_exec(ring, patch_offset); + if (gds_switch_needed) { + id->gds_base = job->gds_base; + id->gds_size = job->gds_size; + id->gws_base = job->gws_base; + id->gws_size = job->gws_size; + id->oa_base = job->oa_base; + id->oa_size = job->oa_size; + amdgpu_ring_emit_gds_switch(ring, job->vm_id, job->gds_base, + job->gds_size, job->gws_base, + job->gws_size, job->oa_base, + job->oa_size); + } - /* the double SWITCH_BUFFER here *cannot* be skipped by COND_EXEC */ - if (ring->funcs->emit_switch_buffer) { - amdgpu_ring_emit_switch_buffer(ring); - amdgpu_ring_emit_switch_buffer(ring); - } + if (ring->funcs->patch_cond_exec) + amdgpu_ring_patch_cond_exec(ring, patch_offset); + + /* the double SWITCH_BUFFER here *cannot* be skipped by COND_EXEC */ + if (ring->funcs->emit_switch_buffer) { + amdgpu_ring_emit_switch_buffer(ring); + amdgpu_ring_emit_switch_buffer(ring); } return 0; } @@ -960,7 +977,7 @@ static struct amdgpu_bo *amdgpu_vm_get_pt(struct amdgpu_pte_update_params *p, unsigned idx, level = p->adev->vm_manager.num_level; while (entry->entries) { - idx = addr >> (amdgpu_vm_block_size * level--); + idx = addr >> (p->adev->vm_manager.block_size * level--); idx %= amdgpu_bo_size(entry->bo) / 8; entry = &entry->entries[idx]; } @@ -987,7 +1004,8 @@ static void amdgpu_vm_update_ptes(struct amdgpu_pte_update_params *params, uint64_t start, uint64_t end, uint64_t dst, uint64_t flags) { - const uint64_t mask = AMDGPU_VM_PTE_COUNT - 1; + struct amdgpu_device *adev = params->adev; + const uint64_t mask = AMDGPU_VM_PTE_COUNT(adev) - 1; uint64_t cur_pe_start, cur_nptes, cur_dst; uint64_t addr; /* next GPU address to be updated */ @@ -1011,7 +1029,7 @@ static void amdgpu_vm_update_ptes(struct amdgpu_pte_update_params *params, if ((addr & ~mask) == (end & ~mask)) nptes = end - addr; else - nptes = AMDGPU_VM_PTE_COUNT - (addr & mask); + nptes = AMDGPU_VM_PTE_COUNT(adev) - (addr & mask); cur_pe_start = amdgpu_bo_gpu_offset(pt); cur_pe_start += (addr & mask) * 8; @@ -1039,7 +1057,7 @@ static void amdgpu_vm_update_ptes(struct amdgpu_pte_update_params *params, if ((addr & ~mask) == (end & ~mask)) nptes = end - addr; else - nptes = AMDGPU_VM_PTE_COUNT - (addr & mask); + nptes = AMDGPU_VM_PTE_COUNT(adev) - (addr & mask); next_pe_start = amdgpu_bo_gpu_offset(pt); next_pe_start += (addr & mask) * 8; @@ -1186,7 +1204,7 @@ static int amdgpu_vm_bo_update_mapping(struct amdgpu_device *adev, * reserve space for one command every (1 << BLOCK_SIZE) * entries or 2k dwords (whatever is smaller) */ - ncmds = (nptes >> min(amdgpu_vm_block_size, 11)) + 1; + ncmds = (nptes >> min(adev->vm_manager.block_size, 11u)) + 1; /* padding, etc. */ ndw = 64; @@ -1301,7 +1319,7 @@ static int amdgpu_vm_bo_split_mapping(struct amdgpu_device *adev, struct drm_mm_node *nodes, struct dma_fence **fence) { - uint64_t pfn, src = 0, start = mapping->it.start; + uint64_t pfn, src = 0, start = mapping->start; int r; /* normally,bo_va->flags only contians READABLE and WIRTEABLE bit go here @@ -1353,7 +1371,7 @@ static int amdgpu_vm_bo_split_mapping(struct amdgpu_device *adev, } addr += pfn << PAGE_SHIFT; - last = min((uint64_t)mapping->it.last, start + max_entries - 1); + last = min((uint64_t)mapping->last, start + max_entries - 1); r = amdgpu_vm_bo_update_mapping(adev, exclusive, src, pages_addr, vm, start, last, flags, addr, @@ -1368,7 +1386,7 @@ static int amdgpu_vm_bo_split_mapping(struct amdgpu_device *adev, } start = last + 1; - } while (unlikely(start != mapping->it.last + 1)); + } while (unlikely(start != mapping->last + 1)); return 0; } @@ -1518,7 +1536,7 @@ static void amdgpu_vm_add_prt_cb(struct amdgpu_device *adev, if (fence) dma_fence_wait(fence, false); - amdgpu_vm_prt_put(cb->adev); + amdgpu_vm_prt_put(adev); } else { cb->adev = adev; if (!fence || dma_fence_add_callback(fence, &cb->cb, @@ -1724,9 +1742,8 @@ int amdgpu_vm_bo_map(struct amdgpu_device *adev, uint64_t saddr, uint64_t offset, uint64_t size, uint64_t flags) { - struct amdgpu_bo_va_mapping *mapping; + struct amdgpu_bo_va_mapping *mapping, *tmp; struct amdgpu_vm *vm = bo_va->vm; - struct interval_tree_node *it; uint64_t eaddr; /* validate the parameters */ @@ -1743,14 +1760,12 @@ int amdgpu_vm_bo_map(struct amdgpu_device *adev, saddr /= AMDGPU_GPU_PAGE_SIZE; eaddr /= AMDGPU_GPU_PAGE_SIZE; - it = interval_tree_iter_first(&vm->va, saddr, eaddr); - if (it) { - struct amdgpu_bo_va_mapping *tmp; - tmp = container_of(it, struct amdgpu_bo_va_mapping, it); + tmp = amdgpu_vm_it_iter_first(&vm->va, saddr, eaddr); + if (tmp) { /* bo and tmp overlap, invalid addr */ dev_err(adev->dev, "bo %p va 0x%010Lx-0x%010Lx conflict with " - "0x%010lx-0x%010lx\n", bo_va->bo, saddr, eaddr, - tmp->it.start, tmp->it.last + 1); + "0x%010Lx-0x%010Lx\n", bo_va->bo, saddr, eaddr, + tmp->start, tmp->last + 1); return -EINVAL; } @@ -1759,13 +1774,13 @@ int amdgpu_vm_bo_map(struct amdgpu_device *adev, return -ENOMEM; INIT_LIST_HEAD(&mapping->list); - mapping->it.start = saddr; - mapping->it.last = eaddr; + mapping->start = saddr; + mapping->last = eaddr; mapping->offset = offset; mapping->flags = flags; list_add(&mapping->list, &bo_va->invalids); - interval_tree_insert(&mapping->it, &vm->va); + amdgpu_vm_it_insert(mapping, &vm->va); if (flags & AMDGPU_PTE_PRT) amdgpu_vm_prt_get(adev); @@ -1823,13 +1838,13 @@ int amdgpu_vm_bo_replace_map(struct amdgpu_device *adev, saddr /= AMDGPU_GPU_PAGE_SIZE; eaddr /= AMDGPU_GPU_PAGE_SIZE; - mapping->it.start = saddr; - mapping->it.last = eaddr; + mapping->start = saddr; + mapping->last = eaddr; mapping->offset = offset; mapping->flags = flags; list_add(&mapping->list, &bo_va->invalids); - interval_tree_insert(&mapping->it, &vm->va); + amdgpu_vm_it_insert(mapping, &vm->va); if (flags & AMDGPU_PTE_PRT) amdgpu_vm_prt_get(adev); @@ -1860,7 +1875,7 @@ int amdgpu_vm_bo_unmap(struct amdgpu_device *adev, saddr /= AMDGPU_GPU_PAGE_SIZE; list_for_each_entry(mapping, &bo_va->valids, list) { - if (mapping->it.start == saddr) + if (mapping->start == saddr) break; } @@ -1868,7 +1883,7 @@ int amdgpu_vm_bo_unmap(struct amdgpu_device *adev, valid = false; list_for_each_entry(mapping, &bo_va->invalids, list) { - if (mapping->it.start == saddr) + if (mapping->start == saddr) break; } @@ -1877,7 +1892,7 @@ int amdgpu_vm_bo_unmap(struct amdgpu_device *adev, } list_del(&mapping->list); - interval_tree_remove(&mapping->it, &vm->va); + amdgpu_vm_it_remove(mapping, &vm->va); trace_amdgpu_vm_bo_unmap(bo_va, mapping); if (valid) @@ -1905,7 +1920,6 @@ int amdgpu_vm_bo_clear_mappings(struct amdgpu_device *adev, uint64_t saddr, uint64_t size) { struct amdgpu_bo_va_mapping *before, *after, *tmp, *next; - struct interval_tree_node *it; LIST_HEAD(removed); uint64_t eaddr; @@ -1927,43 +1941,42 @@ int amdgpu_vm_bo_clear_mappings(struct amdgpu_device *adev, INIT_LIST_HEAD(&after->list); /* Now gather all removed mappings */ - it = interval_tree_iter_first(&vm->va, saddr, eaddr); - while (it) { - tmp = container_of(it, struct amdgpu_bo_va_mapping, it); - it = interval_tree_iter_next(it, saddr, eaddr); - + tmp = amdgpu_vm_it_iter_first(&vm->va, saddr, eaddr); + while (tmp) { /* Remember mapping split at the start */ - if (tmp->it.start < saddr) { - before->it.start = tmp->it.start; - before->it.last = saddr - 1; + if (tmp->start < saddr) { + before->start = tmp->start; + before->last = saddr - 1; before->offset = tmp->offset; before->flags = tmp->flags; list_add(&before->list, &tmp->list); } /* Remember mapping split at the end */ - if (tmp->it.last > eaddr) { - after->it.start = eaddr + 1; - after->it.last = tmp->it.last; + if (tmp->last > eaddr) { + after->start = eaddr + 1; + after->last = tmp->last; after->offset = tmp->offset; - after->offset += after->it.start - tmp->it.start; + after->offset += after->start - tmp->start; after->flags = tmp->flags; list_add(&after->list, &tmp->list); } list_del(&tmp->list); list_add(&tmp->list, &removed); + + tmp = amdgpu_vm_it_iter_next(tmp, saddr, eaddr); } /* And free them up */ list_for_each_entry_safe(tmp, next, &removed, list) { - interval_tree_remove(&tmp->it, &vm->va); + amdgpu_vm_it_remove(tmp, &vm->va); list_del(&tmp->list); - if (tmp->it.start < saddr) - tmp->it.start = saddr; - if (tmp->it.last > eaddr) - tmp->it.last = eaddr; + if (tmp->start < saddr) + tmp->start = saddr; + if (tmp->last > eaddr) + tmp->last = eaddr; list_add(&tmp->list, &vm->freed); trace_amdgpu_vm_bo_unmap(NULL, tmp); @@ -1971,7 +1984,7 @@ int amdgpu_vm_bo_clear_mappings(struct amdgpu_device *adev, /* Insert partial mapping before the range */ if (!list_empty(&before->list)) { - interval_tree_insert(&before->it, &vm->va); + amdgpu_vm_it_insert(before, &vm->va); if (before->flags & AMDGPU_PTE_PRT) amdgpu_vm_prt_get(adev); } else { @@ -1980,7 +1993,7 @@ int amdgpu_vm_bo_clear_mappings(struct amdgpu_device *adev, /* Insert partial mapping after the range */ if (!list_empty(&after->list)) { - interval_tree_insert(&after->it, &vm->va); + amdgpu_vm_it_insert(after, &vm->va); if (after->flags & AMDGPU_PTE_PRT) amdgpu_vm_prt_get(adev); } else { @@ -2014,13 +2027,13 @@ void amdgpu_vm_bo_rmv(struct amdgpu_device *adev, list_for_each_entry_safe(mapping, next, &bo_va->valids, list) { list_del(&mapping->list); - interval_tree_remove(&mapping->it, &vm->va); + amdgpu_vm_it_remove(mapping, &vm->va); trace_amdgpu_vm_bo_unmap(bo_va, mapping); list_add(&mapping->list, &vm->freed); } list_for_each_entry_safe(mapping, next, &bo_va->invalids, list) { list_del(&mapping->list); - interval_tree_remove(&mapping->it, &vm->va); + amdgpu_vm_it_remove(mapping, &vm->va); amdgpu_vm_free_mapping(adev, vm, mapping, bo_va->last_pt_update); } @@ -2051,6 +2064,44 @@ void amdgpu_vm_bo_invalidate(struct amdgpu_device *adev, } } +static uint32_t amdgpu_vm_get_block_size(uint64_t vm_size) +{ + /* Total bits covered by PD + PTs */ + unsigned bits = ilog2(vm_size) + 18; + + /* Make sure the PD is 4K in size up to 8GB address space. + Above that split equal between PD and PTs */ + if (vm_size <= 8) + return (bits - 9); + else + return ((bits + 3) / 2); +} + +/** + * amdgpu_vm_adjust_size - adjust vm size and block size + * + * @adev: amdgpu_device pointer + * @vm_size: the default vm size if it's set auto + */ +void amdgpu_vm_adjust_size(struct amdgpu_device *adev, uint64_t vm_size) +{ + /* adjust vm size firstly */ + if (amdgpu_vm_size == -1) + adev->vm_manager.vm_size = vm_size; + else + adev->vm_manager.vm_size = amdgpu_vm_size; + + /* block size depends on vm size */ + if (amdgpu_vm_block_size == -1) + adev->vm_manager.block_size = + amdgpu_vm_get_block_size(adev->vm_manager.vm_size); + else + adev->vm_manager.block_size = amdgpu_vm_block_size; + + DRM_INFO("vm size is %llu GB, block size is %u-bit\n", + adev->vm_manager.vm_size, adev->vm_manager.block_size); +} + /** * amdgpu_vm_init - initialize a vm instance * @@ -2062,7 +2113,7 @@ void amdgpu_vm_bo_invalidate(struct amdgpu_device *adev, int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm) { const unsigned align = min(AMDGPU_VM_PTB_ALIGN_SIZE, - AMDGPU_VM_PTE_COUNT * 8); + AMDGPU_VM_PTE_COUNT(adev) * 8); unsigned ring_instance; struct amdgpu_ring *ring; struct amd_sched_rq *rq; @@ -2162,9 +2213,9 @@ void amdgpu_vm_fini(struct amdgpu_device *adev, struct amdgpu_vm *vm) if (!RB_EMPTY_ROOT(&vm->va)) { dev_err(adev->dev, "still active bo inside vm\n"); } - rbtree_postorder_for_each_entry_safe(mapping, tmp, &vm->va, it.rb) { + rbtree_postorder_for_each_entry_safe(mapping, tmp, &vm->va, rb) { list_del(&mapping->list); - interval_tree_remove(&mapping->it, &vm->va); + amdgpu_vm_it_remove(mapping, &vm->va); kfree(mapping); } list_for_each_entry_safe(mapping, tmp, &vm->freed, list) { @@ -2227,7 +2278,6 @@ void amdgpu_vm_manager_fini(struct amdgpu_device *adev) for (i = 0; i < AMDGPU_NUM_VM; ++i) { struct amdgpu_vm_id *id = &adev->vm_manager.ids[i]; - dma_fence_put(adev->vm_manager.ids[i].first); amdgpu_sync_free(&adev->vm_manager.ids[i].active); dma_fence_put(id->flushed_updates); dma_fence_put(id->last_flush); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h index fbe17bf73a00..d9e57290dc71 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h @@ -45,7 +45,7 @@ struct amdgpu_bo_list_entry; #define AMDGPU_VM_MAX_UPDATE_SIZE 0x3FFFF /* number of entries in page table */ -#define AMDGPU_VM_PTE_COUNT (1 << amdgpu_vm_block_size) +#define AMDGPU_VM_PTE_COUNT(adev) (1 << (adev)->vm_manager.block_size) /* PTBs (Page Table Blocks) need to be aligned to 32K */ #define AMDGPU_VM_PTB_ALIGN_SIZE 32768 @@ -76,6 +76,14 @@ struct amdgpu_bo_list_entry; #define AMDGPU_VM_FAULT_STOP_FIRST 1 #define AMDGPU_VM_FAULT_STOP_ALWAYS 2 +/* max number of VMHUB */ +#define AMDGPU_MAX_VMHUBS 2 +#define AMDGPU_GFXHUB 0 +#define AMDGPU_MMHUB 1 + +/* hardcode that limit for now */ +#define AMDGPU_VA_RESERVED_SIZE (8 << 20) + struct amdgpu_vm_pt { struct amdgpu_bo *bo; uint64_t addr; @@ -123,7 +131,6 @@ struct amdgpu_vm { struct amdgpu_vm_id { struct list_head list; - struct dma_fence *first; struct amdgpu_sync active; struct dma_fence *last_flush; atomic64_t owner; @@ -155,6 +162,8 @@ struct amdgpu_vm_manager { uint64_t max_pfn; uint32_t num_level; + uint64_t vm_size; + uint32_t block_size; /* vram base address for page table entry */ u64 vram_base_offset; /* is vm enabled? */ @@ -225,5 +234,6 @@ int amdgpu_vm_bo_clear_mappings(struct amdgpu_device *adev, uint64_t saddr, uint64_t size); void amdgpu_vm_bo_rmv(struct amdgpu_device *adev, struct amdgpu_bo_va *bo_va); +void amdgpu_vm_adjust_size(struct amdgpu_device *adev, uint64_t vm_size); #endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c index 9e577e3d3147..a4831fe0223b 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c @@ -93,7 +93,6 @@ static int amdgpu_vram_mgr_new(struct ttm_mem_type_manager *man, const struct ttm_place *place, struct ttm_mem_reg *mem) { - struct amdgpu_bo *bo = container_of(tbo, struct amdgpu_bo, tbo); struct amdgpu_vram_mgr *mgr = man->priv; struct drm_mm *mm = &mgr->mm; struct drm_mm_node *nodes; @@ -106,8 +105,8 @@ static int amdgpu_vram_mgr_new(struct ttm_mem_type_manager *man, if (!lpfn) lpfn = man->size; - if (bo->flags & AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS || - place->lpfn || amdgpu_vram_page_split == -1) { + if (place->flags & TTM_PL_FLAG_CONTIGUOUS || + amdgpu_vram_page_split == -1) { pages_per_node = ~0ul; num_nodes = 1; } else { @@ -124,12 +123,14 @@ static int amdgpu_vram_mgr_new(struct ttm_mem_type_manager *man, if (place->flags & TTM_PL_FLAG_TOPDOWN) mode = DRM_MM_INSERT_HIGH; + mem->start = 0; pages_left = mem->num_pages; spin_lock(&mgr->lock); for (i = 0; i < num_nodes; ++i) { unsigned long pages = min(pages_left, pages_per_node); uint32_t alignment = mem->page_alignment; + unsigned long start; if (pages == pages_per_node) alignment = pages_per_node; @@ -141,11 +142,19 @@ static int amdgpu_vram_mgr_new(struct ttm_mem_type_manager *man, if (unlikely(r)) goto error; + /* Calculate a virtual BO start address to easily check if + * everything is CPU accessible. + */ + start = nodes[i].start + nodes[i].size; + if (start > mem->num_pages) + start -= mem->num_pages; + else + start = 0; + mem->start = max(mem->start, start); pages_left -= pages; } spin_unlock(&mgr->lock); - mem->start = num_nodes == 1 ? nodes[0].start : AMDGPU_BO_INVALID_OFFSET; mem->mm_node = nodes; return 0; diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c index daf003dd2351..ba98d35340a3 100644 --- a/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c +++ b/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c @@ -1090,23 +1090,10 @@ static u32 dce_v10_0_latency_watermark(struct dce10_wm_params *wm) a.full = dfixed_const(available_bandwidth); b.full = dfixed_const(wm->num_heads); a.full = dfixed_div(a, b); + tmp = div_u64((u64) dmif_size * (u64) wm->disp_clk, mc_latency + 512); + tmp = min(dfixed_trunc(a), tmp); - b.full = dfixed_const(mc_latency + 512); - c.full = dfixed_const(wm->disp_clk); - b.full = dfixed_div(b, c); - - c.full = dfixed_const(dmif_size); - b.full = dfixed_div(c, b); - - tmp = min(dfixed_trunc(a), dfixed_trunc(b)); - - b.full = dfixed_const(1000); - c.full = dfixed_const(wm->disp_clk); - b.full = dfixed_div(c, b); - c.full = dfixed_const(wm->bytes_per_pixel); - b.full = dfixed_mul(b, c); - - lb_fill_bw = min(tmp, dfixed_trunc(b)); + lb_fill_bw = min(tmp, wm->disp_clk * wm->bytes_per_pixel / 1000); a.full = dfixed_const(max_src_lines_per_dst_line * wm->src_width * wm->bytes_per_pixel); b.full = dfixed_const(1000); @@ -1214,14 +1201,14 @@ static void dce_v10_0_program_watermarks(struct amdgpu_device *adev, { struct drm_display_mode *mode = &amdgpu_crtc->base.mode; struct dce10_wm_params wm_low, wm_high; - u32 pixel_period; + u32 active_time; u32 line_time = 0; u32 latency_watermark_a = 0, latency_watermark_b = 0; u32 tmp, wm_mask, lb_vblank_lead_lines = 0; if (amdgpu_crtc->base.enabled && num_heads && mode) { - pixel_period = 1000000 / (u32)mode->clock; - line_time = min((u32)mode->crtc_htotal * pixel_period, (u32)65535); + active_time = 1000000UL * (u32)mode->crtc_hdisplay / (u32)mode->clock; + line_time = min((u32) (1000000UL * (u32)mode->crtc_htotal / (u32)mode->clock), (u32)65535); /* watermark for high clocks */ if (adev->pm.dpm_enabled) { @@ -1236,7 +1223,7 @@ static void dce_v10_0_program_watermarks(struct amdgpu_device *adev, wm_high.disp_clk = mode->clock; wm_high.src_width = mode->crtc_hdisplay; - wm_high.active_time = mode->crtc_hdisplay * pixel_period; + wm_high.active_time = active_time; wm_high.blank_time = line_time - wm_high.active_time; wm_high.interlaced = false; if (mode->flags & DRM_MODE_FLAG_INTERLACE) @@ -1275,7 +1262,7 @@ static void dce_v10_0_program_watermarks(struct amdgpu_device *adev, wm_low.disp_clk = mode->clock; wm_low.src_width = mode->crtc_hdisplay; - wm_low.active_time = mode->crtc_hdisplay * pixel_period; + wm_low.active_time = active_time; wm_low.blank_time = line_time - wm_low.active_time; wm_low.interlaced = false; if (mode->flags & DRM_MODE_FLAG_INTERLACE) diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c index 3a7296724457..e59bc42df18c 100644 --- a/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c +++ b/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c @@ -1059,23 +1059,10 @@ static u32 dce_v11_0_latency_watermark(struct dce10_wm_params *wm) a.full = dfixed_const(available_bandwidth); b.full = dfixed_const(wm->num_heads); a.full = dfixed_div(a, b); + tmp = div_u64((u64) dmif_size * (u64) wm->disp_clk, mc_latency + 512); + tmp = min(dfixed_trunc(a), tmp); - b.full = dfixed_const(mc_latency + 512); - c.full = dfixed_const(wm->disp_clk); - b.full = dfixed_div(b, c); - - c.full = dfixed_const(dmif_size); - b.full = dfixed_div(c, b); - - tmp = min(dfixed_trunc(a), dfixed_trunc(b)); - - b.full = dfixed_const(1000); - c.full = dfixed_const(wm->disp_clk); - b.full = dfixed_div(c, b); - c.full = dfixed_const(wm->bytes_per_pixel); - b.full = dfixed_mul(b, c); - - lb_fill_bw = min(tmp, dfixed_trunc(b)); + lb_fill_bw = min(tmp, wm->disp_clk * wm->bytes_per_pixel / 1000); a.full = dfixed_const(max_src_lines_per_dst_line * wm->src_width * wm->bytes_per_pixel); b.full = dfixed_const(1000); @@ -1183,14 +1170,14 @@ static void dce_v11_0_program_watermarks(struct amdgpu_device *adev, { struct drm_display_mode *mode = &amdgpu_crtc->base.mode; struct dce10_wm_params wm_low, wm_high; - u32 pixel_period; + u32 active_time; u32 line_time = 0; u32 latency_watermark_a = 0, latency_watermark_b = 0; u32 tmp, wm_mask, lb_vblank_lead_lines = 0; if (amdgpu_crtc->base.enabled && num_heads && mode) { - pixel_period = 1000000 / (u32)mode->clock; - line_time = min((u32)mode->crtc_htotal * pixel_period, (u32)65535); + active_time = 1000000UL * (u32)mode->crtc_hdisplay / (u32)mode->clock; + line_time = min((u32) (1000000UL * (u32)mode->crtc_htotal / (u32)mode->clock), (u32)65535); /* watermark for high clocks */ if (adev->pm.dpm_enabled) { @@ -1205,7 +1192,7 @@ static void dce_v11_0_program_watermarks(struct amdgpu_device *adev, wm_high.disp_clk = mode->clock; wm_high.src_width = mode->crtc_hdisplay; - wm_high.active_time = mode->crtc_hdisplay * pixel_period; + wm_high.active_time = active_time; wm_high.blank_time = line_time - wm_high.active_time; wm_high.interlaced = false; if (mode->flags & DRM_MODE_FLAG_INTERLACE) @@ -1244,7 +1231,7 @@ static void dce_v11_0_program_watermarks(struct amdgpu_device *adev, wm_low.disp_clk = mode->clock; wm_low.src_width = mode->crtc_hdisplay; - wm_low.active_time = mode->crtc_hdisplay * pixel_period; + wm_low.active_time = active_time; wm_low.blank_time = line_time - wm_low.active_time; wm_low.interlaced = false; if (mode->flags & DRM_MODE_FLAG_INTERLACE) diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c index 8ccada5d6f39..307269bda4fa 100644 --- a/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c +++ b/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c @@ -861,23 +861,10 @@ static u32 dce_v6_0_latency_watermark(struct dce6_wm_params *wm) a.full = dfixed_const(available_bandwidth); b.full = dfixed_const(wm->num_heads); a.full = dfixed_div(a, b); + tmp = div_u64((u64) dmif_size * (u64) wm->disp_clk, mc_latency + 512); + tmp = min(dfixed_trunc(a), tmp); - b.full = dfixed_const(mc_latency + 512); - c.full = dfixed_const(wm->disp_clk); - b.full = dfixed_div(b, c); - - c.full = dfixed_const(dmif_size); - b.full = dfixed_div(c, b); - - tmp = min(dfixed_trunc(a), dfixed_trunc(b)); - - b.full = dfixed_const(1000); - c.full = dfixed_const(wm->disp_clk); - b.full = dfixed_div(c, b); - c.full = dfixed_const(wm->bytes_per_pixel); - b.full = dfixed_mul(b, c); - - lb_fill_bw = min(tmp, dfixed_trunc(b)); + lb_fill_bw = min(tmp, wm->disp_clk * wm->bytes_per_pixel / 1000); a.full = dfixed_const(max_src_lines_per_dst_line * wm->src_width * wm->bytes_per_pixel); b.full = dfixed_const(1000); @@ -986,7 +973,7 @@ static void dce_v6_0_program_watermarks(struct amdgpu_device *adev, struct drm_display_mode *mode = &amdgpu_crtc->base.mode; struct dce6_wm_params wm_low, wm_high; u32 dram_channels; - u32 pixel_period; + u32 active_time; u32 line_time = 0; u32 latency_watermark_a = 0, latency_watermark_b = 0; u32 priority_a_mark = 0, priority_b_mark = 0; @@ -996,8 +983,8 @@ static void dce_v6_0_program_watermarks(struct amdgpu_device *adev, fixed20_12 a, b, c; if (amdgpu_crtc->base.enabled && num_heads && mode) { - pixel_period = 1000000 / (u32)mode->clock; - line_time = min((u32)mode->crtc_htotal * pixel_period, (u32)65535); + active_time = 1000000UL * (u32)mode->crtc_hdisplay / (u32)mode->clock; + line_time = min((u32) (1000000UL * (u32)mode->crtc_htotal / (u32)mode->clock), (u32)65535); priority_a_cnt = 0; priority_b_cnt = 0; @@ -1016,7 +1003,7 @@ static void dce_v6_0_program_watermarks(struct amdgpu_device *adev, wm_high.disp_clk = mode->clock; wm_high.src_width = mode->crtc_hdisplay; - wm_high.active_time = mode->crtc_hdisplay * pixel_period; + wm_high.active_time = active_time; wm_high.blank_time = line_time - wm_high.active_time; wm_high.interlaced = false; if (mode->flags & DRM_MODE_FLAG_INTERLACE) @@ -1043,7 +1030,7 @@ static void dce_v6_0_program_watermarks(struct amdgpu_device *adev, wm_low.disp_clk = mode->clock; wm_low.src_width = mode->crtc_hdisplay; - wm_low.active_time = mode->crtc_hdisplay * pixel_period; + wm_low.active_time = active_time; wm_low.blank_time = line_time - wm_low.active_time; wm_low.interlaced = false; if (mode->flags & DRM_MODE_FLAG_INTERLACE) diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c index 6943f2641c90..6df7a28e8aac 100644 --- a/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c +++ b/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c @@ -974,23 +974,10 @@ static u32 dce_v8_0_latency_watermark(struct dce8_wm_params *wm) a.full = dfixed_const(available_bandwidth); b.full = dfixed_const(wm->num_heads); a.full = dfixed_div(a, b); + tmp = div_u64((u64) dmif_size * (u64) wm->disp_clk, mc_latency + 512); + tmp = min(dfixed_trunc(a), tmp); - b.full = dfixed_const(mc_latency + 512); - c.full = dfixed_const(wm->disp_clk); - b.full = dfixed_div(b, c); - - c.full = dfixed_const(dmif_size); - b.full = dfixed_div(c, b); - - tmp = min(dfixed_trunc(a), dfixed_trunc(b)); - - b.full = dfixed_const(1000); - c.full = dfixed_const(wm->disp_clk); - b.full = dfixed_div(c, b); - c.full = dfixed_const(wm->bytes_per_pixel); - b.full = dfixed_mul(b, c); - - lb_fill_bw = min(tmp, dfixed_trunc(b)); + lb_fill_bw = min(tmp, wm->disp_clk * wm->bytes_per_pixel / 1000); a.full = dfixed_const(max_src_lines_per_dst_line * wm->src_width * wm->bytes_per_pixel); b.full = dfixed_const(1000); @@ -1098,14 +1085,14 @@ static void dce_v8_0_program_watermarks(struct amdgpu_device *adev, { struct drm_display_mode *mode = &amdgpu_crtc->base.mode; struct dce8_wm_params wm_low, wm_high; - u32 pixel_period; + u32 active_time; u32 line_time = 0; u32 latency_watermark_a = 0, latency_watermark_b = 0; u32 tmp, wm_mask, lb_vblank_lead_lines = 0; if (amdgpu_crtc->base.enabled && num_heads && mode) { - pixel_period = 1000000 / (u32)mode->clock; - line_time = min((u32)mode->crtc_htotal * pixel_period, (u32)65535); + active_time = 1000000UL * (u32)mode->crtc_hdisplay / (u32)mode->clock; + line_time = min((u32) (1000000UL * (u32)mode->crtc_htotal / (u32)mode->clock), (u32)65535); /* watermark for high clocks */ if (adev->pm.dpm_enabled) { @@ -1120,7 +1107,7 @@ static void dce_v8_0_program_watermarks(struct amdgpu_device *adev, wm_high.disp_clk = mode->clock; wm_high.src_width = mode->crtc_hdisplay; - wm_high.active_time = mode->crtc_hdisplay * pixel_period; + wm_high.active_time = active_time; wm_high.blank_time = line_time - wm_high.active_time; wm_high.interlaced = false; if (mode->flags & DRM_MODE_FLAG_INTERLACE) @@ -1159,7 +1146,7 @@ static void dce_v8_0_program_watermarks(struct amdgpu_device *adev, wm_low.disp_clk = mode->clock; wm_low.src_width = mode->crtc_hdisplay; - wm_low.active_time = mode->crtc_hdisplay * pixel_period; + wm_low.active_time = active_time; wm_low.blank_time = line_time - wm_low.active_time; wm_low.interlaced = false; if (mode->flags & DRM_MODE_FLAG_INTERLACE) diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c index e0fa0d30e162..dad8a4cd1b37 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c @@ -4565,6 +4565,7 @@ static void gfx_v8_0_cp_compute_enable(struct amdgpu_device *adev, bool enable) WREG32(mmCP_MEC_CNTL, (CP_MEC_CNTL__MEC_ME1_HALT_MASK | CP_MEC_CNTL__MEC_ME2_HALT_MASK)); for (i = 0; i < adev->gfx.num_compute_rings; i++) adev->gfx.compute_ring[i].ready = false; + adev->gfx.kiq.ring.ready = false; } udelay(50); } @@ -4721,14 +4722,10 @@ static int gfx_v8_0_mqd_init(struct amdgpu_ring *ring) mqd->cp_hqd_eop_control = tmp; /* enable doorbell? */ - tmp = RREG32(mmCP_HQD_PQ_DOORBELL_CONTROL); - - if (ring->use_doorbell) - tmp = REG_SET_FIELD(tmp, CP_HQD_PQ_DOORBELL_CONTROL, - DOORBELL_EN, 1); - else - tmp = REG_SET_FIELD(tmp, CP_HQD_PQ_DOORBELL_CONTROL, - DOORBELL_EN, 0); + tmp = REG_SET_FIELD(RREG32(mmCP_HQD_PQ_DOORBELL_CONTROL), + CP_HQD_PQ_DOORBELL_CONTROL, + DOORBELL_EN, + ring->use_doorbell ? 1 : 0); mqd->cp_hqd_pq_doorbell_control = tmp; @@ -4816,13 +4813,10 @@ static int gfx_v8_0_kiq_init_register(struct amdgpu_ring *ring) { struct amdgpu_device *adev = ring->adev; struct vi_mqd *mqd = ring->mqd_ptr; - uint32_t tmp; int j; /* disable wptr polling */ - tmp = RREG32(mmCP_PQ_WPTR_POLL_CNTL); - tmp = REG_SET_FIELD(tmp, CP_PQ_WPTR_POLL_CNTL, EN, 0); - WREG32(mmCP_PQ_WPTR_POLL_CNTL, tmp); + WREG32_FIELD(CP_PQ_WPTR_POLL_CNTL, EN, 0); WREG32(mmCP_HQD_EOP_BASE_ADDR, mqd->cp_hqd_eop_base_addr_lo); WREG32(mmCP_HQD_EOP_BASE_ADDR_HI, mqd->cp_hqd_eop_base_addr_hi); @@ -4834,10 +4828,10 @@ static int gfx_v8_0_kiq_init_register(struct amdgpu_ring *ring) WREG32(mmCP_HQD_PQ_DOORBELL_CONTROL, mqd->cp_hqd_pq_doorbell_control); /* disable the queue if it's active */ - if (RREG32(mmCP_HQD_ACTIVE) & 1) { + if (RREG32(mmCP_HQD_ACTIVE) & CP_HQD_ACTIVE__ACTIVE_MASK) { WREG32(mmCP_HQD_DEQUEUE_REQUEST, 1); for (j = 0; j < adev->usec_timeout; j++) { - if (!(RREG32(mmCP_HQD_ACTIVE) & 1)) + if (!(RREG32(mmCP_HQD_ACTIVE) & CP_HQD_ACTIVE__ACTIVE_MASK)) break; udelay(1); } @@ -4894,11 +4888,8 @@ static int gfx_v8_0_kiq_init_register(struct amdgpu_ring *ring) /* activate the queue */ WREG32(mmCP_HQD_ACTIVE, mqd->cp_hqd_active); - if (ring->use_doorbell) { - tmp = RREG32(mmCP_PQ_STATUS); - tmp = REG_SET_FIELD(tmp, CP_PQ_STATUS, DOORBELL_ENABLE, 1); - WREG32(mmCP_PQ_STATUS, tmp); - } + if (ring->use_doorbell) + WREG32_FIELD(CP_PQ_STATUS, DOORBELL_ENABLE, 1); return 0; } @@ -5471,19 +5462,18 @@ static void gfx_v8_0_inactive_hqd(struct amdgpu_device *adev, { int i; + mutex_lock(&adev->srbm_mutex); vi_srbm_select(adev, ring->me, ring->pipe, ring->queue, 0); if (RREG32(mmCP_HQD_ACTIVE) & CP_HQD_ACTIVE__ACTIVE_MASK) { - u32 tmp; - tmp = RREG32(mmCP_HQD_DEQUEUE_REQUEST); - tmp = REG_SET_FIELD(tmp, CP_HQD_DEQUEUE_REQUEST, - DEQUEUE_REQ, 2); - WREG32(mmCP_HQD_DEQUEUE_REQUEST, tmp); + WREG32_FIELD(CP_HQD_DEQUEUE_REQUEST, DEQUEUE_REQ, 2); for (i = 0; i < adev->usec_timeout; i++) { if (!(RREG32(mmCP_HQD_ACTIVE) & CP_HQD_ACTIVE__ACTIVE_MASK)) break; udelay(1); } } + vi_srbm_select(adev, 0, 0, 0, 0); + mutex_unlock(&adev->srbm_mutex); } static int gfx_v8_0_pre_soft_reset(void *handle) @@ -5589,11 +5579,13 @@ static int gfx_v8_0_soft_reset(void *handle) static void gfx_v8_0_init_hqd(struct amdgpu_device *adev, struct amdgpu_ring *ring) { + mutex_lock(&adev->srbm_mutex); vi_srbm_select(adev, ring->me, ring->pipe, ring->queue, 0); WREG32(mmCP_HQD_DEQUEUE_REQUEST, 0); WREG32(mmCP_HQD_PQ_RPTR, 0); WREG32(mmCP_HQD_PQ_WPTR, 0); vi_srbm_select(adev, 0, 0, 0, 0); + mutex_unlock(&adev->srbm_mutex); } static int gfx_v8_0_post_soft_reset(void *handle) @@ -6986,40 +6978,24 @@ static int gfx_v8_0_kiq_set_interrupt_state(struct amdgpu_device *adev, unsigned int type, enum amdgpu_interrupt_state state) { - uint32_t tmp, target; struct amdgpu_ring *ring = &(adev->gfx.kiq.ring); BUG_ON(ring->funcs->type != AMDGPU_RING_TYPE_KIQ); - if (ring->me == 1) - target = mmCP_ME1_PIPE0_INT_CNTL; - else - target = mmCP_ME2_PIPE0_INT_CNTL; - target += ring->pipe; - switch (type) { case AMDGPU_CP_KIQ_IRQ_DRIVER0: - if (state == AMDGPU_IRQ_STATE_DISABLE) { - tmp = RREG32(mmCPC_INT_CNTL); - tmp = REG_SET_FIELD(tmp, CPC_INT_CNTL, - GENERIC2_INT_ENABLE, 0); - WREG32(mmCPC_INT_CNTL, tmp); - - tmp = RREG32(target); - tmp = REG_SET_FIELD(tmp, CP_ME2_PIPE0_INT_CNTL, - GENERIC2_INT_ENABLE, 0); - WREG32(target, tmp); - } else { - tmp = RREG32(mmCPC_INT_CNTL); - tmp = REG_SET_FIELD(tmp, CPC_INT_CNTL, - GENERIC2_INT_ENABLE, 1); - WREG32(mmCPC_INT_CNTL, tmp); - - tmp = RREG32(target); - tmp = REG_SET_FIELD(tmp, CP_ME2_PIPE0_INT_CNTL, - GENERIC2_INT_ENABLE, 1); - WREG32(target, tmp); - } + WREG32_FIELD(CPC_INT_CNTL, GENERIC2_INT_ENABLE, + state == AMDGPU_IRQ_STATE_DISABLE ? 0 : 1); + if (ring->me == 1) + WREG32_FIELD_OFFSET(CP_ME1_PIPE0_INT_CNTL, + ring->pipe, + GENERIC2_INT_ENABLE, + state == AMDGPU_IRQ_STATE_DISABLE ? 0 : 1); + else + WREG32_FIELD_OFFSET(CP_ME2_PIPE0_INT_CNTL, + ring->pipe, + GENERIC2_INT_ENABLE, + state == AMDGPU_IRQ_STATE_DISABLE ? 0 : 1); break; default: BUG(); /* kiq only support GENERIC2_INT now */ @@ -7159,8 +7135,6 @@ static const struct amdgpu_ring_funcs gfx_v8_0_ring_funcs_kiq = { .emit_ib_size = 4, /* gfx_v8_0_ring_emit_ib_compute */ .emit_ib = gfx_v8_0_ring_emit_ib_compute, .emit_fence = gfx_v8_0_ring_emit_fence_kiq, - .emit_hdp_flush = gfx_v8_0_ring_emit_hdp_flush, - .emit_hdp_invalidate = gfx_v8_0_ring_emit_hdp_invalidate, .test_ring = gfx_v8_0_ring_test_ring, .test_ib = gfx_v8_0_ring_test_ib, .insert_nop = amdgpu_ring_insert_nop, diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c index 669bb98fc45d..a447b70841c9 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c @@ -1288,9 +1288,7 @@ static void gfx_v9_0_gpu_init(struct amdgpu_device *adev) u32 tmp; int i; - tmp = RREG32(SOC15_REG_OFFSET(GC, 0, mmGRBM_CNTL)); - tmp = REG_SET_FIELD(tmp, GRBM_CNTL, READ_TIMEOUT, 0xff); - WREG32(SOC15_REG_OFFSET(GC, 0, mmGRBM_CNTL), tmp); + WREG32_FIELD15(GC, 0, GRBM_CNTL, READ_TIMEOUT, 0xff); gfx_v9_0_tiling_mode_table_init(adev); @@ -1395,13 +1393,9 @@ void gfx_v9_0_rlc_stop(struct amdgpu_device *adev) static void gfx_v9_0_rlc_reset(struct amdgpu_device *adev) { - u32 tmp = RREG32(SOC15_REG_OFFSET(GC, 0, mmGRBM_SOFT_RESET)); - - tmp = REG_SET_FIELD(tmp, GRBM_SOFT_RESET, SOFT_RESET_RLC, 1); - WREG32(SOC15_REG_OFFSET(GC, 0, mmGRBM_SOFT_RESET), tmp); + WREG32_FIELD15(GC, 0, GRBM_SOFT_RESET, SOFT_RESET_RLC, 1); udelay(50); - tmp = REG_SET_FIELD(tmp, GRBM_SOFT_RESET, SOFT_RESET_RLC, 0); - WREG32(SOC15_REG_OFFSET(GC, 0, mmGRBM_SOFT_RESET), tmp); + WREG32_FIELD15(GC, 0, GRBM_SOFT_RESET, SOFT_RESET_RLC, 0); udelay(50); } @@ -1410,10 +1404,8 @@ static void gfx_v9_0_rlc_start(struct amdgpu_device *adev) #ifdef AMDGPU_RLC_DEBUG_RETRY u32 rlc_ucode_ver; #endif - u32 tmp = RREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_CNTL)); - tmp = REG_SET_FIELD(tmp, RLC_CNTL, RLC_ENABLE_F32, 1); - WREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_CNTL), tmp); + WREG32_FIELD15(GC, 0, RLC_CNTL, RLC_ENABLE_F32, 1); /* carrizo do enable cp interrupt after cp inited */ if (!(adev->flags & AMD_IS_APU)) @@ -1497,14 +1489,10 @@ static void gfx_v9_0_cp_gfx_enable(struct amdgpu_device *adev, bool enable) int i; u32 tmp = RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_ME_CNTL)); - if (enable) { - tmp = REG_SET_FIELD(tmp, CP_ME_CNTL, ME_HALT, 0); - tmp = REG_SET_FIELD(tmp, CP_ME_CNTL, PFP_HALT, 0); - tmp = REG_SET_FIELD(tmp, CP_ME_CNTL, CE_HALT, 0); - } else { - tmp = REG_SET_FIELD(tmp, CP_ME_CNTL, ME_HALT, 1); - tmp = REG_SET_FIELD(tmp, CP_ME_CNTL, PFP_HALT, 1); - tmp = REG_SET_FIELD(tmp, CP_ME_CNTL, CE_HALT, 1); + tmp = REG_SET_FIELD(tmp, CP_ME_CNTL, ME_HALT, enable ? 0 : 1); + tmp = REG_SET_FIELD(tmp, CP_ME_CNTL, PFP_HALT, enable ? 0 : 1); + tmp = REG_SET_FIELD(tmp, CP_ME_CNTL, CE_HALT, enable ? 0 : 1); + if (!enable) { for (i = 0; i < adev->gfx.num_gfx_rings; i++) adev->gfx.gfx_ring[i].ready = false; } @@ -2020,13 +2008,10 @@ static int gfx_v9_0_kiq_init_register(struct amdgpu_ring *ring) { struct amdgpu_device *adev = ring->adev; struct v9_mqd *mqd = ring->mqd_ptr; - uint32_t tmp; int j; /* disable wptr polling */ - tmp = RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_PQ_WPTR_POLL_CNTL)); - tmp = REG_SET_FIELD(tmp, CP_PQ_WPTR_POLL_CNTL, EN, 0); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_PQ_WPTR_POLL_CNTL), tmp); + WREG32_FIELD15(GC, 0, CP_PQ_WPTR_POLL_CNTL, EN, 0); WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_EOP_BASE_ADDR), mqd->cp_hqd_eop_base_addr_lo); @@ -2118,11 +2103,8 @@ static int gfx_v9_0_kiq_init_register(struct amdgpu_ring *ring) WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_ACTIVE), mqd->cp_hqd_active); - if (ring->use_doorbell) { - tmp = RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_PQ_STATUS)); - tmp = REG_SET_FIELD(tmp, CP_PQ_STATUS, DOORBELL_ENABLE, 1); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_PQ_STATUS), tmp); - } + if (ring->use_doorbell) + WREG32_FIELD15(GC, 0, CP_PQ_STATUS, DOORBELL_ENABLE, 1); return 0; } @@ -2366,177 +2348,6 @@ static int gfx_v9_0_wait_for_idle(void *handle) return -ETIMEDOUT; } -static void gfx_v9_0_print_status(void *handle) -{ - int i; - struct amdgpu_device *adev = (struct amdgpu_device *)handle; - - dev_info(adev->dev, "GFX 9.x registers\n"); - dev_info(adev->dev, " GRBM_STATUS=0x%08X\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmGRBM_STATUS))); - dev_info(adev->dev, " GRBM_STATUS2=0x%08X\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmGRBM_STATUS2))); - dev_info(adev->dev, " GRBM_STATUS_SE0=0x%08X\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmGRBM_STATUS_SE0))); - dev_info(adev->dev, " GRBM_STATUS_SE1=0x%08X\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmGRBM_STATUS_SE1))); - dev_info(adev->dev, " GRBM_STATUS_SE2=0x%08X\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmGRBM_STATUS_SE2))); - dev_info(adev->dev, " GRBM_STATUS_SE3=0x%08X\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmGRBM_STATUS_SE3))); - dev_info(adev->dev, " CP_STAT = 0x%08x\n", RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_STAT))); - dev_info(adev->dev, " CP_STALLED_STAT1 = 0x%08x\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_STALLED_STAT1))); - dev_info(adev->dev, " CP_STALLED_STAT2 = 0x%08x\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_STALLED_STAT2))); - dev_info(adev->dev, " CP_STALLED_STAT3 = 0x%08x\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_STALLED_STAT3))); - dev_info(adev->dev, " CP_CPF_BUSY_STAT = 0x%08x\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_CPF_BUSY_STAT))); - dev_info(adev->dev, " CP_CPF_STALLED_STAT1 = 0x%08x\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_CPF_STALLED_STAT1))); - dev_info(adev->dev, " CP_CPF_STATUS = 0x%08x\n", RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_CPF_STATUS))); - dev_info(adev->dev, " CP_CPC_BUSY_STAT = 0x%08x\n", RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_CPC_BUSY_STAT))); - dev_info(adev->dev, " CP_CPC_STALLED_STAT1 = 0x%08x\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_CPC_STALLED_STAT1))); - dev_info(adev->dev, " CP_CPC_STATUS = 0x%08x\n", RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_CPC_STATUS))); - - for (i = 0; i < 32; i++) { - dev_info(adev->dev, " GB_TILE_MODE%d=0x%08X\n", - i, RREG32(SOC15_REG_OFFSET(GC, 0, mmGB_TILE_MODE0 ) + i*4)); - } - for (i = 0; i < 16; i++) { - dev_info(adev->dev, " GB_MACROTILE_MODE%d=0x%08X\n", - i, RREG32(SOC15_REG_OFFSET(GC, 0, mmGB_MACROTILE_MODE0) + i*4)); - } - for (i = 0; i < adev->gfx.config.max_shader_engines; i++) { - dev_info(adev->dev, " se: %d\n", i); - gfx_v9_0_select_se_sh(adev, i, 0xffffffff, 0xffffffff); - dev_info(adev->dev, " PA_SC_RASTER_CONFIG=0x%08X\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmPA_SC_RASTER_CONFIG))); - dev_info(adev->dev, " PA_SC_RASTER_CONFIG_1=0x%08X\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmPA_SC_RASTER_CONFIG_1))); - } - gfx_v9_0_select_se_sh(adev, 0xffffffff, 0xffffffff, 0xffffffff); - - dev_info(adev->dev, " GB_ADDR_CONFIG=0x%08X\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmGB_ADDR_CONFIG))); - - dev_info(adev->dev, " CP_MEQ_THRESHOLDS=0x%08X\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_MEQ_THRESHOLDS))); - dev_info(adev->dev, " SX_DEBUG_1=0x%08X\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmSX_DEBUG_1))); - dev_info(adev->dev, " TA_CNTL_AUX=0x%08X\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmTA_CNTL_AUX))); - dev_info(adev->dev, " SPI_CONFIG_CNTL=0x%08X\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmSPI_CONFIG_CNTL))); - dev_info(adev->dev, " SQ_CONFIG=0x%08X\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmSQ_CONFIG))); - dev_info(adev->dev, " DB_DEBUG=0x%08X\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmDB_DEBUG))); - dev_info(adev->dev, " DB_DEBUG2=0x%08X\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmDB_DEBUG2))); - dev_info(adev->dev, " DB_DEBUG3=0x%08X\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmDB_DEBUG3))); - dev_info(adev->dev, " CB_HW_CONTROL=0x%08X\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmCB_HW_CONTROL))); - dev_info(adev->dev, " SPI_CONFIG_CNTL_1=0x%08X\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmSPI_CONFIG_CNTL_1))); - dev_info(adev->dev, " PA_SC_FIFO_SIZE=0x%08X\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmPA_SC_FIFO_SIZE))); - dev_info(adev->dev, " VGT_NUM_INSTANCES=0x%08X\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmVGT_NUM_INSTANCES))); - dev_info(adev->dev, " CP_PERFMON_CNTL=0x%08X\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_PERFMON_CNTL))); - dev_info(adev->dev, " PA_SC_FORCE_EOV_MAX_CNTS=0x%08X\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmPA_SC_FORCE_EOV_MAX_CNTS))); - dev_info(adev->dev, " VGT_CACHE_INVALIDATION=0x%08X\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmVGT_CACHE_INVALIDATION))); - dev_info(adev->dev, " VGT_GS_VERTEX_REUSE=0x%08X\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmVGT_GS_VERTEX_REUSE))); - dev_info(adev->dev, " PA_SC_LINE_STIPPLE_STATE=0x%08X\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmPA_SC_LINE_STIPPLE_STATE))); - dev_info(adev->dev, " PA_CL_ENHANCE=0x%08X\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmPA_CL_ENHANCE))); - dev_info(adev->dev, " PA_SC_ENHANCE=0x%08X\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmPA_SC_ENHANCE))); - - dev_info(adev->dev, " CP_ME_CNTL=0x%08X\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_ME_CNTL))); - dev_info(adev->dev, " CP_MAX_CONTEXT=0x%08X\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_MAX_CONTEXT))); - dev_info(adev->dev, " CP_DEVICE_ID=0x%08X\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_DEVICE_ID))); - - dev_info(adev->dev, " CP_SEM_WAIT_TIMER=0x%08X\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_SEM_WAIT_TIMER))); - - dev_info(adev->dev, " CP_RB_WPTR_DELAY=0x%08X\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_RB_WPTR_DELAY))); - dev_info(adev->dev, " CP_RB_VMID=0x%08X\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_RB_VMID))); - dev_info(adev->dev, " CP_RB0_CNTL=0x%08X\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_RB0_CNTL))); - dev_info(adev->dev, " CP_RB0_WPTR=0x%08X\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_RB0_WPTR))); - dev_info(adev->dev, " CP_RB0_RPTR_ADDR=0x%08X\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_RB0_RPTR_ADDR))); - dev_info(adev->dev, " CP_RB0_RPTR_ADDR_HI=0x%08X\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_RB0_RPTR_ADDR_HI))); - dev_info(adev->dev, " CP_RB0_CNTL=0x%08X\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_RB0_CNTL))); - dev_info(adev->dev, " CP_RB0_BASE=0x%08X\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_RB0_BASE))); - dev_info(adev->dev, " CP_RB0_BASE_HI=0x%08X\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_RB0_BASE_HI))); - dev_info(adev->dev, " CP_MEC_CNTL=0x%08X\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_MEC_CNTL))); - - dev_info(adev->dev, " SCRATCH_ADDR=0x%08X\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmSCRATCH_ADDR))); - dev_info(adev->dev, " SCRATCH_UMSK=0x%08X\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmSCRATCH_UMSK))); - - dev_info(adev->dev, " CP_INT_CNTL_RING0=0x%08X\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_INT_CNTL_RING0))); - dev_info(adev->dev, " RLC_LB_CNTL=0x%08X\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_LB_CNTL))); - dev_info(adev->dev, " RLC_CNTL=0x%08X\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_CNTL))); - dev_info(adev->dev, " RLC_CGCG_CGLS_CTRL=0x%08X\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_CGCG_CGLS_CTRL))); - dev_info(adev->dev, " RLC_LB_CNTR_INIT=0x%08X\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_LB_CNTR_INIT))); - dev_info(adev->dev, " RLC_LB_CNTR_MAX=0x%08X\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_LB_CNTR_MAX))); - dev_info(adev->dev, " RLC_LB_INIT_CU_MASK=0x%08X\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_LB_INIT_CU_MASK))); - dev_info(adev->dev, " RLC_LB_PARAMS=0x%08X\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_LB_PARAMS))); - dev_info(adev->dev, " RLC_LB_CNTL=0x%08X\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_LB_CNTL))); - dev_info(adev->dev, " RLC_UCODE_CNTL=0x%08X\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_UCODE_CNTL))); - - dev_info(adev->dev, " RLC_GPM_GENERAL_6=0x%08X\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_GPM_GENERAL_6))); - dev_info(adev->dev, " RLC_GPM_GENERAL_12=0x%08X\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_GPM_GENERAL_12))); - dev_info(adev->dev, " RLC_GPM_TIMER_INT_3=0x%08X\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_GPM_TIMER_INT_3))); - mutex_lock(&adev->srbm_mutex); - for (i = 0; i < 16; i++) { - soc15_grbm_select(adev, 0, 0, 0, i); - dev_info(adev->dev, " VM %d:\n", i); - dev_info(adev->dev, " SH_MEM_CONFIG=0x%08X\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmSH_MEM_CONFIG))); - dev_info(adev->dev, " SH_MEM_BASES=0x%08X\n", - RREG32(SOC15_REG_OFFSET(GC, 0, mmSH_MEM_BASES))); - } - soc15_grbm_select(adev, 0, 0, 0, 0); - mutex_unlock(&adev->srbm_mutex); -} - static int gfx_v9_0_soft_reset(void *handle) { u32 grbm_soft_reset = 0; @@ -2569,8 +2380,7 @@ static int gfx_v9_0_soft_reset(void *handle) GRBM_SOFT_RESET, SOFT_RESET_RLC, 1); - if (grbm_soft_reset ) { - gfx_v9_0_print_status((void *)adev); + if (grbm_soft_reset) { /* stop the rlc */ gfx_v9_0_rlc_stop(adev); @@ -2596,7 +2406,6 @@ static int gfx_v9_0_soft_reset(void *handle) /* Wait a little for things to settle down */ udelay(50); - gfx_v9_0_print_status((void *)adev); } return 0; } @@ -3148,6 +2957,7 @@ static void gfx_v9_0_ring_emit_vm_flush(struct amdgpu_ring *ring, unsigned vm_id, uint64_t pd_addr) { int usepfp = (ring->funcs->type == AMDGPU_RING_TYPE_GFX); + uint32_t req = ring->adev->gart.gart_funcs->get_invalidate_req(vm_id); unsigned eng = ring->idx; unsigned i; @@ -3157,7 +2967,6 @@ static void gfx_v9_0_ring_emit_vm_flush(struct amdgpu_ring *ring, for (i = 0; i < AMDGPU_MAX_VMHUBS; ++i) { struct amdgpu_vmhub *hub = &ring->adev->vmhub[i]; - uint32_t req = hub->get_invalidate_req(vm_id); gfx_v9_0_write_data_to_reg(ring, usepfp, true, hub->ctx0_ptb_addr_lo32 @@ -3376,21 +3185,12 @@ static void gfx_v9_0_ring_emit_wreg(struct amdgpu_ring *ring, uint32_t reg, static void gfx_v9_0_set_gfx_eop_interrupt_state(struct amdgpu_device *adev, enum amdgpu_interrupt_state state) { - u32 cp_int_cntl; - switch (state) { case AMDGPU_IRQ_STATE_DISABLE: - cp_int_cntl = RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_INT_CNTL_RING0)); - cp_int_cntl = REG_SET_FIELD(cp_int_cntl, CP_INT_CNTL_RING0, - TIME_STAMP_INT_ENABLE, 0); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_INT_CNTL_RING0), cp_int_cntl); - break; case AMDGPU_IRQ_STATE_ENABLE: - cp_int_cntl = RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_INT_CNTL_RING0)); - cp_int_cntl = - REG_SET_FIELD(cp_int_cntl, CP_INT_CNTL_RING0, - TIME_STAMP_INT_ENABLE, 1); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_INT_CNTL_RING0), cp_int_cntl); + WREG32_FIELD15(GC, 0, CP_INT_CNTL_RING0, + TIME_STAMP_INT_ENABLE, + state == AMDGPU_IRQ_STATE_ENABLE ? 1 : 0); break; default: break; @@ -3446,20 +3246,12 @@ static int gfx_v9_0_set_priv_reg_fault_state(struct amdgpu_device *adev, unsigned type, enum amdgpu_interrupt_state state) { - u32 cp_int_cntl; - switch (state) { case AMDGPU_IRQ_STATE_DISABLE: - cp_int_cntl = RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_INT_CNTL_RING0)); - cp_int_cntl = REG_SET_FIELD(cp_int_cntl, CP_INT_CNTL_RING0, - PRIV_REG_INT_ENABLE, 0); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_INT_CNTL_RING0), cp_int_cntl); - break; case AMDGPU_IRQ_STATE_ENABLE: - cp_int_cntl = RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_INT_CNTL_RING0)); - cp_int_cntl = REG_SET_FIELD(cp_int_cntl, CP_INT_CNTL_RING0, - PRIV_REG_INT_ENABLE, 1); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_INT_CNTL_RING0), cp_int_cntl); + WREG32_FIELD15(GC, 0, CP_INT_CNTL_RING0, + PRIV_REG_INT_ENABLE, + state == AMDGPU_IRQ_STATE_ENABLE ? 1 : 0); break; default: break; @@ -3473,21 +3265,12 @@ static int gfx_v9_0_set_priv_inst_fault_state(struct amdgpu_device *adev, unsigned type, enum amdgpu_interrupt_state state) { - u32 cp_int_cntl; - switch (state) { case AMDGPU_IRQ_STATE_DISABLE: - cp_int_cntl = RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_INT_CNTL_RING0)); - cp_int_cntl = REG_SET_FIELD(cp_int_cntl, CP_INT_CNTL_RING0, - PRIV_INSTR_INT_ENABLE, 0); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_INT_CNTL_RING0), cp_int_cntl); - break; case AMDGPU_IRQ_STATE_ENABLE: - cp_int_cntl = RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_INT_CNTL_RING0)); - cp_int_cntl = REG_SET_FIELD(cp_int_cntl, CP_INT_CNTL_RING0, - PRIV_INSTR_INT_ENABLE, 1); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_INT_CNTL_RING0), cp_int_cntl); - break; + WREG32_FIELD15(GC, 0, CP_INT_CNTL_RING0, + PRIV_INSTR_INT_ENABLE, + state == AMDGPU_IRQ_STATE_ENABLE ? 1 : 0); default: break; } @@ -3759,8 +3542,6 @@ static const struct amdgpu_ring_funcs gfx_v9_0_ring_funcs_kiq = { .emit_ib_size = 4, /* gfx_v9_0_ring_emit_ib_compute */ .emit_ib = gfx_v9_0_ring_emit_ib_compute, .emit_fence = gfx_v9_0_ring_emit_fence_kiq, - .emit_hdp_flush = gfx_v9_0_ring_emit_hdp_flush, - .emit_hdp_invalidate = gfx_v9_0_ring_emit_hdp_invalidate, .test_ring = gfx_v9_0_ring_test_ring, .test_ib = gfx_v9_0_ring_test_ib, .insert_nop = amdgpu_ring_insert_nop, @@ -3975,9 +3756,7 @@ static int gfx_v9_0_init_queue(struct amdgpu_ring *ring) ring->pipe, ring->queue, 0); /* disable wptr polling */ - tmp = RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_PQ_WPTR_POLL_CNTL)); - tmp = REG_SET_FIELD(tmp, CP_PQ_WPTR_POLL_CNTL, EN, 0); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_PQ_WPTR_POLL_CNTL), tmp); + WREG32_FIELD15(GC, 0, CP_PQ_WPTR_POLL_CNTL, EN, 0); /* write the EOP addr */ BUG_ON(ring->me != 1 || ring->pipe != 0); /* can't handle other cases eop address */ @@ -4121,11 +3900,8 @@ static int gfx_v9_0_init_queue(struct amdgpu_ring *ring) amdgpu_bo_kunmap(ring->mqd_obj); amdgpu_bo_unreserve(ring->mqd_obj); - if (use_doorbell) { - tmp = RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_PQ_STATUS)); - tmp = REG_SET_FIELD(tmp, CP_PQ_STATUS, DOORBELL_ENABLE, 1); - WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_PQ_STATUS), tmp); - } + if (use_doorbell) + WREG32_FIELD15(GC, 0, CP_PQ_STATUS, DOORBELL_ENABLE, 1); return 0; } diff --git a/drivers/gpu/drm/amd/amdgpu/gfxhub_v1_0.c b/drivers/gpu/drm/amd/amdgpu/gfxhub_v1_0.c index 30ef3126c8a9..005075ff00f7 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfxhub_v1_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfxhub_v1_0.c @@ -222,7 +222,7 @@ int gfxhub_v1_0_gart_enable(struct amdgpu_device *adev) EXECUTE_PROTECTION_FAULT_ENABLE_DEFAULT, 1); tmp = REG_SET_FIELD(tmp, VM_CONTEXT1_CNTL, PAGE_TABLE_BLOCK_SIZE, - amdgpu_vm_block_size - 9); + adev->vm_manager.block_size - 9); WREG32(SOC15_REG_OFFSET(GC, 0, mmVM_CONTEXT1_CNTL) + i, tmp); WREG32(SOC15_REG_OFFSET(GC, 0, mmVM_CONTEXT1_PAGE_TABLE_START_ADDR_LO32) + i*2, 0); WREG32(SOC15_REG_OFFSET(GC, 0, mmVM_CONTEXT1_PAGE_TABLE_START_ADDR_HI32) + i*2, 0); @@ -299,36 +299,6 @@ void gfxhub_v1_0_set_fault_enable_default(struct amdgpu_device *adev, WREG32(SOC15_REG_OFFSET(GC, 0, mmVM_L2_PROTECTION_FAULT_CNTL), tmp); } -static uint32_t gfxhub_v1_0_get_invalidate_req(unsigned int vm_id) -{ - u32 req = 0; - - /* invalidate using legacy mode on vm_id*/ - req = REG_SET_FIELD(req, VM_INVALIDATE_ENG0_REQ, - PER_VMID_INVALIDATE_REQ, 1 << vm_id); - req = REG_SET_FIELD(req, VM_INVALIDATE_ENG0_REQ, FLUSH_TYPE, 0); - req = REG_SET_FIELD(req, VM_INVALIDATE_ENG0_REQ, INVALIDATE_L2_PTES, 1); - req = REG_SET_FIELD(req, VM_INVALIDATE_ENG0_REQ, INVALIDATE_L2_PDE0, 1); - req = REG_SET_FIELD(req, VM_INVALIDATE_ENG0_REQ, INVALIDATE_L2_PDE1, 1); - req = REG_SET_FIELD(req, VM_INVALIDATE_ENG0_REQ, INVALIDATE_L2_PDE2, 1); - req = REG_SET_FIELD(req, VM_INVALIDATE_ENG0_REQ, INVALIDATE_L1_PTES, 1); - req = REG_SET_FIELD(req, VM_INVALIDATE_ENG0_REQ, - CLEAR_PROTECTION_FAULT_STATUS_ADDR, 0); - - return req; -} - -static uint32_t gfxhub_v1_0_get_vm_protection_bits(void) -{ - return (VM_CONTEXT1_CNTL__RANGE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK | - VM_CONTEXT1_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK | - VM_CONTEXT1_CNTL__PDE0_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK | - VM_CONTEXT1_CNTL__VALID_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK | - VM_CONTEXT1_CNTL__READ_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK | - VM_CONTEXT1_CNTL__WRITE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK | - VM_CONTEXT1_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK); -} - static int gfxhub_v1_0_early_init(void *handle) { return 0; @@ -361,9 +331,6 @@ static int gfxhub_v1_0_sw_init(void *handle) hub->vm_l2_pro_fault_cntl = SOC15_REG_OFFSET(GC, 0, mmVM_L2_PROTECTION_FAULT_CNTL); - hub->get_invalidate_req = gfxhub_v1_0_get_invalidate_req; - hub->get_vm_protection_bits = gfxhub_v1_0_get_vm_protection_bits; - return 0; } diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c index d9586601a437..631aef38126d 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c @@ -543,7 +543,8 @@ static int gmc_v6_0_gart_enable(struct amdgpu_device *adev) WREG32(mmVM_CONTEXT1_CNTL, VM_CONTEXT1_CNTL__ENABLE_CONTEXT_MASK | (1UL << VM_CONTEXT1_CNTL__PAGE_TABLE_DEPTH__SHIFT) | - ((amdgpu_vm_block_size - 9) << VM_CONTEXT1_CNTL__PAGE_TABLE_BLOCK_SIZE__SHIFT)); + ((adev->vm_manager.block_size - 9) + << VM_CONTEXT1_CNTL__PAGE_TABLE_BLOCK_SIZE__SHIFT)); if (amdgpu_vm_fault_stop == AMDGPU_VM_FAULT_STOP_ALWAYS) gmc_v6_0_set_fault_enable_default(adev, false); else @@ -848,7 +849,8 @@ static int gmc_v6_0_sw_init(void *handle) if (r) return r; - adev->vm_manager.max_pfn = amdgpu_vm_size << 18; + amdgpu_vm_adjust_size(adev, 64); + adev->vm_manager.max_pfn = adev->vm_manager.vm_size << 18; adev->mc.mc_mask = 0xffffffffffULL; diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c index 0c0a6015cca5..92abe12d92bb 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c @@ -37,6 +37,8 @@ #include "oss/oss_2_0_d.h" #include "oss/oss_2_0_sh_mask.h" +#include "amdgpu_atombios.h" + static void gmc_v7_0_set_gart_funcs(struct amdgpu_device *adev); static void gmc_v7_0_set_irq_funcs(struct amdgpu_device *adev); static int gmc_v7_0_wait_for_idle(void *handle); @@ -325,48 +327,51 @@ static void gmc_v7_0_mc_program(struct amdgpu_device *adev) */ static int gmc_v7_0_mc_init(struct amdgpu_device *adev) { - u32 tmp; - int chansize, numchan; - - /* Get VRAM informations */ - tmp = RREG32(mmMC_ARB_RAMCFG); - if (REG_GET_FIELD(tmp, MC_ARB_RAMCFG, CHANSIZE)) { - chansize = 64; - } else { - chansize = 32; - } - tmp = RREG32(mmMC_SHARED_CHMAP); - switch (REG_GET_FIELD(tmp, MC_SHARED_CHMAP, NOOFCHAN)) { - case 0: - default: - numchan = 1; - break; - case 1: - numchan = 2; - break; - case 2: - numchan = 4; - break; - case 3: - numchan = 8; - break; - case 4: - numchan = 3; - break; - case 5: - numchan = 6; - break; - case 6: - numchan = 10; - break; - case 7: - numchan = 12; - break; - case 8: - numchan = 16; - break; + adev->mc.vram_width = amdgpu_atombios_get_vram_width(adev); + if (!adev->mc.vram_width) { + u32 tmp; + int chansize, numchan; + + /* Get VRAM informations */ + tmp = RREG32(mmMC_ARB_RAMCFG); + if (REG_GET_FIELD(tmp, MC_ARB_RAMCFG, CHANSIZE)) { + chansize = 64; + } else { + chansize = 32; + } + tmp = RREG32(mmMC_SHARED_CHMAP); + switch (REG_GET_FIELD(tmp, MC_SHARED_CHMAP, NOOFCHAN)) { + case 0: + default: + numchan = 1; + break; + case 1: + numchan = 2; + break; + case 2: + numchan = 4; + break; + case 3: + numchan = 8; + break; + case 4: + numchan = 3; + break; + case 5: + numchan = 6; + break; + case 6: + numchan = 10; + break; + case 7: + numchan = 12; + break; + case 8: + numchan = 16; + break; + } + adev->mc.vram_width = numchan * chansize; } - adev->mc.vram_width = numchan * chansize; /* Could aper size report 0 ? */ adev->mc.aper_base = pci_resource_start(adev->pdev, 0); adev->mc.aper_size = pci_resource_len(adev->pdev, 0); @@ -639,7 +644,7 @@ static int gmc_v7_0_gart_enable(struct amdgpu_device *adev) tmp = REG_SET_FIELD(tmp, VM_CONTEXT1_CNTL, ENABLE_CONTEXT, 1); tmp = REG_SET_FIELD(tmp, VM_CONTEXT1_CNTL, PAGE_TABLE_DEPTH, 1); tmp = REG_SET_FIELD(tmp, VM_CONTEXT1_CNTL, PAGE_TABLE_BLOCK_SIZE, - amdgpu_vm_block_size - 9); + adev->vm_manager.block_size - 9); WREG32(mmVM_CONTEXT1_CNTL, tmp); if (amdgpu_vm_fault_stop == AMDGPU_VM_FAULT_STOP_ALWAYS) gmc_v7_0_set_fault_enable_default(adev, false); @@ -998,7 +1003,8 @@ static int gmc_v7_0_sw_init(void *handle) * Currently set to 4GB ((1 << 20) 4k pages). * Max GPUVM size for cayman and SI is 40 bits. */ - adev->vm_manager.max_pfn = amdgpu_vm_size << 18; + amdgpu_vm_adjust_size(adev, 64); + adev->vm_manager.max_pfn = adev->vm_manager.vm_size << 18; /* Set the internal MC address mask * This is the max address of the GPU's diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c index d19d1c5e2847..f2ccefc66fd4 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c @@ -38,6 +38,8 @@ #include "vid.h" #include "vi.h" +#include "amdgpu_atombios.h" + static void gmc_v8_0_set_gart_funcs(struct amdgpu_device *adev); static void gmc_v8_0_set_irq_funcs(struct amdgpu_device *adev); @@ -487,48 +489,51 @@ static void gmc_v8_0_mc_program(struct amdgpu_device *adev) */ static int gmc_v8_0_mc_init(struct amdgpu_device *adev) { - u32 tmp; - int chansize, numchan; - - /* Get VRAM informations */ - tmp = RREG32(mmMC_ARB_RAMCFG); - if (REG_GET_FIELD(tmp, MC_ARB_RAMCFG, CHANSIZE)) { - chansize = 64; - } else { - chansize = 32; - } - tmp = RREG32(mmMC_SHARED_CHMAP); - switch (REG_GET_FIELD(tmp, MC_SHARED_CHMAP, NOOFCHAN)) { - case 0: - default: - numchan = 1; - break; - case 1: - numchan = 2; - break; - case 2: - numchan = 4; - break; - case 3: - numchan = 8; - break; - case 4: - numchan = 3; - break; - case 5: - numchan = 6; - break; - case 6: - numchan = 10; - break; - case 7: - numchan = 12; - break; - case 8: - numchan = 16; - break; + adev->mc.vram_width = amdgpu_atombios_get_vram_width(adev); + if (!adev->mc.vram_width) { + u32 tmp; + int chansize, numchan; + + /* Get VRAM informations */ + tmp = RREG32(mmMC_ARB_RAMCFG); + if (REG_GET_FIELD(tmp, MC_ARB_RAMCFG, CHANSIZE)) { + chansize = 64; + } else { + chansize = 32; + } + tmp = RREG32(mmMC_SHARED_CHMAP); + switch (REG_GET_FIELD(tmp, MC_SHARED_CHMAP, NOOFCHAN)) { + case 0: + default: + numchan = 1; + break; + case 1: + numchan = 2; + break; + case 2: + numchan = 4; + break; + case 3: + numchan = 8; + break; + case 4: + numchan = 3; + break; + case 5: + numchan = 6; + break; + case 6: + numchan = 10; + break; + case 7: + numchan = 12; + break; + case 8: + numchan = 16; + break; + } + adev->mc.vram_width = numchan * chansize; } - adev->mc.vram_width = numchan * chansize; /* Could aper size report 0 ? */ adev->mc.aper_base = pci_resource_start(adev->pdev, 0); adev->mc.aper_size = pci_resource_len(adev->pdev, 0); @@ -848,7 +853,7 @@ static int gmc_v8_0_gart_enable(struct amdgpu_device *adev) tmp = REG_SET_FIELD(tmp, VM_CONTEXT1_CNTL, WRITE_PROTECTION_FAULT_ENABLE_DEFAULT, 1); tmp = REG_SET_FIELD(tmp, VM_CONTEXT1_CNTL, EXECUTE_PROTECTION_FAULT_ENABLE_DEFAULT, 1); tmp = REG_SET_FIELD(tmp, VM_CONTEXT1_CNTL, PAGE_TABLE_BLOCK_SIZE, - amdgpu_vm_block_size - 9); + adev->vm_manager.block_size - 9); WREG32(mmVM_CONTEXT1_CNTL, tmp); if (amdgpu_vm_fault_stop == AMDGPU_VM_FAULT_STOP_ALWAYS) gmc_v8_0_set_fault_enable_default(adev, false); @@ -1082,7 +1087,8 @@ static int gmc_v8_0_sw_init(void *handle) * Currently set to 4GB ((1 << 20) 4k pages). * Max GPUVM size for cayman and SI is 40 bits. */ - adev->vm_manager.max_pfn = amdgpu_vm_size << 18; + amdgpu_vm_adjust_size(adev, 64); + adev->vm_manager.max_pfn = adev->vm_manager.vm_size << 18; /* Set the internal MC address mask * This is the max address of the GPU's diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c index df69aae99df4..3b045e0b114e 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c @@ -75,11 +75,18 @@ static int gmc_v9_0_vm_fault_interrupt_state(struct amdgpu_device *adev, struct amdgpu_vmhub *hub; u32 tmp, reg, bits, i; + bits = VM_CONTEXT1_CNTL__RANGE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK | + VM_CONTEXT1_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK | + VM_CONTEXT1_CNTL__PDE0_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK | + VM_CONTEXT1_CNTL__VALID_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK | + VM_CONTEXT1_CNTL__READ_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK | + VM_CONTEXT1_CNTL__WRITE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK | + VM_CONTEXT1_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK; + switch (state) { case AMDGPU_IRQ_STATE_DISABLE: /* MM HUB */ hub = &adev->vmhub[AMDGPU_MMHUB]; - bits = hub->get_vm_protection_bits(); for (i = 0; i< 16; i++) { reg = hub->vm_context0_cntl + i; tmp = RREG32(reg); @@ -89,7 +96,6 @@ static int gmc_v9_0_vm_fault_interrupt_state(struct amdgpu_device *adev, /* GFX HUB */ hub = &adev->vmhub[AMDGPU_GFXHUB]; - bits = hub->get_vm_protection_bits(); for (i = 0; i < 16; i++) { reg = hub->vm_context0_cntl + i; tmp = RREG32(reg); @@ -100,7 +106,6 @@ static int gmc_v9_0_vm_fault_interrupt_state(struct amdgpu_device *adev, case AMDGPU_IRQ_STATE_ENABLE: /* MM HUB */ hub = &adev->vmhub[AMDGPU_MMHUB]; - bits = hub->get_vm_protection_bits(); for (i = 0; i< 16; i++) { reg = hub->vm_context0_cntl + i; tmp = RREG32(reg); @@ -110,7 +115,6 @@ static int gmc_v9_0_vm_fault_interrupt_state(struct amdgpu_device *adev, /* GFX HUB */ hub = &adev->vmhub[AMDGPU_GFXHUB]; - bits = hub->get_vm_protection_bits(); for (i = 0; i < 16; i++) { reg = hub->vm_context0_cntl + i; tmp = RREG32(reg); @@ -129,8 +133,7 @@ static int gmc_v9_0_process_interrupt(struct amdgpu_device *adev, struct amdgpu_irq_src *source, struct amdgpu_iv_entry *entry) { - struct amdgpu_vmhub *gfxhub = &adev->vmhub[AMDGPU_GFXHUB]; - struct amdgpu_vmhub *mmhub = &adev->vmhub[AMDGPU_MMHUB]; + struct amdgpu_vmhub *hub = &adev->vmhub[entry->vm_id_src]; uint32_t status = 0; u64 addr; @@ -138,13 +141,8 @@ static int gmc_v9_0_process_interrupt(struct amdgpu_device *adev, addr |= ((u64)entry->src_data[1] & 0xf) << 44; if (!amdgpu_sriov_vf(adev)) { - if (entry->vm_id_src) { - status = RREG32(mmhub->vm_l2_pro_fault_status); - WREG32_P(mmhub->vm_l2_pro_fault_cntl, 1, ~1); - } else { - status = RREG32(gfxhub->vm_l2_pro_fault_status); - WREG32_P(gfxhub->vm_l2_pro_fault_cntl, 1, ~1); - } + status = RREG32(hub->vm_l2_pro_fault_status); + WREG32_P(hub->vm_l2_pro_fault_cntl, 1, ~1); } if (printk_ratelimit()) { @@ -175,6 +173,25 @@ static void gmc_v9_0_set_irq_funcs(struct amdgpu_device *adev) adev->mc.vm_fault.funcs = &gmc_v9_0_irq_funcs; } +static uint32_t gmc_v9_0_get_invalidate_req(unsigned int vm_id) +{ + u32 req = 0; + + /* invalidate using legacy mode on vm_id*/ + req = REG_SET_FIELD(req, VM_INVALIDATE_ENG0_REQ, + PER_VMID_INVALIDATE_REQ, 1 << vm_id); + req = REG_SET_FIELD(req, VM_INVALIDATE_ENG0_REQ, FLUSH_TYPE, 0); + req = REG_SET_FIELD(req, VM_INVALIDATE_ENG0_REQ, INVALIDATE_L2_PTES, 1); + req = REG_SET_FIELD(req, VM_INVALIDATE_ENG0_REQ, INVALIDATE_L2_PDE0, 1); + req = REG_SET_FIELD(req, VM_INVALIDATE_ENG0_REQ, INVALIDATE_L2_PDE1, 1); + req = REG_SET_FIELD(req, VM_INVALIDATE_ENG0_REQ, INVALIDATE_L2_PDE2, 1); + req = REG_SET_FIELD(req, VM_INVALIDATE_ENG0_REQ, INVALIDATE_L1_PTES, 1); + req = REG_SET_FIELD(req, VM_INVALIDATE_ENG0_REQ, + CLEAR_PROTECTION_FAULT_STATUS_ADDR, 0); + + return req; +} + /* * GART * VMID 0 is the physical GPU addresses as used by the kernel. @@ -204,7 +221,7 @@ static void gmc_v9_0_gart_flush_gpu_tlb(struct amdgpu_device *adev, for (i = 0; i < AMDGPU_MAX_VMHUBS; ++i) { struct amdgpu_vmhub *hub = &adev->vmhub[i]; - u32 tmp = hub->get_invalidate_req(vmid); + u32 tmp = gmc_v9_0_get_invalidate_req(vmid); WREG32_NO_KIQ(hub->vm_inv_eng0_req + eng, tmp); @@ -337,30 +354,23 @@ static uint64_t gmc_v9_0_get_vm_pte_flags(struct amdgpu_device *adev, return pte_flag; } -static const struct amdgpu_gart_funcs gmc_v9_0_gart_funcs = { - .flush_gpu_tlb = gmc_v9_0_gart_flush_gpu_tlb, - .set_pte_pde = gmc_v9_0_gart_set_pte_pde, - .get_vm_pte_flags = gmc_v9_0_get_vm_pte_flags -}; - -static void gmc_v9_0_set_gart_funcs(struct amdgpu_device *adev) -{ - if (adev->gart.gart_funcs == NULL) - adev->gart.gart_funcs = &gmc_v9_0_gart_funcs; -} - static u64 gmc_v9_0_adjust_mc_addr(struct amdgpu_device *adev, u64 mc_addr) { return adev->vm_manager.vram_base_offset + mc_addr - adev->mc.vram_start; } -static const struct amdgpu_mc_funcs gmc_v9_0_mc_funcs = { +static const struct amdgpu_gart_funcs gmc_v9_0_gart_funcs = { + .flush_gpu_tlb = gmc_v9_0_gart_flush_gpu_tlb, + .set_pte_pde = gmc_v9_0_gart_set_pte_pde, + .get_vm_pte_flags = gmc_v9_0_get_vm_pte_flags, .adjust_mc_addr = gmc_v9_0_adjust_mc_addr, + .get_invalidate_req = gmc_v9_0_get_invalidate_req, }; -static void gmc_v9_0_set_mc_funcs(struct amdgpu_device *adev) +static void gmc_v9_0_set_gart_funcs(struct amdgpu_device *adev) { - adev->mc.mc_funcs = &gmc_v9_0_mc_funcs; + if (adev->gart.gart_funcs == NULL) + adev->gart.gart_funcs = &gmc_v9_0_gart_funcs; } static int gmc_v9_0_early_init(void *handle) @@ -368,7 +378,6 @@ static int gmc_v9_0_early_init(void *handle) struct amdgpu_device *adev = (struct amdgpu_device *)handle; gmc_v9_0_set_gart_funcs(adev); - gmc_v9_0_set_mc_funcs(adev); gmc_v9_0_set_irq_funcs(adev); return 0; @@ -511,7 +520,12 @@ static int gmc_v9_0_vm_init(struct amdgpu_device *adev) * amdkfd will use VMIDs 8-15 */ adev->vm_manager.num_ids = AMDGPU_NUM_OF_VMIDS; - adev->vm_manager.num_level = 3; + + /* 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); /* base offset of vram pages */ @@ -543,9 +557,20 @@ static int gmc_v9_0_sw_init(void *handle) if (adev->flags & AMD_IS_APU) { adev->mc.vram_type = AMDGPU_VRAM_TYPE_UNKNOWN; + amdgpu_vm_adjust_size(adev, 64); } else { /* XXX Don't know how to get VRAM type yet. */ adev->mc.vram_type = AMDGPU_VRAM_TYPE_HBM; + /* + * To fulfill 4-level page support, + * vm size is 256TB (48bit), maximum size of Vega10, + * block size 512 (9bit) + */ + 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); } /* This interrupt is VMC page fault.*/ @@ -557,14 +582,7 @@ static int gmc_v9_0_sw_init(void *handle) if (r) return r; - /* Because of four level VMPTs, vm size is at least 512GB. - * The maximum size is 256TB (48bit). - */ - if (amdgpu_vm_size < 512) { - DRM_WARN("VM size is at least 512GB!\n"); - amdgpu_vm_size = 512; - } - adev->vm_manager.max_pfn = (uint64_t)amdgpu_vm_size << 18; + adev->vm_manager.max_pfn = adev->vm_manager.vm_size << 18; /* Set the internal MC address mask * This is the max address of the GPU's diff --git a/drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.c b/drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.c index 266a0f47a908..62684510ddcd 100644 --- a/drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.c +++ b/drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.c @@ -242,7 +242,7 @@ int mmhub_v1_0_gart_enable(struct amdgpu_device *adev) EXECUTE_PROTECTION_FAULT_ENABLE_DEFAULT, 1); tmp = REG_SET_FIELD(tmp, VM_CONTEXT1_CNTL, PAGE_TABLE_BLOCK_SIZE, - amdgpu_vm_block_size - 9); + adev->vm_manager.block_size - 9); WREG32(SOC15_REG_OFFSET(MMHUB, 0, mmVM_CONTEXT1_CNTL) + i, tmp); WREG32(SOC15_REG_OFFSET(MMHUB, 0, mmVM_CONTEXT1_PAGE_TABLE_START_ADDR_LO32) + i*2, 0); WREG32(SOC15_REG_OFFSET(MMHUB, 0, mmVM_CONTEXT1_PAGE_TABLE_START_ADDR_HI32) + i*2, 0); @@ -317,36 +317,6 @@ void mmhub_v1_0_set_fault_enable_default(struct amdgpu_device *adev, bool value) WREG32(SOC15_REG_OFFSET(MMHUB, 0, mmVM_L2_PROTECTION_FAULT_CNTL), tmp); } -static uint32_t mmhub_v1_0_get_invalidate_req(unsigned int vm_id) -{ - u32 req = 0; - - /* invalidate using legacy mode on vm_id*/ - req = REG_SET_FIELD(req, VM_INVALIDATE_ENG0_REQ, - PER_VMID_INVALIDATE_REQ, 1 << vm_id); - req = REG_SET_FIELD(req, VM_INVALIDATE_ENG0_REQ, FLUSH_TYPE, 0); - req = REG_SET_FIELD(req, VM_INVALIDATE_ENG0_REQ, INVALIDATE_L2_PTES, 1); - req = REG_SET_FIELD(req, VM_INVALIDATE_ENG0_REQ, INVALIDATE_L2_PDE0, 1); - req = REG_SET_FIELD(req, VM_INVALIDATE_ENG0_REQ, INVALIDATE_L2_PDE1, 1); - req = REG_SET_FIELD(req, VM_INVALIDATE_ENG0_REQ, INVALIDATE_L2_PDE2, 1); - req = REG_SET_FIELD(req, VM_INVALIDATE_ENG0_REQ, INVALIDATE_L1_PTES, 1); - req = REG_SET_FIELD(req, VM_INVALIDATE_ENG0_REQ, - CLEAR_PROTECTION_FAULT_STATUS_ADDR, 0); - - return req; -} - -static uint32_t mmhub_v1_0_get_vm_protection_bits(void) -{ - return (VM_CONTEXT1_CNTL__RANGE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK | - VM_CONTEXT1_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK | - VM_CONTEXT1_CNTL__PDE0_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK | - VM_CONTEXT1_CNTL__VALID_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK | - VM_CONTEXT1_CNTL__READ_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK | - VM_CONTEXT1_CNTL__WRITE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK | - VM_CONTEXT1_CNTL__EXECUTE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK); -} - static int mmhub_v1_0_early_init(void *handle) { return 0; @@ -379,9 +349,6 @@ static int mmhub_v1_0_sw_init(void *handle) hub->vm_l2_pro_fault_cntl = SOC15_REG_OFFSET(MMHUB, 0, mmVM_L2_PROTECTION_FAULT_CNTL); - hub->get_invalidate_req = mmhub_v1_0_get_invalidate_req; - hub->get_vm_protection_bits = mmhub_v1_0_get_vm_protection_bits; - return 0; } diff --git a/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c b/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c index cfd5e54777bb..1493301b6a94 100644 --- a/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c +++ b/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c @@ -28,6 +28,7 @@ #include "vega10/GC/gc_9_0_offset.h" #include "vega10/GC/gc_9_0_sh_mask.h" #include "soc15.h" +#include "vega10_ih.h" #include "soc15_common.h" #include "mxgpu_ai.h" @@ -133,7 +134,7 @@ static int xgpu_ai_poll_ack(struct amdgpu_device *adev) return r; } -static int xgpu_vi_poll_msg(struct amdgpu_device *adev, enum idh_event event) +static int xgpu_ai_poll_msg(struct amdgpu_device *adev, enum idh_event event) { int r = 0, timeout = AI_MAILBOX_TIMEDOUT; @@ -172,7 +173,7 @@ static int xgpu_ai_send_access_requests(struct amdgpu_device *adev, if (req == IDH_REQ_GPU_INIT_ACCESS || req == IDH_REQ_GPU_FINI_ACCESS || req == IDH_REQ_GPU_RESET_ACCESS) { - r = xgpu_vi_poll_msg(adev, IDH_READY_TO_ACCESS_GPU); + r = xgpu_ai_poll_msg(adev, IDH_READY_TO_ACCESS_GPU); if (r) return r; } @@ -180,6 +181,11 @@ static int xgpu_ai_send_access_requests(struct amdgpu_device *adev, return 0; } +static int xgpu_ai_request_reset(struct amdgpu_device *adev) +{ + return xgpu_ai_send_access_requests(adev, IDH_REQ_GPU_RESET_ACCESS); +} + static int xgpu_ai_request_full_gpu_access(struct amdgpu_device *adev, bool init) { @@ -201,7 +207,134 @@ static int xgpu_ai_release_full_gpu_access(struct amdgpu_device *adev, return r; } +static int xgpu_ai_mailbox_ack_irq(struct amdgpu_device *adev, + struct amdgpu_irq_src *source, + struct amdgpu_iv_entry *entry) +{ + DRM_DEBUG("get ack intr and do nothing.\n"); + return 0; +} + +static int xgpu_ai_set_mailbox_ack_irq(struct amdgpu_device *adev, + struct amdgpu_irq_src *source, + unsigned type, + enum amdgpu_interrupt_state state) +{ + u32 tmp = RREG32_NO_KIQ(SOC15_REG_OFFSET(NBIO, 0, mmBIF_BX_PF0_MAILBOX_INT_CNTL)); + + tmp = REG_SET_FIELD(tmp, BIF_BX_PF0_MAILBOX_INT_CNTL, ACK_INT_EN, + (state == AMDGPU_IRQ_STATE_ENABLE) ? 1 : 0); + WREG32_NO_KIQ(SOC15_REG_OFFSET(NBIO, 0, mmBIF_BX_PF0_MAILBOX_INT_CNTL), tmp); + + return 0; +} + +static void xgpu_ai_mailbox_flr_work(struct work_struct *work) +{ + struct amdgpu_virt *virt = container_of(work, struct amdgpu_virt, flr_work); + struct amdgpu_device *adev = container_of(virt, struct amdgpu_device, virt); + + /* wait until RCV_MSG become 3 */ + if (xgpu_ai_poll_msg(adev, IDH_FLR_NOTIFICATION_CMPL)) { + pr_err("failed to recieve FLR_CMPL\n"); + return; + } + + /* Trigger recovery due to world switch failure */ + amdgpu_sriov_gpu_reset(adev, false); +} + +static int xgpu_ai_set_mailbox_rcv_irq(struct amdgpu_device *adev, + struct amdgpu_irq_src *src, + unsigned type, + enum amdgpu_interrupt_state state) +{ + u32 tmp = RREG32_NO_KIQ(SOC15_REG_OFFSET(NBIO, 0, mmBIF_BX_PF0_MAILBOX_INT_CNTL)); + + tmp = REG_SET_FIELD(tmp, BIF_BX_PF0_MAILBOX_INT_CNTL, VALID_INT_EN, + (state == AMDGPU_IRQ_STATE_ENABLE) ? 1 : 0); + WREG32_NO_KIQ(SOC15_REG_OFFSET(NBIO, 0, mmBIF_BX_PF0_MAILBOX_INT_CNTL), tmp); + + return 0; +} + +static int xgpu_ai_mailbox_rcv_irq(struct amdgpu_device *adev, + struct amdgpu_irq_src *source, + struct amdgpu_iv_entry *entry) +{ + int r; + + /* see what event we get */ + r = xgpu_ai_mailbox_rcv_msg(adev, IDH_FLR_NOTIFICATION); + + /* only handle FLR_NOTIFY now */ + if (!r) + schedule_work(&adev->virt.flr_work); + + return 0; +} + +static const struct amdgpu_irq_src_funcs xgpu_ai_mailbox_ack_irq_funcs = { + .set = xgpu_ai_set_mailbox_ack_irq, + .process = xgpu_ai_mailbox_ack_irq, +}; + +static const struct amdgpu_irq_src_funcs xgpu_ai_mailbox_rcv_irq_funcs = { + .set = xgpu_ai_set_mailbox_rcv_irq, + .process = xgpu_ai_mailbox_rcv_irq, +}; + +void xgpu_ai_mailbox_set_irq_funcs(struct amdgpu_device *adev) +{ + adev->virt.ack_irq.num_types = 1; + adev->virt.ack_irq.funcs = &xgpu_ai_mailbox_ack_irq_funcs; + adev->virt.rcv_irq.num_types = 1; + adev->virt.rcv_irq.funcs = &xgpu_ai_mailbox_rcv_irq_funcs; +} + +int xgpu_ai_mailbox_add_irq_id(struct amdgpu_device *adev) +{ + int r; + + r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, 135, &adev->virt.rcv_irq); + if (r) + return r; + + r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, 138, &adev->virt.ack_irq); + if (r) { + amdgpu_irq_put(adev, &adev->virt.rcv_irq, 0); + return r; + } + + return 0; +} + +int xgpu_ai_mailbox_get_irq(struct amdgpu_device *adev) +{ + int r; + + r = amdgpu_irq_get(adev, &adev->virt.rcv_irq, 0); + if (r) + return r; + r = amdgpu_irq_get(adev, &adev->virt.ack_irq, 0); + if (r) { + amdgpu_irq_put(adev, &adev->virt.rcv_irq, 0); + return r; + } + + INIT_WORK(&adev->virt.flr_work, xgpu_ai_mailbox_flr_work); + + return 0; +} + +void xgpu_ai_mailbox_put_irq(struct amdgpu_device *adev) +{ + amdgpu_irq_put(adev, &adev->virt.ack_irq, 0); + amdgpu_irq_put(adev, &adev->virt.rcv_irq, 0); +} + 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, }; diff --git a/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.h b/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.h index bf8ab8fd4367..9aefc44d2c34 100644 --- a/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.h +++ b/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.h @@ -24,7 +24,7 @@ #ifndef __MXGPU_AI_H__ #define __MXGPU_AI_H__ -#define AI_MAILBOX_TIMEDOUT 150000 +#define AI_MAILBOX_TIMEDOUT 5000 enum idh_request { IDH_REQ_GPU_INIT_ACCESS = 1, @@ -44,4 +44,9 @@ enum idh_event { extern const struct amdgpu_virt_ops xgpu_ai_virt_ops; +void xgpu_ai_mailbox_set_irq_funcs(struct amdgpu_device *adev); +int xgpu_ai_mailbox_add_irq_id(struct amdgpu_device *adev); +int xgpu_ai_mailbox_get_irq(struct amdgpu_device *adev); +void xgpu_ai_mailbox_put_irq(struct amdgpu_device *adev); + #endif diff --git a/drivers/gpu/drm/amd/amdgpu/psp_v3_1.c b/drivers/gpu/drm/amd/amdgpu/psp_v3_1.c index 5191c45ffdf3..c3588d1c7cb0 100644 --- a/drivers/gpu/drm/amd/amdgpu/psp_v3_1.c +++ b/drivers/gpu/drm/amd/amdgpu/psp_v3_1.c @@ -491,7 +491,7 @@ bool psp_v3_1_compare_sram_data(struct psp_context *psp, ucode_size = ucode->ucode_size; ucode_mem = (uint32_t *)ucode->kaddr; - while (!ucode_size) { + while (ucode_size) { fw_sram_reg_val = RREG32(fw_sram_data_reg_offset); if (*ucode_mem != fw_sram_reg_val) @@ -508,14 +508,10 @@ bool psp_v3_1_compare_sram_data(struct psp_context *psp, bool psp_v3_1_smu_reload_quirk(struct psp_context *psp) { struct amdgpu_device *adev = psp->adev; - uint32_t reg, reg_val; + uint32_t reg; - reg_val = (smnMP1_FIRMWARE_FLAGS & 0xffffffff) | 0x03b00000; - WREG32(SOC15_REG_OFFSET(NBIO, 0, mmPCIE_INDEX2), reg_val); + reg = smnMP1_FIRMWARE_FLAGS | 0x03b00000; + WREG32(SOC15_REG_OFFSET(NBIO, 0, mmPCIE_INDEX2), reg); reg = RREG32(SOC15_REG_OFFSET(NBIO, 0, mmPCIE_DATA2)); - if ((reg & MP1_FIRMWARE_FLAGS__INTERRUPTS_ENABLED_MASK) >> - MP1_FIRMWARE_FLAGS__INTERRUPTS_ENABLED__SHIFT) - return true; - - return false; + return (reg & MP1_FIRMWARE_FLAGS__INTERRUPTS_ENABLED_MASK) ? true : false; } diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c b/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c index 2dd2b20d727e..21f38d882335 100644 --- a/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c +++ b/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c @@ -1039,6 +1039,7 @@ static void sdma_v4_0_ring_emit_pipeline_sync(struct amdgpu_ring *ring) static void sdma_v4_0_ring_emit_vm_flush(struct amdgpu_ring *ring, unsigned vm_id, uint64_t pd_addr) { + uint32_t req = ring->adev->gart.gart_funcs->get_invalidate_req(vm_id); unsigned eng = ring->idx; unsigned i; @@ -1048,7 +1049,6 @@ static void sdma_v4_0_ring_emit_vm_flush(struct amdgpu_ring *ring, for (i = 0; i < AMDGPU_MAX_VMHUBS; ++i) { struct amdgpu_vmhub *hub = &ring->adev->vmhub[i]; - uint32_t req = hub->get_invalidate_req(vm_id); amdgpu_ring_write(ring, SDMA_PKT_HEADER_OP(SDMA_OP_SRBM_WRITE) | SDMA_PKT_SRBM_WRITE_HEADER_BYTE_EN(0xf)); diff --git a/drivers/gpu/drm/amd/amdgpu/soc15.c b/drivers/gpu/drm/amd/amdgpu/soc15.c index bb14a45997b5..385de8617075 100644 --- a/drivers/gpu/drm/amd/amdgpu/soc15.c +++ b/drivers/gpu/drm/amd/amdgpu/soc15.c @@ -106,6 +106,8 @@ static u32 soc15_pcie_rreg(struct amdgpu_device *adev, u32 reg) if (adev->asic_type == CHIP_VEGA10) nbio_pcie_id = &nbio_v6_1_pcie_index_data; + else + BUG(); address = nbio_pcie_id->index_offset; data = nbio_pcie_id->data_offset; @@ -125,6 +127,8 @@ static void soc15_pcie_wreg(struct amdgpu_device *adev, u32 reg, u32 v) if (adev->asic_type == CHIP_VEGA10) nbio_pcie_id = &nbio_v6_1_pcie_index_data; + else + BUG(); address = nbio_pcie_id->index_offset; data = nbio_pcie_id->data_offset; @@ -493,7 +497,8 @@ int soc15_set_ip_blocks(struct amdgpu_device *adev) amdgpu_ip_block_add(adev, &mmhub_v1_0_ip_block); amdgpu_ip_block_add(adev, &gmc_v9_0_ip_block); amdgpu_ip_block_add(adev, &vega10_ih_ip_block); - amdgpu_ip_block_add(adev, &psp_v3_1_ip_block); + if (amdgpu_fw_load_type == 2 || amdgpu_fw_load_type == -1) + amdgpu_ip_block_add(adev, &psp_v3_1_ip_block); if (!amdgpu_sriov_vf(adev)) amdgpu_ip_block_add(adev, &amdgpu_pp_ip_block); if (adev->enable_virtual_display || amdgpu_sriov_vf(adev)) @@ -558,6 +563,7 @@ static int soc15_common_early_init(void *handle) if (amdgpu_sriov_vf(adev)) { amdgpu_virt_init_setting(adev); + xgpu_ai_mailbox_set_irq_funcs(adev); } /* @@ -610,8 +616,23 @@ static int soc15_common_early_init(void *handle) return 0; } +static int soc15_common_late_init(void *handle) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + if (amdgpu_sriov_vf(adev)) + xgpu_ai_mailbox_get_irq(adev); + + return 0; +} + static int soc15_common_sw_init(void *handle) { + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + if (amdgpu_sriov_vf(adev)) + xgpu_ai_mailbox_add_irq_id(adev); + return 0; } @@ -642,6 +663,8 @@ static int soc15_common_hw_fini(void *handle) /* disable the doorbell aperture */ soc15_enable_doorbell_aperture(adev, false); + if (amdgpu_sriov_vf(adev)) + xgpu_ai_mailbox_put_irq(adev); return 0; } @@ -855,7 +878,7 @@ static int soc15_common_set_powergating_state(void *handle, const struct amd_ip_funcs soc15_common_ip_funcs = { .name = "soc15_common", .early_init = soc15_common_early_init, - .late_init = NULL, + .late_init = soc15_common_late_init, .sw_init = soc15_common_sw_init, .sw_fini = soc15_common_sw_fini, .hw_init = soc15_common_hw_init, diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c b/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c index 9a4129d881aa..8ab0f78794a5 100644 --- a/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c +++ b/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c @@ -135,12 +135,9 @@ static int uvd_v4_2_sw_fini(void *handle) if (r) return r; - r = amdgpu_uvd_sw_fini(adev); - if (r) - return r; - - return r; + return amdgpu_uvd_sw_fini(adev); } + static void uvd_v4_2_enable_mgcg(struct amdgpu_device *adev, bool enable); /** @@ -230,11 +227,7 @@ static int uvd_v4_2_suspend(void *handle) if (r) return r; - r = amdgpu_uvd_suspend(adev); - if (r) - return r; - - return r; + return amdgpu_uvd_suspend(adev); } static int uvd_v4_2_resume(void *handle) @@ -246,11 +239,7 @@ static int uvd_v4_2_resume(void *handle) if (r) return r; - r = uvd_v4_2_hw_init(adev); - if (r) - return r; - - return r; + return uvd_v4_2_hw_init(adev); } /** diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c index e448f7d86bc0..bb6d46e168a3 100644 --- a/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c +++ b/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c @@ -131,11 +131,7 @@ static int uvd_v5_0_sw_fini(void *handle) if (r) return r; - r = amdgpu_uvd_sw_fini(adev); - if (r) - return r; - - return r; + return amdgpu_uvd_sw_fini(adev); } /** @@ -228,11 +224,7 @@ static int uvd_v5_0_suspend(void *handle) return r; uvd_v5_0_set_clockgating_state(adev, AMD_CG_STATE_GATE); - r = amdgpu_uvd_suspend(adev); - if (r) - return r; - - return r; + return amdgpu_uvd_suspend(adev); } static int uvd_v5_0_resume(void *handle) @@ -244,11 +236,7 @@ static int uvd_v5_0_resume(void *handle) if (r) return r; - r = uvd_v5_0_hw_init(adev); - if (r) - return r; - - return r; + return uvd_v5_0_hw_init(adev); } /** diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c index 5679a4249bd9..31db356476f8 100644 --- a/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c +++ b/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c @@ -134,11 +134,7 @@ static int uvd_v6_0_sw_fini(void *handle) if (r) return r; - r = amdgpu_uvd_sw_fini(adev); - if (r) - return r; - - return r; + return amdgpu_uvd_sw_fini(adev); } /** @@ -230,11 +226,8 @@ static int uvd_v6_0_suspend(void *handle) return r; /* Skip this for APU for now */ - if (!(adev->flags & AMD_IS_APU)) { + if (!(adev->flags & AMD_IS_APU)) r = amdgpu_uvd_suspend(adev); - if (r) - return r; - } return r; } @@ -250,11 +243,7 @@ static int uvd_v6_0_resume(void *handle) if (r) return r; } - r = uvd_v6_0_hw_init(adev); - if (r) - return r; - - return r; + return uvd_v6_0_hw_init(adev); } /** diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c index 13f52e0af9b8..9bcf01469282 100644 --- a/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c +++ b/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c @@ -438,11 +438,7 @@ static int uvd_v7_0_sw_fini(void *handle) for (i = 0; i < adev->uvd.num_enc_rings; ++i) amdgpu_ring_fini(&adev->uvd.ring_enc[i]); - r = amdgpu_uvd_sw_fini(adev); - if (r) - return r; - - return r; + return amdgpu_uvd_sw_fini(adev); } /** @@ -547,11 +543,8 @@ static int uvd_v7_0_suspend(void *handle) return r; /* Skip this for APU for now */ - if (!(adev->flags & AMD_IS_APU)) { + if (!(adev->flags & AMD_IS_APU)) r = amdgpu_uvd_suspend(adev); - if (r) - return r; - } return r; } @@ -567,11 +560,7 @@ static int uvd_v7_0_resume(void *handle) if (r) return r; } - r = uvd_v7_0_hw_init(adev); - if (r) - return r; - - return r; + return uvd_v7_0_hw_init(adev); } /** @@ -1045,6 +1034,7 @@ static void uvd_v7_0_vm_reg_wait(struct amdgpu_ring *ring, static void uvd_v7_0_ring_emit_vm_flush(struct amdgpu_ring *ring, unsigned vm_id, uint64_t pd_addr) { + uint32_t req = ring->adev->gart.gart_funcs->get_invalidate_req(vm_id); uint32_t data0, data1, mask; unsigned eng = ring->idx; unsigned i; @@ -1055,7 +1045,6 @@ static void uvd_v7_0_ring_emit_vm_flush(struct amdgpu_ring *ring, for (i = 0; i < AMDGPU_MAX_VMHUBS; ++i) { struct amdgpu_vmhub *hub = &ring->adev->vmhub[i]; - uint32_t req = hub->get_invalidate_req(vm_id); data0 = (hub->ctx0_ptb_addr_hi32 + vm_id * 2) << 2; data1 = upper_32_bits(pd_addr); @@ -1091,6 +1080,7 @@ static void uvd_v7_0_enc_ring_insert_end(struct amdgpu_ring *ring) static void uvd_v7_0_enc_ring_emit_vm_flush(struct amdgpu_ring *ring, unsigned int vm_id, uint64_t pd_addr) { + uint32_t req = ring->adev->gart.gart_funcs->get_invalidate_req(vm_id); unsigned eng = ring->idx; unsigned i; @@ -1100,7 +1090,6 @@ static void uvd_v7_0_enc_ring_emit_vm_flush(struct amdgpu_ring *ring, for (i = 0; i < AMDGPU_MAX_VMHUBS; ++i) { struct amdgpu_vmhub *hub = &ring->adev->vmhub[i]; - uint32_t req = hub->get_invalidate_req(vm_id); amdgpu_ring_write(ring, HEVC_ENC_CMD_REG_WRITE); amdgpu_ring_write(ring, diff --git a/drivers/gpu/drm/amd/amdgpu/vce_v2_0.c b/drivers/gpu/drm/amd/amdgpu/vce_v2_0.c index 49a6c45e65be..47f70827195b 100644 --- a/drivers/gpu/drm/amd/amdgpu/vce_v2_0.c +++ b/drivers/gpu/drm/amd/amdgpu/vce_v2_0.c @@ -451,11 +451,7 @@ static int vce_v2_0_sw_fini(void *handle) if (r) return r; - r = amdgpu_vce_sw_fini(adev); - if (r) - return r; - - return r; + return amdgpu_vce_sw_fini(adev); } static int vce_v2_0_hw_init(void *handle) @@ -495,11 +491,7 @@ static int vce_v2_0_suspend(void *handle) if (r) return r; - r = amdgpu_vce_suspend(adev); - if (r) - return r; - - return r; + return amdgpu_vce_suspend(adev); } static int vce_v2_0_resume(void *handle) @@ -511,11 +503,7 @@ static int vce_v2_0_resume(void *handle) if (r) return r; - r = vce_v2_0_hw_init(adev); - if (r) - return r; - - return r; + return vce_v2_0_hw_init(adev); } static int vce_v2_0_soft_reset(void *handle) diff --git a/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c b/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c index db0adac073c6..fb0819359909 100644 --- a/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c +++ b/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c @@ -417,11 +417,7 @@ static int vce_v3_0_sw_fini(void *handle) if (r) return r; - r = amdgpu_vce_sw_fini(adev); - if (r) - return r; - - return r; + return amdgpu_vce_sw_fini(adev); } static int vce_v3_0_hw_init(void *handle) @@ -471,11 +467,7 @@ static int vce_v3_0_suspend(void *handle) if (r) return r; - r = amdgpu_vce_suspend(adev); - if (r) - return r; - - return r; + return amdgpu_vce_suspend(adev); } static int vce_v3_0_resume(void *handle) @@ -487,11 +479,7 @@ static int vce_v3_0_resume(void *handle) if (r) return r; - r = vce_v3_0_hw_init(adev); - if (r) - return r; - - return r; + return vce_v3_0_hw_init(adev); } static void vce_v3_0_mc_resume(struct amdgpu_device *adev, int idx) diff --git a/drivers/gpu/drm/amd/amdgpu/vce_v4_0.c b/drivers/gpu/drm/amd/amdgpu/vce_v4_0.c index becc5f744a98..edde5fe938d6 100644 --- a/drivers/gpu/drm/amd/amdgpu/vce_v4_0.c +++ b/drivers/gpu/drm/amd/amdgpu/vce_v4_0.c @@ -527,11 +527,7 @@ static int vce_v4_0_sw_fini(void *handle) if (r) return r; - r = amdgpu_vce_sw_fini(adev); - if (r) - return r; - - return r; + return amdgpu_vce_sw_fini(adev); } static int vce_v4_0_hw_init(void *handle) @@ -584,11 +580,7 @@ static int vce_v4_0_suspend(void *handle) if (r) return r; - r = amdgpu_vce_suspend(adev); - if (r) - return r; - - return r; + return amdgpu_vce_suspend(adev); } static int vce_v4_0_resume(void *handle) @@ -600,11 +592,7 @@ static int vce_v4_0_resume(void *handle) if (r) return r; - r = vce_v4_0_hw_init(adev); - if (r) - return r; - - return r; + return vce_v4_0_hw_init(adev); } static void vce_v4_0_mc_resume(struct amdgpu_device *adev) @@ -985,6 +973,7 @@ static void vce_v4_0_ring_insert_end(struct amdgpu_ring *ring) static void vce_v4_0_emit_vm_flush(struct amdgpu_ring *ring, unsigned int vm_id, uint64_t pd_addr) { + uint32_t req = ring->adev->gart.gart_funcs->get_invalidate_req(vm_id); unsigned eng = ring->idx; unsigned i; @@ -994,7 +983,6 @@ static void vce_v4_0_emit_vm_flush(struct amdgpu_ring *ring, for (i = 0; i < AMDGPU_MAX_VMHUBS; ++i) { struct amdgpu_vmhub *hub = &ring->adev->vmhub[i]; - uint32_t req = hub->get_invalidate_req(vm_id); amdgpu_ring_write(ring, VCE_CMD_REG_WRITE); amdgpu_ring_write(ring, diff --git a/drivers/gpu/drm/amd/amdgpu/vid.h b/drivers/gpu/drm/amd/amdgpu/vid.h index b3a86e0e96e6..5f2ab9c1609a 100644 --- a/drivers/gpu/drm/amd/amdgpu/vid.h +++ b/drivers/gpu/drm/amd/amdgpu/vid.h @@ -362,7 +362,89 @@ #define PACKET3_WAIT_ON_DE_COUNTER_DIFF 0x88 #define PACKET3_SWITCH_BUFFER 0x8B #define PACKET3_SET_RESOURCES 0xA0 +/* 1. header + * 2. CONTROL + * 3. QUEUE_MASK_LO [31:0] + * 4. QUEUE_MASK_HI [31:0] + * 5. GWS_MASK_LO [31:0] + * 6. GWS_MASK_HI [31:0] + * 7. OAC_MASK [15:0] + * 8. GDS_HEAP_SIZE [16:11] | GDS_HEAP_BASE [5:0] + */ +# define PACKET3_SET_RESOURCES_VMID_MASK(x) ((x) << 0) +# define PACKET3_SET_RESOURCES_UNMAP_LATENTY(x) ((x) << 16) +# define PACKET3_SET_RESOURCES_QUEUE_TYPE(x) ((x) << 29) #define PACKET3_MAP_QUEUES 0xA2 +/* 1. header + * 2. CONTROL + * 3. CONTROL2 + * 4. MQD_ADDR_LO [31:0] + * 5. MQD_ADDR_HI [31:0] + * 6. WPTR_ADDR_LO [31:0] + * 7. WPTR_ADDR_HI [31:0] + */ +/* CONTROL */ +# define PACKET3_MAP_QUEUES_QUEUE_SEL(x) ((x) << 4) +# define PACKET3_MAP_QUEUES_VMID(x) ((x) << 8) +# define PACKET3_MAP_QUEUES_QUEUE_TYPE(x) ((x) << 21) +# define PACKET3_MAP_QUEUES_ALLOC_FORMAT(x) ((x) << 24) +# define PACKET3_MAP_QUEUES_ENGINE_SEL(x) ((x) << 26) +# define PACKET3_MAP_QUEUES_NUM_QUEUES(x) ((x) << 29) +/* CONTROL2 */ +# define PACKET3_MAP_QUEUES_CHECK_DISABLE(x) ((x) << 1) +# define PACKET3_MAP_QUEUES_DOORBELL_OFFSET(x) ((x) << 2) +# define PACKET3_MAP_QUEUES_QUEUE(x) ((x) << 26) +# define PACKET3_MAP_QUEUES_PIPE(x) ((x) << 29) +# define PACKET3_MAP_QUEUES_ME(x) ((x) << 31) +#define PACKET3_UNMAP_QUEUES 0xA3 +/* 1. header + * 2. CONTROL + * 3. CONTROL2 + * 4. CONTROL3 + * 5. CONTROL4 + * 6. CONTROL5 + */ +/* CONTROL */ +# define PACKET3_UNMAP_QUEUES_ACTION(x) ((x) << 0) + /* 0 - PREEMPT_QUEUES + * 1 - RESET_QUEUES + * 2 - DISABLE_PROCESS_QUEUES + * 3 - PREEMPT_QUEUES_NO_UNMAP + */ +# define PACKET3_UNMAP_QUEUES_QUEUE_SEL(x) ((x) << 4) +# define PACKET3_UNMAP_QUEUES_ENGINE_SEL(x) ((x) << 26) +# define PACKET3_UNMAP_QUEUES_NUM_QUEUES(x) ((x) << 29) +/* CONTROL2a */ +# define PACKET3_UNMAP_QUEUES_PASID(x) ((x) << 0) +/* CONTROL2b */ +# define PACKET3_UNMAP_QUEUES_DOORBELL_OFFSET0(x) ((x) << 2) +/* CONTROL3a */ +# define PACKET3_UNMAP_QUEUES_DOORBELL_OFFSET1(x) ((x) << 2) +/* CONTROL3b */ +# define PACKET3_UNMAP_QUEUES_RB_WPTR(x) ((x) << 0) +/* CONTROL4 */ +# define PACKET3_UNMAP_QUEUES_DOORBELL_OFFSET2(x) ((x) << 2) +/* CONTROL5 */ +# define PACKET3_UNMAP_QUEUES_DOORBELL_OFFSET3(x) ((x) << 2) +#define PACKET3_QUERY_STATUS 0xA4 +/* 1. header + * 2. CONTROL + * 3. CONTROL2 + * 4. ADDR_LO [31:0] + * 5. ADDR_HI [31:0] + * 6. DATA_LO [31:0] + * 7. DATA_HI [31:0] + */ +/* CONTROL */ +# define PACKET3_QUERY_STATUS_CONTEXT_ID(x) ((x) << 0) +# define PACKET3_QUERY_STATUS_INTERRUPT_SEL(x) ((x) << 28) +# define PACKET3_QUERY_STATUS_COMMAND(x) ((x) << 30) +/* CONTROL2a */ +# define PACKET3_QUERY_STATUS_PASID(x) ((x) << 0) +/* CONTROL2b */ +# define PACKET3_QUERY_STATUS_DOORBELL_OFFSET(x) ((x) << 2) +# define PACKET3_QUERY_STATUS_ENG_SEL(x) ((x) << 25) + #define VCE_CMD_NO_OP 0x00000000 #define VCE_CMD_END 0x00000001 diff --git a/drivers/gpu/drm/amd/powerplay/amd_powerplay.c b/drivers/gpu/drm/amd/powerplay/amd_powerplay.c index dfd4fe6f0578..9da5b0bb66d8 100644 --- a/drivers/gpu/drm/amd/powerplay/amd_powerplay.c +++ b/drivers/gpu/drm/amd/powerplay/amd_powerplay.c @@ -493,8 +493,10 @@ static int pp_dpm_dispatch_tasks(void *handle, enum amd_pp_event event_id, { enum amd_pm_state_type ps; - if (input == NULL) - return -EINVAL; + if (input == NULL) { + ret = -EINVAL; + break; + } ps = *(unsigned long *)input; data.requested_ui_label = power_state_convert(ps); @@ -539,15 +541,19 @@ static enum amd_pm_state_type pp_dpm_get_current_power_state(void *handle) switch (state->classification.ui_label) { case PP_StateUILabel_Battery: pm_type = POWER_STATE_TYPE_BATTERY; + break; case PP_StateUILabel_Balanced: pm_type = POWER_STATE_TYPE_BALANCED; + break; case PP_StateUILabel_Performance: pm_type = POWER_STATE_TYPE_PERFORMANCE; + break; default: if (state->classification.flags & PP_StateClassificationFlag_Boot) pm_type = POWER_STATE_TYPE_INTERNAL_BOOT; else pm_type = POWER_STATE_TYPE_DEFAULT; + break; } mutex_unlock(&pp_handle->pp_lock); @@ -894,7 +900,7 @@ static int pp_dpm_set_sclk_od(void *handle, uint32_t value) mutex_lock(&pp_handle->pp_lock); ret = hwmgr->hwmgr_func->set_sclk_od(hwmgr, value); - mutex_lock(&pp_handle->pp_lock); + mutex_unlock(&pp_handle->pp_lock); return ret; } diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_pptable.h b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_pptable.h index 8e53d3a5e725..6a907c93fd9c 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_pptable.h +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_pptable.h @@ -250,6 +250,29 @@ typedef struct _ATOM_Vega10_Fan_Table { USHORT usFanStartTemperature; } ATOM_Vega10_Fan_Table; +typedef struct _ATOM_Vega10_Fan_Table_V2 { + UCHAR ucRevId; + USHORT usFanOutputSensitivity; + USHORT usFanAcousticLimitRpm; + USHORT usThrottlingRPM; + USHORT usTargetTemperature; + USHORT usMinimumPWMLimit; + USHORT usTargetGfxClk; + USHORT usFanGainEdge; + USHORT usFanGainHotspot; + USHORT usFanGainLiquid; + USHORT usFanGainVrVddc; + USHORT usFanGainVrMvdd; + USHORT usFanGainPlx; + USHORT usFanGainHbm; + UCHAR ucEnableZeroRPM; + USHORT usFanStopTemperature; + USHORT usFanStartTemperature; + UCHAR ucFanParameters; + UCHAR ucFanMinRPM; + UCHAR ucFanMaxRPM; +} ATOM_Vega10_Fan_Table_V2; + typedef struct _ATOM_Vega10_Thermal_Controller { UCHAR ucRevId; UCHAR ucType; /* one of ATOM_VEGA10_PP_THERMALCONTROLLER_*/ @@ -305,6 +328,33 @@ typedef struct _ATOM_Vega10_PowerTune_Table { USHORT usTemperatureLimitTedge; } ATOM_Vega10_PowerTune_Table; +typedef struct _ATOM_Vega10_PowerTune_Table_V2 +{ + UCHAR ucRevId; + USHORT usSocketPowerLimit; + USHORT usBatteryPowerLimit; + USHORT usSmallPowerLimit; + USHORT usTdcLimit; + USHORT usEdcLimit; + USHORT usSoftwareShutdownTemp; + USHORT usTemperatureLimitHotSpot; + USHORT usTemperatureLimitLiquid1; + USHORT usTemperatureLimitLiquid2; + USHORT usTemperatureLimitHBM; + USHORT usTemperatureLimitVrSoc; + USHORT usTemperatureLimitVrMem; + USHORT usTemperatureLimitPlx; + USHORT usLoadLineResistance; + UCHAR ucLiquid1_I2C_address; + UCHAR ucLiquid2_I2C_address; + UCHAR ucLiquid_I2C_Line; + UCHAR ucVr_I2C_address; + UCHAR ucVr_I2C_Line; + UCHAR ucPlx_I2C_address; + UCHAR ucPlx_I2C_Line; + USHORT usTemperatureLimitTedge; +} ATOM_Vega10_PowerTune_Table_V2; + typedef struct _ATOM_Vega10_Hard_Limit_Record { ULONG ulSOCCLKLimit; ULONG ulGFXCLKLimit; diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_processpptables.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_processpptables.c index 518634f995e7..8b55ae01132d 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_processpptables.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_processpptables.c @@ -116,14 +116,16 @@ static int init_thermal_controller( const ATOM_Vega10_POWERPLAYTABLE *powerplay_table) { const ATOM_Vega10_Thermal_Controller *thermal_controller; - const ATOM_Vega10_Fan_Table *fan_table; + const Vega10_PPTable_Generic_SubTable_Header *header; + const ATOM_Vega10_Fan_Table *fan_table_v1; + const ATOM_Vega10_Fan_Table_V2 *fan_table_v2; thermal_controller = (ATOM_Vega10_Thermal_Controller *) (((unsigned long)powerplay_table) + le16_to_cpu(powerplay_table->usThermalControllerOffset)); PP_ASSERT_WITH_CODE((powerplay_table->usThermalControllerOffset != 0), - "Thermal controller table not set!", return -1); + "Thermal controller table not set!", return -EINVAL); hwmgr->thermal_controller.ucType = thermal_controller->ucType; hwmgr->thermal_controller.ucI2cLine = thermal_controller->ucI2cLine; @@ -142,6 +144,9 @@ static int init_thermal_controller( hwmgr->thermal_controller.fanInfo.ulMaxRPM = thermal_controller->ucFanMaxRPM * 100UL; + hwmgr->thermal_controller.advanceFanControlParameters.ulCycleDelay + = 100000; + set_hw_cap( hwmgr, ATOM_VEGA10_PP_THERMALCONTROLLER_NONE != hwmgr->thermal_controller.ucType, @@ -150,54 +155,101 @@ static int init_thermal_controller( if (!powerplay_table->usFanTableOffset) return 0; - fan_table = (const ATOM_Vega10_Fan_Table *) + header = (const Vega10_PPTable_Generic_SubTable_Header *) (((unsigned long)powerplay_table) + le16_to_cpu(powerplay_table->usFanTableOffset)); - PP_ASSERT_WITH_CODE((fan_table->ucRevId >= 8), - "Invalid Input Fan Table!", return -1); + if (header->ucRevId == 10) { + fan_table_v1 = (ATOM_Vega10_Fan_Table *)header; - hwmgr->thermal_controller.advanceFanControlParameters.ulCycleDelay - = 100000; - phm_cap_set(hwmgr->platform_descriptor.platformCaps, - PHM_PlatformCaps_MicrocodeFanControl); - - hwmgr->thermal_controller.advanceFanControlParameters.usFanOutputSensitivity = - le16_to_cpu(fan_table->usFanOutputSensitivity); - hwmgr->thermal_controller.advanceFanControlParameters.usMaxFanRPM = - le16_to_cpu(fan_table->usFanRPMMax); - hwmgr->thermal_controller.advanceFanControlParameters.usFanRPMMaxLimit = - le16_to_cpu(fan_table->usThrottlingRPM); - hwmgr->thermal_controller.advanceFanControlParameters.ulMinFanSCLKAcousticLimit = - le32_to_cpu((uint32_t)(fan_table->usFanAcousticLimit)); - hwmgr->thermal_controller.advanceFanControlParameters.usTMax = - le16_to_cpu(fan_table->usTargetTemperature); - hwmgr->thermal_controller.advanceFanControlParameters.usPWMMin = - le16_to_cpu(fan_table->usMinimumPWMLimit); - hwmgr->thermal_controller.advanceFanControlParameters.ulTargetGfxClk = - le32_to_cpu((uint32_t)(fan_table->usTargetGfxClk)); - hwmgr->thermal_controller.advanceFanControlParameters.usFanGainEdge = - le16_to_cpu(fan_table->usFanGainEdge); - hwmgr->thermal_controller.advanceFanControlParameters.usFanGainHotspot = - le16_to_cpu(fan_table->usFanGainHotspot); - hwmgr->thermal_controller.advanceFanControlParameters.usFanGainLiquid = - le16_to_cpu(fan_table->usFanGainLiquid); - hwmgr->thermal_controller.advanceFanControlParameters.usFanGainVrVddc = - le16_to_cpu(fan_table->usFanGainVrVddc); - hwmgr->thermal_controller.advanceFanControlParameters.usFanGainVrMvdd = - le16_to_cpu(fan_table->usFanGainVrMvdd); - hwmgr->thermal_controller.advanceFanControlParameters.usFanGainPlx = - le16_to_cpu(fan_table->usFanGainPlx); - hwmgr->thermal_controller.advanceFanControlParameters.usFanGainHbm = - le16_to_cpu(fan_table->usFanGainHbm); - - hwmgr->thermal_controller.advanceFanControlParameters.ucEnableZeroRPM = - fan_table->ucEnableZeroRPM; - hwmgr->thermal_controller.advanceFanControlParameters.usZeroRPMStopTemperature = - le16_to_cpu(fan_table->usFanStopTemperature); - hwmgr->thermal_controller.advanceFanControlParameters.usZeroRPMStartTemperature = - le16_to_cpu(fan_table->usFanStartTemperature); + PP_ASSERT_WITH_CODE((fan_table_v1->ucRevId >= 8), + "Invalid Input Fan Table!", return -EINVAL); + + phm_cap_set(hwmgr->platform_descriptor.platformCaps, + PHM_PlatformCaps_MicrocodeFanControl); + + hwmgr->thermal_controller.advanceFanControlParameters.usFanOutputSensitivity = + le16_to_cpu(fan_table_v1->usFanOutputSensitivity); + hwmgr->thermal_controller.advanceFanControlParameters.usMaxFanRPM = + le16_to_cpu(fan_table_v1->usFanRPMMax); + hwmgr->thermal_controller.advanceFanControlParameters.usFanRPMMaxLimit = + le16_to_cpu(fan_table_v1->usThrottlingRPM); + hwmgr->thermal_controller.advanceFanControlParameters.ulMinFanSCLKAcousticLimit = + le16_to_cpu(fan_table_v1->usFanAcousticLimit); + hwmgr->thermal_controller.advanceFanControlParameters.usTMax = + le16_to_cpu(fan_table_v1->usTargetTemperature); + hwmgr->thermal_controller.advanceFanControlParameters.usPWMMin = + le16_to_cpu(fan_table_v1->usMinimumPWMLimit); + hwmgr->thermal_controller.advanceFanControlParameters.ulTargetGfxClk = + le16_to_cpu(fan_table_v1->usTargetGfxClk); + hwmgr->thermal_controller.advanceFanControlParameters.usFanGainEdge = + le16_to_cpu(fan_table_v1->usFanGainEdge); + hwmgr->thermal_controller.advanceFanControlParameters.usFanGainHotspot = + le16_to_cpu(fan_table_v1->usFanGainHotspot); + hwmgr->thermal_controller.advanceFanControlParameters.usFanGainLiquid = + le16_to_cpu(fan_table_v1->usFanGainLiquid); + hwmgr->thermal_controller.advanceFanControlParameters.usFanGainVrVddc = + le16_to_cpu(fan_table_v1->usFanGainVrVddc); + hwmgr->thermal_controller.advanceFanControlParameters.usFanGainVrMvdd = + le16_to_cpu(fan_table_v1->usFanGainVrMvdd); + hwmgr->thermal_controller.advanceFanControlParameters.usFanGainPlx = + le16_to_cpu(fan_table_v1->usFanGainPlx); + hwmgr->thermal_controller.advanceFanControlParameters.usFanGainHbm = + le16_to_cpu(fan_table_v1->usFanGainHbm); + + hwmgr->thermal_controller.advanceFanControlParameters.ucEnableZeroRPM = + fan_table_v1->ucEnableZeroRPM; + hwmgr->thermal_controller.advanceFanControlParameters.usZeroRPMStopTemperature = + le16_to_cpu(fan_table_v1->usFanStopTemperature); + hwmgr->thermal_controller.advanceFanControlParameters.usZeroRPMStartTemperature = + le16_to_cpu(fan_table_v1->usFanStartTemperature); + } else if (header->ucRevId > 10) { + fan_table_v2 = (ATOM_Vega10_Fan_Table_V2 *)header; + + hwmgr->thermal_controller.fanInfo.ucTachometerPulsesPerRevolution = + fan_table_v2->ucFanParameters & ATOM_VEGA10_PP_FANPARAMETERS_TACHOMETER_PULSES_PER_REVOLUTION_MASK; + hwmgr->thermal_controller.fanInfo.ulMinRPM = fan_table_v2->ucFanMinRPM * 100UL; + hwmgr->thermal_controller.fanInfo.ulMaxRPM = fan_table_v2->ucFanMaxRPM * 100UL; + phm_cap_set(hwmgr->platform_descriptor.platformCaps, + PHM_PlatformCaps_MicrocodeFanControl); + + hwmgr->thermal_controller.advanceFanControlParameters.usFanOutputSensitivity = + le16_to_cpu(fan_table_v2->usFanOutputSensitivity); + hwmgr->thermal_controller.advanceFanControlParameters.usMaxFanRPM = + fan_table_v2->ucFanMaxRPM * 100UL; + hwmgr->thermal_controller.advanceFanControlParameters.usFanRPMMaxLimit = + le16_to_cpu(fan_table_v2->usThrottlingRPM); + hwmgr->thermal_controller.advanceFanControlParameters.ulMinFanSCLKAcousticLimit = + le16_to_cpu(fan_table_v2->usFanAcousticLimitRpm); + hwmgr->thermal_controller.advanceFanControlParameters.usTMax = + le16_to_cpu(fan_table_v2->usTargetTemperature); + hwmgr->thermal_controller.advanceFanControlParameters.usPWMMin = + le16_to_cpu(fan_table_v2->usMinimumPWMLimit); + hwmgr->thermal_controller.advanceFanControlParameters.ulTargetGfxClk = + le16_to_cpu(fan_table_v2->usTargetGfxClk); + hwmgr->thermal_controller.advanceFanControlParameters.usFanGainEdge = + le16_to_cpu(fan_table_v2->usFanGainEdge); + hwmgr->thermal_controller.advanceFanControlParameters.usFanGainHotspot = + le16_to_cpu(fan_table_v2->usFanGainHotspot); + hwmgr->thermal_controller.advanceFanControlParameters.usFanGainLiquid = + le16_to_cpu(fan_table_v2->usFanGainLiquid); + hwmgr->thermal_controller.advanceFanControlParameters.usFanGainVrVddc = + le16_to_cpu(fan_table_v2->usFanGainVrVddc); + hwmgr->thermal_controller.advanceFanControlParameters.usFanGainVrMvdd = + le16_to_cpu(fan_table_v2->usFanGainVrMvdd); + hwmgr->thermal_controller.advanceFanControlParameters.usFanGainPlx = + le16_to_cpu(fan_table_v2->usFanGainPlx); + hwmgr->thermal_controller.advanceFanControlParameters.usFanGainHbm = + le16_to_cpu(fan_table_v2->usFanGainHbm); + + hwmgr->thermal_controller.advanceFanControlParameters.ucEnableZeroRPM = + fan_table_v2->ucEnableZeroRPM; + hwmgr->thermal_controller.advanceFanControlParameters.usZeroRPMStopTemperature = + le16_to_cpu(fan_table_v2->usFanStopTemperature); + hwmgr->thermal_controller.advanceFanControlParameters.usZeroRPMStartTemperature = + le16_to_cpu(fan_table_v2->usFanStartTemperature); + } return 0; } @@ -261,6 +313,48 @@ static int get_mm_clock_voltage_table( return 0; } +static void get_scl_sda_value(uint8_t line, uint8_t *scl, uint8_t* sda) +{ + switch(line){ + case Vega10_I2CLineID_DDC1: + *scl = Vega10_I2C_DDC1CLK; + *sda = Vega10_I2C_DDC1DATA; + break; + case Vega10_I2CLineID_DDC2: + *scl = Vega10_I2C_DDC2CLK; + *sda = Vega10_I2C_DDC2DATA; + break; + case Vega10_I2CLineID_DDC3: + *scl = Vega10_I2C_DDC3CLK; + *sda = Vega10_I2C_DDC3DATA; + break; + case Vega10_I2CLineID_DDC4: + *scl = Vega10_I2C_DDC4CLK; + *sda = Vega10_I2C_DDC4DATA; + break; + case Vega10_I2CLineID_DDC5: + *scl = Vega10_I2C_DDC5CLK; + *sda = Vega10_I2C_DDC5DATA; + break; + case Vega10_I2CLineID_DDC6: + *scl = Vega10_I2C_DDC6CLK; + *sda = Vega10_I2C_DDC6DATA; + break; + case Vega10_I2CLineID_SCLSDA: + *scl = Vega10_I2C_SCL; + *sda = Vega10_I2C_SDA; + break; + case Vega10_I2CLineID_DDCVGA: + *scl = Vega10_I2C_DDCVGACLK; + *sda = Vega10_I2C_DDCVGADATA; + break; + default: + *scl = 0; + *sda = 0; + break; + } +} + static int get_tdp_table( struct pp_hwmgr *hwmgr, struct phm_tdp_table **info_tdp_table, @@ -268,59 +362,99 @@ static int get_tdp_table( { uint32_t table_size; struct phm_tdp_table *tdp_table; - - const ATOM_Vega10_PowerTune_Table *power_tune_table = - (ATOM_Vega10_PowerTune_Table *)table; - - table_size = sizeof(uint32_t) + sizeof(struct phm_cac_tdp_table); - hwmgr->dyn_state.cac_dtp_table = (struct phm_cac_tdp_table *) - kzalloc(table_size, GFP_KERNEL); - - if (!hwmgr->dyn_state.cac_dtp_table) - return -ENOMEM; + uint8_t scl; + uint8_t sda; + const ATOM_Vega10_PowerTune_Table *power_tune_table; + const ATOM_Vega10_PowerTune_Table_V2 *power_tune_table_v2; table_size = sizeof(uint32_t) + sizeof(struct phm_tdp_table); + tdp_table = kzalloc(table_size, GFP_KERNEL); - if (!tdp_table) { - kfree(hwmgr->dyn_state.cac_dtp_table); - hwmgr->dyn_state.cac_dtp_table = NULL; + if (!tdp_table) return -ENOMEM; - } - tdp_table->usMaximumPowerDeliveryLimit = le16_to_cpu(power_tune_table->usSocketPowerLimit); - tdp_table->usTDC = le16_to_cpu(power_tune_table->usTdcLimit); - tdp_table->usEDCLimit = le16_to_cpu(power_tune_table->usEdcLimit); - tdp_table->usSoftwareShutdownTemp = - le16_to_cpu(power_tune_table->usSoftwareShutdownTemp); - tdp_table->usTemperatureLimitTedge = - le16_to_cpu(power_tune_table->usTemperatureLimitTedge); - tdp_table->usTemperatureLimitHotspot = - le16_to_cpu(power_tune_table->usTemperatureLimitHotSpot); - tdp_table->usTemperatureLimitLiquid1 = - le16_to_cpu(power_tune_table->usTemperatureLimitLiquid1); - tdp_table->usTemperatureLimitLiquid2 = - le16_to_cpu(power_tune_table->usTemperatureLimitLiquid2); - tdp_table->usTemperatureLimitHBM = - le16_to_cpu(power_tune_table->usTemperatureLimitHBM); - tdp_table->usTemperatureLimitVrVddc = - le16_to_cpu(power_tune_table->usTemperatureLimitVrSoc); - tdp_table->usTemperatureLimitVrMvdd = - le16_to_cpu(power_tune_table->usTemperatureLimitVrMem); - tdp_table->usTemperatureLimitPlx = - le16_to_cpu(power_tune_table->usTemperatureLimitPlx); - tdp_table->ucLiquid1_I2C_address = power_tune_table->ucLiquid1_I2C_address; - tdp_table->ucLiquid2_I2C_address = power_tune_table->ucLiquid2_I2C_address; - tdp_table->ucLiquid_I2C_Line = power_tune_table->ucLiquid_I2C_LineSCL; - tdp_table->ucLiquid_I2C_LineSDA = power_tune_table->ucLiquid_I2C_LineSDA; - tdp_table->ucVr_I2C_address = power_tune_table->ucVr_I2C_address; - tdp_table->ucVr_I2C_Line = power_tune_table->ucVr_I2C_LineSCL; - tdp_table->ucVr_I2C_LineSDA = power_tune_table->ucVr_I2C_LineSDA; - tdp_table->ucPlx_I2C_address = power_tune_table->ucPlx_I2C_address; - tdp_table->ucPlx_I2C_Line = power_tune_table->ucPlx_I2C_LineSCL; - tdp_table->ucPlx_I2C_LineSDA = power_tune_table->ucPlx_I2C_LineSDA; - - hwmgr->platform_descriptor.LoadLineSlope = power_tune_table->usLoadLineResistance; + if (table->ucRevId == 5) { + power_tune_table = (ATOM_Vega10_PowerTune_Table *)table; + tdp_table->usMaximumPowerDeliveryLimit = le16_to_cpu(power_tune_table->usSocketPowerLimit); + tdp_table->usTDC = le16_to_cpu(power_tune_table->usTdcLimit); + tdp_table->usEDCLimit = le16_to_cpu(power_tune_table->usEdcLimit); + tdp_table->usSoftwareShutdownTemp = + le16_to_cpu(power_tune_table->usSoftwareShutdownTemp); + tdp_table->usTemperatureLimitTedge = + le16_to_cpu(power_tune_table->usTemperatureLimitTedge); + tdp_table->usTemperatureLimitHotspot = + le16_to_cpu(power_tune_table->usTemperatureLimitHotSpot); + tdp_table->usTemperatureLimitLiquid1 = + le16_to_cpu(power_tune_table->usTemperatureLimitLiquid1); + tdp_table->usTemperatureLimitLiquid2 = + le16_to_cpu(power_tune_table->usTemperatureLimitLiquid2); + tdp_table->usTemperatureLimitHBM = + le16_to_cpu(power_tune_table->usTemperatureLimitHBM); + tdp_table->usTemperatureLimitVrVddc = + le16_to_cpu(power_tune_table->usTemperatureLimitVrSoc); + tdp_table->usTemperatureLimitVrMvdd = + le16_to_cpu(power_tune_table->usTemperatureLimitVrMem); + tdp_table->usTemperatureLimitPlx = + le16_to_cpu(power_tune_table->usTemperatureLimitPlx); + tdp_table->ucLiquid1_I2C_address = power_tune_table->ucLiquid1_I2C_address; + tdp_table->ucLiquid2_I2C_address = power_tune_table->ucLiquid2_I2C_address; + tdp_table->ucLiquid_I2C_Line = power_tune_table->ucLiquid_I2C_LineSCL; + tdp_table->ucLiquid_I2C_LineSDA = power_tune_table->ucLiquid_I2C_LineSDA; + tdp_table->ucVr_I2C_address = power_tune_table->ucVr_I2C_address; + tdp_table->ucVr_I2C_Line = power_tune_table->ucVr_I2C_LineSCL; + tdp_table->ucVr_I2C_LineSDA = power_tune_table->ucVr_I2C_LineSDA; + tdp_table->ucPlx_I2C_address = power_tune_table->ucPlx_I2C_address; + tdp_table->ucPlx_I2C_Line = power_tune_table->ucPlx_I2C_LineSCL; + tdp_table->ucPlx_I2C_LineSDA = power_tune_table->ucPlx_I2C_LineSDA; + hwmgr->platform_descriptor.LoadLineSlope = power_tune_table->usLoadLineResistance; + } else { + power_tune_table_v2 = (ATOM_Vega10_PowerTune_Table_V2 *)table; + tdp_table->usMaximumPowerDeliveryLimit = le16_to_cpu(power_tune_table_v2->usSocketPowerLimit); + tdp_table->usTDC = le16_to_cpu(power_tune_table_v2->usTdcLimit); + tdp_table->usEDCLimit = le16_to_cpu(power_tune_table_v2->usEdcLimit); + tdp_table->usSoftwareShutdownTemp = + le16_to_cpu(power_tune_table_v2->usSoftwareShutdownTemp); + tdp_table->usTemperatureLimitTedge = + le16_to_cpu(power_tune_table_v2->usTemperatureLimitTedge); + tdp_table->usTemperatureLimitHotspot = + le16_to_cpu(power_tune_table_v2->usTemperatureLimitHotSpot); + tdp_table->usTemperatureLimitLiquid1 = + le16_to_cpu(power_tune_table_v2->usTemperatureLimitLiquid1); + tdp_table->usTemperatureLimitLiquid2 = + le16_to_cpu(power_tune_table_v2->usTemperatureLimitLiquid2); + tdp_table->usTemperatureLimitHBM = + le16_to_cpu(power_tune_table_v2->usTemperatureLimitHBM); + tdp_table->usTemperatureLimitVrVddc = + le16_to_cpu(power_tune_table_v2->usTemperatureLimitVrSoc); + tdp_table->usTemperatureLimitVrMvdd = + le16_to_cpu(power_tune_table_v2->usTemperatureLimitVrMem); + tdp_table->usTemperatureLimitPlx = + le16_to_cpu(power_tune_table_v2->usTemperatureLimitPlx); + tdp_table->ucLiquid1_I2C_address = power_tune_table_v2->ucLiquid1_I2C_address; + tdp_table->ucLiquid2_I2C_address = power_tune_table_v2->ucLiquid2_I2C_address; + + get_scl_sda_value(power_tune_table_v2->ucLiquid_I2C_Line, &scl, &sda); + + tdp_table->ucLiquid_I2C_Line = scl; + tdp_table->ucLiquid_I2C_LineSDA = sda; + + tdp_table->ucVr_I2C_address = power_tune_table_v2->ucVr_I2C_address; + + get_scl_sda_value(power_tune_table_v2->ucVr_I2C_Line, &scl, &sda); + + tdp_table->ucVr_I2C_Line = scl; + tdp_table->ucVr_I2C_LineSDA = sda; + tdp_table->ucPlx_I2C_address = power_tune_table_v2->ucPlx_I2C_address; + + get_scl_sda_value(power_tune_table_v2->ucPlx_I2C_Line, &scl, &sda); + + tdp_table->ucPlx_I2C_Line = scl; + tdp_table->ucPlx_I2C_LineSDA = sda; + + hwmgr->platform_descriptor.LoadLineSlope = + power_tune_table_v2->usLoadLineResistance; + } *info_tdp_table = tdp_table; @@ -836,7 +970,7 @@ static int init_dpm_2_parameters( (((unsigned long)powerplay_table) + le16_to_cpu(powerplay_table->usVddcLookupTableOffset)); result = get_vddc_lookup_table(hwmgr, - &pp_table_info->vddc_lookup_table, vddc_table, 16); + &pp_table_info->vddc_lookup_table, vddc_table, 8); } if (powerplay_table->usVddmemLookupTableOffset) { @@ -845,7 +979,7 @@ static int init_dpm_2_parameters( (((unsigned long)powerplay_table) + le16_to_cpu(powerplay_table->usVddmemLookupTableOffset)); result = get_vddc_lookup_table(hwmgr, - &pp_table_info->vddmem_lookup_table, vdd_mem_table, 16); + &pp_table_info->vddmem_lookup_table, vdd_mem_table, 4); } if (powerplay_table->usVddciLookupTableOffset) { @@ -854,7 +988,7 @@ static int init_dpm_2_parameters( (((unsigned long)powerplay_table) + le16_to_cpu(powerplay_table->usVddciLookupTableOffset)); result = get_vddc_lookup_table(hwmgr, - &pp_table_info->vddci_lookup_table, vddci_table, 16); + &pp_table_info->vddci_lookup_table, vddci_table, 4); } return result; diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_processpptables.h b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_processpptables.h index 995d133ba6aa..d83ed2af7aa3 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_processpptables.h +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_processpptables.h @@ -26,6 +26,34 @@ #include "hwmgr.h" +enum Vega10_I2CLineID { + Vega10_I2CLineID_DDC1 = 0x90, + Vega10_I2CLineID_DDC2 = 0x91, + Vega10_I2CLineID_DDC3 = 0x92, + Vega10_I2CLineID_DDC4 = 0x93, + Vega10_I2CLineID_DDC5 = 0x94, + Vega10_I2CLineID_DDC6 = 0x95, + Vega10_I2CLineID_SCLSDA = 0x96, + Vega10_I2CLineID_DDCVGA = 0x97 +}; + +#define Vega10_I2C_DDC1DATA 0 +#define Vega10_I2C_DDC1CLK 1 +#define Vega10_I2C_DDC2DATA 2 +#define Vega10_I2C_DDC2CLK 3 +#define Vega10_I2C_DDC3DATA 4 +#define Vega10_I2C_DDC3CLK 5 +#define Vega10_I2C_SDA 40 +#define Vega10_I2C_SCL 41 +#define Vega10_I2C_DDC4DATA 65 +#define Vega10_I2C_DDC4CLK 66 +#define Vega10_I2C_DDC5DATA 0x48 +#define Vega10_I2C_DDC5CLK 0x49 +#define Vega10_I2C_DDC6DATA 0x4a +#define Vega10_I2C_DDC6CLK 0x4b +#define Vega10_I2C_DDCVGADATA 0x4c +#define Vega10_I2C_DDCVGACLK 0x4d + extern const struct pp_table_func vega10_pptable_funcs; extern int vega10_get_number_of_powerplay_table_entries(struct pp_hwmgr *hwmgr); extern int vega10_get_powerplay_table_entry(struct pp_hwmgr *hwmgr, uint32_t entry_index, 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 aee021451d35..2037910adcb1 100644 --- a/drivers/gpu/drm/amd/powerplay/inc/smu9_driver_if.h +++ b/drivers/gpu/drm/amd/powerplay/inc/smu9_driver_if.h @@ -30,7 +30,9 @@ * SMU TEAM: Always increment the interface version if * any structure is changed in this file */ -#define SMU9_DRIVER_IF_VERSION 0xa +#define SMU9_DRIVER_IF_VERSION 0xB + +#define PPTABLE_V10_SMU_VERSION 1 #define NUM_GFXCLK_DPM_LEVELS 8 #define NUM_UVD_DPM_LEVELS 8 @@ -87,6 +89,11 @@ typedef struct { int32_t a0; int32_t a1; int32_t a2; + + uint8_t a0_shift; + uint8_t a1_shift; + uint8_t a2_shift; + uint8_t padding; } GbVdroopTable_t; typedef struct { @@ -293,7 +300,9 @@ typedef struct { uint16_t Platform_sigma; uint16_t PSM_Age_CompFactor; - uint32_t Reserved[20]; + uint32_t DpmLevelPowerDelta; + + uint32_t Reserved[19]; /* Padding - ignore */ uint32_t MmHubPadding[7]; /* SMU internal use */ @@ -350,8 +359,8 @@ typedef struct { typedef struct { uint16_t avgPsmCount[30]; uint16_t minPsmCount[30]; - uint16_t avgPsmVoltage[30]; /* in mV with 2 fractional bits */ - uint16_t minPsmVoltage[30]; /* in mV with 2 fractional bits */ + float avgPsmVoltage[30]; + float minPsmVoltage[30]; uint32_t MmHubPadding[7]; /* SMU internal use */ } AvfsDebugTable_t; @@ -414,5 +423,45 @@ typedef struct { #define UCLK_SWITCH_SLOW 0 #define UCLK_SWITCH_FAST 1 +/* GFX DIDT Configuration */ +#define SQ_Enable_MASK 0x1 +#define SQ_IR_MASK 0x2 +#define SQ_PCC_MASK 0x4 +#define SQ_EDC_MASK 0x8 + +#define TCP_Enable_MASK 0x100 +#define TCP_IR_MASK 0x200 +#define TCP_PCC_MASK 0x400 +#define TCP_EDC_MASK 0x800 + +#define TD_Enable_MASK 0x10000 +#define TD_IR_MASK 0x20000 +#define TD_PCC_MASK 0x40000 +#define TD_EDC_MASK 0x80000 + +#define DB_Enable_MASK 0x1000000 +#define DB_IR_MASK 0x2000000 +#define DB_PCC_MASK 0x4000000 +#define DB_EDC_MASK 0x8000000 + +#define SQ_Enable_SHIFT 0 +#define SQ_IR_SHIFT 1 +#define SQ_PCC_SHIFT 2 +#define SQ_EDC_SHIFT 3 + +#define TCP_Enable_SHIFT 8 +#define TCP_IR_SHIFT 9 +#define TCP_PCC_SHIFT 10 +#define TCP_EDC_SHIFT 11 + +#define TD_Enable_SHIFT 16 +#define TD_IR_SHIFT 17 +#define TD_PCC_SHIFT 18 +#define TD_EDC_SHIFT 19 + +#define DB_Enable_SHIFT 24 +#define DB_IR_SHIFT 25 +#define DB_PCC_SHIFT 26 +#define DB_EDC_SHIFT 27 #endif diff --git a/drivers/gpu/drm/ast/ast_ttm.c b/drivers/gpu/drm/ast/ast_ttm.c index 50c910efa13d..e879496b8a42 100644 --- a/drivers/gpu/drm/ast/ast_ttm.c +++ b/drivers/gpu/drm/ast/ast_ttm.c @@ -236,6 +236,7 @@ struct ttm_bo_driver ast_bo_driver = { .verify_access = ast_bo_verify_access, .io_mem_reserve = &ast_ttm_io_mem_reserve, .io_mem_free = &ast_ttm_io_mem_free, + .io_mem_pfn = ttm_bo_default_io_mem_pfn, }; int ast_mm_init(struct ast_private *ast) diff --git a/drivers/gpu/drm/bochs/bochs_mm.c b/drivers/gpu/drm/bochs/bochs_mm.c index 857755ac2d70..c4cadb638460 100644 --- a/drivers/gpu/drm/bochs/bochs_mm.c +++ b/drivers/gpu/drm/bochs/bochs_mm.c @@ -205,6 +205,7 @@ struct ttm_bo_driver bochs_bo_driver = { .verify_access = bochs_bo_verify_access, .io_mem_reserve = &bochs_ttm_io_mem_reserve, .io_mem_free = &bochs_ttm_io_mem_free, + .io_mem_pfn = ttm_bo_default_io_mem_pfn, }; int bochs_mm_init(struct bochs_device *bochs) diff --git a/drivers/gpu/drm/cirrus/cirrus_ttm.c b/drivers/gpu/drm/cirrus/cirrus_ttm.c index f53aa8f4a143..93dbcd38355d 100644 --- a/drivers/gpu/drm/cirrus/cirrus_ttm.c +++ b/drivers/gpu/drm/cirrus/cirrus_ttm.c @@ -236,6 +236,7 @@ struct ttm_bo_driver cirrus_bo_driver = { .verify_access = cirrus_bo_verify_access, .io_mem_reserve = &cirrus_ttm_io_mem_reserve, .io_mem_free = &cirrus_ttm_io_mem_free, + .io_mem_pfn = ttm_bo_default_io_mem_pfn, }; int cirrus_mm_init(struct cirrus_device *cirrus) diff --git a/drivers/gpu/drm/etnaviv/Kconfig b/drivers/gpu/drm/etnaviv/Kconfig index cc1731c5289c..71cee4e9fefb 100644 --- a/drivers/gpu/drm/etnaviv/Kconfig +++ b/drivers/gpu/drm/etnaviv/Kconfig @@ -5,6 +5,7 @@ config DRM_ETNAVIV depends on ARCH_MXC || ARCH_DOVE || (ARM && COMPILE_TEST) depends on MMU select SHMEM + select SYNC_FILE select TMPFS select IOMMU_API select IOMMU_SUPPORT diff --git a/drivers/gpu/drm/etnaviv/etnaviv_drv.c b/drivers/gpu/drm/etnaviv/etnaviv_drv.c index 587e45043542..5255278dde56 100644 --- a/drivers/gpu/drm/etnaviv/etnaviv_drv.c +++ b/drivers/gpu/drm/etnaviv/etnaviv_drv.c @@ -111,7 +111,7 @@ static int etnaviv_open(struct drm_device *dev, struct drm_file *file) return 0; } -static void etnaviv_preclose(struct drm_device *dev, struct drm_file *file) +static void etnaviv_postclose(struct drm_device *dev, struct drm_file *file) { struct etnaviv_drm_private *priv = dev->dev_private; struct etnaviv_file_private *ctx = file->driver_priv; @@ -488,7 +488,7 @@ static struct drm_driver etnaviv_drm_driver = { DRIVER_PRIME | DRIVER_RENDER, .open = etnaviv_open, - .preclose = etnaviv_preclose, + .postclose = etnaviv_postclose, .gem_free_object_unlocked = etnaviv_gem_free_object, .gem_vm_ops = &vm_ops, .prime_handle_to_fd = drm_gem_prime_handle_to_fd, @@ -512,7 +512,7 @@ static struct drm_driver etnaviv_drm_driver = { .desc = "etnaviv DRM", .date = "20151214", .major = 1, - .minor = 0, + .minor = 1, }; /* diff --git a/drivers/gpu/drm/etnaviv/etnaviv_gem.h b/drivers/gpu/drm/etnaviv/etnaviv_gem.h index e63ff116a3b3..c4a091e87426 100644 --- a/drivers/gpu/drm/etnaviv/etnaviv_gem.h +++ b/drivers/gpu/drm/etnaviv/etnaviv_gem.h @@ -20,6 +20,7 @@ #include <linux/reservation.h> #include "etnaviv_drv.h" +struct dma_fence; struct etnaviv_gem_ops; struct etnaviv_gem_object; @@ -104,9 +105,10 @@ struct etnaviv_gem_submit { struct drm_device *dev; struct etnaviv_gpu *gpu; struct ww_acquire_ctx ticket; - u32 fence; + struct dma_fence *fence; unsigned int nr_bos; struct etnaviv_gem_submit_bo bos[0]; + u32 flags; }; int etnaviv_gem_wait_bo(struct etnaviv_gpu *gpu, struct drm_gem_object *obj, diff --git a/drivers/gpu/drm/etnaviv/etnaviv_gem_submit.c b/drivers/gpu/drm/etnaviv/etnaviv_gem_submit.c index 726090d7a6ac..e1909429837e 100644 --- a/drivers/gpu/drm/etnaviv/etnaviv_gem_submit.c +++ b/drivers/gpu/drm/etnaviv/etnaviv_gem_submit.c @@ -14,7 +14,9 @@ * this program. If not, see <http://www.gnu.org/licenses/>. */ +#include <linux/dma-fence-array.h> #include <linux/reservation.h> +#include <linux/sync_file.h> #include "etnaviv_cmdbuf.h" #include "etnaviv_drv.h" #include "etnaviv_gpu.h" @@ -169,8 +171,10 @@ static int submit_fence_sync(const struct etnaviv_gem_submit *submit) for (i = 0; i < submit->nr_bos; i++) { struct etnaviv_gem_object *etnaviv_obj = submit->bos[i].obj; bool write = submit->bos[i].flags & ETNA_SUBMIT_BO_WRITE; + bool explicit = !(submit->flags & ETNA_SUBMIT_NO_IMPLICIT); - ret = etnaviv_gpu_fence_sync_obj(etnaviv_obj, context, write); + ret = etnaviv_gpu_fence_sync_obj(etnaviv_obj, context, write, + explicit); if (ret) break; } @@ -290,6 +294,7 @@ static void submit_cleanup(struct etnaviv_gem_submit *submit) } ww_acquire_fini(&submit->ticket); + dma_fence_put(submit->fence); kfree(submit); } @@ -303,6 +308,9 @@ int etnaviv_ioctl_gem_submit(struct drm_device *dev, void *data, struct etnaviv_gem_submit *submit; struct etnaviv_cmdbuf *cmdbuf; struct etnaviv_gpu *gpu; + struct dma_fence *in_fence = NULL; + struct sync_file *sync_file = NULL; + int out_fence_fd = -1; void *stream; int ret; @@ -326,6 +334,11 @@ int etnaviv_ioctl_gem_submit(struct drm_device *dev, void *data, return -EINVAL; } + if (args->flags & ~ETNA_SUBMIT_FLAGS) { + DRM_ERROR("invalid flags: 0x%x\n", args->flags); + return -EINVAL; + } + /* * Copy the command submission and bo array to kernel space in * one go, and do this outside of any locks. @@ -365,12 +378,22 @@ int etnaviv_ioctl_gem_submit(struct drm_device *dev, void *data, goto err_submit_cmds; } + if (args->flags & ETNA_SUBMIT_FENCE_FD_OUT) { + out_fence_fd = get_unused_fd_flags(O_CLOEXEC); + if (out_fence_fd < 0) { + ret = out_fence_fd; + goto err_submit_cmds; + } + } + submit = submit_create(dev, gpu, args->nr_bos); if (!submit) { ret = -ENOMEM; goto err_submit_cmds; } + submit->flags = args->flags; + ret = submit_lookup_objects(submit, file, bos, args->nr_bos); if (ret) goto err_submit_objects; @@ -385,6 +408,24 @@ int etnaviv_ioctl_gem_submit(struct drm_device *dev, void *data, goto err_submit_objects; } + if (args->flags & ETNA_SUBMIT_FENCE_FD_IN) { + in_fence = sync_file_get_fence(args->fence_fd); + if (!in_fence) { + ret = -EINVAL; + goto err_submit_objects; + } + + /* + * Wait if the fence is from a foreign context, or if the fence + * array contains any fence from a foreign context. + */ + if (!dma_fence_match_context(in_fence, gpu->fence_context)) { + ret = dma_fence_wait(in_fence, true); + if (ret) + goto err_submit_objects; + } + } + ret = submit_fence_sync(submit); if (ret) goto err_submit_objects; @@ -405,7 +446,23 @@ int etnaviv_ioctl_gem_submit(struct drm_device *dev, void *data, if (ret == 0) cmdbuf = NULL; - args->fence = submit->fence; + if (args->flags & ETNA_SUBMIT_FENCE_FD_OUT) { + /* + * This can be improved: ideally we want to allocate the sync + * file before kicking off the GPU job and just attach the + * fence to the sync file here, eliminating the ENOMEM + * possibility at this stage. + */ + sync_file = sync_file_create(submit->fence); + if (!sync_file) { + ret = -ENOMEM; + goto out; + } + fd_install(out_fence_fd, sync_file->file); + } + + args->fence_fd = out_fence_fd; + args->fence = submit->fence->seqno; out: submit_unpin_objects(submit); @@ -419,9 +476,13 @@ out: flush_workqueue(priv->wq); err_submit_objects: + if (in_fence) + dma_fence_put(in_fence); submit_cleanup(submit); err_submit_cmds: + if (ret && (out_fence_fd >= 0)) + put_unused_fd(out_fence_fd); /* if we still own the cmdbuf */ if (cmdbuf) etnaviv_cmdbuf_free(cmdbuf); diff --git a/drivers/gpu/drm/etnaviv/etnaviv_gpu.c b/drivers/gpu/drm/etnaviv/etnaviv_gpu.c index da48819ff2e6..4f587058a3aa 100644 --- a/drivers/gpu/drm/etnaviv/etnaviv_gpu.c +++ b/drivers/gpu/drm/etnaviv/etnaviv_gpu.c @@ -18,6 +18,7 @@ #include <linux/dma-fence.h> #include <linux/moduleparam.h> #include <linux/of_device.h> +#include <linux/thermal.h> #include "etnaviv_cmdbuf.h" #include "etnaviv_dump.h" @@ -409,6 +410,17 @@ static void etnaviv_gpu_load_clock(struct etnaviv_gpu *gpu, u32 clock) gpu_write(gpu, VIVS_HI_CLOCK_CONTROL, clock); } +static void etnaviv_gpu_update_clock(struct etnaviv_gpu *gpu) +{ + unsigned int fscale = 1 << (6 - gpu->freq_scale); + u32 clock; + + clock = VIVS_HI_CLOCK_CONTROL_DISABLE_DEBUG_REGISTERS | + VIVS_HI_CLOCK_CONTROL_FSCALE_VAL(fscale); + + etnaviv_gpu_load_clock(gpu, clock); +} + static int etnaviv_hw_reset(struct etnaviv_gpu *gpu) { u32 control, idle; @@ -426,11 +438,10 @@ static int etnaviv_hw_reset(struct etnaviv_gpu *gpu) timeout = jiffies + msecs_to_jiffies(1000); while (time_is_after_jiffies(timeout)) { - control = VIVS_HI_CLOCK_CONTROL_DISABLE_DEBUG_REGISTERS | - VIVS_HI_CLOCK_CONTROL_FSCALE_VAL(0x40); - /* enable clock */ - etnaviv_gpu_load_clock(gpu, control); + etnaviv_gpu_update_clock(gpu); + + control = gpu_read(gpu, VIVS_HI_CLOCK_CONTROL); /* Wait for stable clock. Vivante's code waited for 1ms */ usleep_range(1000, 10000); @@ -490,11 +501,7 @@ static int etnaviv_hw_reset(struct etnaviv_gpu *gpu) } /* We rely on the GPU running, so program the clock */ - control = VIVS_HI_CLOCK_CONTROL_DISABLE_DEBUG_REGISTERS | - VIVS_HI_CLOCK_CONTROL_FSCALE_VAL(0x40); - - /* enable clock */ - etnaviv_gpu_load_clock(gpu, control); + etnaviv_gpu_update_clock(gpu); return 0; } @@ -1051,6 +1058,12 @@ static struct dma_fence *etnaviv_gpu_fence_alloc(struct etnaviv_gpu *gpu) { struct etnaviv_fence *f; + /* + * GPU lock must already be held, otherwise fence completion order might + * not match the seqno order assigned here. + */ + lockdep_assert_held(&gpu->lock); + f = kzalloc(sizeof(*f), GFP_KERNEL); if (!f) return NULL; @@ -1064,7 +1077,7 @@ static struct dma_fence *etnaviv_gpu_fence_alloc(struct etnaviv_gpu *gpu) } int etnaviv_gpu_fence_sync_obj(struct etnaviv_gem_object *etnaviv_obj, - unsigned int context, bool exclusive) + unsigned int context, bool exclusive, bool explicit) { struct reservation_object *robj = etnaviv_obj->resv; struct reservation_object_list *fobj; @@ -1077,6 +1090,9 @@ int etnaviv_gpu_fence_sync_obj(struct etnaviv_gem_object *etnaviv_obj, return ret; } + if (explicit) + return 0; + /* * If we have any shared fences, then the exclusive fence * should be ignored as it will already have been signalled. @@ -1321,8 +1337,8 @@ int etnaviv_gpu_submit(struct etnaviv_gpu *gpu, } gpu->event[event].fence = fence; - submit->fence = fence->seqno; - gpu->active_fence = submit->fence; + submit->fence = dma_fence_get(fence); + gpu->active_fence = submit->fence->seqno; if (gpu->lastctx != cmdbuf->ctx) { gpu->mmu->need_flush = true; @@ -1526,17 +1542,13 @@ static int etnaviv_gpu_hw_suspend(struct etnaviv_gpu *gpu) #ifdef CONFIG_PM static int etnaviv_gpu_hw_resume(struct etnaviv_gpu *gpu) { - u32 clock; int ret; ret = mutex_lock_killable(&gpu->lock); if (ret) return ret; - clock = VIVS_HI_CLOCK_CONTROL_DISABLE_DEBUG_REGISTERS | - VIVS_HI_CLOCK_CONTROL_FSCALE_VAL(0x40); - - etnaviv_gpu_load_clock(gpu, clock); + etnaviv_gpu_update_clock(gpu); etnaviv_gpu_hw_init(gpu); gpu->switch_context = true; @@ -1548,6 +1560,47 @@ static int etnaviv_gpu_hw_resume(struct etnaviv_gpu *gpu) } #endif +static int +etnaviv_gpu_cooling_get_max_state(struct thermal_cooling_device *cdev, + unsigned long *state) +{ + *state = 6; + + return 0; +} + +static int +etnaviv_gpu_cooling_get_cur_state(struct thermal_cooling_device *cdev, + unsigned long *state) +{ + struct etnaviv_gpu *gpu = cdev->devdata; + + *state = gpu->freq_scale; + + return 0; +} + +static int +etnaviv_gpu_cooling_set_cur_state(struct thermal_cooling_device *cdev, + unsigned long state) +{ + struct etnaviv_gpu *gpu = cdev->devdata; + + mutex_lock(&gpu->lock); + gpu->freq_scale = state; + if (!pm_runtime_suspended(gpu->dev)) + etnaviv_gpu_update_clock(gpu); + mutex_unlock(&gpu->lock); + + return 0; +} + +static struct thermal_cooling_device_ops cooling_ops = { + .get_max_state = etnaviv_gpu_cooling_get_max_state, + .get_cur_state = etnaviv_gpu_cooling_get_cur_state, + .set_cur_state = etnaviv_gpu_cooling_set_cur_state, +}; + static int etnaviv_gpu_bind(struct device *dev, struct device *master, void *data) { @@ -1556,13 +1609,20 @@ static int etnaviv_gpu_bind(struct device *dev, struct device *master, struct etnaviv_gpu *gpu = dev_get_drvdata(dev); int ret; + gpu->cooling = thermal_of_cooling_device_register(dev->of_node, + (char *)dev_name(dev), gpu, &cooling_ops); + if (IS_ERR(gpu->cooling)) + return PTR_ERR(gpu->cooling); + #ifdef CONFIG_PM ret = pm_runtime_get_sync(gpu->dev); #else ret = etnaviv_gpu_clk_enable(gpu); #endif - if (ret < 0) + if (ret < 0) { + thermal_cooling_device_unregister(gpu->cooling); return ret; + } gpu->drm = drm; gpu->fence_context = dma_fence_context_alloc(1); @@ -1616,6 +1676,9 @@ static void etnaviv_gpu_unbind(struct device *dev, struct device *master, } gpu->drm = NULL; + + thermal_cooling_device_unregister(gpu->cooling); + gpu->cooling = NULL; } static const struct component_ops gpu_ops = { diff --git a/drivers/gpu/drm/etnaviv/etnaviv_gpu.h b/drivers/gpu/drm/etnaviv/etnaviv_gpu.h index 1c0606ea7d5e..9227a9740447 100644 --- a/drivers/gpu/drm/etnaviv/etnaviv_gpu.h +++ b/drivers/gpu/drm/etnaviv/etnaviv_gpu.h @@ -97,6 +97,7 @@ struct etnaviv_cmdbuf; struct etnaviv_gpu { struct drm_device *drm; + struct thermal_cooling_device *cooling; struct device *dev; struct mutex lock; struct etnaviv_chip_identity identity; @@ -150,6 +151,7 @@ struct etnaviv_gpu { u32 hangcheck_fence; u32 hangcheck_dma_addr; struct work_struct recover_work; + unsigned int freq_scale; }; static inline void gpu_write(struct etnaviv_gpu *gpu, u32 reg, u32 data) @@ -181,7 +183,7 @@ int etnaviv_gpu_debugfs(struct etnaviv_gpu *gpu, struct seq_file *m); #endif int etnaviv_gpu_fence_sync_obj(struct etnaviv_gem_object *etnaviv_obj, - unsigned int context, bool exclusive); + unsigned int context, bool exclusive, bool implicit); void etnaviv_gpu_retire(struct etnaviv_gpu *gpu); int etnaviv_gpu_wait_fence_interruptible(struct etnaviv_gpu *gpu, diff --git a/drivers/gpu/drm/i915/gvt/cmd_parser.c b/drivers/gpu/drm/i915/gvt/cmd_parser.c index 6f972afbdbc3..94f2e701e4d4 100644 --- a/drivers/gpu/drm/i915/gvt/cmd_parser.c +++ b/drivers/gpu/drm/i915/gvt/cmd_parser.c @@ -1215,7 +1215,7 @@ static int gen8_check_mi_display_flip(struct parser_exec_state *s, if (!info->async_flip) return 0; - if (IS_SKYLAKE(dev_priv)) { + if (IS_SKYLAKE(dev_priv) || IS_KABYLAKE(dev_priv)) { stride = vgpu_vreg(s->vgpu, info->stride_reg) & GENMASK(9, 0); tile = (vgpu_vreg(s->vgpu, info->ctrl_reg) & GENMASK(12, 10)) >> 10; @@ -1243,7 +1243,7 @@ static int gen8_update_plane_mmio_from_mi_display_flip( set_mask_bits(&vgpu_vreg(vgpu, info->surf_reg), GENMASK(31, 12), info->surf_val << 12); - if (IS_SKYLAKE(dev_priv)) { + if (IS_SKYLAKE(dev_priv) || IS_KABYLAKE(dev_priv)) { set_mask_bits(&vgpu_vreg(vgpu, info->stride_reg), GENMASK(9, 0), info->stride_val); set_mask_bits(&vgpu_vreg(vgpu, info->ctrl_reg), GENMASK(12, 10), @@ -1267,7 +1267,7 @@ static int decode_mi_display_flip(struct parser_exec_state *s, if (IS_BROADWELL(dev_priv)) return gen8_decode_mi_display_flip(s, info); - if (IS_SKYLAKE(dev_priv)) + if (IS_SKYLAKE(dev_priv) || IS_KABYLAKE(dev_priv)) return skl_decode_mi_display_flip(s, info); return -ENODEV; @@ -1278,7 +1278,9 @@ static int check_mi_display_flip(struct parser_exec_state *s, { struct drm_i915_private *dev_priv = s->vgpu->gvt->dev_priv; - if (IS_BROADWELL(dev_priv) || IS_SKYLAKE(dev_priv)) + if (IS_BROADWELL(dev_priv) + || IS_SKYLAKE(dev_priv) + || IS_KABYLAKE(dev_priv)) return gen8_check_mi_display_flip(s, info); return -ENODEV; } @@ -1289,7 +1291,9 @@ static int update_plane_mmio_from_mi_display_flip( { struct drm_i915_private *dev_priv = s->vgpu->gvt->dev_priv; - if (IS_BROADWELL(dev_priv) || IS_SKYLAKE(dev_priv)) + if (IS_BROADWELL(dev_priv) + || IS_SKYLAKE(dev_priv) + || IS_KABYLAKE(dev_priv)) return gen8_update_plane_mmio_from_mi_display_flip(s, info); return -ENODEV; } @@ -1569,7 +1573,8 @@ static int batch_buffer_needs_scan(struct parser_exec_state *s) { struct intel_gvt *gvt = s->vgpu->gvt; - if (IS_BROADWELL(gvt->dev_priv) || IS_SKYLAKE(gvt->dev_priv)) { + if (IS_BROADWELL(gvt->dev_priv) || IS_SKYLAKE(gvt->dev_priv) + || IS_KABYLAKE(gvt->dev_priv)) { /* BDW decides privilege based on address space */ if (cmd_val(s, 0) & (1 << 8)) return 0; @@ -2604,6 +2609,9 @@ static int scan_wa_ctx(struct intel_shadow_wa_ctx *wa_ctx) unsigned long gma_head, gma_tail, gma_bottom, ring_size, ring_tail; struct parser_exec_state s; int ret = 0; + struct intel_vgpu_workload *workload = container_of(wa_ctx, + struct intel_vgpu_workload, + wa_ctx); /* ring base is page aligned */ if (WARN_ON(!IS_ALIGNED(wa_ctx->indirect_ctx.guest_gma, GTT_PAGE_SIZE))) @@ -2618,14 +2626,14 @@ static int scan_wa_ctx(struct intel_shadow_wa_ctx *wa_ctx) s.buf_type = RING_BUFFER_INSTRUCTION; s.buf_addr_type = GTT_BUFFER; - s.vgpu = wa_ctx->workload->vgpu; - s.ring_id = wa_ctx->workload->ring_id; + s.vgpu = workload->vgpu; + s.ring_id = workload->ring_id; s.ring_start = wa_ctx->indirect_ctx.guest_gma; s.ring_size = ring_size; s.ring_head = gma_head; s.ring_tail = gma_tail; s.rb_va = wa_ctx->indirect_ctx.shadow_va; - s.workload = wa_ctx->workload; + s.workload = workload; ret = ip_gma_set(&s, gma_head); if (ret) @@ -2708,12 +2716,15 @@ static int shadow_indirect_ctx(struct intel_shadow_wa_ctx *wa_ctx) { int ctx_size = wa_ctx->indirect_ctx.size; unsigned long guest_gma = wa_ctx->indirect_ctx.guest_gma; - struct intel_vgpu *vgpu = wa_ctx->workload->vgpu; + struct intel_vgpu_workload *workload = container_of(wa_ctx, + struct intel_vgpu_workload, + wa_ctx); + struct intel_vgpu *vgpu = workload->vgpu; struct drm_i915_gem_object *obj; int ret = 0; void *map; - obj = i915_gem_object_create(wa_ctx->workload->vgpu->gvt->dev_priv, + obj = i915_gem_object_create(workload->vgpu->gvt->dev_priv, roundup(ctx_size + CACHELINE_BYTES, PAGE_SIZE)); if (IS_ERR(obj)) @@ -2733,8 +2744,8 @@ static int shadow_indirect_ctx(struct intel_shadow_wa_ctx *wa_ctx) goto unmap_src; } - ret = copy_gma_to_hva(wa_ctx->workload->vgpu, - wa_ctx->workload->vgpu->gtt.ggtt_mm, + ret = copy_gma_to_hva(workload->vgpu, + workload->vgpu->gtt.ggtt_mm, guest_gma, guest_gma + ctx_size, map); if (ret < 0) { @@ -2772,7 +2783,10 @@ static int combine_wa_ctx(struct intel_shadow_wa_ctx *wa_ctx) int intel_gvt_scan_and_shadow_wa_ctx(struct intel_shadow_wa_ctx *wa_ctx) { int ret; - struct intel_vgpu *vgpu = wa_ctx->workload->vgpu; + struct intel_vgpu_workload *workload = container_of(wa_ctx, + struct intel_vgpu_workload, + wa_ctx); + struct intel_vgpu *vgpu = workload->vgpu; if (wa_ctx->indirect_ctx.size == 0) return 0; diff --git a/drivers/gpu/drm/i915/gvt/display.c b/drivers/gpu/drm/i915/gvt/display.c index 5419ae6ec633..4cf2b29fbaa1 100644 --- a/drivers/gpu/drm/i915/gvt/display.c +++ b/drivers/gpu/drm/i915/gvt/display.c @@ -161,8 +161,9 @@ static unsigned char virtual_dp_monitor_edid[GVT_EDID_NUM][EDID_SIZE] = { #define DPCD_HEADER_SIZE 0xb +/* let the virtual display supports DP1.2 */ static u8 dpcd_fix_data[DPCD_HEADER_SIZE] = { - 0x11, 0x0a, 0x04, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 + 0x12, 0x014, 0x04, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 }; static void emulate_monitor_status_change(struct intel_vgpu *vgpu) @@ -172,9 +173,20 @@ static void emulate_monitor_status_change(struct intel_vgpu *vgpu) SDE_PORTC_HOTPLUG_CPT | SDE_PORTD_HOTPLUG_CPT); - if (IS_SKYLAKE(dev_priv)) + if (IS_SKYLAKE(dev_priv) || IS_KABYLAKE(dev_priv)) { vgpu_vreg(vgpu, SDEISR) &= ~(SDE_PORTA_HOTPLUG_SPT | SDE_PORTE_HOTPLUG_SPT); + vgpu_vreg(vgpu, SKL_FUSE_STATUS) |= + SKL_FUSE_DOWNLOAD_STATUS | + SKL_FUSE_PG0_DIST_STATUS | + SKL_FUSE_PG1_DIST_STATUS | + SKL_FUSE_PG2_DIST_STATUS; + vgpu_vreg(vgpu, LCPLL1_CTL) |= + LCPLL_PLL_ENABLE | + LCPLL_PLL_LOCK; + vgpu_vreg(vgpu, LCPLL2_CTL) |= LCPLL_PLL_ENABLE; + + } if (intel_vgpu_has_monitor_on_port(vgpu, PORT_B)) { vgpu_vreg(vgpu, SDEISR) |= SDE_PORTB_HOTPLUG_CPT; @@ -191,7 +203,7 @@ static void emulate_monitor_status_change(struct intel_vgpu *vgpu) vgpu_vreg(vgpu, SFUSE_STRAP) |= SFUSE_STRAP_DDID_DETECTED; } - if (IS_SKYLAKE(dev_priv) && + if ((IS_SKYLAKE(dev_priv) || IS_KABYLAKE(dev_priv)) && intel_vgpu_has_monitor_on_port(vgpu, PORT_E)) { vgpu_vreg(vgpu, SDEISR) |= SDE_PORTE_HOTPLUG_SPT; } @@ -353,7 +365,7 @@ void intel_vgpu_clean_display(struct intel_vgpu *vgpu) { struct drm_i915_private *dev_priv = vgpu->gvt->dev_priv; - if (IS_SKYLAKE(dev_priv)) + if (IS_SKYLAKE(dev_priv) || IS_KABYLAKE(dev_priv)) clean_virtual_dp_monitor(vgpu, PORT_D); else clean_virtual_dp_monitor(vgpu, PORT_B); @@ -375,7 +387,7 @@ int intel_vgpu_init_display(struct intel_vgpu *vgpu, u64 resolution) intel_vgpu_init_i2c_edid(vgpu); - if (IS_SKYLAKE(dev_priv)) + if (IS_SKYLAKE(dev_priv) || IS_KABYLAKE(dev_priv)) return setup_virtual_dp_monitor(vgpu, PORT_D, GVT_DP_D, resolution); else diff --git a/drivers/gpu/drm/i915/gvt/execlist.c b/drivers/gpu/drm/i915/gvt/execlist.c index f1f426a97aa9..ce4276a7cf9c 100644 --- a/drivers/gpu/drm/i915/gvt/execlist.c +++ b/drivers/gpu/drm/i915/gvt/execlist.c @@ -394,9 +394,11 @@ static void prepare_shadow_batch_buffer(struct intel_vgpu_workload *workload) static int update_wa_ctx_2_shadow_ctx(struct intel_shadow_wa_ctx *wa_ctx) { - int ring_id = wa_ctx->workload->ring_id; - struct i915_gem_context *shadow_ctx = - wa_ctx->workload->vgpu->shadow_ctx; + struct intel_vgpu_workload *workload = container_of(wa_ctx, + struct intel_vgpu_workload, + wa_ctx); + int ring_id = workload->ring_id; + struct i915_gem_context *shadow_ctx = workload->vgpu->shadow_ctx; struct drm_i915_gem_object *ctx_obj = shadow_ctx->engine[ring_id].state->obj; struct execlist_ring_context *shadow_ring_context; @@ -680,7 +682,6 @@ static int submit_context(struct intel_vgpu *vgpu, int ring_id, CACHELINE_BYTES; workload->wa_ctx.per_ctx.guest_gma = per_ctx & PER_CTX_ADDR_MASK; - workload->wa_ctx.workload = workload; WARN_ON(workload->wa_ctx.indirect_ctx.size && !(per_ctx & 0x1)); } diff --git a/drivers/gpu/drm/i915/gvt/gtt.c b/drivers/gpu/drm/i915/gvt/gtt.c index b832bea64e03..6da4e444e572 100644 --- a/drivers/gpu/drm/i915/gvt/gtt.c +++ b/drivers/gpu/drm/i915/gvt/gtt.c @@ -2224,7 +2224,8 @@ int intel_gvt_init_gtt(struct intel_gvt *gvt) gvt_dbg_core("init gtt\n"); - if (IS_BROADWELL(gvt->dev_priv) || IS_SKYLAKE(gvt->dev_priv)) { + if (IS_BROADWELL(gvt->dev_priv) || IS_SKYLAKE(gvt->dev_priv) + || IS_KABYLAKE(gvt->dev_priv)) { gvt->gtt.pte_ops = &gen8_gtt_pte_ops; gvt->gtt.gma_ops = &gen8_gtt_gma_ops; gvt->gtt.mm_alloc_page_table = gen8_mm_alloc_page_table; diff --git a/drivers/gpu/drm/i915/gvt/gvt.c b/drivers/gpu/drm/i915/gvt/gvt.c index 3b9d59e457ba..0f3a98865a58 100644 --- a/drivers/gpu/drm/i915/gvt/gvt.c +++ b/drivers/gpu/drm/i915/gvt/gvt.c @@ -106,7 +106,8 @@ static void init_device_info(struct intel_gvt *gvt) struct intel_gvt_device_info *info = &gvt->device_info; struct pci_dev *pdev = gvt->dev_priv->drm.pdev; - if (IS_BROADWELL(gvt->dev_priv) || IS_SKYLAKE(gvt->dev_priv)) { + if (IS_BROADWELL(gvt->dev_priv) || IS_SKYLAKE(gvt->dev_priv) + || IS_KABYLAKE(gvt->dev_priv)) { info->max_support_vgpus = 8; info->cfg_space_size = 256; info->mmio_size = 2 * 1024 * 1024; @@ -143,6 +144,11 @@ static int gvt_service_thread(void *data) intel_gvt_emulate_vblank(gvt); mutex_unlock(&gvt->lock); } + + if (test_and_clear_bit(INTEL_GVT_REQUEST_SCHED, + (void *)&gvt->service_request)) { + intel_gvt_schedule(gvt); + } } return 0; @@ -196,6 +202,8 @@ void intel_gvt_clean_device(struct drm_i915_private *dev_priv) idr_destroy(&gvt->vgpu_idr); + intel_gvt_destroy_idle_vgpu(gvt->idle_vgpu); + kfree(dev_priv->gvt); dev_priv->gvt = NULL; } @@ -214,6 +222,7 @@ void intel_gvt_clean_device(struct drm_i915_private *dev_priv) int intel_gvt_init_device(struct drm_i915_private *dev_priv) { struct intel_gvt *gvt; + struct intel_vgpu *vgpu; int ret; /* @@ -286,6 +295,14 @@ int intel_gvt_init_device(struct drm_i915_private *dev_priv) goto out_clean_types; } + vgpu = intel_gvt_create_idle_vgpu(gvt); + if (IS_ERR(vgpu)) { + ret = PTR_ERR(vgpu); + gvt_err("failed to create idle vgpu\n"); + goto out_clean_types; + } + gvt->idle_vgpu = vgpu; + gvt_dbg_core("gvt device initialization is done\n"); dev_priv->gvt = gvt; return 0; diff --git a/drivers/gpu/drm/i915/gvt/gvt.h b/drivers/gpu/drm/i915/gvt/gvt.h index 6dfc48b63b71..806da96b6a92 100644 --- a/drivers/gpu/drm/i915/gvt/gvt.h +++ b/drivers/gpu/drm/i915/gvt/gvt.h @@ -138,6 +138,10 @@ struct intel_vgpu_display { struct intel_vgpu_sbi sbi; }; +struct vgpu_sched_ctl { + int weight; +}; + struct intel_vgpu { struct intel_gvt *gvt; int id; @@ -147,6 +151,7 @@ struct intel_vgpu { bool failsafe; bool resetting; void *sched_data; + struct vgpu_sched_ctl sched_ctl; struct intel_vgpu_fence fence; struct intel_vgpu_gm gm; @@ -160,6 +165,7 @@ struct intel_vgpu { struct list_head workload_q_head[I915_NUM_ENGINES]; struct kmem_cache *workloads; atomic_t running_workload_num; + ktime_t last_ctx_submit_time; DECLARE_BITMAP(tlb_handle_pending, I915_NUM_ENGINES); struct i915_gem_context *shadow_ctx; @@ -215,6 +221,7 @@ struct intel_vgpu_type { unsigned int low_gm_size; unsigned int high_gm_size; unsigned int fence; + unsigned int weight; enum intel_vgpu_edid resolution; }; @@ -236,6 +243,7 @@ struct intel_gvt { DECLARE_HASHTABLE(cmd_table, GVT_CMD_HASH_BITS); struct intel_vgpu_type *types; unsigned int num_types; + struct intel_vgpu *idle_vgpu; struct task_struct *service_thread; wait_queue_head_t service_thread_wq; @@ -249,6 +257,7 @@ static inline struct intel_gvt *to_gvt(struct drm_i915_private *i915) enum { INTEL_GVT_REQUEST_EMULATE_VBLANK = 0, + INTEL_GVT_REQUEST_SCHED = 1, }; static inline void intel_gvt_request_service(struct intel_gvt *gvt, @@ -322,6 +331,8 @@ struct intel_vgpu_creation_params { __u64 resolution; __s32 primary; __u64 vgpu_id; + + __u32 weight; }; int intel_vgpu_alloc_resource(struct intel_vgpu *vgpu, @@ -376,6 +387,8 @@ static inline void intel_vgpu_write_pci_bar(struct intel_vgpu *vgpu, int intel_gvt_init_vgpu_types(struct intel_gvt *gvt); void intel_gvt_clean_vgpu_types(struct intel_gvt *gvt); +struct intel_vgpu *intel_gvt_create_idle_vgpu(struct intel_gvt *gvt); +void intel_gvt_destroy_idle_vgpu(struct intel_vgpu *vgpu); struct intel_vgpu *intel_gvt_create_vgpu(struct intel_gvt *gvt, struct intel_vgpu_type *type); void intel_gvt_destroy_vgpu(struct intel_vgpu *vgpu); diff --git a/drivers/gpu/drm/i915/gvt/handlers.c b/drivers/gpu/drm/i915/gvt/handlers.c index 6da9ae1618e3..0ad1a508e2af 100644 --- a/drivers/gpu/drm/i915/gvt/handlers.c +++ b/drivers/gpu/drm/i915/gvt/handlers.c @@ -68,6 +68,8 @@ unsigned long intel_gvt_get_device_type(struct intel_gvt *gvt) return D_BDW; else if (IS_SKYLAKE(gvt->dev_priv)) return D_SKL; + else if (IS_KABYLAKE(gvt->dev_priv)) + return D_KBL; return 0; } @@ -234,7 +236,8 @@ static int mul_force_wake_write(struct intel_vgpu *vgpu, old = vgpu_vreg(vgpu, offset); new = CALC_MODE_MASK_REG(old, *(u32 *)p_data); - if (IS_SKYLAKE(vgpu->gvt->dev_priv)) { + if (IS_SKYLAKE(vgpu->gvt->dev_priv) + || IS_KABYLAKE(vgpu->gvt->dev_priv)) { switch (offset) { case FORCEWAKE_RENDER_GEN9_REG: ack_reg_offset = FORCEWAKE_ACK_RENDER_GEN9_REG; @@ -823,8 +826,9 @@ static int dp_aux_ch_ctl_mmio_write(struct intel_vgpu *vgpu, write_vreg(vgpu, offset, p_data, bytes); data = vgpu_vreg(vgpu, offset); - if (IS_SKYLAKE(vgpu->gvt->dev_priv) && - offset != _REG_SKL_DP_AUX_CH_CTL(port_index)) { + if ((IS_SKYLAKE(vgpu->gvt->dev_priv) + || IS_KABYLAKE(vgpu->gvt->dev_priv)) + && offset != _REG_SKL_DP_AUX_CH_CTL(port_index)) { /* SKL DPB/C/D aux ctl register changed */ return 0; } else if (IS_BROADWELL(vgpu->gvt->dev_priv) && @@ -1311,7 +1315,8 @@ static int mailbox_write(struct intel_vgpu *vgpu, unsigned int offset, switch (cmd) { case GEN9_PCODE_READ_MEM_LATENCY: - if (IS_SKYLAKE(vgpu->gvt->dev_priv)) { + if (IS_SKYLAKE(vgpu->gvt->dev_priv) + || IS_KABYLAKE(vgpu->gvt->dev_priv)) { /** * "Read memory latency" command on gen9. * Below memory latency values are read @@ -1324,7 +1329,8 @@ static int mailbox_write(struct intel_vgpu *vgpu, unsigned int offset, } break; case SKL_PCODE_CDCLK_CONTROL: - if (IS_SKYLAKE(vgpu->gvt->dev_priv)) + if (IS_SKYLAKE(vgpu->gvt->dev_priv) + || IS_KABYLAKE(vgpu->gvt->dev_priv)) *data0 = SKL_CDCLK_READY_FOR_CHANGE; break; case GEN6_PCODE_READ_RC6VIDS: @@ -1418,6 +1424,7 @@ static int elsp_mmio_write(struct intel_vgpu *vgpu, unsigned int offset, execlist->elsp_dwords.data[execlist->elsp_dwords.index] = data; if (execlist->elsp_dwords.index == 3) { + vgpu->last_ctx_submit_time = ktime_get(); ret = intel_vgpu_submit_execlist(vgpu, ring_id); if(ret) gvt_vgpu_err("fail submit workload on ring %d\n", @@ -2592,219 +2599,232 @@ static int init_skl_mmio_info(struct intel_gvt *gvt) MMIO_DH(FORCEWAKE_MEDIA_GEN9, D_SKL_PLUS, NULL, mul_force_wake_write); MMIO_DH(FORCEWAKE_ACK_MEDIA_GEN9, D_SKL_PLUS, NULL, NULL); - MMIO_F(_DPB_AUX_CH_CTL, 6 * 4, 0, 0, 0, D_SKL, NULL, dp_aux_ch_ctl_mmio_write); - MMIO_F(_DPC_AUX_CH_CTL, 6 * 4, 0, 0, 0, D_SKL, NULL, dp_aux_ch_ctl_mmio_write); - MMIO_F(_DPD_AUX_CH_CTL, 6 * 4, 0, 0, 0, D_SKL, NULL, dp_aux_ch_ctl_mmio_write); + MMIO_F(_DPB_AUX_CH_CTL, 6 * 4, 0, 0, 0, D_SKL_PLUS, NULL, + dp_aux_ch_ctl_mmio_write); + MMIO_F(_DPC_AUX_CH_CTL, 6 * 4, 0, 0, 0, D_SKL_PLUS, NULL, + dp_aux_ch_ctl_mmio_write); + MMIO_F(_DPD_AUX_CH_CTL, 6 * 4, 0, 0, 0, D_SKL_PLUS, NULL, + dp_aux_ch_ctl_mmio_write); - MMIO_D(HSW_PWR_WELL_BIOS, D_SKL); - MMIO_DH(HSW_PWR_WELL_DRIVER, D_SKL, NULL, skl_power_well_ctl_write); + MMIO_D(HSW_PWR_WELL_BIOS, D_SKL_PLUS); + MMIO_DH(HSW_PWR_WELL_DRIVER, D_SKL_PLUS, NULL, + skl_power_well_ctl_write); + MMIO_DH(GEN6_PCODE_MAILBOX, D_SKL_PLUS, NULL, mailbox_write); MMIO_D(0xa210, D_SKL_PLUS); MMIO_D(GEN9_MEDIA_PG_IDLE_HYSTERESIS, D_SKL_PLUS); MMIO_D(GEN9_RENDER_PG_IDLE_HYSTERESIS, D_SKL_PLUS); MMIO_DFH(GEN9_GAMT_ECO_REG_RW_IA, D_SKL_PLUS, F_CMD_ACCESS, NULL, NULL); - MMIO_DH(0x4ddc, D_SKL, NULL, skl_misc_ctl_write); - MMIO_DH(0x42080, D_SKL, NULL, skl_misc_ctl_write); - MMIO_D(0x45504, D_SKL); - MMIO_D(0x45520, D_SKL); - MMIO_D(0x46000, D_SKL); - MMIO_DH(0x46010, D_SKL, NULL, skl_lcpll_write); - MMIO_DH(0x46014, D_SKL, NULL, skl_lcpll_write); - MMIO_D(0x6C040, D_SKL); - MMIO_D(0x6C048, D_SKL); - MMIO_D(0x6C050, D_SKL); - MMIO_D(0x6C044, D_SKL); - MMIO_D(0x6C04C, D_SKL); - MMIO_D(0x6C054, D_SKL); - MMIO_D(0x6c058, D_SKL); - MMIO_D(0x6c05c, D_SKL); - MMIO_DH(0X6c060, D_SKL, dpll_status_read, NULL); - - MMIO_DH(SKL_PS_WIN_POS(PIPE_A, 0), D_SKL, NULL, pf_write); - MMIO_DH(SKL_PS_WIN_POS(PIPE_A, 1), D_SKL, NULL, pf_write); - MMIO_DH(SKL_PS_WIN_POS(PIPE_B, 0), D_SKL, NULL, pf_write); - MMIO_DH(SKL_PS_WIN_POS(PIPE_B, 1), D_SKL, NULL, pf_write); - MMIO_DH(SKL_PS_WIN_POS(PIPE_C, 0), D_SKL, NULL, pf_write); - MMIO_DH(SKL_PS_WIN_POS(PIPE_C, 1), D_SKL, NULL, pf_write); - - MMIO_DH(SKL_PS_WIN_SZ(PIPE_A, 0), D_SKL, NULL, pf_write); - MMIO_DH(SKL_PS_WIN_SZ(PIPE_A, 1), D_SKL, NULL, pf_write); - MMIO_DH(SKL_PS_WIN_SZ(PIPE_B, 0), D_SKL, NULL, pf_write); - MMIO_DH(SKL_PS_WIN_SZ(PIPE_B, 1), D_SKL, NULL, pf_write); - MMIO_DH(SKL_PS_WIN_SZ(PIPE_C, 0), D_SKL, NULL, pf_write); - MMIO_DH(SKL_PS_WIN_SZ(PIPE_C, 1), D_SKL, NULL, pf_write); - - MMIO_DH(SKL_PS_CTRL(PIPE_A, 0), D_SKL, NULL, pf_write); - MMIO_DH(SKL_PS_CTRL(PIPE_A, 1), D_SKL, NULL, pf_write); - MMIO_DH(SKL_PS_CTRL(PIPE_B, 0), D_SKL, NULL, pf_write); - MMIO_DH(SKL_PS_CTRL(PIPE_B, 1), D_SKL, NULL, pf_write); - MMIO_DH(SKL_PS_CTRL(PIPE_C, 0), D_SKL, NULL, pf_write); - MMIO_DH(SKL_PS_CTRL(PIPE_C, 1), D_SKL, NULL, pf_write); - - MMIO_DH(PLANE_BUF_CFG(PIPE_A, 0), D_SKL, NULL, NULL); - MMIO_DH(PLANE_BUF_CFG(PIPE_A, 1), D_SKL, NULL, NULL); - MMIO_DH(PLANE_BUF_CFG(PIPE_A, 2), D_SKL, NULL, NULL); - MMIO_DH(PLANE_BUF_CFG(PIPE_A, 3), D_SKL, NULL, NULL); - - MMIO_DH(PLANE_BUF_CFG(PIPE_B, 0), D_SKL, NULL, NULL); - MMIO_DH(PLANE_BUF_CFG(PIPE_B, 1), D_SKL, NULL, NULL); - MMIO_DH(PLANE_BUF_CFG(PIPE_B, 2), D_SKL, NULL, NULL); - MMIO_DH(PLANE_BUF_CFG(PIPE_B, 3), D_SKL, NULL, NULL); - - MMIO_DH(PLANE_BUF_CFG(PIPE_C, 0), D_SKL, NULL, NULL); - MMIO_DH(PLANE_BUF_CFG(PIPE_C, 1), D_SKL, NULL, NULL); - MMIO_DH(PLANE_BUF_CFG(PIPE_C, 2), D_SKL, NULL, NULL); - MMIO_DH(PLANE_BUF_CFG(PIPE_C, 3), D_SKL, NULL, NULL); - - MMIO_DH(CUR_BUF_CFG(PIPE_A), D_SKL, NULL, NULL); - MMIO_DH(CUR_BUF_CFG(PIPE_B), D_SKL, NULL, NULL); - MMIO_DH(CUR_BUF_CFG(PIPE_C), D_SKL, NULL, NULL); - - MMIO_F(PLANE_WM(PIPE_A, 0, 0), 4 * 8, 0, 0, 0, D_SKL, NULL, NULL); - MMIO_F(PLANE_WM(PIPE_A, 1, 0), 4 * 8, 0, 0, 0, D_SKL, NULL, NULL); - MMIO_F(PLANE_WM(PIPE_A, 2, 0), 4 * 8, 0, 0, 0, D_SKL, NULL, NULL); - - MMIO_F(PLANE_WM(PIPE_B, 0, 0), 4 * 8, 0, 0, 0, D_SKL, NULL, NULL); - MMIO_F(PLANE_WM(PIPE_B, 1, 0), 4 * 8, 0, 0, 0, D_SKL, NULL, NULL); - MMIO_F(PLANE_WM(PIPE_B, 2, 0), 4 * 8, 0, 0, 0, D_SKL, NULL, NULL); - - MMIO_F(PLANE_WM(PIPE_C, 0, 0), 4 * 8, 0, 0, 0, D_SKL, NULL, NULL); - MMIO_F(PLANE_WM(PIPE_C, 1, 0), 4 * 8, 0, 0, 0, D_SKL, NULL, NULL); - MMIO_F(PLANE_WM(PIPE_C, 2, 0), 4 * 8, 0, 0, 0, D_SKL, NULL, NULL); - - MMIO_F(CUR_WM(PIPE_A, 0), 4 * 8, 0, 0, 0, D_SKL, NULL, NULL); - MMIO_F(CUR_WM(PIPE_B, 0), 4 * 8, 0, 0, 0, D_SKL, NULL, NULL); - MMIO_F(CUR_WM(PIPE_C, 0), 4 * 8, 0, 0, 0, D_SKL, NULL, NULL); - - MMIO_DH(PLANE_WM_TRANS(PIPE_A, 0), D_SKL, NULL, NULL); - MMIO_DH(PLANE_WM_TRANS(PIPE_A, 1), D_SKL, NULL, NULL); - MMIO_DH(PLANE_WM_TRANS(PIPE_A, 2), D_SKL, NULL, NULL); - - MMIO_DH(PLANE_WM_TRANS(PIPE_B, 0), D_SKL, NULL, NULL); - MMIO_DH(PLANE_WM_TRANS(PIPE_B, 1), D_SKL, NULL, NULL); - MMIO_DH(PLANE_WM_TRANS(PIPE_B, 2), D_SKL, NULL, NULL); - - MMIO_DH(PLANE_WM_TRANS(PIPE_C, 0), D_SKL, NULL, NULL); - MMIO_DH(PLANE_WM_TRANS(PIPE_C, 1), D_SKL, NULL, NULL); - MMIO_DH(PLANE_WM_TRANS(PIPE_C, 2), D_SKL, NULL, NULL); - - MMIO_DH(CUR_WM_TRANS(PIPE_A), D_SKL, NULL, NULL); - MMIO_DH(CUR_WM_TRANS(PIPE_B), D_SKL, NULL, NULL); - MMIO_DH(CUR_WM_TRANS(PIPE_C), D_SKL, NULL, NULL); - - MMIO_DH(PLANE_NV12_BUF_CFG(PIPE_A, 0), D_SKL, NULL, NULL); - MMIO_DH(PLANE_NV12_BUF_CFG(PIPE_A, 1), D_SKL, NULL, NULL); - MMIO_DH(PLANE_NV12_BUF_CFG(PIPE_A, 2), D_SKL, NULL, NULL); - MMIO_DH(PLANE_NV12_BUF_CFG(PIPE_A, 3), D_SKL, NULL, NULL); - - MMIO_DH(PLANE_NV12_BUF_CFG(PIPE_B, 0), D_SKL, NULL, NULL); - MMIO_DH(PLANE_NV12_BUF_CFG(PIPE_B, 1), D_SKL, NULL, NULL); - MMIO_DH(PLANE_NV12_BUF_CFG(PIPE_B, 2), D_SKL, NULL, NULL); - MMIO_DH(PLANE_NV12_BUF_CFG(PIPE_B, 3), D_SKL, NULL, NULL); - - MMIO_DH(PLANE_NV12_BUF_CFG(PIPE_C, 0), D_SKL, NULL, NULL); - MMIO_DH(PLANE_NV12_BUF_CFG(PIPE_C, 1), D_SKL, NULL, NULL); - MMIO_DH(PLANE_NV12_BUF_CFG(PIPE_C, 2), D_SKL, NULL, NULL); - MMIO_DH(PLANE_NV12_BUF_CFG(PIPE_C, 3), D_SKL, NULL, NULL); - - MMIO_DH(_REG_701C0(PIPE_A, 1), D_SKL, NULL, NULL); - MMIO_DH(_REG_701C0(PIPE_A, 2), D_SKL, NULL, NULL); - MMIO_DH(_REG_701C0(PIPE_A, 3), D_SKL, NULL, NULL); - MMIO_DH(_REG_701C0(PIPE_A, 4), D_SKL, NULL, NULL); - - MMIO_DH(_REG_701C0(PIPE_B, 1), D_SKL, NULL, NULL); - MMIO_DH(_REG_701C0(PIPE_B, 2), D_SKL, NULL, NULL); - MMIO_DH(_REG_701C0(PIPE_B, 3), D_SKL, NULL, NULL); - MMIO_DH(_REG_701C0(PIPE_B, 4), D_SKL, NULL, NULL); - - MMIO_DH(_REG_701C0(PIPE_C, 1), D_SKL, NULL, NULL); - MMIO_DH(_REG_701C0(PIPE_C, 2), D_SKL, NULL, NULL); - MMIO_DH(_REG_701C0(PIPE_C, 3), D_SKL, NULL, NULL); - MMIO_DH(_REG_701C0(PIPE_C, 4), D_SKL, NULL, NULL); - - MMIO_DH(_REG_701C4(PIPE_A, 1), D_SKL, NULL, NULL); - MMIO_DH(_REG_701C4(PIPE_A, 2), D_SKL, NULL, NULL); - MMIO_DH(_REG_701C4(PIPE_A, 3), D_SKL, NULL, NULL); - MMIO_DH(_REG_701C4(PIPE_A, 4), D_SKL, NULL, NULL); - - MMIO_DH(_REG_701C4(PIPE_B, 1), D_SKL, NULL, NULL); - MMIO_DH(_REG_701C4(PIPE_B, 2), D_SKL, NULL, NULL); - MMIO_DH(_REG_701C4(PIPE_B, 3), D_SKL, NULL, NULL); - MMIO_DH(_REG_701C4(PIPE_B, 4), D_SKL, NULL, NULL); - - MMIO_DH(_REG_701C4(PIPE_C, 1), D_SKL, NULL, NULL); - MMIO_DH(_REG_701C4(PIPE_C, 2), D_SKL, NULL, NULL); - MMIO_DH(_REG_701C4(PIPE_C, 3), D_SKL, NULL, NULL); - MMIO_DH(_REG_701C4(PIPE_C, 4), D_SKL, NULL, NULL); - - MMIO_D(0x70380, D_SKL); - MMIO_D(0x71380, D_SKL); - MMIO_D(0x72380, D_SKL); - MMIO_D(0x7039c, D_SKL); - - MMIO_F(0x80000, 0x3000, 0, 0, 0, D_SKL, NULL, NULL); - MMIO_D(0x8f074, D_SKL); - MMIO_D(0x8f004, D_SKL); - MMIO_D(0x8f034, D_SKL); - - MMIO_D(0xb11c, D_SKL); - - MMIO_D(0x51000, D_SKL); - MMIO_D(0x6c00c, D_SKL); - - MMIO_F(0xc800, 0x7f8, F_CMD_ACCESS, 0, 0, D_SKL, NULL, NULL); - MMIO_F(0xb020, 0x80, F_CMD_ACCESS, 0, 0, D_SKL, NULL, NULL); - - MMIO_D(0xd08, D_SKL); - MMIO_DFH(0x20e0, D_SKL, F_MODE_MASK, NULL, NULL); - MMIO_DFH(0x20ec, D_SKL, F_MODE_MASK | F_CMD_ACCESS, NULL, NULL); + MMIO_DH(0x4ddc, D_SKL_PLUS, NULL, skl_misc_ctl_write); + MMIO_DH(0x42080, D_SKL_PLUS, NULL, skl_misc_ctl_write); + MMIO_D(0x45504, D_SKL_PLUS); + MMIO_D(0x45520, D_SKL_PLUS); + MMIO_D(0x46000, D_SKL_PLUS); + MMIO_DH(0x46010, D_SKL | D_KBL, NULL, skl_lcpll_write); + MMIO_DH(0x46014, D_SKL | D_KBL, NULL, skl_lcpll_write); + MMIO_D(0x6C040, D_SKL | D_KBL); + MMIO_D(0x6C048, D_SKL | D_KBL); + MMIO_D(0x6C050, D_SKL | D_KBL); + MMIO_D(0x6C044, D_SKL | D_KBL); + MMIO_D(0x6C04C, D_SKL | D_KBL); + MMIO_D(0x6C054, D_SKL | D_KBL); + MMIO_D(0x6c058, D_SKL | D_KBL); + MMIO_D(0x6c05c, D_SKL | D_KBL); + MMIO_DH(0X6c060, D_SKL | D_KBL, dpll_status_read, NULL); + + MMIO_DH(SKL_PS_WIN_POS(PIPE_A, 0), D_SKL_PLUS, NULL, pf_write); + MMIO_DH(SKL_PS_WIN_POS(PIPE_A, 1), D_SKL_PLUS, NULL, pf_write); + MMIO_DH(SKL_PS_WIN_POS(PIPE_B, 0), D_SKL_PLUS, NULL, pf_write); + MMIO_DH(SKL_PS_WIN_POS(PIPE_B, 1), D_SKL_PLUS, NULL, pf_write); + MMIO_DH(SKL_PS_WIN_POS(PIPE_C, 0), D_SKL_PLUS, NULL, pf_write); + MMIO_DH(SKL_PS_WIN_POS(PIPE_C, 1), D_SKL_PLUS, NULL, pf_write); + + MMIO_DH(SKL_PS_WIN_SZ(PIPE_A, 0), D_SKL_PLUS, NULL, pf_write); + MMIO_DH(SKL_PS_WIN_SZ(PIPE_A, 1), D_SKL_PLUS, NULL, pf_write); + MMIO_DH(SKL_PS_WIN_SZ(PIPE_B, 0), D_SKL_PLUS, NULL, pf_write); + MMIO_DH(SKL_PS_WIN_SZ(PIPE_B, 1), D_SKL_PLUS, NULL, pf_write); + MMIO_DH(SKL_PS_WIN_SZ(PIPE_C, 0), D_SKL_PLUS, NULL, pf_write); + MMIO_DH(SKL_PS_WIN_SZ(PIPE_C, 1), D_SKL_PLUS, NULL, pf_write); + + MMIO_DH(SKL_PS_CTRL(PIPE_A, 0), D_SKL_PLUS, NULL, pf_write); + MMIO_DH(SKL_PS_CTRL(PIPE_A, 1), D_SKL_PLUS, NULL, pf_write); + MMIO_DH(SKL_PS_CTRL(PIPE_B, 0), D_SKL_PLUS, NULL, pf_write); + MMIO_DH(SKL_PS_CTRL(PIPE_B, 1), D_SKL_PLUS, NULL, pf_write); + MMIO_DH(SKL_PS_CTRL(PIPE_C, 0), D_SKL_PLUS, NULL, pf_write); + MMIO_DH(SKL_PS_CTRL(PIPE_C, 1), D_SKL_PLUS, NULL, pf_write); + + MMIO_DH(PLANE_BUF_CFG(PIPE_A, 0), D_SKL_PLUS, NULL, NULL); + MMIO_DH(PLANE_BUF_CFG(PIPE_A, 1), D_SKL_PLUS, NULL, NULL); + MMIO_DH(PLANE_BUF_CFG(PIPE_A, 2), D_SKL_PLUS, NULL, NULL); + MMIO_DH(PLANE_BUF_CFG(PIPE_A, 3), D_SKL_PLUS, NULL, NULL); + + MMIO_DH(PLANE_BUF_CFG(PIPE_B, 0), D_SKL_PLUS, NULL, NULL); + MMIO_DH(PLANE_BUF_CFG(PIPE_B, 1), D_SKL_PLUS, NULL, NULL); + MMIO_DH(PLANE_BUF_CFG(PIPE_B, 2), D_SKL_PLUS, NULL, NULL); + MMIO_DH(PLANE_BUF_CFG(PIPE_B, 3), D_SKL_PLUS, NULL, NULL); + + MMIO_DH(PLANE_BUF_CFG(PIPE_C, 0), D_SKL_PLUS, NULL, NULL); + MMIO_DH(PLANE_BUF_CFG(PIPE_C, 1), D_SKL_PLUS, NULL, NULL); + MMIO_DH(PLANE_BUF_CFG(PIPE_C, 2), D_SKL_PLUS, NULL, NULL); + MMIO_DH(PLANE_BUF_CFG(PIPE_C, 3), D_SKL_PLUS, NULL, NULL); + + MMIO_DH(CUR_BUF_CFG(PIPE_A), D_SKL_PLUS, NULL, NULL); + MMIO_DH(CUR_BUF_CFG(PIPE_B), D_SKL_PLUS, NULL, NULL); + MMIO_DH(CUR_BUF_CFG(PIPE_C), D_SKL_PLUS, NULL, NULL); + + MMIO_F(PLANE_WM(PIPE_A, 0, 0), 4 * 8, 0, 0, 0, D_SKL_PLUS, NULL, NULL); + MMIO_F(PLANE_WM(PIPE_A, 1, 0), 4 * 8, 0, 0, 0, D_SKL_PLUS, NULL, NULL); + MMIO_F(PLANE_WM(PIPE_A, 2, 0), 4 * 8, 0, 0, 0, D_SKL_PLUS, NULL, NULL); + + MMIO_F(PLANE_WM(PIPE_B, 0, 0), 4 * 8, 0, 0, 0, D_SKL_PLUS, NULL, NULL); + MMIO_F(PLANE_WM(PIPE_B, 1, 0), 4 * 8, 0, 0, 0, D_SKL_PLUS, NULL, NULL); + MMIO_F(PLANE_WM(PIPE_B, 2, 0), 4 * 8, 0, 0, 0, D_SKL_PLUS, NULL, NULL); + + MMIO_F(PLANE_WM(PIPE_C, 0, 0), 4 * 8, 0, 0, 0, D_SKL_PLUS, NULL, NULL); + MMIO_F(PLANE_WM(PIPE_C, 1, 0), 4 * 8, 0, 0, 0, D_SKL_PLUS, NULL, NULL); + MMIO_F(PLANE_WM(PIPE_C, 2, 0), 4 * 8, 0, 0, 0, D_SKL_PLUS, NULL, NULL); + + MMIO_F(CUR_WM(PIPE_A, 0), 4 * 8, 0, 0, 0, D_SKL_PLUS, NULL, NULL); + MMIO_F(CUR_WM(PIPE_B, 0), 4 * 8, 0, 0, 0, D_SKL_PLUS, NULL, NULL); + MMIO_F(CUR_WM(PIPE_C, 0), 4 * 8, 0, 0, 0, D_SKL_PLUS, NULL, NULL); + + MMIO_DH(PLANE_WM_TRANS(PIPE_A, 0), D_SKL_PLUS, NULL, NULL); + MMIO_DH(PLANE_WM_TRANS(PIPE_A, 1), D_SKL_PLUS, NULL, NULL); + MMIO_DH(PLANE_WM_TRANS(PIPE_A, 2), D_SKL_PLUS, NULL, NULL); + + MMIO_DH(PLANE_WM_TRANS(PIPE_B, 0), D_SKL_PLUS, NULL, NULL); + MMIO_DH(PLANE_WM_TRANS(PIPE_B, 1), D_SKL_PLUS, NULL, NULL); + MMIO_DH(PLANE_WM_TRANS(PIPE_B, 2), D_SKL_PLUS, NULL, NULL); + + MMIO_DH(PLANE_WM_TRANS(PIPE_C, 0), D_SKL_PLUS, NULL, NULL); + MMIO_DH(PLANE_WM_TRANS(PIPE_C, 1), D_SKL_PLUS, NULL, NULL); + MMIO_DH(PLANE_WM_TRANS(PIPE_C, 2), D_SKL_PLUS, NULL, NULL); + + MMIO_DH(CUR_WM_TRANS(PIPE_A), D_SKL_PLUS, NULL, NULL); + MMIO_DH(CUR_WM_TRANS(PIPE_B), D_SKL_PLUS, NULL, NULL); + MMIO_DH(CUR_WM_TRANS(PIPE_C), D_SKL_PLUS, NULL, NULL); + + MMIO_DH(PLANE_NV12_BUF_CFG(PIPE_A, 0), D_SKL_PLUS, NULL, NULL); + MMIO_DH(PLANE_NV12_BUF_CFG(PIPE_A, 1), D_SKL_PLUS, NULL, NULL); + MMIO_DH(PLANE_NV12_BUF_CFG(PIPE_A, 2), D_SKL_PLUS, NULL, NULL); + MMIO_DH(PLANE_NV12_BUF_CFG(PIPE_A, 3), D_SKL_PLUS, NULL, NULL); + + MMIO_DH(PLANE_NV12_BUF_CFG(PIPE_B, 0), D_SKL_PLUS, NULL, NULL); + MMIO_DH(PLANE_NV12_BUF_CFG(PIPE_B, 1), D_SKL_PLUS, NULL, NULL); + MMIO_DH(PLANE_NV12_BUF_CFG(PIPE_B, 2), D_SKL_PLUS, NULL, NULL); + MMIO_DH(PLANE_NV12_BUF_CFG(PIPE_B, 3), D_SKL_PLUS, NULL, NULL); + + MMIO_DH(PLANE_NV12_BUF_CFG(PIPE_C, 0), D_SKL_PLUS, NULL, NULL); + MMIO_DH(PLANE_NV12_BUF_CFG(PIPE_C, 1), D_SKL_PLUS, NULL, NULL); + MMIO_DH(PLANE_NV12_BUF_CFG(PIPE_C, 2), D_SKL_PLUS, NULL, NULL); + MMIO_DH(PLANE_NV12_BUF_CFG(PIPE_C, 3), D_SKL_PLUS, NULL, NULL); + + MMIO_DH(_REG_701C0(PIPE_A, 1), D_SKL_PLUS, NULL, NULL); + MMIO_DH(_REG_701C0(PIPE_A, 2), D_SKL_PLUS, NULL, NULL); + MMIO_DH(_REG_701C0(PIPE_A, 3), D_SKL_PLUS, NULL, NULL); + MMIO_DH(_REG_701C0(PIPE_A, 4), D_SKL_PLUS, NULL, NULL); + + MMIO_DH(_REG_701C0(PIPE_B, 1), D_SKL_PLUS, NULL, NULL); + MMIO_DH(_REG_701C0(PIPE_B, 2), D_SKL_PLUS, NULL, NULL); + MMIO_DH(_REG_701C0(PIPE_B, 3), D_SKL_PLUS, NULL, NULL); + MMIO_DH(_REG_701C0(PIPE_B, 4), D_SKL_PLUS, NULL, NULL); + + MMIO_DH(_REG_701C0(PIPE_C, 1), D_SKL_PLUS, NULL, NULL); + MMIO_DH(_REG_701C0(PIPE_C, 2), D_SKL_PLUS, NULL, NULL); + MMIO_DH(_REG_701C0(PIPE_C, 3), D_SKL_PLUS, NULL, NULL); + MMIO_DH(_REG_701C0(PIPE_C, 4), D_SKL_PLUS, NULL, NULL); + + MMIO_DH(_REG_701C4(PIPE_A, 1), D_SKL_PLUS, NULL, NULL); + MMIO_DH(_REG_701C4(PIPE_A, 2), D_SKL_PLUS, NULL, NULL); + MMIO_DH(_REG_701C4(PIPE_A, 3), D_SKL_PLUS, NULL, NULL); + MMIO_DH(_REG_701C4(PIPE_A, 4), D_SKL_PLUS, NULL, NULL); + + MMIO_DH(_REG_701C4(PIPE_B, 1), D_SKL_PLUS, NULL, NULL); + MMIO_DH(_REG_701C4(PIPE_B, 2), D_SKL_PLUS, NULL, NULL); + MMIO_DH(_REG_701C4(PIPE_B, 3), D_SKL_PLUS, NULL, NULL); + MMIO_DH(_REG_701C4(PIPE_B, 4), D_SKL_PLUS, NULL, NULL); + + MMIO_DH(_REG_701C4(PIPE_C, 1), D_SKL_PLUS, NULL, NULL); + MMIO_DH(_REG_701C4(PIPE_C, 2), D_SKL_PLUS, NULL, NULL); + MMIO_DH(_REG_701C4(PIPE_C, 3), D_SKL_PLUS, NULL, NULL); + MMIO_DH(_REG_701C4(PIPE_C, 4), D_SKL_PLUS, NULL, NULL); + + MMIO_D(0x70380, D_SKL_PLUS); + MMIO_D(0x71380, D_SKL_PLUS); + MMIO_D(0x72380, D_SKL_PLUS); + MMIO_D(0x7039c, D_SKL_PLUS); + + MMIO_F(0x80000, 0x3000, 0, 0, 0, D_SKL_PLUS, NULL, NULL); + MMIO_D(0x8f074, D_SKL | D_KBL); + MMIO_D(0x8f004, D_SKL | D_KBL); + MMIO_D(0x8f034, D_SKL | D_KBL); + + MMIO_D(0xb11c, D_SKL | D_KBL); + + MMIO_D(0x51000, D_SKL | D_KBL); + MMIO_D(0x6c00c, D_SKL_PLUS); + + MMIO_F(0xc800, 0x7f8, F_CMD_ACCESS, 0, 0, D_SKL | D_KBL, NULL, NULL); + MMIO_F(0xb020, 0x80, F_CMD_ACCESS, 0, 0, D_SKL | D_KBL, NULL, NULL); + + MMIO_D(0xd08, D_SKL_PLUS); + MMIO_DFH(0x20e0, D_SKL_PLUS, F_MODE_MASK, NULL, NULL); + MMIO_DFH(0x20ec, D_SKL_PLUS, F_MODE_MASK | F_CMD_ACCESS, NULL, NULL); /* TRTT */ - MMIO_DFH(0x4de0, D_SKL, F_CMD_ACCESS, NULL, NULL); - MMIO_DFH(0x4de4, D_SKL, F_CMD_ACCESS, NULL, NULL); - MMIO_DFH(0x4de8, D_SKL, F_CMD_ACCESS, NULL, NULL); - MMIO_DFH(0x4dec, D_SKL, F_CMD_ACCESS, NULL, NULL); - MMIO_DFH(0x4df0, D_SKL, F_CMD_ACCESS, NULL, NULL); - MMIO_DFH(0x4df4, D_SKL, F_CMD_ACCESS, NULL, gen9_trtte_write); - MMIO_DH(0x4dfc, D_SKL, NULL, gen9_trtt_chicken_write); + MMIO_DFH(0x4de0, D_SKL | D_KBL, F_CMD_ACCESS, NULL, NULL); + MMIO_DFH(0x4de4, D_SKL | D_KBL, F_CMD_ACCESS, NULL, NULL); + MMIO_DFH(0x4de8, D_SKL | D_KBL, F_CMD_ACCESS, NULL, NULL); + MMIO_DFH(0x4dec, D_SKL | D_KBL, F_CMD_ACCESS, NULL, NULL); + MMIO_DFH(0x4df0, D_SKL | D_KBL, F_CMD_ACCESS, NULL, NULL); + MMIO_DFH(0x4df4, D_SKL | D_KBL, F_CMD_ACCESS, NULL, gen9_trtte_write); + MMIO_DH(0x4dfc, D_SKL | D_KBL, NULL, gen9_trtt_chicken_write); - MMIO_D(0x45008, D_SKL); + MMIO_D(0x45008, D_SKL | D_KBL); - MMIO_D(0x46430, D_SKL); + MMIO_D(0x46430, D_SKL | D_KBL); - MMIO_D(0x46520, D_SKL); + MMIO_D(0x46520, D_SKL | D_KBL); - MMIO_D(0xc403c, D_SKL); - MMIO_D(0xb004, D_SKL); + MMIO_D(0xc403c, D_SKL | D_KBL); + MMIO_D(0xb004, D_SKL_PLUS); MMIO_DH(DMA_CTRL, D_SKL_PLUS, NULL, dma_ctrl_write); - MMIO_D(0x65900, D_SKL); - MMIO_D(0x1082c0, D_SKL); - MMIO_D(0x4068, D_SKL); - MMIO_D(0x67054, D_SKL); - MMIO_D(0x6e560, D_SKL); - MMIO_D(0x6e554, D_SKL); - MMIO_D(0x2b20, D_SKL); - MMIO_D(0x65f00, D_SKL); - MMIO_D(0x65f08, D_SKL); - MMIO_D(0x320f0, D_SKL); - - MMIO_DFH(_REG_VCS2_EXCC, D_SKL, F_CMD_ACCESS, NULL, NULL); - MMIO_D(0x70034, D_SKL); - MMIO_D(0x71034, D_SKL); - MMIO_D(0x72034, D_SKL); - - MMIO_D(_PLANE_KEYVAL_1(PIPE_A), D_SKL); - MMIO_D(_PLANE_KEYVAL_1(PIPE_B), D_SKL); - MMIO_D(_PLANE_KEYVAL_1(PIPE_C), D_SKL); - MMIO_D(_PLANE_KEYMSK_1(PIPE_A), D_SKL); - MMIO_D(_PLANE_KEYMSK_1(PIPE_B), D_SKL); - MMIO_D(_PLANE_KEYMSK_1(PIPE_C), D_SKL); - - MMIO_D(0x44500, D_SKL); + MMIO_D(0x65900, D_SKL_PLUS); + MMIO_D(0x1082c0, D_SKL | D_KBL); + MMIO_D(0x4068, D_SKL | D_KBL); + MMIO_D(0x67054, D_SKL | D_KBL); + MMIO_D(0x6e560, D_SKL | D_KBL); + MMIO_D(0x6e554, D_SKL | D_KBL); + MMIO_D(0x2b20, D_SKL | D_KBL); + MMIO_D(0x65f00, D_SKL | D_KBL); + MMIO_D(0x65f08, D_SKL | D_KBL); + MMIO_D(0x320f0, D_SKL | D_KBL); + + MMIO_DFH(_REG_VCS2_EXCC, D_SKL_PLUS, F_CMD_ACCESS, NULL, NULL); + MMIO_DFH(_REG_VECS_EXCC, D_SKL_PLUS, F_CMD_ACCESS, NULL, NULL); + MMIO_D(0x70034, D_SKL_PLUS); + MMIO_D(0x71034, D_SKL_PLUS); + MMIO_D(0x72034, D_SKL_PLUS); + + MMIO_D(_PLANE_KEYVAL_1(PIPE_A), D_SKL_PLUS); + MMIO_D(_PLANE_KEYVAL_1(PIPE_B), D_SKL_PLUS); + MMIO_D(_PLANE_KEYVAL_1(PIPE_C), D_SKL_PLUS); + MMIO_D(_PLANE_KEYMSK_1(PIPE_A), D_SKL_PLUS); + MMIO_D(_PLANE_KEYMSK_1(PIPE_B), D_SKL_PLUS); + MMIO_D(_PLANE_KEYMSK_1(PIPE_C), D_SKL_PLUS); + + MMIO_D(0x44500, D_SKL_PLUS); MMIO_DFH(GEN9_CSFE_CHICKEN1_RCS, D_SKL_PLUS, F_CMD_ACCESS, NULL, NULL); - MMIO_DFH(GEN8_HDC_CHICKEN1, D_SKL, F_MODE_MASK | F_CMD_ACCESS, + MMIO_DFH(GEN8_HDC_CHICKEN1, D_SKL | D_KBL, F_MODE_MASK | F_CMD_ACCESS, NULL, NULL); + + MMIO_D(0x4ab8, D_KBL); + MMIO_D(0x940c, D_SKL_PLUS); + MMIO_D(0x2248, D_SKL_PLUS | D_KBL); + MMIO_D(0x4ab0, D_SKL | D_KBL); + MMIO_D(0x20d4, D_SKL | D_KBL); + return 0; } @@ -2881,7 +2901,8 @@ int intel_gvt_setup_mmio_info(struct intel_gvt *gvt) ret = init_broadwell_mmio_info(gvt); if (ret) goto err; - } else if (IS_SKYLAKE(dev_priv)) { + } else if (IS_SKYLAKE(dev_priv) + || IS_KABYLAKE(dev_priv)) { ret = init_broadwell_mmio_info(gvt); if (ret) goto err; diff --git a/drivers/gpu/drm/i915/gvt/interrupt.c b/drivers/gpu/drm/i915/gvt/interrupt.c index 92bb247e3478..9d6812f0957f 100644 --- a/drivers/gpu/drm/i915/gvt/interrupt.c +++ b/drivers/gpu/drm/i915/gvt/interrupt.c @@ -580,7 +580,7 @@ static void gen8_init_irq( SET_BIT_INFO(irq, 4, PRIMARY_C_FLIP_DONE, INTEL_GVT_IRQ_INFO_DE_PIPE_C); SET_BIT_INFO(irq, 5, SPRITE_C_FLIP_DONE, INTEL_GVT_IRQ_INFO_DE_PIPE_C); - } else if (IS_SKYLAKE(gvt->dev_priv)) { + } else if (IS_SKYLAKE(gvt->dev_priv) || IS_KABYLAKE(gvt->dev_priv)) { SET_BIT_INFO(irq, 25, AUX_CHANNEL_B, INTEL_GVT_IRQ_INFO_DE_PORT); SET_BIT_INFO(irq, 26, AUX_CHANNEL_C, INTEL_GVT_IRQ_INFO_DE_PORT); SET_BIT_INFO(irq, 27, AUX_CHANNEL_D, INTEL_GVT_IRQ_INFO_DE_PORT); @@ -690,7 +690,8 @@ int intel_gvt_init_irq(struct intel_gvt *gvt) gvt_dbg_core("init irq framework\n"); - if (IS_BROADWELL(gvt->dev_priv) || IS_SKYLAKE(gvt->dev_priv)) { + if (IS_BROADWELL(gvt->dev_priv) || IS_SKYLAKE(gvt->dev_priv) + || IS_KABYLAKE(gvt->dev_priv)) { irq->ops = &gen8_irq_ops; irq->irq_map = gen8_irq_map; } else { diff --git a/drivers/gpu/drm/i915/gvt/kvmgt.c b/drivers/gpu/drm/i915/gvt/kvmgt.c index d641214578a7..42ff7ffb6066 100644 --- a/drivers/gpu/drm/i915/gvt/kvmgt.c +++ b/drivers/gpu/drm/i915/gvt/kvmgt.c @@ -295,10 +295,12 @@ static ssize_t description_show(struct kobject *kobj, struct device *dev, return 0; return sprintf(buf, "low_gm_size: %dMB\nhigh_gm_size: %dMB\n" - "fence: %d\nresolution: %s\n", + "fence: %d\nresolution: %s\n" + "weight: %d\n", BYTES_TO_MB(type->low_gm_size), BYTES_TO_MB(type->high_gm_size), - type->fence, vgpu_edid_str(type->resolution)); + type->fence, vgpu_edid_str(type->resolution), + type->weight); } static MDEV_TYPE_ATTR_RO(available_instances); @@ -1146,8 +1148,40 @@ static long intel_vgpu_ioctl(struct mdev_device *mdev, unsigned int cmd, return 0; } +static ssize_t +vgpu_id_show(struct device *dev, struct device_attribute *attr, + char *buf) +{ + struct mdev_device *mdev = mdev_from_dev(dev); + + if (mdev) { + struct intel_vgpu *vgpu = (struct intel_vgpu *) + mdev_get_drvdata(mdev); + return sprintf(buf, "%d\n", vgpu->id); + } + return sprintf(buf, "\n"); +} + +static DEVICE_ATTR_RO(vgpu_id); + +static struct attribute *intel_vgpu_attrs[] = { + &dev_attr_vgpu_id.attr, + NULL +}; + +static const struct attribute_group intel_vgpu_group = { + .name = "intel_vgpu", + .attrs = intel_vgpu_attrs, +}; + +static const struct attribute_group *intel_vgpu_groups[] = { + &intel_vgpu_group, + NULL, +}; + static const struct mdev_parent_ops intel_vgpu_ops = { .supported_type_groups = intel_vgpu_type_groups, + .mdev_attr_groups = intel_vgpu_groups, .create = intel_vgpu_create, .remove = intel_vgpu_remove, @@ -1340,13 +1374,6 @@ static int kvmgt_guest_init(struct mdev_device *mdev) static bool kvmgt_guest_exit(struct kvmgt_guest_info *info) { - struct intel_vgpu *vgpu = info->vgpu; - - if (!info) { - gvt_vgpu_err("kvmgt_guest_info invalid\n"); - return false; - } - kvm_page_track_unregister_notifier(info->kvm, &info->track_node); kvm_put_kvm(info->kvm); kvmgt_protect_table_destroy(info); diff --git a/drivers/gpu/drm/i915/gvt/mmio.h b/drivers/gpu/drm/i915/gvt/mmio.h index a3a027025cd0..7edd66f38ef9 100644 --- a/drivers/gpu/drm/i915/gvt/mmio.h +++ b/drivers/gpu/drm/i915/gvt/mmio.h @@ -44,20 +44,21 @@ struct intel_vgpu; #define D_HSW (1 << 2) #define D_BDW (1 << 3) #define D_SKL (1 << 4) +#define D_KBL (1 << 5) -#define D_GEN9PLUS (D_SKL) -#define D_GEN8PLUS (D_BDW | D_SKL) -#define D_GEN75PLUS (D_HSW | D_BDW | D_SKL) -#define D_GEN7PLUS (D_IVB | D_HSW | D_BDW | D_SKL) +#define D_GEN9PLUS (D_SKL | D_KBL) +#define D_GEN8PLUS (D_BDW | D_SKL | D_KBL) +#define D_GEN75PLUS (D_HSW | D_BDW | D_SKL | D_KBL) +#define D_GEN7PLUS (D_IVB | D_HSW | D_BDW | D_SKL | D_KBL) -#define D_SKL_PLUS (D_SKL) -#define D_BDW_PLUS (D_BDW | D_SKL) -#define D_HSW_PLUS (D_HSW | D_BDW | D_SKL) -#define D_IVB_PLUS (D_IVB | D_HSW | D_BDW | D_SKL) +#define D_SKL_PLUS (D_SKL | D_KBL) +#define D_BDW_PLUS (D_BDW | D_SKL | D_KBL) +#define D_HSW_PLUS (D_HSW | D_BDW | D_SKL | D_KBL) +#define D_IVB_PLUS (D_IVB | D_HSW | D_BDW | D_SKL | D_KBL) #define D_PRE_BDW (D_SNB | D_IVB | D_HSW) #define D_PRE_SKL (D_SNB | D_IVB | D_HSW | D_BDW) -#define D_ALL (D_SNB | D_IVB | D_HSW | D_BDW | D_SKL) +#define D_ALL (D_SNB | D_IVB | D_HSW | D_BDW | D_SKL | D_KBL) struct intel_gvt_mmio_info { u32 offset; diff --git a/drivers/gpu/drm/i915/gvt/render.c b/drivers/gpu/drm/i915/gvt/render.c index 0beb83563b08..a7b665e9bbad 100644 --- a/drivers/gpu/drm/i915/gvt/render.c +++ b/drivers/gpu/drm/i915/gvt/render.c @@ -126,6 +126,18 @@ static struct render_mmio gen9_render_mmio_list[] = { {VCS2, _MMIO(0x1c028), 0xffff, false}, {VECS, _MMIO(0x1a028), 0xffff, false}, + + {RCS, _MMIO(0x7304), 0xffff, true}, + {RCS, _MMIO(0x2248), 0x0, false}, + {RCS, _MMIO(0x940c), 0x0, false}, + {RCS, _MMIO(0x4ab8), 0x0, false}, + + {RCS, _MMIO(0x4ab0), 0x0, false}, + {RCS, _MMIO(0x20d4), 0x0, false}, + + {RCS, _MMIO(0xb004), 0x0, false}, + {RCS, _MMIO(0x20a0), 0x0, false}, + {RCS, _MMIO(0x20e4), 0xffff, false}, }; static u32 gen9_render_mocs[I915_NUM_ENGINES][64]; @@ -159,7 +171,7 @@ static void handle_tlb_pending_event(struct intel_vgpu *vgpu, int ring_id) */ fw = intel_uncore_forcewake_for_reg(dev_priv, reg, FW_REG_READ | FW_REG_WRITE); - if (ring_id == RCS && IS_SKYLAKE(dev_priv)) + if (ring_id == RCS && (IS_SKYLAKE(dev_priv) || IS_KABYLAKE(dev_priv))) fw |= FORCEWAKE_RENDER; intel_uncore_forcewake_get(dev_priv, fw); @@ -192,7 +204,7 @@ static void load_mocs(struct intel_vgpu *vgpu, int ring_id) if (WARN_ON(ring_id >= ARRAY_SIZE(regs))) return; - if (!IS_SKYLAKE(dev_priv)) + if (!(IS_SKYLAKE(dev_priv) || IS_KABYLAKE(dev_priv))) return; offset.reg = regs[ring_id]; @@ -230,7 +242,7 @@ static void restore_mocs(struct intel_vgpu *vgpu, int ring_id) if (WARN_ON(ring_id >= ARRAY_SIZE(regs))) return; - if (!IS_SKYLAKE(dev_priv)) + if (!(IS_SKYLAKE(dev_priv) || IS_KABYLAKE(dev_priv))) return; offset.reg = regs[ring_id]; @@ -265,7 +277,8 @@ void intel_gvt_load_render_mmio(struct intel_vgpu *vgpu, int ring_id) u32 inhibit_mask = _MASKED_BIT_ENABLE(CTX_CTRL_ENGINE_CTX_RESTORE_INHIBIT); - if (IS_SKYLAKE(vgpu->gvt->dev_priv)) { + if (IS_SKYLAKE(vgpu->gvt->dev_priv) + || IS_KABYLAKE(vgpu->gvt->dev_priv)) { mmio = gen9_render_mmio_list; array_size = ARRAY_SIZE(gen9_render_mmio_list); load_mocs(vgpu, ring_id); @@ -312,7 +325,7 @@ void intel_gvt_restore_render_mmio(struct intel_vgpu *vgpu, int ring_id) u32 v; int i, array_size; - if (IS_SKYLAKE(dev_priv)) { + if (IS_SKYLAKE(dev_priv) || IS_KABYLAKE(dev_priv)) { mmio = gen9_render_mmio_list; array_size = ARRAY_SIZE(gen9_render_mmio_list); restore_mocs(vgpu, ring_id); diff --git a/drivers/gpu/drm/i915/gvt/sched_policy.c b/drivers/gpu/drm/i915/gvt/sched_policy.c index 34b9acdf3479..f84959170674 100644 --- a/drivers/gpu/drm/i915/gvt/sched_policy.c +++ b/drivers/gpu/drm/i915/gvt/sched_policy.c @@ -47,11 +47,87 @@ static bool vgpu_has_pending_workload(struct intel_vgpu *vgpu) return false; } +struct vgpu_sched_data { + struct list_head lru_list; + struct intel_vgpu *vgpu; + + ktime_t sched_in_time; + ktime_t sched_out_time; + ktime_t sched_time; + ktime_t left_ts; + ktime_t allocated_ts; + + struct vgpu_sched_ctl sched_ctl; +}; + +struct gvt_sched_data { + struct intel_gvt *gvt; + struct hrtimer timer; + unsigned long period; + struct list_head lru_runq_head; +}; + +static void vgpu_update_timeslice(struct intel_vgpu *pre_vgpu) +{ + ktime_t delta_ts; + struct vgpu_sched_data *vgpu_data = pre_vgpu->sched_data; + + delta_ts = vgpu_data->sched_out_time - vgpu_data->sched_in_time; + + vgpu_data->sched_time += delta_ts; + vgpu_data->left_ts -= delta_ts; +} + +#define GVT_TS_BALANCE_PERIOD_MS 100 +#define GVT_TS_BALANCE_STAGE_NUM 10 + +static void gvt_balance_timeslice(struct gvt_sched_data *sched_data) +{ + struct vgpu_sched_data *vgpu_data; + struct list_head *pos; + static uint64_t stage_check; + int stage = stage_check++ % GVT_TS_BALANCE_STAGE_NUM; + + /* The timeslice accumulation reset at stage 0, which is + * allocated again without adding previous debt. + */ + if (stage == 0) { + int total_weight = 0; + ktime_t fair_timeslice; + + list_for_each(pos, &sched_data->lru_runq_head) { + vgpu_data = container_of(pos, struct vgpu_sched_data, lru_list); + total_weight += vgpu_data->sched_ctl.weight; + } + + list_for_each(pos, &sched_data->lru_runq_head) { + vgpu_data = container_of(pos, struct vgpu_sched_data, lru_list); + fair_timeslice = ms_to_ktime(GVT_TS_BALANCE_PERIOD_MS) * + vgpu_data->sched_ctl.weight / + total_weight; + + vgpu_data->allocated_ts = fair_timeslice; + vgpu_data->left_ts = vgpu_data->allocated_ts; + } + } else { + list_for_each(pos, &sched_data->lru_runq_head) { + vgpu_data = container_of(pos, struct vgpu_sched_data, lru_list); + + /* timeslice for next 100ms should add the left/debt + * slice of previous stages. + */ + vgpu_data->left_ts += vgpu_data->allocated_ts; + } + } +} + static void try_to_schedule_next_vgpu(struct intel_gvt *gvt) { struct intel_gvt_workload_scheduler *scheduler = &gvt->scheduler; enum intel_engine_id i; struct intel_engine_cs *engine; + struct vgpu_sched_data *vgpu_data; + ktime_t cur_time; /* no target to schedule */ if (!scheduler->next_vgpu) @@ -77,6 +153,15 @@ static void try_to_schedule_next_vgpu(struct intel_gvt *gvt) gvt_dbg_sched("switch to next vgpu %d\n", scheduler->next_vgpu->id); + cur_time = ktime_get(); + if (scheduler->current_vgpu) { + vgpu_data = scheduler->current_vgpu->sched_data; + vgpu_data->sched_out_time = cur_time; + vgpu_update_timeslice(scheduler->current_vgpu); + } + vgpu_data = scheduler->next_vgpu->sched_data; + vgpu_data->sched_in_time = cur_time; + /* switch current vgpu */ scheduler->current_vgpu = scheduler->next_vgpu; scheduler->next_vgpu = NULL; @@ -88,62 +173,61 @@ static void try_to_schedule_next_vgpu(struct intel_gvt *gvt) wake_up(&scheduler->waitq[i]); } -struct tbs_vgpu_data { - struct list_head list; - struct intel_vgpu *vgpu; - /* put some per-vgpu sched stats here */ -}; - -struct tbs_sched_data { - struct intel_gvt *gvt; - struct delayed_work work; - unsigned long period; - struct list_head runq_head; -}; - -#define GVT_DEFAULT_TIME_SLICE (msecs_to_jiffies(1)) - -static void tbs_sched_func(struct work_struct *work) +static struct intel_vgpu *find_busy_vgpu(struct gvt_sched_data *sched_data) { - struct tbs_sched_data *sched_data = container_of(work, - struct tbs_sched_data, work.work); - struct tbs_vgpu_data *vgpu_data; - - struct intel_gvt *gvt = sched_data->gvt; - struct intel_gvt_workload_scheduler *scheduler = &gvt->scheduler; - + struct vgpu_sched_data *vgpu_data; struct intel_vgpu *vgpu = NULL; - struct list_head *pos, *head; - - mutex_lock(&gvt->lock); - - /* no vgpu or has already had a target */ - if (list_empty(&sched_data->runq_head) || scheduler->next_vgpu) - goto out; - - if (scheduler->current_vgpu) { - vgpu_data = scheduler->current_vgpu->sched_data; - head = &vgpu_data->list; - } else { - head = &sched_data->runq_head; - } + struct list_head *head = &sched_data->lru_runq_head; + struct list_head *pos; /* search a vgpu with pending workload */ list_for_each(pos, head) { - if (pos == &sched_data->runq_head) - continue; - vgpu_data = container_of(pos, struct tbs_vgpu_data, list); + vgpu_data = container_of(pos, struct vgpu_sched_data, lru_list); if (!vgpu_has_pending_workload(vgpu_data->vgpu)) continue; - vgpu = vgpu_data->vgpu; - break; + /* Return the vGPU only if it has time slice left */ + if (vgpu_data->left_ts > 0) { + vgpu = vgpu_data->vgpu; + break; + } } + return vgpu; +} + +/* in nanosecond */ +#define GVT_DEFAULT_TIME_SLICE 1000000 + +static void tbs_sched_func(struct gvt_sched_data *sched_data) +{ + struct intel_gvt *gvt = sched_data->gvt; + struct intel_gvt_workload_scheduler *scheduler = &gvt->scheduler; + struct vgpu_sched_data *vgpu_data; + struct intel_vgpu *vgpu = NULL; + static uint64_t timer_check; + + if (!(timer_check++ % GVT_TS_BALANCE_PERIOD_MS)) + gvt_balance_timeslice(sched_data); + + /* no active vgpu or has already had a target */ + if (list_empty(&sched_data->lru_runq_head) || scheduler->next_vgpu) + goto out; + + vgpu = find_busy_vgpu(sched_data); if (vgpu) { scheduler->next_vgpu = vgpu; + + /* Move the last used vGPU to the tail of lru_list */ + vgpu_data = vgpu->sched_data; + list_del_init(&vgpu_data->lru_list); + list_add_tail(&vgpu_data->lru_list, + &sched_data->lru_runq_head); + gvt_dbg_sched("pick next vgpu %d\n", vgpu->id); + } else { + scheduler->next_vgpu = gvt->idle_vgpu; } out: if (scheduler->next_vgpu) { @@ -151,34 +235,49 @@ out: scheduler->next_vgpu->id); try_to_schedule_next_vgpu(gvt); } +} - /* - * still have vgpu on runq - * or last schedule haven't finished due to running workload - */ - if (!list_empty(&sched_data->runq_head) || scheduler->next_vgpu) - schedule_delayed_work(&sched_data->work, sched_data->period); +void intel_gvt_schedule(struct intel_gvt *gvt) +{ + struct gvt_sched_data *sched_data = gvt->scheduler.sched_data; + mutex_lock(&gvt->lock); + tbs_sched_func(sched_data); mutex_unlock(&gvt->lock); } +static enum hrtimer_restart tbs_timer_fn(struct hrtimer *timer_data) +{ + struct gvt_sched_data *data; + + data = container_of(timer_data, struct gvt_sched_data, timer); + + intel_gvt_request_service(data->gvt, INTEL_GVT_REQUEST_SCHED); + + hrtimer_add_expires_ns(&data->timer, data->period); + + return HRTIMER_RESTART; +} + static int tbs_sched_init(struct intel_gvt *gvt) { struct intel_gvt_workload_scheduler *scheduler = &gvt->scheduler; - struct tbs_sched_data *data; + struct gvt_sched_data *data; data = kzalloc(sizeof(*data), GFP_KERNEL); if (!data) return -ENOMEM; - INIT_LIST_HEAD(&data->runq_head); - INIT_DELAYED_WORK(&data->work, tbs_sched_func); + INIT_LIST_HEAD(&data->lru_runq_head); + hrtimer_init(&data->timer, CLOCK_MONOTONIC, HRTIMER_MODE_ABS); + data->timer.function = tbs_timer_fn; data->period = GVT_DEFAULT_TIME_SLICE; data->gvt = gvt; scheduler->sched_data = data; + return 0; } @@ -186,25 +285,28 @@ static void tbs_sched_clean(struct intel_gvt *gvt) { struct intel_gvt_workload_scheduler *scheduler = &gvt->scheduler; - struct tbs_sched_data *data = scheduler->sched_data; + struct gvt_sched_data *data = scheduler->sched_data; + + hrtimer_cancel(&data->timer); - cancel_delayed_work(&data->work); kfree(data); scheduler->sched_data = NULL; } static int tbs_sched_init_vgpu(struct intel_vgpu *vgpu) { - struct tbs_vgpu_data *data; + struct vgpu_sched_data *data; data = kzalloc(sizeof(*data), GFP_KERNEL); if (!data) return -ENOMEM; + data->sched_ctl.weight = vgpu->sched_ctl.weight; data->vgpu = vgpu; - INIT_LIST_HEAD(&data->list); + INIT_LIST_HEAD(&data->lru_list); vgpu->sched_data = data; + return 0; } @@ -216,21 +318,24 @@ static void tbs_sched_clean_vgpu(struct intel_vgpu *vgpu) static void tbs_sched_start_schedule(struct intel_vgpu *vgpu) { - struct tbs_sched_data *sched_data = vgpu->gvt->scheduler.sched_data; - struct tbs_vgpu_data *vgpu_data = vgpu->sched_data; + struct gvt_sched_data *sched_data = vgpu->gvt->scheduler.sched_data; + struct vgpu_sched_data *vgpu_data = vgpu->sched_data; - if (!list_empty(&vgpu_data->list)) + if (!list_empty(&vgpu_data->lru_list)) return; - list_add_tail(&vgpu_data->list, &sched_data->runq_head); - schedule_delayed_work(&sched_data->work, 0); + list_add_tail(&vgpu_data->lru_list, &sched_data->lru_runq_head); + + if (!hrtimer_active(&sched_data->timer)) + hrtimer_start(&sched_data->timer, ktime_add_ns(ktime_get(), + sched_data->period), HRTIMER_MODE_ABS); } static void tbs_sched_stop_schedule(struct intel_vgpu *vgpu) { - struct tbs_vgpu_data *vgpu_data = vgpu->sched_data; + struct vgpu_sched_data *vgpu_data = vgpu->sched_data; - list_del_init(&vgpu_data->list); + list_del_init(&vgpu_data->lru_list); } static struct intel_gvt_sched_policy_ops tbs_schedule_ops = { diff --git a/drivers/gpu/drm/i915/gvt/sched_policy.h b/drivers/gpu/drm/i915/gvt/sched_policy.h index bb8b9097e41a..ba00a5f7455f 100644 --- a/drivers/gpu/drm/i915/gvt/sched_policy.h +++ b/drivers/gpu/drm/i915/gvt/sched_policy.h @@ -43,6 +43,8 @@ struct intel_gvt_sched_policy_ops { void (*stop_schedule)(struct intel_vgpu *vgpu); }; +void intel_gvt_schedule(struct intel_gvt *gvt); + int intel_gvt_init_sched_policy(struct intel_gvt *gvt); void intel_gvt_clean_sched_policy(struct intel_gvt *gvt); diff --git a/drivers/gpu/drm/i915/gvt/scheduler.c b/drivers/gpu/drm/i915/gvt/scheduler.c index d9a735310e75..a77db2332e68 100644 --- a/drivers/gpu/drm/i915/gvt/scheduler.c +++ b/drivers/gpu/drm/i915/gvt/scheduler.c @@ -453,7 +453,8 @@ static int workload_thread(void *priv) struct intel_vgpu_workload *workload = NULL; struct intel_vgpu *vgpu = NULL; int ret; - bool need_force_wake = IS_SKYLAKE(gvt->dev_priv); + bool need_force_wake = IS_SKYLAKE(gvt->dev_priv) + || IS_KABYLAKE(gvt->dev_priv); DEFINE_WAIT_FUNC(wait, woken_wake_function); kfree(p); diff --git a/drivers/gpu/drm/i915/gvt/scheduler.h b/drivers/gpu/drm/i915/gvt/scheduler.h index 2833dfa8c9ae..2cd725c0573e 100644 --- a/drivers/gpu/drm/i915/gvt/scheduler.h +++ b/drivers/gpu/drm/i915/gvt/scheduler.h @@ -67,7 +67,6 @@ struct shadow_per_ctx { }; struct intel_shadow_wa_ctx { - struct intel_vgpu_workload *workload; struct shadow_indirect_ctx indirect_ctx; struct shadow_per_ctx per_ctx; diff --git a/drivers/gpu/drm/i915/gvt/vgpu.c b/drivers/gpu/drm/i915/gvt/vgpu.c index 41cfa5ccae84..6ba02525e905 100644 --- a/drivers/gpu/drm/i915/gvt/vgpu.c +++ b/drivers/gpu/drm/i915/gvt/vgpu.c @@ -64,18 +64,28 @@ void populate_pvinfo_page(struct intel_vgpu *vgpu) WARN_ON(sizeof(struct vgt_if) != VGT_PVINFO_SIZE); } +#define VGPU_MAX_WEIGHT 16 +#define VGPU_WEIGHT(vgpu_num) \ + (VGPU_MAX_WEIGHT / (vgpu_num)) + static struct { unsigned int low_mm; unsigned int high_mm; unsigned int fence; + + /* A vGPU with a weight of 8 will get twice as much GPU as a vGPU + * with a weight of 4 on a contended host, different vGPU type has + * different weight set. Legal weights range from 1 to 16. + */ + unsigned int weight; enum intel_vgpu_edid edid; char *name; } vgpu_types[] = { /* Fixed vGPU type table */ - { MB_TO_BYTES(64), MB_TO_BYTES(512), 4, GVT_EDID_1024_768, "8" }, - { MB_TO_BYTES(128), MB_TO_BYTES(512), 4, GVT_EDID_1920_1200, "4" }, - { MB_TO_BYTES(256), MB_TO_BYTES(1024), 4, GVT_EDID_1920_1200, "2" }, - { MB_TO_BYTES(512), MB_TO_BYTES(2048), 4, GVT_EDID_1920_1200, "1" }, + { MB_TO_BYTES(64), MB_TO_BYTES(384), 4, VGPU_WEIGHT(8), GVT_EDID_1024_768, "8" }, + { MB_TO_BYTES(128), MB_TO_BYTES(512), 4, VGPU_WEIGHT(4), GVT_EDID_1920_1200, "4" }, + { MB_TO_BYTES(256), MB_TO_BYTES(1024), 4, VGPU_WEIGHT(2), GVT_EDID_1920_1200, "2" }, + { MB_TO_BYTES(512), MB_TO_BYTES(2048), 4, VGPU_WEIGHT(1), GVT_EDID_1920_1200, "1" }, }; /** @@ -120,6 +130,12 @@ int intel_gvt_init_vgpu_types(struct intel_gvt *gvt) gvt->types[i].low_gm_size = vgpu_types[i].low_mm; gvt->types[i].high_gm_size = vgpu_types[i].high_mm; gvt->types[i].fence = vgpu_types[i].fence; + + if (vgpu_types[i].weight < 1 || + vgpu_types[i].weight > VGPU_MAX_WEIGHT) + return -EINVAL; + + gvt->types[i].weight = vgpu_types[i].weight; gvt->types[i].resolution = vgpu_types[i].edid; gvt->types[i].avail_instance = min(low_avail / vgpu_types[i].low_mm, high_avail / vgpu_types[i].high_mm); @@ -131,11 +147,12 @@ int intel_gvt_init_vgpu_types(struct intel_gvt *gvt) sprintf(gvt->types[i].name, "GVTg_V5_%s", vgpu_types[i].name); - gvt_dbg_core("type[%d]: %s avail %u low %u high %u fence %u res %s\n", + gvt_dbg_core("type[%d]: %s avail %u low %u high %u fence %u weight %u res %s\n", i, gvt->types[i].name, gvt->types[i].avail_instance, gvt->types[i].low_gm_size, gvt->types[i].high_gm_size, gvt->types[i].fence, + gvt->types[i].weight, vgpu_edid_str(gvt->types[i].resolution)); } @@ -216,6 +233,59 @@ void intel_gvt_destroy_vgpu(struct intel_vgpu *vgpu) mutex_unlock(&gvt->lock); } +#define IDLE_VGPU_IDR 0 + +/** + * intel_gvt_create_idle_vgpu - create an idle virtual GPU + * @gvt: GVT device + * + * This function is called when user wants to create an idle virtual GPU. + * + * Returns: + * pointer to intel_vgpu, error pointer if failed. + */ +struct intel_vgpu *intel_gvt_create_idle_vgpu(struct intel_gvt *gvt) +{ + struct intel_vgpu *vgpu; + enum intel_engine_id i; + int ret; + + vgpu = vzalloc(sizeof(*vgpu)); + if (!vgpu) + return ERR_PTR(-ENOMEM); + + vgpu->id = IDLE_VGPU_IDR; + vgpu->gvt = gvt; + + for (i = 0; i < I915_NUM_ENGINES; i++) + INIT_LIST_HEAD(&vgpu->workload_q_head[i]); + + ret = intel_vgpu_init_sched_policy(vgpu); + if (ret) + goto out_free_vgpu; + + vgpu->active = false; + + return vgpu; + +out_free_vgpu: + vfree(vgpu); + return ERR_PTR(ret); +} + +/** + * intel_gvt_destroy_vgpu - destroy an idle virtual GPU + * @vgpu: virtual GPU + * + * This function is called when user wants to destroy an idle virtual GPU. + * + */ +void intel_gvt_destroy_idle_vgpu(struct intel_vgpu *vgpu) +{ + intel_vgpu_clean_sched_policy(vgpu); + vfree(vgpu); +} + static struct intel_vgpu *__intel_gvt_create_vgpu(struct intel_gvt *gvt, struct intel_vgpu_creation_params *param) { @@ -232,13 +302,15 @@ static struct intel_vgpu *__intel_gvt_create_vgpu(struct intel_gvt *gvt, mutex_lock(&gvt->lock); - ret = idr_alloc(&gvt->vgpu_idr, vgpu, 1, GVT_MAX_VGPU, GFP_KERNEL); + ret = idr_alloc(&gvt->vgpu_idr, vgpu, IDLE_VGPU_IDR + 1, GVT_MAX_VGPU, + GFP_KERNEL); if (ret < 0) goto out_free_vgpu; vgpu->id = ret; vgpu->handle = param->handle; vgpu->gvt = gvt; + vgpu->sched_ctl.weight = param->weight; bitmap_zero(vgpu->tlb_handle_pending, I915_NUM_ENGINES); intel_vgpu_init_cfg_space(vgpu, param->primary); @@ -325,6 +397,7 @@ struct intel_vgpu *intel_gvt_create_vgpu(struct intel_gvt *gvt, param.low_gm_sz = type->low_gm_size; param.high_gm_sz = type->high_gm_size; param.fence_sz = type->fence; + param.weight = type->weight; param.resolution = type->resolution; /* XXX current param based on MB */ diff --git a/drivers/gpu/drm/i915/i915_debugfs.c b/drivers/gpu/drm/i915/i915_debugfs.c index 47e707d83c4d..d689e511744e 100644 --- a/drivers/gpu/drm/i915/i915_debugfs.c +++ b/drivers/gpu/drm/i915/i915_debugfs.c @@ -1012,9 +1012,12 @@ static int gpu_state_release(struct inode *inode, struct file *file) static int i915_gpu_info_open(struct inode *inode, struct file *file) { + struct drm_i915_private *i915 = inode->i_private; struct i915_gpu_state *gpu; - gpu = i915_capture_gpu_state(inode->i_private); + intel_runtime_pm_get(i915); + gpu = i915_capture_gpu_state(i915); + intel_runtime_pm_put(i915); if (!gpu) return -ENOMEM; @@ -1459,16 +1462,14 @@ static int ironlake_drpc_info(struct seq_file *m) static int i915_forcewake_domains(struct seq_file *m, void *data) { - struct drm_i915_private *dev_priv = node_to_i915(m->private); + struct drm_i915_private *i915 = node_to_i915(m->private); struct intel_uncore_forcewake_domain *fw_domain; + unsigned int tmp; - spin_lock_irq(&dev_priv->uncore.lock); - for_each_fw_domain(fw_domain, dev_priv) { + for_each_fw_domain(fw_domain, i915, tmp) seq_printf(m, "%s.wake_count = %u\n", intel_uncore_forcewake_domain_to_str(fw_domain->id), - fw_domain->wake_count); - } - spin_unlock_irq(&dev_priv->uncore.lock); + READ_ONCE(fw_domain->wake_count)); return 0; } @@ -1938,9 +1939,8 @@ static int i915_gem_framebuffer_info(struct seq_file *m, void *data) static void describe_ctx_ring(struct seq_file *m, struct intel_ring *ring) { - seq_printf(m, " (ringbuffer, space: %d, head: %u, tail: %u, last head: %d)", - ring->space, ring->head, ring->tail, - ring->last_retired_head); + seq_printf(m, " (ringbuffer, space: %d, head: %u, tail: %u)", + ring->space, ring->head, ring->tail); } static int i915_context_status(struct seq_file *m, void *unused) @@ -2474,9 +2474,9 @@ static void i915_guc_client_info(struct seq_file *m, enum intel_engine_id id; uint64_t tot = 0; - seq_printf(m, "\tPriority %d, GuC ctx index: %u, PD offset 0x%x\n", - client->priority, client->ctx_index, client->proc_desc_offset); - seq_printf(m, "\tDoorbell id %d, offset: 0x%x, cookie 0x%x\n", + seq_printf(m, "\tPriority %d, GuC stage index: %u, PD offset 0x%x\n", + client->priority, client->stage_id, client->proc_desc_offset); + seq_printf(m, "\tDoorbell id %d, offset: 0x%lx, cookie 0x%x\n", client->doorbell_id, client->doorbell_offset, client->doorbell_cookie); seq_printf(m, "\tWQ size %d, offset: 0x%x, tail %d\n", client->wq_size, client->wq_offset, client->wq_tail); @@ -2511,7 +2511,7 @@ static int i915_guc_info(struct seq_file *m, void *data) } seq_printf(m, "Doorbell map:\n"); - seq_printf(m, "\t%*pb\n", GUC_MAX_DOORBELLS, guc->doorbell_bitmap); + seq_printf(m, "\t%*pb\n", GUC_NUM_DOORBELLS, guc->doorbell_bitmap); seq_printf(m, "Doorbell next cacheline: 0x%x\n\n", guc->db_cacheline); seq_printf(m, "GuC total action count: %llu\n", guc->action_count); @@ -4129,7 +4129,9 @@ i915_wedged_get(void *data, u64 *val) static int i915_wedged_set(void *data, u64 val) { - struct drm_i915_private *dev_priv = data; + struct drm_i915_private *i915 = data; + struct intel_engine_cs *engine; + unsigned int tmp; /* * There is no safeguard against this debugfs entry colliding @@ -4139,13 +4141,17 @@ i915_wedged_set(void *data, u64 val) * while it is writing to 'i915_wedged' */ - if (i915_reset_backoff(&dev_priv->gpu_error)) + if (i915_reset_backoff(&i915->gpu_error)) return -EAGAIN; - i915_handle_error(dev_priv, val, - "Manually setting wedged to %llu", val); + for_each_engine_masked(engine, i915, val, tmp) { + engine->hangcheck.seqno = intel_engine_get_seqno(engine); + engine->hangcheck.stalled = true; + } + + i915_handle_error(i915, val, "Manually setting wedged to %llu", val); - wait_on_bit(&dev_priv->gpu_error.flags, + wait_on_bit(&i915->gpu_error.flags, I915_RESET_HANDOFF, TASK_UNINTERRUPTIBLE); @@ -4173,10 +4179,6 @@ fault_irq_set(struct drm_i915_private *i915, if (err) goto err_unlock; - /* Retire to kick idle work */ - i915_gem_retire_requests(i915); - GEM_BUG_ON(i915->gt.active_requests); - *irq = val; mutex_unlock(&i915->drm.struct_mutex); @@ -4280,7 +4282,7 @@ i915_drop_caches_set(void *data, u64 val) goto unlock; } - if (val & (DROP_RETIRE | DROP_ACTIVE)) + if (val & DROP_RETIRE) i915_gem_retire_requests(dev_priv); lockdep_set_current_reclaim_state(GFP_KERNEL); diff --git a/drivers/gpu/drm/i915/i915_drv.c b/drivers/gpu/drm/i915/i915_drv.c index 98b17070a123..c616b4e755bc 100644 --- a/drivers/gpu/drm/i915/i915_drv.c +++ b/drivers/gpu/drm/i915/i915_drv.c @@ -549,6 +549,7 @@ static const struct vga_switcheroo_client_ops i915_switcheroo_ops = { static void i915_gem_fini(struct drm_i915_private *dev_priv) { mutex_lock(&dev_priv->drm.struct_mutex); + intel_uc_fini_hw(dev_priv); i915_gem_cleanup_engines(dev_priv); i915_gem_context_fini(dev_priv); mutex_unlock(&dev_priv->drm.struct_mutex); @@ -609,7 +610,7 @@ static int i915_load_modeset_init(struct drm_device *dev) ret = i915_gem_init(dev_priv); if (ret) - goto cleanup_irq; + goto cleanup_uc; intel_modeset_gem_init(dev); @@ -631,9 +632,9 @@ cleanup_gem: if (i915_gem_suspend(dev_priv)) DRM_ERROR("failed to idle hardware; continuing to unload!\n"); i915_gem_fini(dev_priv); +cleanup_uc: + intel_uc_fini_fw(dev_priv); cleanup_irq: - intel_guc_fini(dev_priv); - intel_huc_fini(dev_priv); drm_irq_uninstall(dev); intel_teardown_gmbus(dev_priv); cleanup_csr: @@ -1351,9 +1352,8 @@ void i915_driver_unload(struct drm_device *dev) /* Flush any outstanding unpin_work. */ drain_workqueue(dev_priv->wq); - intel_guc_fini(dev_priv); - intel_huc_fini(dev_priv); i915_gem_fini(dev_priv); + intel_uc_fini_fw(dev_priv); intel_fbc_cleanup_cfb(dev_priv); intel_power_domains_fini(dev_priv); diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h index a5947a496d0a..c9b0949f6c1a 100644 --- a/drivers/gpu/drm/i915/i915_drv.h +++ b/drivers/gpu/drm/i915/i915_drv.h @@ -79,26 +79,8 @@ #define DRIVER_NAME "i915" #define DRIVER_DESC "Intel Graphics" -#define DRIVER_DATE "20170320" -#define DRIVER_TIMESTAMP 1489994464 - -#undef WARN_ON -/* Many gcc seem to no see through this and fall over :( */ -#if 0 -#define WARN_ON(x) ({ \ - bool __i915_warn_cond = (x); \ - if (__builtin_constant_p(__i915_warn_cond)) \ - BUILD_BUG_ON(__i915_warn_cond); \ - WARN(__i915_warn_cond, "WARN_ON(" #x ")"); }) -#else -#define WARN_ON(x) WARN((x), "%s", "WARN_ON(" __stringify(x) ")") -#endif - -#undef WARN_ON_ONCE -#define WARN_ON_ONCE(x) WARN_ONCE((x), "%s", "WARN_ON_ONCE(" __stringify(x) ")") - -#define MISSING_CASE(x) WARN(1, "Missing switch case (%lu) in %s\n", \ - (long) (x), __func__); +#define DRIVER_DATE "20170403" +#define DRIVER_TIMESTAMP 1491198738 /* Use I915_STATE_WARN(x) and I915_STATE_WARN_ON() (rather than WARN() and * WARN_ON()) for hw state sanity checks to check for unexpected conditions @@ -703,9 +685,9 @@ enum forcewake_domain_id { }; enum forcewake_domains { - FORCEWAKE_RENDER = (1 << FW_DOMAIN_ID_RENDER), - FORCEWAKE_BLITTER = (1 << FW_DOMAIN_ID_BLITTER), - FORCEWAKE_MEDIA = (1 << FW_DOMAIN_ID_MEDIA), + FORCEWAKE_RENDER = BIT(FW_DOMAIN_ID_RENDER), + FORCEWAKE_BLITTER = BIT(FW_DOMAIN_ID_BLITTER), + FORCEWAKE_MEDIA = BIT(FW_DOMAIN_ID_MEDIA), FORCEWAKE_ALL = (FORCEWAKE_RENDER | FORCEWAKE_BLITTER | FORCEWAKE_MEDIA) @@ -732,21 +714,25 @@ intel_uncore_forcewake_for_reg(struct drm_i915_private *dev_priv, struct intel_uncore_funcs { void (*force_wake_get)(struct drm_i915_private *dev_priv, - enum forcewake_domains domains); + enum forcewake_domains domains); void (*force_wake_put)(struct drm_i915_private *dev_priv, - enum forcewake_domains domains); - - uint8_t (*mmio_readb)(struct drm_i915_private *dev_priv, i915_reg_t r, bool trace); - uint16_t (*mmio_readw)(struct drm_i915_private *dev_priv, i915_reg_t r, bool trace); - uint32_t (*mmio_readl)(struct drm_i915_private *dev_priv, i915_reg_t r, bool trace); - uint64_t (*mmio_readq)(struct drm_i915_private *dev_priv, i915_reg_t r, bool trace); - - void (*mmio_writeb)(struct drm_i915_private *dev_priv, i915_reg_t r, - uint8_t val, bool trace); - void (*mmio_writew)(struct drm_i915_private *dev_priv, i915_reg_t r, - uint16_t val, bool trace); - void (*mmio_writel)(struct drm_i915_private *dev_priv, i915_reg_t r, - uint32_t val, bool trace); + enum forcewake_domains domains); + + uint8_t (*mmio_readb)(struct drm_i915_private *dev_priv, + i915_reg_t r, bool trace); + uint16_t (*mmio_readw)(struct drm_i915_private *dev_priv, + i915_reg_t r, bool trace); + uint32_t (*mmio_readl)(struct drm_i915_private *dev_priv, + i915_reg_t r, bool trace); + uint64_t (*mmio_readq)(struct drm_i915_private *dev_priv, + i915_reg_t r, bool trace); + + void (*mmio_writeb)(struct drm_i915_private *dev_priv, + i915_reg_t r, uint8_t val, bool trace); + void (*mmio_writew)(struct drm_i915_private *dev_priv, + i915_reg_t r, uint16_t val, bool trace); + void (*mmio_writel)(struct drm_i915_private *dev_priv, + i915_reg_t r, uint32_t val, bool trace); }; struct intel_forcewake_range { @@ -770,32 +756,35 @@ struct intel_uncore { enum forcewake_domains fw_domains; enum forcewake_domains fw_domains_active; + u32 fw_set; + u32 fw_clear; + u32 fw_reset; + struct intel_uncore_forcewake_domain { - struct drm_i915_private *i915; enum forcewake_domain_id id; enum forcewake_domains mask; unsigned wake_count; struct hrtimer timer; i915_reg_t reg_set; - u32 val_set; - u32 val_clear; i915_reg_t reg_ack; - i915_reg_t reg_post; - u32 val_reset; } fw_domain[FW_DOMAIN_ID_COUNT]; int unclaimed_mmio_check; }; +#define __mask_next_bit(mask) ({ \ + int __idx = ffs(mask) - 1; \ + mask &= ~BIT(__idx); \ + __idx; \ +}) + /* Iterate over initialised fw domains */ -#define for_each_fw_domain_masked(domain__, mask__, dev_priv__) \ - for ((domain__) = &(dev_priv__)->uncore.fw_domain[0]; \ - (domain__) < &(dev_priv__)->uncore.fw_domain[FW_DOMAIN_ID_COUNT]; \ - (domain__)++) \ - for_each_if ((mask__) & (domain__)->mask) +#define for_each_fw_domain_masked(domain__, mask__, dev_priv__, tmp__) \ + for (tmp__ = (mask__); \ + tmp__ ? (domain__ = &(dev_priv__)->uncore.fw_domain[__mask_next_bit(tmp__)]), 1 : 0;) -#define for_each_fw_domain(domain__, dev_priv__) \ - for_each_fw_domain_masked(domain__, FORCEWAKE_ALL, dev_priv__) +#define for_each_fw_domain(domain__, dev_priv__, tmp__) \ + for_each_fw_domain_masked(domain__, (dev_priv__)->uncore.fw_domains, dev_priv__, tmp__) #define CSR_VERSION(major, minor) ((major) << 16 | (minor)) #define CSR_VERSION_MAJOR(version) ((version) >> 16) @@ -846,6 +835,7 @@ struct intel_csr { func(has_resource_streamer); \ func(has_runtime_pm); \ func(has_snoop); \ + func(unfenced_needs_alignment); \ func(cursor_needs_physical); \ func(hws_needs_physical); \ func(overlay_needs_physical); \ @@ -2578,12 +2568,6 @@ static inline struct drm_i915_private *huc_to_i915(struct intel_huc *huc) (id__)++) \ for_each_if ((engine__) = (dev_priv__)->engine[(id__)]) -#define __mask_next_bit(mask) ({ \ - int __idx = ffs(mask) - 1; \ - mask &= ~BIT(__idx); \ - __idx; \ -}) - /* Iterator over subset of engines selected by mask */ #define for_each_engine_masked(engine__, dev_priv__, mask__, tmp__) \ for (tmp__ = mask__ & INTEL_INFO(dev_priv__)->ring_mask; \ @@ -3956,14 +3940,14 @@ u64 intel_rc6_residency_us(struct drm_i915_private *dev_priv, #define POSTING_READ16(reg) (void)I915_READ16_NOTRACE(reg) #define __raw_read(x, s) \ -static inline uint##x##_t __raw_i915_read##x(struct drm_i915_private *dev_priv, \ +static inline uint##x##_t __raw_i915_read##x(const struct drm_i915_private *dev_priv, \ i915_reg_t reg) \ { \ return read##s(dev_priv->regs + i915_mmio_reg_offset(reg)); \ } #define __raw_write(x, s) \ -static inline void __raw_i915_write##x(struct drm_i915_private *dev_priv, \ +static inline void __raw_i915_write##x(const struct drm_i915_private *dev_priv, \ i915_reg_t reg, uint##x##_t val) \ { \ write##s(val, dev_priv->regs + i915_mmio_reg_offset(reg)); \ diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c index 58e1db77d70e..bbc6f1c9f175 100644 --- a/drivers/gpu/drm/i915/i915_gem.c +++ b/drivers/gpu/drm/i915/i915_gem.c @@ -2321,7 +2321,7 @@ rebuild_st: st->nents = 0; for (i = 0; i < page_count; i++) { page = shmem_read_mapping_page_gfp(mapping, i, gfp); - if (IS_ERR(page)) { + if (unlikely(IS_ERR(page))) { i915_gem_shrink(dev_priv, page_count, I915_SHRINK_BOUND | @@ -2329,12 +2329,21 @@ rebuild_st: I915_SHRINK_PURGEABLE); page = shmem_read_mapping_page_gfp(mapping, i, gfp); } - if (IS_ERR(page)) { + if (unlikely(IS_ERR(page))) { + gfp_t reclaim; + /* We've tried hard to allocate the memory by reaping * our own buffer, now let the real VM do its job and * go down in flames if truly OOM. + * + * However, since graphics tend to be disposable, + * defer the oom here by reporting the ENOMEM back + * to userspace. */ - page = shmem_read_mapping_page(mapping, i); + reclaim = mapping_gfp_constraint(mapping, 0); + reclaim |= __GFP_NORETRY; /* reclaim, but no oom */ + + page = shmem_read_mapping_page_gfp(mapping, i, reclaim); if (IS_ERR(page)) { ret = PTR_ERR(page); goto err_sg; @@ -2989,10 +2998,15 @@ void i915_gem_set_wedged(struct drm_i915_private *dev_priv) lockdep_assert_held(&dev_priv->drm.struct_mutex); set_bit(I915_WEDGED, &dev_priv->gpu_error.flags); + /* Retire completed requests first so the list of inflight/incomplete + * requests is accurate and we don't try and mark successful requests + * as in error during __i915_gem_set_wedged_BKL(). + */ + i915_gem_retire_requests(dev_priv); + stop_machine(__i915_gem_set_wedged_BKL, dev_priv, NULL); i915_gem_context_lost(dev_priv); - i915_gem_retire_requests(dev_priv); mod_delayed_work(dev_priv->wq, &dev_priv->gt.idle_work, 0); } @@ -3098,9 +3112,7 @@ i915_gem_idle_work_handler(struct work_struct *work) * Wait for last execlists context complete, but bail out in case a * new request is submitted. */ - wait_for(READ_ONCE(dev_priv->gt.active_requests) || - intel_engines_are_idle(dev_priv), - 10); + wait_for(intel_engines_are_idle(dev_priv), 10); if (READ_ONCE(dev_priv->gt.active_requests)) return; @@ -3259,6 +3271,29 @@ static int wait_for_timeline(struct i915_gem_timeline *tl, unsigned int flags) return 0; } +static int wait_for_engine(struct intel_engine_cs *engine, int timeout_ms) +{ + return wait_for(intel_engine_is_idle(engine), timeout_ms); +} + +static int wait_for_engines(struct drm_i915_private *i915) +{ + struct intel_engine_cs *engine; + enum intel_engine_id id; + + for_each_engine(engine, i915, id) { + if (GEM_WARN_ON(wait_for_engine(engine, 50))) { + i915_gem_set_wedged(i915); + return -EIO; + } + + GEM_BUG_ON(intel_engine_get_seqno(engine) != + intel_engine_last_submit(engine)); + } + + return 0; +} + int i915_gem_wait_for_idle(struct drm_i915_private *i915, unsigned int flags) { int ret; @@ -3273,13 +3308,16 @@ int i915_gem_wait_for_idle(struct drm_i915_private *i915, unsigned int flags) if (ret) return ret; } + + i915_gem_retire_requests(i915); + GEM_BUG_ON(i915->gt.active_requests); + + ret = wait_for_engines(i915); } else { ret = wait_for_timeline(&i915->gt.global_timeline, flags); - if (ret) - return ret; } - return 0; + return ret; } /** Flushes the GTT write domain for the object if it's dirty. */ @@ -3307,8 +3345,14 @@ i915_gem_object_flush_gtt_write_domain(struct drm_i915_gem_object *obj) * system agents we cannot reproduce this behaviour). */ wmb(); - if (INTEL_GEN(dev_priv) >= 6 && !HAS_LLC(dev_priv)) - POSTING_READ(RING_ACTHD(dev_priv->engine[RCS]->mmio_base)); + if (INTEL_GEN(dev_priv) >= 6 && !HAS_LLC(dev_priv)) { + if (intel_runtime_pm_get_if_in_use(dev_priv)) { + spin_lock_irq(&dev_priv->uncore.lock); + POSTING_READ_FW(RING_ACTHD(dev_priv->engine[RCS]->mmio_base)); + spin_unlock_irq(&dev_priv->uncore.lock); + intel_runtime_pm_put(dev_priv); + } + } intel_fb_obj_flush(obj, write_origin(obj, I915_GEM_DOMAIN_GTT)); @@ -4408,9 +4452,6 @@ int i915_gem_suspend(struct drm_i915_private *dev_priv) if (ret) goto err_unlock; - i915_gem_retire_requests(dev_priv); - GEM_BUG_ON(dev_priv->gt.active_requests); - assert_kernel_context_is_current(dev_priv); i915_gem_context_lost(dev_priv); mutex_unlock(&dev->struct_mutex); diff --git a/drivers/gpu/drm/i915/i915_gem_clflush.c b/drivers/gpu/drm/i915/i915_gem_clflush.c index d925fb582ba7..ffd01e02fe94 100644 --- a/drivers/gpu/drm/i915/i915_gem_clflush.c +++ b/drivers/gpu/drm/i915/i915_gem_clflush.c @@ -168,7 +168,7 @@ void i915_gem_clflush_object(struct drm_i915_gem_object *obj, i915_sw_fence_await_reservation(&clflush->wait, obj->resv, NULL, - false, I915_FENCE_TIMEOUT, + true, I915_FENCE_TIMEOUT, GFP_KERNEL); reservation_object_lock(obj->resv, NULL); diff --git a/drivers/gpu/drm/i915/i915_gem_context.c b/drivers/gpu/drm/i915/i915_gem_context.c index 486051ed681d..8bd0c4966913 100644 --- a/drivers/gpu/drm/i915/i915_gem_context.c +++ b/drivers/gpu/drm/i915/i915_gem_context.c @@ -576,25 +576,25 @@ void i915_gem_context_close(struct drm_device *dev, struct drm_file *file) } static inline int -mi_set_context(struct drm_i915_gem_request *req, u32 hw_flags) +mi_set_context(struct drm_i915_gem_request *req, u32 flags) { struct drm_i915_private *dev_priv = req->i915; struct intel_engine_cs *engine = req->engine; enum intel_engine_id id; - u32 *cs, flags = hw_flags | MI_MM_SPACE_GTT; const int num_rings = - /* Use an extended w/a on ivb+ if signalling from other rings */ - i915.semaphores ? + /* Use an extended w/a on gen7 if signalling from other rings */ + (i915.semaphores && INTEL_GEN(dev_priv) == 7) ? INTEL_INFO(dev_priv)->num_rings - 1 : 0; int len; + u32 *cs; - /* These flags are for resource streamer on HSW+ */ + flags |= MI_MM_SPACE_GTT; if (IS_HASWELL(dev_priv) || INTEL_GEN(dev_priv) >= 8) - flags |= (HSW_MI_RS_SAVE_STATE_EN | HSW_MI_RS_RESTORE_STATE_EN); - else if (INTEL_GEN(dev_priv) < 8) - flags |= (MI_SAVE_EXT_STATE_EN | MI_RESTORE_EXT_STATE_EN); - + /* These flags are for resource streamer on HSW+ */ + flags |= HSW_MI_RS_SAVE_STATE_EN | HSW_MI_RS_RESTORE_STATE_EN; + else + flags |= MI_SAVE_EXT_STATE_EN | MI_RESTORE_EXT_STATE_EN; len = 4; if (INTEL_GEN(dev_priv) >= 7) diff --git a/drivers/gpu/drm/i915/i915_gem_evict.c b/drivers/gpu/drm/i915/i915_gem_evict.c index 2da3a94fc9f3..51e365f70464 100644 --- a/drivers/gpu/drm/i915/i915_gem_evict.c +++ b/drivers/gpu/drm/i915/i915_gem_evict.c @@ -196,7 +196,6 @@ search_again: if (ret) return ret; - i915_gem_retire_requests(dev_priv); goto search_again; found: @@ -383,7 +382,6 @@ int i915_gem_evict_vm(struct i915_address_space *vm, bool do_idle) if (ret) return ret; - i915_gem_retire_requests(dev_priv); WARN_ON(!list_empty(&vm->active_list)); } diff --git a/drivers/gpu/drm/i915/i915_gem_execbuffer.c b/drivers/gpu/drm/i915/i915_gem_execbuffer.c index dd7181ed5eca..a3e59c8ef27b 100644 --- a/drivers/gpu/drm/i915/i915_gem_execbuffer.c +++ b/drivers/gpu/drm/i915/i915_gem_execbuffer.c @@ -890,6 +890,7 @@ i915_gem_execbuffer_reserve(struct intel_engine_cs *engine, struct list_head ordered_vmas; struct list_head pinned_vmas; bool has_fenced_gpu_access = INTEL_GEN(engine->i915) < 4; + bool needs_unfenced_map = INTEL_INFO(engine->i915)->unfenced_needs_alignment; int retry; vm = list_first_entry(vmas, struct i915_vma, exec_list)->vm; @@ -910,7 +911,8 @@ i915_gem_execbuffer_reserve(struct intel_engine_cs *engine, if (!has_fenced_gpu_access) entry->flags &= ~EXEC_OBJECT_NEEDS_FENCE; need_fence = - entry->flags & EXEC_OBJECT_NEEDS_FENCE && + (entry->flags & EXEC_OBJECT_NEEDS_FENCE || + needs_unfenced_map) && i915_gem_object_is_tiled(obj); need_mappable = need_fence || need_reloc_mappable(vma); diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.c b/drivers/gpu/drm/i915/i915_gem_gtt.c index cee9c4fec52a..8bab4aea63e6 100644 --- a/drivers/gpu/drm/i915/i915_gem_gtt.c +++ b/drivers/gpu/drm/i915/i915_gem_gtt.c @@ -2364,7 +2364,7 @@ void i915_gem_gtt_finish_pages(struct drm_i915_gem_object *obj, struct i915_ggtt *ggtt = &dev_priv->ggtt; if (unlikely(ggtt->do_idle_maps)) { - if (i915_gem_wait_for_idle(dev_priv, I915_WAIT_LOCKED)) { + if (i915_gem_wait_for_idle(dev_priv, 0)) { DRM_ERROR("Failed to wait for idle; VT'd may hang.\n"); /* Wait a bit, in hopes it avoids the hang */ udelay(10); diff --git a/drivers/gpu/drm/i915/i915_gem_request.c b/drivers/gpu/drm/i915/i915_gem_request.c index 0e8d1010cecb..6348353b91ec 100644 --- a/drivers/gpu/drm/i915/i915_gem_request.c +++ b/drivers/gpu/drm/i915/i915_gem_request.c @@ -37,6 +37,17 @@ static const char *i915_fence_get_driver_name(struct dma_fence *fence) static const char *i915_fence_get_timeline_name(struct dma_fence *fence) { + /* The timeline struct (as part of the ppgtt underneath a context) + * may be freed when the request is no longer in use by the GPU. + * We could extend the life of a context to beyond that of all + * fences, possibly keeping the hw resource around indefinitely, + * or we just give them a false name. Since + * dma_fence_ops.get_timeline_name is a debug feature, the occasional + * lie seems justifiable. + */ + if (test_bit(DMA_FENCE_FLAG_SIGNALED_BIT, &fence->flags)) + return "signaled"; + return to_request(fence)->timeline->common->name; } @@ -180,7 +191,6 @@ i915_priotree_init(struct i915_priotree *pt) static int reset_all_global_seqno(struct drm_i915_private *i915, u32 seqno) { - struct i915_gem_timeline *timeline = &i915->gt.global_timeline; struct intel_engine_cs *engine; enum intel_engine_id id; int ret; @@ -192,15 +202,10 @@ static int reset_all_global_seqno(struct drm_i915_private *i915, u32 seqno) if (ret) return ret; - i915_gem_retire_requests(i915); - GEM_BUG_ON(i915->gt.active_requests > 1); - /* If the seqno wraps around, we need to clear the breadcrumb rbtree */ for_each_engine(engine, i915, id) { - struct intel_timeline *tl = &timeline->engine[id]; - - if (wait_for(intel_engine_is_idle(engine), 50)) - return -EBUSY; + struct i915_gem_timeline *timeline; + struct intel_timeline *tl = engine->timeline; if (!i915_seqno_passed(seqno, tl->seqno)) { /* spin until threads are complete */ @@ -211,14 +216,10 @@ static int reset_all_global_seqno(struct drm_i915_private *i915, u32 seqno) /* Finally reset hw state */ tl->seqno = seqno; intel_engine_init_global_seqno(engine, seqno); - } - list_for_each_entry(timeline, &i915->gt.timelines, link) { - for_each_engine(engine, i915, id) { - struct intel_timeline *tl = &timeline->engine[id]; - - memset(tl->sync_seqno, 0, sizeof(tl->sync_seqno)); - } + list_for_each_entry(timeline, &i915->gt.timelines, link) + memset(timeline->engine[id].sync_seqno, 0, + sizeof(timeline->engine[id].sync_seqno)); } return 0; @@ -295,7 +296,7 @@ static void i915_gem_request_retire(struct drm_i915_gem_request *request) * completion order. */ list_del(&request->ring_link); - request->ring->last_retired_head = request->postfix; + request->ring->head = request->postfix; if (!--request->i915->gt.active_requests) { GEM_BUG_ON(!request->i915->gt.awake); mod_delayed_work(request->i915->wq, diff --git a/drivers/gpu/drm/i915/i915_guc_submission.c b/drivers/gpu/drm/i915/i915_guc_submission.c index 832ac9e45801..1642fff9cf13 100644 --- a/drivers/gpu/drm/i915/i915_guc_submission.c +++ b/drivers/gpu/drm/i915/i915_guc_submission.c @@ -30,16 +30,25 @@ /** * DOC: GuC-based command submission * - * i915_guc_client: - * We use the term client to avoid confusion with contexts. A i915_guc_client is - * equivalent to GuC object guc_context_desc. This context descriptor is - * allocated from a pool of 1024 entries. Kernel driver will allocate doorbell - * and workqueue for it. Also the process descriptor (guc_process_desc), which - * is mapped to client space. So the client can write Work Item then ring the - * doorbell. + * GuC client: + * A i915_guc_client refers to a submission path through GuC. Currently, there + * is only one of these (the execbuf_client) and this one is charged with all + * submissions to the GuC. This struct is the owner of a doorbell, a process + * descriptor and a workqueue (all of them inside a single gem object that + * contains all required pages for these elements). * - * To simplify the implementation, we allocate one gem object that contains all - * pages for doorbell, process descriptor and workqueue. + * GuC stage descriptor: + * During initialization, the driver allocates a static pool of 1024 such + * descriptors, and shares them with the GuC. + * Currently, there exists a 1:1 mapping between a i915_guc_client and a + * guc_stage_desc (via the client's stage_id), so effectively only one + * gets used. This stage descriptor lets the GuC know about the doorbell, + * workqueue and process descriptor. Theoretically, it also lets the GuC + * know about our HW contexts (context ID, etc...), but we actually + * employ a kind of submission where the GuC uses the LRCA sent via the work + * item instead (the single guc_stage_desc associated to execbuf client + * contains information about the default kernel context only, but this is + * essentially unused). This is called a "proxy" submission. * * The Scratch registers: * There are 16 MMIO-based registers start from 0xC180. The kernel driver writes @@ -62,34 +71,91 @@ * ELSP context descriptor dword into Work Item. * See guc_wq_item_append() * + * ADS: + * The Additional Data Struct (ADS) has pointers for different buffers used by + * the GuC. One single gem object contains the ADS struct itself (guc_ads), the + * scheduling policies (guc_policies), a structure describing a collection of + * register sets (guc_mmio_reg_state) and some extra pages for the GuC to save + * its internal state for sleep. + * */ +static inline bool is_high_priority(struct i915_guc_client* client) +{ + return client->priority <= GUC_CLIENT_PRIORITY_HIGH; +} + +static int __reserve_doorbell(struct i915_guc_client *client) +{ + unsigned long offset; + unsigned long end; + u16 id; + + GEM_BUG_ON(client->doorbell_id != GUC_DOORBELL_INVALID); + + /* + * The bitmap tracks which doorbell registers are currently in use. + * It is split into two halves; the first half is used for normal + * priority contexts, the second half for high-priority ones. + */ + offset = 0; + end = GUC_NUM_DOORBELLS/2; + if (is_high_priority(client)) { + offset = end; + end += offset; + } + + id = find_next_zero_bit(client->guc->doorbell_bitmap, offset, end); + if (id == end) + return -ENOSPC; + + __set_bit(id, client->guc->doorbell_bitmap); + client->doorbell_id = id; + DRM_DEBUG_DRIVER("client %u (high prio=%s) reserved doorbell: %d\n", + client->stage_id, yesno(is_high_priority(client)), + id); + return 0; +} + +static void __unreserve_doorbell(struct i915_guc_client *client) +{ + GEM_BUG_ON(client->doorbell_id == GUC_DOORBELL_INVALID); + + __clear_bit(client->doorbell_id, client->guc->doorbell_bitmap); + client->doorbell_id = GUC_DOORBELL_INVALID; +} + /* * Tell the GuC to allocate or deallocate a specific doorbell */ -static int guc_allocate_doorbell(struct intel_guc *guc, - struct i915_guc_client *client) +static int __guc_allocate_doorbell(struct intel_guc *guc, u32 stage_id) { u32 action[] = { INTEL_GUC_ACTION_ALLOCATE_DOORBELL, - client->ctx_index + stage_id }; return intel_guc_send(guc, action, ARRAY_SIZE(action)); } -static int guc_release_doorbell(struct intel_guc *guc, - struct i915_guc_client *client) +static int __guc_deallocate_doorbell(struct intel_guc *guc, u32 stage_id) { u32 action[] = { INTEL_GUC_ACTION_DEALLOCATE_DOORBELL, - client->ctx_index + stage_id }; return intel_guc_send(guc, action, ARRAY_SIZE(action)); } +static struct guc_stage_desc *__get_stage_desc(struct i915_guc_client *client) +{ + struct guc_stage_desc *base = client->guc->stage_desc_pool_vaddr; + + return &base[client->stage_id]; +} + /* * Initialise, update, or clear doorbell data shared with the GuC * @@ -97,107 +163,129 @@ static int guc_release_doorbell(struct intel_guc *guc, * client object which contains the page being used for the doorbell */ -static int guc_update_doorbell_id(struct intel_guc *guc, - struct i915_guc_client *client, - u16 new_id) +static void __update_doorbell_desc(struct i915_guc_client *client, u16 new_id) { - struct sg_table *sg = guc->ctx_pool_vma->pages; - void *doorbell_bitmap = guc->doorbell_bitmap; - struct guc_doorbell_info *doorbell; - struct guc_context_desc desc; - size_t len; + struct guc_stage_desc *desc; - doorbell = client->vaddr + client->doorbell_offset; + /* Update the GuC's idea of the doorbell ID */ + desc = __get_stage_desc(client); + desc->db_id = new_id; +} - if (client->doorbell_id != GUC_INVALID_DOORBELL_ID && - test_bit(client->doorbell_id, doorbell_bitmap)) { - /* Deactivate the old doorbell */ - doorbell->db_status = GUC_DOORBELL_DISABLED; - (void)guc_release_doorbell(guc, client); - __clear_bit(client->doorbell_id, doorbell_bitmap); - } +static struct guc_doorbell_info *__get_doorbell(struct i915_guc_client *client) +{ + return client->vaddr + client->doorbell_offset; +} - /* Update the GuC's idea of the doorbell ID */ - len = sg_pcopy_to_buffer(sg->sgl, sg->nents, &desc, sizeof(desc), - sizeof(desc) * client->ctx_index); - if (len != sizeof(desc)) - return -EFAULT; - desc.db_id = new_id; - len = sg_pcopy_from_buffer(sg->sgl, sg->nents, &desc, sizeof(desc), - sizeof(desc) * client->ctx_index); - if (len != sizeof(desc)) - return -EFAULT; - - client->doorbell_id = new_id; - if (new_id == GUC_INVALID_DOORBELL_ID) - return 0; +static bool has_doorbell(struct i915_guc_client *client) +{ + if (client->doorbell_id == GUC_DOORBELL_INVALID) + return false; - /* Activate the new doorbell */ - __set_bit(new_id, doorbell_bitmap); + return test_bit(client->doorbell_id, client->guc->doorbell_bitmap); +} + +static int __create_doorbell(struct i915_guc_client *client) +{ + struct guc_doorbell_info *doorbell; + int err; + + doorbell = __get_doorbell(client); doorbell->db_status = GUC_DOORBELL_ENABLED; doorbell->cookie = client->doorbell_cookie; - return guc_allocate_doorbell(guc, client); + + err = __guc_allocate_doorbell(client->guc, client->stage_id); + if (err) { + doorbell->db_status = GUC_DOORBELL_DISABLED; + doorbell->cookie = 0; + } + return err; } -static void guc_disable_doorbell(struct intel_guc *guc, - struct i915_guc_client *client) +static int __destroy_doorbell(struct i915_guc_client *client) { - (void)guc_update_doorbell_id(guc, client, GUC_INVALID_DOORBELL_ID); + struct drm_i915_private *dev_priv = guc_to_i915(client->guc); + struct guc_doorbell_info *doorbell; + u16 db_id = client->doorbell_id; - /* XXX: wait for any interrupts */ - /* XXX: wait for workqueue to drain */ + GEM_BUG_ON(db_id >= GUC_DOORBELL_INVALID); + + doorbell = __get_doorbell(client); + doorbell->db_status = GUC_DOORBELL_DISABLED; + doorbell->cookie = 0; + + /* Doorbell release flow requires that we wait for GEN8_DRB_VALID bit + * to go to zero after updating db_status before we call the GuC to + * release the doorbell */ + if (wait_for_us(!(I915_READ(GEN8_DRBREGL(db_id)) & GEN8_DRB_VALID), 10)) + WARN_ONCE(true, "Doorbell never became invalid after disable\n"); + + return __guc_deallocate_doorbell(client->guc, client->stage_id); } -static uint16_t -select_doorbell_register(struct intel_guc *guc, uint32_t priority) +static int create_doorbell(struct i915_guc_client *client) { - /* - * The bitmap tracks which doorbell registers are currently in use. - * It is split into two halves; the first half is used for normal - * priority contexts, the second half for high-priority ones. - * Note that logically higher priorities are numerically less than - * normal ones, so the test below means "is it high-priority?" - */ - const bool hi_pri = (priority <= GUC_CTX_PRIORITY_HIGH); - const uint16_t half = GUC_MAX_DOORBELLS / 2; - const uint16_t start = hi_pri ? half : 0; - const uint16_t end = start + half; - uint16_t id; + int ret; - id = find_next_zero_bit(guc->doorbell_bitmap, end, start); - if (id == end) - id = GUC_INVALID_DOORBELL_ID; + ret = __reserve_doorbell(client); + if (ret) + return ret; - DRM_DEBUG_DRIVER("assigned %s priority doorbell id 0x%x\n", - hi_pri ? "high" : "normal", id); + __update_doorbell_desc(client, client->doorbell_id); + + ret = __create_doorbell(client); + if (ret) + goto err; + + return 0; - return id; +err: + __update_doorbell_desc(client, GUC_DOORBELL_INVALID); + __unreserve_doorbell(client); + return ret; } -/* - * Select, assign and relase doorbell cachelines - * - * These functions track which doorbell cachelines are in use. - * The data they manipulate is protected by the intel_guc_send lock. - */ +static int destroy_doorbell(struct i915_guc_client *client) +{ + int err; + + GEM_BUG_ON(!has_doorbell(client)); + + /* XXX: wait for any interrupts */ + /* XXX: wait for workqueue to drain */ + + err = __destroy_doorbell(client); + if (err) + return err; + + __update_doorbell_desc(client, GUC_DOORBELL_INVALID); + + __unreserve_doorbell(client); -static uint32_t select_doorbell_cacheline(struct intel_guc *guc) + return 0; +} + +static unsigned long __select_cacheline(struct intel_guc* guc) { - const uint32_t cacheline_size = cache_line_size(); - uint32_t offset; + unsigned long offset; /* Doorbell uses a single cache line within a page */ offset = offset_in_page(guc->db_cacheline); /* Moving to next cache line to reduce contention */ - guc->db_cacheline += cacheline_size; - - DRM_DEBUG_DRIVER("selected doorbell cacheline 0x%x, next 0x%x, linesize %u\n", - offset, guc->db_cacheline, cacheline_size); + guc->db_cacheline += cache_line_size(); + DRM_DEBUG_DRIVER("reserved cacheline 0x%lx, next 0x%x, linesize %u\n", + offset, guc->db_cacheline, cache_line_size()); return offset; } +static inline struct guc_process_desc * +__get_process_desc(struct i915_guc_client *client) +{ + return client->vaddr + client->proc_desc_offset; +} + /* * Initialise the process descriptor shared with the GuC firmware. */ @@ -206,9 +294,7 @@ static void guc_proc_desc_init(struct intel_guc *guc, { struct guc_process_desc *desc; - desc = client->vaddr + client->proc_desc_offset; - - memset(desc, 0, sizeof(*desc)); + desc = memset(__get_process_desc(client), 0, sizeof(*desc)); /* * XXX: pDoorbell and WQVBaseAddress are pointers in process address @@ -219,42 +305,41 @@ static void guc_proc_desc_init(struct intel_guc *guc, desc->wq_base_addr = 0; desc->db_base_addr = 0; - desc->context_id = client->ctx_index; + desc->stage_id = client->stage_id; desc->wq_size_bytes = client->wq_size; desc->wq_status = WQ_STATUS_ACTIVE; desc->priority = client->priority; } /* - * Initialise/clear the context descriptor shared with the GuC firmware. + * Initialise/clear the stage descriptor shared with the GuC firmware. * * This descriptor tells the GuC where (in GGTT space) to find the important * data structures relating to this client (doorbell, process descriptor, * write queue, etc). */ - -static void guc_ctx_desc_init(struct intel_guc *guc, - struct i915_guc_client *client) +static void guc_stage_desc_init(struct intel_guc *guc, + struct i915_guc_client *client) { struct drm_i915_private *dev_priv = guc_to_i915(guc); struct intel_engine_cs *engine; struct i915_gem_context *ctx = client->owner; - struct guc_context_desc desc; - struct sg_table *sg; + struct guc_stage_desc *desc; unsigned int tmp; u32 gfx_addr; - memset(&desc, 0, sizeof(desc)); + desc = __get_stage_desc(client); + memset(desc, 0, sizeof(*desc)); - desc.attribute = GUC_CTX_DESC_ATTR_ACTIVE | GUC_CTX_DESC_ATTR_KERNEL; - desc.context_id = client->ctx_index; - desc.priority = client->priority; - desc.db_id = client->doorbell_id; + desc->attribute = GUC_STAGE_DESC_ATTR_ACTIVE | GUC_STAGE_DESC_ATTR_KERNEL; + desc->stage_id = client->stage_id; + desc->priority = client->priority; + desc->db_id = client->doorbell_id; for_each_engine_masked(engine, dev_priv, client->engines, tmp) { struct intel_context *ce = &ctx->engine[engine->id]; uint32_t guc_engine_id = engine->guc_id; - struct guc_execlist_context *lrc = &desc.lrc[guc_engine_id]; + struct guc_execlist_context *lrc = &desc->lrc[guc_engine_id]; /* TODO: We have a design issue to be solved here. Only when we * receive the first batch, we know which engine is used by the @@ -266,12 +351,22 @@ static void guc_ctx_desc_init(struct intel_guc *guc, if (!ce->state) break; /* XXX: continue? */ + /* + * XXX: When this is a GUC_STAGE_DESC_ATTR_KERNEL client (proxy + * submission or, in other words, not using a direct submission + * model) the KMD's LRCA is not used for any work submission. + * Instead, the GuC uses the LRCA of the user mode context (see + * guc_wq_item_append below). + */ lrc->context_desc = lower_32_bits(ce->lrc_desc); /* The state page is after PPHWSP */ - lrc->ring_lcra = + lrc->ring_lrca = guc_ggtt_offset(ce->state) + LRC_STATE_PN * PAGE_SIZE; - lrc->context_id = (client->ctx_index << GUC_ELC_CTXID_OFFSET) | + + /* XXX: In direct submission, the GuC wants the HW context id + * here. In proxy submission, it wants the stage id */ + lrc->context_id = (client->stage_id << GUC_ELC_CTXID_OFFSET) | (guc_engine_id << GUC_ELC_ENGINE_OFFSET); lrc->ring_begin = guc_ggtt_offset(ce->ring->vma); @@ -279,50 +374,36 @@ static void guc_ctx_desc_init(struct intel_guc *guc, lrc->ring_next_free_location = lrc->ring_begin; lrc->ring_current_tail_pointer_value = 0; - desc.engines_used |= (1 << guc_engine_id); + desc->engines_used |= (1 << guc_engine_id); } DRM_DEBUG_DRIVER("Host engines 0x%x => GuC engines used 0x%x\n", - client->engines, desc.engines_used); - WARN_ON(desc.engines_used == 0); + client->engines, desc->engines_used); + WARN_ON(desc->engines_used == 0); /* * The doorbell, process descriptor, and workqueue are all parts * of the client object, which the GuC will reference via the GGTT */ gfx_addr = guc_ggtt_offset(client->vma); - desc.db_trigger_phy = sg_dma_address(client->vma->pages->sgl) + + desc->db_trigger_phy = sg_dma_address(client->vma->pages->sgl) + client->doorbell_offset; - desc.db_trigger_cpu = - (uintptr_t)client->vaddr + client->doorbell_offset; - desc.db_trigger_uk = gfx_addr + client->doorbell_offset; - desc.process_desc = gfx_addr + client->proc_desc_offset; - desc.wq_addr = gfx_addr + client->wq_offset; - desc.wq_size = client->wq_size; - - /* - * XXX: Take LRCs from an existing context if this is not an - * IsKMDCreatedContext client - */ - desc.desc_private = (uintptr_t)client; + desc->db_trigger_cpu = (uintptr_t)__get_doorbell(client); + desc->db_trigger_uk = gfx_addr + client->doorbell_offset; + desc->process_desc = gfx_addr + client->proc_desc_offset; + desc->wq_addr = gfx_addr + client->wq_offset; + desc->wq_size = client->wq_size; - /* Pool context is pinned already */ - sg = guc->ctx_pool_vma->pages; - sg_pcopy_from_buffer(sg->sgl, sg->nents, &desc, sizeof(desc), - sizeof(desc) * client->ctx_index); + desc->desc_private = (uintptr_t)client; } -static void guc_ctx_desc_fini(struct intel_guc *guc, - struct i915_guc_client *client) +static void guc_stage_desc_fini(struct intel_guc *guc, + struct i915_guc_client *client) { - struct guc_context_desc desc; - struct sg_table *sg; - - memset(&desc, 0, sizeof(desc)); + struct guc_stage_desc *desc; - sg = guc->ctx_pool_vma->pages; - sg_pcopy_from_buffer(sg->sgl, sg->nents, &desc, sizeof(desc), - sizeof(desc) * client->ctx_index); + desc = __get_stage_desc(client); + memset(desc, 0, sizeof(*desc)); } /** @@ -345,8 +426,7 @@ int i915_guc_wq_reserve(struct drm_i915_gem_request *request) { const size_t wqi_size = sizeof(struct guc_wq_item); struct i915_guc_client *client = request->i915->guc.execbuf_client; - struct guc_process_desc *desc = client->vaddr + - client->proc_desc_offset; + struct guc_process_desc *desc = __get_process_desc(client); u32 freespace; int ret; @@ -391,19 +471,17 @@ static void guc_wq_item_append(struct i915_guc_client *client, const size_t wqi_size = sizeof(struct guc_wq_item); const u32 wqi_len = wqi_size/sizeof(u32) - 1; struct intel_engine_cs *engine = rq->engine; - struct guc_process_desc *desc; + struct guc_process_desc *desc = __get_process_desc(client); struct guc_wq_item *wqi; u32 freespace, tail, wq_off; - desc = client->vaddr + client->proc_desc_offset; - /* Free space is guaranteed, see i915_guc_wq_reserve() above */ freespace = CIRC_SPACE(client->wq_tail, desc->head, client->wq_size); GEM_BUG_ON(freespace < wqi_size); /* The GuC firmware wants the tail index in QWords, not bytes */ tail = rq->tail; - GEM_BUG_ON(tail & 7); + assert_ring_tail_valid(rq->ring, rq->tail); tail >>= 3; GEM_BUG_ON(tail > WQ_RING_TAIL_MAX); @@ -436,19 +514,27 @@ static void guc_wq_item_append(struct i915_guc_client *client, /* The GuC wants only the low-order word of the context descriptor */ wqi->context_desc = (u32)intel_lr_context_descriptor(rq->ctx, engine); - wqi->ring_tail = tail << WQ_RING_TAIL_SHIFT; + wqi->submit_element_info = tail << WQ_RING_TAIL_SHIFT; wqi->fence_id = rq->global_seqno; } +static void guc_reset_wq(struct i915_guc_client *client) +{ + struct guc_process_desc *desc = __get_process_desc(client); + + desc->head = 0; + desc->tail = 0; + + client->wq_tail = 0; +} + static int guc_ring_doorbell(struct i915_guc_client *client) { - struct guc_process_desc *desc; + struct guc_process_desc *desc = __get_process_desc(client); union guc_doorbell_qw db_cmp, db_exc, db_ret; union guc_doorbell_qw *db; int attempt = 2, ret = -EAGAIN; - desc = client->vaddr + client->proc_desc_offset; - /* Update the tail so it is visible to GuC */ desc->tail = client->wq_tail; @@ -463,7 +549,7 @@ static int guc_ring_doorbell(struct i915_guc_client *client) db_exc.cookie = 1; /* pointer of current doorbell cacheline */ - db = client->vaddr + client->doorbell_offset; + db = (union guc_doorbell_qw *)__get_doorbell(client); while (attempt--) { /* lets ring the doorbell */ @@ -573,23 +659,10 @@ static bool i915_guc_dequeue(struct intel_engine_cs *engine) { struct execlist_port *port = engine->execlist_port; struct drm_i915_gem_request *last = port[0].request; - unsigned long flags; struct rb_node *rb; bool submit = false; - /* After execlist_first is updated, the tasklet will be rescheduled. - * - * If we are currently running (inside the tasklet) and a third - * party queues a request and so updates engine->execlist_first under - * the spinlock (which we have elided), it will atomically set the - * TASKLET_SCHED flag causing the us to be re-executed and pick up - * the change in state (the update to TASKLET_SCHED incurs a memory - * barrier making this cross-cpu checking safe). - */ - if (!READ_ONCE(engine->execlist_first)) - return false; - - spin_lock_irqsave(&engine->timeline->lock, flags); + spin_lock_irq(&engine->timeline->lock); rb = engine->execlist_first; while (rb) { struct drm_i915_gem_request *rq = @@ -609,8 +682,8 @@ static bool i915_guc_dequeue(struct intel_engine_cs *engine) RB_CLEAR_NODE(&rq->priotree.node); rq->priotree.priority = INT_MAX; - trace_i915_gem_request_in(rq, port - engine->execlist_port); i915_guc_submit(rq); + trace_i915_gem_request_in(rq, port - engine->execlist_port); last = rq; submit = true; } @@ -619,7 +692,7 @@ static bool i915_guc_dequeue(struct intel_engine_cs *engine) nested_enable_signaling(last); engine->execlist_first = rb; } - spin_unlock_irqrestore(&engine->timeline->lock, flags); + spin_unlock_irq(&engine->timeline->lock); return submit; } @@ -695,93 +768,100 @@ err: return vma; } -static void -guc_client_free(struct drm_i915_private *dev_priv, - struct i915_guc_client *client) +/* Check that a doorbell register is in the expected state */ +static bool doorbell_ok(struct intel_guc *guc, u16 db_id) { - struct intel_guc *guc = &dev_priv->guc; - - if (!client) - return; - - /* - * XXX: wait for any outstanding submissions before freeing memory. - * Be sure to drop any locks - */ + struct drm_i915_private *dev_priv = guc_to_i915(guc); + u32 drbregl; + bool valid; - if (client->vaddr) { - /* - * If we got as far as setting up a doorbell, make sure we - * shut it down before unmapping & deallocating the memory. - */ - guc_disable_doorbell(guc, client); + GEM_BUG_ON(db_id >= GUC_DOORBELL_INVALID); - i915_gem_object_unpin_map(client->vma->obj); - } + drbregl = I915_READ(GEN8_DRBREGL(db_id)); + valid = drbregl & GEN8_DRB_VALID; - i915_vma_unpin_and_release(&client->vma); + if (test_bit(db_id, guc->doorbell_bitmap) == valid) + return true; - if (client->ctx_index != GUC_INVALID_CTX_ID) { - guc_ctx_desc_fini(guc, client); - ida_simple_remove(&guc->ctx_ids, client->ctx_index); - } + DRM_DEBUG_DRIVER("Doorbell %d has unexpected state (0x%x): valid=%s\n", + db_id, drbregl, yesno(valid)); - kfree(client); + return false; } -/* Check that a doorbell register is in the expected state */ -static bool guc_doorbell_check(struct intel_guc *guc, uint16_t db_id) +/* + * If the GuC thinks that the doorbell is unassigned (e.g. because we reset and + * reloaded the GuC FW) we can use this function to tell the GuC to reassign the + * doorbell to the rightful owner. + */ +static int __reset_doorbell(struct i915_guc_client* client, u16 db_id) { - struct drm_i915_private *dev_priv = guc_to_i915(guc); - i915_reg_t drbreg = GEN8_DRBREGL(db_id); - uint32_t value = I915_READ(drbreg); - bool enabled = (value & GUC_DOORBELL_ENABLED) != 0; - bool expected = test_bit(db_id, guc->doorbell_bitmap); - - if (enabled == expected) - return true; + int err; - DRM_DEBUG_DRIVER("Doorbell %d (reg 0x%x) 0x%x, should be %s\n", - db_id, drbreg.reg, value, - expected ? "active" : "inactive"); + __update_doorbell_desc(client, db_id); + err = __create_doorbell(client); + if (!err) + err = __destroy_doorbell(client); - return false; + return err; } /* - * Borrow the first client to set up & tear down each unused doorbell - * in turn, to ensure that all doorbell h/w is (re)initialised. + * Set up & tear down each unused doorbell in turn, to ensure that all doorbell + * HW is (re)initialised. For that end, we might have to borrow the first + * client. Also, tell GuC about all the doorbells in use by all clients. + * We do this because the KMD, the GuC and the doorbell HW can easily go out of + * sync (e.g. we can reset the GuC, but not the doorbel HW). */ -static void guc_init_doorbell_hw(struct intel_guc *guc) +static int guc_init_doorbell_hw(struct intel_guc *guc) { struct i915_guc_client *client = guc->execbuf_client; - uint16_t db_id; - int i, err; - - guc_disable_doorbell(guc, client); + bool recreate_first_client = false; + u16 db_id; + int ret; - for (i = 0; i < GUC_MAX_DOORBELLS; ++i) { - /* Skip if doorbell is OK */ - if (guc_doorbell_check(guc, i)) + /* For unused doorbells, make sure they are disabled */ + for_each_clear_bit(db_id, guc->doorbell_bitmap, GUC_NUM_DOORBELLS) { + if (doorbell_ok(guc, db_id)) continue; - err = guc_update_doorbell_id(guc, client, i); - if (err) - DRM_DEBUG_DRIVER("Doorbell %d update failed, err %d\n", - i, err); + if (has_doorbell(client)) { + /* Borrow execbuf_client (we will recreate it later) */ + destroy_doorbell(client); + recreate_first_client = true; + } + + ret = __reset_doorbell(client, db_id); + WARN(ret, "Doorbell %u reset failed, err %d\n", db_id, ret); } - db_id = select_doorbell_register(guc, client->priority); - WARN_ON(db_id == GUC_INVALID_DOORBELL_ID); + if (recreate_first_client) { + ret = __reserve_doorbell(client); + if (unlikely(ret)) { + DRM_ERROR("Couldn't re-reserve first client db: %d\n", ret); + return ret; + } + + __update_doorbell_desc(client, client->doorbell_id); + } - err = guc_update_doorbell_id(guc, client, db_id); - if (err) - DRM_WARN("Failed to restore doorbell to %d, err %d\n", - db_id, err); + /* Now for every client (and not only execbuf_client) make sure their + * doorbells are known by the GuC */ + //for (client = client_list; client != NULL; client = client->next) + { + ret = __create_doorbell(client); + if (ret) { + DRM_ERROR("Couldn't recreate client %u doorbell: %d\n", + client->stage_id, ret); + return ret; + } + } - /* Read back & verify all doorbell registers */ - for (i = 0; i < GUC_MAX_DOORBELLS; ++i) - (void)guc_doorbell_check(guc, i); + /* Read back & verify all (used & unused) doorbell registers */ + for (db_id = 0; db_id < GUC_NUM_DOORBELLS; ++db_id) + WARN_ON(!doorbell_ok(guc, db_id)); + + return 0; } /** @@ -807,49 +887,46 @@ guc_client_alloc(struct drm_i915_private *dev_priv, struct intel_guc *guc = &dev_priv->guc; struct i915_vma *vma; void *vaddr; - uint16_t db_id; + int ret; client = kzalloc(sizeof(*client), GFP_KERNEL); if (!client) - return NULL; + return ERR_PTR(-ENOMEM); - client->owner = ctx; client->guc = guc; + client->owner = ctx; client->engines = engines; client->priority = priority; - client->doorbell_id = GUC_INVALID_DOORBELL_ID; + client->doorbell_id = GUC_DOORBELL_INVALID; + client->wq_offset = GUC_DB_SIZE; + client->wq_size = GUC_WQ_SIZE; + spin_lock_init(&client->wq_lock); - client->ctx_index = (uint32_t)ida_simple_get(&guc->ctx_ids, 0, - GUC_MAX_GPU_CONTEXTS, GFP_KERNEL); - if (client->ctx_index >= GUC_MAX_GPU_CONTEXTS) { - client->ctx_index = GUC_INVALID_CTX_ID; - goto err; - } + ret = ida_simple_get(&guc->stage_ids, 0, GUC_MAX_STAGE_DESCRIPTORS, + GFP_KERNEL); + if (ret < 0) + goto err_client; + + client->stage_id = ret; /* The first page is doorbell/proc_desc. Two followed pages are wq. */ vma = intel_guc_allocate_vma(guc, GUC_DB_SIZE + GUC_WQ_SIZE); - if (IS_ERR(vma)) - goto err; + if (IS_ERR(vma)) { + ret = PTR_ERR(vma); + goto err_id; + } /* We'll keep just the first (doorbell/proc) page permanently kmap'd. */ client->vma = vma; vaddr = i915_gem_object_pin_map(vma->obj, I915_MAP_WB); - if (IS_ERR(vaddr)) - goto err; - + if (IS_ERR(vaddr)) { + ret = PTR_ERR(vaddr); + goto err_vma; + } client->vaddr = vaddr; - spin_lock_init(&client->wq_lock); - client->wq_offset = GUC_DB_SIZE; - client->wq_size = GUC_WQ_SIZE; - - db_id = select_doorbell_register(guc, client->priority); - if (db_id == GUC_INVALID_DOORBELL_ID) - /* XXX: evict a doorbell instead? */ - goto err; - - client->doorbell_offset = select_doorbell_cacheline(guc); + client->doorbell_offset = __select_cacheline(guc); /* * Since the doorbell only requires a single cacheline, we can save @@ -862,28 +939,47 @@ guc_client_alloc(struct drm_i915_private *dev_priv, client->proc_desc_offset = (GUC_DB_SIZE / 2); guc_proc_desc_init(guc, client); - guc_ctx_desc_init(guc, client); + guc_stage_desc_init(guc, client); - /* For runtime client allocation we need to enable the doorbell. Not - * required yet for the static execbuf_client as this special kernel - * client is enabled from i915_guc_submission_enable(). - * - * guc_update_doorbell_id(guc, client, db_id); - */ + ret = create_doorbell(client); + if (ret) + goto err_vaddr; - DRM_DEBUG_DRIVER("new priority %u client %p for engine(s) 0x%x: ctx_index %u\n", - priority, client, client->engines, client->ctx_index); - DRM_DEBUG_DRIVER("doorbell id %u, cacheline offset 0x%x\n", - client->doorbell_id, client->doorbell_offset); + DRM_DEBUG_DRIVER("new priority %u client %p for engine(s) 0x%x: stage_id %u\n", + priority, client, client->engines, client->stage_id); + DRM_DEBUG_DRIVER("doorbell id %u, cacheline offset 0x%lx\n", + client->doorbell_id, client->doorbell_offset); return client; -err: - guc_client_free(dev_priv, client); - return NULL; +err_vaddr: + i915_gem_object_unpin_map(client->vma->obj); +err_vma: + i915_vma_unpin_and_release(&client->vma); +err_id: + ida_simple_remove(&guc->stage_ids, client->stage_id); +err_client: + kfree(client); + return ERR_PTR(ret); } +static void guc_client_free(struct i915_guc_client *client) +{ + /* + * XXX: wait for any outstanding submissions before freeing memory. + * Be sure to drop any locks + */ + /* FIXME: in many cases, by the time we get here the GuC has been + * reset, so we cannot destroy the doorbell properly. Ignore the + * error message for now */ + destroy_doorbell(client); + guc_stage_desc_fini(client->guc, client); + i915_gem_object_unpin_map(client->vma->obj); + i915_vma_unpin_and_release(&client->vma); + ida_simple_remove(&client->guc->stage_ids, client->stage_id); + kfree(client); +} static void guc_policies_init(struct guc_policies *policies) { @@ -893,7 +989,7 @@ static void guc_policies_init(struct guc_policies *policies) policies->dpc_promote_time = 500000; policies->max_num_work_items = POLICY_MAX_NUM_WI; - for (p = 0; p < GUC_CTX_PRIORITY_NUM; p++) { + for (p = 0; p < GUC_CLIENT_PRIORITY_NUM; p++) { for (i = GUC_RENDER_ENGINE; i < GUC_MAX_ENGINES_NUM; i++) { policy = &policies->policy[p][i]; @@ -907,7 +1003,7 @@ static void guc_policies_init(struct guc_policies *policies) policies->is_valid = 1; } -static void guc_addon_create(struct intel_guc *guc) +static int guc_ads_create(struct intel_guc *guc) { struct drm_i915_private *dev_priv = guc_to_i915(guc); struct i915_vma *vma; @@ -923,14 +1019,13 @@ static void guc_addon_create(struct intel_guc *guc) enum intel_engine_id id; u32 base; - vma = guc->ads_vma; - if (!vma) { - vma = intel_guc_allocate_vma(guc, PAGE_ALIGN(sizeof(*blob))); - if (IS_ERR(vma)) - return; + GEM_BUG_ON(guc->ads_vma); - guc->ads_vma = vma; - } + vma = intel_guc_allocate_vma(guc, PAGE_ALIGN(sizeof(*blob))); + if (IS_ERR(vma)) + return PTR_ERR(vma); + + guc->ads_vma = vma; page = i915_vma_first_page(vma); blob = kmap(page); @@ -940,11 +1035,11 @@ static void guc_addon_create(struct intel_guc *guc) /* MMIO reg state */ for_each_engine(engine, dev_priv, id) { - blob->reg_state.mmio_white_list[engine->guc_id].mmio_start = + blob->reg_state.white_list[engine->guc_id].mmio_start = engine->mmio_base + GUC_MMIO_WHITE_LIST_START; /* Nothing to be saved or restored for now. */ - blob->reg_state.mmio_white_list[engine->guc_id].count = 0; + blob->reg_state.white_list[engine->guc_id].count = 0; } /* @@ -967,67 +1062,75 @@ static void guc_addon_create(struct intel_guc *guc) blob->ads.reg_state_addr = base + ptr_offset(blob, reg_state); kunmap(page); + + return 0; +} + +static void guc_ads_destroy(struct intel_guc *guc) +{ + i915_vma_unpin_and_release(&guc->ads_vma); } /* - * Set up the memory resources to be shared with the GuC. At this point, - * we require just one object that can be mapped through the GGTT. + * Set up the memory resources to be shared with the GuC (via the GGTT) + * at firmware loading time. */ int i915_guc_submission_init(struct drm_i915_private *dev_priv) { - const size_t ctxsize = sizeof(struct guc_context_desc); - const size_t poolsize = GUC_MAX_GPU_CONTEXTS * ctxsize; - const size_t gemsize = round_up(poolsize, PAGE_SIZE); struct intel_guc *guc = &dev_priv->guc; struct i915_vma *vma; + void *vaddr; + int ret; - if (!HAS_GUC_SCHED(dev_priv)) + if (guc->stage_desc_pool) return 0; - /* Wipe bitmap & delete client in case of reinitialisation */ - bitmap_clear(guc->doorbell_bitmap, 0, GUC_MAX_DOORBELLS); - i915_guc_submission_disable(dev_priv); - - if (!i915.enable_guc_submission) - return 0; /* not enabled */ - - if (guc->ctx_pool_vma) - return 0; /* already allocated */ - - vma = intel_guc_allocate_vma(guc, gemsize); + vma = intel_guc_allocate_vma(guc, + PAGE_ALIGN(sizeof(struct guc_stage_desc) * + GUC_MAX_STAGE_DESCRIPTORS)); if (IS_ERR(vma)) return PTR_ERR(vma); - guc->ctx_pool_vma = vma; - ida_init(&guc->ctx_ids); - intel_guc_log_create(guc); - guc_addon_create(guc); - - guc->execbuf_client = guc_client_alloc(dev_priv, - INTEL_INFO(dev_priv)->ring_mask, - GUC_CTX_PRIORITY_KMD_NORMAL, - dev_priv->kernel_context); - if (!guc->execbuf_client) { - DRM_ERROR("Failed to create GuC client for execbuf!\n"); - goto err; + guc->stage_desc_pool = vma; + + vaddr = i915_gem_object_pin_map(guc->stage_desc_pool->obj, I915_MAP_WB); + if (IS_ERR(vaddr)) { + ret = PTR_ERR(vaddr); + goto err_vma; } + guc->stage_desc_pool_vaddr = vaddr; + + ret = intel_guc_log_create(guc); + if (ret < 0) + goto err_vaddr; + + ret = guc_ads_create(guc); + if (ret < 0) + goto err_log; + + ida_init(&guc->stage_ids); + return 0; -err: - i915_guc_submission_fini(dev_priv); - return -ENOMEM; +err_log: + intel_guc_log_destroy(guc); +err_vaddr: + i915_gem_object_unpin_map(guc->stage_desc_pool->obj); +err_vma: + i915_vma_unpin_and_release(&guc->stage_desc_pool); + return ret; } -static void guc_reset_wq(struct i915_guc_client *client) +void i915_guc_submission_fini(struct drm_i915_private *dev_priv) { - struct guc_process_desc *desc = client->vaddr + - client->proc_desc_offset; - - desc->head = 0; - desc->tail = 0; + struct intel_guc *guc = &dev_priv->guc; - client->wq_tail = 0; + ida_destroy(&guc->stage_ids); + guc_ads_destroy(guc); + intel_guc_log_destroy(guc); + i915_gem_object_unpin_map(guc->stage_desc_pool->obj); + i915_vma_unpin_and_release(&guc->stage_desc_pool); } static void guc_interrupts_capture(struct drm_i915_private *dev_priv) @@ -1072,20 +1175,60 @@ static void guc_interrupts_capture(struct drm_i915_private *dev_priv) dev_priv->rps.pm_intrmsk_mbz &= ~GEN8_PMINTR_DISABLE_REDIRECT_TO_GUC; } +static void guc_interrupts_release(struct drm_i915_private *dev_priv) +{ + struct intel_engine_cs *engine; + enum intel_engine_id id; + int irqs; + + /* + * tell all command streamers NOT to forward interrupts or vblank + * to GuC. + */ + irqs = _MASKED_FIELD(GFX_FORWARD_VBLANK_MASK, GFX_FORWARD_VBLANK_NEVER); + irqs |= _MASKED_BIT_DISABLE(GFX_INTERRUPT_STEERING); + for_each_engine(engine, dev_priv, id) + I915_WRITE(RING_MODE_GEN7(engine), irqs); + + /* route all GT interrupts to the host */ + I915_WRITE(GUC_BCS_RCS_IER, 0); + I915_WRITE(GUC_VCS2_VCS1_IER, 0); + I915_WRITE(GUC_WD_VECS_IER, 0); + + dev_priv->rps.pm_intrmsk_mbz |= GEN8_PMINTR_DISABLE_REDIRECT_TO_GUC; + dev_priv->rps.pm_intrmsk_mbz &= ~ARAT_EXPIRED_INTRMSK; +} + int i915_guc_submission_enable(struct drm_i915_private *dev_priv) { struct intel_guc *guc = &dev_priv->guc; struct i915_guc_client *client = guc->execbuf_client; struct intel_engine_cs *engine; enum intel_engine_id id; + int err; + + if (!client) { + client = guc_client_alloc(dev_priv, + INTEL_INFO(dev_priv)->ring_mask, + GUC_CLIENT_PRIORITY_KMD_NORMAL, + dev_priv->kernel_context); + if (IS_ERR(client)) { + DRM_ERROR("Failed to create GuC client for execbuf!\n"); + return PTR_ERR(client); + } - if (!client) - return -ENODEV; + guc->execbuf_client = client; + } - intel_guc_sample_forcewake(guc); + err = intel_guc_sample_forcewake(guc); + if (err) + goto err_execbuf_client; guc_reset_wq(client); - guc_init_doorbell_hw(guc); + + err = guc_init_doorbell_hw(guc); + if (err) + goto err_execbuf_client; /* Take over from manual control of ELSP (execlists) */ guc_interrupts_capture(dev_priv); @@ -1112,30 +1255,11 @@ int i915_guc_submission_enable(struct drm_i915_private *dev_priv) } return 0; -} -static void guc_interrupts_release(struct drm_i915_private *dev_priv) -{ - struct intel_engine_cs *engine; - enum intel_engine_id id; - int irqs; - - /* - * tell all command streamers NOT to forward interrupts or vblank - * to GuC. - */ - irqs = _MASKED_FIELD(GFX_FORWARD_VBLANK_MASK, GFX_FORWARD_VBLANK_NEVER); - irqs |= _MASKED_BIT_DISABLE(GFX_INTERRUPT_STEERING); - for_each_engine(engine, dev_priv, id) - I915_WRITE(RING_MODE_GEN7(engine), irqs); - - /* route all GT interrupts to the host */ - I915_WRITE(GUC_BCS_RCS_IER, 0); - I915_WRITE(GUC_VCS2_VCS1_IER, 0); - I915_WRITE(GUC_WD_VECS_IER, 0); - - dev_priv->rps.pm_intrmsk_mbz |= GEN8_PMINTR_DISABLE_REDIRECT_TO_GUC; - dev_priv->rps.pm_intrmsk_mbz &= ~ARAT_EXPIRED_INTRMSK; +err_execbuf_client: + guc_client_free(guc->execbuf_client); + guc->execbuf_client = NULL; + return err; } void i915_guc_submission_disable(struct drm_i915_private *dev_priv) @@ -1144,30 +1268,11 @@ void i915_guc_submission_disable(struct drm_i915_private *dev_priv) guc_interrupts_release(dev_priv); - if (!guc->execbuf_client) - return; - /* Revert back to manual ELSP submission */ intel_engines_reset_default_submission(dev_priv); -} - -void i915_guc_submission_fini(struct drm_i915_private *dev_priv) -{ - struct intel_guc *guc = &dev_priv->guc; - struct i915_guc_client *client; - client = fetch_and_zero(&guc->execbuf_client); - if (!client) - return; - - guc_client_free(dev_priv, client); - - i915_vma_unpin_and_release(&guc->ads_vma); - i915_vma_unpin_and_release(&guc->log.vma); - - if (guc->ctx_pool_vma) - ida_destroy(&guc->ctx_ids); - i915_vma_unpin_and_release(&guc->ctx_pool_vma); + guc_client_free(guc->execbuf_client); + guc->execbuf_client = NULL; } /** @@ -1196,7 +1301,6 @@ int intel_guc_suspend(struct drm_i915_private *dev_priv) return intel_guc_send(guc, data, ARRAY_SIZE(data)); } - /** * intel_guc_resume() - notify GuC resuming from suspend state * @dev_priv: i915 device private @@ -1222,5 +1326,3 @@ int intel_guc_resume(struct drm_i915_private *dev_priv) return intel_guc_send(guc, data, ARRAY_SIZE(data)); } - - diff --git a/drivers/gpu/drm/i915/i915_irq.c b/drivers/gpu/drm/i915/i915_irq.c index 8163d5024ff8..d9d196977f4a 100644 --- a/drivers/gpu/drm/i915/i915_irq.c +++ b/drivers/gpu/drm/i915/i915_irq.c @@ -1742,8 +1742,8 @@ static void gen9_guc_irq_handler(struct drm_i915_private *dev_priv, u32 gt_iir) I915_WRITE(SOFT_SCRATCH(15), msg & ~flush); /* Handle flush interrupt in bottom half */ - queue_work(dev_priv->guc.log.flush_wq, - &dev_priv->guc.log.flush_work); + queue_work(dev_priv->guc.log.runtime.flush_wq, + &dev_priv->guc.log.runtime.flush_work); dev_priv->guc.log.flush_interrupt_count++; } else { diff --git a/drivers/gpu/drm/i915/i915_pci.c b/drivers/gpu/drm/i915/i915_pci.c index 732101ed57fb..f87b0c4e564d 100644 --- a/drivers/gpu/drm/i915/i915_pci.c +++ b/drivers/gpu/drm/i915/i915_pci.c @@ -61,6 +61,7 @@ .has_overlay = 1, .overlay_needs_physical = 1, \ .has_gmch_display = 1, \ .hws_needs_physical = 1, \ + .unfenced_needs_alignment = 1, \ .ring_mask = RENDER_RING, \ GEN_DEFAULT_PIPEOFFSETS, \ CURSOR_OFFSETS @@ -102,6 +103,7 @@ static const struct intel_device_info intel_i915g_info = { .platform = INTEL_I915G, .cursor_needs_physical = 1, .has_overlay = 1, .overlay_needs_physical = 1, .hws_needs_physical = 1, + .unfenced_needs_alignment = 1, }; static const struct intel_device_info intel_i915gm_info = { @@ -113,6 +115,7 @@ static const struct intel_device_info intel_i915gm_info = { .supports_tv = 1, .has_fbc = 1, .hws_needs_physical = 1, + .unfenced_needs_alignment = 1, }; static const struct intel_device_info intel_i945g_info = { @@ -121,6 +124,7 @@ static const struct intel_device_info intel_i945g_info = { .has_hotplug = 1, .cursor_needs_physical = 1, .has_overlay = 1, .overlay_needs_physical = 1, .hws_needs_physical = 1, + .unfenced_needs_alignment = 1, }; static const struct intel_device_info intel_i945gm_info = { @@ -131,6 +135,7 @@ static const struct intel_device_info intel_i945gm_info = { .supports_tv = 1, .has_fbc = 1, .hws_needs_physical = 1, + .unfenced_needs_alignment = 1, }; static const struct intel_device_info intel_g33_info = { diff --git a/drivers/gpu/drm/i915/i915_perf.c b/drivers/gpu/drm/i915/i915_perf.c index 8c121187ff39..060b171480d5 100644 --- a/drivers/gpu/drm/i915/i915_perf.c +++ b/drivers/gpu/drm/i915/i915_perf.c @@ -1705,7 +1705,7 @@ i915_perf_open_ioctl_locked(struct drm_i915_private *dev_priv, */ if (WARN_ON(stream->sample_flags != props->sample_flags)) { ret = -ENODEV; - goto err_alloc; + goto err_flags; } list_add(&stream->link, &dev_priv->perf.streams); @@ -1728,6 +1728,7 @@ i915_perf_open_ioctl_locked(struct drm_i915_private *dev_priv, err_open: list_del(&stream->link); +err_flags: if (stream->ops->destroy) stream->ops->destroy(stream); err_alloc: @@ -1793,6 +1794,11 @@ static int read_properties_unlocked(struct drm_i915_private *dev_priv, if (ret) return ret; + if (id == 0 || id >= DRM_I915_PERF_PROP_MAX) { + DRM_DEBUG("Unknown i915 perf property ID\n"); + return -EINVAL; + } + switch ((enum drm_i915_perf_property_id)id) { case DRM_I915_PERF_PROP_CTX_HANDLE: props->single_context = 1; @@ -1862,9 +1868,8 @@ static int read_properties_unlocked(struct drm_i915_private *dev_priv, props->oa_periodic = true; props->oa_period_exponent = value; break; - default: + case DRM_I915_PERF_PROP_MAX: MISSING_CASE(id); - DRM_DEBUG("Unknown i915 perf property ID\n"); return -EINVAL; } diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h index 04c8f69fcc62..11b12f412492 100644 --- a/drivers/gpu/drm/i915/i915_reg.h +++ b/drivers/gpu/drm/i915/i915_reg.h @@ -7829,7 +7829,14 @@ enum { #define TRANS_DDI_EDP_INPUT_B_ONOFF (5<<12) #define TRANS_DDI_EDP_INPUT_C_ONOFF (6<<12) #define TRANS_DDI_DP_VC_PAYLOAD_ALLOC (1<<8) +#define TRANS_DDI_HDMI_SCRAMBLER_CTS_ENABLE (1<<7) +#define TRANS_DDI_HDMI_SCRAMBLER_RESET_FREQ (1<<6) #define TRANS_DDI_BFI_ENABLE (1<<4) +#define TRANS_DDI_HIGH_TMDS_CHAR_RATE (1<<4) +#define TRANS_DDI_HDMI_SCRAMBLING (1<<0) +#define TRANS_DDI_HDMI_SCRAMBLING_MASK (TRANS_DDI_HDMI_SCRAMBLER_CTS_ENABLE \ + | TRANS_DDI_HDMI_SCRAMBLER_RESET_FREQ \ + | TRANS_DDI_HDMI_SCRAMBLING) /* DisplayPort Transport Control */ #define _DP_TP_CTL_A 0x64040 diff --git a/drivers/gpu/drm/i915/i915_utils.h b/drivers/gpu/drm/i915/i915_utils.h index 94a3a3299910..c5455d36b617 100644 --- a/drivers/gpu/drm/i915/i915_utils.h +++ b/drivers/gpu/drm/i915/i915_utils.h @@ -25,6 +25,24 @@ #ifndef __I915_UTILS_H #define __I915_UTILS_H +#undef WARN_ON +/* Many gcc seem to no see through this and fall over :( */ +#if 0 +#define WARN_ON(x) ({ \ + bool __i915_warn_cond = (x); \ + if (__builtin_constant_p(__i915_warn_cond)) \ + BUILD_BUG_ON(__i915_warn_cond); \ + WARN(__i915_warn_cond, "WARN_ON(" #x ")"); }) +#else +#define WARN_ON(x) WARN((x), "%s", "WARN_ON(" __stringify(x) ")") +#endif + +#undef WARN_ON_ONCE +#define WARN_ON_ONCE(x) WARN_ONCE((x), "%s", "WARN_ON_ONCE(" __stringify(x) ")") + +#define MISSING_CASE(x) WARN(1, "Missing switch case (%lu) in %s\n", \ + (long)(x), __func__) + #if GCC_VERSION >= 70000 #define add_overflows(A, B) \ __builtin_add_overflow_p((A), (B), (typeof((A) + (B)))0) diff --git a/drivers/gpu/drm/i915/intel_breadcrumbs.c b/drivers/gpu/drm/i915/intel_breadcrumbs.c index ba986edee312..b6ea192ad550 100644 --- a/drivers/gpu/drm/i915/intel_breadcrumbs.c +++ b/drivers/gpu/drm/i915/intel_breadcrumbs.c @@ -47,11 +47,12 @@ static unsigned int __intel_breadcrumbs_wakeup(struct intel_breadcrumbs *b) unsigned int intel_engine_wakeup(struct intel_engine_cs *engine) { struct intel_breadcrumbs *b = &engine->breadcrumbs; + unsigned long flags; unsigned int result; - spin_lock_irq(&b->irq_lock); + spin_lock_irqsave(&b->irq_lock, flags); result = __intel_breadcrumbs_wakeup(b); - spin_unlock_irq(&b->irq_lock); + spin_unlock_irqrestore(&b->irq_lock, flags); return result; } diff --git a/drivers/gpu/drm/i915/intel_cdclk.c b/drivers/gpu/drm/i915/intel_cdclk.c index c2cc33f3d888..dd3ad52b7dfe 100644 --- a/drivers/gpu/drm/i915/intel_cdclk.c +++ b/drivers/gpu/drm/i915/intel_cdclk.c @@ -1442,16 +1442,33 @@ static int bdw_adjust_min_pipe_pixel_rate(struct intel_crtc_state *crtc_state, if (IS_BROADWELL(dev_priv) && crtc_state->ips_enabled) pixel_rate = DIV_ROUND_UP(pixel_rate * 100, 95); - /* BSpec says "Do not use DisplayPort with CDCLK less than - * 432 MHz, audio enabled, port width x4, and link rate - * HBR2 (5.4 GHz), or else there may be audio corruption or - * screen corruption." + /* BSpec says "Do not use DisplayPort with CDCLK less than 432 MHz, + * audio enabled, port width x4, and link rate HBR2 (5.4 GHz), or else + * there may be audio corruption or screen corruption." This cdclk + * restriction for GLK is 316.8 MHz and since GLK can output two + * pixels per clock, the pixel rate becomes 2 * 316.8 MHz. */ if (intel_crtc_has_dp_encoder(crtc_state) && crtc_state->has_audio && crtc_state->port_clock >= 540000 && - crtc_state->lane_count == 4) - pixel_rate = max(432000, pixel_rate); + crtc_state->lane_count == 4) { + if (IS_GEMINILAKE(dev_priv)) + pixel_rate = max(2 * 316800, pixel_rate); + else + pixel_rate = max(432000, pixel_rate); + } + + /* According to BSpec, "The CD clock frequency must be at least twice + * the frequency of the Azalia BCLK." and BCLK is 96 MHz by default. + * The check for GLK has to be adjusted as the platform can output + * two pixels per clock. + */ + if (crtc_state->has_audio && INTEL_GEN(dev_priv) >= 9) { + if (IS_GEMINILAKE(dev_priv)) + pixel_rate = max(2 * 2 * 96000, pixel_rate); + else + pixel_rate = max(2 * 96000, pixel_rate); + } return pixel_rate; } diff --git a/drivers/gpu/drm/i915/intel_csr.c b/drivers/gpu/drm/i915/intel_csr.c index 36832257cc9b..1575bde0cf90 100644 --- a/drivers/gpu/drm/i915/intel_csr.c +++ b/drivers/gpu/drm/i915/intel_csr.c @@ -49,7 +49,7 @@ MODULE_FIRMWARE(I915_CSR_SKL); MODULE_FIRMWARE(I915_CSR_BXT); #define BXT_CSR_VERSION_REQUIRED CSR_VERSION(1, 7) -#define FIRMWARE_URL "https://01.org/linuxgraphics/intel-linux-graphics-firmwares" +#define FIRMWARE_URL "https://01.org/linuxgraphics/downloads/firmware" diff --git a/drivers/gpu/drm/i915/intel_ddi.c b/drivers/gpu/drm/i915/intel_ddi.c index d8214ba8da14..0914ad96a71b 100644 --- a/drivers/gpu/drm/i915/intel_ddi.c +++ b/drivers/gpu/drm/i915/intel_ddi.c @@ -539,7 +539,7 @@ intel_ddi_get_buf_trans_fdi(struct drm_i915_private *dev_priv, * values in advance. This function programs the correct values for * DP/eDP/FDI use cases. */ -void intel_prepare_dp_ddi_buffers(struct intel_encoder *encoder) +static void intel_prepare_dp_ddi_buffers(struct intel_encoder *encoder) { struct drm_i915_private *dev_priv = to_i915(encoder->base.dev); u32 iboost_bit = 0; @@ -806,7 +806,7 @@ void hsw_fdi_link_train(struct intel_crtc *crtc, DP_TP_CTL_ENABLE); } -void intel_ddi_init_dp_buf_reg(struct intel_encoder *encoder) +static void intel_ddi_init_dp_buf_reg(struct intel_encoder *encoder) { struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base); struct intel_digital_port *intel_dig_port = @@ -837,7 +837,8 @@ intel_ddi_get_crtc_encoder(struct intel_crtc *crtc) return ret; } -static struct intel_encoder * +/* Finds the only possible encoder associated with the given CRTC. */ +struct intel_encoder * intel_ddi_get_crtc_new_encoder(struct intel_crtc_state *crtc_state) { struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc); @@ -1127,72 +1128,6 @@ void intel_ddi_clock_get(struct intel_encoder *encoder, bxt_ddi_clock_get(encoder, pipe_config); } -static bool -hsw_ddi_pll_select(struct intel_crtc *intel_crtc, - struct intel_crtc_state *crtc_state, - struct intel_encoder *encoder) -{ - struct intel_shared_dpll *pll; - - pll = intel_get_shared_dpll(intel_crtc, crtc_state, - encoder); - if (!pll) - DRM_DEBUG_DRIVER("failed to find PLL for pipe %c\n", - pipe_name(intel_crtc->pipe)); - - return pll; -} - -static bool -skl_ddi_pll_select(struct intel_crtc *intel_crtc, - struct intel_crtc_state *crtc_state, - struct intel_encoder *encoder) -{ - struct intel_shared_dpll *pll; - - pll = intel_get_shared_dpll(intel_crtc, crtc_state, encoder); - if (pll == NULL) { - DRM_DEBUG_DRIVER("failed to find PLL for pipe %c\n", - pipe_name(intel_crtc->pipe)); - return false; - } - - return true; -} - -static bool -bxt_ddi_pll_select(struct intel_crtc *intel_crtc, - struct intel_crtc_state *crtc_state, - struct intel_encoder *encoder) -{ - return !!intel_get_shared_dpll(intel_crtc, crtc_state, encoder); -} - -/* - * Tries to find a *shared* PLL for the CRTC and store it in - * intel_crtc->ddi_pll_sel. - * - * For private DPLLs, compute_config() should do the selection for us. This - * function should be folded into compute_config() eventually. - */ -bool intel_ddi_pll_select(struct intel_crtc *intel_crtc, - struct intel_crtc_state *crtc_state) -{ - struct drm_i915_private *dev_priv = to_i915(intel_crtc->base.dev); - struct intel_encoder *encoder = - intel_ddi_get_crtc_new_encoder(crtc_state); - - if (IS_GEN9_BC(dev_priv)) - return skl_ddi_pll_select(intel_crtc, crtc_state, - encoder); - else if (IS_GEN9_LP(dev_priv)) - return bxt_ddi_pll_select(intel_crtc, crtc_state, - encoder); - else - return hsw_ddi_pll_select(intel_crtc, crtc_state, - encoder); -} - void intel_ddi_set_pipe_settings(const struct intel_crtc_state *crtc_state) { struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc); @@ -1309,6 +1244,11 @@ void intel_ddi_enable_transcoder_func(const struct intel_crtc_state *crtc_state) temp |= TRANS_DDI_MODE_SELECT_HDMI; else temp |= TRANS_DDI_MODE_SELECT_DVI; + + if (crtc_state->hdmi_scrambling) + temp |= TRANS_DDI_HDMI_SCRAMBLING_MASK; + if (crtc_state->hdmi_high_tmds_clock_ratio) + temp |= TRANS_DDI_HIGH_TMDS_CHAR_RATE; } else if (type == INTEL_OUTPUT_ANALOG) { temp |= TRANS_DDI_MODE_SELECT_FDI; temp |= (crtc_state->fdi_lanes - 1) << 1; @@ -1676,8 +1616,8 @@ uint32_t ddi_signal_levels(struct intel_dp *intel_dp) return DDI_BUF_TRANS_SELECT(level); } -void intel_ddi_clk_select(struct intel_encoder *encoder, - struct intel_shared_dpll *pll) +static void intel_ddi_clk_select(struct intel_encoder *encoder, + struct intel_shared_dpll *pll) { struct drm_i915_private *dev_priv = to_i915(encoder->base.dev); enum port port = intel_ddi_get_encoder_port(encoder); @@ -1881,6 +1821,12 @@ static void intel_enable_ddi(struct intel_encoder *intel_encoder, if (type == INTEL_OUTPUT_HDMI) { struct intel_digital_port *intel_dig_port = enc_to_dig_port(encoder); + bool clock_ratio = pipe_config->hdmi_high_tmds_clock_ratio; + bool scrambling = pipe_config->hdmi_scrambling; + + intel_hdmi_handle_sink_scrambling(intel_encoder, + conn_state->connector, + clock_ratio, scrambling); /* In HDMI/DVI mode, the port width, and swing/emphasis values * are ignored so nothing special needs to be done besides @@ -1914,6 +1860,12 @@ static void intel_disable_ddi(struct intel_encoder *intel_encoder, if (old_crtc_state->has_audio) intel_audio_codec_disable(intel_encoder); + if (type == INTEL_OUTPUT_HDMI) { + intel_hdmi_handle_sink_scrambling(intel_encoder, + old_conn_state->connector, + false, false); + } + if (type == INTEL_OUTPUT_EDP) { struct intel_dp *intel_dp = enc_to_intel_dp(encoder); @@ -2040,6 +1992,12 @@ void intel_ddi_get_config(struct intel_encoder *encoder, if (intel_hdmi->infoframe_enabled(&encoder->base, pipe_config)) pipe_config->has_infoframe = true; + + if ((temp & TRANS_DDI_HDMI_SCRAMBLING_MASK) == + TRANS_DDI_HDMI_SCRAMBLING_MASK) + pipe_config->hdmi_scrambling = true; + if (temp & TRANS_DDI_HIGH_TMDS_CHAR_RATE) + pipe_config->hdmi_high_tmds_clock_ratio = true; /* fall through */ case TRANS_DDI_MODE_SELECT_DVI: pipe_config->lane_count = 4; diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index e217d04133e6..881dec88df6e 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -1997,7 +1997,7 @@ intel_tile_width_bytes(const struct drm_framebuffer *fb, int plane) unsigned int cpp = fb->format->cpp[plane]; switch (fb->modifier) { - case DRM_FORMAT_MOD_NONE: + case DRM_FORMAT_MOD_LINEAR: return cpp; case I915_FORMAT_MOD_X_TILED: if (IS_GEN2(dev_priv)) @@ -2033,7 +2033,7 @@ intel_tile_width_bytes(const struct drm_framebuffer *fb, int plane) static unsigned int intel_tile_height(const struct drm_framebuffer *fb, int plane) { - if (fb->modifier == DRM_FORMAT_MOD_NONE) + if (fb->modifier == DRM_FORMAT_MOD_LINEAR) return 1; else return intel_tile_size(to_i915(fb->dev)) / @@ -2107,7 +2107,7 @@ static unsigned int intel_surf_alignment(const struct drm_framebuffer *fb, return 4096; switch (fb->modifier) { - case DRM_FORMAT_MOD_NONE: + case DRM_FORMAT_MOD_LINEAR: return intel_linear_alignment(dev_priv); case I915_FORMAT_MOD_X_TILED: if (INTEL_GEN(dev_priv) >= 9) @@ -2290,7 +2290,7 @@ static u32 intel_adjust_tile_offset(int *x, int *y, WARN_ON(new_offset > old_offset); - if (fb->modifier != DRM_FORMAT_MOD_NONE) { + if (fb->modifier != DRM_FORMAT_MOD_LINEAR) { unsigned int tile_size, tile_width, tile_height; unsigned int pitch_tiles; @@ -2345,7 +2345,7 @@ static u32 _intel_compute_tile_offset(const struct drm_i915_private *dev_priv, if (alignment) alignment--; - if (fb_modifier != DRM_FORMAT_MOD_NONE) { + if (fb_modifier != DRM_FORMAT_MOD_LINEAR) { unsigned int tile_size, tile_width, tile_height; unsigned int tile_rows, tiles, pitch_tiles; @@ -2471,7 +2471,7 @@ intel_fill_fb_info(struct drm_i915_private *dev_priv, DRM_ROTATE_0, tile_size); offset /= tile_size; - if (fb->modifier != DRM_FORMAT_MOD_NONE) { + if (fb->modifier != DRM_FORMAT_MOD_LINEAR) { unsigned int tile_width, tile_height; unsigned int pitch_tiles; struct drm_rect r; @@ -2803,7 +2803,7 @@ static int skl_max_plane_width(const struct drm_framebuffer *fb, int plane, int cpp = fb->format->cpp[plane]; switch (fb->modifier) { - case DRM_FORMAT_MOD_NONE: + case DRM_FORMAT_MOD_LINEAR: case I915_FORMAT_MOD_X_TILED: switch (cpp) { case 8: @@ -2962,28 +2962,27 @@ int skl_check_plane_surface(struct intel_plane_state *plane_state) return 0; } -static void i9xx_update_primary_plane(struct drm_plane *primary, - const struct intel_crtc_state *crtc_state, - const struct intel_plane_state *plane_state) +static u32 i9xx_plane_ctl(const struct intel_crtc_state *crtc_state, + const struct intel_plane_state *plane_state) { - struct drm_i915_private *dev_priv = to_i915(primary->dev); - struct intel_crtc *intel_crtc = to_intel_crtc(crtc_state->base.crtc); - struct drm_framebuffer *fb = plane_state->base.fb; - int plane = intel_crtc->plane; - u32 linear_offset; - u32 dspcntr; - i915_reg_t reg = DSPCNTR(plane); + struct drm_i915_private *dev_priv = + to_i915(plane_state->base.plane->dev); + struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc); + const struct drm_framebuffer *fb = plane_state->base.fb; unsigned int rotation = plane_state->base.rotation; - int x = plane_state->base.src.x1 >> 16; - int y = plane_state->base.src.y1 >> 16; - unsigned long irqflags; + u32 dspcntr; - dspcntr = DISPPLANE_GAMMA_ENABLE; + dspcntr = DISPLAY_PLANE_ENABLE | DISPPLANE_GAMMA_ENABLE; + + if (IS_G4X(dev_priv) || IS_GEN5(dev_priv) || + IS_GEN6(dev_priv) || IS_IVYBRIDGE(dev_priv)) + dspcntr |= DISPPLANE_TRICKLE_FEED_DISABLE; - dspcntr |= DISPLAY_PLANE_ENABLE; + if (IS_HASWELL(dev_priv) || IS_BROADWELL(dev_priv)) + dspcntr |= DISPPLANE_PIPE_CSC_ENABLE; if (INTEL_GEN(dev_priv) < 4) { - if (intel_crtc->pipe == PIPE_B) + if (crtc->pipe == PIPE_B) dspcntr |= DISPPLANE_SEL_PIPE_B; } @@ -3010,7 +3009,8 @@ static void i9xx_update_primary_plane(struct drm_plane *primary, dspcntr |= DISPPLANE_RGBX101010; break; default: - BUG(); + MISSING_CASE(fb->format->format); + return 0; } if (INTEL_GEN(dev_priv) >= 4 && @@ -3023,25 +3023,66 @@ static void i9xx_update_primary_plane(struct drm_plane *primary, if (rotation & DRM_REFLECT_X) dspcntr |= DISPPLANE_MIRROR; - if (IS_G4X(dev_priv)) - dspcntr |= DISPPLANE_TRICKLE_FEED_DISABLE; + return dspcntr; +} - intel_add_fb_offsets(&x, &y, plane_state, 0); +int i9xx_check_plane_surface(struct intel_plane_state *plane_state) +{ + struct drm_i915_private *dev_priv = + to_i915(plane_state->base.plane->dev); + int src_x = plane_state->base.src.x1 >> 16; + int src_y = plane_state->base.src.y1 >> 16; + u32 offset; + + intel_add_fb_offsets(&src_x, &src_y, plane_state, 0); if (INTEL_GEN(dev_priv) >= 4) - intel_crtc->dspaddr_offset = - intel_compute_tile_offset(&x, &y, plane_state, 0); + offset = intel_compute_tile_offset(&src_x, &src_y, + plane_state, 0); + else + offset = 0; + + /* HSW/BDW do this automagically in hardware */ + if (!IS_HASWELL(dev_priv) && !IS_BROADWELL(dev_priv)) { + unsigned int rotation = plane_state->base.rotation; + int src_w = drm_rect_width(&plane_state->base.src) >> 16; + int src_h = drm_rect_height(&plane_state->base.src) >> 16; - if (rotation & DRM_ROTATE_180) { - x += crtc_state->pipe_src_w - 1; - y += crtc_state->pipe_src_h - 1; - } else if (rotation & DRM_REFLECT_X) { - x += crtc_state->pipe_src_w - 1; + if (rotation & DRM_ROTATE_180) { + src_x += src_w - 1; + src_y += src_h - 1; + } else if (rotation & DRM_REFLECT_X) { + src_x += src_w - 1; + } } + plane_state->main.offset = offset; + plane_state->main.x = src_x; + plane_state->main.y = src_y; + + return 0; +} + +static void i9xx_update_primary_plane(struct drm_plane *primary, + const struct intel_crtc_state *crtc_state, + const struct intel_plane_state *plane_state) +{ + struct drm_i915_private *dev_priv = to_i915(primary->dev); + struct intel_crtc *intel_crtc = to_intel_crtc(crtc_state->base.crtc); + struct drm_framebuffer *fb = plane_state->base.fb; + int plane = intel_crtc->plane; + u32 linear_offset; + u32 dspcntr = plane_state->ctl; + i915_reg_t reg = DSPCNTR(plane); + int x = plane_state->main.x; + int y = plane_state->main.y; + unsigned long irqflags; + linear_offset = intel_fb_xy_to_linear(x, y, plane_state, 0); - if (INTEL_GEN(dev_priv) < 4) + if (INTEL_GEN(dev_priv) >= 4) + intel_crtc->dspaddr_offset = plane_state->main.offset; + else intel_crtc->dspaddr_offset = linear_offset; intel_crtc->adjusted_x = x; @@ -3068,7 +3109,12 @@ static void i9xx_update_primary_plane(struct drm_plane *primary, I915_WRITE_FW(reg, dspcntr); I915_WRITE_FW(DSPSTRIDE(plane), fb->pitches[0]); - if (INTEL_GEN(dev_priv) >= 4) { + if (IS_HASWELL(dev_priv) || IS_BROADWELL(dev_priv)) { + I915_WRITE_FW(DSPSURF(plane), + intel_plane_ggtt_offset(plane_state) + + intel_crtc->dspaddr_offset); + I915_WRITE_FW(DSPOFFSET(plane), (y << 16) | x); + } else if (INTEL_GEN(dev_priv) >= 4) { I915_WRITE_FW(DSPSURF(plane), intel_plane_ggtt_offset(plane_state) + intel_crtc->dspaddr_offset); @@ -3105,101 +3151,10 @@ static void i9xx_disable_primary_plane(struct drm_plane *primary, spin_unlock_irqrestore(&dev_priv->uncore.lock, irqflags); } -static void ironlake_update_primary_plane(struct drm_plane *primary, - const struct intel_crtc_state *crtc_state, - const struct intel_plane_state *plane_state) -{ - struct drm_device *dev = primary->dev; - struct drm_i915_private *dev_priv = to_i915(dev); - struct intel_crtc *intel_crtc = to_intel_crtc(crtc_state->base.crtc); - struct drm_framebuffer *fb = plane_state->base.fb; - int plane = intel_crtc->plane; - u32 linear_offset; - u32 dspcntr; - i915_reg_t reg = DSPCNTR(plane); - unsigned int rotation = plane_state->base.rotation; - int x = plane_state->base.src.x1 >> 16; - int y = plane_state->base.src.y1 >> 16; - unsigned long irqflags; - - dspcntr = DISPPLANE_GAMMA_ENABLE; - dspcntr |= DISPLAY_PLANE_ENABLE; - - if (IS_HASWELL(dev_priv) || IS_BROADWELL(dev_priv)) - dspcntr |= DISPPLANE_PIPE_CSC_ENABLE; - - switch (fb->format->format) { - case DRM_FORMAT_C8: - dspcntr |= DISPPLANE_8BPP; - break; - case DRM_FORMAT_RGB565: - dspcntr |= DISPPLANE_BGRX565; - break; - case DRM_FORMAT_XRGB8888: - dspcntr |= DISPPLANE_BGRX888; - break; - case DRM_FORMAT_XBGR8888: - dspcntr |= DISPPLANE_RGBX888; - break; - case DRM_FORMAT_XRGB2101010: - dspcntr |= DISPPLANE_BGRX101010; - break; - case DRM_FORMAT_XBGR2101010: - dspcntr |= DISPPLANE_RGBX101010; - break; - default: - BUG(); - } - - if (fb->modifier == I915_FORMAT_MOD_X_TILED) - dspcntr |= DISPPLANE_TILED; - - if (rotation & DRM_ROTATE_180) - dspcntr |= DISPPLANE_ROTATE_180; - - if (!IS_HASWELL(dev_priv) && !IS_BROADWELL(dev_priv)) - dspcntr |= DISPPLANE_TRICKLE_FEED_DISABLE; - - intel_add_fb_offsets(&x, &y, plane_state, 0); - - intel_crtc->dspaddr_offset = - intel_compute_tile_offset(&x, &y, plane_state, 0); - - /* HSW+ does this automagically in hardware */ - if (!IS_HASWELL(dev_priv) && !IS_BROADWELL(dev_priv) && - rotation & DRM_ROTATE_180) { - x += crtc_state->pipe_src_w - 1; - y += crtc_state->pipe_src_h - 1; - } - - linear_offset = intel_fb_xy_to_linear(x, y, plane_state, 0); - - intel_crtc->adjusted_x = x; - intel_crtc->adjusted_y = y; - - spin_lock_irqsave(&dev_priv->uncore.lock, irqflags); - - I915_WRITE_FW(reg, dspcntr); - - I915_WRITE_FW(DSPSTRIDE(plane), fb->pitches[0]); - I915_WRITE_FW(DSPSURF(plane), - intel_plane_ggtt_offset(plane_state) + - intel_crtc->dspaddr_offset); - if (IS_HASWELL(dev_priv) || IS_BROADWELL(dev_priv)) { - I915_WRITE_FW(DSPOFFSET(plane), (y << 16) | x); - } else { - I915_WRITE_FW(DSPTILEOFF(plane), (y << 16) | x); - I915_WRITE_FW(DSPLINOFF(plane), linear_offset); - } - POSTING_READ_FW(reg); - - spin_unlock_irqrestore(&dev_priv->uncore.lock, irqflags); -} - static u32 intel_fb_stride_alignment(const struct drm_framebuffer *fb, int plane) { - if (fb->modifier == DRM_FORMAT_MOD_NONE) + if (fb->modifier == DRM_FORMAT_MOD_LINEAR) return 64; else return intel_tile_width_bytes(fb, plane); @@ -3254,7 +3209,7 @@ u32 skl_plane_stride(const struct drm_framebuffer *fb, int plane, return stride; } -u32 skl_plane_ctl_format(uint32_t pixel_format) +static u32 skl_plane_ctl_format(uint32_t pixel_format) { switch (pixel_format) { case DRM_FORMAT_C8: @@ -3295,10 +3250,10 @@ u32 skl_plane_ctl_format(uint32_t pixel_format) return 0; } -u32 skl_plane_ctl_tiling(uint64_t fb_modifier) +static u32 skl_plane_ctl_tiling(uint64_t fb_modifier) { switch (fb_modifier) { - case DRM_FORMAT_MOD_NONE: + case DRM_FORMAT_MOD_LINEAR: break; case I915_FORMAT_MOD_X_TILED: return PLANE_CTL_TILED_X; @@ -3313,7 +3268,7 @@ u32 skl_plane_ctl_tiling(uint64_t fb_modifier) return 0; } -u32 skl_plane_ctl_rotation(unsigned int rotation) +static u32 skl_plane_ctl_rotation(unsigned int rotation) { switch (rotation) { case DRM_ROTATE_0: @@ -3335,6 +3290,37 @@ u32 skl_plane_ctl_rotation(unsigned int rotation) return 0; } +u32 skl_plane_ctl(const struct intel_crtc_state *crtc_state, + const struct intel_plane_state *plane_state) +{ + struct drm_i915_private *dev_priv = + to_i915(plane_state->base.plane->dev); + const struct drm_framebuffer *fb = plane_state->base.fb; + unsigned int rotation = plane_state->base.rotation; + const struct drm_intel_sprite_colorkey *key = &plane_state->ckey; + u32 plane_ctl; + + plane_ctl = PLANE_CTL_ENABLE; + + if (!IS_GEMINILAKE(dev_priv)) { + plane_ctl |= + PLANE_CTL_PIPE_GAMMA_ENABLE | + PLANE_CTL_PIPE_CSC_ENABLE | + PLANE_CTL_PLANE_GAMMA_DISABLE; + } + + plane_ctl |= skl_plane_ctl_format(fb->format->format); + plane_ctl |= skl_plane_ctl_tiling(fb->modifier); + plane_ctl |= skl_plane_ctl_rotation(rotation); + + if (key->flags & I915_SET_COLORKEY_DESTINATION) + plane_ctl |= PLANE_CTL_KEY_ENABLE_DESTINATION; + else if (key->flags & I915_SET_COLORKEY_SOURCE) + plane_ctl |= PLANE_CTL_KEY_ENABLE_SOURCE; + + return plane_ctl; +} + static void skylake_update_primary_plane(struct drm_plane *plane, const struct intel_crtc_state *crtc_state, const struct intel_plane_state *plane_state) @@ -3345,7 +3331,7 @@ static void skylake_update_primary_plane(struct drm_plane *plane, struct drm_framebuffer *fb = plane_state->base.fb; enum plane_id plane_id = to_intel_plane(plane)->id; enum pipe pipe = to_intel_plane(plane)->pipe; - u32 plane_ctl; + u32 plane_ctl = plane_state->ctl; unsigned int rotation = plane_state->base.rotation; u32 stride = skl_plane_stride(fb, 0, rotation); u32 surf_addr = plane_state->main.offset; @@ -3360,19 +3346,6 @@ static void skylake_update_primary_plane(struct drm_plane *plane, int dst_h = drm_rect_height(&plane_state->base.dst); unsigned long irqflags; - plane_ctl = PLANE_CTL_ENABLE; - - if (!IS_GEMINILAKE(dev_priv)) { - plane_ctl |= - PLANE_CTL_PIPE_GAMMA_ENABLE | - PLANE_CTL_PIPE_CSC_ENABLE | - PLANE_CTL_PLANE_GAMMA_DISABLE; - } - - plane_ctl |= skl_plane_ctl_format(fb->format->format); - plane_ctl |= skl_plane_ctl_tiling(fb->modifier); - plane_ctl |= skl_plane_ctl_rotation(rotation); - /* Sizes are 0 based */ src_w--; src_h--; @@ -6306,6 +6279,17 @@ intel_reduce_m_n_ratio(uint32_t *num, uint32_t *den) static void compute_m_n(unsigned int m, unsigned int n, uint32_t *ret_m, uint32_t *ret_n) { + /* + * Reduce M/N as much as possible without loss in precision. Several DP + * dongles in particular seem to be fussy about too large *link* M/N + * values. The passed in values are more likely to have the least + * significant bits zero than M after rounding below, so do this first. + */ + while ((m & 1) == 0 && (n & 1) == 0) { + m >>= 1; + n >>= 1; + } + *ret_n = min_t(unsigned int, roundup_pow_of_two(n), DATA_LINK_N_MAX); *ret_m = div_u64((uint64_t) m * *ret_n, n); intel_reduce_m_n_ratio(ret_m, ret_n); @@ -8395,7 +8379,7 @@ skylake_get_initial_plane_config(struct intel_crtc *crtc, tiling = val & PLANE_CTL_TILED_MASK; switch (tiling) { case PLANE_CTL_TILED_LINEAR: - fb->modifier = DRM_FORMAT_MOD_NONE; + fb->modifier = DRM_FORMAT_MOD_LINEAR; break; case PLANE_CTL_TILED_X: plane_config->tiling = I915_TILING_X; @@ -8851,8 +8835,14 @@ static int haswell_crtc_compute_clock(struct intel_crtc *crtc, struct intel_crtc_state *crtc_state) { if (!intel_crtc_has_type(crtc_state, INTEL_OUTPUT_DSI)) { - if (!intel_ddi_pll_select(crtc, crtc_state)) + struct intel_encoder *encoder = + intel_ddi_get_crtc_new_encoder(crtc_state); + + if (!intel_get_shared_dpll(crtc, crtc_state, encoder)) { + DRM_DEBUG_DRIVER("failed to find PLL for pipe %c\n", + pipe_name(crtc->pipe)); return -EINVAL; + } } crtc->lowfreq_avail = false; @@ -9148,6 +9138,31 @@ out: return active; } +static u32 i845_cursor_ctl(const struct intel_crtc_state *crtc_state, + const struct intel_plane_state *plane_state) +{ + unsigned int width = plane_state->base.crtc_w; + unsigned int stride = roundup_pow_of_two(width) * 4; + + switch (stride) { + default: + WARN_ONCE(1, "Invalid cursor width/stride, width=%u, stride=%u\n", + width, stride); + stride = 256; + /* fallthrough */ + case 256: + case 512: + case 1024: + case 2048: + break; + } + + return CURSOR_ENABLE | + CURSOR_GAMMA_ENABLE | + CURSOR_FORMAT_ARGB | + CURSOR_STRIDE(stride); +} + static void i845_update_cursor(struct drm_crtc *crtc, u32 base, const struct intel_plane_state *plane_state) { @@ -9159,26 +9174,8 @@ static void i845_update_cursor(struct drm_crtc *crtc, u32 base, if (plane_state && plane_state->base.visible) { unsigned int width = plane_state->base.crtc_w; unsigned int height = plane_state->base.crtc_h; - unsigned int stride = roundup_pow_of_two(width) * 4; - - switch (stride) { - default: - WARN_ONCE(1, "Invalid cursor width/stride, width=%u, stride=%u\n", - width, stride); - stride = 256; - /* fallthrough */ - case 256: - case 512: - case 1024: - case 2048: - break; - } - - cntl |= CURSOR_ENABLE | - CURSOR_GAMMA_ENABLE | - CURSOR_FORMAT_ARGB | - CURSOR_STRIDE(stride); + cntl = plane_state->ctl; size = (height << 12) | width; } @@ -9211,6 +9208,43 @@ static void i845_update_cursor(struct drm_crtc *crtc, u32 base, } } +static u32 i9xx_cursor_ctl(const struct intel_crtc_state *crtc_state, + const struct intel_plane_state *plane_state) +{ + struct drm_i915_private *dev_priv = + to_i915(plane_state->base.plane->dev); + struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc); + enum pipe pipe = crtc->pipe; + u32 cntl; + + cntl = MCURSOR_GAMMA_ENABLE; + + if (HAS_DDI(dev_priv)) + cntl |= CURSOR_PIPE_CSC_ENABLE; + + cntl |= pipe << 28; /* Connect to correct pipe */ + + switch (plane_state->base.crtc_w) { + case 64: + cntl |= CURSOR_MODE_64_ARGB_AX; + break; + case 128: + cntl |= CURSOR_MODE_128_ARGB_AX; + break; + case 256: + cntl |= CURSOR_MODE_256_ARGB_AX; + break; + default: + MISSING_CASE(plane_state->base.crtc_w); + return 0; + } + + if (plane_state->base.rotation & DRM_ROTATE_180) + cntl |= CURSOR_ROTATE_180; + + return cntl; +} + static void i9xx_update_cursor(struct drm_crtc *crtc, u32 base, const struct intel_plane_state *plane_state) { @@ -9220,30 +9254,8 @@ static void i9xx_update_cursor(struct drm_crtc *crtc, u32 base, int pipe = intel_crtc->pipe; uint32_t cntl = 0; - if (plane_state && plane_state->base.visible) { - cntl = MCURSOR_GAMMA_ENABLE; - switch (plane_state->base.crtc_w) { - case 64: - cntl |= CURSOR_MODE_64_ARGB_AX; - break; - case 128: - cntl |= CURSOR_MODE_128_ARGB_AX; - break; - case 256: - cntl |= CURSOR_MODE_256_ARGB_AX; - break; - default: - MISSING_CASE(plane_state->base.crtc_w); - return; - } - cntl |= pipe << 28; /* Connect to correct pipe */ - - if (HAS_DDI(dev_priv)) - cntl |= CURSOR_PIPE_CSC_ENABLE; - - if (plane_state->base.rotation & DRM_ROTATE_180) - cntl |= CURSOR_ROTATE_180; - } + if (plane_state && plane_state->base.visible) + cntl = plane_state->ctl; if (intel_crtc->cursor_cntl != cntl) { I915_WRITE_FW(CURCNTR(pipe), cntl); @@ -10338,7 +10350,7 @@ static void skl_do_mmio_flip(struct intel_crtc *intel_crtc, ctl = I915_READ(PLANE_CTL(pipe, 0)); ctl &= ~PLANE_CTL_TILED_MASK; switch (fb->modifier) { - case DRM_FORMAT_MOD_NONE: + case DRM_FORMAT_MOD_LINEAR: break; case I915_FORMAT_MOD_X_TILED: ctl |= PLANE_CTL_TILED_X; @@ -11692,6 +11704,9 @@ intel_pipe_config_compare(struct drm_i915_private *dev_priv, if ((INTEL_GEN(dev_priv) < 8 && !IS_HASWELL(dev_priv)) || IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv)) PIPE_CONF_CHECK_I(limited_color_range); + + PIPE_CONF_CHECK_I(hdmi_scrambling); + PIPE_CONF_CHECK_I(hdmi_high_tmds_clock_ratio); PIPE_CONF_CHECK_I(has_infoframe); PIPE_CONF_CHECK_I(has_audio); @@ -13285,6 +13300,14 @@ intel_check_primary_plane(struct drm_plane *plane, ret = skl_check_plane_surface(state); if (ret) return ret; + + state->ctl = skl_plane_ctl(crtc_state, state); + } else { + ret = i9xx_check_plane_surface(state); + if (ret) + return ret; + + state->ctl = i9xx_plane_ctl(crtc_state, state); } return 0; @@ -13544,12 +13567,6 @@ intel_primary_plane_create(struct drm_i915_private *dev_priv, enum pipe pipe) primary->update_plane = skylake_update_primary_plane; primary->disable_plane = skylake_disable_primary_plane; - } else if (HAS_PCH_SPLIT(dev_priv)) { - intel_primary_formats = i965_primary_formats; - num_formats = ARRAY_SIZE(i965_primary_formats); - - primary->update_plane = ironlake_update_primary_plane; - primary->disable_plane = i9xx_disable_primary_plane; } else if (INTEL_GEN(dev_priv) >= 4) { intel_primary_formats = i965_primary_formats; num_formats = ARRAY_SIZE(i965_primary_formats); @@ -13621,6 +13638,7 @@ intel_check_cursor_plane(struct drm_plane *plane, struct intel_crtc_state *crtc_state, struct intel_plane_state *state) { + struct drm_i915_private *dev_priv = to_i915(plane->dev); struct drm_framebuffer *fb = state->base.fb; struct drm_i915_gem_object *obj = intel_fb_obj(fb); enum pipe pipe = to_intel_plane(plane)->pipe; @@ -13640,7 +13658,7 @@ intel_check_cursor_plane(struct drm_plane *plane, return 0; /* Check for which cursor types we support */ - if (!cursor_size_ok(to_i915(plane->dev), state->base.crtc_w, + if (!cursor_size_ok(dev_priv, state->base.crtc_w, state->base.crtc_h)) { DRM_DEBUG("Cursor dimension %dx%d not supported\n", state->base.crtc_w, state->base.crtc_h); @@ -13653,7 +13671,7 @@ intel_check_cursor_plane(struct drm_plane *plane, return -ENOMEM; } - if (fb->modifier != DRM_FORMAT_MOD_NONE) { + if (fb->modifier != DRM_FORMAT_MOD_LINEAR) { DRM_DEBUG_KMS("cursor cannot be tiled\n"); return -EINVAL; } @@ -13668,12 +13686,17 @@ intel_check_cursor_plane(struct drm_plane *plane, * display power well must be turned off and on again. * Refuse the put the cursor into that compromised position. */ - if (IS_CHERRYVIEW(to_i915(plane->dev)) && pipe == PIPE_C && + if (IS_CHERRYVIEW(dev_priv) && pipe == PIPE_C && state->base.visible && state->base.crtc_x < 0) { DRM_DEBUG_KMS("CHV cursor C not allowed to straddle the left screen edge\n"); return -EINVAL; } + if (IS_I845G(dev_priv) || IS_I865G(dev_priv)) + state->ctl = i845_cursor_ctl(crtc_state, state); + else + state->ctl = i9xx_cursor_ctl(crtc_state, state); + return 0; } @@ -14309,7 +14332,7 @@ static int intel_framebuffer_init(struct intel_framebuffer *intel_fb, mode_cmd->modifier[0]); goto err; } - case DRM_FORMAT_MOD_NONE: + case DRM_FORMAT_MOD_LINEAR: case I915_FORMAT_MOD_X_TILED: break; default: @@ -14332,7 +14355,7 @@ static int intel_framebuffer_init(struct intel_framebuffer *intel_fb, mode_cmd->pixel_format); if (mode_cmd->pitches[0] > pitch_limit) { DRM_DEBUG_KMS("%s pitch (%u) must be at most %d\n", - mode_cmd->modifier[0] != DRM_FORMAT_MOD_NONE ? + mode_cmd->modifier[0] != DRM_FORMAT_MOD_LINEAR ? "tiled" : "linear", mode_cmd->pitches[0], pitch_limit); goto err; diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h index fb7afcf9e8be..aaee3949a422 100644 --- a/drivers/gpu/drm/i915/intel_drv.h +++ b/drivers/gpu/drm/i915/intel_drv.h @@ -398,6 +398,9 @@ struct intel_plane_state { int x, y; } aux; + /* plane control register */ + u32 ctl; + /* * scaler_id * = -1 : not using a scaler @@ -729,6 +732,12 @@ struct intel_crtc_state { /* bitmask of visible planes (enum plane_id) */ u8 active_planes; + + /* HDMI scrambling status */ + bool hdmi_scrambling; + + /* HDMI High TMDS char rate ratio */ + bool hdmi_high_tmds_clock_ratio; }; struct intel_crtc { @@ -1220,12 +1229,9 @@ void intel_crt_init(struct drm_i915_private *dev_priv); void intel_crt_reset(struct drm_encoder *encoder); /* intel_ddi.c */ -void intel_ddi_clk_select(struct intel_encoder *encoder, - struct intel_shared_dpll *pll); void intel_ddi_fdi_post_disable(struct intel_encoder *intel_encoder, struct intel_crtc_state *old_crtc_state, struct drm_connector_state *old_conn_state); -void intel_prepare_dp_ddi_buffers(struct intel_encoder *encoder); void hsw_fdi_link_train(struct intel_crtc *crtc, const struct intel_crtc_state *crtc_state); void intel_ddi_init(struct drm_i915_private *dev_priv, enum port port); @@ -1236,8 +1242,8 @@ void intel_ddi_disable_transcoder_func(struct drm_i915_private *dev_priv, enum transcoder cpu_transcoder); void intel_ddi_enable_pipe_clock(const struct intel_crtc_state *crtc_state); void intel_ddi_disable_pipe_clock(const struct intel_crtc_state *crtc_state); -bool intel_ddi_pll_select(struct intel_crtc *crtc, - struct intel_crtc_state *crtc_state); +struct intel_encoder * +intel_ddi_get_crtc_new_encoder(struct intel_crtc_state *crtc_state); void intel_ddi_set_pipe_settings(const struct intel_crtc_state *crtc_state); void intel_ddi_prepare_link_retrain(struct intel_dp *intel_dp); bool intel_ddi_connector_get_hw_state(struct intel_connector *intel_connector); @@ -1246,7 +1252,6 @@ bool intel_ddi_is_audio_enabled(struct drm_i915_private *dev_priv, void intel_ddi_get_config(struct intel_encoder *encoder, struct intel_crtc_state *pipe_config); -void intel_ddi_init_dp_buf_reg(struct intel_encoder *encoder); void intel_ddi_clock_get(struct intel_encoder *encoder, struct intel_crtc_state *pipe_config); void intel_ddi_set_vc_payload_alloc(const struct intel_crtc_state *crtc_state, @@ -1445,12 +1450,12 @@ static inline u32 intel_plane_ggtt_offset(const struct intel_plane_state *state) return i915_ggtt_offset(state->vma); } -u32 skl_plane_ctl_format(uint32_t pixel_format); -u32 skl_plane_ctl_tiling(uint64_t fb_modifier); -u32 skl_plane_ctl_rotation(unsigned int rotation); +u32 skl_plane_ctl(const struct intel_crtc_state *crtc_state, + const struct intel_plane_state *plane_state); u32 skl_plane_stride(const struct drm_framebuffer *fb, int plane, unsigned int rotation); int skl_check_plane_surface(struct intel_plane_state *plane_state); +int i9xx_check_plane_surface(struct intel_plane_state *plane_state); /* intel_csr.c */ void intel_csr_ucode_init(struct drm_i915_private *); @@ -1620,6 +1625,10 @@ struct intel_hdmi *enc_to_intel_hdmi(struct drm_encoder *encoder); bool intel_hdmi_compute_config(struct intel_encoder *encoder, struct intel_crtc_state *pipe_config, struct drm_connector_state *conn_state); +void intel_hdmi_handle_sink_scrambling(struct intel_encoder *intel_encoder, + struct drm_connector *connector, + bool high_tmds_clock_ratio, + bool scrambling); void intel_dp_dual_mode_set_tmds_output(struct intel_hdmi *hdmi, bool enable); diff --git a/drivers/gpu/drm/i915/intel_engine_cs.c b/drivers/gpu/drm/i915/intel_engine_cs.c index 4200faa520c7..854e8e0c836b 100644 --- a/drivers/gpu/drm/i915/intel_engine_cs.c +++ b/drivers/gpu/drm/i915/intel_engine_cs.c @@ -36,45 +36,45 @@ static const struct engine_info { int (*init_execlists)(struct intel_engine_cs *engine); } intel_engines[] = { [RCS] = { - .name = "render ring", - .exec_id = I915_EXEC_RENDER, + .name = "rcs", .hw_id = RCS_HW, + .exec_id = I915_EXEC_RENDER, .mmio_base = RENDER_RING_BASE, .irq_shift = GEN8_RCS_IRQ_SHIFT, .init_execlists = logical_render_ring_init, .init_legacy = intel_init_render_ring_buffer, }, [BCS] = { - .name = "blitter ring", - .exec_id = I915_EXEC_BLT, + .name = "bcs", .hw_id = BCS_HW, + .exec_id = I915_EXEC_BLT, .mmio_base = BLT_RING_BASE, .irq_shift = GEN8_BCS_IRQ_SHIFT, .init_execlists = logical_xcs_ring_init, .init_legacy = intel_init_blt_ring_buffer, }, [VCS] = { - .name = "bsd ring", - .exec_id = I915_EXEC_BSD, + .name = "vcs", .hw_id = VCS_HW, + .exec_id = I915_EXEC_BSD, .mmio_base = GEN6_BSD_RING_BASE, .irq_shift = GEN8_VCS1_IRQ_SHIFT, .init_execlists = logical_xcs_ring_init, .init_legacy = intel_init_bsd_ring_buffer, }, [VCS2] = { - .name = "bsd2 ring", - .exec_id = I915_EXEC_BSD, + .name = "vcs2", .hw_id = VCS2_HW, + .exec_id = I915_EXEC_BSD, .mmio_base = GEN8_BSD2_RING_BASE, .irq_shift = GEN8_VCS2_IRQ_SHIFT, .init_execlists = logical_xcs_ring_init, .init_legacy = intel_init_bsd2_ring_buffer, }, [VECS] = { - .name = "video enhancement ring", - .exec_id = I915_EXEC_VEBOX, + .name = "vecs", .hw_id = VECS_HW, + .exec_id = I915_EXEC_VEBOX, .mmio_base = VEBOX_RING_BASE, .irq_shift = GEN8_VECS_IRQ_SHIFT, .init_execlists = logical_xcs_ring_init, @@ -242,12 +242,12 @@ void intel_engine_init_global_seqno(struct intel_engine_cs *engine, u32 seqno) void *semaphores; /* Semaphores are in noncoherent memory, flush to be safe */ - semaphores = kmap(page); + semaphores = kmap_atomic(page); memset(semaphores + GEN8_SEMAPHORE_OFFSET(engine->id, 0), 0, I915_NUM_ENGINES * gen8_semaphore_seqno_size); drm_clflush_virt_range(semaphores + GEN8_SEMAPHORE_OFFSET(engine->id, 0), I915_NUM_ENGINES * gen8_semaphore_seqno_size); - kunmap(page); + kunmap_atomic(semaphores); } intel_write_status_page(engine, I915_GEM_HWS_INDEX, seqno); @@ -1111,6 +1111,15 @@ bool intel_engines_are_idle(struct drm_i915_private *dev_priv) struct intel_engine_cs *engine; enum intel_engine_id id; + if (READ_ONCE(dev_priv->gt.active_requests)) + return false; + + /* If the driver is wedged, HW state may be very inconsistent and + * report that it is still busy, even though we have stopped using it. + */ + if (i915_terminally_wedged(&dev_priv->gpu_error)) + return true; + for_each_engine(engine, dev_priv, id) { if (!intel_engine_is_idle(engine)) return false; diff --git a/drivers/gpu/drm/i915/intel_guc_fwif.h b/drivers/gpu/drm/i915/intel_guc_fwif.h index 25691f0e4c50..cb36cbf3818f 100644 --- a/drivers/gpu/drm/i915/intel_guc_fwif.h +++ b/drivers/gpu/drm/i915/intel_guc_fwif.h @@ -26,14 +26,14 @@ #define GFXCORE_FAMILY_GEN9 12 #define GFXCORE_FAMILY_UNKNOWN 0x7fffffff -#define GUC_CTX_PRIORITY_KMD_HIGH 0 -#define GUC_CTX_PRIORITY_HIGH 1 -#define GUC_CTX_PRIORITY_KMD_NORMAL 2 -#define GUC_CTX_PRIORITY_NORMAL 3 -#define GUC_CTX_PRIORITY_NUM 4 +#define GUC_CLIENT_PRIORITY_KMD_HIGH 0 +#define GUC_CLIENT_PRIORITY_HIGH 1 +#define GUC_CLIENT_PRIORITY_KMD_NORMAL 2 +#define GUC_CLIENT_PRIORITY_NORMAL 3 +#define GUC_CLIENT_PRIORITY_NUM 4 -#define GUC_MAX_GPU_CONTEXTS 1024 -#define GUC_INVALID_CTX_ID GUC_MAX_GPU_CONTEXTS +#define GUC_MAX_STAGE_DESCRIPTORS 1024 +#define GUC_INVALID_STAGE_ID GUC_MAX_STAGE_DESCRIPTORS #define GUC_RENDER_ENGINE 0 #define GUC_VIDEO_ENGINE 1 @@ -68,14 +68,14 @@ #define GUC_DOORBELL_ENABLED 1 #define GUC_DOORBELL_DISABLED 0 -#define GUC_CTX_DESC_ATTR_ACTIVE (1 << 0) -#define GUC_CTX_DESC_ATTR_PENDING_DB (1 << 1) -#define GUC_CTX_DESC_ATTR_KERNEL (1 << 2) -#define GUC_CTX_DESC_ATTR_PREEMPT (1 << 3) -#define GUC_CTX_DESC_ATTR_RESET (1 << 4) -#define GUC_CTX_DESC_ATTR_WQLOCKED (1 << 5) -#define GUC_CTX_DESC_ATTR_PCH (1 << 6) -#define GUC_CTX_DESC_ATTR_TERMINATED (1 << 7) +#define GUC_STAGE_DESC_ATTR_ACTIVE BIT(0) +#define GUC_STAGE_DESC_ATTR_PENDING_DB BIT(1) +#define GUC_STAGE_DESC_ATTR_KERNEL BIT(2) +#define GUC_STAGE_DESC_ATTR_PREEMPT BIT(3) +#define GUC_STAGE_DESC_ATTR_RESET BIT(4) +#define GUC_STAGE_DESC_ATTR_WQLOCKED BIT(5) +#define GUC_STAGE_DESC_ATTR_PCH BIT(6) +#define GUC_STAGE_DESC_ATTR_TERMINATED BIT(7) /* The guc control data is 10 DWORDs */ #define GUC_CTL_CTXINFO 0 @@ -241,8 +241,8 @@ union guc_doorbell_qw { u64 value_qw; } __packed; -#define GUC_MAX_DOORBELLS 256 -#define GUC_INVALID_DOORBELL_ID (GUC_MAX_DOORBELLS) +#define GUC_NUM_DOORBELLS 256 +#define GUC_DOORBELL_INVALID (GUC_NUM_DOORBELLS) #define GUC_DB_SIZE (PAGE_SIZE) #define GUC_WQ_SIZE (PAGE_SIZE * 2) @@ -251,12 +251,12 @@ union guc_doorbell_qw { struct guc_wq_item { u32 header; u32 context_desc; - u32 ring_tail; + u32 submit_element_info; u32 fence_id; } __packed; struct guc_process_desc { - u32 context_id; + u32 stage_id; u64 db_base_addr; u32 head; u32 tail; @@ -278,7 +278,7 @@ struct guc_execlist_context { u32 context_desc; u32 context_id; u32 ring_status; - u32 ring_lcra; + u32 ring_lrca; u32 ring_begin; u32 ring_end; u32 ring_next_free_location; @@ -289,10 +289,18 @@ struct guc_execlist_context { u16 engine_submit_queue_count; } __packed; -/*Context descriptor for communicating between uKernel and Driver*/ -struct guc_context_desc { +/* + * This structure describes a stage set arranged for a particular communication + * between uKernel (GuC) and Driver (KMD). Technically, this is known as a + * "GuC Context descriptor" in the specs, but we use the term "stage descriptor" + * to avoid confusion with all the other things already named "context" in the + * driver. A static pool of these descriptors are stored inside a GEM object + * (stage_desc_pool) which is held for the entire lifetime of our interaction + * with the GuC, being allocated before the GuC is loaded with its firmware. + */ +struct guc_stage_desc { u32 sched_common_area; - u32 context_id; + u32 stage_id; u32 pas_id; u8 engines_used; u64 db_trigger_cpu; @@ -359,7 +367,7 @@ struct guc_policy { } __packed; struct guc_policies { - struct guc_policy policy[GUC_CTX_PRIORITY_NUM][GUC_MAX_ENGINES_NUM]; + struct guc_policy policy[GUC_CLIENT_PRIORITY_NUM][GUC_MAX_ENGINES_NUM]; /* In micro seconds. How much time to allow before DPC processing is * called back via interrupt (to prevent DPC queue drain starving). @@ -401,16 +409,17 @@ struct guc_mmio_regset { u32 number_of_registers; } __packed; +/* MMIO registers that are set as non privileged */ +struct mmio_white_list { + u32 mmio_start; + u32 offsets[GUC_MMIO_WHITE_LIST_MAX]; + u32 count; +} __packed; + struct guc_mmio_reg_state { struct guc_mmio_regset global_reg; struct guc_mmio_regset engine_reg[GUC_MAX_ENGINES_NUM]; - - /* MMIO registers that are set as non privileged */ - struct __packed { - u32 mmio_start; - u32 offsets[GUC_MMIO_WHITE_LIST_MAX]; - u32 count; - } mmio_white_list[GUC_MAX_ENGINES_NUM]; + struct mmio_white_list white_list[GUC_MAX_ENGINES_NUM]; } __packed; /* GuC Additional Data Struct */ diff --git a/drivers/gpu/drm/i915/intel_guc_loader.c b/drivers/gpu/drm/i915/intel_guc_loader.c index 2f270d02894c..8a1a023e48b2 100644 --- a/drivers/gpu/drm/i915/intel_guc_loader.c +++ b/drivers/gpu/drm/i915/intel_guc_loader.c @@ -73,22 +73,6 @@ MODULE_FIRMWARE(I915_BXT_GUC_UCODE); #define I915_KBL_GUC_UCODE GUC_FW_PATH(kbl, KBL_FW_MAJOR, KBL_FW_MINOR) MODULE_FIRMWARE(I915_KBL_GUC_UCODE); -/* User-friendly representation of an enum */ -const char *intel_uc_fw_status_repr(enum intel_uc_fw_status status) -{ - switch (status) { - case INTEL_UC_FIRMWARE_FAIL: - return "FAIL"; - case INTEL_UC_FIRMWARE_NONE: - return "NONE"; - case INTEL_UC_FIRMWARE_PENDING: - return "PENDING"; - case INTEL_UC_FIRMWARE_SUCCESS: - return "SUCCESS"; - default: - return "UNKNOWN!"; - } -}; static u32 get_gttype(struct drm_i915_private *dev_priv) { @@ -148,16 +132,14 @@ static void guc_params_init(struct drm_i915_private *dev_priv) } else params[GUC_CTL_DEBUG] = GUC_LOG_DISABLED; - if (guc->ads_vma) { + /* If GuC submission is enabled, set up additional parameters here */ + if (i915.enable_guc_submission) { u32 ads = guc_ggtt_offset(guc->ads_vma) >> PAGE_SHIFT; + u32 pgs = guc_ggtt_offset(dev_priv->guc.stage_desc_pool); + u32 ctx_in_16 = GUC_MAX_STAGE_DESCRIPTORS / 16; + params[GUC_CTL_DEBUG] |= ads << GUC_ADS_ADDR_SHIFT; params[GUC_CTL_DEBUG] |= GUC_ADS_ENABLED; - } - - /* If GuC submission is enabled, set up additional parameters here */ - if (i915.enable_guc_submission) { - u32 pgs = guc_ggtt_offset(dev_priv->guc.ctx_pool_vma); - u32 ctx_in_16 = GUC_MAX_GPU_CONTEXTS / 16; pgs >>= PAGE_SHIFT; params[GUC_CTL_CTXINFO] = (pgs << GUC_CTL_BASE_ADDR_SHIFT) | @@ -430,24 +412,3 @@ int intel_guc_select_fw(struct intel_guc *guc) return 0; } - -/** - * intel_guc_fini() - clean up all allocated resources - * @dev_priv: i915 device private - */ -void intel_guc_fini(struct drm_i915_private *dev_priv) -{ - struct intel_uc_fw *guc_fw = &dev_priv->guc.fw; - struct drm_i915_gem_object *obj; - - mutex_lock(&dev_priv->drm.struct_mutex); - i915_guc_submission_disable(dev_priv); - i915_guc_submission_fini(dev_priv); - mutex_unlock(&dev_priv->drm.struct_mutex); - - obj = fetch_and_zero(&guc_fw->obj); - if (obj) - i915_gem_object_put(obj); - - guc_fw->fetch_status = INTEL_UC_FIRMWARE_NONE; -} diff --git a/drivers/gpu/drm/i915/intel_guc_log.c b/drivers/gpu/drm/i915/intel_guc_log.c index 5c0f9a49da0e..6fb63a3c65b0 100644 --- a/drivers/gpu/drm/i915/intel_guc_log.c +++ b/drivers/gpu/drm/i915/intel_guc_log.c @@ -66,7 +66,6 @@ static int guc_log_control(struct intel_guc *guc, u32 control_val) return intel_guc_send(guc, action, ARRAY_SIZE(action)); } - /* * Sub buffer switch callback. Called whenever relay has to switch to a new * sub buffer, relay stays on the same sub buffer if 0 is returned. @@ -139,45 +138,15 @@ static struct rchan_callbacks relay_callbacks = { .remove_buf_file = remove_buf_file_callback, }; -static void guc_log_remove_relay_file(struct intel_guc *guc) -{ - relay_close(guc->log.relay_chan); -} - -static int guc_log_create_relay_channel(struct intel_guc *guc) -{ - struct drm_i915_private *dev_priv = guc_to_i915(guc); - struct rchan *guc_log_relay_chan; - size_t n_subbufs, subbuf_size; - - /* Keep the size of sub buffers same as shared log buffer */ - subbuf_size = guc->log.vma->obj->base.size; - - /* Store up to 8 snapshots, which is large enough to buffer sufficient - * boot time logs and provides enough leeway to User, in terms of - * latency, for consuming the logs from relay. Also doesn't take - * up too much memory. - */ - n_subbufs = 8; - - guc_log_relay_chan = relay_open(NULL, NULL, subbuf_size, - n_subbufs, &relay_callbacks, dev_priv); - if (!guc_log_relay_chan) { - DRM_ERROR("Couldn't create relay chan for GuC logging\n"); - return -ENOMEM; - } - - GEM_BUG_ON(guc_log_relay_chan->subbuf_size < subbuf_size); - guc->log.relay_chan = guc_log_relay_chan; - return 0; -} - -static int guc_log_create_relay_file(struct intel_guc *guc) +static int guc_log_relay_file_create(struct intel_guc *guc) { struct drm_i915_private *dev_priv = guc_to_i915(guc); struct dentry *log_dir; int ret; + if (i915.guc_log_level < 0) + return 0; + /* For now create the log file in /sys/kernel/debug/dri/0 dir */ log_dir = dev_priv->drm.primary->debugfs_root; @@ -197,8 +166,8 @@ static int guc_log_create_relay_file(struct intel_guc *guc) return -ENODEV; } - ret = relay_late_setup_files(guc->log.relay_chan, "guc_log", log_dir); - if (ret) { + ret = relay_late_setup_files(guc->log.runtime.relay_chan, "guc_log", log_dir); + if (ret < 0 && ret != -EEXIST) { DRM_ERROR("Couldn't associate relay chan with file %d\n", ret); return ret; } @@ -214,15 +183,15 @@ static void guc_move_to_next_buf(struct intel_guc *guc) smp_wmb(); /* All data has been written, so now move the offset of sub buffer. */ - relay_reserve(guc->log.relay_chan, guc->log.vma->obj->base.size); + relay_reserve(guc->log.runtime.relay_chan, guc->log.vma->obj->base.size); /* Switch to the next sub buffer */ - relay_flush(guc->log.relay_chan); + relay_flush(guc->log.runtime.relay_chan); } static void *guc_get_write_buffer(struct intel_guc *guc) { - if (!guc->log.relay_chan) + if (!guc->log.runtime.relay_chan) return NULL; /* Just get the base address of a new sub buffer and copy data into it @@ -233,7 +202,7 @@ static void *guc_get_write_buffer(struct intel_guc *guc) * done without using relay_reserve() along with relay_write(). So its * better to use relay_reserve() alone. */ - return relay_reserve(guc->log.relay_chan, 0); + return relay_reserve(guc->log.runtime.relay_chan, 0); } static bool guc_check_log_buf_overflow(struct intel_guc *guc, @@ -284,11 +253,11 @@ static void guc_read_update_log_buffer(struct intel_guc *guc) void *src_data, *dst_data; bool new_overflow; - if (WARN_ON(!guc->log.buf_addr)) + if (WARN_ON(!guc->log.runtime.buf_addr)) return; /* Get the pointer to shared GuC log buffer */ - log_buf_state = src_data = guc->log.buf_addr; + log_buf_state = src_data = guc->log.runtime.buf_addr; /* Get the pointer to local buffer to store the logs */ log_buf_snapshot_state = dst_data = guc_get_write_buffer(guc); @@ -371,153 +340,113 @@ static void guc_read_update_log_buffer(struct intel_guc *guc) } } -static void guc_log_cleanup(struct intel_guc *guc) -{ - struct drm_i915_private *dev_priv = guc_to_i915(guc); - - lockdep_assert_held(&dev_priv->drm.struct_mutex); - - /* First disable the flush interrupt */ - gen9_disable_guc_interrupts(dev_priv); - - if (guc->log.flush_wq) - destroy_workqueue(guc->log.flush_wq); - - guc->log.flush_wq = NULL; - - if (guc->log.relay_chan) - guc_log_remove_relay_file(guc); - - guc->log.relay_chan = NULL; - - if (guc->log.buf_addr) - i915_gem_object_unpin_map(guc->log.vma->obj); - - guc->log.buf_addr = NULL; -} - static void capture_logs_work(struct work_struct *work) { struct intel_guc *guc = - container_of(work, struct intel_guc, log.flush_work); + container_of(work, struct intel_guc, log.runtime.flush_work); guc_log_capture_logs(guc); } -static int guc_log_create_extras(struct intel_guc *guc) +static bool guc_log_has_runtime(struct intel_guc *guc) +{ + return guc->log.runtime.buf_addr != NULL; +} + +static int guc_log_runtime_create(struct intel_guc *guc) { struct drm_i915_private *dev_priv = guc_to_i915(guc); void *vaddr; - int ret; + struct rchan *guc_log_relay_chan; + size_t n_subbufs, subbuf_size; + int ret = 0; lockdep_assert_held(&dev_priv->drm.struct_mutex); - /* Nothing to do */ - if (i915.guc_log_level < 0) - return 0; - - if (!guc->log.buf_addr) { - /* Create a WC (Uncached for read) vmalloc mapping of log - * buffer pages, so that we can directly get the data - * (up-to-date) from memory. - */ - vaddr = i915_gem_object_pin_map(guc->log.vma->obj, I915_MAP_WC); - if (IS_ERR(vaddr)) { - ret = PTR_ERR(vaddr); - DRM_ERROR("Couldn't map log buffer pages %d\n", ret); - return ret; - } + GEM_BUG_ON(guc_log_has_runtime(guc)); - guc->log.buf_addr = vaddr; + /* Create a WC (Uncached for read) vmalloc mapping of log + * buffer pages, so that we can directly get the data + * (up-to-date) from memory. + */ + vaddr = i915_gem_object_pin_map(guc->log.vma->obj, I915_MAP_WC); + if (IS_ERR(vaddr)) { + DRM_ERROR("Couldn't map log buffer pages %d\n", ret); + return PTR_ERR(vaddr); } - if (!guc->log.relay_chan) { - /* Create a relay channel, so that we have buffers for storing - * the GuC firmware logs, the channel will be linked with a file - * later on when debugfs is registered. - */ - ret = guc_log_create_relay_channel(guc); - if (ret) - return ret; - } + guc->log.runtime.buf_addr = vaddr; - if (!guc->log.flush_wq) { - INIT_WORK(&guc->log.flush_work, capture_logs_work); - - /* - * GuC log buffer flush work item has to do register access to - * send the ack to GuC and this work item, if not synced before - * suspend, can potentially get executed after the GFX device is - * suspended. - * By marking the WQ as freezable, we don't have to bother about - * flushing of this work item from the suspend hooks, the pending - * work item if any will be either executed before the suspend - * or scheduled later on resume. This way the handling of work - * item can be kept same between system suspend & rpm suspend. - */ - guc->log.flush_wq = alloc_ordered_workqueue("i915-guc_log", - WQ_HIGHPRI | WQ_FREEZABLE); - if (guc->log.flush_wq == NULL) { - DRM_ERROR("Couldn't allocate the wq for GuC logging\n"); - return -ENOMEM; - } - } - - return 0; -} - -void intel_guc_log_create(struct intel_guc *guc) -{ - struct i915_vma *vma; - unsigned long offset; - uint32_t size, flags; + /* Keep the size of sub buffers same as shared log buffer */ + subbuf_size = guc->log.vma->obj->base.size; - if (i915.guc_log_level > GUC_LOG_VERBOSITY_MAX) - i915.guc_log_level = GUC_LOG_VERBOSITY_MAX; + /* Store up to 8 snapshots, which is large enough to buffer sufficient + * boot time logs and provides enough leeway to User, in terms of + * latency, for consuming the logs from relay. Also doesn't take + * up too much memory. + */ + n_subbufs = 8; - /* The first page is to save log buffer state. Allocate one - * extra page for others in case for overlap */ - size = (1 + GUC_LOG_DPC_PAGES + 1 + - GUC_LOG_ISR_PAGES + 1 + - GUC_LOG_CRASH_PAGES + 1) << PAGE_SHIFT; + /* Create a relay channel, so that we have buffers for storing + * the GuC firmware logs, the channel will be linked with a file + * later on when debugfs is registered. + */ + guc_log_relay_chan = relay_open(NULL, NULL, subbuf_size, + n_subbufs, &relay_callbacks, dev_priv); + if (!guc_log_relay_chan) { + DRM_ERROR("Couldn't create relay chan for GuC logging\n"); - vma = guc->log.vma; - if (!vma) { - /* We require SSE 4.1 for fast reads from the GuC log buffer and - * it should be present on the chipsets supporting GuC based - * submisssions. - */ - if (WARN_ON(!i915_has_memcpy_from_wc())) { - /* logging will not be enabled */ - i915.guc_log_level = -1; - return; - } + ret = -ENOMEM; + goto err_vaddr; + } - vma = intel_guc_allocate_vma(guc, size); - if (IS_ERR(vma)) { - /* logging will be off */ - i915.guc_log_level = -1; - return; - } + GEM_BUG_ON(guc_log_relay_chan->subbuf_size < subbuf_size); + guc->log.runtime.relay_chan = guc_log_relay_chan; + + INIT_WORK(&guc->log.runtime.flush_work, capture_logs_work); + + /* + * GuC log buffer flush work item has to do register access to + * send the ack to GuC and this work item, if not synced before + * suspend, can potentially get executed after the GFX device is + * suspended. + * By marking the WQ as freezable, we don't have to bother about + * flushing of this work item from the suspend hooks, the pending + * work item if any will be either executed before the suspend + * or scheduled later on resume. This way the handling of work + * item can be kept same between system suspend & rpm suspend. + */ + guc->log.runtime.flush_wq = alloc_ordered_workqueue("i915-guc_log", + WQ_HIGHPRI | WQ_FREEZABLE); + if (!guc->log.runtime.flush_wq) { + DRM_ERROR("Couldn't allocate the wq for GuC logging\n"); + ret = -ENOMEM; + goto err_relaychan; + } - guc->log.vma = vma; + return 0; - if (guc_log_create_extras(guc)) { - guc_log_cleanup(guc); - i915_vma_unpin_and_release(&guc->log.vma); - i915.guc_log_level = -1; - return; - } - } +err_relaychan: + relay_close(guc->log.runtime.relay_chan); +err_vaddr: + i915_gem_object_unpin_map(guc->log.vma->obj); + guc->log.runtime.buf_addr = NULL; + return ret; +} - /* each allocated unit is a page */ - flags = GUC_LOG_VALID | GUC_LOG_NOTIFY_ON_HALF_FULL | - (GUC_LOG_DPC_PAGES << GUC_LOG_DPC_SHIFT) | - (GUC_LOG_ISR_PAGES << GUC_LOG_ISR_SHIFT) | - (GUC_LOG_CRASH_PAGES << GUC_LOG_CRASH_SHIFT); +static void guc_log_runtime_destroy(struct intel_guc *guc) +{ + /* + * It's possible that the runtime stuff was never allocated because + * guc_log_level was < 0 at the time + **/ + if (!guc_log_has_runtime(guc)) + return; - offset = guc_ggtt_offset(vma) >> PAGE_SHIFT; /* in pages */ - guc->log.flags = (offset << GUC_LOG_BUF_ADDR_SHIFT) | flags; + destroy_workqueue(guc->log.runtime.flush_wq); + relay_close(guc->log.runtime.relay_chan); + i915_gem_object_unpin_map(guc->log.vma->obj); + guc->log.runtime.buf_addr = NULL; } static int guc_log_late_setup(struct intel_guc *guc) @@ -527,24 +456,25 @@ static int guc_log_late_setup(struct intel_guc *guc) lockdep_assert_held(&dev_priv->drm.struct_mutex); - if (i915.guc_log_level < 0) - return -EINVAL; - - /* If log_level was set as -1 at boot time, then setup needed to - * handle log buffer flush interrupts would not have been done yet, - * so do that now. - */ - ret = guc_log_create_extras(guc); - if (ret) - goto err; + if (!guc_log_has_runtime(guc)) { + /* If log_level was set as -1 at boot time, then setup needed to + * handle log buffer flush interrupts would not have been done yet, + * so do that now. + */ + ret = guc_log_runtime_create(guc); + if (ret) + goto err; + } - ret = guc_log_create_relay_file(guc); + ret = guc_log_relay_file_create(guc); if (ret) - goto err; + goto err_runtime; return 0; + +err_runtime: + guc_log_runtime_destroy(guc); err: - guc_log_cleanup(guc); /* logging will remain off */ i915.guc_log_level = -1; return ret; @@ -577,7 +507,7 @@ static void guc_flush_logs(struct intel_guc *guc) /* Before initiating the forceful flush, wait for any pending/ongoing * flush to complete otherwise forceful flush may not actually happen. */ - flush_work(&guc->log.flush_work); + flush_work(&guc->log.runtime.flush_work); /* Ask GuC to update the log buffer state */ guc_log_flush(guc); @@ -586,6 +516,72 @@ static void guc_flush_logs(struct intel_guc *guc) guc_log_capture_logs(guc); } +int intel_guc_log_create(struct intel_guc *guc) +{ + struct i915_vma *vma; + unsigned long offset; + uint32_t size, flags; + int ret; + + GEM_BUG_ON(guc->log.vma); + + if (i915.guc_log_level > GUC_LOG_VERBOSITY_MAX) + i915.guc_log_level = GUC_LOG_VERBOSITY_MAX; + + /* The first page is to save log buffer state. Allocate one + * extra page for others in case for overlap */ + size = (1 + GUC_LOG_DPC_PAGES + 1 + + GUC_LOG_ISR_PAGES + 1 + + GUC_LOG_CRASH_PAGES + 1) << PAGE_SHIFT; + + /* We require SSE 4.1 for fast reads from the GuC log buffer and + * it should be present on the chipsets supporting GuC based + * submisssions. + */ + if (WARN_ON(!i915_has_memcpy_from_wc())) { + ret = -EINVAL; + goto err; + } + + vma = intel_guc_allocate_vma(guc, size); + if (IS_ERR(vma)) { + ret = PTR_ERR(vma); + goto err; + } + + guc->log.vma = vma; + + if (i915.guc_log_level >= 0) { + ret = guc_log_runtime_create(guc); + if (ret < 0) + goto err_vma; + } + + /* each allocated unit is a page */ + flags = GUC_LOG_VALID | GUC_LOG_NOTIFY_ON_HALF_FULL | + (GUC_LOG_DPC_PAGES << GUC_LOG_DPC_SHIFT) | + (GUC_LOG_ISR_PAGES << GUC_LOG_ISR_SHIFT) | + (GUC_LOG_CRASH_PAGES << GUC_LOG_CRASH_SHIFT); + + offset = guc_ggtt_offset(vma) >> PAGE_SHIFT; /* in pages */ + guc->log.flags = (offset << GUC_LOG_BUF_ADDR_SHIFT) | flags; + + return 0; + +err_vma: + i915_vma_unpin_and_release(&guc->log.vma); +err: + /* logging will be off */ + i915.guc_log_level = -1; + return ret; +} + +void intel_guc_log_destroy(struct intel_guc *guc) +{ + guc_log_runtime_destroy(guc); + i915_vma_unpin_and_release(&guc->log.vma); +} + int i915_guc_log_control(struct drm_i915_private *dev_priv, u64 control_val) { struct intel_guc *guc = &dev_priv->guc; @@ -609,17 +605,22 @@ int i915_guc_log_control(struct drm_i915_private *dev_priv, u64 control_val) return ret; } - i915.guc_log_level = log_param.verbosity; + if (log_param.logging_enabled) { + i915.guc_log_level = log_param.verbosity; - /* If log_level was set as -1 at boot time, then the relay channel file - * wouldn't have been created by now and interrupts also would not have - * been enabled. - */ - if (!dev_priv->guc.log.relay_chan) { + /* If log_level was set as -1 at boot time, then the relay channel file + * wouldn't have been created by now and interrupts also would not have + * been enabled. Try again now, just in case. + */ ret = guc_log_late_setup(guc); - if (!ret) - gen9_enable_guc_interrupts(dev_priv); - } else if (!log_param.logging_enabled) { + if (ret < 0) { + DRM_DEBUG_DRIVER("GuC log late setup failed %d\n", ret); + return ret; + } + + /* GuC logging is currently the only user of Guc2Host interrupts */ + gen9_enable_guc_interrupts(dev_priv); + } else { /* Once logging is disabled, GuC won't generate logs & send an * interrupt. But there could be some data in the log buffer * which is yet to be captured. So request GuC to update the log @@ -629,9 +630,6 @@ int i915_guc_log_control(struct drm_i915_private *dev_priv, u64 control_val) /* As logging is disabled, update log level to reflect that */ i915.guc_log_level = -1; - } else { - /* In case interrupts were disabled, enable them now */ - gen9_enable_guc_interrupts(dev_priv); } return ret; @@ -639,7 +637,7 @@ int i915_guc_log_control(struct drm_i915_private *dev_priv, u64 control_val) void i915_guc_log_register(struct drm_i915_private *dev_priv) { - if (!i915.enable_guc_submission) + if (!i915.enable_guc_submission || i915.guc_log_level < 0) return; mutex_lock(&dev_priv->drm.struct_mutex); @@ -653,6 +651,8 @@ void i915_guc_log_unregister(struct drm_i915_private *dev_priv) return; mutex_lock(&dev_priv->drm.struct_mutex); - guc_log_cleanup(&dev_priv->guc); + /* GuC logging is currently the only user of Guc2Host interrupts */ + gen9_disable_guc_interrupts(dev_priv); + guc_log_runtime_destroy(&dev_priv->guc); mutex_unlock(&dev_priv->drm.struct_mutex); } diff --git a/drivers/gpu/drm/i915/intel_gvt.c b/drivers/gpu/drm/i915/intel_gvt.c index 8c04eca84351..e1ab6432a914 100644 --- a/drivers/gpu/drm/i915/intel_gvt.c +++ b/drivers/gpu/drm/i915/intel_gvt.c @@ -45,6 +45,8 @@ static bool is_supported_device(struct drm_i915_private *dev_priv) return true; if (IS_SKYLAKE(dev_priv)) return true; + if (IS_KABYLAKE(dev_priv) && INTEL_DEVID(dev_priv) == 0x591D) + return true; return false; } diff --git a/drivers/gpu/drm/i915/intel_hdmi.c b/drivers/gpu/drm/i915/intel_hdmi.c index 3eec74ca5116..1d623b5e09d6 100644 --- a/drivers/gpu/drm/i915/intel_hdmi.c +++ b/drivers/gpu/drm/i915/intel_hdmi.c @@ -34,6 +34,7 @@ #include <drm/drm_atomic_helper.h> #include <drm/drm_crtc.h> #include <drm/drm_edid.h> +#include <drm/drm_scdc_helper.h> #include "intel_drv.h" #include <drm/i915_drm.h> #include <drm/intel_lpe_audio.h> @@ -1208,6 +1209,8 @@ static int intel_hdmi_source_max_tmds_clock(struct drm_i915_private *dev_priv) { if (IS_G4X(dev_priv)) return 165000; + else if (IS_GEMINILAKE(dev_priv)) + return 594000; else if (IS_HASWELL(dev_priv) || INTEL_INFO(dev_priv)->gen >= 8) return 300000; else @@ -1334,6 +1337,7 @@ bool intel_hdmi_compute_config(struct intel_encoder *encoder, struct intel_hdmi *intel_hdmi = enc_to_intel_hdmi(&encoder->base); struct drm_i915_private *dev_priv = to_i915(encoder->base.dev); struct drm_display_mode *adjusted_mode = &pipe_config->base.adjusted_mode; + struct drm_scdc *scdc = &conn_state->connector->display_info.hdmi.scdc; int clock_8bpc = pipe_config->base.adjusted_mode.crtc_clock; int clock_12bpc = clock_8bpc * 3 / 2; int desired_bpp; @@ -1403,6 +1407,16 @@ bool intel_hdmi_compute_config(struct intel_encoder *encoder, pipe_config->lane_count = 4; + if (scdc->scrambling.supported && IS_GEMINILAKE(dev_priv)) { + if (scdc->scrambling.low_rates) + pipe_config->hdmi_scrambling = true; + + if (pipe_config->port_clock > 340000) { + pipe_config->hdmi_scrambling = true; + pipe_config->hdmi_high_tmds_clock_ratio = true; + } + } + return true; } @@ -1812,6 +1826,57 @@ intel_hdmi_add_properties(struct intel_hdmi *intel_hdmi, struct drm_connector *c intel_hdmi->aspect_ratio = HDMI_PICTURE_ASPECT_NONE; } +/* + * intel_hdmi_handle_sink_scrambling: handle sink scrambling/clock ratio setup + * @encoder: intel_encoder + * @connector: drm_connector + * @high_tmds_clock_ratio = bool to indicate if the function needs to set + * or reset the high tmds clock ratio for scrambling + * @scrambling: bool to Indicate if the function needs to set or reset + * sink scrambling + * + * This function handles scrambling on HDMI 2.0 capable sinks. + * If required clock rate is > 340 Mhz && scrambling is supported by sink + * it enables scrambling. This should be called before enabling the HDMI + * 2.0 port, as the sink can choose to disable the scrambling if it doesn't + * detect a scrambled clock within 100 ms. + */ +void intel_hdmi_handle_sink_scrambling(struct intel_encoder *encoder, + struct drm_connector *connector, + bool high_tmds_clock_ratio, + bool scrambling) +{ + struct intel_hdmi *intel_hdmi = enc_to_intel_hdmi(&encoder->base); + struct drm_i915_private *dev_priv = connector->dev->dev_private; + struct drm_scrambling *sink_scrambling = + &connector->display_info.hdmi.scdc.scrambling; + struct i2c_adapter *adptr = intel_gmbus_get_adapter(dev_priv, + intel_hdmi->ddc_bus); + bool ret; + + if (!sink_scrambling->supported) + return; + + DRM_DEBUG_KMS("Setting sink scrambling for enc:%s connector:%s\n", + encoder->base.name, connector->name); + + /* Set TMDS bit clock ratio to 1/40 or 1/10 */ + ret = drm_scdc_set_high_tmds_clock_ratio(adptr, high_tmds_clock_ratio); + if (!ret) { + DRM_ERROR("Set TMDS ratio failed\n"); + return; + } + + /* Enable/disable sink scrambling */ + ret = drm_scdc_set_scrambling(adptr, scrambling); + if (!ret) { + DRM_ERROR("Set sink scrambling failed\n"); + return; + } + + DRM_DEBUG_KMS("sink scrambling handled\n"); +} + static u8 intel_hdmi_ddc_pin(struct drm_i915_private *dev_priv, enum port port) { diff --git a/drivers/gpu/drm/i915/intel_huc.c b/drivers/gpu/drm/i915/intel_huc.c index 7af900bcdc05..9ee819666a4c 100644 --- a/drivers/gpu/drm/i915/intel_huc.c +++ b/drivers/gpu/drm/i915/intel_huc.c @@ -251,24 +251,6 @@ fail: } /** - * intel_huc_fini() - clean up resources allocated for HuC - * @dev_priv: the drm_i915_private device - * - * Cleans up by releasing the huc firmware GEM obj. - */ -void intel_huc_fini(struct drm_i915_private *dev_priv) -{ - struct intel_uc_fw *huc_fw = &dev_priv->huc.fw; - struct drm_i915_gem_object *obj; - - obj = fetch_and_zero(&huc_fw->obj); - if (obj) - i915_gem_object_put(obj); - - huc_fw->fetch_status = INTEL_UC_FIRMWARE_NONE; -} - -/** * intel_guc_auth_huc() - authenticate ucode * @dev_priv: the drm_i915_device * diff --git a/drivers/gpu/drm/i915/intel_lpe_audio.c b/drivers/gpu/drm/i915/intel_lpe_audio.c index 7a5b41b1c024..d8ca187ae001 100644 --- a/drivers/gpu/drm/i915/intel_lpe_audio.c +++ b/drivers/gpu/drm/i915/intel_lpe_audio.c @@ -331,6 +331,7 @@ void intel_lpe_audio_teardown(struct drm_i915_private *dev_priv) * audio driver and i915 * @dev_priv: the i915 drm device private data * @eld : ELD data + * @pipe: pipe id * @port: port id * @tmds_clk_speed: tmds clock frequency in Hz * diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c index 77168e673e0a..c8f7c631fc1f 100644 --- a/drivers/gpu/drm/i915/intel_lrc.c +++ b/drivers/gpu/drm/i915/intel_lrc.c @@ -326,7 +326,7 @@ static u64 execlists_update_context(struct drm_i915_gem_request *rq) rq->ctx->ppgtt ?: rq->i915->mm.aliasing_ppgtt; u32 *reg_state = ce->lrc_reg_state; - GEM_BUG_ON(!IS_ALIGNED(rq->tail, 8)); + assert_ring_tail_valid(rq->ring, rq->tail); reg_state[CTX_RING_TAIL+1] = rq->tail; /* True 32b PPGTT with dynamic page allocation: update PDP @@ -399,22 +399,9 @@ static void execlists_dequeue(struct intel_engine_cs *engine) { struct drm_i915_gem_request *last; struct execlist_port *port = engine->execlist_port; - unsigned long flags; struct rb_node *rb; bool submit = false; - /* After execlist_first is updated, the tasklet will be rescheduled. - * - * If we are currently running (inside the tasklet) and a third - * party queues a request and so updates engine->execlist_first under - * the spinlock (which we have elided), it will atomically set the - * TASKLET_SCHED flag causing the us to be re-executed and pick up - * the change in state (the update to TASKLET_SCHED incurs a memory - * barrier making this cross-cpu checking safe). - */ - if (!READ_ONCE(engine->execlist_first)) - return; - last = port->request; if (last) /* WaIdleLiteRestore:bdw,skl @@ -448,7 +435,7 @@ static void execlists_dequeue(struct intel_engine_cs *engine) * and context switches) submission. */ - spin_lock_irqsave(&engine->timeline->lock, flags); + spin_lock_irq(&engine->timeline->lock); rb = engine->execlist_first; while (rb) { struct drm_i915_gem_request *cursor = @@ -500,7 +487,7 @@ static void execlists_dequeue(struct intel_engine_cs *engine) i915_gem_request_assign(&port->request, last); engine->execlist_first = rb; } - spin_unlock_irqrestore(&engine->timeline->lock, flags); + spin_unlock_irq(&engine->timeline->lock); if (submit) execlists_submit_ports(engine); @@ -530,24 +517,36 @@ static void intel_lrc_irq_handler(unsigned long data) intel_uncore_forcewake_get(dev_priv, engine->fw_domains); - while (test_and_clear_bit(ENGINE_IRQ_EXECLIST, &engine->irq_posted)) { + /* Prefer doing test_and_clear_bit() as a two stage operation to avoid + * imposing the cost of a locked atomic transaction when submitting a + * new request (outside of the context-switch interrupt). + */ + while (test_bit(ENGINE_IRQ_EXECLIST, &engine->irq_posted)) { u32 __iomem *csb_mmio = dev_priv->regs + i915_mmio_reg_offset(RING_CONTEXT_STATUS_PTR(engine)); u32 __iomem *buf = dev_priv->regs + i915_mmio_reg_offset(RING_CONTEXT_STATUS_BUF_LO(engine, 0)); - unsigned int csb, head, tail; - - csb = readl(csb_mmio); - head = GEN8_CSB_READ_PTR(csb); - tail = GEN8_CSB_WRITE_PTR(csb); - if (head == tail) - break; + unsigned int head, tail; + + /* The write will be ordered by the uncached read (itself + * a memory barrier), so we do not need another in the form + * of a locked instruction. The race between the interrupt + * handler and the split test/clear is harmless as we order + * our clear before the CSB read. If the interrupt arrived + * first between the test and the clear, we read the updated + * CSB and clear the bit. If the interrupt arrives as we read + * the CSB or later (i.e. after we had cleared the bit) the bit + * is set and we do a new loop. + */ + __clear_bit(ENGINE_IRQ_EXECLIST, &engine->irq_posted); + head = readl(csb_mmio); + tail = GEN8_CSB_WRITE_PTR(head); + head = GEN8_CSB_READ_PTR(head); + while (head != tail) { + unsigned int status; - if (tail < head) - tail += GEN8_CSB_ENTRIES; - do { - unsigned int idx = ++head % GEN8_CSB_ENTRIES; - unsigned int status = readl(buf + 2 * idx); + if (++head == GEN8_CSB_ENTRIES) + head = 0; /* We are flying near dragons again. * @@ -566,11 +565,12 @@ static void intel_lrc_irq_handler(unsigned long data) * status notifier. */ + status = readl(buf + 2 * head); if (!(status & GEN8_CTX_STATUS_COMPLETED_MASK)) continue; /* Check the context/desc id for this event matches */ - GEM_DEBUG_BUG_ON(readl(buf + 2 * idx + 1) != + GEM_DEBUG_BUG_ON(readl(buf + 2 * head + 1) != port[0].context_id); GEM_BUG_ON(port[0].count == 0); @@ -588,10 +588,9 @@ static void intel_lrc_irq_handler(unsigned long data) GEM_BUG_ON(port[0].count == 0 && !(status & GEN8_CTX_STATUS_ACTIVE_IDLE)); - } while (head < tail); + } - writel(_MASKED_FIELD(GEN8_CSB_READ_PTR_MASK, - GEN8_CSB_WRITE_PTR(csb) << 8), + writel(_MASKED_FIELD(GEN8_CSB_READ_PTR_MASK, head << 8), csb_mmio); } @@ -647,15 +646,14 @@ static void execlists_submit_request(struct drm_i915_gem_request *request) static struct intel_engine_cs * pt_lock_engine(struct i915_priotree *pt, struct intel_engine_cs *locked) { - struct intel_engine_cs *engine; + struct intel_engine_cs *engine = + container_of(pt, struct drm_i915_gem_request, priotree)->engine; + + GEM_BUG_ON(!locked); - engine = container_of(pt, - struct drm_i915_gem_request, - priotree)->engine; if (engine != locked) { - if (locked) - spin_unlock_irq(&locked->timeline->lock); - spin_lock_irq(&engine->timeline->lock); + spin_unlock(&locked->timeline->lock); + spin_lock(&engine->timeline->lock); } return engine; @@ -663,7 +661,7 @@ pt_lock_engine(struct i915_priotree *pt, struct intel_engine_cs *locked) static void execlists_schedule(struct drm_i915_gem_request *request, int prio) { - struct intel_engine_cs *engine = NULL; + struct intel_engine_cs *engine; struct i915_dependency *dep, *p; struct i915_dependency stack; LIST_HEAD(dfs); @@ -697,26 +695,23 @@ static void execlists_schedule(struct drm_i915_gem_request *request, int prio) list_for_each_entry_safe(dep, p, &dfs, dfs_link) { struct i915_priotree *pt = dep->signaler; - list_for_each_entry(p, &pt->signalers_list, signal_link) + /* Within an engine, there can be no cycle, but we may + * refer to the same dependency chain multiple times + * (redundant dependencies are not eliminated) and across + * engines. + */ + list_for_each_entry(p, &pt->signalers_list, signal_link) { + GEM_BUG_ON(p->signaler->priority < pt->priority); if (prio > READ_ONCE(p->signaler->priority)) list_move_tail(&p->dfs_link, &dfs); + } list_safe_reset_next(dep, p, dfs_link); - if (!RB_EMPTY_NODE(&pt->node)) - continue; - - engine = pt_lock_engine(pt, engine); - - /* If it is not already in the rbtree, we can update the - * priority inplace and skip over it (and its dependencies) - * if it is referenced *again* as we descend the dfs. - */ - if (prio > pt->priority && RB_EMPTY_NODE(&pt->node)) { - pt->priority = prio; - list_del_init(&dep->dfs_link); - } } + engine = request->engine; + spin_lock_irq(&engine->timeline->lock); + /* Fifo and depth-first replacement ensure our deps execute before us */ list_for_each_entry_safe_reverse(dep, p, &dfs, dfs_link) { struct i915_priotree *pt = dep->signaler; @@ -728,16 +723,15 @@ static void execlists_schedule(struct drm_i915_gem_request *request, int prio) if (prio <= pt->priority) continue; - GEM_BUG_ON(RB_EMPTY_NODE(&pt->node)); - pt->priority = prio; - rb_erase(&pt->node, &engine->execlist_queue); - if (insert_request(pt, &engine->execlist_queue)) - engine->execlist_first = &pt->node; + if (!RB_EMPTY_NODE(&pt->node)) { + rb_erase(&pt->node, &engine->execlist_queue); + if (insert_request(pt, &engine->execlist_queue)) + engine->execlist_first = &pt->node; + } } - if (engine) - spin_unlock_irq(&engine->timeline->lock); + spin_unlock_irq(&engine->timeline->lock); /* XXX Do we need to preempt to make room for us and our deps? */ } @@ -1255,7 +1249,6 @@ static void reset_common_ring(struct intel_engine_cs *engine, ce->lrc_reg_state[CTX_RING_HEAD+1] = request->postfix; request->ring->head = request->postfix; - request->ring->last_retired_head = -1; intel_ring_update_space(request->ring); /* Catch up with any missed context-switch interrupts */ @@ -1268,8 +1261,10 @@ static void reset_common_ring(struct intel_engine_cs *engine, GEM_BUG_ON(request->ctx != port[0].request->ctx); /* Reset WaIdleLiteRestore:bdw,skl as well */ - request->tail = request->wa_tail - WA_TAIL_DWORDS * sizeof(u32); - GEM_BUG_ON(!IS_ALIGNED(request->tail, 8)); + request->tail = + intel_ring_wrap(request->ring, + request->wa_tail - WA_TAIL_DWORDS*sizeof(u32)); + assert_ring_tail_valid(request->ring, request->tail); } static int intel_logical_ring_emit_pdps(struct drm_i915_gem_request *req) @@ -1480,7 +1475,7 @@ static void gen8_emit_breadcrumb(struct drm_i915_gem_request *request, u32 *cs) *cs++ = MI_USER_INTERRUPT; *cs++ = MI_NOOP; request->tail = intel_ring_offset(request, cs); - GEM_BUG_ON(!IS_ALIGNED(request->tail, 8)); + assert_ring_tail_valid(request->ring, request->tail); gen8_emit_wa_tail(request, cs); } @@ -1508,7 +1503,7 @@ static void gen8_emit_breadcrumb_render(struct drm_i915_gem_request *request, *cs++ = MI_USER_INTERRUPT; *cs++ = MI_NOOP; request->tail = intel_ring_offset(request, cs); - GEM_BUG_ON(!IS_ALIGNED(request->tail, 8)); + assert_ring_tail_valid(request->ring, request->tail); gen8_emit_wa_tail(request, cs); } @@ -1575,6 +1570,7 @@ static void execlists_set_default_submission(struct intel_engine_cs *engine) { engine->submit_request = execlists_submit_request; engine->schedule = execlists_schedule; + engine->irq_tasklet.func = intel_lrc_irq_handler; } static void @@ -2041,7 +2037,6 @@ void intel_lr_context_resume(struct drm_i915_private *dev_priv) i915_gem_object_unpin_map(ce->state->obj); ce->ring->head = ce->ring->tail = 0; - ce->ring->last_retired_head = -1; intel_ring_update_space(ce->ring); } } diff --git a/drivers/gpu/drm/i915/intel_opregion.c b/drivers/gpu/drm/i915/intel_opregion.c index 441c01466384..d44465190dc1 100644 --- a/drivers/gpu/drm/i915/intel_opregion.c +++ b/drivers/gpu/drm/i915/intel_opregion.c @@ -920,6 +920,8 @@ int intel_opregion_setup(struct drm_i915_private *dev_priv) char buf[sizeof(OPREGION_SIGNATURE)]; int err = 0; void *base; + const void *vbt; + u32 vbt_size; BUILD_BUG_ON(sizeof(struct opregion_header) != 0x100); BUILD_BUG_ON(sizeof(struct opregion_acpi) != 0x100); @@ -972,45 +974,46 @@ int intel_opregion_setup(struct drm_i915_private *dev_priv) if (mboxes & MBOX_ASLE_EXT) DRM_DEBUG_DRIVER("ASLE extension supported\n"); - if (!dmi_check_system(intel_no_opregion_vbt)) { - const void *vbt = NULL; - u32 vbt_size = 0; - - if (opregion->header->opregion_ver >= 2 && opregion->asle && - opregion->asle->rvda && opregion->asle->rvds) { - opregion->rvda = memremap(opregion->asle->rvda, - opregion->asle->rvds, - MEMREMAP_WB); - vbt = opregion->rvda; - vbt_size = opregion->asle->rvds; - } + if (dmi_check_system(intel_no_opregion_vbt)) + goto out; + if (opregion->header->opregion_ver >= 2 && opregion->asle && + opregion->asle->rvda && opregion->asle->rvds) { + opregion->rvda = memremap(opregion->asle->rvda, + opregion->asle->rvds, + MEMREMAP_WB); + vbt = opregion->rvda; + vbt_size = opregion->asle->rvds; if (intel_bios_is_valid_vbt(vbt, vbt_size)) { DRM_DEBUG_KMS("Found valid VBT in ACPI OpRegion (RVDA)\n"); opregion->vbt = vbt; opregion->vbt_size = vbt_size; + goto out; } else { - vbt = base + OPREGION_VBT_OFFSET; - /* - * The VBT specification says that if the ASLE ext - * mailbox is not used its area is reserved, but - * on some CHT boards the VBT extends into the - * ASLE ext area. Allow this even though it is - * against the spec, so we do not end up rejecting - * the VBT on those boards (and end up not finding the - * LCD panel because of this). - */ - vbt_size = (mboxes & MBOX_ASLE_EXT) ? - OPREGION_ASLE_EXT_OFFSET : OPREGION_SIZE; - vbt_size -= OPREGION_VBT_OFFSET; - if (intel_bios_is_valid_vbt(vbt, vbt_size)) { - DRM_DEBUG_KMS("Found valid VBT in ACPI OpRegion (Mailbox #4)\n"); - opregion->vbt = vbt; - opregion->vbt_size = vbt_size; - } + DRM_DEBUG_KMS("Invalid VBT in ACPI OpRegion (RVDA)\n"); } } + vbt = base + OPREGION_VBT_OFFSET; + /* + * The VBT specification says that if the ASLE ext mailbox is not used + * its area is reserved, but on some CHT boards the VBT extends into the + * ASLE ext area. Allow this even though it is against the spec, so we + * do not end up rejecting the VBT on those boards (and end up not + * finding the LCD panel because of this). + */ + vbt_size = (mboxes & MBOX_ASLE_EXT) ? + OPREGION_ASLE_EXT_OFFSET : OPREGION_SIZE; + vbt_size -= OPREGION_VBT_OFFSET; + if (intel_bios_is_valid_vbt(vbt, vbt_size)) { + DRM_DEBUG_KMS("Found valid VBT in ACPI OpRegion (Mailbox #4)\n"); + opregion->vbt = vbt; + opregion->vbt_size = vbt_size; + } else { + DRM_DEBUG_KMS("Invalid VBT in ACPI OpRegion (Mailbox #4)\n"); + } + +out: return 0; err_out: diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index aece0ff88a5d..570bd603f401 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -655,6 +655,29 @@ static unsigned long intel_calculate_wm(unsigned long clock_in_khz, return wm_size; } +static bool intel_wm_plane_visible(const struct intel_crtc_state *crtc_state, + const struct intel_plane_state *plane_state) +{ + struct intel_plane *plane = to_intel_plane(plane_state->base.plane); + + /* FIXME check the 'enable' instead */ + if (!crtc_state->base.active) + return false; + + /* + * Treat cursor with fb as always visible since cursor updates + * can happen faster than the vrefresh rate, and the current + * watermark code doesn't handle that correctly. Cursor updates + * which set/clear the fb or change the cursor size are going + * to get throttled by intel_legacy_cursor_update() to work + * around this problem with the watermark code. + */ + if (plane->id == PLANE_CURSOR) + return plane_state->base.fb != NULL; + else + return plane_state->base.visible; +} + static struct intel_crtc *single_enabled_crtc(struct drm_i915_private *dev_priv) { struct intel_crtc *crtc, *enabled = NULL; @@ -1961,7 +1984,7 @@ static uint32_t ilk_compute_pri_wm(const struct intel_crtc_state *cstate, uint32_t method1, method2; int cpp; - if (!cstate->base.active || !pstate->base.visible) + if (!intel_wm_plane_visible(cstate, pstate)) return 0; cpp = pstate->base.fb->format->cpp[0]; @@ -1990,7 +2013,7 @@ static uint32_t ilk_compute_spr_wm(const struct intel_crtc_state *cstate, uint32_t method1, method2; int cpp; - if (!cstate->base.active || !pstate->base.visible) + if (!intel_wm_plane_visible(cstate, pstate)) return 0; cpp = pstate->base.fb->format->cpp[0]; @@ -2013,15 +2036,7 @@ static uint32_t ilk_compute_cur_wm(const struct intel_crtc_state *cstate, { int cpp; - /* - * Treat cursor with fb as always visible since cursor updates - * can happen faster than the vrefresh rate, and the current - * watermark code doesn't handle that correctly. Cursor updates - * which set/clear the fb or change the cursor size are going - * to get throttled by intel_legacy_cursor_update() to work - * around this problem with the watermark code. - */ - if (!cstate->base.active || !pstate->base.fb) + if (!intel_wm_plane_visible(cstate, pstate)) return 0; cpp = pstate->base.fb->format->cpp[0]; @@ -2038,7 +2053,7 @@ static uint32_t ilk_compute_fbc_wm(const struct intel_crtc_state *cstate, { int cpp; - if (!cstate->base.active || !pstate->base.visible) + if (!intel_wm_plane_visible(cstate, pstate)) return 0; cpp = pstate->base.fb->format->cpp[0]; @@ -3346,19 +3361,29 @@ void skl_ddb_get_hw_state(struct drm_i915_private *dev_priv, * Caller should take care of dividing & rounding off the value. */ static uint32_t -skl_plane_downscale_amount(const struct intel_plane_state *pstate) +skl_plane_downscale_amount(const struct intel_crtc_state *cstate, + const struct intel_plane_state *pstate) { + struct intel_plane *plane = to_intel_plane(pstate->base.plane); uint32_t downscale_h, downscale_w; uint32_t src_w, src_h, dst_w, dst_h; - if (WARN_ON(!pstate->base.visible)) + if (WARN_ON(!intel_wm_plane_visible(cstate, pstate))) return DRM_PLANE_HELPER_NO_SCALING; /* n.b., src is 16.16 fixed point, dst is whole integer */ - src_w = drm_rect_width(&pstate->base.src); - src_h = drm_rect_height(&pstate->base.src); - dst_w = drm_rect_width(&pstate->base.dst); - dst_h = drm_rect_height(&pstate->base.dst); + if (plane->id == PLANE_CURSOR) { + src_w = pstate->base.src_w; + src_h = pstate->base.src_h; + dst_w = pstate->base.crtc_w; + dst_h = pstate->base.crtc_h; + } else { + src_w = drm_rect_width(&pstate->base.src); + src_h = drm_rect_height(&pstate->base.src); + dst_w = drm_rect_width(&pstate->base.dst); + dst_h = drm_rect_height(&pstate->base.dst); + } + if (drm_rotation_90_or_270(pstate->base.rotation)) swap(dst_w, dst_h); @@ -3374,6 +3399,7 @@ skl_plane_relative_data_rate(const struct intel_crtc_state *cstate, const struct drm_plane_state *pstate, int y) { + struct intel_plane *plane = to_intel_plane(pstate->plane); struct intel_plane_state *intel_pstate = to_intel_plane_state(pstate); uint32_t down_scale_amount, data_rate; uint32_t width = 0, height = 0; @@ -3386,7 +3412,7 @@ skl_plane_relative_data_rate(const struct intel_crtc_state *cstate, fb = pstate->fb; format = fb->format->format; - if (pstate->plane->type == DRM_PLANE_TYPE_CURSOR) + if (plane->id == PLANE_CURSOR) return 0; if (y && format != DRM_FORMAT_NV12) return 0; @@ -3410,7 +3436,7 @@ skl_plane_relative_data_rate(const struct intel_crtc_state *cstate, data_rate = width * height * fb->format->cpp[0]; } - down_scale_amount = skl_plane_downscale_amount(intel_pstate); + down_scale_amount = skl_plane_downscale_amount(cstate, intel_pstate); return (uint64_t)data_rate * down_scale_amount >> 16; } @@ -3702,7 +3728,7 @@ static uint32_t skl_adjusted_plane_pixel_rate(const struct intel_crtc_state *cst uint64_t pixel_rate; /* Shouldn't reach here on disabled planes... */ - if (WARN_ON(!pstate->base.visible)) + if (WARN_ON(!intel_wm_plane_visible(cstate, pstate))) return 0; /* @@ -3710,7 +3736,7 @@ static uint32_t skl_adjusted_plane_pixel_rate(const struct intel_crtc_state *cst * with additional adjustments for plane-specific scaling. */ adjusted_pixel_rate = cstate->pixel_rate; - downscale_amount = skl_plane_downscale_amount(pstate); + downscale_amount = skl_plane_downscale_amount(cstate, pstate); pixel_rate = adjusted_pixel_rate * downscale_amount >> 16; WARN_ON(pixel_rate != clamp_t(uint32_t, pixel_rate, 0, ~0)); @@ -3727,6 +3753,7 @@ static int skl_compute_plane_wm(const struct drm_i915_private *dev_priv, uint8_t *out_lines, /* out */ bool *enabled /* out */) { + struct intel_plane *plane = to_intel_plane(intel_pstate->base.plane); struct drm_plane_state *pstate = &intel_pstate->base; struct drm_framebuffer *fb = pstate->fb; uint32_t latency = dev_priv->wm.skl_latency[level]; @@ -3746,7 +3773,8 @@ static int skl_compute_plane_wm(const struct drm_i915_private *dev_priv, bool apply_memory_bw_wa = skl_needs_memory_bw_wa(state); bool y_tiled, x_tiled; - if (latency == 0 || !cstate->base.active || !intel_pstate->base.visible) { + if (latency == 0 || + !intel_wm_plane_visible(cstate, intel_pstate)) { *enabled = false; return 0; } @@ -3762,8 +3790,13 @@ static int skl_compute_plane_wm(const struct drm_i915_private *dev_priv, if (apply_memory_bw_wa && x_tiled) latency += 15; - width = drm_rect_width(&intel_pstate->base.src) >> 16; - height = drm_rect_height(&intel_pstate->base.src) >> 16; + if (plane->id == PLANE_CURSOR) { + width = intel_pstate->base.crtc_w; + height = intel_pstate->base.crtc_h; + } else { + width = drm_rect_width(&intel_pstate->base.src) >> 16; + height = drm_rect_height(&intel_pstate->base.src) >> 16; + } if (drm_rotation_90_or_270(pstate->rotation)) swap(width, height); @@ -8055,7 +8088,7 @@ static inline int gen6_check_mailbox_status(struct drm_i915_private *dev_priv) case GEN6_PCODE_TIMEOUT: return -ETIMEDOUT; default: - MISSING_CASE(flags) + MISSING_CASE(flags); return 0; } } @@ -8355,6 +8388,7 @@ static u64 vlv_residency_raw(struct drm_i915_private *dev_priv, const i915_reg_t reg) { u32 lower, upper, tmp; + int loop = 2; /* The register accessed do not need forcewake. We borrow * uncore lock to prevent concurrent access to range reg. @@ -8383,7 +8417,7 @@ static u64 vlv_residency_raw(struct drm_i915_private *dev_priv, I915_WRITE_FW(VLV_COUNTER_CONTROL, _MASKED_BIT_ENABLE(VLV_COUNT_RANGE_HIGH)); upper = I915_READ_FW(reg); - } while (upper != tmp); + } while (upper != tmp && --loop); /* Everywhere else we always use VLV_COUNTER_CONTROL with the * VLV_COUNT_RANGE_HIGH bit set - so it is safe to leave it set diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c index f7c6383ecd06..66a2b8b83972 100644 --- a/drivers/gpu/drm/i915/intel_ringbuffer.c +++ b/drivers/gpu/drm/i915/intel_ringbuffer.c @@ -49,13 +49,7 @@ static int __intel_ring_space(int head, int tail, int size) void intel_ring_update_space(struct intel_ring *ring) { - if (ring->last_retired_head != -1) { - ring->head = ring->last_retired_head; - ring->last_retired_head = -1; - } - - ring->space = __intel_ring_space(ring->head & HEAD_ADDR, - ring->tail, ring->size); + ring->space = __intel_ring_space(ring->head, ring->tail, ring->size); } static int @@ -618,12 +612,8 @@ static void reset_ring_common(struct intel_engine_cs *engine, } /* If the rq hung, jump to its breadcrumb and skip the batch */ - if (request->fence.error == -EIO) { - struct intel_ring *ring = request->ring; - - ring->head = request->postfix; - ring->last_retired_head = -1; - } + if (request->fence.error == -EIO) + request->ring->head = request->postfix; } else { engine->legacy_active_context = NULL; } @@ -784,7 +774,7 @@ static void i9xx_submit_request(struct drm_i915_gem_request *request) i915_gem_request_submit(request); - GEM_BUG_ON(!IS_ALIGNED(request->tail, 8)); + assert_ring_tail_valid(request->ring, request->tail); I915_WRITE_TAIL(request->engine, request->tail); } @@ -796,7 +786,7 @@ static void i9xx_emit_breadcrumb(struct drm_i915_gem_request *req, u32 *cs) *cs++ = MI_USER_INTERRUPT; req->tail = intel_ring_offset(req, cs); - GEM_BUG_ON(!IS_ALIGNED(req->tail, 8)); + assert_ring_tail_valid(req->ring, req->tail); } static const int i9xx_emit_breadcrumb_sz = 4; @@ -835,7 +825,7 @@ static void gen8_render_emit_breadcrumb(struct drm_i915_gem_request *req, *cs++ = MI_NOOP; req->tail = intel_ring_offset(req, cs); - GEM_BUG_ON(!IS_ALIGNED(req->tail, 8)); + assert_ring_tail_valid(req->ring, req->tail); } static const int gen8_render_emit_breadcrumb_sz = 8; @@ -1392,7 +1382,6 @@ intel_engine_create_ring(struct intel_engine_cs *engine, int size) if (IS_I830(engine->i915) || IS_I845G(engine->i915)) ring->effective_size -= 2 * CACHELINE_BYTES; - ring->last_retired_head = -1; intel_ring_update_space(ring); vma = intel_ring_create_vma(engine->i915, size); @@ -1573,10 +1562,8 @@ void intel_legacy_submission_resume(struct drm_i915_private *dev_priv) struct intel_engine_cs *engine; enum intel_engine_id id; - for_each_engine(engine, dev_priv, id) { + for_each_engine(engine, dev_priv, id) engine->buffer->head = engine->buffer->tail; - engine->buffer->last_retired_head = -1; - } } static int ring_request_alloc(struct drm_i915_gem_request *request) @@ -2130,7 +2117,7 @@ int intel_init_render_ring_buffer(struct intel_engine_cs *engine) num_rings = hweight32(INTEL_INFO(dev_priv)->ring_mask) - 1; - engine->emit_breadcrumb_sz += num_rings * 6; + engine->emit_breadcrumb_sz += num_rings * 8; } } else if (INTEL_GEN(dev_priv) >= 6) { engine->init_context = intel_rcs_ctx_init; diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.h b/drivers/gpu/drm/i915/intel_ringbuffer.h index 847aea554464..a82a0807f64d 100644 --- a/drivers/gpu/drm/i915/intel_ringbuffer.h +++ b/drivers/gpu/drm/i915/intel_ringbuffer.h @@ -149,16 +149,6 @@ struct intel_ring { int space; int size; int effective_size; - - /** We track the position of the requests in the ring buffer, and - * when each is retired we increment last_retired_head as the GPU - * must have finished processing the request and so we know we - * can advance the ringbuffer up to that position. - * - * last_retired_head is set to -1 after the value is consumed so - * we can detect new retirements. - */ - u32 last_retired_head; }; struct i915_gem_context; @@ -442,18 +432,10 @@ struct intel_engine_cs { u32 (*get_cmd_length_mask)(u32 cmd_header); }; -static inline unsigned +static inline unsigned int intel_engine_flag(const struct intel_engine_cs *engine) { - return 1 << engine->id; -} - -static inline void -intel_flush_status_page(struct intel_engine_cs *engine, int reg) -{ - mb(); - clflush(&engine->status_page.page_addr[reg]); - mb(); + return BIT(engine->id); } static inline u32 @@ -464,14 +446,22 @@ intel_read_status_page(struct intel_engine_cs *engine, int reg) } static inline void -intel_write_status_page(struct intel_engine_cs *engine, - int reg, u32 value) +intel_write_status_page(struct intel_engine_cs *engine, int reg, u32 value) { - mb(); - clflush(&engine->status_page.page_addr[reg]); - engine->status_page.page_addr[reg] = value; - clflush(&engine->status_page.page_addr[reg]); - mb(); + /* Writing into the status page should be done sparingly. Since + * we do when we are uncertain of the device state, we take a bit + * of extra paranoia to try and ensure that the HWS takes the value + * we give and that it doesn't end up trapped inside the CPU! + */ + if (static_cpu_has(X86_FEATURE_CLFLUSH)) { + mb(); + clflush(&engine->status_page.page_addr[reg]); + engine->status_page.page_addr[reg] = value; + clflush(&engine->status_page.page_addr[reg]); + mb(); + } else { + WRITE_ONCE(engine->status_page.page_addr[reg], value); + } } /* @@ -525,12 +515,29 @@ intel_ring_advance(struct drm_i915_gem_request *req, u32 *cs) } static inline u32 -intel_ring_offset(struct drm_i915_gem_request *req, void *addr) +intel_ring_wrap(const struct intel_ring *ring, u32 pos) +{ + return pos & (ring->size - 1); +} + +static inline u32 +intel_ring_offset(const struct drm_i915_gem_request *req, void *addr) { /* Don't write ring->size (equivalent to 0) as that hangs some GPUs. */ u32 offset = addr - req->ring->vaddr; GEM_BUG_ON(offset > req->ring->size); - return offset & (req->ring->size - 1); + return intel_ring_wrap(req->ring, offset); +} + +static inline void +assert_ring_tail_valid(const struct intel_ring *ring, unsigned int tail) +{ + /* We could combine these into a single tail operation, but keeping + * them as seperate tests will help identify the cause should one + * ever fire. + */ + GEM_BUG_ON(!IS_ALIGNED(tail, 8)); + GEM_BUG_ON(tail >= ring->size); } void intel_ring_update_space(struct intel_ring *ring); diff --git a/drivers/gpu/drm/i915/intel_runtime_pm.c b/drivers/gpu/drm/i915/intel_runtime_pm.c index 012bc358a33a..f8a375f8dde6 100644 --- a/drivers/gpu/drm/i915/intel_runtime_pm.c +++ b/drivers/gpu/drm/i915/intel_runtime_pm.c @@ -2840,8 +2840,10 @@ void intel_runtime_pm_get(struct drm_i915_private *dev_priv) { struct pci_dev *pdev = dev_priv->drm.pdev; struct device *kdev = &pdev->dev; + int ret; - pm_runtime_get_sync(kdev); + ret = pm_runtime_get_sync(kdev); + WARN_ONCE(ret < 0, "pm_runtime_get_sync() failed: %d\n", ret); atomic_inc(&dev_priv->pm.wakeref_count); assert_rpm_wakelock_held(dev_priv); @@ -2871,7 +2873,8 @@ bool intel_runtime_pm_get_if_in_use(struct drm_i915_private *dev_priv) * function, since the power state is undefined. This applies * atm to the late/early system suspend/resume handlers. */ - WARN_ON_ONCE(ret < 0); + WARN_ONCE(ret < 0, + "pm_runtime_get_if_in_use() failed: %d\n", ret); if (ret <= 0) return false; } @@ -2955,8 +2958,11 @@ void intel_runtime_pm_enable(struct drm_i915_private *dev_priv) * platforms without RPM support. */ if (!HAS_RUNTIME_PM(dev_priv)) { + int ret; + pm_runtime_dont_use_autosuspend(kdev); - pm_runtime_get_sync(kdev); + ret = pm_runtime_get_sync(kdev); + WARN(ret < 0, "pm_runtime_get_sync() failed: %d\n", ret); } else { pm_runtime_use_autosuspend(kdev); } diff --git a/drivers/gpu/drm/i915/intel_sprite.c b/drivers/gpu/drm/i915/intel_sprite.c index b931d0bd7a64..f7d431427115 100644 --- a/drivers/gpu/drm/i915/intel_sprite.c +++ b/drivers/gpu/drm/i915/intel_sprite.c @@ -217,7 +217,7 @@ skl_update_plane(struct drm_plane *drm_plane, struct drm_framebuffer *fb = plane_state->base.fb; enum plane_id plane_id = intel_plane->id; enum pipe pipe = intel_plane->pipe; - u32 plane_ctl; + u32 plane_ctl = plane_state->ctl; const struct drm_intel_sprite_colorkey *key = &plane_state->ckey; u32 surf_addr = plane_state->main.offset; unsigned int rotation = plane_state->base.rotation; @@ -232,24 +232,6 @@ skl_update_plane(struct drm_plane *drm_plane, uint32_t src_h = drm_rect_height(&plane_state->base.src) >> 16; unsigned long irqflags; - plane_ctl = PLANE_CTL_ENABLE; - - if (!IS_GEMINILAKE(dev_priv)) { - plane_ctl |= - PLANE_CTL_PIPE_GAMMA_ENABLE | - PLANE_CTL_PIPE_CSC_ENABLE | - PLANE_CTL_PLANE_GAMMA_DISABLE; - } - - plane_ctl |= skl_plane_ctl_format(fb->format->format); - plane_ctl |= skl_plane_ctl_tiling(fb->modifier); - plane_ctl |= skl_plane_ctl_rotation(rotation); - - if (key->flags & I915_SET_COLORKEY_DESTINATION) - plane_ctl |= PLANE_CTL_KEY_ENABLE_DESTINATION; - else if (key->flags & I915_SET_COLORKEY_SOURCE) - plane_ctl |= PLANE_CTL_KEY_ENABLE_SOURCE; - /* Sizes are 0 based */ src_w--; src_h--; @@ -361,32 +343,15 @@ chv_update_csc(struct intel_plane *intel_plane, uint32_t format) I915_WRITE_FW(SPCSCCROCLAMP(plane_id), SPCSC_OMAX(1023) | SPCSC_OMIN(0)); } -static void -vlv_update_plane(struct drm_plane *dplane, - const struct intel_crtc_state *crtc_state, - const struct intel_plane_state *plane_state) +static u32 vlv_sprite_ctl(const struct intel_crtc_state *crtc_state, + const struct intel_plane_state *plane_state) { - struct drm_device *dev = dplane->dev; - struct drm_i915_private *dev_priv = to_i915(dev); - struct intel_plane *intel_plane = to_intel_plane(dplane); - struct drm_framebuffer *fb = plane_state->base.fb; - enum pipe pipe = intel_plane->pipe; - enum plane_id plane_id = intel_plane->id; - u32 sprctl; - u32 sprsurf_offset, linear_offset; + const struct drm_framebuffer *fb = plane_state->base.fb; unsigned int rotation = plane_state->base.rotation; const struct drm_intel_sprite_colorkey *key = &plane_state->ckey; - int crtc_x = plane_state->base.dst.x1; - int crtc_y = plane_state->base.dst.y1; - uint32_t crtc_w = drm_rect_width(&plane_state->base.dst); - uint32_t crtc_h = drm_rect_height(&plane_state->base.dst); - uint32_t x = plane_state->base.src.x1 >> 16; - uint32_t y = plane_state->base.src.y1 >> 16; - uint32_t src_w = drm_rect_width(&plane_state->base.src) >> 16; - uint32_t src_h = drm_rect_height(&plane_state->base.src) >> 16; - unsigned long irqflags; + u32 sprctl; - sprctl = SP_ENABLE; + sprctl = SP_ENABLE | SP_GAMMA_ENABLE; switch (fb->format->format) { case DRM_FORMAT_YUYV: @@ -423,20 +388,10 @@ vlv_update_plane(struct drm_plane *dplane, sprctl |= SP_FORMAT_RGBA8888; break; default: - /* - * If we get here one of the upper layers failed to filter - * out the unsupported plane formats - */ - BUG(); - break; + MISSING_CASE(fb->format->format); + return 0; } - /* - * Enable gamma to match primary/cursor plane behaviour. - * FIXME should be user controllable via propertiesa. - */ - sprctl |= SP_GAMMA_ENABLE; - if (fb->modifier == I915_FORMAT_MOD_X_TILED) sprctl |= SP_TILED; @@ -449,22 +404,36 @@ vlv_update_plane(struct drm_plane *dplane, if (key->flags & I915_SET_COLORKEY_SOURCE) sprctl |= SP_SOURCE_KEY; + return sprctl; +} + +static void +vlv_update_plane(struct drm_plane *dplane, + const struct intel_crtc_state *crtc_state, + const struct intel_plane_state *plane_state) +{ + struct drm_device *dev = dplane->dev; + struct drm_i915_private *dev_priv = to_i915(dev); + struct intel_plane *intel_plane = to_intel_plane(dplane); + struct drm_framebuffer *fb = plane_state->base.fb; + enum pipe pipe = intel_plane->pipe; + enum plane_id plane_id = intel_plane->id; + u32 sprctl = plane_state->ctl; + u32 sprsurf_offset = plane_state->main.offset; + u32 linear_offset; + const struct drm_intel_sprite_colorkey *key = &plane_state->ckey; + int crtc_x = plane_state->base.dst.x1; + int crtc_y = plane_state->base.dst.y1; + uint32_t crtc_w = drm_rect_width(&plane_state->base.dst); + uint32_t crtc_h = drm_rect_height(&plane_state->base.dst); + uint32_t x = plane_state->main.x; + uint32_t y = plane_state->main.y; + unsigned long irqflags; + /* Sizes are 0 based */ - src_w--; - src_h--; crtc_w--; crtc_h--; - intel_add_fb_offsets(&x, &y, plane_state, 0); - sprsurf_offset = intel_compute_tile_offset(&x, &y, plane_state, 0); - - if (rotation & DRM_ROTATE_180) { - x += src_w; - y += src_h; - } else if (rotation & DRM_REFLECT_X) { - x += src_w; - } - linear_offset = intel_fb_xy_to_linear(x, y, plane_state, 0); spin_lock_irqsave(&dev_priv->uncore.lock, irqflags); @@ -516,31 +485,23 @@ vlv_disable_plane(struct drm_plane *dplane, struct drm_crtc *crtc) spin_unlock_irqrestore(&dev_priv->uncore.lock, irqflags); } -static void -ivb_update_plane(struct drm_plane *plane, - const struct intel_crtc_state *crtc_state, - const struct intel_plane_state *plane_state) +static u32 ivb_sprite_ctl(const struct intel_crtc_state *crtc_state, + const struct intel_plane_state *plane_state) { - struct drm_device *dev = plane->dev; - struct drm_i915_private *dev_priv = to_i915(dev); - struct intel_plane *intel_plane = to_intel_plane(plane); - struct drm_framebuffer *fb = plane_state->base.fb; - enum pipe pipe = intel_plane->pipe; - u32 sprctl, sprscale = 0; - u32 sprsurf_offset, linear_offset; + struct drm_i915_private *dev_priv = + to_i915(plane_state->base.plane->dev); + const struct drm_framebuffer *fb = plane_state->base.fb; unsigned int rotation = plane_state->base.rotation; const struct drm_intel_sprite_colorkey *key = &plane_state->ckey; - int crtc_x = plane_state->base.dst.x1; - int crtc_y = plane_state->base.dst.y1; - uint32_t crtc_w = drm_rect_width(&plane_state->base.dst); - uint32_t crtc_h = drm_rect_height(&plane_state->base.dst); - uint32_t x = plane_state->base.src.x1 >> 16; - uint32_t y = plane_state->base.src.y1 >> 16; - uint32_t src_w = drm_rect_width(&plane_state->base.src) >> 16; - uint32_t src_h = drm_rect_height(&plane_state->base.src) >> 16; - unsigned long irqflags; + u32 sprctl; + + sprctl = SPRITE_ENABLE | SPRITE_GAMMA_ENABLE; + + if (IS_IVYBRIDGE(dev_priv)) + sprctl |= SPRITE_TRICKLE_FEED_DISABLE; - sprctl = SPRITE_ENABLE; + if (IS_HASWELL(dev_priv) || IS_BROADWELL(dev_priv)) + sprctl |= SPRITE_PIPE_CSC_ENABLE; switch (fb->format->format) { case DRM_FORMAT_XBGR8888: @@ -562,34 +523,48 @@ ivb_update_plane(struct drm_plane *plane, sprctl |= SPRITE_FORMAT_YUV422 | SPRITE_YUV_ORDER_VYUY; break; default: - BUG(); + MISSING_CASE(fb->format->format); + return 0; } - /* - * Enable gamma to match primary/cursor plane behaviour. - * FIXME should be user controllable via propertiesa. - */ - sprctl |= SPRITE_GAMMA_ENABLE; - if (fb->modifier == I915_FORMAT_MOD_X_TILED) sprctl |= SPRITE_TILED; if (rotation & DRM_ROTATE_180) sprctl |= SPRITE_ROTATE_180; - if (IS_HASWELL(dev_priv) || IS_BROADWELL(dev_priv)) - sprctl &= ~SPRITE_TRICKLE_FEED_DISABLE; - else - sprctl |= SPRITE_TRICKLE_FEED_DISABLE; - - if (IS_HASWELL(dev_priv) || IS_BROADWELL(dev_priv)) - sprctl |= SPRITE_PIPE_CSC_ENABLE; - if (key->flags & I915_SET_COLORKEY_DESTINATION) sprctl |= SPRITE_DEST_KEY; else if (key->flags & I915_SET_COLORKEY_SOURCE) sprctl |= SPRITE_SOURCE_KEY; + return sprctl; +} + +static void +ivb_update_plane(struct drm_plane *plane, + const struct intel_crtc_state *crtc_state, + const struct intel_plane_state *plane_state) +{ + struct drm_device *dev = plane->dev; + struct drm_i915_private *dev_priv = to_i915(dev); + struct intel_plane *intel_plane = to_intel_plane(plane); + struct drm_framebuffer *fb = plane_state->base.fb; + enum pipe pipe = intel_plane->pipe; + u32 sprctl = plane_state->ctl, sprscale = 0; + u32 sprsurf_offset = plane_state->main.offset; + u32 linear_offset; + const struct drm_intel_sprite_colorkey *key = &plane_state->ckey; + int crtc_x = plane_state->base.dst.x1; + int crtc_y = plane_state->base.dst.y1; + uint32_t crtc_w = drm_rect_width(&plane_state->base.dst); + uint32_t crtc_h = drm_rect_height(&plane_state->base.dst); + uint32_t x = plane_state->main.x; + uint32_t y = plane_state->main.y; + uint32_t src_w = drm_rect_width(&plane_state->base.src) >> 16; + uint32_t src_h = drm_rect_height(&plane_state->base.src) >> 16; + unsigned long irqflags; + /* Sizes are 0 based */ src_w--; src_h--; @@ -599,16 +574,6 @@ ivb_update_plane(struct drm_plane *plane, if (crtc_w != src_w || crtc_h != src_h) sprscale = SPRITE_SCALE_ENABLE | (src_w << 16) | src_h; - intel_add_fb_offsets(&x, &y, plane_state, 0); - sprsurf_offset = intel_compute_tile_offset(&x, &y, plane_state, 0); - - /* HSW+ does this automagically in hardware */ - if (!IS_HASWELL(dev_priv) && !IS_BROADWELL(dev_priv) && - rotation & DRM_ROTATE_180) { - x += src_w; - y += src_h; - } - linear_offset = intel_fb_xy_to_linear(x, y, plane_state, 0); spin_lock_irqsave(&dev_priv->uncore.lock, irqflags); @@ -664,31 +629,20 @@ ivb_disable_plane(struct drm_plane *plane, struct drm_crtc *crtc) spin_unlock_irqrestore(&dev_priv->uncore.lock, irqflags); } -static void -ilk_update_plane(struct drm_plane *plane, - const struct intel_crtc_state *crtc_state, - const struct intel_plane_state *plane_state) +static u32 ilk_sprite_ctl(const struct intel_crtc_state *crtc_state, + const struct intel_plane_state *plane_state) { - struct drm_device *dev = plane->dev; - struct drm_i915_private *dev_priv = to_i915(dev); - struct intel_plane *intel_plane = to_intel_plane(plane); - struct drm_framebuffer *fb = plane_state->base.fb; - int pipe = intel_plane->pipe; - u32 dvscntr, dvsscale; - u32 dvssurf_offset, linear_offset; + struct drm_i915_private *dev_priv = + to_i915(plane_state->base.plane->dev); + const struct drm_framebuffer *fb = plane_state->base.fb; unsigned int rotation = plane_state->base.rotation; const struct drm_intel_sprite_colorkey *key = &plane_state->ckey; - int crtc_x = plane_state->base.dst.x1; - int crtc_y = plane_state->base.dst.y1; - uint32_t crtc_w = drm_rect_width(&plane_state->base.dst); - uint32_t crtc_h = drm_rect_height(&plane_state->base.dst); - uint32_t x = plane_state->base.src.x1 >> 16; - uint32_t y = plane_state->base.src.y1 >> 16; - uint32_t src_w = drm_rect_width(&plane_state->base.src) >> 16; - uint32_t src_h = drm_rect_height(&plane_state->base.src) >> 16; - unsigned long irqflags; + u32 dvscntr; - dvscntr = DVS_ENABLE; + dvscntr = DVS_ENABLE | DVS_GAMMA_ENABLE; + + if (IS_GEN6(dev_priv)) + dvscntr |= DVS_TRICKLE_FEED_DISABLE; switch (fb->format->format) { case DRM_FORMAT_XBGR8888: @@ -710,47 +664,57 @@ ilk_update_plane(struct drm_plane *plane, dvscntr |= DVS_FORMAT_YUV422 | DVS_YUV_ORDER_VYUY; break; default: - BUG(); + MISSING_CASE(fb->format->format); + return 0; } - /* - * Enable gamma to match primary/cursor plane behaviour. - * FIXME should be user controllable via propertiesa. - */ - dvscntr |= DVS_GAMMA_ENABLE; - if (fb->modifier == I915_FORMAT_MOD_X_TILED) dvscntr |= DVS_TILED; if (rotation & DRM_ROTATE_180) dvscntr |= DVS_ROTATE_180; - if (IS_GEN6(dev_priv)) - dvscntr |= DVS_TRICKLE_FEED_DISABLE; /* must disable */ - if (key->flags & I915_SET_COLORKEY_DESTINATION) dvscntr |= DVS_DEST_KEY; else if (key->flags & I915_SET_COLORKEY_SOURCE) dvscntr |= DVS_SOURCE_KEY; + return dvscntr; +} + +static void +ilk_update_plane(struct drm_plane *plane, + const struct intel_crtc_state *crtc_state, + const struct intel_plane_state *plane_state) +{ + struct drm_device *dev = plane->dev; + struct drm_i915_private *dev_priv = to_i915(dev); + struct intel_plane *intel_plane = to_intel_plane(plane); + struct drm_framebuffer *fb = plane_state->base.fb; + int pipe = intel_plane->pipe; + u32 dvscntr = plane_state->ctl, dvsscale = 0; + u32 dvssurf_offset = plane_state->main.offset; + u32 linear_offset; + const struct drm_intel_sprite_colorkey *key = &plane_state->ckey; + int crtc_x = plane_state->base.dst.x1; + int crtc_y = plane_state->base.dst.y1; + uint32_t crtc_w = drm_rect_width(&plane_state->base.dst); + uint32_t crtc_h = drm_rect_height(&plane_state->base.dst); + uint32_t x = plane_state->main.x; + uint32_t y = plane_state->main.y; + uint32_t src_w = drm_rect_width(&plane_state->base.src) >> 16; + uint32_t src_h = drm_rect_height(&plane_state->base.src) >> 16; + unsigned long irqflags; + /* Sizes are 0 based */ src_w--; src_h--; crtc_w--; crtc_h--; - dvsscale = 0; if (crtc_w != src_w || crtc_h != src_h) dvsscale = DVS_SCALE_ENABLE | (src_w << 16) | src_h; - intel_add_fb_offsets(&x, &y, plane_state, 0); - dvssurf_offset = intel_compute_tile_offset(&x, &y, plane_state, 0); - - if (rotation & DRM_ROTATE_180) { - x += src_w; - y += src_h; - } - linear_offset = intel_fb_xy_to_linear(x, y, plane_state, 0); spin_lock_irqsave(&dev_priv->uncore.lock, irqflags); @@ -981,6 +945,26 @@ intel_check_sprite_plane(struct drm_plane *plane, ret = skl_check_plane_surface(state); if (ret) return ret; + + state->ctl = skl_plane_ctl(crtc_state, state); + } else if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv)) { + ret = i9xx_check_plane_surface(state); + if (ret) + return ret; + + state->ctl = vlv_sprite_ctl(crtc_state, state); + } else if (INTEL_GEN(dev_priv) >= 7) { + ret = i9xx_check_plane_surface(state); + if (ret) + return ret; + + state->ctl = ivb_sprite_ctl(crtc_state, state); + } else { + ret = i9xx_check_plane_surface(state); + if (ret) + return ret; + + state->ctl = ilk_sprite_ctl(crtc_state, state); } return 0; diff --git a/drivers/gpu/drm/i915/intel_uc.c b/drivers/gpu/drm/i915/intel_uc.c index d15a7d9d4eb0..c117424f1f50 100644 --- a/drivers/gpu/drm/i915/intel_uc.c +++ b/drivers/gpu/drm/i915/intel_uc.c @@ -26,6 +26,19 @@ #include "intel_uc.h" #include <linux/firmware.h> +/* Cleans up uC firmware by releasing the firmware GEM obj. + */ +static void __intel_uc_fw_fini(struct intel_uc_fw *uc_fw) +{ + struct drm_i915_gem_object *obj; + + obj = fetch_and_zero(&uc_fw->obj); + if (obj) + i915_gem_object_put(obj); + + uc_fw->fetch_status = INTEL_UC_FIRMWARE_NONE; +} + /* Reset GuC providing us with fresh state for both GuC and HuC. */ static int __intel_uc_reset_hw(struct drm_i915_private *dev_priv) @@ -83,23 +96,166 @@ void intel_uc_sanitize_options(struct drm_i915_private *dev_priv) void intel_uc_init_early(struct drm_i915_private *dev_priv) { - mutex_init(&dev_priv->guc.send_mutex); + struct intel_guc *guc = &dev_priv->guc; + + mutex_init(&guc->send_mutex); + guc->send = intel_guc_send_mmio; +} + +static void fetch_uc_fw(struct drm_i915_private *dev_priv, + struct intel_uc_fw *uc_fw) +{ + struct pci_dev *pdev = dev_priv->drm.pdev; + struct drm_i915_gem_object *obj; + const struct firmware *fw = NULL; + struct uc_css_header *css; + size_t size; + int err; + + if (!uc_fw->path) + return; + + uc_fw->fetch_status = INTEL_UC_FIRMWARE_PENDING; + + DRM_DEBUG_DRIVER("before requesting firmware: uC fw fetch status %s\n", + intel_uc_fw_status_repr(uc_fw->fetch_status)); + + err = request_firmware(&fw, uc_fw->path, &pdev->dev); + if (err) + goto fail; + if (!fw) + goto fail; + + DRM_DEBUG_DRIVER("fetch uC fw from %s succeeded, fw %p\n", + uc_fw->path, fw); + + /* Check the size of the blob before examining buffer contents */ + if (fw->size < sizeof(struct uc_css_header)) { + DRM_NOTE("Firmware header is missing\n"); + goto fail; + } + + css = (struct uc_css_header *)fw->data; + + /* Firmware bits always start from header */ + uc_fw->header_offset = 0; + uc_fw->header_size = (css->header_size_dw - css->modulus_size_dw - + css->key_size_dw - css->exponent_size_dw) * sizeof(u32); + + if (uc_fw->header_size != sizeof(struct uc_css_header)) { + DRM_NOTE("CSS header definition mismatch\n"); + goto fail; + } + + /* then, uCode */ + uc_fw->ucode_offset = uc_fw->header_offset + uc_fw->header_size; + uc_fw->ucode_size = (css->size_dw - css->header_size_dw) * sizeof(u32); + + /* now RSA */ + if (css->key_size_dw != UOS_RSA_SCRATCH_MAX_COUNT) { + DRM_NOTE("RSA key size is bad\n"); + goto fail; + } + uc_fw->rsa_offset = uc_fw->ucode_offset + uc_fw->ucode_size; + uc_fw->rsa_size = css->key_size_dw * sizeof(u32); + + /* At least, it should have header, uCode and RSA. Size of all three. */ + size = uc_fw->header_size + uc_fw->ucode_size + uc_fw->rsa_size; + if (fw->size < size) { + DRM_NOTE("Missing firmware components\n"); + goto fail; + } + + /* + * The GuC firmware image has the version number embedded at a + * well-known offset within the firmware blob; note that major / minor + * version are TWO bytes each (i.e. u16), although all pointers and + * offsets are defined in terms of bytes (u8). + */ + switch (uc_fw->type) { + case INTEL_UC_FW_TYPE_GUC: + /* Header and uCode will be loaded to WOPCM. Size of the two. */ + size = uc_fw->header_size + uc_fw->ucode_size; + + /* Top 32k of WOPCM is reserved (8K stack + 24k RC6 context). */ + if (size > intel_guc_wopcm_size(dev_priv)) { + DRM_ERROR("Firmware is too large to fit in WOPCM\n"); + goto fail; + } + uc_fw->major_ver_found = css->guc.sw_version >> 16; + uc_fw->minor_ver_found = css->guc.sw_version & 0xFFFF; + break; + + case INTEL_UC_FW_TYPE_HUC: + uc_fw->major_ver_found = css->huc.sw_version >> 16; + uc_fw->minor_ver_found = css->huc.sw_version & 0xFFFF; + break; + + default: + DRM_ERROR("Unknown firmware type %d\n", uc_fw->type); + err = -ENOEXEC; + goto fail; + } + + if (uc_fw->major_ver_wanted == 0 && uc_fw->minor_ver_wanted == 0) { + DRM_NOTE("Skipping %s firmware version check\n", + intel_uc_fw_type_repr(uc_fw->type)); + } else if (uc_fw->major_ver_found != uc_fw->major_ver_wanted || + uc_fw->minor_ver_found < uc_fw->minor_ver_wanted) { + DRM_NOTE("%s firmware version %d.%d, required %d.%d\n", + intel_uc_fw_type_repr(uc_fw->type), + uc_fw->major_ver_found, uc_fw->minor_ver_found, + uc_fw->major_ver_wanted, uc_fw->minor_ver_wanted); + err = -ENOEXEC; + goto fail; + } + + DRM_DEBUG_DRIVER("firmware version %d.%d OK (minimum %d.%d)\n", + uc_fw->major_ver_found, uc_fw->minor_ver_found, + uc_fw->major_ver_wanted, uc_fw->minor_ver_wanted); + + obj = i915_gem_object_create_from_data(dev_priv, fw->data, fw->size); + if (IS_ERR(obj)) { + err = PTR_ERR(obj); + goto fail; + } + + uc_fw->obj = obj; + uc_fw->size = fw->size; + + DRM_DEBUG_DRIVER("uC fw fetch status SUCCESS, obj %p\n", + uc_fw->obj); + + release_firmware(fw); + uc_fw->fetch_status = INTEL_UC_FIRMWARE_SUCCESS; + return; + +fail: + DRM_WARN("Failed to fetch valid uC firmware from %s (error %d)\n", + uc_fw->path, err); + DRM_DEBUG_DRIVER("uC fw fetch status FAIL; err %d, fw %p, obj %p\n", + err, fw, uc_fw->obj); + + release_firmware(fw); /* OK even if fw is NULL */ + uc_fw->fetch_status = INTEL_UC_FIRMWARE_FAIL; } void intel_uc_init_fw(struct drm_i915_private *dev_priv) { - if (dev_priv->huc.fw.path) - intel_uc_prepare_fw(dev_priv, &dev_priv->huc.fw); + fetch_uc_fw(dev_priv, &dev_priv->huc.fw); + fetch_uc_fw(dev_priv, &dev_priv->guc.fw); +} - if (dev_priv->guc.fw.path) - intel_uc_prepare_fw(dev_priv, &dev_priv->guc.fw); +void intel_uc_fini_fw(struct drm_i915_private *dev_priv) +{ + __intel_uc_fw_fini(&dev_priv->guc.fw); + __intel_uc_fw_fini(&dev_priv->huc.fw); } int intel_uc_init_hw(struct drm_i915_private *dev_priv) { int ret, attempts; - /* GuC not enabled, nothing to do */ if (!i915.enable_guc_loading) return 0; @@ -109,9 +265,13 @@ int intel_uc_init_hw(struct drm_i915_private *dev_priv) i915_ggtt_enable_guc(dev_priv); if (i915.enable_guc_submission) { + /* + * This is stuff we need to have available at fw load time + * if we are planning to enable submission later + */ ret = i915_guc_submission_init(dev_priv); if (ret) - goto err; + goto err_guc; } /* WaEnableuKernelHeaderValidFix:skl */ @@ -150,7 +310,7 @@ int intel_uc_init_hw(struct drm_i915_private *dev_priv) ret = i915_guc_submission_enable(dev_priv); if (ret) - goto err_submission; + goto err_interrupts; } return 0; @@ -164,11 +324,12 @@ int intel_uc_init_hw(struct drm_i915_private *dev_priv) * nonfatal error (i.e. it doesn't prevent driver load, but * marks the GPU as wedged until reset). */ +err_interrupts: + gen9_disable_guc_interrupts(dev_priv); err_submission: if (i915.enable_guc_submission) i915_guc_submission_fini(dev_priv); - -err: +err_guc: i915_ggtt_disable_guc(dev_priv); DRM_ERROR("GuC init failed\n"); @@ -185,11 +346,24 @@ err: return ret; } +void intel_uc_fini_hw(struct drm_i915_private *dev_priv) +{ + if (!i915.enable_guc_loading) + return; + + if (i915.enable_guc_submission) { + i915_guc_submission_disable(dev_priv); + gen9_disable_guc_interrupts(dev_priv); + i915_guc_submission_fini(dev_priv); + } + i915_ggtt_disable_guc(dev_priv); +} + /* * Read GuC command/status register (SOFT_SCRATCH_0) * Return true if it contains a response rather than a command */ -static bool intel_guc_recv(struct intel_guc *guc, u32 *status) +static bool guc_recv(struct intel_guc *guc, u32 *status) { struct drm_i915_private *dev_priv = guc_to_i915(guc); @@ -198,7 +372,10 @@ static bool intel_guc_recv(struct intel_guc *guc, u32 *status) return INTEL_GUC_RECV_IS_RESPONSE(val); } -int intel_guc_send(struct intel_guc *guc, const u32 *action, u32 len) +/* + * This function implements the MMIO based host to GuC interface. + */ +int intel_guc_send_mmio(struct intel_guc *guc, const u32 *action, u32 len) { struct drm_i915_private *dev_priv = guc_to_i915(guc); u32 status; @@ -209,7 +386,7 @@ int intel_guc_send(struct intel_guc *guc, const u32 *action, u32 len) return -EINVAL; mutex_lock(&guc->send_mutex); - intel_uncore_forcewake_get(dev_priv, FORCEWAKE_ALL); + intel_uncore_forcewake_get(dev_priv, FORCEWAKE_BLITTER); dev_priv->guc.action_count += 1; dev_priv->guc.action_cmd = action[0]; @@ -226,9 +403,9 @@ int intel_guc_send(struct intel_guc *guc, const u32 *action, u32 len) * up to that length of time, then switch to a slower sleep-wait loop. * No inte_guc_send command should ever take longer than 10ms. */ - ret = wait_for_us(intel_guc_recv(guc, &status), 10); + ret = wait_for_us(guc_recv(guc, &status), 10); if (ret) - ret = wait_for(intel_guc_recv(guc, &status), 10); + ret = wait_for(guc_recv(guc, &status), 10); if (status != INTEL_GUC_STATUS_SUCCESS) { /* * Either the GuC explicitly returned an error (which @@ -247,7 +424,7 @@ int intel_guc_send(struct intel_guc *guc, const u32 *action, u32 len) } dev_priv->guc.action_status = status; - intel_uncore_forcewake_put(dev_priv, FORCEWAKE_ALL); + intel_uncore_forcewake_put(dev_priv, FORCEWAKE_BLITTER); mutex_unlock(&guc->send_mutex); return ret; @@ -268,136 +445,3 @@ int intel_guc_sample_forcewake(struct intel_guc *guc) return intel_guc_send(guc, action, ARRAY_SIZE(action)); } - -void intel_uc_prepare_fw(struct drm_i915_private *dev_priv, - struct intel_uc_fw *uc_fw) -{ - struct pci_dev *pdev = dev_priv->drm.pdev; - struct drm_i915_gem_object *obj; - const struct firmware *fw = NULL; - struct uc_css_header *css; - size_t size; - int err; - - uc_fw->fetch_status = INTEL_UC_FIRMWARE_PENDING; - - DRM_DEBUG_DRIVER("before requesting firmware: uC fw fetch status %s\n", - intel_uc_fw_status_repr(uc_fw->fetch_status)); - - err = request_firmware(&fw, uc_fw->path, &pdev->dev); - if (err) - goto fail; - if (!fw) - goto fail; - - DRM_DEBUG_DRIVER("fetch uC fw from %s succeeded, fw %p\n", - uc_fw->path, fw); - - /* Check the size of the blob before examining buffer contents */ - if (fw->size < sizeof(struct uc_css_header)) { - DRM_NOTE("Firmware header is missing\n"); - goto fail; - } - - css = (struct uc_css_header *)fw->data; - - /* Firmware bits always start from header */ - uc_fw->header_offset = 0; - uc_fw->header_size = (css->header_size_dw - css->modulus_size_dw - - css->key_size_dw - css->exponent_size_dw) * sizeof(u32); - - if (uc_fw->header_size != sizeof(struct uc_css_header)) { - DRM_NOTE("CSS header definition mismatch\n"); - goto fail; - } - - /* then, uCode */ - uc_fw->ucode_offset = uc_fw->header_offset + uc_fw->header_size; - uc_fw->ucode_size = (css->size_dw - css->header_size_dw) * sizeof(u32); - - /* now RSA */ - if (css->key_size_dw != UOS_RSA_SCRATCH_MAX_COUNT) { - DRM_NOTE("RSA key size is bad\n"); - goto fail; - } - uc_fw->rsa_offset = uc_fw->ucode_offset + uc_fw->ucode_size; - uc_fw->rsa_size = css->key_size_dw * sizeof(u32); - - /* At least, it should have header, uCode and RSA. Size of all three. */ - size = uc_fw->header_size + uc_fw->ucode_size + uc_fw->rsa_size; - if (fw->size < size) { - DRM_NOTE("Missing firmware components\n"); - goto fail; - } - - /* - * The GuC firmware image has the version number embedded at a - * well-known offset within the firmware blob; note that major / minor - * version are TWO bytes each (i.e. u16), although all pointers and - * offsets are defined in terms of bytes (u8). - */ - switch (uc_fw->type) { - case INTEL_UC_FW_TYPE_GUC: - /* Header and uCode will be loaded to WOPCM. Size of the two. */ - size = uc_fw->header_size + uc_fw->ucode_size; - - /* Top 32k of WOPCM is reserved (8K stack + 24k RC6 context). */ - if (size > intel_guc_wopcm_size(dev_priv)) { - DRM_ERROR("Firmware is too large to fit in WOPCM\n"); - goto fail; - } - uc_fw->major_ver_found = css->guc.sw_version >> 16; - uc_fw->minor_ver_found = css->guc.sw_version & 0xFFFF; - break; - - case INTEL_UC_FW_TYPE_HUC: - uc_fw->major_ver_found = css->huc.sw_version >> 16; - uc_fw->minor_ver_found = css->huc.sw_version & 0xFFFF; - break; - - default: - DRM_ERROR("Unknown firmware type %d\n", uc_fw->type); - err = -ENOEXEC; - goto fail; - } - - if (uc_fw->major_ver_wanted == 0 && uc_fw->minor_ver_wanted == 0) { - DRM_NOTE("Skipping uC firmware version check\n"); - } else if (uc_fw->major_ver_found != uc_fw->major_ver_wanted || - uc_fw->minor_ver_found < uc_fw->minor_ver_wanted) { - DRM_NOTE("uC firmware version %d.%d, required %d.%d\n", - uc_fw->major_ver_found, uc_fw->minor_ver_found, - uc_fw->major_ver_wanted, uc_fw->minor_ver_wanted); - err = -ENOEXEC; - goto fail; - } - - DRM_DEBUG_DRIVER("firmware version %d.%d OK (minimum %d.%d)\n", - uc_fw->major_ver_found, uc_fw->minor_ver_found, - uc_fw->major_ver_wanted, uc_fw->minor_ver_wanted); - - obj = i915_gem_object_create_from_data(dev_priv, fw->data, fw->size); - if (IS_ERR(obj)) { - err = PTR_ERR(obj); - goto fail; - } - - uc_fw->obj = obj; - uc_fw->size = fw->size; - - DRM_DEBUG_DRIVER("uC fw fetch status SUCCESS, obj %p\n", - uc_fw->obj); - - release_firmware(fw); - uc_fw->fetch_status = INTEL_UC_FIRMWARE_SUCCESS; - return; - -fail: - DRM_WARN("Failed to fetch valid uC firmware from %s (error %d)\n", - uc_fw->path, err); - DRM_DEBUG_DRIVER("uC fw fetch status FAIL; err %d, fw %p, obj %p\n", - err, fw, uc_fw->obj); - - release_firmware(fw); /* OK even if fw is NULL */ - uc_fw->fetch_status = INTEL_UC_FIRMWARE_FAIL; -} diff --git a/drivers/gpu/drm/i915/intel_uc.h b/drivers/gpu/drm/i915/intel_uc.h index a35ededfaa40..4b7f73aeddac 100644 --- a/drivers/gpu/drm/i915/intel_uc.h +++ b/drivers/gpu/drm/i915/intel_uc.h @@ -34,7 +34,9 @@ struct drm_i915_gem_request; /* * This structure primarily describes the GEM object shared with the GuC. - * The GEM object is held for the entire lifetime of our interaction with + * The specs sometimes refer to this object as a "GuC context", but we use + * the term "client" to avoid confusion with hardware contexts. This + * GEM object is held for the entire lifetime of our interaction with * the GuC, being allocated before the GuC is loaded with its firmware. * Because there's no way to update the address used by the GuC after * initialisation, the shared object must stay pinned into the GGTT as @@ -44,7 +46,7 @@ struct drm_i915_gem_request; * * The single GEM object described here is actually made up of several * separate areas, as far as the GuC is concerned. The first page (kept - * kmap'd) includes the "process decriptor" which holds sequence data for + * kmap'd) includes the "process descriptor" which holds sequence data for * the doorbell, and one cacheline which actually *is* the doorbell; a * write to this will "ring the doorbell" (i.e. send an interrupt to the * GuC). The subsequent pages of the client object constitute the work @@ -72,13 +74,12 @@ struct i915_guc_client { uint32_t engines; /* bitmap of (host) engine ids */ uint32_t priority; - uint32_t ctx_index; + u32 stage_id; uint32_t proc_desc_offset; - uint32_t doorbell_offset; - uint32_t doorbell_cookie; - uint16_t doorbell_id; - uint16_t padding[3]; /* Maintain alignment */ + u16 doorbell_id; + unsigned long doorbell_offset; + u32 doorbell_cookie; spinlock_t wq_lock; uint32_t wq_offset; @@ -100,11 +101,40 @@ enum intel_uc_fw_status { INTEL_UC_FIRMWARE_SUCCESS }; +/* User-friendly representation of an enum */ +static inline +const char *intel_uc_fw_status_repr(enum intel_uc_fw_status status) +{ + switch (status) { + case INTEL_UC_FIRMWARE_FAIL: + return "FAIL"; + case INTEL_UC_FIRMWARE_NONE: + return "NONE"; + case INTEL_UC_FIRMWARE_PENDING: + return "PENDING"; + case INTEL_UC_FIRMWARE_SUCCESS: + return "SUCCESS"; + } + return "<invalid>"; +} + enum intel_uc_fw_type { INTEL_UC_FW_TYPE_GUC, INTEL_UC_FW_TYPE_HUC }; +/* User-friendly representation of an enum */ +static inline const char *intel_uc_fw_type_repr(enum intel_uc_fw_type type) +{ + switch (type) { + case INTEL_UC_FW_TYPE_GUC: + return "GuC"; + case INTEL_UC_FW_TYPE_HUC: + return "HuC"; + } + return "uC"; +} + /* * This structure encapsulates all the data needed during the process * of fetching, caching, and loading the firmware image into the GuC. @@ -133,11 +163,13 @@ struct intel_uc_fw { struct intel_guc_log { uint32_t flags; struct i915_vma *vma; - void *buf_addr; - struct workqueue_struct *flush_wq; - struct work_struct flush_work; - struct rchan *relay_chan; - + /* The runtime stuff gets created only when GuC logging gets enabled */ + struct { + void *buf_addr; + struct workqueue_struct *flush_wq; + struct work_struct flush_work; + struct rchan *relay_chan; + } runtime; /* logging related stats */ u32 capture_miss_count; u32 flush_interrupt_count; @@ -154,12 +186,13 @@ struct intel_guc { bool interrupts_enabled; struct i915_vma *ads_vma; - struct i915_vma *ctx_pool_vma; - struct ida ctx_ids; + struct i915_vma *stage_desc_pool; + void *stage_desc_pool_vaddr; + struct ida stage_ids; struct i915_guc_client *execbuf_client; - DECLARE_BITMAP(doorbell_bitmap, GUC_MAX_DOORBELLS); + DECLARE_BITMAP(doorbell_bitmap, GUC_NUM_DOORBELLS); uint32_t db_cacheline; /* Cyclic counter mod pagesize */ /* Action status & statistics */ @@ -174,6 +207,9 @@ struct intel_guc { /* To serialize the intel_guc_send actions */ struct mutex send_mutex; + + /* GuC's FW specific send function */ + int (*send)(struct intel_guc *guc, const u32 *data, u32 len); }; struct intel_huc { @@ -187,17 +223,19 @@ struct intel_huc { void intel_uc_sanitize_options(struct drm_i915_private *dev_priv); void intel_uc_init_early(struct drm_i915_private *dev_priv); void intel_uc_init_fw(struct drm_i915_private *dev_priv); +void intel_uc_fini_fw(struct drm_i915_private *dev_priv); int intel_uc_init_hw(struct drm_i915_private *dev_priv); -void intel_uc_prepare_fw(struct drm_i915_private *dev_priv, - struct intel_uc_fw *uc_fw); -int intel_guc_send(struct intel_guc *guc, const u32 *action, u32 len); +void intel_uc_fini_hw(struct drm_i915_private *dev_priv); int intel_guc_sample_forcewake(struct intel_guc *guc); +int intel_guc_send_mmio(struct intel_guc *guc, const u32 *action, u32 len); +static inline int intel_guc_send(struct intel_guc *guc, const u32 *action, u32 len) +{ + return guc->send(guc, action, len); +} /* intel_guc_loader.c */ int intel_guc_select_fw(struct intel_guc *guc); int intel_guc_init_hw(struct intel_guc *guc); -void intel_guc_fini(struct drm_i915_private *dev_priv); -const char *intel_uc_fw_status_repr(enum intel_uc_fw_status status); int intel_guc_suspend(struct drm_i915_private *dev_priv); int intel_guc_resume(struct drm_i915_private *dev_priv); u32 intel_guc_wopcm_size(struct drm_i915_private *dev_priv); @@ -212,10 +250,11 @@ void i915_guc_submission_fini(struct drm_i915_private *dev_priv); struct i915_vma *intel_guc_allocate_vma(struct intel_guc *guc, u32 size); /* intel_guc_log.c */ -void intel_guc_log_create(struct intel_guc *guc); +int intel_guc_log_create(struct intel_guc *guc); +void intel_guc_log_destroy(struct intel_guc *guc); +int i915_guc_log_control(struct drm_i915_private *dev_priv, u64 control_val); void i915_guc_log_register(struct drm_i915_private *dev_priv); void i915_guc_log_unregister(struct drm_i915_private *dev_priv); -int i915_guc_log_control(struct drm_i915_private *dev_priv, u64 control_val); static inline u32 guc_ggtt_offset(struct i915_vma *vma) { @@ -227,7 +266,6 @@ static inline u32 guc_ggtt_offset(struct i915_vma *vma) /* intel_huc.c */ void intel_huc_select_fw(struct intel_huc *huc); -void intel_huc_fini(struct drm_i915_private *dev_priv); int intel_huc_init_hw(struct intel_huc *huc); void intel_guc_auth_huc(struct drm_i915_private *dev_priv); diff --git a/drivers/gpu/drm/i915/intel_uncore.c b/drivers/gpu/drm/i915/intel_uncore.c index 09f5f02d7901..6d1ea26b2493 100644 --- a/drivers/gpu/drm/i915/intel_uncore.c +++ b/drivers/gpu/drm/i915/intel_uncore.c @@ -52,10 +52,10 @@ intel_uncore_forcewake_domain_to_str(const enum forcewake_domain_id id) } static inline void -fw_domain_reset(const struct intel_uncore_forcewake_domain *d) +fw_domain_reset(struct drm_i915_private *i915, + const struct intel_uncore_forcewake_domain *d) { - WARN_ON(!i915_mmio_reg_valid(d->reg_set)); - __raw_i915_write32(d->i915, d->reg_set, d->val_reset); + __raw_i915_write32(i915, d->reg_set, i915->uncore.fw_reset); } static inline void @@ -69,9 +69,10 @@ fw_domain_arm_timer(struct intel_uncore_forcewake_domain *d) } static inline void -fw_domain_wait_ack_clear(const struct intel_uncore_forcewake_domain *d) +fw_domain_wait_ack_clear(const struct drm_i915_private *i915, + const struct intel_uncore_forcewake_domain *d) { - if (wait_for_atomic((__raw_i915_read32(d->i915, d->reg_ack) & + if (wait_for_atomic((__raw_i915_read32(i915, d->reg_ack) & FORCEWAKE_KERNEL) == 0, FORCEWAKE_ACK_TIMEOUT_MS)) DRM_ERROR("%s: timed out waiting for forcewake ack to clear.\n", @@ -79,15 +80,17 @@ fw_domain_wait_ack_clear(const struct intel_uncore_forcewake_domain *d) } static inline void -fw_domain_get(const struct intel_uncore_forcewake_domain *d) +fw_domain_get(struct drm_i915_private *i915, + const struct intel_uncore_forcewake_domain *d) { - __raw_i915_write32(d->i915, d->reg_set, d->val_set); + __raw_i915_write32(i915, d->reg_set, i915->uncore.fw_set); } static inline void -fw_domain_wait_ack(const struct intel_uncore_forcewake_domain *d) +fw_domain_wait_ack(const struct drm_i915_private *i915, + const struct intel_uncore_forcewake_domain *d) { - if (wait_for_atomic((__raw_i915_read32(d->i915, d->reg_ack) & + if (wait_for_atomic((__raw_i915_read32(i915, d->reg_ack) & FORCEWAKE_KERNEL), FORCEWAKE_ACK_TIMEOUT_MS)) DRM_ERROR("%s: timed out waiting for forcewake ack request.\n", @@ -95,72 +98,59 @@ fw_domain_wait_ack(const struct intel_uncore_forcewake_domain *d) } static inline void -fw_domain_put(const struct intel_uncore_forcewake_domain *d) +fw_domain_put(const struct drm_i915_private *i915, + const struct intel_uncore_forcewake_domain *d) { - __raw_i915_write32(d->i915, d->reg_set, d->val_clear); -} - -static inline void -fw_domain_posting_read(const struct intel_uncore_forcewake_domain *d) -{ - /* something from same cacheline, but not from the set register */ - if (i915_mmio_reg_valid(d->reg_post)) - __raw_posting_read(d->i915, d->reg_post); + __raw_i915_write32(i915, d->reg_set, i915->uncore.fw_clear); } static void -fw_domains_get(struct drm_i915_private *dev_priv, enum forcewake_domains fw_domains) +fw_domains_get(struct drm_i915_private *i915, enum forcewake_domains fw_domains) { struct intel_uncore_forcewake_domain *d; + unsigned int tmp; - for_each_fw_domain_masked(d, fw_domains, dev_priv) { - fw_domain_wait_ack_clear(d); - fw_domain_get(d); + GEM_BUG_ON(fw_domains & ~i915->uncore.fw_domains); + + for_each_fw_domain_masked(d, fw_domains, i915, tmp) { + fw_domain_wait_ack_clear(i915, d); + fw_domain_get(i915, d); } - for_each_fw_domain_masked(d, fw_domains, dev_priv) - fw_domain_wait_ack(d); + for_each_fw_domain_masked(d, fw_domains, i915, tmp) + fw_domain_wait_ack(i915, d); - dev_priv->uncore.fw_domains_active |= fw_domains; + i915->uncore.fw_domains_active |= fw_domains; } static void -fw_domains_put(struct drm_i915_private *dev_priv, enum forcewake_domains fw_domains) +fw_domains_put(struct drm_i915_private *i915, enum forcewake_domains fw_domains) { struct intel_uncore_forcewake_domain *d; + unsigned int tmp; - for_each_fw_domain_masked(d, fw_domains, dev_priv) { - fw_domain_put(d); - fw_domain_posting_read(d); - } - - dev_priv->uncore.fw_domains_active &= ~fw_domains; -} + GEM_BUG_ON(fw_domains & ~i915->uncore.fw_domains); -static void -fw_domains_posting_read(struct drm_i915_private *dev_priv) -{ - struct intel_uncore_forcewake_domain *d; + for_each_fw_domain_masked(d, fw_domains, i915, tmp) + fw_domain_put(i915, d); - /* No need to do for all, just do for first found */ - for_each_fw_domain(d, dev_priv) { - fw_domain_posting_read(d); - break; - } + i915->uncore.fw_domains_active &= ~fw_domains; } static void -fw_domains_reset(struct drm_i915_private *dev_priv, enum forcewake_domains fw_domains) +fw_domains_reset(struct drm_i915_private *i915, + enum forcewake_domains fw_domains) { struct intel_uncore_forcewake_domain *d; + unsigned int tmp; - if (dev_priv->uncore.fw_domains == 0) + if (!fw_domains) return; - for_each_fw_domain_masked(d, fw_domains, dev_priv) - fw_domain_reset(d); + GEM_BUG_ON(fw_domains & ~i915->uncore.fw_domains); - fw_domains_posting_read(dev_priv); + for_each_fw_domain_masked(d, fw_domains, i915, tmp) + fw_domain_reset(i915, d); } static void __gen6_gt_wait_for_thread_c0(struct drm_i915_private *dev_priv) @@ -236,7 +226,8 @@ intel_uncore_fw_release_timer(struct hrtimer *timer) { struct intel_uncore_forcewake_domain *domain = container_of(timer, struct intel_uncore_forcewake_domain, timer); - struct drm_i915_private *dev_priv = domain->i915; + struct drm_i915_private *dev_priv = + container_of(domain, struct drm_i915_private, uncore.fw_domain[domain->id]); unsigned long irqflags; assert_rpm_device_not_suspended(dev_priv); @@ -266,9 +257,11 @@ static void intel_uncore_forcewake_reset(struct drm_i915_private *dev_priv, * timers are run before holding. */ while (1) { + unsigned int tmp; + active_domains = 0; - for_each_fw_domain(domain, dev_priv) { + for_each_fw_domain(domain, dev_priv, tmp) { if (hrtimer_cancel(&domain->timer) == 0) continue; @@ -277,7 +270,7 @@ static void intel_uncore_forcewake_reset(struct drm_i915_private *dev_priv, spin_lock_irqsave(&dev_priv->uncore.lock, irqflags); - for_each_fw_domain(domain, dev_priv) { + for_each_fw_domain(domain, dev_priv, tmp) { if (hrtimer_active(&domain->timer)) active_domains |= domain->mask; } @@ -300,7 +293,7 @@ static void intel_uncore_forcewake_reset(struct drm_i915_private *dev_priv, if (fw) dev_priv->uncore.funcs.force_wake_put(dev_priv, fw); - fw_domains_reset(dev_priv, FORCEWAKE_ALL); + fw_domains_reset(dev_priv, dev_priv->uncore.fw_domains); if (restore) { /* If reset with a user forcewake, try to restore */ if (fw) @@ -457,13 +450,13 @@ static void __intel_uncore_forcewake_get(struct drm_i915_private *dev_priv, enum forcewake_domains fw_domains) { struct intel_uncore_forcewake_domain *domain; + unsigned int tmp; fw_domains &= dev_priv->uncore.fw_domains; - for_each_fw_domain_masked(domain, fw_domains, dev_priv) { + for_each_fw_domain_masked(domain, fw_domains, dev_priv, tmp) if (domain->wake_count++) fw_domains &= ~domain->mask; - } if (fw_domains) dev_priv->uncore.funcs.force_wake_get(dev_priv, fw_domains); @@ -520,10 +513,11 @@ static void __intel_uncore_forcewake_put(struct drm_i915_private *dev_priv, enum forcewake_domains fw_domains) { struct intel_uncore_forcewake_domain *domain; + unsigned int tmp; fw_domains &= dev_priv->uncore.fw_domains; - for_each_fw_domain_masked(domain, fw_domains, dev_priv) { + for_each_fw_domain_masked(domain, fw_domains, dev_priv, tmp) { if (WARN_ON(domain->wake_count == 0)) continue; @@ -928,8 +922,11 @@ static noinline void ___force_wake_auto(struct drm_i915_private *dev_priv, enum forcewake_domains fw_domains) { struct intel_uncore_forcewake_domain *domain; + unsigned int tmp; + + GEM_BUG_ON(fw_domains & ~dev_priv->uncore.fw_domains); - for_each_fw_domain_masked(domain, fw_domains, dev_priv) + for_each_fw_domain_masked(domain, fw_domains, dev_priv, tmp) fw_domain_arm_timer(domain); dev_priv->uncore.funcs.force_wake_get(dev_priv, fw_domains); @@ -1141,41 +1138,27 @@ static void fw_domain_init(struct drm_i915_private *dev_priv, WARN_ON(d->wake_count); + WARN_ON(!i915_mmio_reg_valid(reg_set)); + WARN_ON(!i915_mmio_reg_valid(reg_ack)); + d->wake_count = 0; d->reg_set = reg_set; d->reg_ack = reg_ack; - if (IS_GEN6(dev_priv)) { - d->val_reset = 0; - d->val_set = FORCEWAKE_KERNEL; - d->val_clear = 0; - } else { - /* WaRsClearFWBitsAtReset:bdw,skl */ - d->val_reset = _MASKED_BIT_DISABLE(0xffff); - d->val_set = _MASKED_BIT_ENABLE(FORCEWAKE_KERNEL); - d->val_clear = _MASKED_BIT_DISABLE(FORCEWAKE_KERNEL); - } - - if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv)) - d->reg_post = FORCEWAKE_ACK_VLV; - else if (IS_GEN6(dev_priv) || IS_GEN7(dev_priv) || IS_GEN8(dev_priv)) - d->reg_post = ECOBUS; - - d->i915 = dev_priv; d->id = domain_id; BUILD_BUG_ON(FORCEWAKE_RENDER != (1 << FW_DOMAIN_ID_RENDER)); BUILD_BUG_ON(FORCEWAKE_BLITTER != (1 << FW_DOMAIN_ID_BLITTER)); BUILD_BUG_ON(FORCEWAKE_MEDIA != (1 << FW_DOMAIN_ID_MEDIA)); - d->mask = 1 << domain_id; + d->mask = BIT(domain_id); hrtimer_init(&d->timer, CLOCK_MONOTONIC, HRTIMER_MODE_REL); d->timer.function = intel_uncore_fw_release_timer; - dev_priv->uncore.fw_domains |= (1 << domain_id); + dev_priv->uncore.fw_domains |= BIT(domain_id); - fw_domain_reset(d); + fw_domain_reset(dev_priv, d); } static void intel_uncore_fw_domains_init(struct drm_i915_private *dev_priv) @@ -1183,6 +1166,17 @@ static void intel_uncore_fw_domains_init(struct drm_i915_private *dev_priv) if (INTEL_GEN(dev_priv) <= 5 || intel_vgpu_active(dev_priv)) return; + if (IS_GEN6(dev_priv)) { + dev_priv->uncore.fw_reset = 0; + dev_priv->uncore.fw_set = FORCEWAKE_KERNEL; + dev_priv->uncore.fw_clear = 0; + } else { + /* WaRsClearFWBitsAtReset:bdw,skl */ + dev_priv->uncore.fw_reset = _MASKED_BIT_DISABLE(0xffff); + dev_priv->uncore.fw_set = _MASKED_BIT_ENABLE(FORCEWAKE_KERNEL); + dev_priv->uncore.fw_clear = _MASKED_BIT_DISABLE(FORCEWAKE_KERNEL); + } + if (IS_GEN9(dev_priv)) { dev_priv->uncore.funcs.force_wake_get = fw_domains_get; dev_priv->uncore.funcs.force_wake_put = fw_domains_put; @@ -1246,9 +1240,9 @@ static void intel_uncore_fw_domains_init(struct drm_i915_private *dev_priv) FORCEWAKE_MT, FORCEWAKE_MT_ACK); spin_lock_irq(&dev_priv->uncore.lock); - fw_domains_get_with_thread_status(dev_priv, FORCEWAKE_ALL); + fw_domains_get_with_thread_status(dev_priv, FORCEWAKE_RENDER); ecobus = __raw_i915_read32(dev_priv, ECOBUS); - fw_domains_put_with_fifo(dev_priv, FORCEWAKE_ALL); + fw_domains_put_with_fifo(dev_priv, FORCEWAKE_RENDER); spin_unlock_irq(&dev_priv->uncore.lock); if (!(ecobus & FORCEWAKE_MT_ENABLE)) { diff --git a/drivers/gpu/drm/i915/selftests/i915_gem_request.c b/drivers/gpu/drm/i915/selftests/i915_gem_request.c index 926b24c117d6..98b7aac41eec 100644 --- a/drivers/gpu/drm/i915/selftests/i915_gem_request.c +++ b/drivers/gpu/drm/i915/selftests/i915_gem_request.c @@ -291,8 +291,6 @@ static int begin_live_test(struct live_test *t, return err; } - i915_gem_retire_requests(i915); - i915->gpu_error.missed_irq_rings = 0; t->reset_count = i915_reset_count(&i915->gpu_error); @@ -303,7 +301,9 @@ static int end_live_test(struct live_test *t) { struct drm_i915_private *i915 = t->i915; - if (wait_for(intel_engines_are_idle(i915), 1)) { + i915_gem_retire_requests(i915); + + if (wait_for(intel_engines_are_idle(i915), 10)) { pr_err("%s(%s): GPU not idle\n", t->func, t->name); return -EIO; } diff --git a/drivers/gpu/drm/i915/selftests/intel_hangcheck.c b/drivers/gpu/drm/i915/selftests/intel_hangcheck.c index 6ec7c731a267..aa31d6c0cdfb 100644 --- a/drivers/gpu/drm/i915/selftests/intel_hangcheck.c +++ b/drivers/gpu/drm/i915/selftests/intel_hangcheck.c @@ -235,7 +235,6 @@ static void hang_fini(struct hang *h) i915_gem_object_put(h->hws); i915_gem_wait_for_idle(h->i915, I915_WAIT_LOCKED); - i915_gem_retire_requests(h->i915); } static int igt_hang_sanitycheck(void *arg) diff --git a/drivers/gpu/drm/i915/selftests/mock_engine.c b/drivers/gpu/drm/i915/selftests/mock_engine.c index 8d5ba037064c..0ad624a1db90 100644 --- a/drivers/gpu/drm/i915/selftests/mock_engine.c +++ b/drivers/gpu/drm/i915/selftests/mock_engine.c @@ -118,7 +118,6 @@ static struct intel_ring *mock_ring(struct intel_engine_cs *engine) ring->vaddr = (void *)(ring + 1); INIT_LIST_HEAD(&ring->request_list); - ring->last_retired_head = -1; intel_ring_update_space(ring); return ring; diff --git a/drivers/gpu/drm/i915/selftests/scatterlist.c b/drivers/gpu/drm/i915/selftests/scatterlist.c index eb2cda8e2b9f..1cc5d2931753 100644 --- a/drivers/gpu/drm/i915/selftests/scatterlist.c +++ b/drivers/gpu/drm/i915/selftests/scatterlist.c @@ -189,6 +189,13 @@ static unsigned int random(unsigned long n, return 1 + (prandom_u32_state(rnd) % 1024); } +static inline bool page_contiguous(struct page *first, + struct page *last, + unsigned long npages) +{ + return first + npages == last; +} + static int alloc_table(struct pfn_table *pt, unsigned long count, unsigned long max, npages_fn_t npages_fn, @@ -216,7 +223,9 @@ static int alloc_table(struct pfn_table *pt, unsigned long npages = npages_fn(n, count, rnd); /* Nobody expects the Sparse Memmap! */ - if (pfn_to_page(pfn + npages) != pfn_to_page(pfn) + npages) { + if (!page_contiguous(pfn_to_page(pfn), + pfn_to_page(pfn + npages), + npages)) { sg_free_table(&pt->st); return -ENOSPC; } diff --git a/drivers/gpu/drm/imx/Kconfig b/drivers/gpu/drm/imx/Kconfig index f2c9ae822149..c9e439c82241 100644 --- a/drivers/gpu/drm/imx/Kconfig +++ b/drivers/gpu/drm/imx/Kconfig @@ -31,13 +31,6 @@ config DRM_IMX_LDB Choose this to enable the internal LVDS Display Bridge (LDB) found on i.MX53 and i.MX6 processors. -config DRM_IMX_IPUV3 - tristate - depends on DRM_IMX - depends on IMX_IPUV3_CORE - default y if DRM_IMX=y - default m if DRM_IMX=m - config DRM_IMX_HDMI tristate "Freescale i.MX DRM HDMI" select DRM_DW_HDMI diff --git a/drivers/gpu/drm/imx/Makefile b/drivers/gpu/drm/imx/Makefile index f3ecd8903d97..16ecef33e008 100644 --- a/drivers/gpu/drm/imx/Makefile +++ b/drivers/gpu/drm/imx/Makefile @@ -1,5 +1,5 @@ -imxdrm-objs := imx-drm-core.o +imxdrm-objs := imx-drm-core.o ipuv3-crtc.o ipuv3-plane.o obj-$(CONFIG_DRM_IMX) += imxdrm.o @@ -7,6 +7,5 @@ obj-$(CONFIG_DRM_IMX_PARALLEL_DISPLAY) += parallel-display.o obj-$(CONFIG_DRM_IMX_TVE) += imx-tve.o obj-$(CONFIG_DRM_IMX_LDB) += imx-ldb.o -imx-ipuv3-crtc-objs := ipuv3-crtc.o ipuv3-plane.o obj-$(CONFIG_DRM_IMX_IPUV3) += imx-ipuv3-crtc.o obj-$(CONFIG_DRM_IMX_HDMI) += dw_hdmi-imx.o diff --git a/drivers/gpu/drm/imx/imx-drm-core.c b/drivers/gpu/drm/imx/imx-drm-core.c index 1888bf3920fc..50add2f9e250 100644 --- a/drivers/gpu/drm/imx/imx-drm-core.c +++ b/drivers/gpu/drm/imx/imx-drm-core.c @@ -422,7 +422,23 @@ static struct platform_driver imx_drm_pdrv = { .of_match_table = imx_drm_dt_ids, }, }; -module_platform_driver(imx_drm_pdrv); + +static struct platform_driver * const drivers[] = { + &imx_drm_pdrv, + &ipu_drm_driver, +}; + +static int __init imx_drm_init(void) +{ + return platform_register_drivers(drivers, ARRAY_SIZE(drivers)); +} +module_init(imx_drm_init); + +static void __exit imx_drm_exit(void) +{ + platform_unregister_drivers(drivers, ARRAY_SIZE(drivers)); +} +module_exit(imx_drm_exit); MODULE_AUTHOR("Sascha Hauer <s.hauer@pengutronix.de>"); MODULE_DESCRIPTION("i.MX drm driver core"); diff --git a/drivers/gpu/drm/imx/imx-drm.h b/drivers/gpu/drm/imx/imx-drm.h index 295434b199db..f6dd64be9cd5 100644 --- a/drivers/gpu/drm/imx/imx-drm.h +++ b/drivers/gpu/drm/imx/imx-drm.h @@ -29,6 +29,8 @@ int imx_drm_init_drm(struct platform_device *pdev, int preferred_bpp); int imx_drm_exit_drm(void); +extern struct platform_driver ipu_drm_driver; + void imx_drm_mode_config_init(struct drm_device *drm); struct drm_gem_cma_object *imx_drm_fb_get_obj(struct drm_framebuffer *fb); diff --git a/drivers/gpu/drm/imx/ipuv3-crtc.c b/drivers/gpu/drm/imx/ipuv3-crtc.c index dab9d50ffd8c..5456c15d962c 100644 --- a/drivers/gpu/drm/imx/ipuv3-crtc.c +++ b/drivers/gpu/drm/imx/ipuv3-crtc.c @@ -465,16 +465,10 @@ static int ipu_drm_remove(struct platform_device *pdev) return 0; } -static struct platform_driver ipu_drm_driver = { +struct platform_driver ipu_drm_driver = { .driver = { .name = "imx-ipuv3-crtc", }, .probe = ipu_drm_probe, .remove = ipu_drm_remove, }; -module_platform_driver(ipu_drm_driver); - -MODULE_AUTHOR("Sascha Hauer <s.hauer@pengutronix.de>"); -MODULE_DESCRIPTION(DRIVER_DESC); -MODULE_LICENSE("GPL"); -MODULE_ALIAS("platform:imx-ipuv3-crtc"); diff --git a/drivers/gpu/drm/mediatek/mtk_disp_ovl.c b/drivers/gpu/drm/mediatek/mtk_disp_ovl.c index c70310206ac5..a14d7d64d7b1 100644 --- a/drivers/gpu/drm/mediatek/mtk_disp_ovl.c +++ b/drivers/gpu/drm/mediatek/mtk_disp_ovl.c @@ -35,18 +35,28 @@ #define DISP_REG_OVL_PITCH(n) (0x0044 + 0x20 * (n)) #define DISP_REG_OVL_RDMA_CTRL(n) (0x00c0 + 0x20 * (n)) #define DISP_REG_OVL_RDMA_GMC(n) (0x00c8 + 0x20 * (n)) -#define DISP_REG_OVL_ADDR(n) (0x0f40 + 0x20 * (n)) +#define DISP_REG_OVL_ADDR_MT2701 0x0040 +#define DISP_REG_OVL_ADDR_MT8173 0x0f40 +#define DISP_REG_OVL_ADDR(ovl, n) ((ovl)->data->addr + 0x20 * (n)) #define OVL_RDMA_MEM_GMC 0x40402020 #define OVL_CON_BYTE_SWAP BIT(24) -#define OVL_CON_CLRFMT_RGB565 (0 << 12) -#define OVL_CON_CLRFMT_RGB888 (1 << 12) +#define OVL_CON_CLRFMT_RGB (1 << 12) #define OVL_CON_CLRFMT_RGBA8888 (2 << 12) #define OVL_CON_CLRFMT_ARGB8888 (3 << 12) +#define OVL_CON_CLRFMT_RGB565(ovl) ((ovl)->data->fmt_rgb565_is_0 ? \ + 0 : OVL_CON_CLRFMT_RGB) +#define OVL_CON_CLRFMT_RGB888(ovl) ((ovl)->data->fmt_rgb565_is_0 ? \ + OVL_CON_CLRFMT_RGB : 0) #define OVL_CON_AEN BIT(8) #define OVL_CON_ALPHA 0xff +struct mtk_disp_ovl_data { + unsigned int addr; + bool fmt_rgb565_is_0; +}; + /** * struct mtk_disp_ovl - DISP_OVL driver structure * @ddp_comp - structure containing type enum and hardware resources @@ -55,8 +65,14 @@ struct mtk_disp_ovl { struct mtk_ddp_comp ddp_comp; struct drm_crtc *crtc; + const struct mtk_disp_ovl_data *data; }; +static inline struct mtk_disp_ovl *comp_to_ovl(struct mtk_ddp_comp *comp) +{ + return container_of(comp, struct mtk_disp_ovl, ddp_comp); +} + static irqreturn_t mtk_disp_ovl_irq_handler(int irq, void *dev_id) { struct mtk_disp_ovl *priv = dev_id; @@ -76,20 +92,18 @@ static irqreturn_t mtk_disp_ovl_irq_handler(int irq, void *dev_id) static void mtk_ovl_enable_vblank(struct mtk_ddp_comp *comp, struct drm_crtc *crtc) { - struct mtk_disp_ovl *priv = container_of(comp, struct mtk_disp_ovl, - ddp_comp); + struct mtk_disp_ovl *ovl = comp_to_ovl(comp); - priv->crtc = crtc; + ovl->crtc = crtc; writel(0x0, comp->regs + DISP_REG_OVL_INTSTA); writel_relaxed(OVL_FME_CPL_INT, comp->regs + DISP_REG_OVL_INTEN); } static void mtk_ovl_disable_vblank(struct mtk_ddp_comp *comp) { - struct mtk_disp_ovl *priv = container_of(comp, struct mtk_disp_ovl, - ddp_comp); + struct mtk_disp_ovl *ovl = comp_to_ovl(comp); - priv->crtc = NULL; + ovl->crtc = NULL; writel_relaxed(0x0, comp->regs + DISP_REG_OVL_INTEN); } @@ -138,18 +152,18 @@ static void mtk_ovl_layer_off(struct mtk_ddp_comp *comp, unsigned int idx) writel(0x0, comp->regs + DISP_REG_OVL_RDMA_CTRL(idx)); } -static unsigned int ovl_fmt_convert(unsigned int fmt) +static unsigned int ovl_fmt_convert(struct mtk_disp_ovl *ovl, unsigned int fmt) { switch (fmt) { default: case DRM_FORMAT_RGB565: - return OVL_CON_CLRFMT_RGB565; + return OVL_CON_CLRFMT_RGB565(ovl); case DRM_FORMAT_BGR565: - return OVL_CON_CLRFMT_RGB565 | OVL_CON_BYTE_SWAP; + return OVL_CON_CLRFMT_RGB565(ovl) | OVL_CON_BYTE_SWAP; case DRM_FORMAT_RGB888: - return OVL_CON_CLRFMT_RGB888; + return OVL_CON_CLRFMT_RGB888(ovl); case DRM_FORMAT_BGR888: - return OVL_CON_CLRFMT_RGB888 | OVL_CON_BYTE_SWAP; + return OVL_CON_CLRFMT_RGB888(ovl) | OVL_CON_BYTE_SWAP; case DRM_FORMAT_RGBX8888: case DRM_FORMAT_RGBA8888: return OVL_CON_CLRFMT_ARGB8888; @@ -168,6 +182,7 @@ static unsigned int ovl_fmt_convert(unsigned int fmt) static void mtk_ovl_layer_config(struct mtk_ddp_comp *comp, unsigned int idx, struct mtk_plane_state *state) { + struct mtk_disp_ovl *ovl = comp_to_ovl(comp); struct mtk_plane_pending_state *pending = &state->pending; unsigned int addr = pending->addr; unsigned int pitch = pending->pitch & 0xffff; @@ -179,7 +194,7 @@ static void mtk_ovl_layer_config(struct mtk_ddp_comp *comp, unsigned int idx, if (!pending->enable) mtk_ovl_layer_off(comp, idx); - con = ovl_fmt_convert(fmt); + con = ovl_fmt_convert(ovl, fmt); if (idx != 0) con |= OVL_CON_AEN | OVL_CON_ALPHA; @@ -187,7 +202,7 @@ static void mtk_ovl_layer_config(struct mtk_ddp_comp *comp, unsigned int idx, writel_relaxed(pitch, comp->regs + DISP_REG_OVL_PITCH(idx)); writel_relaxed(src_size, comp->regs + DISP_REG_OVL_SRC_SIZE(idx)); writel_relaxed(offset, comp->regs + DISP_REG_OVL_OFFSET(idx)); - writel_relaxed(addr, comp->regs + DISP_REG_OVL_ADDR(idx)); + writel_relaxed(addr, comp->regs + DISP_REG_OVL_ADDR(ovl, idx)); if (pending->enable) mtk_ovl_layer_on(comp, idx); @@ -264,6 +279,8 @@ static int mtk_disp_ovl_probe(struct platform_device *pdev) return ret; } + priv->data = of_device_get_match_data(dev); + platform_set_drvdata(pdev, priv); ret = devm_request_irq(dev, irq, mtk_disp_ovl_irq_handler, @@ -287,8 +304,21 @@ static int mtk_disp_ovl_remove(struct platform_device *pdev) return 0; } +static const struct mtk_disp_ovl_data mt2701_ovl_driver_data = { + .addr = DISP_REG_OVL_ADDR_MT2701, + .fmt_rgb565_is_0 = false, +}; + +static const struct mtk_disp_ovl_data mt8173_ovl_driver_data = { + .addr = DISP_REG_OVL_ADDR_MT8173, + .fmt_rgb565_is_0 = true, +}; + static const struct of_device_id mtk_disp_ovl_driver_dt_match[] = { - { .compatible = "mediatek,mt8173-disp-ovl", }, + { .compatible = "mediatek,mt2701-disp-ovl", + .data = &mt2701_ovl_driver_data}, + { .compatible = "mediatek,mt8173-disp-ovl", + .data = &mt8173_ovl_driver_data}, {}, }; MODULE_DEVICE_TABLE(of, mtk_disp_ovl_driver_dt_match); diff --git a/drivers/gpu/drm/mediatek/mtk_disp_rdma.c b/drivers/gpu/drm/mediatek/mtk_disp_rdma.c index 0df05f95b916..b68a51376f83 100644 --- a/drivers/gpu/drm/mediatek/mtk_disp_rdma.c +++ b/drivers/gpu/drm/mediatek/mtk_disp_rdma.c @@ -38,6 +38,11 @@ #define RDMA_FIFO_UNDERFLOW_EN BIT(31) #define RDMA_FIFO_PSEUDO_SIZE(bytes) (((bytes) / 16) << 16) #define RDMA_OUTPUT_VALID_FIFO_THRESHOLD(bytes) ((bytes) / 16) +#define RDMA_FIFO_SIZE(rdma) ((rdma)->data->fifo_size) + +struct mtk_disp_rdma_data { + unsigned int fifo_size; +}; /** * struct mtk_disp_rdma - DISP_RDMA driver structure @@ -47,8 +52,14 @@ struct mtk_disp_rdma { struct mtk_ddp_comp ddp_comp; struct drm_crtc *crtc; + const struct mtk_disp_rdma_data *data; }; +static inline struct mtk_disp_rdma *comp_to_rdma(struct mtk_ddp_comp *comp) +{ + return container_of(comp, struct mtk_disp_rdma, ddp_comp); +} + static irqreturn_t mtk_disp_rdma_irq_handler(int irq, void *dev_id) { struct mtk_disp_rdma *priv = dev_id; @@ -77,20 +88,18 @@ static void rdma_update_bits(struct mtk_ddp_comp *comp, unsigned int reg, static void mtk_rdma_enable_vblank(struct mtk_ddp_comp *comp, struct drm_crtc *crtc) { - struct mtk_disp_rdma *priv = container_of(comp, struct mtk_disp_rdma, - ddp_comp); + struct mtk_disp_rdma *rdma = comp_to_rdma(comp); - priv->crtc = crtc; + rdma->crtc = crtc; rdma_update_bits(comp, DISP_REG_RDMA_INT_ENABLE, RDMA_FRAME_END_INT, RDMA_FRAME_END_INT); } static void mtk_rdma_disable_vblank(struct mtk_ddp_comp *comp) { - struct mtk_disp_rdma *priv = container_of(comp, struct mtk_disp_rdma, - ddp_comp); + struct mtk_disp_rdma *rdma = comp_to_rdma(comp); - priv->crtc = NULL; + rdma->crtc = NULL; rdma_update_bits(comp, DISP_REG_RDMA_INT_ENABLE, RDMA_FRAME_END_INT, 0); } @@ -111,6 +120,7 @@ static void mtk_rdma_config(struct mtk_ddp_comp *comp, unsigned int width, { unsigned int threshold; unsigned int reg; + struct mtk_disp_rdma *rdma = comp_to_rdma(comp); rdma_update_bits(comp, DISP_REG_RDMA_SIZE_CON_0, 0xfff, width); rdma_update_bits(comp, DISP_REG_RDMA_SIZE_CON_1, 0xfffff, height); @@ -123,7 +133,7 @@ static void mtk_rdma_config(struct mtk_ddp_comp *comp, unsigned int width, */ threshold = width * height * vrefresh * 4 * 7 / 1000000; reg = RDMA_FIFO_UNDERFLOW_EN | - RDMA_FIFO_PSEUDO_SIZE(SZ_8K) | + RDMA_FIFO_PSEUDO_SIZE(RDMA_FIFO_SIZE(rdma)) | RDMA_OUTPUT_VALID_FIFO_THRESHOLD(threshold); writel(reg, comp->regs + DISP_REG_RDMA_FIFO_CON); } @@ -208,6 +218,8 @@ static int mtk_disp_rdma_probe(struct platform_device *pdev) return ret; } + priv->data = of_device_get_match_data(dev); + platform_set_drvdata(pdev, priv); ret = component_add(dev, &mtk_disp_rdma_component_ops); @@ -224,8 +236,19 @@ static int mtk_disp_rdma_remove(struct platform_device *pdev) return 0; } +static const struct mtk_disp_rdma_data mt2701_rdma_driver_data = { + .fifo_size = SZ_4K, +}; + +static const struct mtk_disp_rdma_data mt8173_rdma_driver_data = { + .fifo_size = SZ_8K, +}; + static const struct of_device_id mtk_disp_rdma_driver_dt_match[] = { - { .compatible = "mediatek,mt8173-disp-rdma", }, + { .compatible = "mediatek,mt2701-disp-rdma", + .data = &mt2701_rdma_driver_data}, + { .compatible = "mediatek,mt8173-disp-rdma", + .data = &mt8173_rdma_driver_data}, {}, }; MODULE_DEVICE_TABLE(of, mtk_disp_rdma_driver_dt_match); diff --git a/drivers/gpu/drm/mediatek/mtk_drm_crtc.c b/drivers/gpu/drm/mediatek/mtk_drm_crtc.c index 69982f5a6198..6b08774e5501 100644 --- a/drivers/gpu/drm/mediatek/mtk_drm_crtc.c +++ b/drivers/gpu/drm/mediatek/mtk_drm_crtc.c @@ -327,6 +327,42 @@ static void mtk_crtc_ddp_hw_fini(struct mtk_drm_crtc *mtk_crtc) pm_runtime_put(drm->dev); } +static void mtk_crtc_ddp_config(struct drm_crtc *crtc) +{ + struct mtk_drm_crtc *mtk_crtc = to_mtk_crtc(crtc); + struct mtk_crtc_state *state = to_mtk_crtc_state(mtk_crtc->base.state); + struct mtk_ddp_comp *ovl = mtk_crtc->ddp_comp[0]; + unsigned int i; + + /* + * TODO: instead of updating the registers here, we should prepare + * working registers in atomic_commit and let the hardware command + * queue update module registers on vblank. + */ + if (state->pending_config) { + mtk_ddp_comp_config(ovl, state->pending_width, + state->pending_height, + state->pending_vrefresh, 0); + + state->pending_config = false; + } + + if (mtk_crtc->pending_planes) { + for (i = 0; i < OVL_LAYER_NR; i++) { + struct drm_plane *plane = &mtk_crtc->planes[i]; + struct mtk_plane_state *plane_state; + + plane_state = to_mtk_plane_state(plane->state); + + if (plane_state->pending.config) { + mtk_ddp_comp_layer_config(ovl, i, plane_state); + plane_state->pending.config = false; + } + } + mtk_crtc->pending_planes = false; + } +} + static void mtk_drm_crtc_enable(struct drm_crtc *crtc) { struct mtk_drm_crtc *mtk_crtc = to_mtk_crtc(crtc); @@ -403,6 +439,7 @@ static void mtk_drm_crtc_atomic_flush(struct drm_crtc *crtc, struct drm_crtc_state *old_crtc_state) { struct mtk_drm_crtc *mtk_crtc = to_mtk_crtc(crtc); + struct mtk_drm_private *priv = crtc->dev->dev_private; unsigned int pending_planes = 0; int i; @@ -424,6 +461,12 @@ static void mtk_drm_crtc_atomic_flush(struct drm_crtc *crtc, if (crtc->state->color_mgmt_changed) for (i = 0; i < mtk_crtc->ddp_comp_nr; i++) mtk_ddp_gamma_set(mtk_crtc->ddp_comp[i], crtc->state); + + if (priv->data->shadow_register) { + mtk_disp_mutex_acquire(mtk_crtc->mutex); + mtk_crtc_ddp_config(crtc); + mtk_disp_mutex_release(mtk_crtc->mutex); + } } static const struct drm_crtc_funcs mtk_crtc_funcs = { @@ -471,36 +514,10 @@ err_cleanup_crtc: void mtk_crtc_ddp_irq(struct drm_crtc *crtc, struct mtk_ddp_comp *ovl) { struct mtk_drm_crtc *mtk_crtc = to_mtk_crtc(crtc); - struct mtk_crtc_state *state = to_mtk_crtc_state(mtk_crtc->base.state); - unsigned int i; + struct mtk_drm_private *priv = crtc->dev->dev_private; - /* - * TODO: instead of updating the registers here, we should prepare - * working registers in atomic_commit and let the hardware command - * queue update module registers on vblank. - */ - if (state->pending_config) { - mtk_ddp_comp_config(ovl, state->pending_width, - state->pending_height, - state->pending_vrefresh, 0); - - state->pending_config = false; - } - - if (mtk_crtc->pending_planes) { - for (i = 0; i < OVL_LAYER_NR; i++) { - struct drm_plane *plane = &mtk_crtc->planes[i]; - struct mtk_plane_state *plane_state; - - plane_state = to_mtk_plane_state(plane->state); - - if (plane_state->pending.config) { - mtk_ddp_comp_layer_config(ovl, i, plane_state); - plane_state->pending.config = false; - } - } - mtk_crtc->pending_planes = false; - } + if (!priv->data->shadow_register) + mtk_crtc_ddp_config(crtc); mtk_drm_finish_page_flip(mtk_crtc); } diff --git a/drivers/gpu/drm/mediatek/mtk_drm_ddp.c b/drivers/gpu/drm/mediatek/mtk_drm_ddp.c index 17ba9355a49c..8130f3dab661 100644 --- a/drivers/gpu/drm/mediatek/mtk_drm_ddp.c +++ b/drivers/gpu/drm/mediatek/mtk_drm_ddp.c @@ -12,6 +12,7 @@ */ #include <linux/clk.h> +#include <linux/iopoll.h> #include <linux/module.h> #include <linux/of_device.h> #include <linux/platform_device.h> @@ -31,26 +32,40 @@ #define DISP_REG_CONFIG_DISP_RDMA1_MOUT_EN 0x0c8 #define DISP_REG_CONFIG_MMSYS_CG_CON0 0x100 +#define DISP_REG_CONFIG_DISP_OVL_MOUT_EN 0x030 +#define DISP_REG_CONFIG_OUT_SEL 0x04c +#define DISP_REG_CONFIG_DSI_SEL 0x050 + #define DISP_REG_MUTEX_EN(n) (0x20 + 0x20 * (n)) +#define DISP_REG_MUTEX(n) (0x24 + 0x20 * (n)) #define DISP_REG_MUTEX_RST(n) (0x28 + 0x20 * (n)) #define DISP_REG_MUTEX_MOD(n) (0x2c + 0x20 * (n)) #define DISP_REG_MUTEX_SOF(n) (0x30 + 0x20 * (n)) -#define MUTEX_MOD_DISP_OVL0 BIT(11) -#define MUTEX_MOD_DISP_OVL1 BIT(12) -#define MUTEX_MOD_DISP_RDMA0 BIT(13) -#define MUTEX_MOD_DISP_RDMA1 BIT(14) -#define MUTEX_MOD_DISP_RDMA2 BIT(15) -#define MUTEX_MOD_DISP_WDMA0 BIT(16) -#define MUTEX_MOD_DISP_WDMA1 BIT(17) -#define MUTEX_MOD_DISP_COLOR0 BIT(18) -#define MUTEX_MOD_DISP_COLOR1 BIT(19) -#define MUTEX_MOD_DISP_AAL BIT(20) -#define MUTEX_MOD_DISP_GAMMA BIT(21) -#define MUTEX_MOD_DISP_UFOE BIT(22) -#define MUTEX_MOD_DISP_PWM0 BIT(23) -#define MUTEX_MOD_DISP_PWM1 BIT(24) -#define MUTEX_MOD_DISP_OD BIT(25) +#define INT_MUTEX BIT(1) + +#define MT8173_MUTEX_MOD_DISP_OVL0 BIT(11) +#define MT8173_MUTEX_MOD_DISP_OVL1 BIT(12) +#define MT8173_MUTEX_MOD_DISP_RDMA0 BIT(13) +#define MT8173_MUTEX_MOD_DISP_RDMA1 BIT(14) +#define MT8173_MUTEX_MOD_DISP_RDMA2 BIT(15) +#define MT8173_MUTEX_MOD_DISP_WDMA0 BIT(16) +#define MT8173_MUTEX_MOD_DISP_WDMA1 BIT(17) +#define MT8173_MUTEX_MOD_DISP_COLOR0 BIT(18) +#define MT8173_MUTEX_MOD_DISP_COLOR1 BIT(19) +#define MT8173_MUTEX_MOD_DISP_AAL BIT(20) +#define MT8173_MUTEX_MOD_DISP_GAMMA BIT(21) +#define MT8173_MUTEX_MOD_DISP_UFOE BIT(22) +#define MT8173_MUTEX_MOD_DISP_PWM0 BIT(23) +#define MT8173_MUTEX_MOD_DISP_PWM1 BIT(24) +#define MT8173_MUTEX_MOD_DISP_OD BIT(25) + +#define MT2701_MUTEX_MOD_DISP_OVL BIT(3) +#define MT2701_MUTEX_MOD_DISP_WDMA BIT(6) +#define MT2701_MUTEX_MOD_DISP_COLOR BIT(7) +#define MT2701_MUTEX_MOD_DISP_BLS BIT(9) +#define MT2701_MUTEX_MOD_DISP_RDMA0 BIT(10) +#define MT2701_MUTEX_MOD_DISP_RDMA1 BIT(12) #define MUTEX_SOF_SINGLE_MODE 0 #define MUTEX_SOF_DSI0 1 @@ -67,6 +82,10 @@ #define DPI0_SEL_IN_RDMA1 0x1 #define COLOR1_SEL_IN_OVL1 0x1 +#define OVL_MOUT_EN_RDMA 0x1 +#define BLS_TO_DSI_RDMA1_TO_DPI1 0x8 +#define DSI_SEL_IN_BLS 0x0 + struct mtk_disp_mutex { int id; bool claimed; @@ -77,24 +96,34 @@ struct mtk_ddp { struct clk *clk; void __iomem *regs; struct mtk_disp_mutex mutex[10]; + const unsigned int *mutex_mod; +}; + +static const unsigned int mt2701_mutex_mod[DDP_COMPONENT_ID_MAX] = { + [DDP_COMPONENT_BLS] = MT2701_MUTEX_MOD_DISP_BLS, + [DDP_COMPONENT_COLOR0] = MT2701_MUTEX_MOD_DISP_COLOR, + [DDP_COMPONENT_OVL0] = MT2701_MUTEX_MOD_DISP_OVL, + [DDP_COMPONENT_RDMA0] = MT2701_MUTEX_MOD_DISP_RDMA0, + [DDP_COMPONENT_RDMA1] = MT2701_MUTEX_MOD_DISP_RDMA1, + [DDP_COMPONENT_WDMA0] = MT2701_MUTEX_MOD_DISP_WDMA, }; -static const unsigned int mutex_mod[DDP_COMPONENT_ID_MAX] = { - [DDP_COMPONENT_AAL] = MUTEX_MOD_DISP_AAL, - [DDP_COMPONENT_COLOR0] = MUTEX_MOD_DISP_COLOR0, - [DDP_COMPONENT_COLOR1] = MUTEX_MOD_DISP_COLOR1, - [DDP_COMPONENT_GAMMA] = MUTEX_MOD_DISP_GAMMA, - [DDP_COMPONENT_OD] = MUTEX_MOD_DISP_OD, - [DDP_COMPONENT_OVL0] = MUTEX_MOD_DISP_OVL0, - [DDP_COMPONENT_OVL1] = MUTEX_MOD_DISP_OVL1, - [DDP_COMPONENT_PWM0] = MUTEX_MOD_DISP_PWM0, - [DDP_COMPONENT_PWM1] = MUTEX_MOD_DISP_PWM1, - [DDP_COMPONENT_RDMA0] = MUTEX_MOD_DISP_RDMA0, - [DDP_COMPONENT_RDMA1] = MUTEX_MOD_DISP_RDMA1, - [DDP_COMPONENT_RDMA2] = MUTEX_MOD_DISP_RDMA2, - [DDP_COMPONENT_UFOE] = MUTEX_MOD_DISP_UFOE, - [DDP_COMPONENT_WDMA0] = MUTEX_MOD_DISP_WDMA0, - [DDP_COMPONENT_WDMA1] = MUTEX_MOD_DISP_WDMA1, +static const unsigned int mt8173_mutex_mod[DDP_COMPONENT_ID_MAX] = { + [DDP_COMPONENT_AAL] = MT8173_MUTEX_MOD_DISP_AAL, + [DDP_COMPONENT_COLOR0] = MT8173_MUTEX_MOD_DISP_COLOR0, + [DDP_COMPONENT_COLOR1] = MT8173_MUTEX_MOD_DISP_COLOR1, + [DDP_COMPONENT_GAMMA] = MT8173_MUTEX_MOD_DISP_GAMMA, + [DDP_COMPONENT_OD] = MT8173_MUTEX_MOD_DISP_OD, + [DDP_COMPONENT_OVL0] = MT8173_MUTEX_MOD_DISP_OVL0, + [DDP_COMPONENT_OVL1] = MT8173_MUTEX_MOD_DISP_OVL1, + [DDP_COMPONENT_PWM0] = MT8173_MUTEX_MOD_DISP_PWM0, + [DDP_COMPONENT_PWM1] = MT8173_MUTEX_MOD_DISP_PWM1, + [DDP_COMPONENT_RDMA0] = MT8173_MUTEX_MOD_DISP_RDMA0, + [DDP_COMPONENT_RDMA1] = MT8173_MUTEX_MOD_DISP_RDMA1, + [DDP_COMPONENT_RDMA2] = MT8173_MUTEX_MOD_DISP_RDMA2, + [DDP_COMPONENT_UFOE] = MT8173_MUTEX_MOD_DISP_UFOE, + [DDP_COMPONENT_WDMA0] = MT8173_MUTEX_MOD_DISP_WDMA0, + [DDP_COMPONENT_WDMA1] = MT8173_MUTEX_MOD_DISP_WDMA1, }; static unsigned int mtk_ddp_mout_en(enum mtk_ddp_comp_id cur, @@ -106,6 +135,9 @@ static unsigned int mtk_ddp_mout_en(enum mtk_ddp_comp_id cur, if (cur == DDP_COMPONENT_OVL0 && next == DDP_COMPONENT_COLOR0) { *addr = DISP_REG_CONFIG_DISP_OVL0_MOUT_EN; value = OVL0_MOUT_EN_COLOR0; + } else if (cur == DDP_COMPONENT_OVL0 && next == DDP_COMPONENT_RDMA0) { + *addr = DISP_REG_CONFIG_DISP_OVL_MOUT_EN; + value = OVL_MOUT_EN_RDMA; } else if (cur == DDP_COMPONENT_OD && next == DDP_COMPONENT_RDMA0) { *addr = DISP_REG_CONFIG_DISP_OD_MOUT_EN; value = OD_MOUT_EN_RDMA0; @@ -143,6 +175,9 @@ static unsigned int mtk_ddp_sel_in(enum mtk_ddp_comp_id cur, } else if (cur == DDP_COMPONENT_OVL1 && next == DDP_COMPONENT_COLOR1) { *addr = DISP_REG_CONFIG_DISP_COLOR1_SEL_IN; value = COLOR1_SEL_IN_OVL1; + } else if (cur == DDP_COMPONENT_BLS && next == DDP_COMPONENT_DSI0) { + *addr = DISP_REG_CONFIG_DSI_SEL; + value = DSI_SEL_IN_BLS; } else { value = 0; } @@ -150,6 +185,15 @@ static unsigned int mtk_ddp_sel_in(enum mtk_ddp_comp_id cur, return value; } +static void mtk_ddp_sout_sel(void __iomem *config_regs, + enum mtk_ddp_comp_id cur, + enum mtk_ddp_comp_id next) +{ + if (cur == DDP_COMPONENT_BLS && next == DDP_COMPONENT_DSI0) + writel_relaxed(BLS_TO_DSI_RDMA1_TO_DPI1, + config_regs + DISP_REG_CONFIG_OUT_SEL); +} + void mtk_ddp_add_comp_to_path(void __iomem *config_regs, enum mtk_ddp_comp_id cur, enum mtk_ddp_comp_id next) @@ -162,6 +206,8 @@ void mtk_ddp_add_comp_to_path(void __iomem *config_regs, writel_relaxed(reg, config_regs + addr); } + mtk_ddp_sout_sel(config_regs, cur, next); + value = mtk_ddp_sel_in(cur, next, &addr); if (value) { reg = readl_relaxed(config_regs + addr) | value; @@ -247,7 +293,7 @@ void mtk_disp_mutex_add_comp(struct mtk_disp_mutex *mutex, break; default: reg = readl_relaxed(ddp->regs + DISP_REG_MUTEX_MOD(mutex->id)); - reg |= mutex_mod[id]; + reg |= ddp->mutex_mod[id]; writel_relaxed(reg, ddp->regs + DISP_REG_MUTEX_MOD(mutex->id)); return; } @@ -273,7 +319,7 @@ void mtk_disp_mutex_remove_comp(struct mtk_disp_mutex *mutex, break; default: reg = readl_relaxed(ddp->regs + DISP_REG_MUTEX_MOD(mutex->id)); - reg &= ~mutex_mod[id]; + reg &= ~(ddp->mutex_mod[id]); writel_relaxed(reg, ddp->regs + DISP_REG_MUTEX_MOD(mutex->id)); break; } @@ -299,6 +345,27 @@ void mtk_disp_mutex_disable(struct mtk_disp_mutex *mutex) writel(0, ddp->regs + DISP_REG_MUTEX_EN(mutex->id)); } +void mtk_disp_mutex_acquire(struct mtk_disp_mutex *mutex) +{ + struct mtk_ddp *ddp = container_of(mutex, struct mtk_ddp, + mutex[mutex->id]); + u32 tmp; + + writel(1, ddp->regs + DISP_REG_MUTEX_EN(mutex->id)); + writel(1, ddp->regs + DISP_REG_MUTEX(mutex->id)); + if (readl_poll_timeout_atomic(ddp->regs + DISP_REG_MUTEX(mutex->id), + tmp, tmp & INT_MUTEX, 1, 10000)) + pr_err("could not acquire mutex %d\n", mutex->id); +} + +void mtk_disp_mutex_release(struct mtk_disp_mutex *mutex) +{ + struct mtk_ddp *ddp = container_of(mutex, struct mtk_ddp, + mutex[mutex->id]); + + writel(0, ddp->regs + DISP_REG_MUTEX(mutex->id)); +} + static int mtk_ddp_probe(struct platform_device *pdev) { struct device *dev = &pdev->dev; @@ -326,6 +393,8 @@ static int mtk_ddp_probe(struct platform_device *pdev) return PTR_ERR(ddp->regs); } + ddp->mutex_mod = of_device_get_match_data(dev); + platform_set_drvdata(pdev, ddp); return 0; @@ -337,7 +406,8 @@ static int mtk_ddp_remove(struct platform_device *pdev) } static const struct of_device_id ddp_driver_dt_match[] = { - { .compatible = "mediatek,mt8173-disp-mutex" }, + { .compatible = "mediatek,mt2701-disp-mutex", .data = mt2701_mutex_mod}, + { .compatible = "mediatek,mt8173-disp-mutex", .data = mt8173_mutex_mod}, {}, }; MODULE_DEVICE_TABLE(of, ddp_driver_dt_match); diff --git a/drivers/gpu/drm/mediatek/mtk_drm_ddp.h b/drivers/gpu/drm/mediatek/mtk_drm_ddp.h index 92c11752ff65..f9a799168077 100644 --- a/drivers/gpu/drm/mediatek/mtk_drm_ddp.h +++ b/drivers/gpu/drm/mediatek/mtk_drm_ddp.h @@ -37,5 +37,7 @@ void mtk_disp_mutex_remove_comp(struct mtk_disp_mutex *mutex, enum mtk_ddp_comp_id id); void mtk_disp_mutex_unprepare(struct mtk_disp_mutex *mutex); void mtk_disp_mutex_put(struct mtk_disp_mutex *mutex); +void mtk_disp_mutex_acquire(struct mtk_disp_mutex *mutex); +void mtk_disp_mutex_release(struct mtk_disp_mutex *mutex); #endif /* MTK_DRM_DDP_H */ diff --git a/drivers/gpu/drm/mediatek/mtk_drm_ddp_comp.c b/drivers/gpu/drm/mediatek/mtk_drm_ddp_comp.c index 48cc01fd20c7..8b52416b6e41 100644 --- a/drivers/gpu/drm/mediatek/mtk_drm_ddp_comp.c +++ b/drivers/gpu/drm/mediatek/mtk_drm_ddp_comp.c @@ -39,9 +39,11 @@ #define DISP_REG_UFO_START 0x0000 #define DISP_COLOR_CFG_MAIN 0x0400 -#define DISP_COLOR_START 0x0c00 -#define DISP_COLOR_WIDTH 0x0c50 -#define DISP_COLOR_HEIGHT 0x0c54 +#define DISP_COLOR_START_MT2701 0x0f00 +#define DISP_COLOR_START_MT8173 0x0c00 +#define DISP_COLOR_START(comp) ((comp)->data->color_offset) +#define DISP_COLOR_WIDTH(comp) (DISP_COLOR_START(comp) + 0x50) +#define DISP_COLOR_HEIGHT(comp) (DISP_COLOR_START(comp) + 0x54) #define DISP_AAL_EN 0x0000 #define DISP_AAL_SIZE 0x0030 @@ -80,6 +82,20 @@ #define DITHER_ADD_LSHIFT_G(x) (((x) & 0x7) << 4) #define DITHER_ADD_RSHIFT_G(x) (((x) & 0x7) << 0) +struct mtk_disp_color_data { + unsigned int color_offset; +}; + +struct mtk_disp_color { + struct mtk_ddp_comp ddp_comp; + const struct mtk_disp_color_data *data; +}; + +static inline struct mtk_disp_color *comp_to_color(struct mtk_ddp_comp *comp) +{ + return container_of(comp, struct mtk_disp_color, ddp_comp); +} + void mtk_dither_set(struct mtk_ddp_comp *comp, unsigned int bpc, unsigned int CFG) { @@ -107,15 +123,19 @@ static void mtk_color_config(struct mtk_ddp_comp *comp, unsigned int w, unsigned int h, unsigned int vrefresh, unsigned int bpc) { - writel(w, comp->regs + DISP_COLOR_WIDTH); - writel(h, comp->regs + DISP_COLOR_HEIGHT); + struct mtk_disp_color *color = comp_to_color(comp); + + writel(w, comp->regs + DISP_COLOR_WIDTH(color)); + writel(h, comp->regs + DISP_COLOR_HEIGHT(color)); } static void mtk_color_start(struct mtk_ddp_comp *comp) { + struct mtk_disp_color *color = comp_to_color(comp); + writel(COLOR_BYPASS_ALL | COLOR_SEQ_SEL, comp->regs + DISP_COLOR_CFG_MAIN); - writel(0x1, comp->regs + DISP_COLOR_START); + writel(0x1, comp->regs + DISP_COLOR_START(color)); } static void mtk_od_config(struct mtk_ddp_comp *comp, unsigned int w, @@ -236,6 +256,7 @@ static const char * const mtk_ddp_comp_stem[MTK_DDP_COMP_TYPE_MAX] = { [MTK_DISP_PWM] = "pwm", [MTK_DISP_MUTEX] = "mutex", [MTK_DISP_OD] = "od", + [MTK_DISP_BLS] = "bls", }; struct mtk_ddp_comp_match { @@ -246,6 +267,7 @@ struct mtk_ddp_comp_match { static const struct mtk_ddp_comp_match mtk_ddp_matches[DDP_COMPONENT_ID_MAX] = { [DDP_COMPONENT_AAL] = { MTK_DISP_AAL, 0, &ddp_aal }, + [DDP_COMPONENT_BLS] = { MTK_DISP_BLS, 0, NULL }, [DDP_COMPONENT_COLOR0] = { MTK_DISP_COLOR, 0, &ddp_color }, [DDP_COMPONENT_COLOR1] = { MTK_DISP_COLOR, 1, &ddp_color }, [DDP_COMPONENT_DPI0] = { MTK_DPI, 0, NULL }, @@ -264,6 +286,22 @@ static const struct mtk_ddp_comp_match mtk_ddp_matches[DDP_COMPONENT_ID_MAX] = { [DDP_COMPONENT_WDMA1] = { MTK_DISP_WDMA, 1, NULL }, }; +static const struct mtk_disp_color_data mt2701_color_driver_data = { + .color_offset = DISP_COLOR_START_MT2701, +}; + +static const struct mtk_disp_color_data mt8173_color_driver_data = { + .color_offset = DISP_COLOR_START_MT8173, +}; + +static const struct of_device_id mtk_disp_color_driver_dt_match[] = { + { .compatible = "mediatek,mt2701-disp-color", + .data = &mt2701_color_driver_data}, + { .compatible = "mediatek,mt8173-disp-color", + .data = &mt8173_color_driver_data}, + {}, +}; + int mtk_ddp_comp_get_id(struct device_node *node, enum mtk_ddp_comp_type comp_type) { @@ -286,14 +324,29 @@ int mtk_ddp_comp_init(struct device *dev, struct device_node *node, enum mtk_ddp_comp_type type; struct device_node *larb_node; struct platform_device *larb_pdev; + const struct of_device_id *match; + struct mtk_disp_color *color; if (comp_id < 0 || comp_id >= DDP_COMPONENT_ID_MAX) return -EINVAL; + type = mtk_ddp_matches[comp_id].type; + if (type == MTK_DISP_COLOR) { + devm_kfree(dev, comp); + color = devm_kzalloc(dev, sizeof(*color), GFP_KERNEL); + if (!color) + return -ENOMEM; + + match = of_match_node(mtk_disp_color_driver_dt_match, node); + color->data = match->data; + comp = &color->ddp_comp; + } + comp->id = comp_id; comp->funcs = funcs ?: mtk_ddp_matches[comp_id].funcs; - if (comp_id == DDP_COMPONENT_DPI0 || + if (comp_id == DDP_COMPONENT_BLS || + comp_id == DDP_COMPONENT_DPI0 || comp_id == DDP_COMPONENT_DSI0 || comp_id == DDP_COMPONENT_PWM0) { comp->regs = NULL; @@ -308,8 +361,6 @@ int mtk_ddp_comp_init(struct device *dev, struct device_node *node, if (IS_ERR(comp->clk)) comp->clk = NULL; - type = mtk_ddp_matches[comp_id].type; - /* Only DMA capable components need the LARB property */ comp->larb_dev = NULL; if (type != MTK_DISP_OVL && diff --git a/drivers/gpu/drm/mediatek/mtk_drm_ddp_comp.h b/drivers/gpu/drm/mediatek/mtk_drm_ddp_comp.h index 22a33ee451c4..0828cf8bf85c 100644 --- a/drivers/gpu/drm/mediatek/mtk_drm_ddp_comp.h +++ b/drivers/gpu/drm/mediatek/mtk_drm_ddp_comp.h @@ -36,11 +36,13 @@ enum mtk_ddp_comp_type { MTK_DISP_PWM, MTK_DISP_MUTEX, MTK_DISP_OD, + MTK_DISP_BLS, MTK_DDP_COMP_TYPE_MAX, }; enum mtk_ddp_comp_id { DDP_COMPONENT_AAL, + DDP_COMPONENT_BLS, DDP_COMPONENT_COLOR0, DDP_COMPONENT_COLOR1, DDP_COMPONENT_DPI0, diff --git a/drivers/gpu/drm/mediatek/mtk_drm_drv.c b/drivers/gpu/drm/mediatek/mtk_drm_drv.c index f5a1fd9b3ecc..f6c8ec4c7dbc 100644 --- a/drivers/gpu/drm/mediatek/mtk_drm_drv.c +++ b/drivers/gpu/drm/mediatek/mtk_drm_drv.c @@ -128,7 +128,20 @@ static const struct drm_mode_config_funcs mtk_drm_mode_config_funcs = { .atomic_commit = mtk_atomic_commit, }; -static const enum mtk_ddp_comp_id mtk_ddp_main[] = { +static const enum mtk_ddp_comp_id mt2701_mtk_ddp_main[] = { + DDP_COMPONENT_OVL0, + DDP_COMPONENT_RDMA0, + DDP_COMPONENT_COLOR0, + DDP_COMPONENT_BLS, + DDP_COMPONENT_DSI0, +}; + +static const enum mtk_ddp_comp_id mt2701_mtk_ddp_ext[] = { + DDP_COMPONENT_RDMA1, + DDP_COMPONENT_DPI0, +}; + +static const enum mtk_ddp_comp_id mt8173_mtk_ddp_main[] = { DDP_COMPONENT_OVL0, DDP_COMPONENT_COLOR0, DDP_COMPONENT_AAL, @@ -139,7 +152,7 @@ static const enum mtk_ddp_comp_id mtk_ddp_main[] = { DDP_COMPONENT_PWM0, }; -static const enum mtk_ddp_comp_id mtk_ddp_ext[] = { +static const enum mtk_ddp_comp_id mt8173_mtk_ddp_ext[] = { DDP_COMPONENT_OVL1, DDP_COMPONENT_COLOR1, DDP_COMPONENT_GAMMA, @@ -147,6 +160,21 @@ static const enum mtk_ddp_comp_id mtk_ddp_ext[] = { DDP_COMPONENT_DPI0, }; +static const struct mtk_mmsys_driver_data mt2701_mmsys_driver_data = { + .main_path = mt2701_mtk_ddp_main, + .main_len = ARRAY_SIZE(mt2701_mtk_ddp_main), + .ext_path = mt2701_mtk_ddp_ext, + .ext_len = ARRAY_SIZE(mt2701_mtk_ddp_ext), + .shadow_register = true, +}; + +static const struct mtk_mmsys_driver_data mt8173_mmsys_driver_data = { + .main_path = mt8173_mtk_ddp_main, + .main_len = ARRAY_SIZE(mt8173_mtk_ddp_main), + .ext_path = mt8173_mtk_ddp_ext, + .ext_len = ARRAY_SIZE(mt8173_mtk_ddp_ext), +}; + static int mtk_drm_kms_init(struct drm_device *drm) { struct mtk_drm_private *private = drm->dev_private; @@ -189,17 +217,19 @@ static int mtk_drm_kms_init(struct drm_device *drm) * and each statically assigned to a crtc: * OVL0 -> COLOR0 -> AAL -> OD -> RDMA0 -> UFOE -> DSI0 ... */ - ret = mtk_drm_crtc_create(drm, mtk_ddp_main, ARRAY_SIZE(mtk_ddp_main)); + ret = mtk_drm_crtc_create(drm, private->data->main_path, + private->data->main_len); if (ret < 0) goto err_component_unbind; /* ... and OVL1 -> COLOR1 -> GAMMA -> RDMA1 -> DPI0. */ - ret = mtk_drm_crtc_create(drm, mtk_ddp_ext, ARRAY_SIZE(mtk_ddp_ext)); + ret = mtk_drm_crtc_create(drm, private->data->ext_path, + private->data->ext_len); if (ret < 0) goto err_component_unbind; /* Use OVL device for all DMA memory allocations */ - np = private->comp_node[mtk_ddp_main[0]] ?: - private->comp_node[mtk_ddp_ext[0]]; + np = private->comp_node[private->data->main_path[0]] ?: + private->comp_node[private->data->ext_path[0]]; pdev = of_find_device_by_node(np); if (!pdev) { ret = -ENODEV; @@ -328,16 +358,22 @@ static const struct component_master_ops mtk_drm_ops = { }; static const struct of_device_id mtk_ddp_comp_dt_ids[] = { + { .compatible = "mediatek,mt2701-disp-ovl", .data = (void *)MTK_DISP_OVL }, { .compatible = "mediatek,mt8173-disp-ovl", .data = (void *)MTK_DISP_OVL }, + { .compatible = "mediatek,mt2701-disp-rdma", .data = (void *)MTK_DISP_RDMA }, { .compatible = "mediatek,mt8173-disp-rdma", .data = (void *)MTK_DISP_RDMA }, { .compatible = "mediatek,mt8173-disp-wdma", .data = (void *)MTK_DISP_WDMA }, + { .compatible = "mediatek,mt2701-disp-color", .data = (void *)MTK_DISP_COLOR }, { .compatible = "mediatek,mt8173-disp-color", .data = (void *)MTK_DISP_COLOR }, { .compatible = "mediatek,mt8173-disp-aal", .data = (void *)MTK_DISP_AAL}, { .compatible = "mediatek,mt8173-disp-gamma", .data = (void *)MTK_DISP_GAMMA, }, { .compatible = "mediatek,mt8173-disp-ufoe", .data = (void *)MTK_DISP_UFOE }, + { .compatible = "mediatek,mt2701-dsi", .data = (void *)MTK_DSI }, { .compatible = "mediatek,mt8173-dsi", .data = (void *)MTK_DSI }, { .compatible = "mediatek,mt8173-dpi", .data = (void *)MTK_DPI }, + { .compatible = "mediatek,mt2701-disp-mutex", .data = (void *)MTK_DISP_MUTEX }, { .compatible = "mediatek,mt8173-disp-mutex", .data = (void *)MTK_DISP_MUTEX }, + { .compatible = "mediatek,mt2701-disp-pwm", .data = (void *)MTK_DISP_BLS }, { .compatible = "mediatek,mt8173-disp-pwm", .data = (void *)MTK_DISP_PWM }, { .compatible = "mediatek,mt8173-disp-od", .data = (void *)MTK_DISP_OD }, { } @@ -359,6 +395,7 @@ static int mtk_drm_probe(struct platform_device *pdev) mutex_init(&private->commit.lock); INIT_WORK(&private->commit.work, mtk_atomic_work); + private->data = of_device_get_match_data(dev); mem = platform_get_resource(pdev, IORESOURCE_MEM, 0); private->config_regs = devm_ioremap_resource(dev, mem); @@ -510,7 +547,10 @@ static SIMPLE_DEV_PM_OPS(mtk_drm_pm_ops, mtk_drm_sys_suspend, mtk_drm_sys_resume); static const struct of_device_id mtk_drm_of_ids[] = { - { .compatible = "mediatek,mt8173-mmsys", }, + { .compatible = "mediatek,mt2701-mmsys", + .data = &mt2701_mmsys_driver_data}, + { .compatible = "mediatek,mt8173-mmsys", + .data = &mt8173_mmsys_driver_data}, { } }; diff --git a/drivers/gpu/drm/mediatek/mtk_drm_drv.h b/drivers/gpu/drm/mediatek/mtk_drm_drv.h index df322a7a5fcb..aef8747d810b 100644 --- a/drivers/gpu/drm/mediatek/mtk_drm_drv.h +++ b/drivers/gpu/drm/mediatek/mtk_drm_drv.h @@ -28,6 +28,14 @@ struct drm_fb_helper; struct drm_property; struct regmap; +struct mtk_mmsys_driver_data { + const enum mtk_ddp_comp_id *main_path; + unsigned int main_len; + const enum mtk_ddp_comp_id *ext_path; + unsigned int ext_len; + bool shadow_register; +}; + struct mtk_drm_private { struct drm_device *drm; struct device *dma_dev; @@ -39,6 +47,7 @@ struct mtk_drm_private { void __iomem *config_regs; struct device_node *comp_node[DDP_COMPONENT_ID_MAX]; struct mtk_ddp_comp *ddp_comp[DDP_COMPONENT_ID_MAX]; + const struct mtk_mmsys_driver_data *data; struct { struct drm_atomic_state *state; diff --git a/drivers/gpu/drm/mediatek/mtk_dsi.c b/drivers/gpu/drm/mediatek/mtk_dsi.c index 4d837b29ee9e..808b995a990f 100644 --- a/drivers/gpu/drm/mediatek/mtk_dsi.c +++ b/drivers/gpu/drm/mediatek/mtk_dsi.c @@ -19,19 +19,28 @@ #include <drm/drm_of.h> #include <linux/clk.h> #include <linux/component.h> +#include <linux/irq.h> #include <linux/of.h> #include <linux/of_platform.h> #include <linux/phy/phy.h> #include <linux/platform_device.h> +#include <video/mipi_display.h> #include <video/videomode.h> #include "mtk_drm_ddp_comp.h" -#define DSI_VIDEO_FIFO_DEPTH (1920 / 4) -#define DSI_HOST_FIFO_DEPTH 64 - #define DSI_START 0x00 +#define DSI_INTEN 0x08 + +#define DSI_INTSTA 0x0c +#define LPRX_RD_RDY_INT_FLAG BIT(0) +#define CMD_DONE_INT_FLAG BIT(1) +#define TE_RDY_INT_FLAG BIT(2) +#define VM_DONE_INT_FLAG BIT(3) +#define EXT_TE_RDY_INT_FLAG BIT(4) +#define DSI_BUSY BIT(31) + #define DSI_CON_CTRL 0x10 #define DSI_RESET BIT(0) #define DSI_EN BIT(1) @@ -46,7 +55,7 @@ #define MIX_MODE BIT(17) #define DSI_TXRX_CTRL 0x18 -#define VC_NUM (2 << 0) +#define VC_NUM BIT(1) #define LANE_NUM (0xf << 2) #define DIS_EOT BIT(6) #define NULL_EN BIT(7) @@ -72,8 +81,19 @@ #define DSI_HBP_WC 0x54 #define DSI_HFP_WC 0x58 +#define DSI_CMDQ_SIZE 0x60 +#define CMDQ_SIZE 0x3f + #define DSI_HSTX_CKL_WC 0x64 +#define DSI_RX_DATA0 0x74 +#define DSI_RX_DATA1 0x78 +#define DSI_RX_DATA2 0x7c +#define DSI_RX_DATA3 0x80 + +#define DSI_RACK 0x84 +#define RACK BIT(0) + #define DSI_PHY_LCCON 0x104 #define LC_HS_TX_EN BIT(0) #define LC_ULPM_EN BIT(1) @@ -106,6 +126,19 @@ #define CLK_HS_POST (0xff << 8) #define CLK_HS_EXIT (0xff << 16) +#define DSI_VM_CMD_CON 0x130 +#define VM_CMD_EN BIT(0) +#define TS_VFP_EN BIT(5) + +#define DSI_CMDQ0 0x180 +#define CONFIG (0xff << 0) +#define SHORT_PACKET 0 +#define LONG_PACKET 2 +#define BTA BIT(2) +#define DATA_ID (0xff << 8) +#define DATA_0 (0xff << 16) +#define DATA_1 (0xff << 24) + #define T_LPX 5 #define T_HS_PREP 6 #define T_HS_TRAIL 8 @@ -114,6 +147,12 @@ #define NS_TO_CYCLE(n, c) ((n) / (c) + (((n) % (c)) ? 1 : 0)) +#define MTK_DSI_HOST_IS_READ(type) \ + ((type == MIPI_DSI_GENERIC_READ_REQUEST_0_PARAM) || \ + (type == MIPI_DSI_GENERIC_READ_REQUEST_1_PARAM) || \ + (type == MIPI_DSI_GENERIC_READ_REQUEST_2_PARAM) || \ + (type == MIPI_DSI_DCS_READ)) + struct phy; struct mtk_dsi { @@ -140,6 +179,8 @@ struct mtk_dsi { struct videomode vm; int refcount; bool enabled; + u32 irq_data; + wait_queue_head_t irq_wait_queue; }; static inline struct mtk_dsi *encoder_to_dsi(struct drm_encoder *e) @@ -164,7 +205,7 @@ static void mtk_dsi_mask(struct mtk_dsi *dsi, u32 offset, u32 mask, u32 data) writel((temp & ~mask) | (data & mask), dsi->regs + offset); } -static void dsi_phy_timconfig(struct mtk_dsi *dsi) +static void mtk_dsi_phy_timconfig(struct mtk_dsi *dsi) { u32 timcon0, timcon1, timcon2, timcon3; u32 ui, cycle_time; @@ -196,118 +237,39 @@ static void mtk_dsi_disable(struct mtk_dsi *dsi) mtk_dsi_mask(dsi, DSI_CON_CTRL, DSI_EN, 0); } -static void mtk_dsi_reset(struct mtk_dsi *dsi) +static void mtk_dsi_reset_engine(struct mtk_dsi *dsi) { mtk_dsi_mask(dsi, DSI_CON_CTRL, DSI_RESET, DSI_RESET); mtk_dsi_mask(dsi, DSI_CON_CTRL, DSI_RESET, 0); } -static int mtk_dsi_poweron(struct mtk_dsi *dsi) -{ - struct device *dev = dsi->dev; - int ret; - u64 pixel_clock, total_bits; - u32 htotal, htotal_bits, bit_per_pixel, overhead_cycles, overhead_bits; - - if (++dsi->refcount != 1) - return 0; - - switch (dsi->format) { - case MIPI_DSI_FMT_RGB565: - bit_per_pixel = 16; - break; - case MIPI_DSI_FMT_RGB666_PACKED: - bit_per_pixel = 18; - break; - case MIPI_DSI_FMT_RGB666: - case MIPI_DSI_FMT_RGB888: - default: - bit_per_pixel = 24; - break; - } - - /** - * vm.pixelclock is in kHz, pixel_clock unit is Hz, so multiply by 1000 - * htotal_time = htotal * byte_per_pixel / num_lanes - * overhead_time = lpx + hs_prepare + hs_zero + hs_trail + hs_exit - * mipi_ratio = (htotal_time + overhead_time) / htotal_time - * data_rate = pixel_clock * bit_per_pixel * mipi_ratio / num_lanes; - */ - pixel_clock = dsi->vm.pixelclock * 1000; - htotal = dsi->vm.hactive + dsi->vm.hback_porch + dsi->vm.hfront_porch + - dsi->vm.hsync_len; - htotal_bits = htotal * bit_per_pixel; - - overhead_cycles = T_LPX + T_HS_PREP + T_HS_ZERO + T_HS_TRAIL + - T_HS_EXIT; - overhead_bits = overhead_cycles * dsi->lanes * 8; - total_bits = htotal_bits + overhead_bits; - - dsi->data_rate = DIV_ROUND_UP_ULL(pixel_clock * total_bits, - htotal * dsi->lanes); - - ret = clk_set_rate(dsi->hs_clk, dsi->data_rate); - if (ret < 0) { - dev_err(dev, "Failed to set data rate: %d\n", ret); - goto err_refcount; - } - - phy_power_on(dsi->phy); - - ret = clk_prepare_enable(dsi->engine_clk); - if (ret < 0) { - dev_err(dev, "Failed to enable engine clock: %d\n", ret); - goto err_phy_power_off; - } - - ret = clk_prepare_enable(dsi->digital_clk); - if (ret < 0) { - dev_err(dev, "Failed to enable digital clock: %d\n", ret); - goto err_disable_engine_clk; - } - - mtk_dsi_enable(dsi); - mtk_dsi_reset(dsi); - dsi_phy_timconfig(dsi); - - return 0; - -err_disable_engine_clk: - clk_disable_unprepare(dsi->engine_clk); -err_phy_power_off: - phy_power_off(dsi->phy); -err_refcount: - dsi->refcount--; - return ret; -} - -static void dsi_clk_ulp_mode_enter(struct mtk_dsi *dsi) +static void mtk_dsi_clk_ulp_mode_enter(struct mtk_dsi *dsi) { mtk_dsi_mask(dsi, DSI_PHY_LCCON, LC_HS_TX_EN, 0); mtk_dsi_mask(dsi, DSI_PHY_LCCON, LC_ULPM_EN, 0); } -static void dsi_clk_ulp_mode_leave(struct mtk_dsi *dsi) +static void mtk_dsi_clk_ulp_mode_leave(struct mtk_dsi *dsi) { mtk_dsi_mask(dsi, DSI_PHY_LCCON, LC_ULPM_EN, 0); mtk_dsi_mask(dsi, DSI_PHY_LCCON, LC_WAKEUP_EN, LC_WAKEUP_EN); mtk_dsi_mask(dsi, DSI_PHY_LCCON, LC_WAKEUP_EN, 0); } -static void dsi_lane0_ulp_mode_enter(struct mtk_dsi *dsi) +static void mtk_dsi_lane0_ulp_mode_enter(struct mtk_dsi *dsi) { mtk_dsi_mask(dsi, DSI_PHY_LD0CON, LD0_HS_TX_EN, 0); mtk_dsi_mask(dsi, DSI_PHY_LD0CON, LD0_ULPM_EN, 0); } -static void dsi_lane0_ulp_mode_leave(struct mtk_dsi *dsi) +static void mtk_dsi_lane0_ulp_mode_leave(struct mtk_dsi *dsi) { mtk_dsi_mask(dsi, DSI_PHY_LD0CON, LD0_ULPM_EN, 0); mtk_dsi_mask(dsi, DSI_PHY_LD0CON, LD0_WAKEUP_EN, LD0_WAKEUP_EN); mtk_dsi_mask(dsi, DSI_PHY_LD0CON, LD0_WAKEUP_EN, 0); } -static bool dsi_clk_hs_state(struct mtk_dsi *dsi) +static bool mtk_dsi_clk_hs_state(struct mtk_dsi *dsi) { u32 tmp_reg1; @@ -315,30 +277,37 @@ static bool dsi_clk_hs_state(struct mtk_dsi *dsi) return ((tmp_reg1 & LC_HS_TX_EN) == 1) ? true : false; } -static void dsi_clk_hs_mode(struct mtk_dsi *dsi, bool enter) +static void mtk_dsi_clk_hs_mode(struct mtk_dsi *dsi, bool enter) { - if (enter && !dsi_clk_hs_state(dsi)) + if (enter && !mtk_dsi_clk_hs_state(dsi)) mtk_dsi_mask(dsi, DSI_PHY_LCCON, LC_HS_TX_EN, LC_HS_TX_EN); - else if (!enter && dsi_clk_hs_state(dsi)) + else if (!enter && mtk_dsi_clk_hs_state(dsi)) mtk_dsi_mask(dsi, DSI_PHY_LCCON, LC_HS_TX_EN, 0); } -static void dsi_set_mode(struct mtk_dsi *dsi) +static void mtk_dsi_set_mode(struct mtk_dsi *dsi) { u32 vid_mode = CMD_MODE; if (dsi->mode_flags & MIPI_DSI_MODE_VIDEO) { - vid_mode = SYNC_PULSE_MODE; - - if ((dsi->mode_flags & MIPI_DSI_MODE_VIDEO_BURST) && - !(dsi->mode_flags & MIPI_DSI_MODE_VIDEO_SYNC_PULSE)) + if (dsi->mode_flags & MIPI_DSI_MODE_VIDEO_BURST) vid_mode = BURST_MODE; + else if (dsi->mode_flags & MIPI_DSI_MODE_VIDEO_SYNC_PULSE) + vid_mode = SYNC_PULSE_MODE; + else + vid_mode = SYNC_EVENT_MODE; } writel(vid_mode, dsi->regs + DSI_MODE_CTRL); } -static void dsi_ps_control_vact(struct mtk_dsi *dsi) +static void mtk_dsi_set_vm_cmd(struct mtk_dsi *dsi) +{ + mtk_dsi_mask(dsi, DSI_VM_CMD_CON, VM_CMD_EN, VM_CMD_EN); + mtk_dsi_mask(dsi, DSI_VM_CMD_CON, TS_VFP_EN, TS_VFP_EN); +} + +static void mtk_dsi_ps_control_vact(struct mtk_dsi *dsi) { struct videomode *vm = &dsi->vm; u32 dsi_buf_bpp, ps_wc; @@ -372,7 +341,7 @@ static void dsi_ps_control_vact(struct mtk_dsi *dsi) writel(ps_wc, dsi->regs + DSI_HSTX_CKL_WC); } -static void dsi_rxtx_control(struct mtk_dsi *dsi) +static void mtk_dsi_rxtx_control(struct mtk_dsi *dsi) { u32 tmp_reg; @@ -394,12 +363,15 @@ static void dsi_rxtx_control(struct mtk_dsi *dsi) break; } + tmp_reg |= (dsi->mode_flags & MIPI_DSI_CLOCK_NON_CONTINUOUS) << 6; + tmp_reg |= (dsi->mode_flags & MIPI_DSI_MODE_EOT_PACKET) >> 3; + writel(tmp_reg, dsi->regs + DSI_TXRX_CTRL); } -static void dsi_ps_control(struct mtk_dsi *dsi) +static void mtk_dsi_ps_control(struct mtk_dsi *dsi) { - unsigned int dsi_tmp_buf_bpp; + u32 dsi_tmp_buf_bpp; u32 tmp_reg; switch (dsi->format) { @@ -429,12 +401,12 @@ static void dsi_ps_control(struct mtk_dsi *dsi) writel(tmp_reg, dsi->regs + DSI_PSCTRL); } -static void dsi_config_vdo_timing(struct mtk_dsi *dsi) +static void mtk_dsi_config_vdo_timing(struct mtk_dsi *dsi) { - unsigned int horizontal_sync_active_byte; - unsigned int horizontal_backporch_byte; - unsigned int horizontal_frontporch_byte; - unsigned int dsi_tmp_buf_bpp; + u32 horizontal_sync_active_byte; + u32 horizontal_backporch_byte; + u32 horizontal_frontporch_byte; + u32 dsi_tmp_buf_bpp; struct videomode *vm = &dsi->vm; @@ -463,7 +435,7 @@ static void dsi_config_vdo_timing(struct mtk_dsi *dsi) writel(horizontal_backporch_byte, dsi->regs + DSI_HBP_WC); writel(horizontal_frontporch_byte, dsi->regs + DSI_HFP_WC); - dsi_ps_control(dsi); + mtk_dsi_ps_control(dsi); } static void mtk_dsi_start(struct mtk_dsi *dsi) @@ -472,6 +444,184 @@ static void mtk_dsi_start(struct mtk_dsi *dsi) writel(1, dsi->regs + DSI_START); } +static void mtk_dsi_stop(struct mtk_dsi *dsi) +{ + writel(0, dsi->regs + DSI_START); +} + +static void mtk_dsi_set_cmd_mode(struct mtk_dsi *dsi) +{ + writel(CMD_MODE, dsi->regs + DSI_MODE_CTRL); +} + +static void mtk_dsi_set_interrupt_enable(struct mtk_dsi *dsi) +{ + u32 inten = LPRX_RD_RDY_INT_FLAG | CMD_DONE_INT_FLAG | VM_DONE_INT_FLAG; + + writel(inten, dsi->regs + DSI_INTEN); +} + +static void mtk_dsi_irq_data_set(struct mtk_dsi *dsi, u32 irq_bit) +{ + dsi->irq_data |= irq_bit; +} + +static void mtk_dsi_irq_data_clear(struct mtk_dsi *dsi, u32 irq_bit) +{ + dsi->irq_data &= ~irq_bit; +} + +static s32 mtk_dsi_wait_for_irq_done(struct mtk_dsi *dsi, u32 irq_flag, + unsigned int timeout) +{ + s32 ret = 0; + unsigned long jiffies = msecs_to_jiffies(timeout); + + ret = wait_event_interruptible_timeout(dsi->irq_wait_queue, + dsi->irq_data & irq_flag, + jiffies); + if (ret == 0) { + DRM_WARN("Wait DSI IRQ(0x%08x) Timeout\n", irq_flag); + + mtk_dsi_enable(dsi); + mtk_dsi_reset_engine(dsi); + } + + return ret; +} + +static irqreturn_t mtk_dsi_irq(int irq, void *dev_id) +{ + struct mtk_dsi *dsi = dev_id; + u32 status, tmp; + u32 flag = LPRX_RD_RDY_INT_FLAG | CMD_DONE_INT_FLAG | VM_DONE_INT_FLAG; + + status = readl(dsi->regs + DSI_INTSTA) & flag; + + if (status) { + do { + mtk_dsi_mask(dsi, DSI_RACK, RACK, RACK); + tmp = readl(dsi->regs + DSI_INTSTA); + } while (tmp & DSI_BUSY); + + mtk_dsi_mask(dsi, DSI_INTSTA, status, 0); + mtk_dsi_irq_data_set(dsi, status); + wake_up_interruptible(&dsi->irq_wait_queue); + } + + return IRQ_HANDLED; +} + +static s32 mtk_dsi_switch_to_cmd_mode(struct mtk_dsi *dsi, u8 irq_flag, u32 t) +{ + mtk_dsi_irq_data_clear(dsi, irq_flag); + mtk_dsi_set_cmd_mode(dsi); + + if (!mtk_dsi_wait_for_irq_done(dsi, irq_flag, t)) { + DRM_ERROR("failed to switch cmd mode\n"); + return -ETIME; + } else { + return 0; + } +} + +static int mtk_dsi_poweron(struct mtk_dsi *dsi) +{ + struct device *dev = dsi->dev; + int ret; + u64 pixel_clock, total_bits; + u32 htotal, htotal_bits, bit_per_pixel, overhead_cycles, overhead_bits; + + if (++dsi->refcount != 1) + return 0; + + switch (dsi->format) { + case MIPI_DSI_FMT_RGB565: + bit_per_pixel = 16; + break; + case MIPI_DSI_FMT_RGB666_PACKED: + bit_per_pixel = 18; + break; + case MIPI_DSI_FMT_RGB666: + case MIPI_DSI_FMT_RGB888: + default: + bit_per_pixel = 24; + break; + } + + /** + * vm.pixelclock is in kHz, pixel_clock unit is Hz, so multiply by 1000 + * htotal_time = htotal * byte_per_pixel / num_lanes + * overhead_time = lpx + hs_prepare + hs_zero + hs_trail + hs_exit + * mipi_ratio = (htotal_time + overhead_time) / htotal_time + * data_rate = pixel_clock * bit_per_pixel * mipi_ratio / num_lanes; + */ + pixel_clock = dsi->vm.pixelclock * 1000; + htotal = dsi->vm.hactive + dsi->vm.hback_porch + dsi->vm.hfront_porch + + dsi->vm.hsync_len; + htotal_bits = htotal * bit_per_pixel; + + overhead_cycles = T_LPX + T_HS_PREP + T_HS_ZERO + T_HS_TRAIL + + T_HS_EXIT; + overhead_bits = overhead_cycles * dsi->lanes * 8; + total_bits = htotal_bits + overhead_bits; + + dsi->data_rate = DIV_ROUND_UP_ULL(pixel_clock * total_bits, + htotal * dsi->lanes); + + ret = clk_set_rate(dsi->hs_clk, dsi->data_rate); + if (ret < 0) { + dev_err(dev, "Failed to set data rate: %d\n", ret); + goto err_refcount; + } + + phy_power_on(dsi->phy); + + ret = clk_prepare_enable(dsi->engine_clk); + if (ret < 0) { + dev_err(dev, "Failed to enable engine clock: %d\n", ret); + goto err_phy_power_off; + } + + ret = clk_prepare_enable(dsi->digital_clk); + if (ret < 0) { + dev_err(dev, "Failed to enable digital clock: %d\n", ret); + goto err_disable_engine_clk; + } + + mtk_dsi_enable(dsi); + mtk_dsi_reset_engine(dsi); + mtk_dsi_phy_timconfig(dsi); + + mtk_dsi_rxtx_control(dsi); + mtk_dsi_ps_control_vact(dsi); + mtk_dsi_set_vm_cmd(dsi); + mtk_dsi_config_vdo_timing(dsi); + mtk_dsi_set_interrupt_enable(dsi); + + mtk_dsi_clk_ulp_mode_leave(dsi); + mtk_dsi_lane0_ulp_mode_leave(dsi); + mtk_dsi_clk_hs_mode(dsi, 0); + + if (dsi->panel) { + if (drm_panel_prepare(dsi->panel)) { + DRM_ERROR("failed to prepare the panel\n"); + goto err_disable_digital_clk; + } + } + + return 0; +err_disable_digital_clk: + clk_disable_unprepare(dsi->digital_clk); +err_disable_engine_clk: + clk_disable_unprepare(dsi->engine_clk); +err_phy_power_off: + phy_power_off(dsi->phy); +err_refcount: + dsi->refcount--; + return ret; +} + static void mtk_dsi_poweroff(struct mtk_dsi *dsi) { if (WARN_ON(dsi->refcount == 0)) @@ -480,8 +630,18 @@ static void mtk_dsi_poweroff(struct mtk_dsi *dsi) if (--dsi->refcount != 0) return; - dsi_lane0_ulp_mode_enter(dsi); - dsi_clk_ulp_mode_enter(dsi); + if (!mtk_dsi_switch_to_cmd_mode(dsi, VM_DONE_INT_FLAG, 500)) { + if (dsi->panel) { + if (drm_panel_unprepare(dsi->panel)) { + DRM_ERROR("failed to unprepare the panel\n"); + return; + } + } + } + + mtk_dsi_reset_engine(dsi); + mtk_dsi_lane0_ulp_mode_enter(dsi); + mtk_dsi_clk_ulp_mode_enter(dsi); mtk_dsi_disable(dsi); @@ -498,35 +658,30 @@ static void mtk_output_dsi_enable(struct mtk_dsi *dsi) if (dsi->enabled) return; - if (dsi->panel) { - if (drm_panel_prepare(dsi->panel)) { - DRM_ERROR("failed to setup the panel\n"); - return; - } - } - ret = mtk_dsi_poweron(dsi); if (ret < 0) { DRM_ERROR("failed to power on dsi\n"); return; } - dsi_rxtx_control(dsi); - - dsi_clk_ulp_mode_leave(dsi); - dsi_lane0_ulp_mode_leave(dsi); - dsi_clk_hs_mode(dsi, 0); - dsi_set_mode(dsi); - - dsi_ps_control_vact(dsi); - dsi_config_vdo_timing(dsi); - - dsi_set_mode(dsi); - dsi_clk_hs_mode(dsi, 1); + mtk_dsi_set_mode(dsi); + mtk_dsi_clk_hs_mode(dsi, 1); mtk_dsi_start(dsi); + if (dsi->panel) { + if (drm_panel_enable(dsi->panel)) { + DRM_ERROR("failed to enable the panel\n"); + goto err_dsi_power_off; + } + } + dsi->enabled = true; + + return; +err_dsi_power_off: + mtk_dsi_stop(dsi); + mtk_dsi_poweroff(dsi); } static void mtk_output_dsi_disable(struct mtk_dsi *dsi) @@ -541,6 +696,7 @@ static void mtk_output_dsi_disable(struct mtk_dsi *dsi) } } + mtk_dsi_stop(dsi); mtk_dsi_poweroff(dsi); dsi->enabled = false; @@ -742,9 +898,149 @@ static int mtk_dsi_host_detach(struct mipi_dsi_host *host, return 0; } +static void mtk_dsi_wait_for_idle(struct mtk_dsi *dsi) +{ + u32 timeout_ms = 500000; /* total 1s ~ 2s timeout */ + + while (timeout_ms--) { + if (!(readl(dsi->regs + DSI_INTSTA) & DSI_BUSY)) + break; + + usleep_range(2, 4); + } + + if (timeout_ms == 0) { + DRM_WARN("polling dsi wait not busy timeout!\n"); + + mtk_dsi_enable(dsi); + mtk_dsi_reset_engine(dsi); + } +} + +static u32 mtk_dsi_recv_cnt(u8 type, u8 *read_data) +{ + switch (type) { + case MIPI_DSI_RX_GENERIC_SHORT_READ_RESPONSE_1BYTE: + case MIPI_DSI_RX_DCS_SHORT_READ_RESPONSE_1BYTE: + return 1; + case MIPI_DSI_RX_GENERIC_SHORT_READ_RESPONSE_2BYTE: + case MIPI_DSI_RX_DCS_SHORT_READ_RESPONSE_2BYTE: + return 2; + case MIPI_DSI_RX_GENERIC_LONG_READ_RESPONSE: + case MIPI_DSI_RX_DCS_LONG_READ_RESPONSE: + return read_data[1] + read_data[2] * 16; + case MIPI_DSI_RX_ACKNOWLEDGE_AND_ERROR_REPORT: + DRM_INFO("type is 0x02, try again\n"); + break; + default: + DRM_INFO("type(0x%x) cannot be non-recognite\n", type); + break; + } + + return 0; +} + +static void mtk_dsi_cmdq(struct mtk_dsi *dsi, const struct mipi_dsi_msg *msg) +{ + const char *tx_buf = msg->tx_buf; + u8 config, cmdq_size, cmdq_off, type = msg->type; + u32 reg_val, cmdq_mask, i; + + if (MTK_DSI_HOST_IS_READ(type)) + config = BTA; + else + config = (msg->tx_len > 2) ? LONG_PACKET : SHORT_PACKET; + + if (msg->tx_len > 2) { + cmdq_size = 1 + (msg->tx_len + 3) / 4; + cmdq_off = 4; + cmdq_mask = CONFIG | DATA_ID | DATA_0 | DATA_1; + reg_val = (msg->tx_len << 16) | (type << 8) | config; + } else { + cmdq_size = 1; + cmdq_off = 2; + cmdq_mask = CONFIG | DATA_ID; + reg_val = (type << 8) | config; + } + + for (i = 0; i < msg->tx_len; i++) + writeb(tx_buf[i], dsi->regs + DSI_CMDQ0 + cmdq_off + i); + + mtk_dsi_mask(dsi, DSI_CMDQ0, cmdq_mask, reg_val); + mtk_dsi_mask(dsi, DSI_CMDQ_SIZE, CMDQ_SIZE, cmdq_size); +} + +static ssize_t mtk_dsi_host_send_cmd(struct mtk_dsi *dsi, + const struct mipi_dsi_msg *msg, u8 flag) +{ + mtk_dsi_wait_for_idle(dsi); + mtk_dsi_irq_data_clear(dsi, flag); + mtk_dsi_cmdq(dsi, msg); + mtk_dsi_start(dsi); + + if (!mtk_dsi_wait_for_irq_done(dsi, flag, 2000)) + return -ETIME; + else + return 0; +} + +static ssize_t mtk_dsi_host_transfer(struct mipi_dsi_host *host, + const struct mipi_dsi_msg *msg) +{ + struct mtk_dsi *dsi = host_to_dsi(host); + u32 recv_cnt, i; + u8 read_data[16]; + void *src_addr; + u8 irq_flag = CMD_DONE_INT_FLAG; + + if (readl(dsi->regs + DSI_MODE_CTRL) & MODE) { + DRM_ERROR("dsi engine is not command mode\n"); + return -EINVAL; + } + + if (MTK_DSI_HOST_IS_READ(msg->type)) + irq_flag |= LPRX_RD_RDY_INT_FLAG; + + if (mtk_dsi_host_send_cmd(dsi, msg, irq_flag) < 0) + return -ETIME; + + if (!MTK_DSI_HOST_IS_READ(msg->type)) + return 0; + + if (!msg->rx_buf) { + DRM_ERROR("dsi receive buffer size may be NULL\n"); + return -EINVAL; + } + + for (i = 0; i < 16; i++) + *(read_data + i) = readb(dsi->regs + DSI_RX_DATA0 + i); + + recv_cnt = mtk_dsi_recv_cnt(read_data[0], read_data); + + if (recv_cnt > 2) + src_addr = &read_data[4]; + else + src_addr = &read_data[1]; + + if (recv_cnt > 10) + recv_cnt = 10; + + if (recv_cnt > msg->rx_len) + recv_cnt = msg->rx_len; + + if (recv_cnt) + memcpy(msg->rx_buf, src_addr, recv_cnt); + + DRM_INFO("dsi get %d byte data from the panel address(0x%x)\n", + recv_cnt, *((u8 *)(msg->tx_buf))); + + return recv_cnt; +} + static const struct mipi_dsi_host_ops mtk_dsi_ops = { .attach = mtk_dsi_host_attach, .detach = mtk_dsi_host_detach, + .transfer = mtk_dsi_host_transfer, }; static int mtk_dsi_bind(struct device *dev, struct device *master, void *data) @@ -802,6 +1098,7 @@ static int mtk_dsi_probe(struct platform_device *pdev) struct mtk_dsi *dsi; struct device *dev = &pdev->dev; struct resource *regs; + int irq_num; int comp_id; int ret; @@ -866,6 +1163,22 @@ static int mtk_dsi_probe(struct platform_device *pdev) return ret; } + irq_num = platform_get_irq(pdev, 0); + if (irq_num < 0) { + dev_err(&pdev->dev, "failed to request dsi irq resource\n"); + return -EPROBE_DEFER; + } + + irq_set_status_flags(irq_num, IRQ_TYPE_LEVEL_LOW); + ret = devm_request_irq(&pdev->dev, irq_num, mtk_dsi_irq, + IRQF_TRIGGER_LOW, dev_name(&pdev->dev), dsi); + if (ret) { + dev_err(&pdev->dev, "failed to request mediatek dsi irq\n"); + return -EPROBE_DEFER; + } + + init_waitqueue_head(&dsi->irq_wait_queue); + platform_set_drvdata(pdev, dsi); return component_add(&pdev->dev, &mtk_dsi_component_ops); @@ -882,6 +1195,7 @@ static int mtk_dsi_remove(struct platform_device *pdev) } static const struct of_device_id mtk_dsi_of_match[] = { + { .compatible = "mediatek,mt2701-dsi" }, { .compatible = "mediatek,mt8173-dsi" }, { }, }; diff --git a/drivers/gpu/drm/mediatek/mtk_mipi_tx.c b/drivers/gpu/drm/mediatek/mtk_mipi_tx.c index 1c366f8cb2d0..90e913108950 100644 --- a/drivers/gpu/drm/mediatek/mtk_mipi_tx.c +++ b/drivers/gpu/drm/mediatek/mtk_mipi_tx.c @@ -16,6 +16,7 @@ #include <linux/delay.h> #include <linux/io.h> #include <linux/module.h> +#include <linux/of_device.h> #include <linux/platform_device.h> #include <linux/phy/phy.h> @@ -87,6 +88,9 @@ #define MIPITX_DSI_PLL_CON2 0x58 +#define MIPITX_DSI_PLL_TOP 0x64 +#define RG_DSI_MPPLL_PRESERVE (0xff << 8) + #define MIPITX_DSI_PLL_PWR 0x68 #define RG_DSI_MPPLL_SDM_PWR_ON BIT(0) #define RG_DSI_MPPLL_SDM_ISO_EN BIT(1) @@ -123,10 +127,15 @@ #define SW_LNT2_HSTX_PRE_OE BIT(24) #define SW_LNT2_HSTX_OE BIT(25) +struct mtk_mipitx_data { + const u32 mppll_preserve; +}; + struct mtk_mipi_tx { struct device *dev; void __iomem *regs; - unsigned int data_rate; + u32 data_rate; + const struct mtk_mipitx_data *driver_data; struct clk_hw pll_hw; struct clk *pll; }; @@ -163,7 +172,7 @@ static void mtk_mipi_tx_update_bits(struct mtk_mipi_tx *mipi_tx, u32 offset, static int mtk_mipi_tx_pll_prepare(struct clk_hw *hw) { struct mtk_mipi_tx *mipi_tx = mtk_mipi_tx_from_clk_hw(hw); - unsigned int txdiv, txdiv0, txdiv1; + u8 txdiv, txdiv0, txdiv1; u64 pcw; dev_dbg(mipi_tx->dev, "prepare: %u Hz\n", mipi_tx->data_rate); @@ -243,6 +252,10 @@ static int mtk_mipi_tx_pll_prepare(struct clk_hw *hw) mtk_mipi_tx_clear_bits(mipi_tx, MIPITX_DSI_PLL_CON1, RG_DSI_MPPLL_SDM_SSC_EN); + mtk_mipi_tx_update_bits(mipi_tx, MIPITX_DSI_PLL_TOP, + RG_DSI_MPPLL_PRESERVE, + mipi_tx->driver_data->mppll_preserve); + return 0; } @@ -255,6 +268,9 @@ static void mtk_mipi_tx_pll_unprepare(struct clk_hw *hw) mtk_mipi_tx_clear_bits(mipi_tx, MIPITX_DSI_PLL_CON0, RG_DSI_MPPLL_PLL_EN); + mtk_mipi_tx_update_bits(mipi_tx, MIPITX_DSI_PLL_TOP, + RG_DSI_MPPLL_PRESERVE, 0); + mtk_mipi_tx_update_bits(mipi_tx, MIPITX_DSI_PLL_PWR, RG_DSI_MPPLL_SDM_ISO_EN | RG_DSI_MPPLL_SDM_PWR_ON, @@ -310,7 +326,7 @@ static const struct clk_ops mtk_mipi_tx_pll_ops = { static int mtk_mipi_tx_power_on_signal(struct phy *phy) { struct mtk_mipi_tx *mipi_tx = phy_get_drvdata(phy); - unsigned int reg; + u32 reg; for (reg = MIPITX_DSI_CLOCK_LANE; reg <= MIPITX_DSI_DATA_LANE3; reg += 4) @@ -341,7 +357,7 @@ static int mtk_mipi_tx_power_on(struct phy *phy) static void mtk_mipi_tx_power_off_signal(struct phy *phy) { struct mtk_mipi_tx *mipi_tx = phy_get_drvdata(phy); - unsigned int reg; + u32 reg; mtk_mipi_tx_set_bits(mipi_tx, MIPITX_DSI_TOP_CON, RG_DSI_PAD_TIE_LOW_EN); @@ -391,6 +407,7 @@ static int mtk_mipi_tx_probe(struct platform_device *pdev) if (!mipi_tx) return -ENOMEM; + mipi_tx->driver_data = of_device_get_match_data(dev); mem = platform_get_resource(pdev, IORESOURCE_MEM, 0); mipi_tx->regs = devm_ioremap_resource(dev, mem); if (IS_ERR(mipi_tx->regs)) { @@ -448,8 +465,19 @@ static int mtk_mipi_tx_remove(struct platform_device *pdev) return 0; } +static const struct mtk_mipitx_data mt2701_mipitx_data = { + .mppll_preserve = (3 << 8) +}; + +static const struct mtk_mipitx_data mt8173_mipitx_data = { + .mppll_preserve = (0 << 8) +}; + static const struct of_device_id mtk_mipi_tx_match[] = { - { .compatible = "mediatek,mt8173-mipi-tx", }, + { .compatible = "mediatek,mt2701-mipi-tx", + .data = &mt2701_mipitx_data }, + { .compatible = "mediatek,mt8173-mipi-tx", + .data = &mt8173_mipitx_data }, {}, }; diff --git a/drivers/gpu/drm/mgag200/mgag200_ttm.c b/drivers/gpu/drm/mgag200/mgag200_ttm.c index 657598bb1e6b..565a217b46f2 100644 --- a/drivers/gpu/drm/mgag200/mgag200_ttm.c +++ b/drivers/gpu/drm/mgag200/mgag200_ttm.c @@ -236,6 +236,7 @@ struct ttm_bo_driver mgag200_bo_driver = { .verify_access = mgag200_bo_verify_access, .io_mem_reserve = &mgag200_ttm_io_mem_reserve, .io_mem_free = &mgag200_ttm_io_mem_free, + .io_mem_pfn = ttm_bo_default_io_mem_pfn, }; int mgag200_mm_init(struct mga_device *mdev) diff --git a/drivers/gpu/drm/msm/adreno/a5xx_gpu.c b/drivers/gpu/drm/msm/adreno/a5xx_gpu.c index 4414cf73735d..36602ac7e248 100644 --- a/drivers/gpu/drm/msm/adreno/a5xx_gpu.c +++ b/drivers/gpu/drm/msm/adreno/a5xx_gpu.c @@ -1,4 +1,4 @@ -/* Copyright (c) 2016 The Linux Foundation. All rights reserved. +/* Copyright (c) 2016-2017 The Linux Foundation. All rights reserved. * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License version 2 and @@ -534,7 +534,7 @@ static void a5xx_destroy(struct msm_gpu *gpu) } if (a5xx_gpu->gpmu_bo) { - if (a5xx_gpu->gpmu_bo) + if (a5xx_gpu->gpmu_iova) msm_gem_put_iova(a5xx_gpu->gpmu_bo, gpu->id); drm_gem_object_unreference_unlocked(a5xx_gpu->gpmu_bo); } @@ -860,7 +860,9 @@ static const struct adreno_gpu_funcs funcs = { .idle = a5xx_idle, .irq = a5xx_irq, .destroy = a5xx_destroy, +#ifdef CONFIG_DEBUG_FS .show = a5xx_show, +#endif }, .get_timestamp = a5xx_get_timestamp, }; diff --git a/drivers/gpu/drm/msm/adreno/adreno_gpu.c b/drivers/gpu/drm/msm/adreno/adreno_gpu.c index c9bd1e6225f4..5ae65426b4e5 100644 --- a/drivers/gpu/drm/msm/adreno/adreno_gpu.c +++ b/drivers/gpu/drm/msm/adreno/adreno_gpu.c @@ -418,18 +418,27 @@ int adreno_gpu_init(struct drm_device *drm, struct platform_device *pdev, return 0; } -void adreno_gpu_cleanup(struct adreno_gpu *gpu) +void adreno_gpu_cleanup(struct adreno_gpu *adreno_gpu) { - if (gpu->memptrs_bo) { - if (gpu->memptrs) - msm_gem_put_vaddr(gpu->memptrs_bo); + struct msm_gpu *gpu = &adreno_gpu->base; + + if (adreno_gpu->memptrs_bo) { + if (adreno_gpu->memptrs) + msm_gem_put_vaddr(adreno_gpu->memptrs_bo); + + if (adreno_gpu->memptrs_iova) + msm_gem_put_iova(adreno_gpu->memptrs_bo, gpu->id); + + drm_gem_object_unreference_unlocked(adreno_gpu->memptrs_bo); + } + release_firmware(adreno_gpu->pm4); + release_firmware(adreno_gpu->pfp); - if (gpu->memptrs_iova) - msm_gem_put_iova(gpu->memptrs_bo, gpu->base.id); + msm_gpu_cleanup(gpu); - drm_gem_object_unreference_unlocked(gpu->memptrs_bo); + if (gpu->aspace) { + gpu->aspace->mmu->funcs->detach(gpu->aspace->mmu, + iommu_ports, ARRAY_SIZE(iommu_ports)); + msm_gem_address_space_destroy(gpu->aspace); } - release_firmware(gpu->pm4); - release_firmware(gpu->pfp); - msm_gpu_cleanup(&gpu->base); } diff --git a/drivers/gpu/drm/msm/dsi/dsi_manager.c b/drivers/gpu/drm/msm/dsi/dsi_manager.c index 921270ea6059..a879ffa534b4 100644 --- a/drivers/gpu/drm/msm/dsi/dsi_manager.c +++ b/drivers/gpu/drm/msm/dsi/dsi_manager.c @@ -171,7 +171,7 @@ dsi_mgr_phy_enable(int id, } } } else { - msm_dsi_host_reset_phy(mdsi->host); + msm_dsi_host_reset_phy(msm_dsi->host); ret = enable_phy(msm_dsi, src_pll_id, &shared_timings[id]); if (ret) return ret; diff --git a/drivers/gpu/drm/msm/hdmi/hdmi_audio.c b/drivers/gpu/drm/msm/hdmi/hdmi_audio.c index a54d3bb5baad..8177e8511afd 100644 --- a/drivers/gpu/drm/msm/hdmi/hdmi_audio.c +++ b/drivers/gpu/drm/msm/hdmi/hdmi_audio.c @@ -18,13 +18,6 @@ #include <linux/hdmi.h> #include "hdmi.h" - -/* Supported HDMI Audio channels */ -#define MSM_HDMI_AUDIO_CHANNEL_2 0 -#define MSM_HDMI_AUDIO_CHANNEL_4 1 -#define MSM_HDMI_AUDIO_CHANNEL_6 2 -#define MSM_HDMI_AUDIO_CHANNEL_8 3 - /* maps MSM_HDMI_AUDIO_CHANNEL_n consts used by audio driver to # of channels: */ static int nchannels[] = { 2, 4, 6, 8 }; diff --git a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_pipe.h b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_pipe.h index 611da7a660c9..238901987e00 100644 --- a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_pipe.h +++ b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_pipe.h @@ -18,7 +18,8 @@ #ifndef __MDP5_PIPE_H__ #define __MDP5_PIPE_H__ -#define SSPP_MAX (SSPP_RGB3 + 1) /* TODO: Add SSPP_MAX in mdp5.xml.h */ +/* TODO: Add SSPP_MAX in mdp5.xml.h */ +#define SSPP_MAX (SSPP_CURSOR1 + 1) /* represents a hw pipe, which is dynamically assigned to a plane */ struct mdp5_hw_pipe { diff --git a/drivers/gpu/drm/msm/msm_gem.c b/drivers/gpu/drm/msm/msm_gem.c index 59811f29607d..68e509b3b9e4 100644 --- a/drivers/gpu/drm/msm/msm_gem.c +++ b/drivers/gpu/drm/msm/msm_gem.c @@ -812,6 +812,12 @@ struct drm_gem_object *msm_gem_new(struct drm_device *dev, size = PAGE_ALIGN(size); + /* Disallow zero sized objects as they make the underlying + * infrastructure grumpy + */ + if (size == 0) + return ERR_PTR(-EINVAL); + ret = msm_gem_new_impl(dev, size, flags, NULL, &obj); if (ret) goto fail; diff --git a/drivers/gpu/drm/msm/msm_gpu.c b/drivers/gpu/drm/msm/msm_gpu.c index 99e05aacbee1..af5b6ba4095b 100644 --- a/drivers/gpu/drm/msm/msm_gpu.c +++ b/drivers/gpu/drm/msm/msm_gpu.c @@ -706,9 +706,6 @@ void msm_gpu_cleanup(struct msm_gpu *gpu) msm_ringbuffer_destroy(gpu->rb); } - if (gpu->aspace) - msm_gem_address_space_destroy(gpu->aspace); - if (gpu->fctx) msm_fence_context_free(gpu->fctx); } diff --git a/drivers/gpu/drm/nouveau/include/nvkm/core/msgqueue.h b/drivers/gpu/drm/nouveau/include/nvkm/core/msgqueue.h index fac0824197f1..bf3e532665fb 100644 --- a/drivers/gpu/drm/nouveau/include/nvkm/core/msgqueue.h +++ b/drivers/gpu/drm/nouveau/include/nvkm/core/msgqueue.h @@ -22,17 +22,14 @@ #ifndef __NVKM_CORE_MSGQUEUE_H #define __NVKM_CORE_MSGQUEUE_H - -#include <core/os.h> - -struct nvkm_falcon; +#include <subdev/secboot.h> struct nvkm_msgqueue; -enum nvkm_secboot_falcon; /* Hopefully we will never have firmware arguments larger than that... */ #define NVKM_MSGQUEUE_CMDLINE_SIZE 0x100 -int nvkm_msgqueue_new(u32, struct nvkm_falcon *, struct nvkm_msgqueue **); +int nvkm_msgqueue_new(u32, struct nvkm_falcon *, const struct nvkm_secboot *, + struct nvkm_msgqueue **); void nvkm_msgqueue_del(struct nvkm_msgqueue **); void nvkm_msgqueue_recv(struct nvkm_msgqueue *); int nvkm_msgqueue_reinit(struct nvkm_msgqueue *); @@ -41,7 +38,6 @@ int nvkm_msgqueue_reinit(struct nvkm_msgqueue *); void nvkm_msgqueue_write_cmdline(struct nvkm_msgqueue *, void *); /* interface to ACR unit running on falcon (NVIDIA signed firmware) */ -int nvkm_msgqueue_acr_boot_falcon(struct nvkm_msgqueue *, - enum nvkm_secboot_falcon); +int nvkm_msgqueue_acr_boot_falcons(struct nvkm_msgqueue *, unsigned long); #endif diff --git a/drivers/gpu/drm/nouveau/include/nvkm/core/tegra.h b/drivers/gpu/drm/nouveau/include/nvkm/core/tegra.h index e5c9b6268dcc..7c7d91cad09a 100644 --- a/drivers/gpu/drm/nouveau/include/nvkm/core/tegra.h +++ b/drivers/gpu/drm/nouveau/include/nvkm/core/tegra.h @@ -42,6 +42,10 @@ struct nvkm_device_tegra_func { * Whether the chip requires a reference clock */ bool require_ref_clk; + /* + * Whether the chip requires the VDD regulator + */ + bool require_vdd; }; int nvkm_device_tegra_new(const struct nvkm_device_tegra_func *, diff --git a/drivers/gpu/drm/nouveau/include/nvkm/engine/fifo.h b/drivers/gpu/drm/nouveau/include/nvkm/engine/fifo.h index 24efa900d8ca..f00527b36acc 100644 --- a/drivers/gpu/drm/nouveau/include/nvkm/engine/fifo.h +++ b/drivers/gpu/drm/nouveau/include/nvkm/engine/fifo.h @@ -68,4 +68,5 @@ int gm107_fifo_new(struct nvkm_device *, int, struct nvkm_fifo **); int gm200_fifo_new(struct nvkm_device *, int, struct nvkm_fifo **); int gm20b_fifo_new(struct nvkm_device *, int, struct nvkm_fifo **); int gp100_fifo_new(struct nvkm_device *, int, struct nvkm_fifo **); +int gp10b_fifo_new(struct nvkm_device *, int, struct nvkm_fifo **); #endif diff --git a/drivers/gpu/drm/nouveau/include/nvkm/engine/gr.h b/drivers/gpu/drm/nouveau/include/nvkm/engine/gr.h index 0a636833e0eb..c7944b19bed8 100644 --- a/drivers/gpu/drm/nouveau/include/nvkm/engine/gr.h +++ b/drivers/gpu/drm/nouveau/include/nvkm/engine/gr.h @@ -44,4 +44,6 @@ int gm200_gr_new(struct nvkm_device *, int, struct nvkm_gr **); int gm20b_gr_new(struct nvkm_device *, int, struct nvkm_gr **); int gp100_gr_new(struct nvkm_device *, int, struct nvkm_gr **); int gp102_gr_new(struct nvkm_device *, int, struct nvkm_gr **); +int gp107_gr_new(struct nvkm_device *, int, struct nvkm_gr **); +int gp10b_gr_new(struct nvkm_device *, int, struct nvkm_gr **); #endif diff --git a/drivers/gpu/drm/nouveau/include/nvkm/subdev/fb.h b/drivers/gpu/drm/nouveau/include/nvkm/subdev/fb.h index 891497a0fe3b..28d513fbf44c 100644 --- a/drivers/gpu/drm/nouveau/include/nvkm/subdev/fb.h +++ b/drivers/gpu/drm/nouveau/include/nvkm/subdev/fb.h @@ -97,6 +97,7 @@ int gm200_fb_new(struct nvkm_device *, int, struct nvkm_fb **); int gm20b_fb_new(struct nvkm_device *, int, struct nvkm_fb **); int gp100_fb_new(struct nvkm_device *, int, struct nvkm_fb **); int gp102_fb_new(struct nvkm_device *, int, struct nvkm_fb **); +int gp10b_fb_new(struct nvkm_device *, int, struct nvkm_fb **); #include <subdev/bios.h> #include <subdev/bios/ramcfg.h> diff --git a/drivers/gpu/drm/nouveau/include/nvkm/subdev/ibus.h b/drivers/gpu/drm/nouveau/include/nvkm/subdev/ibus.h index c4ecf255ff39..6e2b70bd2f41 100644 --- a/drivers/gpu/drm/nouveau/include/nvkm/subdev/ibus.h +++ b/drivers/gpu/drm/nouveau/include/nvkm/subdev/ibus.h @@ -7,4 +7,5 @@ int gf117_ibus_new(struct nvkm_device *, int, struct nvkm_subdev **); int gk104_ibus_new(struct nvkm_device *, int, struct nvkm_subdev **); int gk20a_ibus_new(struct nvkm_device *, int, struct nvkm_subdev **); int gm200_ibus_new(struct nvkm_device *, int, struct nvkm_subdev **); +int gp10b_ibus_new(struct nvkm_device *, int, struct nvkm_subdev **); #endif diff --git a/drivers/gpu/drm/nouveau/include/nvkm/subdev/mc.h b/drivers/gpu/drm/nouveau/include/nvkm/subdev/mc.h index e68ba636741b..58f10890c3b6 100644 --- a/drivers/gpu/drm/nouveau/include/nvkm/subdev/mc.h +++ b/drivers/gpu/drm/nouveau/include/nvkm/subdev/mc.h @@ -29,4 +29,5 @@ int gf100_mc_new(struct nvkm_device *, int, struct nvkm_mc **); int gk104_mc_new(struct nvkm_device *, int, struct nvkm_mc **); int gk20a_mc_new(struct nvkm_device *, int, struct nvkm_mc **); int gp100_mc_new(struct nvkm_device *, int, struct nvkm_mc **); +int gp10b_mc_new(struct nvkm_device *, int, struct nvkm_mc **); #endif diff --git a/drivers/gpu/drm/nouveau/include/nvkm/subdev/secboot.h b/drivers/gpu/drm/nouveau/include/nvkm/subdev/secboot.h index d6a4bdb6573b..59f3ba551681 100644 --- a/drivers/gpu/drm/nouveau/include/nvkm/subdev/secboot.h +++ b/drivers/gpu/drm/nouveau/include/nvkm/subdev/secboot.h @@ -55,10 +55,11 @@ struct nvkm_secboot { #define nvkm_secboot(p) container_of((p), struct nvkm_secboot, subdev) bool nvkm_secboot_is_managed(struct nvkm_secboot *, enum nvkm_secboot_falcon); -int nvkm_secboot_reset(struct nvkm_secboot *, enum nvkm_secboot_falcon); +int nvkm_secboot_reset(struct nvkm_secboot *, unsigned long); int gm200_secboot_new(struct nvkm_device *, int, struct nvkm_secboot **); int gm20b_secboot_new(struct nvkm_device *, int, struct nvkm_secboot **); int gp102_secboot_new(struct nvkm_device *, int, struct nvkm_secboot **); +int gp10b_secboot_new(struct nvkm_device *, int, struct nvkm_secboot **); #endif diff --git a/drivers/gpu/drm/nouveau/nouveau_bo.c b/drivers/gpu/drm/nouveau/nouveau_bo.c index 548f36d33924..e427f80344c4 100644 --- a/drivers/gpu/drm/nouveau/nouveau_bo.c +++ b/drivers/gpu/drm/nouveau/nouveau_bo.c @@ -1574,6 +1574,7 @@ struct ttm_bo_driver nouveau_bo_driver = { .fault_reserve_notify = &nouveau_ttm_fault_reserve_notify, .io_mem_reserve = &nouveau_ttm_io_mem_reserve, .io_mem_free = &nouveau_ttm_io_mem_free, + .io_mem_pfn = ttm_bo_default_io_mem_pfn, }; struct nvkm_vma * diff --git a/drivers/gpu/drm/nouveau/nouveau_platform.c b/drivers/gpu/drm/nouveau/nouveau_platform.c index 4c4cc2260257..1ada186fab77 100644 --- a/drivers/gpu/drm/nouveau/nouveau_platform.c +++ b/drivers/gpu/drm/nouveau/nouveau_platform.c @@ -53,13 +53,21 @@ static int nouveau_platform_remove(struct platform_device *pdev) #if IS_ENABLED(CONFIG_OF) static const struct nvkm_device_tegra_func gk20a_platform_data = { .iommu_bit = 34, + .require_vdd = true, }; static const struct nvkm_device_tegra_func gm20b_platform_data = { .iommu_bit = 34, + .require_vdd = true, .require_ref_clk = true, }; +static const struct nvkm_device_tegra_func gp10b_platform_data = { + .iommu_bit = 36, + /* power provided by generic PM domains */ + .require_vdd = false, +}; + static const struct of_device_id nouveau_platform_match[] = { { .compatible = "nvidia,gk20a", @@ -69,6 +77,10 @@ static const struct of_device_id nouveau_platform_match[] = { .compatible = "nvidia,gm20b", .data = &gm20b_platform_data, }, + { + .compatible = "nvidia,gp10b", + .data = &gp10b_platform_data, + }, { } }; diff --git a/drivers/gpu/drm/nouveau/nv50_display.c b/drivers/gpu/drm/nouveau/nv50_display.c index bf504788bce6..0e58537352fe 100644 --- a/drivers/gpu/drm/nouveau/nv50_display.c +++ b/drivers/gpu/drm/nouveau/nv50_display.c @@ -1002,7 +1002,6 @@ nv50_wndw_atomic_destroy_state(struct drm_plane *plane, { struct nv50_wndw_atom *asyw = nv50_wndw_atom(state); __drm_atomic_helper_plane_destroy_state(&asyw->state); - dma_fence_put(asyw->state.fence); kfree(asyw); } @@ -1014,7 +1013,6 @@ nv50_wndw_atomic_duplicate_state(struct drm_plane *plane) if (!(asyw = kmalloc(sizeof(*asyw), GFP_KERNEL))) return NULL; __drm_atomic_helper_plane_duplicate_state(plane, &asyw->state); - asyw->state.fence = NULL; asyw->interval = 1; asyw->sema = armw->sema; asyw->ntfy = armw->ntfy; @@ -2043,6 +2041,7 @@ nv50_head_atomic_check_mode(struct nv50_head *head, struct nv50_head_atom *asyh) u32 vbackp = (mode->vtotal - mode->vsync_end) * vscan / ilace; u32 hfrontp = mode->hsync_start - mode->hdisplay; u32 vfrontp = (mode->vsync_start - mode->vdisplay) * vscan / ilace; + u32 blankus; struct nv50_head_mode *m = &asyh->mode; m->h.active = mode->htotal; @@ -2056,9 +2055,10 @@ nv50_head_atomic_check_mode(struct nv50_head *head, struct nv50_head_atom *asyh) m->v.blanks = m->v.active - vfrontp - 1; /*XXX: Safe underestimate, even "0" works */ - m->v.blankus = (m->v.active - mode->vdisplay - 2) * m->h.active; - m->v.blankus *= 1000; - m->v.blankus /= mode->clock; + blankus = (m->v.active - mode->vdisplay - 2) * m->h.active; + blankus *= 1000; + blankus /= mode->clock; + m->v.blankus = blankus; if (mode->flags & DRM_MODE_FLAG_INTERLACE) { m->v.blank2e = m->v.active + m->v.synce + vbackp; diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c index 1076949b802a..b690bc12a5b7 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c @@ -714,7 +714,7 @@ nv4a_chipset = { .i2c = nv04_i2c_new, .imem = nv40_instmem_new, .mc = nv44_mc_new, - .mmu = nv44_mmu_new, + .mmu = nv04_mmu_new, .pci = nv40_pci_new, .therm = nv40_therm_new, .timer = nv41_timer_new, @@ -2201,8 +2201,6 @@ nv132_chipset = { .mc = gp100_mc_new, .mmu = gf100_mmu_new, .secboot = gp102_secboot_new, - .sec2 = gp102_sec2_new, - .nvdec = gp102_nvdec_new, .pci = gp100_pci_new, .pmu = gp102_pmu_new, .timer = gk20a_timer_new, @@ -2215,6 +2213,8 @@ nv132_chipset = { .dma = gf119_dma_new, .fifo = gp100_fifo_new, .gr = gp102_gr_new, + .nvdec = gp102_nvdec_new, + .sec2 = gp102_sec2_new, .sw = gf100_sw_new, }; @@ -2235,8 +2235,6 @@ nv134_chipset = { .mc = gp100_mc_new, .mmu = gf100_mmu_new, .secboot = gp102_secboot_new, - .sec2 = gp102_sec2_new, - .nvdec = gp102_nvdec_new, .pci = gp100_pci_new, .pmu = gp102_pmu_new, .timer = gk20a_timer_new, @@ -2249,6 +2247,8 @@ nv134_chipset = { .dma = gf119_dma_new, .fifo = gp100_fifo_new, .gr = gp102_gr_new, + .nvdec = gp102_nvdec_new, + .sec2 = gp102_sec2_new, .sw = gf100_sw_new, }; @@ -2269,8 +2269,6 @@ nv136_chipset = { .mc = gp100_mc_new, .mmu = gf100_mmu_new, .secboot = gp102_secboot_new, - .sec2 = gp102_sec2_new, - .nvdec = gp102_nvdec_new, .pci = gp100_pci_new, .pmu = gp102_pmu_new, .timer = gk20a_timer_new, @@ -2283,6 +2281,65 @@ nv136_chipset = { .dma = gf119_dma_new, .fifo = gp100_fifo_new, .gr = gp102_gr_new, + .nvdec = gp102_nvdec_new, + .sec2 = gp102_sec2_new, + .sw = gf100_sw_new, +}; + +static const struct nvkm_device_chip +nv137_chipset = { + .name = "GP107", + .bar = gf100_bar_new, + .bios = nvkm_bios_new, + .bus = gf100_bus_new, + .devinit = gm200_devinit_new, + .fb = gp102_fb_new, + .fuse = gm107_fuse_new, + .gpio = gk104_gpio_new, + .i2c = gm200_i2c_new, + .ibus = gm200_ibus_new, + .imem = nv50_instmem_new, + .ltc = gp100_ltc_new, + .mc = gp100_mc_new, + .mmu = gf100_mmu_new, + .secboot = gp102_secboot_new, + .pci = gp100_pci_new, + .pmu = gp102_pmu_new, + .timer = gk20a_timer_new, + .top = gk104_top_new, + .ce[0] = gp102_ce_new, + .ce[1] = gp102_ce_new, + .ce[2] = gp102_ce_new, + .ce[3] = gp102_ce_new, + .disp = gp102_disp_new, + .dma = gf119_dma_new, + .fifo = gp100_fifo_new, + .gr = gp107_gr_new, + .nvdec = gp102_nvdec_new, + .sec2 = gp102_sec2_new, + .sw = gf100_sw_new, +}; + +static const struct nvkm_device_chip +nv13b_chipset = { + .name = "GP10B", + .bar = gk20a_bar_new, + .bus = gf100_bus_new, + .fb = gp10b_fb_new, + .fuse = gm107_fuse_new, + .ibus = gp10b_ibus_new, + .imem = gk20a_instmem_new, + .ltc = gp100_ltc_new, + .mc = gp10b_mc_new, + .mmu = gf100_mmu_new, + .secboot = gp10b_secboot_new, + .pmu = gm20b_pmu_new, + .timer = gk20a_timer_new, + .top = gk104_top_new, + .ce[2] = gp102_ce_new, + .dma = gf119_dma_new, + .fifo = gp10b_fifo_new, + .gr = gp10b_gr_new, .sw = gf100_sw_new, }; @@ -2724,6 +2781,8 @@ nvkm_device_ctor(const struct nvkm_device_func *func, case 0x132: device->chip = &nv132_chipset; break; case 0x134: device->chip = &nv134_chipset; break; case 0x136: device->chip = &nv136_chipset; break; + case 0x137: device->chip = &nv137_chipset; break; + case 0x13b: device->chip = &nv13b_chipset; break; default: nvdev_error(device, "unknown chipset (%08x)\n", boot0); goto done; diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/device/tegra.c b/drivers/gpu/drm/nouveau/nvkm/engine/device/tegra.c index f2bc0b7d9b93..6474bd2a6d07 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/device/tegra.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/device/tegra.c @@ -28,9 +28,11 @@ nvkm_device_tegra_power_up(struct nvkm_device_tegra *tdev) { int ret; - ret = regulator_enable(tdev->vdd); - if (ret) - goto err_power; + if (tdev->vdd) { + ret = regulator_enable(tdev->vdd); + if (ret) + goto err_power; + } ret = clk_prepare_enable(tdev->clk); if (ret) @@ -67,7 +69,8 @@ err_clk_pwr: err_clk_ref: clk_disable_unprepare(tdev->clk); err_clk: - regulator_disable(tdev->vdd); + if (tdev->vdd) + regulator_disable(tdev->vdd); err_power: return ret; } @@ -75,6 +78,8 @@ err_power: static int nvkm_device_tegra_power_down(struct nvkm_device_tegra *tdev) { + int ret; + reset_control_assert(tdev->rst); udelay(10); @@ -84,7 +89,13 @@ nvkm_device_tegra_power_down(struct nvkm_device_tegra *tdev) clk_disable_unprepare(tdev->clk); udelay(10); - return regulator_disable(tdev->vdd); + if (tdev->vdd) { + ret = regulator_disable(tdev->vdd); + if (ret) + return ret; + } + + return 0; } static void @@ -264,10 +275,12 @@ nvkm_device_tegra_new(const struct nvkm_device_tegra_func *func, tdev->func = func; tdev->pdev = pdev; - tdev->vdd = devm_regulator_get(&pdev->dev, "vdd"); - if (IS_ERR(tdev->vdd)) { - ret = PTR_ERR(tdev->vdd); - goto free; + if (func->require_vdd) { + tdev->vdd = devm_regulator_get(&pdev->dev, "vdd"); + if (IS_ERR(tdev->vdd)) { + ret = PTR_ERR(tdev->vdd); + goto free; + } } tdev->rst = devm_reset_control_get(&pdev->dev, "gpu"); diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/Kbuild b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/Kbuild index 98651a43bc12..64e51838edf8 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/Kbuild +++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/Kbuild @@ -14,6 +14,7 @@ nvkm-y += nvkm/engine/fifo/gm107.o nvkm-y += nvkm/engine/fifo/gm200.o nvkm-y += nvkm/engine/fifo/gm20b.o nvkm-y += nvkm/engine/fifo/gp100.o +nvkm-y += nvkm/engine/fifo/gp10b.o nvkm-y += nvkm/engine/fifo/chan.o nvkm-y += nvkm/engine/fifo/channv50.o diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.h b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.h index 679f3ec311e9..44bff98d6725 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.h +++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.h @@ -83,4 +83,5 @@ extern const struct nvkm_enum gk104_fifo_fault_hubclient[]; extern const struct nvkm_enum gk104_fifo_fault_gpcclient[]; extern const struct nvkm_enum gm107_fifo_fault_engine[]; +extern const struct nvkm_enum gp100_fifo_fault_engine[]; #endif diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp100.c index b2635aea9f6e..41f16cf5a918 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp100.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp100.c @@ -24,7 +24,7 @@ #include "gk104.h" #include "changk104.h" -static const struct nvkm_enum +const struct nvkm_enum gp100_fifo_fault_engine[] = { { 0x01, "DISPLAY" }, { 0x03, "IFB", NULL, NVKM_ENGINE_IFB }, diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp10b.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp10b.c new file mode 100644 index 000000000000..4af96c3e69ff --- /dev/null +++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp10b.c @@ -0,0 +1,41 @@ +/* + * Copyright (c) 2017, NVIDIA CORPORATION. All rights reserved. + * + * 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 AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + */ +#include "gk104.h" +#include "changk104.h" + +static const struct gk104_fifo_func +gp10b_fifo = { + .fault.engine = gp100_fifo_fault_engine, + .fault.reason = gk104_fifo_fault_reason, + .fault.hubclient = gk104_fifo_fault_hubclient, + .fault.gpcclient = gk104_fifo_fault_gpcclient, + .chan = { + &gp100_fifo_gpfifo_oclass, + NULL + }, +}; + +int +gp10b_fifo_new(struct nvkm_device *device, int index, struct nvkm_fifo **pfifo) +{ + return gk104_fifo_new_(&gp10b_fifo, device, index, 512, pfifo); +} diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/Kbuild b/drivers/gpu/drm/nouveau/nvkm/engine/gr/Kbuild index 2938ad5aca40..8a22558b7b52 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/Kbuild +++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/Kbuild @@ -33,6 +33,8 @@ nvkm-y += nvkm/engine/gr/gm200.o nvkm-y += nvkm/engine/gr/gm20b.o nvkm-y += nvkm/engine/gr/gp100.o nvkm-y += nvkm/engine/gr/gp102.o +nvkm-y += nvkm/engine/gr/gp107.o +nvkm-y += nvkm/engine/gr/gp10b.o nvkm-y += nvkm/engine/gr/ctxnv40.o nvkm-y += nvkm/engine/gr/ctxnv50.o @@ -52,3 +54,4 @@ nvkm-y += nvkm/engine/gr/ctxgm200.o nvkm-y += nvkm/engine/gr/ctxgm20b.o nvkm-y += nvkm/engine/gr/ctxgp100.o nvkm-y += nvkm/engine/gr/ctxgp102.o +nvkm-y += nvkm/engine/gr/ctxgp107.o diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h index 0ae032fa2909..017180d147cf 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h +++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h @@ -106,6 +106,9 @@ void gp100_grctx_generate_main(struct gf100_gr *, struct gf100_grctx *); void gp100_grctx_generate_pagepool(struct gf100_grctx *); extern const struct gf100_grctx_func gp102_grctx; +void gp102_grctx_generate_attrib(struct gf100_grctx *); + +extern const struct gf100_grctx_func gp107_grctx; /* context init value lists */ diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp102.c index ee26d64af73a..80b7ab0bee3a 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp102.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp102.c @@ -29,7 +29,7 @@ * PGRAPH context implementation ******************************************************************************/ -static void +void gp102_grctx_generate_attrib(struct gf100_grctx *info) { struct gf100_gr *gr = info->gr; diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp107.c new file mode 100644 index 000000000000..8da91a0b3bd2 --- /dev/null +++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp107.c @@ -0,0 +1,47 @@ +/* + * Copyright 2017 Red Hat Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + * Authors: Ben Skeggs <bskeggs@redhat.com> + */ +#include "ctxgf100.h" + +#include <subdev/fb.h> + +/******************************************************************************* + * PGRAPH context implementation + ******************************************************************************/ + +const struct gf100_grctx_func +gp107_grctx = { + .main = gp100_grctx_generate_main, + .unkn = gk104_grctx_generate_unkn, + .bundle = gm107_grctx_generate_bundle, + .bundle_size = 0x3000, + .bundle_min_gpm_fifo_depth = 0x180, + .bundle_token_limit = 0x300, + .pagepool = gp100_grctx_generate_pagepool, + .pagepool_size = 0x20000, + .attrib = gp102_grctx_generate_attrib, + .attrib_nr_max = 0x15de, + .attrib_nr = 0x540, + .alpha_nr_max = 0xc00, + .alpha_nr = 0x800, +}; diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c index a4410ef19db5..99689f4de502 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c @@ -1463,25 +1463,27 @@ gf100_gr_init_ctxctl_ext(struct gf100_gr *gr) struct nvkm_subdev *subdev = &gr->base.engine.subdev; struct nvkm_device *device = subdev->device; struct nvkm_secboot *sb = device->secboot; - int ret = 0; + u32 secboot_mask = 0; /* load fuc microcode */ nvkm_mc_unk260(device, 0); /* securely-managed falcons must be reset using secure boot */ if (nvkm_secboot_is_managed(sb, NVKM_SECBOOT_FALCON_FECS)) - ret = nvkm_secboot_reset(sb, NVKM_SECBOOT_FALCON_FECS); + secboot_mask |= BIT(NVKM_SECBOOT_FALCON_FECS); else gf100_gr_init_fw(gr->fecs, &gr->fuc409c, &gr->fuc409d); - if (ret) - return ret; if (nvkm_secboot_is_managed(sb, NVKM_SECBOOT_FALCON_GPCCS)) - ret = nvkm_secboot_reset(sb, NVKM_SECBOOT_FALCON_GPCCS); + secboot_mask |= BIT(NVKM_SECBOOT_FALCON_GPCCS); else gf100_gr_init_fw(gr->gpccs, &gr->fuc41ac, &gr->fuc41ad); - if (ret) - return ret; + + if (secboot_mask != 0) { + int ret = nvkm_secboot_reset(sb, secboot_mask); + if (ret) + return ret; + } nvkm_mc_unk260(device, 1); diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h index 1d2101af2a87..a36e45a4a635 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h +++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h @@ -125,6 +125,7 @@ struct gf100_gr_func { void (*init_rop_active_fbps)(struct gf100_gr *); void (*init_ppc_exceptions)(struct gf100_gr *); void (*init_swdx_pes_mask)(struct gf100_gr *); + void (*init_num_active_ltcs)(struct gf100_gr *); void (*set_hww_esr_report_mask)(struct gf100_gr *); const struct gf100_gr_pack *mmio; struct { @@ -301,4 +302,8 @@ extern const struct gf100_gr_init gm107_gr_init_cbm_0[]; void gm107_gr_init_bios(struct gf100_gr *); void gm200_gr_init_gpc_mmu(struct gf100_gr *); + +void gp100_gr_init_num_active_ltcs(struct gf100_gr *gr); + +void gp102_gr_init_swdx_pes_mask(struct gf100_gr *); #endif diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c index 94ed7debb714..867a5f7cc5bc 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c @@ -40,6 +40,15 @@ gp100_gr_init_rop_active_fbps(struct gf100_gr *gr) nvkm_mask(device, 0x408958, 0x0000000f, fbp_count); /* crop */ } +void +gp100_gr_init_num_active_ltcs(struct gf100_gr *gr) +{ + struct nvkm_device *device = gr->base.engine.subdev.device; + + nvkm_wr32(device, GPC_BCAST(0x08ac), nvkm_rd32(device, 0x100800)); + nvkm_wr32(device, GPC_BCAST(0x033c), nvkm_rd32(device, 0x100804)); +} + int gp100_gr_init(struct gf100_gr *gr) { @@ -81,8 +90,7 @@ gp100_gr_init(struct gf100_gr *gr) } nvkm_wr32(device, GPC_BCAST(0x3fd4), magicgpc918); - nvkm_wr32(device, GPC_BCAST(0x08ac), nvkm_rd32(device, 0x100800)); - nvkm_wr32(device, GPC_BCAST(0x033c), nvkm_rd32(device, 0x100804)); + gr->func->init_num_active_ltcs(gr); gr->func->init_rop_active_fbps(gr); if (gr->func->init_swdx_pes_mask) @@ -154,6 +162,7 @@ gp100_gr = { .init_gpc_mmu = gm200_gr_init_gpc_mmu, .init_rop_active_fbps = gp100_gr_init_rop_active_fbps, .init_ppc_exceptions = gk104_gr_init_ppc_exceptions, + .init_num_active_ltcs = gp100_gr_init_num_active_ltcs, .rops = gm200_gr_rops, .ppc_nr = 2, .grctx = &gp100_grctx, diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c index 1d5117a16299..61e3a0b08559 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c @@ -26,7 +26,7 @@ #include <nvif/class.h> -static void +void gp102_gr_init_swdx_pes_mask(struct gf100_gr *gr) { struct nvkm_device *device = gr->base.engine.subdev.device; @@ -47,6 +47,7 @@ gp102_gr = { .init_rop_active_fbps = gp100_gr_init_rop_active_fbps, .init_ppc_exceptions = gk104_gr_init_ppc_exceptions, .init_swdx_pes_mask = gp102_gr_init_swdx_pes_mask, + .init_num_active_ltcs = gp100_gr_init_num_active_ltcs, .rops = gm200_gr_rops, .ppc_nr = 3, .grctx = &gp102_grctx, diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c new file mode 100644 index 000000000000..f7272323f694 --- /dev/null +++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c @@ -0,0 +1,53 @@ +/* + * Copyright 2017 Red Hat Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + * Authors: Ben Skeggs <bskeggs@redhat.com> + */ +#include "gf100.h" +#include "ctxgf100.h" + +#include <nvif/class.h> + +static const struct gf100_gr_func +gp107_gr = { + .init = gp100_gr_init, + .init_gpc_mmu = gm200_gr_init_gpc_mmu, + .init_rop_active_fbps = gp100_gr_init_rop_active_fbps, + .init_ppc_exceptions = gk104_gr_init_ppc_exceptions, + .init_swdx_pes_mask = gp102_gr_init_swdx_pes_mask, + .init_num_active_ltcs = gp100_gr_init_num_active_ltcs, + .rops = gm200_gr_rops, + .ppc_nr = 1, + .grctx = &gp107_grctx, + .sclass = { + { -1, -1, FERMI_TWOD_A }, + { -1, -1, KEPLER_INLINE_TO_MEMORY_B }, + { -1, -1, PASCAL_B, &gf100_fermi }, + { -1, -1, PASCAL_COMPUTE_B }, + {} + } +}; + +int +gp107_gr_new(struct nvkm_device *device, int index, struct nvkm_gr **pgr) +{ + return gm200_gr_new_(&gp107_gr, device, index, pgr); +} diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c new file mode 100644 index 000000000000..5f3d161a0842 --- /dev/null +++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c @@ -0,0 +1,59 @@ +/* + * Copyright (c) 2017, NVIDIA CORPORATION. All rights reserved. + * + * 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 AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + */ + +#include "gf100.h" +#include "ctxgf100.h" + +#include <nvif/class.h> + +static void +gp10b_gr_init_num_active_ltcs(struct gf100_gr *gr) +{ + struct nvkm_device *device = gr->base.engine.subdev.device; + + nvkm_wr32(device, GPC_BCAST(0x08ac), nvkm_rd32(device, 0x100800)); +} + +static const struct gf100_gr_func +gp10b_gr = { + .init = gp100_gr_init, + .init_gpc_mmu = gm200_gr_init_gpc_mmu, + .init_rop_active_fbps = gp100_gr_init_rop_active_fbps, + .init_ppc_exceptions = gk104_gr_init_ppc_exceptions, + .init_num_active_ltcs = gp10b_gr_init_num_active_ltcs, + .rops = gm200_gr_rops, + .ppc_nr = 1, + .grctx = &gp102_grctx, + .sclass = { + { -1, -1, FERMI_TWOD_A }, + { -1, -1, KEPLER_INLINE_TO_MEMORY_B }, + { -1, -1, PASCAL_A, &gf100_fermi }, + { -1, -1, PASCAL_COMPUTE_A }, + {} + } +}; + +int +gp10b_gr_new(struct nvkm_device *device, int index, struct nvkm_gr **pgr) +{ + return gm200_gr_new_(&gp10b_gr, device, index, pgr); +} diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/mpeg/nv31.c b/drivers/gpu/drm/nouveau/nvkm/engine/mpeg/nv31.c index 003ac915eaad..8a8895246d26 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/mpeg/nv31.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/mpeg/nv31.c @@ -198,7 +198,7 @@ nv31_mpeg_intr(struct nvkm_engine *engine) } if (type == 0x00000010) { - if (!nv31_mpeg_mthd(mpeg, mthd, data)) + if (nv31_mpeg_mthd(mpeg, mthd, data)) show &= ~0x01000000; } } diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/mpeg/nv44.c b/drivers/gpu/drm/nouveau/nvkm/engine/mpeg/nv44.c index e536f37e24b0..c3cf02ed468e 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/mpeg/nv44.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/mpeg/nv44.c @@ -172,7 +172,7 @@ nv44_mpeg_intr(struct nvkm_engine *engine) } if (type == 0x00000010) { - if (!nv44_mpeg_mthd(subdev->device, mthd, data)) + if (nv44_mpeg_mthd(subdev->device, mthd, data)) show &= ~0x01000000; } } diff --git a/drivers/gpu/drm/nouveau/nvkm/falcon/msgqueue.c b/drivers/gpu/drm/nouveau/nvkm/falcon/msgqueue.c index 982efedb4b13..d45d7947a964 100644 --- a/drivers/gpu/drm/nouveau/nvkm/falcon/msgqueue.c +++ b/drivers/gpu/drm/nouveau/nvkm/falcon/msgqueue.c @@ -463,26 +463,49 @@ nvkm_msgqueue_write_cmdline(struct nvkm_msgqueue *queue, void *buf) } int -nvkm_msgqueue_acr_boot_falcon(struct nvkm_msgqueue *queue, enum nvkm_secboot_falcon falcon) +nvkm_msgqueue_acr_boot_falcons(struct nvkm_msgqueue *queue, + unsigned long falcon_mask) { - if (!queue || !queue->func->acr_func || !queue->func->acr_func->boot_falcon) + unsigned long falcon; + + if (!queue || !queue->func->acr_func) return -ENODEV; - return queue->func->acr_func->boot_falcon(queue, falcon); + /* Does the firmware support booting multiple falcons? */ + if (queue->func->acr_func->boot_multiple_falcons) + return queue->func->acr_func->boot_multiple_falcons(queue, + falcon_mask); + + /* Else boot all requested falcons individually */ + if (!queue->func->acr_func->boot_falcon) + return -ENODEV; + + for_each_set_bit(falcon, &falcon_mask, NVKM_SECBOOT_FALCON_END) { + int ret = queue->func->acr_func->boot_falcon(queue, falcon); + + if (ret) + return ret; + } + + return 0; } int -nvkm_msgqueue_new(u32 version, struct nvkm_falcon *falcon, struct nvkm_msgqueue **queue) +nvkm_msgqueue_new(u32 version, struct nvkm_falcon *falcon, + const struct nvkm_secboot *sb, struct nvkm_msgqueue **queue) { const struct nvkm_subdev *subdev = falcon->owner; int ret = -EINVAL; switch (version) { case 0x0137c63d: - ret = msgqueue_0137c63d_new(falcon, queue); + ret = msgqueue_0137c63d_new(falcon, sb, queue); + break; + case 0x0137bca5: + ret = msgqueue_0137bca5_new(falcon, sb, queue); break; case 0x0148cdec: - ret = msgqueue_0148cdec_new(falcon, queue); + ret = msgqueue_0148cdec_new(falcon, sb, queue); break; default: nvkm_error(subdev, "unhandled firmware version 0x%08x\n", diff --git a/drivers/gpu/drm/nouveau/nvkm/falcon/msgqueue.h b/drivers/gpu/drm/nouveau/nvkm/falcon/msgqueue.h index f37afe963d3e..13b54f8d8e04 100644 --- a/drivers/gpu/drm/nouveau/nvkm/falcon/msgqueue.h +++ b/drivers/gpu/drm/nouveau/nvkm/falcon/msgqueue.h @@ -101,9 +101,11 @@ struct nvkm_msgqueue_init_func { * struct nvkm_msgqueue_acr_func - msgqueue functions related to ACR * * @boot_falcon: build and send the command to reset a given falcon + * @boot_multiple_falcons: build and send the command to reset several falcons */ struct nvkm_msgqueue_acr_func { int (*boot_falcon)(struct nvkm_msgqueue *, enum nvkm_secboot_falcon); + int (*boot_multiple_falcons)(struct nvkm_msgqueue *, unsigned long); }; struct nvkm_msgqueue_func { @@ -201,7 +203,11 @@ int nvkm_msgqueue_post(struct nvkm_msgqueue *, enum msgqueue_msg_priority, void nvkm_msgqueue_process_msgs(struct nvkm_msgqueue *, struct nvkm_msgqueue_queue *); -int msgqueue_0137c63d_new(struct nvkm_falcon *, struct nvkm_msgqueue **); -int msgqueue_0148cdec_new(struct nvkm_falcon *, struct nvkm_msgqueue **); +int msgqueue_0137c63d_new(struct nvkm_falcon *, const struct nvkm_secboot *, + struct nvkm_msgqueue **); +int msgqueue_0137bca5_new(struct nvkm_falcon *, const struct nvkm_secboot *, + struct nvkm_msgqueue **); +int msgqueue_0148cdec_new(struct nvkm_falcon *, const struct nvkm_secboot *, + struct nvkm_msgqueue **); #endif diff --git a/drivers/gpu/drm/nouveau/nvkm/falcon/msgqueue_0137c63d.c b/drivers/gpu/drm/nouveau/nvkm/falcon/msgqueue_0137c63d.c index bba91207fb18..fec0273158f6 100644 --- a/drivers/gpu/drm/nouveau/nvkm/falcon/msgqueue_0137c63d.c +++ b/drivers/gpu/drm/nouveau/nvkm/falcon/msgqueue_0137c63d.c @@ -43,6 +43,15 @@ struct msgqueue_0137c63d { #define msgqueue_0137c63d(q) \ container_of(q, struct msgqueue_0137c63d, base) +struct msgqueue_0137bca5 { + struct msgqueue_0137c63d base; + + u64 wpr_addr; +}; +#define msgqueue_0137bca5(q) \ + container_of(container_of(q, struct msgqueue_0137c63d, base), \ + struct msgqueue_0137bca5, base); + static struct nvkm_msgqueue_queue * msgqueue_0137c63d_cmd_queue(struct nvkm_msgqueue *queue, enum msgqueue_msg_priority priority) @@ -180,6 +189,7 @@ msgqueue_0137c63d_init_func = { enum { ACR_CMD_INIT_WPR_REGION = 0x00, ACR_CMD_BOOTSTRAP_FALCON = 0x01, + ACR_CMD_BOOTSTRAP_MULTIPLE_FALCONS = 0x03, }; static void @@ -286,11 +296,81 @@ acr_boot_falcon(struct nvkm_msgqueue *priv, enum nvkm_secboot_falcon falcon) return 0; } +static void +acr_boot_multiple_falcons_callback(struct nvkm_msgqueue *priv, + struct nvkm_msgqueue_hdr *hdr) +{ + struct acr_bootstrap_falcon_msg { + struct nvkm_msgqueue_msg base; + + u32 falcon_mask; + } *msg = (void *)hdr; + const struct nvkm_subdev *subdev = priv->falcon->owner; + unsigned long falcon_mask = msg->falcon_mask; + u32 falcon_id, falcon_treated = 0; + + for_each_set_bit(falcon_id, &falcon_mask, NVKM_SECBOOT_FALCON_END) { + nvkm_debug(subdev, "%s booted\n", + nvkm_secboot_falcon_name[falcon_id]); + falcon_treated |= BIT(falcon_id); + } + + if (falcon_treated != msg->falcon_mask) { + nvkm_error(subdev, "in bootstrap falcon callback:\n"); + nvkm_error(subdev, "invalid falcon mask 0x%x\n", + msg->falcon_mask); + return; + } +} + +static int +acr_boot_multiple_falcons(struct nvkm_msgqueue *priv, unsigned long falcon_mask) +{ + DECLARE_COMPLETION_ONSTACK(completed); + /* + * flags - Flag specifying RESET or no RESET. + * falcon id - Falcon id specifying falcon to bootstrap. + */ + struct { + struct nvkm_msgqueue_hdr hdr; + u8 cmd_type; + u32 flags; + u32 falcon_mask; + u32 use_va_mask; + u32 wpr_lo; + u32 wpr_hi; + } cmd; + struct msgqueue_0137bca5 *queue = msgqueue_0137bca5(priv); + + memset(&cmd, 0, sizeof(cmd)); + + cmd.hdr.unit_id = MSGQUEUE_0137C63D_UNIT_ACR; + cmd.hdr.size = sizeof(cmd); + cmd.cmd_type = ACR_CMD_BOOTSTRAP_MULTIPLE_FALCONS; + cmd.flags = ACR_CMD_BOOTSTRAP_FALCON_FLAGS_RESET_YES; + cmd.falcon_mask = falcon_mask; + cmd.wpr_lo = lower_32_bits(queue->wpr_addr); + cmd.wpr_hi = upper_32_bits(queue->wpr_addr); + nvkm_msgqueue_post(priv, MSGQUEUE_MSG_PRIORITY_HIGH, &cmd.hdr, + acr_boot_multiple_falcons_callback, &completed, true); + + if (!wait_for_completion_timeout(&completed, msecs_to_jiffies(1000))) + return -ETIMEDOUT; + + return 0; +} + static const struct nvkm_msgqueue_acr_func msgqueue_0137c63d_acr_func = { .boot_falcon = acr_boot_falcon, }; +static const struct nvkm_msgqueue_acr_func +msgqueue_0137bca5_acr_func = { + .boot_falcon = acr_boot_falcon, + .boot_multiple_falcons = acr_boot_multiple_falcons, +}; + static void msgqueue_0137c63d_dtor(struct nvkm_msgqueue *queue) { @@ -307,7 +387,8 @@ msgqueue_0137c63d_func = { }; int -msgqueue_0137c63d_new(struct nvkm_falcon *falcon, struct nvkm_msgqueue **queue) +msgqueue_0137c63d_new(struct nvkm_falcon *falcon, const struct nvkm_secboot *sb, + struct nvkm_msgqueue **queue) { struct msgqueue_0137c63d *ret; @@ -321,3 +402,35 @@ msgqueue_0137c63d_new(struct nvkm_falcon *falcon, struct nvkm_msgqueue **queue) return 0; } + +static const struct nvkm_msgqueue_func +msgqueue_0137bca5_func = { + .init_func = &msgqueue_0137c63d_init_func, + .acr_func = &msgqueue_0137bca5_acr_func, + .cmd_queue = msgqueue_0137c63d_cmd_queue, + .recv = msgqueue_0137c63d_process_msgs, + .dtor = msgqueue_0137c63d_dtor, +}; + +int +msgqueue_0137bca5_new(struct nvkm_falcon *falcon, const struct nvkm_secboot *sb, + struct nvkm_msgqueue **queue) +{ + struct msgqueue_0137bca5 *ret; + + ret = kzalloc(sizeof(*ret), GFP_KERNEL); + if (!ret) + return -ENOMEM; + + *queue = &ret->base.base; + + /* + * FIXME this must be set to the address of a *GPU* mapping within the + * ACR address space! + */ + /* ret->wpr_addr = sb->wpr_addr; */ + + nvkm_msgqueue_ctor(&msgqueue_0137bca5_func, falcon, &ret->base.base); + + return 0; +} diff --git a/drivers/gpu/drm/nouveau/nvkm/falcon/msgqueue_0148cdec.c b/drivers/gpu/drm/nouveau/nvkm/falcon/msgqueue_0148cdec.c index ed5d0da4f4e9..9424803b9ef4 100644 --- a/drivers/gpu/drm/nouveau/nvkm/falcon/msgqueue_0148cdec.c +++ b/drivers/gpu/drm/nouveau/nvkm/falcon/msgqueue_0148cdec.c @@ -247,7 +247,8 @@ msgqueue_0148cdec_func = { }; int -msgqueue_0148cdec_new(struct nvkm_falcon *falcon, struct nvkm_msgqueue **queue) +msgqueue_0148cdec_new(struct nvkm_falcon *falcon, const struct nvkm_secboot *sb, + struct nvkm_msgqueue **queue) { struct msgqueue_0148cdec *ret; diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/Kbuild b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/Kbuild index 1c5e5ba487a8..2571530e82f1 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/Kbuild +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/Kbuild @@ -28,6 +28,7 @@ nvkm-y += nvkm/subdev/fb/gm200.o nvkm-y += nvkm/subdev/fb/gm20b.o nvkm-y += nvkm/subdev/fb/gp100.o nvkm-y += nvkm/subdev/fb/gp102.o +nvkm-y += nvkm/subdev/fb/gp10b.o nvkm-y += nvkm/subdev/fb/ram.o nvkm-y += nvkm/subdev/fb/ramnv04.o diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gp10b.c b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gp10b.c new file mode 100644 index 000000000000..f2b1fbf428d5 --- /dev/null +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gp10b.c @@ -0,0 +1,38 @@ +/* + * Copyright (c) 2016, NVIDIA CORPORATION. All rights reserved. + * + * 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 AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + */ +#include "gf100.h" + +static const struct nvkm_fb_func +gp10b_fb = { + .dtor = gf100_fb_dtor, + .oneinit = gf100_fb_oneinit, + .init = gm200_fb_init, + .init_page = gm200_fb_init_page, + .intr = gf100_fb_intr, + .memtype_valid = gf100_fb_memtype_valid, +}; + +int +gp10b_fb_new(struct nvkm_device *device, int index, struct nvkm_fb **pfb) +{ + return gf100_fb_new_(&gp10b_fb, device, index, pfb); +} diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/gpio/base.c b/drivers/gpu/drm/nouveau/nvkm/subdev/gpio/base.c index 77c649723ad7..4a57defc99b3 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/gpio/base.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/gpio/base.c @@ -164,7 +164,7 @@ static int nvkm_gpio_fini(struct nvkm_subdev *subdev, bool suspend) { struct nvkm_gpio *gpio = nvkm_gpio(subdev); - u32 mask = (1 << gpio->func->lines) - 1; + u32 mask = (1ULL << gpio->func->lines) - 1; gpio->func->intr_mask(gpio, NVKM_GPIO_TOGGLED, mask, 0); gpio->func->intr_stat(gpio, &mask, &mask); diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/ibus/Kbuild b/drivers/gpu/drm/nouveau/nvkm/subdev/ibus/Kbuild index ad572d3b5466..7f5883b78efa 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/ibus/Kbuild +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/ibus/Kbuild @@ -3,3 +3,4 @@ nvkm-y += nvkm/subdev/ibus/gf117.o nvkm-y += nvkm/subdev/ibus/gk104.o nvkm-y += nvkm/subdev/ibus/gk20a.o nvkm-y += nvkm/subdev/ibus/gm200.o +nvkm-y += nvkm/subdev/ibus/gp10b.o diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/ibus/gp10b.c b/drivers/gpu/drm/nouveau/nvkm/subdev/ibus/gp10b.c new file mode 100644 index 000000000000..39db90aa2c80 --- /dev/null +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/ibus/gp10b.c @@ -0,0 +1,59 @@ +/* + * Copyright (c) 2017, NVIDIA CORPORATION. All rights reserved. + * + * 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 AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + */ +#include <subdev/ibus.h> + +#include "priv.h" + +static int +gp10b_ibus_init(struct nvkm_subdev *ibus) +{ + struct nvkm_device *device = ibus->device; + + nvkm_wr32(device, 0x1200a8, 0x0); + + /* init ring */ + nvkm_wr32(device, 0x12004c, 0x4); + nvkm_wr32(device, 0x122204, 0x2); + nvkm_rd32(device, 0x122204); + + /* timeout configuration */ + nvkm_wr32(device, 0x009080, 0x800186a0); + + return 0; +} + +static const struct nvkm_subdev_func +gp10b_ibus = { + .init = gp10b_ibus_init, + .intr = gk104_ibus_intr, +}; + +int +gp10b_ibus_new(struct nvkm_device *device, int index, + struct nvkm_subdev **pibus) +{ + struct nvkm_subdev *ibus; + if (!(ibus = *pibus = kzalloc(sizeof(*ibus), GFP_KERNEL))) + return -ENOMEM; + nvkm_subdev_ctor(&gp10b_ibus, device, index, ibus); + return 0; +} diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/instmem/gk20a.c b/drivers/gpu/drm/nouveau/nvkm/subdev/instmem/gk20a.c index 9dec58ec3d9f..cd5adbec5e57 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/instmem/gk20a.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/instmem/gk20a.c @@ -94,7 +94,7 @@ struct gk20a_instmem { struct nvkm_instmem base; /* protects vaddr_* and gk20a_instobj::vaddr* */ - spinlock_t lock; + struct mutex lock; /* CPU mappings LRU */ unsigned int vaddr_use; @@ -184,11 +184,10 @@ gk20a_instobj_acquire_iommu(struct nvkm_memory *memory) struct gk20a_instmem *imem = node->base.imem; struct nvkm_ltc *ltc = imem->base.subdev.device->ltc; const u64 size = nvkm_memory_size(memory); - unsigned long flags; nvkm_ltc_flush(ltc); - spin_lock_irqsave(&imem->lock, flags); + mutex_lock(&imem->lock); if (node->base.vaddr) { if (!node->use_cpt) { @@ -216,7 +215,7 @@ gk20a_instobj_acquire_iommu(struct nvkm_memory *memory) out: node->use_cpt++; - spin_unlock_irqrestore(&imem->lock, flags); + mutex_unlock(&imem->lock); return node->base.vaddr; } @@ -239,9 +238,8 @@ gk20a_instobj_release_iommu(struct nvkm_memory *memory) struct gk20a_instobj_iommu *node = gk20a_instobj_iommu(memory); struct gk20a_instmem *imem = node->base.imem; struct nvkm_ltc *ltc = imem->base.subdev.device->ltc; - unsigned long flags; - spin_lock_irqsave(&imem->lock, flags); + mutex_lock(&imem->lock); /* we should at least have one user to release... */ if (WARN_ON(node->use_cpt == 0)) @@ -252,7 +250,7 @@ gk20a_instobj_release_iommu(struct nvkm_memory *memory) list_add_tail(&node->vaddr_node, &imem->vaddr_lru); out: - spin_unlock_irqrestore(&imem->lock, flags); + mutex_unlock(&imem->lock); wmb(); nvkm_ltc_invalidate(ltc); @@ -306,19 +304,18 @@ gk20a_instobj_dtor_iommu(struct nvkm_memory *memory) struct gk20a_instmem *imem = node->base.imem; struct device *dev = imem->base.subdev.device->dev; struct nvkm_mm_node *r = node->base.mem.mem; - unsigned long flags; int i; if (unlikely(!r)) goto out; - spin_lock_irqsave(&imem->lock, flags); + mutex_lock(&imem->lock); /* vaddr has already been recycled */ if (node->base.vaddr) gk20a_instobj_iommu_recycle_vaddr(node); - spin_unlock_irqrestore(&imem->lock, flags); + mutex_unlock(&imem->lock); /* clear IOMMU bit to unmap pages */ r->offset &= ~BIT(imem->iommu_bit - imem->iommu_pgshift); @@ -571,7 +568,7 @@ gk20a_instmem_new(struct nvkm_device *device, int index, if (!(imem = kzalloc(sizeof(*imem), GFP_KERNEL))) return -ENOMEM; nvkm_instmem_ctor(&gk20a_instmem, device, index, &imem->base); - spin_lock_init(&imem->lock); + mutex_init(&imem->lock); *pimem = &imem->base; /* do not allow more than 1MB of CPU-mapped instmem */ diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/mc/Kbuild b/drivers/gpu/drm/nouveau/nvkm/subdev/mc/Kbuild index 12943f92c206..2befbe36dc28 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/mc/Kbuild +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/mc/Kbuild @@ -11,3 +11,4 @@ nvkm-y += nvkm/subdev/mc/gf100.o nvkm-y += nvkm/subdev/mc/gk104.o nvkm-y += nvkm/subdev/mc/gk20a.o nvkm-y += nvkm/subdev/mc/gp100.o +nvkm-y += nvkm/subdev/mc/gp10b.o diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/mc/gp100.c b/drivers/gpu/drm/nouveau/nvkm/subdev/mc/gp100.c index 4d22f4abd6de..7321ad3758c3 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/mc/gp100.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/mc/gp100.c @@ -42,7 +42,7 @@ gp100_mc_intr_update(struct gp100_mc *mc) } } -static void +void gp100_mc_intr_unarm(struct nvkm_mc *base) { struct gp100_mc *mc = gp100_mc(base); @@ -53,7 +53,7 @@ gp100_mc_intr_unarm(struct nvkm_mc *base) spin_unlock_irqrestore(&mc->lock, flags); } -static void +void gp100_mc_intr_rearm(struct nvkm_mc *base) { struct gp100_mc *mc = gp100_mc(base); @@ -64,7 +64,7 @@ gp100_mc_intr_rearm(struct nvkm_mc *base) spin_unlock_irqrestore(&mc->lock, flags); } -static void +void gp100_mc_intr_mask(struct nvkm_mc *base, u32 mask, u32 intr) { struct gp100_mc *mc = gp100_mc(base); @@ -87,13 +87,14 @@ gp100_mc = { }; int -gp100_mc_new(struct nvkm_device *device, int index, struct nvkm_mc **pmc) +gp100_mc_new_(const struct nvkm_mc_func *func, struct nvkm_device *device, + int index, struct nvkm_mc **pmc) { struct gp100_mc *mc; if (!(mc = kzalloc(sizeof(*mc), GFP_KERNEL))) return -ENOMEM; - nvkm_mc_ctor(&gp100_mc, device, index, &mc->base); + nvkm_mc_ctor(func, device, index, &mc->base); *pmc = &mc->base; spin_lock_init(&mc->lock); @@ -101,3 +102,9 @@ gp100_mc_new(struct nvkm_device *device, int index, struct nvkm_mc **pmc) mc->mask = 0x7fffffff; return 0; } + +int +gp100_mc_new(struct nvkm_device *device, int index, struct nvkm_mc **pmc) +{ + return gp100_mc_new_(&gp100_mc, device, index, pmc); +} diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/mc/gp10b.c b/drivers/gpu/drm/nouveau/nvkm/subdev/mc/gp10b.c new file mode 100644 index 000000000000..2283e3b74277 --- /dev/null +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/mc/gp10b.c @@ -0,0 +1,49 @@ +/* + * Copyright (c) 2017, NVIDIA CORPORATION. All rights reserved. + * + * 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 "priv.h" + +void +gp10b_mc_init(struct nvkm_mc *mc) +{ + struct nvkm_device *device = mc->subdev.device; + nvkm_wr32(device, 0x000200, 0xffffffff); /* everything on */ + nvkm_wr32(device, 0x00020c, 0xffffffff); /* everything out of ELPG */ +} + +static const struct nvkm_mc_func +gp10b_mc = { + .init = gp10b_mc_init, + .intr = gk104_mc_intr, + .intr_unarm = gp100_mc_intr_unarm, + .intr_rearm = gp100_mc_intr_rearm, + .intr_mask = gp100_mc_intr_mask, + .intr_stat = gf100_mc_intr_stat, + .reset = gk104_mc_reset, +}; + +int +gp10b_mc_new(struct nvkm_device *device, int index, struct nvkm_mc **pmc) +{ + return gp100_mc_new_(&gp10b_mc, device, index, pmc); +} diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/mc/priv.h b/drivers/gpu/drm/nouveau/nvkm/subdev/mc/priv.h index 4f0576a06d24..3be4126441e4 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/mc/priv.h +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/mc/priv.h @@ -41,12 +41,18 @@ extern const struct nvkm_mc_map nv17_mc_reset[]; void nv44_mc_init(struct nvkm_mc *); void nv50_mc_init(struct nvkm_mc *); +void gk104_mc_init(struct nvkm_mc *); void gf100_mc_intr_unarm(struct nvkm_mc *); void gf100_mc_intr_rearm(struct nvkm_mc *); void gf100_mc_intr_mask(struct nvkm_mc *, u32, u32); u32 gf100_mc_intr_stat(struct nvkm_mc *); void gf100_mc_unk260(struct nvkm_mc *, u32); +void gp100_mc_intr_unarm(struct nvkm_mc *); +void gp100_mc_intr_rearm(struct nvkm_mc *); +void gp100_mc_intr_mask(struct nvkm_mc *, u32, u32); +int gp100_mc_new_(const struct nvkm_mc_func *, struct nvkm_device *, int, + struct nvkm_mc **); extern const struct nvkm_mc_map gk104_mc_intr[]; extern const struct nvkm_mc_map gk104_mc_reset[]; diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/Kbuild b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/Kbuild index ac7f50ae53c6..e698f4836521 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/Kbuild +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/Kbuild @@ -11,3 +11,4 @@ nvkm-y += nvkm/subdev/secboot/acr_r375.o nvkm-y += nvkm/subdev/secboot/gm200.o nvkm-y += nvkm/subdev/secboot/gm20b.o nvkm-y += nvkm/subdev/secboot/gp102.o +nvkm-y += nvkm/subdev/secboot/gp10b.o diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/acr.h b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/acr.h index 93d804652d44..b615fc81aca4 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/acr.h +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/acr.h @@ -39,8 +39,7 @@ struct nvkm_acr_func { int (*fini)(struct nvkm_acr *, struct nvkm_secboot *, bool); int (*load)(struct nvkm_acr *, struct nvkm_falcon *, struct nvkm_gpuobj *, u64); - int (*reset)(struct nvkm_acr *, struct nvkm_secboot *, - enum nvkm_secboot_falcon); + int (*reset)(struct nvkm_acr *, struct nvkm_secboot *, unsigned long); }; /** diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/acr_r352.c b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/acr_r352.c index 993a38eb3ed5..a721354249ce 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/acr_r352.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/acr_r352.c @@ -26,8 +26,6 @@ #include <core/gpuobj.h> #include <core/firmware.h> #include <engine/falcon.h> -#include <subdev/mc.h> -#include <subdev/timer.h> #include <subdev/pmu.h> #include <core/msgqueue.h> #include <engine/sec2.h> @@ -241,6 +239,7 @@ struct ls_ucode_img_r352 { */ struct ls_ucode_img * acr_r352_ls_ucode_img_load(const struct acr_r352 *acr, + const struct nvkm_secboot *sb, enum nvkm_secboot_falcon falcon_id) { const struct nvkm_subdev *subdev = acr->base.subdev; @@ -253,7 +252,7 @@ acr_r352_ls_ucode_img_load(const struct acr_r352 *acr, img->base.falcon_id = falcon_id; - ret = acr->func->ls_func[falcon_id]->load(subdev, &img->base); + ret = acr->func->ls_func[falcon_id]->load(sb, &img->base); if (ret) { kfree(img->base.ucode_data); @@ -462,12 +461,14 @@ acr_r352_ls_write_wpr(struct acr_r352 *acr, struct list_head *imgs, * will be copied into the WPR region by the HS firmware. */ static int -acr_r352_prepare_ls_blob(struct acr_r352 *acr, u64 wpr_addr, u32 wpr_size) +acr_r352_prepare_ls_blob(struct acr_r352 *acr, struct nvkm_secboot *sb) { const struct nvkm_subdev *subdev = acr->base.subdev; struct list_head imgs; struct ls_ucode_img *img, *t; unsigned long managed_falcons = acr->base.managed_falcons; + u64 wpr_addr = sb->wpr_addr; + u32 wpr_size = sb->wpr_size; int managed_count = 0; u32 image_wpr_size, ls_blob_size; int falcon_id; @@ -479,7 +480,7 @@ acr_r352_prepare_ls_blob(struct acr_r352 *acr, u64 wpr_addr, u32 wpr_size) for_each_set_bit(falcon_id, &managed_falcons, NVKM_SECBOOT_FALCON_END) { struct ls_ucode_img *img; - img = acr->func->ls_ucode_img_load(acr, falcon_id); + img = acr->func->ls_ucode_img_load(acr, sb, falcon_id); if (IS_ERR(img)) { if (acr->base.optional_falcons & BIT(falcon_id)) { managed_falcons &= ~BIT(falcon_id); @@ -704,7 +705,7 @@ acr_r352_load_blobs(struct acr_r352 *acr, struct nvkm_secboot *sb) return 0; /* Load and prepare the managed falcon's firmwares */ - ret = acr_r352_prepare_ls_blob(acr, sb->wpr_addr, sb->wpr_size); + ret = acr_r352_prepare_ls_blob(acr, sb); if (ret) return ret; @@ -882,7 +883,6 @@ acr_r352_bootstrap(struct acr_r352 *acr, struct nvkm_secboot *sb) { const struct nvkm_subdev *subdev = &sb->subdev; unsigned long managed_falcons = acr->base.managed_falcons; - u32 reg; int falcon_id; int ret; @@ -917,54 +917,13 @@ acr_r352_bootstrap(struct acr_r352 *acr, struct nvkm_secboot *sb) const struct acr_r352_ls_func *func = acr->func->ls_func[falcon_id]; - if (func->post_run) - func->post_run(&acr->base, sb); - } - - /* Re-start ourselves if we are managed */ - if (!nvkm_secboot_is_managed(sb, acr->base.boot_falcon)) - return 0; - - /* Enable interrupts */ - nvkm_falcon_wr32(sb->boot_falcon, 0x10, 0xff); - nvkm_mc_intr_mask(subdev->device, sb->boot_falcon->owner->index, true); - - /* Start LS firmware on boot falcon */ - nvkm_falcon_start(sb->boot_falcon); - - /* - * There is a bug where the LS firmware sometimes require to be started - * twice (this happens only on SEC). Detect and workaround that - * condition. - * - * Once started, the falcon will end up in STOPPED condition (bit 5) - * if successful, or in HALT condition (bit 4) if not. - */ - nvkm_msec(subdev->device, 1, - if ((reg = nvkm_rd32(subdev->device, - sb->boot_falcon->addr + 0x100) - & 0x30) != 0) - break; - ); - if (reg & BIT(4)) { - nvkm_debug(subdev, "applying workaround for start bug..."); - nvkm_falcon_start(sb->boot_falcon); - nvkm_msec(subdev->device, 1, - if ((reg = nvkm_rd32(subdev->device, - sb->boot_falcon->addr + 0x100) - & 0x30) != 0) - break; - ); - if (reg & BIT(4)) { - nvkm_error(subdev, "%s failed to start\n", - nvkm_secboot_falcon_name[acr->base.boot_falcon]); - return -EINVAL; + if (func->post_run) { + ret = func->post_run(&acr->base, sb); + if (ret) + return ret; } } - nvkm_debug(subdev, "%s started\n", - nvkm_secboot_falcon_name[acr->base.boot_falcon]); - return 0; } @@ -976,15 +935,16 @@ acr_r352_bootstrap(struct acr_r352 *acr, struct nvkm_secboot *sb) */ static int acr_r352_reset_nopmu(struct acr_r352 *acr, struct nvkm_secboot *sb, - enum nvkm_secboot_falcon falcon) + unsigned long falcon_mask) { + int falcon; int ret; /* * Perform secure boot each time we are called on FECS. Since only FECS * and GPCCS are managed and started together, this ought to be safe. */ - if (falcon != NVKM_SECBOOT_FALCON_FECS) + if (!(falcon_mask & BIT(NVKM_SECBOOT_FALCON_FECS))) goto end; ret = acr_r352_shutdown(acr, sb); @@ -996,7 +956,9 @@ acr_r352_reset_nopmu(struct acr_r352 *acr, struct nvkm_secboot *sb, return ret; end: - acr->falcon_state[falcon] = RESET; + for_each_set_bit(falcon, &falcon_mask, NVKM_SECBOOT_FALCON_END) { + acr->falcon_state[falcon] = RESET; + } return 0; } @@ -1009,11 +971,11 @@ end: */ static int acr_r352_reset(struct nvkm_acr *_acr, struct nvkm_secboot *sb, - enum nvkm_secboot_falcon falcon) + unsigned long falcon_mask) { struct acr_r352 *acr = acr_r352(_acr); struct nvkm_msgqueue *queue; - const char *fname = nvkm_secboot_falcon_name[falcon]; + int falcon; bool wpr_already_set = sb->wpr_set; int ret; @@ -1026,7 +988,7 @@ acr_r352_reset(struct nvkm_acr *_acr, struct nvkm_secboot *sb, if (!nvkm_secboot_is_managed(sb, _acr->boot_falcon)) { /* Redo secure boot entirely if it was already done */ if (wpr_already_set) - return acr_r352_reset_nopmu(acr, sb, falcon); + return acr_r352_reset_nopmu(acr, sb, falcon_mask); /* Else return the result of the initial invokation */ else return ret; @@ -1044,13 +1006,15 @@ acr_r352_reset(struct nvkm_acr *_acr, struct nvkm_secboot *sb, } /* Otherwise just ask the LS firmware to reset the falcon */ - nvkm_debug(&sb->subdev, "resetting %s falcon\n", fname); - ret = nvkm_msgqueue_acr_boot_falcon(queue, falcon); + for_each_set_bit(falcon, &falcon_mask, NVKM_SECBOOT_FALCON_END) + nvkm_debug(&sb->subdev, "resetting %s falcon\n", + nvkm_secboot_falcon_name[falcon]); + ret = nvkm_msgqueue_acr_boot_falcons(queue, falcon_mask); if (ret) { - nvkm_error(&sb->subdev, "cannot boot %s falcon\n", fname); + nvkm_error(&sb->subdev, "error during falcon reset: %d\n", ret); return ret; } - nvkm_debug(&sb->subdev, "falcon %s reset\n", fname); + nvkm_debug(&sb->subdev, "falcon reset done\n"); return 0; } diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/acr_r352.h b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/acr_r352.h index 6e88520566c9..3d58ab871563 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/acr_r352.h +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/acr_r352.h @@ -57,11 +57,11 @@ hsf_load_header_app_size(const struct hsf_load_header *hdr, u32 app) * @lhdr_flags: LS flags */ struct acr_r352_ls_func { - int (*load)(const struct nvkm_subdev *, struct ls_ucode_img *); + int (*load)(const struct nvkm_secboot *, struct ls_ucode_img *); void (*generate_bl_desc)(const struct nvkm_acr *, const struct ls_ucode_img *, u64, void *); u32 bl_desc_size; - void (*post_run)(const struct nvkm_acr *, const struct nvkm_secboot *); + int (*post_run)(const struct nvkm_acr *, const struct nvkm_secboot *); u32 lhdr_flags; }; @@ -82,6 +82,7 @@ struct acr_r352_func { bool shadow_blob; struct ls_ucode_img *(*ls_ucode_img_load)(const struct acr_r352 *, + const struct nvkm_secboot *, enum nvkm_secboot_falcon); int (*ls_fill_headers)(struct acr_r352 *, struct list_head *); int (*ls_write_wpr)(struct acr_r352 *, struct list_head *, @@ -145,6 +146,7 @@ struct nvkm_acr *acr_r352_new_(const struct acr_r352_func *, enum nvkm_secboot_falcon, unsigned long); struct ls_ucode_img *acr_r352_ls_ucode_img_load(const struct acr_r352 *, + const struct nvkm_secboot *, enum nvkm_secboot_falcon); int acr_r352_ls_fill_headers(struct acr_r352 *, struct list_head *); int acr_r352_ls_write_wpr(struct acr_r352 *, struct list_head *, diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/acr_r367.c b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/acr_r367.c index f860713642f1..866877b88797 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/acr_r367.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/acr_r367.c @@ -107,6 +107,7 @@ struct ls_ucode_img_r367 { struct ls_ucode_img * acr_r367_ls_ucode_img_load(const struct acr_r352 *acr, + const struct nvkm_secboot *sb, enum nvkm_secboot_falcon falcon_id) { const struct nvkm_subdev *subdev = acr->base.subdev; @@ -119,7 +120,7 @@ acr_r367_ls_ucode_img_load(const struct acr_r352 *acr, img->base.falcon_id = falcon_id; - ret = acr->func->ls_func[falcon_id]->load(subdev, &img->base); + ret = acr->func->ls_func[falcon_id]->load(sb, &img->base); if (ret) { kfree(img->base.ucode_data); kfree(img->base.sig); diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/acr_r367.h b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/acr_r367.h index ec6a71ca36be..8bdfb3e5cd1c 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/acr_r367.h +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/acr_r367.h @@ -28,6 +28,7 @@ void acr_r367_fixup_hs_desc(struct acr_r352 *, struct nvkm_secboot *, void *); struct ls_ucode_img *acr_r367_ls_ucode_img_load(const struct acr_r352 *, + const struct nvkm_secboot *, enum nvkm_secboot_falcon); int acr_r367_ls_fill_headers(struct acr_r352 *, struct list_head *); int acr_r367_ls_write_wpr(struct acr_r352 *, struct list_head *, diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/base.c b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/base.c index 5c11e8c50964..ee29c6c11afd 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/base.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/base.c @@ -102,15 +102,15 @@ nvkm_secboot_falcon_name[] = { * nvkm_secboot_reset() - reset specified falcon */ int -nvkm_secboot_reset(struct nvkm_secboot *sb, enum nvkm_secboot_falcon falcon) +nvkm_secboot_reset(struct nvkm_secboot *sb, unsigned long falcon_mask) { /* Unmanaged falcon? */ - if (!(BIT(falcon) & sb->acr->managed_falcons)) { + if ((falcon_mask | sb->acr->managed_falcons) != sb->acr->managed_falcons) { nvkm_error(&sb->subdev, "cannot reset unmanaged falcon!\n"); return -EINVAL; } - return sb->acr->func->reset(sb->acr, sb, falcon); + return sb->acr->func->reset(sb->acr, sb, falcon_mask); } /** diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/gm200.h b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/gm200.h index 6dc9fc384f24..c8ab3d76bdef 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/gm200.h +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/gm200.h @@ -41,4 +41,7 @@ void *gm200_secboot_dtor(struct nvkm_secboot *); int gm200_secboot_run_blob(struct nvkm_secboot *, struct nvkm_gpuobj *, struct nvkm_falcon *); +/* Tegra-only */ +int gm20b_secboot_tegra_read_wpr(struct gm200_secboot *, u32); + #endif diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/gm20b.c b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/gm20b.c index 29e6f73dfd7e..b10ed59a4911 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/gm20b.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/gm20b.c @@ -23,28 +23,29 @@ #include "acr.h" #include "gm200.h" +#define TEGRA210_MC_BASE 0x70019000 + #ifdef CONFIG_ARCH_TEGRA -#define TEGRA_MC_BASE 0x70019000 #define MC_SECURITY_CARVEOUT2_CFG0 0xc58 #define MC_SECURITY_CARVEOUT2_BOM_0 0xc5c #define MC_SECURITY_CARVEOUT2_BOM_HI_0 0xc60 #define MC_SECURITY_CARVEOUT2_SIZE_128K 0xc64 #define TEGRA_MC_SECURITY_CARVEOUT_CFG_LOCKED (1 << 1) /** - * sb_tegra_read_wpr() - read the WPR registers on Tegra + * gm20b_secboot_tegra_read_wpr() - read the WPR registers on Tegra * * On dGPU, we can manage the WPR region ourselves, but on Tegra the WPR region * is reserved from system memory by the bootloader and irreversibly locked. * This function reads the address and size of the pre-configured WPR region. */ -static int -gm20b_tegra_read_wpr(struct gm200_secboot *gsb) +int +gm20b_secboot_tegra_read_wpr(struct gm200_secboot *gsb, u32 mc_base) { struct nvkm_secboot *sb = &gsb->base; void __iomem *mc; u32 cfg; - mc = ioremap(TEGRA_MC_BASE, 0xd00); + mc = ioremap(mc_base, 0xd00); if (!mc) { nvkm_error(&sb->subdev, "Cannot map Tegra MC registers\n"); return PTR_ERR(mc); @@ -70,8 +71,8 @@ gm20b_tegra_read_wpr(struct gm200_secboot *gsb) return 0; } #else -static int -gm20b_tegra_read_wpr(struct gm200_secboot *gsb) +int +gm20b_secboot_tegra_read_wpr(struct gm200_secboot *gsb, u32 mc_base) { nvkm_error(&gsb->base.subdev, "Tegra support not compiled in\n"); return -EINVAL; @@ -84,7 +85,7 @@ gm20b_secboot_oneinit(struct nvkm_secboot *sb) struct gm200_secboot *gsb = gm200_secboot(sb); int ret; - ret = gm20b_tegra_read_wpr(gsb); + ret = gm20b_secboot_tegra_read_wpr(gsb, TEGRA210_MC_BASE); if (ret) return ret; diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/gp10b.c b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/gp10b.c new file mode 100644 index 000000000000..632e9545e292 --- /dev/null +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/gp10b.c @@ -0,0 +1,93 @@ +/* + * Copyright (c) 2017, NVIDIA CORPORATION. All rights reserved. + * + * 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 AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER + * DEALINGS IN THE SOFTWARE. + */ + +#include "acr.h" +#include "gm200.h" + +#define TEGRA186_MC_BASE 0x02c10000 + +static int +gp10b_secboot_oneinit(struct nvkm_secboot *sb) +{ + struct gm200_secboot *gsb = gm200_secboot(sb); + int ret; + + ret = gm20b_secboot_tegra_read_wpr(gsb, TEGRA186_MC_BASE); + if (ret) + return ret; + + return gm200_secboot_oneinit(sb); +} + +static const struct nvkm_secboot_func +gp10b_secboot = { + .dtor = gm200_secboot_dtor, + .oneinit = gp10b_secboot_oneinit, + .fini = gm200_secboot_fini, + .run_blob = gm200_secboot_run_blob, +}; + +int +gp10b_secboot_new(struct nvkm_device *device, int index, + struct nvkm_secboot **psb) +{ + int ret; + struct gm200_secboot *gsb; + struct nvkm_acr *acr; + + acr = acr_r352_new(BIT(NVKM_SECBOOT_FALCON_FECS) | + BIT(NVKM_SECBOOT_FALCON_GPCCS) | + BIT(NVKM_SECBOOT_FALCON_PMU)); + if (IS_ERR(acr)) + return PTR_ERR(acr); + + gsb = kzalloc(sizeof(*gsb), GFP_KERNEL); + if (!gsb) { + psb = NULL; + return -ENOMEM; + } + *psb = &gsb->base; + + ret = nvkm_secboot_ctor(&gp10b_secboot, acr, device, index, &gsb->base); + if (ret) + return ret; + + return 0; +} + +MODULE_FIRMWARE("nvidia/gp10b/acr/bl.bin"); +MODULE_FIRMWARE("nvidia/gp10b/acr/ucode_load.bin"); +MODULE_FIRMWARE("nvidia/gp10b/gr/fecs_bl.bin"); +MODULE_FIRMWARE("nvidia/gp10b/gr/fecs_inst.bin"); +MODULE_FIRMWARE("nvidia/gp10b/gr/fecs_data.bin"); +MODULE_FIRMWARE("nvidia/gp10b/gr/fecs_sig.bin"); +MODULE_FIRMWARE("nvidia/gp10b/gr/gpccs_bl.bin"); +MODULE_FIRMWARE("nvidia/gp10b/gr/gpccs_inst.bin"); +MODULE_FIRMWARE("nvidia/gp10b/gr/gpccs_data.bin"); +MODULE_FIRMWARE("nvidia/gp10b/gr/gpccs_sig.bin"); +MODULE_FIRMWARE("nvidia/gp10b/gr/sw_ctx.bin"); +MODULE_FIRMWARE("nvidia/gp10b/gr/sw_nonctx.bin"); +MODULE_FIRMWARE("nvidia/gp10b/gr/sw_bundle_init.bin"); +MODULE_FIRMWARE("nvidia/gp10b/gr/sw_method_init.bin"); +MODULE_FIRMWARE("nvidia/gp10b/pmu/desc.bin"); +MODULE_FIRMWARE("nvidia/gp10b/pmu/image.bin"); +MODULE_FIRMWARE("nvidia/gp10b/pmu/sig.bin"); diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/ls_ucode.h b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/ls_ucode.h index 4ff9138a2a83..9b7c402594e8 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/ls_ucode.h +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/ls_ucode.h @@ -147,11 +147,11 @@ struct fw_bl_desc { u32 data_size; }; -int acr_ls_ucode_load_fecs(const struct nvkm_subdev *, struct ls_ucode_img *); -int acr_ls_ucode_load_gpccs(const struct nvkm_subdev *, struct ls_ucode_img *); -int acr_ls_ucode_load_pmu(const struct nvkm_subdev *, struct ls_ucode_img *); -void acr_ls_pmu_post_run(const struct nvkm_acr *, const struct nvkm_secboot *); -int acr_ls_ucode_load_sec2(const struct nvkm_subdev *, struct ls_ucode_img *); -void acr_ls_sec2_post_run(const struct nvkm_acr *, const struct nvkm_secboot *); +int acr_ls_ucode_load_fecs(const struct nvkm_secboot *, struct ls_ucode_img *); +int acr_ls_ucode_load_gpccs(const struct nvkm_secboot *, struct ls_ucode_img *); +int acr_ls_ucode_load_pmu(const struct nvkm_secboot *, struct ls_ucode_img *); +int acr_ls_pmu_post_run(const struct nvkm_acr *, const struct nvkm_secboot *); +int acr_ls_ucode_load_sec2(const struct nvkm_secboot *, struct ls_ucode_img *); +int acr_ls_sec2_post_run(const struct nvkm_acr *, const struct nvkm_secboot *); #endif diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/ls_ucode_gr.c b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/ls_ucode_gr.c index 40a6df77bb8a..d1cf02d22db1 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/ls_ucode_gr.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/ls_ucode_gr.c @@ -144,15 +144,13 @@ error: } int -acr_ls_ucode_load_fecs(const struct nvkm_subdev *subdev, - struct ls_ucode_img *img) +acr_ls_ucode_load_fecs(const struct nvkm_secboot *sb, struct ls_ucode_img *img) { - return ls_ucode_img_load_gr(subdev, img, "fecs"); + return ls_ucode_img_load_gr(&sb->subdev, img, "fecs"); } int -acr_ls_ucode_load_gpccs(const struct nvkm_subdev *subdev, - struct ls_ucode_img *img) +acr_ls_ucode_load_gpccs(const struct nvkm_secboot *sb, struct ls_ucode_img *img) { - return ls_ucode_img_load_gr(subdev, img, "gpccs"); + return ls_ucode_img_load_gr(&sb->subdev, img, "gpccs"); } diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/ls_ucode_msgqueue.c b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/ls_ucode_msgqueue.c index ef0b298b70d7..ee989210725e 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/ls_ucode_msgqueue.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/ls_ucode_msgqueue.c @@ -28,6 +28,8 @@ #include <core/msgqueue.h> #include <subdev/pmu.h> #include <engine/sec2.h> +#include <subdev/mc.h> +#include <subdev/timer.h> /** * acr_ls_ucode_load_msgqueue - load and prepare a ucode img for a msgqueue fw @@ -73,10 +75,11 @@ acr_ls_ucode_load_msgqueue(const struct nvkm_subdev *subdev, const char *name, return 0; } -static void +static int acr_ls_msgqueue_post_run(struct nvkm_msgqueue *queue, struct nvkm_falcon *falcon, u32 addr_args) { + struct nvkm_device *device = falcon->owner->device; u32 cmdline_size = NVKM_MSGQUEUE_CMDLINE_SIZE; u8 buf[cmdline_size]; @@ -85,65 +88,118 @@ acr_ls_msgqueue_post_run(struct nvkm_msgqueue *queue, nvkm_falcon_load_dmem(falcon, buf, addr_args, cmdline_size, 0); /* rearm the queue so it will wait for the init message */ nvkm_msgqueue_reinit(queue); + + /* Enable interrupts */ + nvkm_falcon_wr32(falcon, 0x10, 0xff); + nvkm_mc_intr_mask(device, falcon->owner->index, true); + + /* Start LS firmware on boot falcon */ + nvkm_falcon_start(falcon); + + return 0; } int -acr_ls_ucode_load_pmu(const struct nvkm_subdev *subdev, - struct ls_ucode_img *img) +acr_ls_ucode_load_pmu(const struct nvkm_secboot *sb, struct ls_ucode_img *img) { - struct nvkm_pmu *pmu = subdev->device->pmu; + struct nvkm_pmu *pmu = sb->subdev.device->pmu; int ret; - ret = acr_ls_ucode_load_msgqueue(subdev, "pmu", img); + ret = acr_ls_ucode_load_msgqueue(&sb->subdev, "pmu", img); if (ret) return ret; /* Allocate the PMU queue corresponding to the FW version */ ret = nvkm_msgqueue_new(img->ucode_desc.app_version, pmu->falcon, - &pmu->queue); + sb, &pmu->queue); if (ret) return ret; return 0; } -void +int acr_ls_pmu_post_run(const struct nvkm_acr *acr, const struct nvkm_secboot *sb) { struct nvkm_device *device = sb->subdev.device; struct nvkm_pmu *pmu = device->pmu; u32 addr_args = pmu->falcon->data.limit - NVKM_MSGQUEUE_CMDLINE_SIZE; + int ret; + + ret = acr_ls_msgqueue_post_run(pmu->queue, pmu->falcon, addr_args); + if (ret) + return ret; - acr_ls_msgqueue_post_run(pmu->queue, pmu->falcon, addr_args); + nvkm_debug(&sb->subdev, "%s started\n", + nvkm_secboot_falcon_name[acr->boot_falcon]); + + return 0; } int -acr_ls_ucode_load_sec2(const struct nvkm_subdev *subdev, - struct ls_ucode_img *img) +acr_ls_ucode_load_sec2(const struct nvkm_secboot *sb, struct ls_ucode_img *img) { - struct nvkm_sec2 *sec = subdev->device->sec2; + struct nvkm_sec2 *sec = sb->subdev.device->sec2; int ret; - ret = acr_ls_ucode_load_msgqueue(subdev, "sec2", img); + ret = acr_ls_ucode_load_msgqueue(&sb->subdev, "sec2", img); if (ret) return ret; /* Allocate the PMU queue corresponding to the FW version */ ret = nvkm_msgqueue_new(img->ucode_desc.app_version, sec->falcon, - &sec->queue); + sb, &sec->queue); if (ret) return ret; return 0; } -void +int acr_ls_sec2_post_run(const struct nvkm_acr *acr, const struct nvkm_secboot *sb) { - struct nvkm_device *device = sb->subdev.device; + const struct nvkm_subdev *subdev = &sb->subdev; + struct nvkm_device *device = subdev->device; struct nvkm_sec2 *sec = device->sec2; /* on SEC arguments are always at the beginning of EMEM */ - u32 addr_args = 0x01000000; + const u32 addr_args = 0x01000000; + u32 reg; + int ret; - acr_ls_msgqueue_post_run(sec->queue, sec->falcon, addr_args); + ret = acr_ls_msgqueue_post_run(sec->queue, sec->falcon, addr_args); + if (ret) + return ret; + + /* + * There is a bug where the LS firmware sometimes require to be started + * twice (this happens only on SEC). Detect and workaround that + * condition. + * + * Once started, the falcon will end up in STOPPED condition (bit 5) + * if successful, or in HALT condition (bit 4) if not. + */ + nvkm_msec(device, 1, + if ((reg = nvkm_falcon_rd32(sb->boot_falcon, 0x100) & 0x30) != 0) + break; + ); + if (reg & BIT(4)) { + nvkm_debug(subdev, "applying workaround for start bug..."); + nvkm_falcon_start(sb->boot_falcon); + nvkm_msec(subdev->device, 1, + if ((reg = nvkm_rd32(subdev->device, + sb->boot_falcon->addr + 0x100) + & 0x30) != 0) + break; + ); + if (reg & BIT(4)) { + nvkm_error(subdev, "%s failed to start\n", + nvkm_secboot_falcon_name[acr->boot_falcon]); + return -EINVAL; + } + } + + nvkm_debug(&sb->subdev, "%s started\n", + nvkm_secboot_falcon_name[acr->boot_falcon]); + + return 0; } diff --git a/drivers/gpu/drm/panel/Kconfig b/drivers/gpu/drm/panel/Kconfig index 62aba976e744..5dc2106da2bc 100644 --- a/drivers/gpu/drm/panel/Kconfig +++ b/drivers/gpu/drm/panel/Kconfig @@ -7,6 +7,16 @@ config DRM_PANEL menu "Display Panels" depends on DRM && DRM_PANEL +config DRM_PANEL_LVDS + tristate "Generic LVDS panel driver" + depends on OF + depends on BACKLIGHT_CLASS_DEVICE + select VIDEOMODE_HELPERS + help + This driver supports LVDS panels that don't require device-specific + handling of power supplies or control signals. It implements automatic + backlight handling if the panel is attached to a backlight controller. + config DRM_PANEL_SIMPLE tristate "support for simple panels" depends on OF diff --git a/drivers/gpu/drm/panel/Makefile b/drivers/gpu/drm/panel/Makefile index a5c7ec0236e0..20b5060d1f47 100644 --- a/drivers/gpu/drm/panel/Makefile +++ b/drivers/gpu/drm/panel/Makefile @@ -1,3 +1,4 @@ +obj-$(CONFIG_DRM_PANEL_LVDS) += panel-lvds.o obj-$(CONFIG_DRM_PANEL_SIMPLE) += panel-simple.o obj-$(CONFIG_DRM_PANEL_JDI_LT070ME05000) += panel-jdi-lt070me05000.o obj-$(CONFIG_DRM_PANEL_LG_LG4573) += panel-lg-lg4573.o diff --git a/drivers/gpu/drm/panel/panel-lvds.c b/drivers/gpu/drm/panel/panel-lvds.c new file mode 100644 index 000000000000..3216aa9a88d6 --- /dev/null +++ b/drivers/gpu/drm/panel/panel-lvds.c @@ -0,0 +1,286 @@ +/* + * rcar_du_crtc.c -- R-Car Display Unit CRTCs + * + * Copyright (C) 2016 Laurent Pinchart + * Copyright (C) 2016 Renesas Electronics Corporation + * + * Contact: Laurent Pinchart (laurent.pinchart@ideasonboard.com) + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + */ + +#include <linux/backlight.h> +#include <linux/gpio/consumer.h> +#include <linux/module.h> +#include <linux/of_platform.h> +#include <linux/platform_device.h> +#include <linux/slab.h> + +#include <drm/drmP.h> +#include <drm/drm_crtc.h> +#include <drm/drm_panel.h> + +#include <video/display_timing.h> +#include <video/of_display_timing.h> +#include <video/videomode.h> + +struct panel_lvds { + struct drm_panel panel; + struct device *dev; + + const char *label; + unsigned int width; + unsigned int height; + struct videomode video_mode; + unsigned int bus_format; + bool data_mirror; + + struct backlight_device *backlight; + + struct gpio_desc *enable_gpio; + struct gpio_desc *reset_gpio; +}; + +static inline struct panel_lvds *to_panel_lvds(struct drm_panel *panel) +{ + return container_of(panel, struct panel_lvds, panel); +} + +static int panel_lvds_disable(struct drm_panel *panel) +{ + struct panel_lvds *lvds = to_panel_lvds(panel); + + if (lvds->backlight) { + lvds->backlight->props.power = FB_BLANK_POWERDOWN; + lvds->backlight->props.state |= BL_CORE_FBBLANK; + backlight_update_status(lvds->backlight); + } + + return 0; +} + +static int panel_lvds_unprepare(struct drm_panel *panel) +{ + struct panel_lvds *lvds = to_panel_lvds(panel); + + if (lvds->enable_gpio) + gpiod_set_value_cansleep(lvds->enable_gpio, 0); + + return 0; +} + +static int panel_lvds_prepare(struct drm_panel *panel) +{ + struct panel_lvds *lvds = to_panel_lvds(panel); + + if (lvds->enable_gpio) + gpiod_set_value_cansleep(lvds->enable_gpio, 1); + + return 0; +} + +static int panel_lvds_enable(struct drm_panel *panel) +{ + struct panel_lvds *lvds = to_panel_lvds(panel); + + if (lvds->backlight) { + lvds->backlight->props.state &= ~BL_CORE_FBBLANK; + lvds->backlight->props.power = FB_BLANK_UNBLANK; + backlight_update_status(lvds->backlight); + } + + return 0; +} + +static int panel_lvds_get_modes(struct drm_panel *panel) +{ + struct panel_lvds *lvds = to_panel_lvds(panel); + struct drm_connector *connector = lvds->panel.connector; + struct drm_display_mode *mode; + + mode = drm_mode_create(lvds->panel.drm); + if (!mode) + return 0; + + drm_display_mode_from_videomode(&lvds->video_mode, mode); + mode->type |= DRM_MODE_TYPE_DRIVER | DRM_MODE_TYPE_PREFERRED; + drm_mode_probed_add(connector, mode); + + connector->display_info.width_mm = lvds->width; + connector->display_info.height_mm = lvds->height; + drm_display_info_set_bus_formats(&connector->display_info, + &lvds->bus_format, 1); + connector->display_info.bus_flags = lvds->data_mirror + ? DRM_BUS_FLAG_DATA_LSB_TO_MSB + : DRM_BUS_FLAG_DATA_MSB_TO_LSB; + + return 1; +} + +static const struct drm_panel_funcs panel_lvds_funcs = { + .disable = panel_lvds_disable, + .unprepare = panel_lvds_unprepare, + .prepare = panel_lvds_prepare, + .enable = panel_lvds_enable, + .get_modes = panel_lvds_get_modes, +}; + +static int panel_lvds_parse_dt(struct panel_lvds *lvds) +{ + struct device_node *np = lvds->dev->of_node; + struct display_timing timing; + const char *mapping; + int ret; + + ret = of_get_display_timing(np, "panel-timing", &timing); + if (ret < 0) + return ret; + + videomode_from_timing(&timing, &lvds->video_mode); + + ret = of_property_read_u32(np, "width-mm", &lvds->width); + if (ret < 0) { + dev_err(lvds->dev, "%s: invalid or missing %s DT property\n", + of_node_full_name(np), "width-mm"); + return -ENODEV; + } + ret = of_property_read_u32(np, "height-mm", &lvds->height); + if (ret < 0) { + dev_err(lvds->dev, "%s: invalid or missing %s DT property\n", + of_node_full_name(np), "height-mm"); + return -ENODEV; + } + + of_property_read_string(np, "label", &lvds->label); + + ret = of_property_read_string(np, "data-mapping", &mapping); + if (ret < 0) { + dev_err(lvds->dev, "%s: invalid or missing %s DT property\n", + of_node_full_name(np), "data-mapping"); + return -ENODEV; + } + + if (!strcmp(mapping, "jeida-18")) { + lvds->bus_format = MEDIA_BUS_FMT_RGB666_1X7X3_SPWG; + } else if (!strcmp(mapping, "jeida-24")) { + lvds->bus_format = MEDIA_BUS_FMT_RGB888_1X7X4_JEIDA; + } else if (!strcmp(mapping, "vesa-24")) { + lvds->bus_format = MEDIA_BUS_FMT_RGB888_1X7X4_SPWG; + } else { + dev_err(lvds->dev, "%s: invalid or missing %s DT property\n", + of_node_full_name(np), "data-mapping"); + return -EINVAL; + } + + lvds->data_mirror = of_property_read_bool(np, "data-mirror"); + + return 0; +} + +static int panel_lvds_probe(struct platform_device *pdev) +{ + struct panel_lvds *lvds; + struct device_node *np; + int ret; + + lvds = devm_kzalloc(&pdev->dev, sizeof(*lvds), GFP_KERNEL); + if (!lvds) + return -ENOMEM; + + lvds->dev = &pdev->dev; + + ret = panel_lvds_parse_dt(lvds); + if (ret < 0) + return ret; + + /* Get GPIOs and backlight controller. */ + lvds->enable_gpio = devm_gpiod_get_optional(lvds->dev, "enable", + GPIOD_OUT_LOW); + if (IS_ERR(lvds->enable_gpio)) { + ret = PTR_ERR(lvds->enable_gpio); + dev_err(lvds->dev, "failed to request %s GPIO: %d\n", + "enable", ret); + return ret; + } + + lvds->reset_gpio = devm_gpiod_get_optional(lvds->dev, "reset", + GPIOD_OUT_HIGH); + if (IS_ERR(lvds->reset_gpio)) { + ret = PTR_ERR(lvds->reset_gpio); + dev_err(lvds->dev, "failed to request %s GPIO: %d\n", + "reset", ret); + return ret; + } + + np = of_parse_phandle(lvds->dev->of_node, "backlight", 0); + if (np) { + lvds->backlight = of_find_backlight_by_node(np); + of_node_put(np); + + if (!lvds->backlight) + return -EPROBE_DEFER; + } + + /* + * TODO: Handle all power supplies specified in the DT node in a generic + * way for panels that don't care about power supply ordering. LVDS + * panels that require a specific power sequence will need a dedicated + * driver. + */ + + /* Register the panel. */ + drm_panel_init(&lvds->panel); + lvds->panel.dev = lvds->dev; + lvds->panel.funcs = &panel_lvds_funcs; + + ret = drm_panel_add(&lvds->panel); + if (ret < 0) + goto error; + + dev_set_drvdata(lvds->dev, lvds); + return 0; + +error: + put_device(&lvds->backlight->dev); + return ret; +} + +static int panel_lvds_remove(struct platform_device *pdev) +{ + struct panel_lvds *lvds = dev_get_drvdata(&pdev->dev); + + drm_panel_detach(&lvds->panel); + drm_panel_remove(&lvds->panel); + + panel_lvds_disable(&lvds->panel); + + if (lvds->backlight) + put_device(&lvds->backlight->dev); + + return 0; +} + +static const struct of_device_id panel_lvds_of_table[] = { + { .compatible = "panel-lvds", }, + { /* Sentinel */ }, +}; + +MODULE_DEVICE_TABLE(of, panel_lvds_of_table); + +static struct platform_driver panel_lvds_driver = { + .probe = panel_lvds_probe, + .remove = panel_lvds_remove, + .driver = { + .name = "panel-lvds", + .of_match_table = panel_lvds_of_table, + }, +}; + +module_platform_driver(panel_lvds_driver); + +MODULE_AUTHOR("Laurent Pinchart <laurent.pinchart@ideasonboard.com>"); +MODULE_DESCRIPTION("LVDS Panel Driver"); +MODULE_LICENSE("GPL"); diff --git a/drivers/gpu/drm/qxl/qxl_ttm.c b/drivers/gpu/drm/qxl/qxl_ttm.c index 7d1cab57c89e..0fdedee4509d 100644 --- a/drivers/gpu/drm/qxl/qxl_ttm.c +++ b/drivers/gpu/drm/qxl/qxl_ttm.c @@ -393,6 +393,7 @@ static struct ttm_bo_driver qxl_bo_driver = { .verify_access = &qxl_verify_access, .io_mem_reserve = &qxl_ttm_io_mem_reserve, .io_mem_free = &qxl_ttm_io_mem_free, + .io_mem_pfn = ttm_bo_default_io_mem_pfn, .move_notify = &qxl_bo_move_notify, }; diff --git a/drivers/gpu/drm/radeon/r100.c b/drivers/gpu/drm/radeon/r100.c index aefca0b03f38..c31e660e35db 100644 --- a/drivers/gpu/drm/radeon/r100.c +++ b/drivers/gpu/drm/radeon/r100.c @@ -3295,7 +3295,7 @@ void r100_bandwidth_update(struct radeon_device *rdev) mem_trp = ((temp >> 8) & 0x7) + 1; mem_tras = ((temp >> 11) & 0xf) + 4; } else if (rdev->family == CHIP_RV350 || - rdev->family <= CHIP_RV380) { + rdev->family == CHIP_RV380) { /* rv3x0 */ mem_trcd = (temp & 0x7) + 3; mem_trp = ((temp >> 8) & 0x7) + 3; diff --git a/drivers/gpu/drm/radeon/radeon.h b/drivers/gpu/drm/radeon/radeon.h index 2e400dc414e3..c1c8e2208a21 100644 --- a/drivers/gpu/drm/radeon/radeon.h +++ b/drivers/gpu/drm/radeon/radeon.h @@ -499,6 +499,7 @@ struct radeon_bo { u32 tiling_flags; u32 pitch; int surface_reg; + unsigned prime_shared_count; /* list of all virtual address to which this bo * is associated to */ diff --git a/drivers/gpu/drm/radeon/radeon_cs.c b/drivers/gpu/drm/radeon/radeon_cs.c index a8442f7196d6..df6b58c08544 100644 --- a/drivers/gpu/drm/radeon/radeon_cs.c +++ b/drivers/gpu/drm/radeon/radeon_cs.c @@ -164,6 +164,16 @@ static int radeon_cs_parser_relocs(struct radeon_cs_parser *p) p->relocs[i].allowed_domains = domain; } + /* Objects shared as dma-bufs cannot be moved to VRAM */ + if (p->relocs[i].robj->prime_shared_count) { + p->relocs[i].allowed_domains &= ~RADEON_GEM_DOMAIN_VRAM; + if (!p->relocs[i].allowed_domains) { + DRM_ERROR("BO associated with dma-buf cannot " + "be moved to VRAM\n"); + return -EINVAL; + } + } + p->relocs[i].tv.bo = &p->relocs[i].robj->tbo; p->relocs[i].tv.shared = !r->write_domain; diff --git a/drivers/gpu/drm/radeon/radeon_display.c b/drivers/gpu/drm/radeon/radeon_display.c index 981385eb5389..17d3dafc8319 100644 --- a/drivers/gpu/drm/radeon/radeon_display.c +++ b/drivers/gpu/drm/radeon/radeon_display.c @@ -1355,6 +1355,12 @@ radeon_user_framebuffer_create(struct drm_device *dev, return ERR_PTR(-ENOENT); } + /* Handle is imported dma-buf, so cannot be migrated to VRAM for scanout */ + if (obj->import_attach) { + DRM_DEBUG_KMS("Cannot create framebuffer from imported dma_buf\n"); + return ERR_PTR(-EINVAL); + } + radeon_fb = kzalloc(sizeof(*radeon_fb), GFP_KERNEL); if (radeon_fb == NULL) { drm_gem_object_unreference_unlocked(obj); diff --git a/drivers/gpu/drm/radeon/radeon_gem.c b/drivers/gpu/drm/radeon/radeon_gem.c index 9b0b123ce079..dddb372de2b9 100644 --- a/drivers/gpu/drm/radeon/radeon_gem.c +++ b/drivers/gpu/drm/radeon/radeon_gem.c @@ -120,6 +120,10 @@ static int radeon_gem_set_domain(struct drm_gem_object *gobj, return r; } } + if (domain == RADEON_GEM_DOMAIN_VRAM && robj->prime_shared_count) { + /* A BO that is associated with a dma-buf cannot be sensibly migrated to VRAM */ + return -EINVAL; + } return 0; } diff --git a/drivers/gpu/drm/radeon/radeon_object.c b/drivers/gpu/drm/radeon/radeon_object.c index 74b276060c20..bec2ec056de4 100644 --- a/drivers/gpu/drm/radeon/radeon_object.c +++ b/drivers/gpu/drm/radeon/radeon_object.c @@ -352,6 +352,11 @@ int radeon_bo_pin_restricted(struct radeon_bo *bo, u32 domain, u64 max_offset, return 0; } + if (bo->prime_shared_count && domain == RADEON_GEM_DOMAIN_VRAM) { + /* A BO shared as a dma-buf cannot be sensibly migrated to VRAM */ + return -EINVAL; + } + radeon_ttm_placement_from_domain(bo, domain); for (i = 0; i < bo->placement.num_placement; i++) { /* force to pin into visible video ram */ diff --git a/drivers/gpu/drm/radeon/radeon_prime.c b/drivers/gpu/drm/radeon/radeon_prime.c index f3609c97496b..7110d403322c 100644 --- a/drivers/gpu/drm/radeon/radeon_prime.c +++ b/drivers/gpu/drm/radeon/radeon_prime.c @@ -77,6 +77,7 @@ struct drm_gem_object *radeon_gem_prime_import_sg_table(struct drm_device *dev, list_add_tail(&bo->list, &rdev->gem.objects); mutex_unlock(&rdev->gem.mutex); + bo->prime_shared_count = 1; return &bo->gem_base; } @@ -91,6 +92,9 @@ int radeon_gem_prime_pin(struct drm_gem_object *obj) /* pin buffer into GTT */ ret = radeon_bo_pin(bo, RADEON_GEM_DOMAIN_GTT, NULL); + if (likely(ret == 0)) + bo->prime_shared_count++; + radeon_bo_unreserve(bo); return ret; } @@ -105,6 +109,8 @@ void radeon_gem_prime_unpin(struct drm_gem_object *obj) return; radeon_bo_unpin(bo); + if (bo->prime_shared_count) + bo->prime_shared_count--; radeon_bo_unreserve(bo); } diff --git a/drivers/gpu/drm/radeon/radeon_ttm.c b/drivers/gpu/drm/radeon/radeon_ttm.c index aaa3e80fecb4..8b7623b5a624 100644 --- a/drivers/gpu/drm/radeon/radeon_ttm.c +++ b/drivers/gpu/drm/radeon/radeon_ttm.c @@ -873,6 +873,7 @@ static struct ttm_bo_driver radeon_bo_driver = { .fault_reserve_notify = &radeon_bo_fault_reserve_notify, .io_mem_reserve = &radeon_ttm_io_mem_reserve, .io_mem_free = &radeon_ttm_io_mem_free, + .io_mem_pfn = ttm_bo_default_io_mem_pfn, }; int radeon_ttm_init(struct radeon_device *rdev) diff --git a/drivers/gpu/drm/rcar-du/Kconfig b/drivers/gpu/drm/rcar-du/Kconfig index 4c2fd056dd6d..8a50dab19e5c 100644 --- a/drivers/gpu/drm/rcar-du/Kconfig +++ b/drivers/gpu/drm/rcar-du/Kconfig @@ -11,15 +11,17 @@ config DRM_RCAR_DU Choose this option if you have an R-Car chipset. If M is selected the module will be called rcar-du-drm. -config DRM_RCAR_HDMI - bool "R-Car DU HDMI Encoder Support" - depends on DRM_RCAR_DU +config DRM_RCAR_DW_HDMI + tristate "R-Car DU Gen3 HDMI Encoder Support" + depends on DRM && OF + select DRM_DW_HDMI help - Enable support for external HDMI encoders. + Enable support for R-Car Gen3 internal HDMI encoder. config DRM_RCAR_LVDS bool "R-Car DU LVDS Encoder Support" depends on DRM_RCAR_DU + select DRM_PANEL help Enable support for the R-Car Display Unit embedded LVDS encoders. diff --git a/drivers/gpu/drm/rcar-du/Makefile b/drivers/gpu/drm/rcar-du/Makefile index d3b44651061a..2131e722de3b 100644 --- a/drivers/gpu/drm/rcar-du/Makefile +++ b/drivers/gpu/drm/rcar-du/Makefile @@ -4,13 +4,11 @@ rcar-du-drm-y := rcar_du_crtc.o \ rcar_du_group.o \ rcar_du_kms.o \ rcar_du_lvdscon.o \ - rcar_du_plane.o \ - rcar_du_vgacon.o - -rcar-du-drm-$(CONFIG_DRM_RCAR_HDMI) += rcar_du_hdmienc.o + rcar_du_plane.o rcar-du-drm-$(CONFIG_DRM_RCAR_LVDS) += rcar_du_lvdsenc.o rcar-du-drm-$(CONFIG_DRM_RCAR_VSP) += rcar_du_vsp.o obj-$(CONFIG_DRM_RCAR_DU) += rcar-du-drm.o +obj-$(CONFIG_DRM_RCAR_DW_HDMI) += rcar_dw_hdmi.o diff --git a/drivers/gpu/drm/rcar-du/rcar_du_crtc.c b/drivers/gpu/drm/rcar-du/rcar_du_crtc.c index edcbe2e3625d..4ed6f2340af0 100644 --- a/drivers/gpu/drm/rcar-du/rcar_du_crtc.c +++ b/drivers/gpu/drm/rcar-du/rcar_du_crtc.c @@ -106,9 +106,62 @@ static void rcar_du_crtc_put(struct rcar_du_crtc *rcrtc) * Hardware Setup */ +struct dpll_info { + unsigned int output; + unsigned int fdpll; + unsigned int n; + unsigned int m; +}; + +static void rcar_du_dpll_divider(struct rcar_du_crtc *rcrtc, + struct dpll_info *dpll, + unsigned long input, + unsigned long target) +{ + unsigned long best_diff = (unsigned long)-1; + unsigned long diff; + unsigned int fdpll; + unsigned int m; + unsigned int n; + + for (n = 39; n < 120; n++) { + for (m = 0; m < 4; m++) { + 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; + + if (output >= 400000000) + continue; + + diff = abs((long)output - (long)target); + if (best_diff > diff) { + best_diff = diff; + dpll->n = n; + dpll->m = m; + dpll->fdpll = fdpll; + dpll->output = output; + } + + if (diff == 0) + goto done; + } + } + } + +done: + dev_dbg(rcrtc->group->dev->dev, + "output:%u, fdpll:%u, n:%u, m:%u, diff:%lu\n", + dpll->output, dpll->fdpll, dpll->n, dpll->m, + best_diff); +} + static void rcar_du_crtc_set_display_timing(struct rcar_du_crtc *rcrtc) { const struct drm_display_mode *mode = &rcrtc->crtc.state->adjusted_mode; + struct rcar_du_device *rcdu = rcrtc->group->dev; unsigned long mode_clock = mode->clock * 1000; unsigned long clk; u32 value; @@ -124,12 +177,18 @@ static void rcar_du_crtc_set_display_timing(struct rcar_du_crtc *rcrtc) escr = div | ESCR_DCLKSEL_CLKS; if (rcrtc->extclock) { + struct dpll_info dpll = { 0 }; unsigned long extclk; unsigned long extrate; unsigned long rate; u32 extdiv; extclk = clk_get_rate(rcrtc->extclock); + if (rcdu->info->dpll_ch & (1 << rcrtc->index)) { + rcar_du_dpll_divider(rcrtc, &dpll, extclk, mode_clock); + extclk = dpll.output; + } + extdiv = DIV_ROUND_CLOSEST(extclk, mode_clock); extdiv = clamp(extdiv, 1U, 64U) - 1; @@ -140,7 +199,27 @@ static void rcar_du_crtc_set_display_timing(struct rcar_du_crtc *rcrtc) abs((long)rate - (long)mode_clock)) { dev_dbg(rcrtc->group->dev->dev, "crtc%u: using external clock\n", rcrtc->index); - escr = extdiv | ESCR_DCLKSEL_DCLKIN; + + if (rcdu->info->dpll_ch & (1 << rcrtc->index)) { + u32 dpllcr = DPLLCR_CODE | DPLLCR_CLKE + | DPLLCR_FDPLL(dpll.fdpll) + | DPLLCR_N(dpll.n) | DPLLCR_M(dpll.m) + | DPLLCR_STBY; + + if (rcrtc->index == 1) + dpllcr |= DPLLCR_PLCS1 + | DPLLCR_INCS_DOTCLKIN1; + else + dpllcr |= DPLLCR_PLCS0 + | DPLLCR_INCS_DOTCLKIN0; + + rcar_du_group_write(rcrtc->group, DPLLCR, + dpllcr); + + escr = ESCR_DCLKSEL_DCLKIN | 1; + } else { + escr = ESCR_DCLKSEL_DCLKIN | extdiv; + } } } @@ -488,22 +567,29 @@ static void rcar_du_crtc_disable(struct drm_crtc *crtc) rcar_du_crtc_stop(rcrtc); rcar_du_crtc_put(rcrtc); + spin_lock_irq(&crtc->dev->event_lock); + if (crtc->state->event) { + drm_crtc_send_vblank_event(crtc, crtc->state->event); + crtc->state->event = NULL; + } + spin_unlock_irq(&crtc->dev->event_lock); + rcrtc->outputs = 0; } static void rcar_du_crtc_atomic_begin(struct drm_crtc *crtc, struct drm_crtc_state *old_crtc_state) { - struct drm_pending_vblank_event *event = crtc->state->event; struct rcar_du_crtc *rcrtc = to_rcar_crtc(crtc); struct drm_device *dev = rcrtc->crtc.dev; unsigned long flags; - if (event) { + if (crtc->state->event) { WARN_ON(drm_crtc_vblank_get(crtc) != 0); spin_lock_irqsave(&dev->event_lock, flags); - rcrtc->event = event; + rcrtc->event = crtc->state->event; + crtc->state->event = NULL; spin_unlock_irqrestore(&dev->event_lock, flags); } diff --git a/drivers/gpu/drm/rcar-du/rcar_du_crtc.h b/drivers/gpu/drm/rcar-du/rcar_du_crtc.h index a7194812997e..15871fae7445 100644 --- a/drivers/gpu/drm/rcar-du/rcar_du_crtc.h +++ b/drivers/gpu/drm/rcar-du/rcar_du_crtc.h @@ -1,7 +1,7 @@ /* * rcar_du_crtc.h -- R-Car Display Unit CRTCs * - * Copyright (C) 2013-2014 Renesas Electronics Corporation + * Copyright (C) 2013-2015 Renesas Electronics Corporation * * Contact: Laurent Pinchart (laurent.pinchart@ideasonboard.com) * @@ -61,6 +61,8 @@ enum rcar_du_output { RCAR_DU_OUTPUT_DPAD1, RCAR_DU_OUTPUT_LVDS0, RCAR_DU_OUTPUT_LVDS1, + RCAR_DU_OUTPUT_HDMI0, + RCAR_DU_OUTPUT_HDMI1, RCAR_DU_OUTPUT_TCON, RCAR_DU_OUTPUT_MAX, }; diff --git a/drivers/gpu/drm/rcar-du/rcar_du_drv.c b/drivers/gpu/drm/rcar-du/rcar_du_drv.c index 62a3b3e32153..d6a0255181cc 100644 --- a/drivers/gpu/drm/rcar-du/rcar_du_drv.c +++ b/drivers/gpu/drm/rcar-du/rcar_du_drv.c @@ -44,12 +44,10 @@ static const struct rcar_du_device_info rcar_du_r8a7779_info = { */ [RCAR_DU_OUTPUT_DPAD0] = { .possible_crtcs = BIT(0), - .encoder_type = DRM_MODE_ENCODER_NONE, .port = 0, }, [RCAR_DU_OUTPUT_DPAD1] = { .possible_crtcs = BIT(1) | BIT(0), - .encoder_type = DRM_MODE_ENCODER_NONE, .port = 1, }, }, @@ -68,17 +66,14 @@ static const struct rcar_du_device_info rcar_du_r8a7790_info = { */ [RCAR_DU_OUTPUT_DPAD0] = { .possible_crtcs = BIT(2) | BIT(1) | BIT(0), - .encoder_type = DRM_MODE_ENCODER_NONE, .port = 0, }, [RCAR_DU_OUTPUT_LVDS0] = { .possible_crtcs = BIT(0), - .encoder_type = DRM_MODE_ENCODER_LVDS, .port = 1, }, [RCAR_DU_OUTPUT_LVDS1] = { .possible_crtcs = BIT(2) | BIT(1), - .encoder_type = DRM_MODE_ENCODER_LVDS, .port = 2, }, }, @@ -97,12 +92,10 @@ static const struct rcar_du_device_info rcar_du_r8a7791_info = { */ [RCAR_DU_OUTPUT_DPAD0] = { .possible_crtcs = BIT(1) | BIT(0), - .encoder_type = DRM_MODE_ENCODER_NONE, .port = 0, }, [RCAR_DU_OUTPUT_LVDS0] = { .possible_crtcs = BIT(0), - .encoder_type = DRM_MODE_ENCODER_LVDS, .port = 1, }, }, @@ -118,12 +111,10 @@ static const struct rcar_du_device_info rcar_du_r8a7792_info = { /* R8A7792 has two RGB outputs. */ [RCAR_DU_OUTPUT_DPAD0] = { .possible_crtcs = BIT(0), - .encoder_type = DRM_MODE_ENCODER_NONE, .port = 0, }, [RCAR_DU_OUTPUT_DPAD1] = { .possible_crtcs = BIT(1), - .encoder_type = DRM_MODE_ENCODER_NONE, .port = 1, }, }, @@ -141,12 +132,10 @@ static const struct rcar_du_device_info rcar_du_r8a7794_info = { */ [RCAR_DU_OUTPUT_DPAD0] = { .possible_crtcs = BIT(0), - .encoder_type = DRM_MODE_ENCODER_NONE, .port = 0, }, [RCAR_DU_OUTPUT_DPAD1] = { .possible_crtcs = BIT(1), - .encoder_type = DRM_MODE_ENCODER_NONE, .port = 1, }, }, @@ -160,21 +149,28 @@ 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, one LVDS output and two - * (currently unsupported) HDMI outputs. + /* R8A7795 has one RGB output, two HDMI outputs and one + * LVDS output. */ [RCAR_DU_OUTPUT_DPAD0] = { .possible_crtcs = BIT(3), - .encoder_type = DRM_MODE_ENCODER_NONE, .port = 0, }, + [RCAR_DU_OUTPUT_HDMI0] = { + .possible_crtcs = BIT(1), + .port = 1, + }, + [RCAR_DU_OUTPUT_HDMI1] = { + .possible_crtcs = BIT(2), + .port = 2, + }, [RCAR_DU_OUTPUT_LVDS0] = { .possible_crtcs = BIT(0), - .encoder_type = DRM_MODE_ENCODER_LVDS, .port = 3, }, }, .num_lvds = 1, + .dpll_ch = BIT(1) | BIT(2), }; static const struct rcar_du_device_info rcar_du_r8a7796_info = { @@ -189,12 +185,10 @@ static const struct rcar_du_device_info rcar_du_r8a7796_info = { */ [RCAR_DU_OUTPUT_DPAD0] = { .possible_crtcs = BIT(2), - .encoder_type = DRM_MODE_ENCODER_NONE, .port = 0, }, [RCAR_DU_OUTPUT_LVDS0] = { .possible_crtcs = BIT(0), - .encoder_type = DRM_MODE_ENCODER_LVDS, .port = 2, }, }, @@ -318,10 +312,8 @@ static int rcar_du_probe(struct platform_device *pdev) if (rcdu == NULL) return -ENOMEM; - init_waitqueue_head(&rcdu->commit.wait); - rcdu->dev = &pdev->dev; - rcdu->info = of_match_device(rcar_du_of_table, rcdu->dev)->data; + rcdu->info = of_device_get_match_data(rcdu->dev); platform_set_drvdata(pdev, rcdu); diff --git a/drivers/gpu/drm/rcar-du/rcar_du_drv.h b/drivers/gpu/drm/rcar-du/rcar_du_drv.h index c843c3134498..f8cd79488ece 100644 --- a/drivers/gpu/drm/rcar-du/rcar_du_drv.h +++ b/drivers/gpu/drm/rcar-du/rcar_du_drv.h @@ -38,7 +38,6 @@ struct rcar_du_lvdsenc; /* * struct rcar_du_output_routing - Output routing specification * @possible_crtcs: bitmask of possible CRTCs for the output - * @encoder_type: DRM type of the internal encoder associated with the output * @port: device tree port number corresponding to this output route * * The DU has 5 possible outputs (DPAD0/1, LVDS0/1, TCON). Output routing data @@ -47,7 +46,6 @@ struct rcar_du_lvdsenc; */ struct rcar_du_output_routing { unsigned int possible_crtcs; - unsigned int encoder_type; unsigned int port; }; @@ -67,6 +65,7 @@ struct rcar_du_device_info { unsigned int num_crtcs; struct rcar_du_output_routing routes[RCAR_DU_OUTPUT_MAX]; unsigned int num_lvds; + unsigned int dpll_ch; }; #define RCAR_DU_MAX_CRTCS 4 @@ -98,11 +97,6 @@ struct rcar_du_device { unsigned int vspd1_sink; struct rcar_du_lvdsenc *lvds[RCAR_DU_MAX_LVDS]; - - struct { - wait_queue_head_t wait; - u32 pending; - } commit; }; static inline bool rcar_du_has(struct rcar_du_device *rcdu, diff --git a/drivers/gpu/drm/rcar-du/rcar_du_encoder.c b/drivers/gpu/drm/rcar-du/rcar_du_encoder.c index ab8645c57e2d..3e048dd98b64 100644 --- a/drivers/gpu/drm/rcar-du/rcar_du_encoder.c +++ b/drivers/gpu/drm/rcar-du/rcar_du_encoder.c @@ -16,14 +16,13 @@ #include <drm/drmP.h> #include <drm/drm_crtc.h> #include <drm/drm_crtc_helper.h> +#include <drm/drm_panel.h> #include "rcar_du_drv.h" #include "rcar_du_encoder.h" -#include "rcar_du_hdmienc.h" #include "rcar_du_kms.h" #include "rcar_du_lvdscon.h" #include "rcar_du_lvdsenc.h" -#include "rcar_du_vgacon.h" /* ----------------------------------------------------------------------------- * Encoder @@ -33,6 +32,11 @@ static void rcar_du_encoder_disable(struct drm_encoder *encoder) { struct rcar_du_encoder *renc = to_rcar_encoder(encoder); + if (renc->connector && renc->connector->panel) { + drm_panel_disable(renc->connector->panel); + drm_panel_unprepare(renc->connector->panel); + } + if (renc->lvds) rcar_du_lvdsenc_enable(renc->lvds, encoder->crtc, false); } @@ -43,6 +47,11 @@ static void rcar_du_encoder_enable(struct drm_encoder *encoder) if (renc->lvds) rcar_du_lvdsenc_enable(renc->lvds, encoder->crtc, true); + + if (renc->connector && renc->connector->panel) { + drm_panel_prepare(renc->connector->panel); + drm_panel_enable(renc->connector->panel); + } } static int rcar_du_encoder_atomic_check(struct drm_encoder *encoder, @@ -52,30 +61,36 @@ static int rcar_du_encoder_atomic_check(struct drm_encoder *encoder, struct rcar_du_encoder *renc = to_rcar_encoder(encoder); struct drm_display_mode *adjusted_mode = &crtc_state->adjusted_mode; const struct drm_display_mode *mode = &crtc_state->mode; - const struct drm_display_mode *panel_mode; struct drm_connector *connector = conn_state->connector; struct drm_device *dev = encoder->dev; - /* DAC encoders have currently no restriction on the mode. */ - if (encoder->encoder_type == DRM_MODE_ENCODER_DAC) - return 0; - - if (list_empty(&connector->modes)) { - dev_dbg(dev->dev, "encoder: empty modes list\n"); - return -EINVAL; + /* + * Only panel-related encoder types require validation here, everything + * else is handled by the bridge drivers. + */ + if (connector->connector_type == DRM_MODE_CONNECTOR_LVDS) { + const struct drm_display_mode *panel_mode; + + if (list_empty(&connector->modes)) { + dev_dbg(dev->dev, "encoder: empty modes list\n"); + return -EINVAL; + } + + panel_mode = list_first_entry(&connector->modes, + struct drm_display_mode, head); + + /* We're not allowed to modify the resolution. */ + if (mode->hdisplay != panel_mode->hdisplay || + mode->vdisplay != panel_mode->vdisplay) + return -EINVAL; + + /* + * The flat panel mode is fixed, just copy it to the adjusted + * mode. + */ + drm_mode_copy(adjusted_mode, panel_mode); } - panel_mode = list_first_entry(&connector->modes, - struct drm_display_mode, head); - - /* We're not allowed to modify the resolution. */ - if (mode->hdisplay != panel_mode->hdisplay || - mode->vdisplay != panel_mode->vdisplay) - return -EINVAL; - - /* The flat panel mode is fixed, just copy it to the adjusted mode. */ - drm_mode_copy(adjusted_mode, panel_mode); - if (renc->lvds) rcar_du_lvdsenc_atomic_check(renc->lvds, adjusted_mode); @@ -83,16 +98,54 @@ static int rcar_du_encoder_atomic_check(struct drm_encoder *encoder, } static void rcar_du_encoder_mode_set(struct drm_encoder *encoder, - struct drm_display_mode *mode, - struct drm_display_mode *adjusted_mode) + struct drm_crtc_state *crtc_state, + struct drm_connector_state *conn_state) { struct rcar_du_encoder *renc = to_rcar_encoder(encoder); + struct drm_display_info *info = &conn_state->connector->display_info; + enum rcar_lvds_mode mode; + + rcar_du_crtc_route_output(crtc_state->crtc, renc->output); + + if (!renc->lvds) { + /* + * The DU driver creates connectors only for the outputs of the + * internal LVDS encoders. + */ + renc->connector = NULL; + return; + } + + renc->connector = to_rcar_connector(conn_state->connector); + + if (!info->num_bus_formats || !info->bus_formats) { + dev_err(encoder->dev->dev, "no LVDS bus format reported\n"); + return; + } + + switch (info->bus_formats[0]) { + case MEDIA_BUS_FMT_RGB666_1X7X3_SPWG: + case MEDIA_BUS_FMT_RGB888_1X7X4_JEIDA: + mode = RCAR_LVDS_MODE_JEIDA; + break; + case MEDIA_BUS_FMT_RGB888_1X7X4_SPWG: + mode = RCAR_LVDS_MODE_VESA; + break; + default: + dev_err(encoder->dev->dev, + "unsupported LVDS bus format 0x%04x\n", + info->bus_formats[0]); + return; + } + + if (info->bus_flags & DRM_BUS_FLAG_DATA_LSB_TO_MSB) + mode |= RCAR_LVDS_MODE_MIRROR; - rcar_du_crtc_route_output(encoder->crtc, renc->output); + rcar_du_lvdsenc_set_mode(renc->lvds, mode); } static const struct drm_encoder_helper_funcs encoder_helper_funcs = { - .mode_set = rcar_du_encoder_mode_set, + .atomic_mode_set = rcar_du_encoder_mode_set, .disable = rcar_du_encoder_disable, .enable = rcar_du_encoder_enable, .atomic_check = rcar_du_encoder_atomic_check, @@ -103,14 +156,13 @@ static const struct drm_encoder_funcs encoder_funcs = { }; int rcar_du_encoder_init(struct rcar_du_device *rcdu, - enum rcar_du_encoder_type type, enum rcar_du_output output, struct device_node *enc_node, struct device_node *con_node) { struct rcar_du_encoder *renc; struct drm_encoder *encoder; - unsigned int encoder_type; + struct drm_bridge *bridge = NULL; int ret; renc = devm_kzalloc(rcdu->dev, sizeof(*renc), GFP_KERNEL); @@ -133,52 +185,51 @@ int rcar_du_encoder_init(struct rcar_du_device *rcdu, break; } - switch (type) { - case RCAR_DU_ENCODER_VGA: - encoder_type = DRM_MODE_ENCODER_DAC; - break; - case RCAR_DU_ENCODER_LVDS: - encoder_type = DRM_MODE_ENCODER_LVDS; - break; - case RCAR_DU_ENCODER_HDMI: - encoder_type = DRM_MODE_ENCODER_TMDS; - break; - case RCAR_DU_ENCODER_NONE: - default: - /* No external encoder, use the internal encoder type. */ - encoder_type = rcdu->info->routes[output].encoder_type; - break; - } + if (enc_node) { + dev_dbg(rcdu->dev, "initializing encoder %s for output %u\n", + of_node_full_name(enc_node), output); - if (type == RCAR_DU_ENCODER_HDMI) { - ret = rcar_du_hdmienc_init(rcdu, renc, enc_node); - if (ret < 0) + /* Locate the DRM bridge from the encoder DT node. */ + bridge = of_drm_find_bridge(enc_node); + if (!bridge) { + ret = -EPROBE_DEFER; goto done; + } } else { - ret = drm_encoder_init(rcdu->ddev, encoder, &encoder_funcs, - encoder_type, NULL); - if (ret < 0) - goto done; - - drm_encoder_helper_add(encoder, &encoder_helper_funcs); + dev_dbg(rcdu->dev, + "initializing internal encoder for output %u\n", + output); } - switch (encoder_type) { - case DRM_MODE_ENCODER_LVDS: - ret = rcar_du_lvds_connector_init(rcdu, renc, con_node); - break; - - case DRM_MODE_ENCODER_DAC: - ret = rcar_du_vga_connector_init(rcdu, renc); - break; - - case DRM_MODE_ENCODER_TMDS: - /* connector managed by the bridge driver */ - break; - - default: - ret = -EINVAL; - break; + ret = drm_encoder_init(rcdu->ddev, encoder, &encoder_funcs, + DRM_MODE_ENCODER_NONE, NULL); + if (ret < 0) + goto done; + + drm_encoder_helper_add(encoder, &encoder_helper_funcs); + + if (bridge) { + /* + * Attach the bridge to the encoder. The bridge will create the + * connector. + */ + ret = drm_bridge_attach(encoder, bridge, NULL); + if (ret) { + drm_encoder_cleanup(encoder); + return ret; + } + } else { + /* There's no bridge, create the connector manually. */ + switch (output) { + case RCAR_DU_OUTPUT_LVDS0: + case RCAR_DU_OUTPUT_LVDS1: + ret = rcar_du_lvds_connector_init(rcdu, renc, con_node); + break; + + default: + ret = -EINVAL; + break; + } } done: diff --git a/drivers/gpu/drm/rcar-du/rcar_du_encoder.h b/drivers/gpu/drm/rcar-du/rcar_du_encoder.h index a050a3699857..5422fa4df272 100644 --- a/drivers/gpu/drm/rcar-du/rcar_du_encoder.h +++ b/drivers/gpu/drm/rcar-du/rcar_du_encoder.h @@ -17,22 +17,14 @@ #include <drm/drm_crtc.h> #include <drm/drm_encoder.h> +struct drm_panel; struct rcar_du_device; -struct rcar_du_hdmienc; struct rcar_du_lvdsenc; -enum rcar_du_encoder_type { - RCAR_DU_ENCODER_UNUSED = 0, - RCAR_DU_ENCODER_NONE, - RCAR_DU_ENCODER_VGA, - RCAR_DU_ENCODER_LVDS, - RCAR_DU_ENCODER_HDMI, -}; - struct rcar_du_encoder { struct drm_encoder base; enum rcar_du_output output; - struct rcar_du_hdmienc *hdmi; + struct rcar_du_connector *connector; struct rcar_du_lvdsenc *lvds; }; @@ -44,13 +36,13 @@ struct rcar_du_encoder { struct rcar_du_connector { struct drm_connector connector; struct rcar_du_encoder *encoder; + struct drm_panel *panel; }; #define to_rcar_connector(c) \ container_of(c, struct rcar_du_connector, connector) int rcar_du_encoder_init(struct rcar_du_device *rcdu, - enum rcar_du_encoder_type type, enum rcar_du_output output, struct device_node *enc_node, struct device_node *con_node); diff --git a/drivers/gpu/drm/rcar-du/rcar_du_hdmienc.c b/drivers/gpu/drm/rcar-du/rcar_du_hdmienc.c deleted file mode 100644 index c4c5d1abcff8..000000000000 --- a/drivers/gpu/drm/rcar-du/rcar_du_hdmienc.c +++ /dev/null @@ -1,134 +0,0 @@ -/* - * R-Car Display Unit HDMI Encoder - * - * Copyright (C) 2014 Renesas Electronics Corporation - * - * Contact: Laurent Pinchart (laurent.pinchart@ideasonboard.com) - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. - */ - -#include <linux/slab.h> - -#include <drm/drmP.h> -#include <drm/drm_crtc.h> -#include <drm/drm_crtc_helper.h> - -#include "rcar_du_drv.h" -#include "rcar_du_encoder.h" -#include "rcar_du_hdmienc.h" -#include "rcar_du_lvdsenc.h" - -struct rcar_du_hdmienc { - struct rcar_du_encoder *renc; - bool enabled; -}; - -#define to_rcar_hdmienc(e) (to_rcar_encoder(e)->hdmi) - -static void rcar_du_hdmienc_disable(struct drm_encoder *encoder) -{ - struct rcar_du_hdmienc *hdmienc = to_rcar_hdmienc(encoder); - - if (hdmienc->renc->lvds) - rcar_du_lvdsenc_enable(hdmienc->renc->lvds, encoder->crtc, - false); - - hdmienc->enabled = false; -} - -static void rcar_du_hdmienc_enable(struct drm_encoder *encoder) -{ - struct rcar_du_hdmienc *hdmienc = to_rcar_hdmienc(encoder); - - if (hdmienc->renc->lvds) - rcar_du_lvdsenc_enable(hdmienc->renc->lvds, encoder->crtc, - true); - - hdmienc->enabled = true; -} - -static int rcar_du_hdmienc_atomic_check(struct drm_encoder *encoder, - struct drm_crtc_state *crtc_state, - struct drm_connector_state *conn_state) -{ - struct rcar_du_hdmienc *hdmienc = to_rcar_hdmienc(encoder); - struct drm_display_mode *adjusted_mode = &crtc_state->adjusted_mode; - - if (hdmienc->renc->lvds) - rcar_du_lvdsenc_atomic_check(hdmienc->renc->lvds, - adjusted_mode); - - return 0; -} - - -static void rcar_du_hdmienc_mode_set(struct drm_encoder *encoder, - struct drm_display_mode *mode, - struct drm_display_mode *adjusted_mode) -{ - struct rcar_du_hdmienc *hdmienc = to_rcar_hdmienc(encoder); - - rcar_du_crtc_route_output(encoder->crtc, hdmienc->renc->output); -} - -static const struct drm_encoder_helper_funcs encoder_helper_funcs = { - .mode_set = rcar_du_hdmienc_mode_set, - .disable = rcar_du_hdmienc_disable, - .enable = rcar_du_hdmienc_enable, - .atomic_check = rcar_du_hdmienc_atomic_check, -}; - -static void rcar_du_hdmienc_cleanup(struct drm_encoder *encoder) -{ - struct rcar_du_hdmienc *hdmienc = to_rcar_hdmienc(encoder); - - if (hdmienc->enabled) - rcar_du_hdmienc_disable(encoder); - - drm_encoder_cleanup(encoder); -} - -static const struct drm_encoder_funcs encoder_funcs = { - .destroy = rcar_du_hdmienc_cleanup, -}; - -int rcar_du_hdmienc_init(struct rcar_du_device *rcdu, - struct rcar_du_encoder *renc, struct device_node *np) -{ - struct drm_encoder *encoder = rcar_encoder_to_drm_encoder(renc); - struct drm_bridge *bridge; - struct rcar_du_hdmienc *hdmienc; - int ret; - - hdmienc = devm_kzalloc(rcdu->dev, sizeof(*hdmienc), GFP_KERNEL); - if (hdmienc == NULL) - return -ENOMEM; - - /* Locate the DRM bridge from the HDMI encoder DT node. */ - bridge = of_drm_find_bridge(np); - if (!bridge) - return -EPROBE_DEFER; - - ret = drm_encoder_init(rcdu->ddev, encoder, &encoder_funcs, - DRM_MODE_ENCODER_TMDS, NULL); - if (ret < 0) - return ret; - - drm_encoder_helper_add(encoder, &encoder_helper_funcs); - - renc->hdmi = hdmienc; - hdmienc->renc = renc; - - /* Link the bridge to the encoder. */ - ret = drm_bridge_attach(encoder, bridge, NULL); - if (ret) { - drm_encoder_cleanup(encoder); - return ret; - } - - return 0; -} diff --git a/drivers/gpu/drm/rcar-du/rcar_du_hdmienc.h b/drivers/gpu/drm/rcar-du/rcar_du_hdmienc.h deleted file mode 100644 index 2ff0128ac8e1..000000000000 --- a/drivers/gpu/drm/rcar-du/rcar_du_hdmienc.h +++ /dev/null @@ -1,35 +0,0 @@ -/* - * R-Car Display Unit HDMI Encoder - * - * Copyright (C) 2014 Renesas Electronics Corporation - * - * Contact: Laurent Pinchart (laurent.pinchart@ideasonboard.com) - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. - */ - -#ifndef __RCAR_DU_HDMIENC_H__ -#define __RCAR_DU_HDMIENC_H__ - -#include <linux/module.h> - -struct device_node; -struct rcar_du_device; -struct rcar_du_encoder; - -#if IS_ENABLED(CONFIG_DRM_RCAR_HDMI) -int rcar_du_hdmienc_init(struct rcar_du_device *rcdu, - struct rcar_du_encoder *renc, struct device_node *np); -#else -static inline int rcar_du_hdmienc_init(struct rcar_du_device *rcdu, - struct rcar_du_encoder *renc, - struct device_node *np) -{ - return -ENOSYS; -} -#endif - -#endif /* __RCAR_DU_HDMIENC_H__ */ diff --git a/drivers/gpu/drm/rcar-du/rcar_du_kms.c b/drivers/gpu/drm/rcar-du/rcar_du_kms.c index ff61f6032f2c..f4125c8ca902 100644 --- a/drivers/gpu/drm/rcar-du/rcar_du_kms.c +++ b/drivers/gpu/drm/rcar-du/rcar_du_kms.c @@ -249,18 +249,9 @@ static int rcar_du_atomic_check(struct drm_device *dev, return rcar_du_atomic_check_planes(dev, state); } -struct rcar_du_commit { - struct work_struct work; - struct drm_device *dev; - struct drm_atomic_state *state; - u32 crtcs; -}; - -static void rcar_du_atomic_complete(struct rcar_du_commit *commit) +static void rcar_du_atomic_commit_tail(struct drm_atomic_state *old_state) { - struct drm_device *dev = commit->dev; - struct rcar_du_device *rcdu = dev->dev_private; - struct drm_atomic_state *old_state = commit->state; + struct drm_device *dev = old_state->dev; /* Apply the atomic update. */ drm_atomic_helper_commit_modeset_disables(dev, old_state); @@ -268,114 +259,31 @@ static void rcar_du_atomic_complete(struct rcar_du_commit *commit) drm_atomic_helper_commit_planes(dev, old_state, DRM_PLANE_COMMIT_ACTIVE_ONLY); + drm_atomic_helper_commit_hw_done(old_state); drm_atomic_helper_wait_for_vblanks(dev, old_state); drm_atomic_helper_cleanup_planes(dev, old_state); - - drm_atomic_state_put(old_state); - - /* Complete the commit, wake up any waiter. */ - spin_lock(&rcdu->commit.wait.lock); - rcdu->commit.pending &= ~commit->crtcs; - wake_up_all_locked(&rcdu->commit.wait); - spin_unlock(&rcdu->commit.wait.lock); - - kfree(commit); -} - -static void rcar_du_atomic_work(struct work_struct *work) -{ - struct rcar_du_commit *commit = - container_of(work, struct rcar_du_commit, work); - - rcar_du_atomic_complete(commit); -} - -static int rcar_du_atomic_commit(struct drm_device *dev, - struct drm_atomic_state *state, - bool nonblock) -{ - struct rcar_du_device *rcdu = dev->dev_private; - struct rcar_du_commit *commit; - struct drm_crtc *crtc; - struct drm_crtc_state *crtc_state; - unsigned int i; - int ret; - - ret = drm_atomic_helper_prepare_planes(dev, state); - if (ret) - return ret; - - /* Allocate the commit object. */ - commit = kzalloc(sizeof(*commit), GFP_KERNEL); - if (commit == NULL) { - ret = -ENOMEM; - goto error; - } - - INIT_WORK(&commit->work, rcar_du_atomic_work); - commit->dev = dev; - commit->state = state; - - /* Wait until all affected CRTCs have completed previous commits and - * mark them as pending. - */ - for_each_crtc_in_state(state, crtc, crtc_state, i) - commit->crtcs |= drm_crtc_mask(crtc); - - spin_lock(&rcdu->commit.wait.lock); - ret = wait_event_interruptible_locked(rcdu->commit.wait, - !(rcdu->commit.pending & commit->crtcs)); - if (ret == 0) - rcdu->commit.pending |= commit->crtcs; - spin_unlock(&rcdu->commit.wait.lock); - - if (ret) { - kfree(commit); - goto error; - } - - /* Swap the state, this is the point of no return. */ - drm_atomic_helper_swap_state(state, true); - - drm_atomic_state_get(state); - if (nonblock) - schedule_work(&commit->work); - else - rcar_du_atomic_complete(commit); - - return 0; - -error: - drm_atomic_helper_cleanup_planes(dev, state); - return ret; } /* ----------------------------------------------------------------------------- * Initialization */ +static const struct drm_mode_config_helper_funcs rcar_du_mode_config_helper = { + .atomic_commit_tail = rcar_du_atomic_commit_tail, +}; + static const struct drm_mode_config_funcs rcar_du_mode_config_funcs = { .fb_create = rcar_du_fb_create, .output_poll_changed = rcar_du_output_poll_changed, .atomic_check = rcar_du_atomic_check, - .atomic_commit = rcar_du_atomic_commit, + .atomic_commit = drm_atomic_helper_commit, }; static int rcar_du_encoders_init_one(struct rcar_du_device *rcdu, enum rcar_du_output output, struct of_endpoint *ep) { - static const struct { - const char *compatible; - enum rcar_du_encoder_type type; - } encoders[] = { - { "adi,adv7123", RCAR_DU_ENCODER_VGA }, - { "adi,adv7511w", RCAR_DU_ENCODER_HDMI }, - { "thine,thc63lvdm83d", RCAR_DU_ENCODER_LVDS }, - }; - - enum rcar_du_encoder_type enc_type = RCAR_DU_ENCODER_NONE; struct device_node *connector = NULL; struct device_node *encoder = NULL; struct device_node *ep_node = NULL; @@ -394,6 +302,13 @@ static int rcar_du_encoders_init_one(struct rcar_du_device *rcdu, return -ENODEV; } + if (!of_device_is_available(entity)) { + dev_dbg(rcdu->dev, + "connected entity %s is disabled, skipping\n", + entity->full_name); + return -ENODEV; + } + entity_ep_node = of_parse_phandle(ep->local_node, "remote-endpoint", 0); for_each_endpoint_of_node(entity, ep_node) { @@ -422,30 +337,7 @@ static int rcar_du_encoders_init_one(struct rcar_du_device *rcdu, of_node_put(entity_ep_node); - if (encoder) { - /* - * If an encoder has been found, get its type based on its - * compatible string. - */ - unsigned int i; - - for (i = 0; i < ARRAY_SIZE(encoders); ++i) { - if (of_device_is_compatible(encoder, - encoders[i].compatible)) { - enc_type = encoders[i].type; - break; - } - } - - if (i == ARRAY_SIZE(encoders)) { - dev_warn(rcdu->dev, - "unknown encoder type for %s, skipping\n", - encoder->full_name); - of_node_put(encoder); - of_node_put(connector); - return -EINVAL; - } - } else { + if (!encoder) { /* * If no encoder has been found the entity must be the * connector. @@ -453,7 +345,7 @@ static int rcar_du_encoders_init_one(struct rcar_du_device *rcdu, connector = entity; } - ret = rcar_du_encoder_init(rcdu, enc_type, output, encoder, connector); + ret = rcar_du_encoder_init(rcdu, output, encoder, connector); if (ret && ret != -EPROBE_DEFER) dev_warn(rcdu->dev, "failed to initialize encoder %s on output %u (%d), skipping\n", @@ -561,6 +453,7 @@ int rcar_du_modeset_init(struct rcar_du_device *rcdu) dev->mode_config.max_width = 4095; dev->mode_config.max_height = 2047; dev->mode_config.funcs = &rcar_du_mode_config_funcs; + dev->mode_config.helper_private = &rcar_du_mode_config_helper; rcdu->num_crtcs = rcdu->info->num_crtcs; diff --git a/drivers/gpu/drm/rcar-du/rcar_du_lvdscon.c b/drivers/gpu/drm/rcar-du/rcar_du_lvdscon.c index 3bcfd161c53f..ee91481131ad 100644 --- a/drivers/gpu/drm/rcar-du/rcar_du_lvdscon.c +++ b/drivers/gpu/drm/rcar-du/rcar_du_lvdscon.c @@ -15,6 +15,7 @@ #include <drm/drm_atomic_helper.h> #include <drm/drm_crtc.h> #include <drm/drm_crtc_helper.h> +#include <drm/drm_panel.h> #include <video/display_timing.h> #include <video/of_display_timing.h> @@ -25,47 +26,30 @@ #include "rcar_du_kms.h" #include "rcar_du_lvdscon.h" -struct rcar_du_lvds_connector { - struct rcar_du_connector connector; - - struct { - unsigned int width_mm; /* Panel width in mm */ - unsigned int height_mm; /* Panel height in mm */ - struct videomode mode; - } panel; -}; - -#define to_rcar_lvds_connector(c) \ - container_of(c, struct rcar_du_lvds_connector, connector.connector) - static int rcar_du_lvds_connector_get_modes(struct drm_connector *connector) { - struct rcar_du_lvds_connector *lvdscon = - to_rcar_lvds_connector(connector); - struct drm_display_mode *mode; - - mode = drm_mode_create(connector->dev); - if (mode == NULL) - return 0; + struct rcar_du_connector *rcon = to_rcar_connector(connector); - mode->type = DRM_MODE_TYPE_PREFERRED | DRM_MODE_TYPE_DRIVER; - - drm_display_mode_from_videomode(&lvdscon->panel.mode, mode); - - drm_mode_probed_add(connector, mode); - - return 1; + return drm_panel_get_modes(rcon->panel); } static const struct drm_connector_helper_funcs connector_helper_funcs = { .get_modes = rcar_du_lvds_connector_get_modes, }; +static void rcar_du_lvds_connector_destroy(struct drm_connector *connector) +{ + struct rcar_du_connector *rcon = to_rcar_connector(connector); + + drm_panel_detach(rcon->panel); + drm_connector_cleanup(connector); +} + static const struct drm_connector_funcs connector_funcs = { .dpms = drm_atomic_helper_connector_dpms, .reset = drm_atomic_helper_connector_reset, .fill_modes = drm_helper_probe_single_connector_modes, - .destroy = drm_connector_cleanup, + .destroy = rcar_du_lvds_connector_destroy, .atomic_duplicate_state = drm_atomic_helper_connector_duplicate_state, .atomic_destroy_state = drm_atomic_helper_connector_destroy_state, }; @@ -75,27 +59,19 @@ int rcar_du_lvds_connector_init(struct rcar_du_device *rcdu, const struct device_node *np) { struct drm_encoder *encoder = rcar_encoder_to_drm_encoder(renc); - struct rcar_du_lvds_connector *lvdscon; + struct rcar_du_connector *rcon; struct drm_connector *connector; - struct display_timing timing; int ret; - lvdscon = devm_kzalloc(rcdu->dev, sizeof(*lvdscon), GFP_KERNEL); - if (lvdscon == NULL) + rcon = devm_kzalloc(rcdu->dev, sizeof(*rcon), GFP_KERNEL); + if (rcon == NULL) return -ENOMEM; - ret = of_get_display_timing(np, "panel-timing", &timing); - if (ret < 0) - return ret; - - videomode_from_timing(&timing, &lvdscon->panel.mode); + connector = &rcon->connector; - of_property_read_u32(np, "width-mm", &lvdscon->panel.width_mm); - of_property_read_u32(np, "height-mm", &lvdscon->panel.height_mm); - - connector = &lvdscon->connector.connector; - connector->display_info.width_mm = lvdscon->panel.width_mm; - connector->display_info.height_mm = lvdscon->panel.height_mm; + rcon->panel = of_drm_find_panel(np); + if (!rcon->panel) + return -EPROBE_DEFER; ret = drm_connector_init(rcdu->ddev, connector, &connector_funcs, DRM_MODE_CONNECTOR_LVDS); @@ -112,7 +88,11 @@ int rcar_du_lvds_connector_init(struct rcar_du_device *rcdu, if (ret < 0) return ret; - lvdscon->connector.encoder = renc; + ret = drm_panel_attach(rcon->panel, connector); + if (ret < 0) + return ret; + + rcon->encoder = renc; return 0; } diff --git a/drivers/gpu/drm/rcar-du/rcar_du_lvdsenc.c b/drivers/gpu/drm/rcar-du/rcar_du_lvdsenc.c index e3a4985f6f3f..1661f6201210 100644 --- a/drivers/gpu/drm/rcar-du/rcar_du_lvdsenc.c +++ b/drivers/gpu/drm/rcar-du/rcar_du_lvdsenc.c @@ -31,6 +31,7 @@ struct rcar_du_lvdsenc { bool enabled; enum rcar_lvds_input input; + enum rcar_lvds_mode mode; }; static void rcar_lvds_write(struct rcar_du_lvdsenc *lvds, u32 reg, u32 data) @@ -61,7 +62,7 @@ static void rcar_du_lvdsenc_start_gen2(struct rcar_du_lvdsenc *lvds, /* Select the input, hardcode mode 0, enable LVDS operation and turn * bias circuitry on. */ - lvdcr0 = LVDCR0_BEN | LVDCR0_LVEN; + lvdcr0 = (lvds->mode << LVDCR0_LVMD_SHIFT) | LVDCR0_BEN | LVDCR0_LVEN; if (rcrtc->index == 2) lvdcr0 |= LVDCR0_DUSEL; rcar_lvds_write(lvds, LVDCR0, lvdcr0); @@ -114,7 +115,7 @@ static void rcar_du_lvdsenc_start_gen3(struct rcar_du_lvdsenc *lvds, * Turn the PLL on, set it to LVDS normal mode, wait for the startup * delay and turn the output on. */ - lvdcr0 = LVDCR0_PLLON; + lvdcr0 = (lvds->mode << LVDCR0_LVMD_SHIFT) | LVDCR0_PLLON; rcar_lvds_write(lvds, LVDCR0, lvdcr0); lvdcr0 |= LVDCR0_PWD; @@ -211,6 +212,12 @@ void rcar_du_lvdsenc_atomic_check(struct rcar_du_lvdsenc *lvds, mode->clock = clamp(mode->clock, 25175, 148500); } +void rcar_du_lvdsenc_set_mode(struct rcar_du_lvdsenc *lvds, + enum rcar_lvds_mode mode) +{ + lvds->mode = mode; +} + static int rcar_du_lvdsenc_get_resources(struct rcar_du_lvdsenc *lvds, struct platform_device *pdev) { diff --git a/drivers/gpu/drm/rcar-du/rcar_du_lvdsenc.h b/drivers/gpu/drm/rcar-du/rcar_du_lvdsenc.h index dfdba746edf4..7218ac89333e 100644 --- a/drivers/gpu/drm/rcar-du/rcar_du_lvdsenc.h +++ b/drivers/gpu/drm/rcar-du/rcar_du_lvdsenc.h @@ -26,8 +26,17 @@ enum rcar_lvds_input { RCAR_LVDS_INPUT_DU2, }; +/* Keep in sync with the LVDCR0.LVMD hardware register values. */ +enum rcar_lvds_mode { + RCAR_LVDS_MODE_JEIDA = 0, + RCAR_LVDS_MODE_MIRROR = 1, + RCAR_LVDS_MODE_VESA = 4, +}; + #if IS_ENABLED(CONFIG_DRM_RCAR_LVDS) int rcar_du_lvdsenc_init(struct rcar_du_device *rcdu); +void rcar_du_lvdsenc_set_mode(struct rcar_du_lvdsenc *lvds, + enum rcar_lvds_mode mode); int rcar_du_lvdsenc_enable(struct rcar_du_lvdsenc *lvds, struct drm_crtc *crtc, bool enable); void rcar_du_lvdsenc_atomic_check(struct rcar_du_lvdsenc *lvds, @@ -37,6 +46,10 @@ static inline int rcar_du_lvdsenc_init(struct rcar_du_device *rcdu) { return 0; } +static inline void rcar_du_lvdsenc_set_mode(struct rcar_du_lvdsenc *lvds, + enum rcar_lvds_mode mode) +{ +} static inline int rcar_du_lvdsenc_enable(struct rcar_du_lvdsenc *lvds, struct drm_crtc *crtc, bool enable) { diff --git a/drivers/gpu/drm/rcar-du/rcar_du_regs.h b/drivers/gpu/drm/rcar-du/rcar_du_regs.h index fedb0161e234..d5bae99d3cfe 100644 --- a/drivers/gpu/drm/rcar-du/rcar_du_regs.h +++ b/drivers/gpu/drm/rcar-du/rcar_du_regs.h @@ -277,6 +277,29 @@ #define DEFR10_TSEL_H3_TCON1 (0 << 1) /* DEFR102 register only (DU2/DU3) */ #define DEFR10_DEFE10 (1 << 0) +#define DPLLCR 0x20044 +#define DPLLCR_CODE (0x95 << 24) +#define DPLLCR_PLCS1 (1 << 23) +/* + * PLCS0 is bit 21, but H3 ES1.x requires bit 20 to be set as well. As bit 20 + * isn't implemented by other SoC in the Gen3 family it can safely be set + * unconditionally. + */ +#define DPLLCR_PLCS0 (3 << 20) +#define DPLLCR_CLKE (1 << 18) +#define DPLLCR_FDPLL(n) ((n) << 12) +#define DPLLCR_N(n) ((n) << 5) +#define DPLLCR_M(n) ((n) << 3) +#define DPLLCR_STBY (1 << 2) +#define DPLLCR_INCS_DOTCLKIN0 (0 << 0) +#define DPLLCR_INCS_DOTCLKIN1 (1 << 1) + +#define DPLLC2R 0x20048 +#define DPLLC2R_CODE (0x95 << 24) +#define DPLLC2R_SELC (1 << 12) +#define DPLLC2R_M(n) ((n) << 8) +#define DPLLC2R_FDPLL(n) ((n) << 0) + /* ----------------------------------------------------------------------------- * Display Timing Generation Registers */ diff --git a/drivers/gpu/drm/rcar-du/rcar_du_vgacon.c b/drivers/gpu/drm/rcar-du/rcar_du_vgacon.c deleted file mode 100644 index 8d6125c1c0f9..000000000000 --- a/drivers/gpu/drm/rcar-du/rcar_du_vgacon.c +++ /dev/null @@ -1,82 +0,0 @@ -/* - * rcar_du_vgacon.c -- R-Car Display Unit VGA Connector - * - * Copyright (C) 2013-2014 Renesas Electronics Corporation - * - * Contact: Laurent Pinchart (laurent.pinchart@ideasonboard.com) - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. - */ - -#include <drm/drmP.h> -#include <drm/drm_atomic_helper.h> -#include <drm/drm_crtc.h> -#include <drm/drm_crtc_helper.h> - -#include "rcar_du_drv.h" -#include "rcar_du_encoder.h" -#include "rcar_du_kms.h" -#include "rcar_du_vgacon.h" - -static int rcar_du_vga_connector_get_modes(struct drm_connector *connector) -{ - return 0; -} - -static const struct drm_connector_helper_funcs connector_helper_funcs = { - .get_modes = rcar_du_vga_connector_get_modes, -}; - -static enum drm_connector_status -rcar_du_vga_connector_detect(struct drm_connector *connector, bool force) -{ - return connector_status_connected; -} - -static const struct drm_connector_funcs connector_funcs = { - .dpms = drm_atomic_helper_connector_dpms, - .reset = drm_atomic_helper_connector_reset, - .detect = rcar_du_vga_connector_detect, - .fill_modes = drm_helper_probe_single_connector_modes, - .destroy = drm_connector_cleanup, - .atomic_duplicate_state = drm_atomic_helper_connector_duplicate_state, - .atomic_destroy_state = drm_atomic_helper_connector_destroy_state, -}; - -int rcar_du_vga_connector_init(struct rcar_du_device *rcdu, - struct rcar_du_encoder *renc) -{ - struct drm_encoder *encoder = rcar_encoder_to_drm_encoder(renc); - struct rcar_du_connector *rcon; - struct drm_connector *connector; - int ret; - - rcon = devm_kzalloc(rcdu->dev, sizeof(*rcon), GFP_KERNEL); - if (rcon == NULL) - return -ENOMEM; - - connector = &rcon->connector; - connector->display_info.width_mm = 0; - connector->display_info.height_mm = 0; - connector->interlace_allowed = true; - - ret = drm_connector_init(rcdu->ddev, connector, &connector_funcs, - DRM_MODE_CONNECTOR_VGA); - if (ret < 0) - return ret; - - drm_connector_helper_add(connector, &connector_helper_funcs); - - connector->dpms = DRM_MODE_DPMS_OFF; - drm_object_property_set_value(&connector->base, - rcdu->ddev->mode_config.dpms_property, DRM_MODE_DPMS_OFF); - - ret = drm_mode_connector_attach_encoder(connector, encoder); - if (ret < 0) - return ret; - - return 0; -} diff --git a/drivers/gpu/drm/rcar-du/rcar_du_vgacon.h b/drivers/gpu/drm/rcar-du/rcar_du_vgacon.h deleted file mode 100644 index 112f50316e01..000000000000 --- a/drivers/gpu/drm/rcar-du/rcar_du_vgacon.h +++ /dev/null @@ -1,23 +0,0 @@ -/* - * rcar_du_vgacon.h -- R-Car Display Unit VGA Connector - * - * Copyright (C) 2013-2014 Renesas Electronics Corporation - * - * Contact: Laurent Pinchart (laurent.pinchart@ideasonboard.com) - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. - */ - -#ifndef __RCAR_DU_VGACON_H__ -#define __RCAR_DU_VGACON_H__ - -struct rcar_du_device; -struct rcar_du_encoder; - -int rcar_du_vga_connector_init(struct rcar_du_device *rcdu, - struct rcar_du_encoder *renc); - -#endif /* __RCAR_DU_VGACON_H__ */ diff --git a/drivers/gpu/drm/rcar-du/rcar_du_vsp.h b/drivers/gpu/drm/rcar-du/rcar_du_vsp.h index 510dcc9c6816..f1d0f1824528 100644 --- a/drivers/gpu/drm/rcar-du/rcar_du_vsp.h +++ b/drivers/gpu/drm/rcar-du/rcar_du_vsp.h @@ -68,7 +68,7 @@ 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 0; }; +static inline int rcar_du_vsp_init(struct rcar_du_vsp *vsp) { 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 new file mode 100644 index 000000000000..7539626b8ebd --- /dev/null +++ b/drivers/gpu/drm/rcar-du/rcar_dw_hdmi.c @@ -0,0 +1,100 @@ +/* + * R-Car Gen3 HDMI PHY + * + * Copyright (C) 2016 Renesas Electronics Corporation + * + * Contact: Laurent Pinchart (laurent.pinchart@ideasonboard.com) + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + */ + +#include <linux/module.h> +#include <linux/platform_device.h> + +#include <drm/bridge/dw_hdmi.h> + +#define RCAR_HDMI_PHY_OPMODE_PLLCFG 0x06 /* Mode of operation and PLL dividers */ +#define RCAR_HDMI_PHY_PLLCURRGMPCTRL 0x10 /* PLL current and Gmp (conductance) */ +#define RCAR_HDMI_PHY_PLLDIVCTRL 0x11 /* PLL dividers */ + +struct rcar_hdmi_phy_params { + unsigned long mpixelclock; + u16 opmode_div; /* Mode of operation and PLL dividers */ + u16 curr_gmp; /* PLL current and Gmp (conductance) */ + u16 div; /* PLL dividers */ +}; + +static const struct rcar_hdmi_phy_params rcar_hdmi_phy_params[] = { + { 35500000, 0x0003, 0x0344, 0x0328 }, + { 44900000, 0x0003, 0x0285, 0x0128 }, + { 71000000, 0x0002, 0x1184, 0x0314 }, + { 90000000, 0x0002, 0x1144, 0x0114 }, + { 140250000, 0x0001, 0x20c4, 0x030a }, + { 182750000, 0x0001, 0x2084, 0x010a }, + { 281250000, 0x0000, 0x0084, 0x0305 }, + { 297000000, 0x0000, 0x0084, 0x0105 }, + { ~0UL, 0x0000, 0x0000, 0x0000 }, +}; + +static int rcar_hdmi_phy_configure(struct dw_hdmi *hdmi, + const struct dw_hdmi_plat_data *pdata, + unsigned long mpixelclock) +{ + const struct rcar_hdmi_phy_params *params = rcar_hdmi_phy_params; + + for (; params && params->mpixelclock != ~0UL; ++params) { + if (mpixelclock <= params->mpixelclock) + break; + } + + if (params->mpixelclock == ~0UL) + return -EINVAL; + + dw_hdmi_phy_i2c_write(hdmi, params->opmode_div, + RCAR_HDMI_PHY_OPMODE_PLLCFG); + dw_hdmi_phy_i2c_write(hdmi, params->curr_gmp, + RCAR_HDMI_PHY_PLLCURRGMPCTRL); + dw_hdmi_phy_i2c_write(hdmi, params->div, RCAR_HDMI_PHY_PLLDIVCTRL); + + return 0; +} + +static const struct dw_hdmi_plat_data rcar_dw_hdmi_plat_data = { + .configure_phy = rcar_hdmi_phy_configure, +}; + +static int rcar_dw_hdmi_probe(struct platform_device *pdev) +{ + return dw_hdmi_probe(pdev, &rcar_dw_hdmi_plat_data); +} + +static int rcar_dw_hdmi_remove(struct platform_device *pdev) +{ + dw_hdmi_remove(pdev); + + return 0; +} + +static const struct of_device_id rcar_dw_hdmi_of_table[] = { + { .compatible = "renesas,rcar-gen3-hdmi" }, + { /* Sentinel */ }, +}; +MODULE_DEVICE_TABLE(of, rcar_dw_hdmi_of_table); + +static struct platform_driver rcar_dw_hdmi_platform_driver = { + .probe = rcar_dw_hdmi_probe, + .remove = rcar_dw_hdmi_remove, + .driver = { + .name = "rcar-dw-hdmi", + .of_match_table = rcar_dw_hdmi_of_table, + }, +}; + +module_platform_driver(rcar_dw_hdmi_platform_driver); + +MODULE_AUTHOR("Laurent Pinchart <laurent.pinchart@ideasonboard.com>"); +MODULE_DESCRIPTION("Renesas R-Car Gen3 HDMI Encoder Driver"); +MODULE_LICENSE("GPL"); diff --git a/drivers/gpu/drm/ttm/ttm_bo.c b/drivers/gpu/drm/ttm/ttm_bo.c index 412240a3ba90..e44626a2e698 100644 --- a/drivers/gpu/drm/ttm/ttm_bo.c +++ b/drivers/gpu/drm/ttm/ttm_bo.c @@ -1020,37 +1020,44 @@ out_unlock: return ret; } -bool ttm_bo_mem_compat(struct ttm_placement *placement, - struct ttm_mem_reg *mem, - uint32_t *new_flags) +static bool ttm_bo_places_compat(const struct ttm_place *places, + unsigned num_placement, + struct ttm_mem_reg *mem, + uint32_t *new_flags) { - int i; - - for (i = 0; i < placement->num_placement; i++) { - const struct ttm_place *heap = &placement->placement[i]; - if (mem->mm_node && - (mem->start < heap->fpfn || - (heap->lpfn != 0 && (mem->start + mem->num_pages) > heap->lpfn))) - continue; + unsigned i; - *new_flags = heap->flags; - if ((*new_flags & mem->placement & TTM_PL_MASK_CACHING) && - (*new_flags & mem->placement & TTM_PL_MASK_MEM)) - return true; - } + for (i = 0; i < num_placement; i++) { + const struct ttm_place *heap = &places[i]; - for (i = 0; i < placement->num_busy_placement; i++) { - const struct ttm_place *heap = &placement->busy_placement[i]; - if (mem->mm_node && - (mem->start < heap->fpfn || + if (mem->mm_node && (mem->start < heap->fpfn || (heap->lpfn != 0 && (mem->start + mem->num_pages) > heap->lpfn))) continue; *new_flags = heap->flags; if ((*new_flags & mem->placement & TTM_PL_MASK_CACHING) && - (*new_flags & mem->placement & TTM_PL_MASK_MEM)) + (*new_flags & mem->placement & TTM_PL_MASK_MEM) && + (!(*new_flags & TTM_PL_FLAG_CONTIGUOUS) || + (mem->placement & TTM_PL_FLAG_CONTIGUOUS))) return true; } + return false; +} + +bool ttm_bo_mem_compat(struct ttm_placement *placement, + struct ttm_mem_reg *mem, + uint32_t *new_flags) +{ + if (ttm_bo_places_compat(placement->placement, placement->num_placement, + mem, new_flags)) + return true; + + if ((placement->busy_placement != placement->placement || + placement->num_busy_placement > placement->num_placement) && + ttm_bo_places_compat(placement->busy_placement, + placement->num_busy_placement, + mem, new_flags)) + return true; return false; } diff --git a/drivers/gpu/drm/ttm/ttm_bo_vm.c b/drivers/gpu/drm/ttm/ttm_bo_vm.c index 35ffb3754feb..9f53df95f35c 100644 --- a/drivers/gpu/drm/ttm/ttm_bo_vm.c +++ b/drivers/gpu/drm/ttm/ttm_bo_vm.c @@ -231,7 +231,7 @@ static int ttm_bo_vm_fault(struct vm_fault *vmf) */ for (i = 0; i < TTM_BO_VM_NUM_PREFAULT; ++i) { if (bo->mem.bus.is_iomem) - pfn = ((bo->mem.bus.base + bo->mem.bus.offset) >> PAGE_SHIFT) + page_offset; + pfn = bdev->driver->io_mem_pfn(bo, page_offset); else { page = ttm->pages[page_offset]; if (unlikely(!page && i == 0)) { @@ -324,6 +324,14 @@ static struct ttm_buffer_object *ttm_bo_vm_lookup(struct ttm_bo_device *bdev, return bo; } +unsigned long ttm_bo_default_io_mem_pfn(struct ttm_buffer_object *bo, + unsigned long page_offset) +{ + return ((bo->mem.bus.base + bo->mem.bus.offset) >> PAGE_SHIFT) + + page_offset; +} +EXPORT_SYMBOL(ttm_bo_default_io_mem_pfn); + int ttm_bo_mmap(struct file *filp, struct vm_area_struct *vma, struct ttm_bo_device *bdev) { diff --git a/drivers/gpu/drm/ttm/ttm_object.c b/drivers/gpu/drm/ttm/ttm_object.c index fdb451e3ec01..26a7ad0f4789 100644 --- a/drivers/gpu/drm/ttm/ttm_object.c +++ b/drivers/gpu/drm/ttm/ttm_object.c @@ -179,7 +179,7 @@ int ttm_base_object_init(struct ttm_object_file *tfile, if (unlikely(ret != 0)) goto out_err0; - ret = ttm_ref_object_add(tfile, base, TTM_REF_USAGE, NULL); + ret = ttm_ref_object_add(tfile, base, TTM_REF_USAGE, NULL, false); if (unlikely(ret != 0)) goto out_err1; @@ -318,7 +318,8 @@ EXPORT_SYMBOL(ttm_ref_object_exists); int ttm_ref_object_add(struct ttm_object_file *tfile, struct ttm_base_object *base, - enum ttm_ref_type ref_type, bool *existed) + enum ttm_ref_type ref_type, bool *existed, + bool require_existed) { struct drm_open_hash *ht = &tfile->ref_hash[ref_type]; struct ttm_ref_object *ref; @@ -345,6 +346,9 @@ int ttm_ref_object_add(struct ttm_object_file *tfile, } rcu_read_unlock(); + if (require_existed) + return -EPERM; + ret = ttm_mem_global_alloc(mem_glob, sizeof(*ref), false, false); if (unlikely(ret != 0)) @@ -449,10 +453,10 @@ void ttm_object_file_release(struct ttm_object_file **p_tfile) ttm_ref_object_release(&ref->kref); } + spin_unlock(&tfile->lock); for (i = 0; i < TTM_REF_NUM; ++i) drm_ht_remove(&tfile->ref_hash[i]); - spin_unlock(&tfile->lock); ttm_object_file_unref(&tfile); } EXPORT_SYMBOL(ttm_object_file_release); @@ -529,9 +533,7 @@ void ttm_object_device_release(struct ttm_object_device **p_tdev) *p_tdev = NULL; - spin_lock(&tdev->object_lock); drm_ht_remove(&tdev->object_hash); - spin_unlock(&tdev->object_lock); kfree(tdev); } @@ -635,7 +637,7 @@ int ttm_prime_fd_to_handle(struct ttm_object_file *tfile, prime = (struct ttm_prime_object *) dma_buf->priv; base = &prime->base; *handle = base->hash.key; - ret = ttm_ref_object_add(tfile, base, TTM_REF_USAGE, NULL); + ret = ttm_ref_object_add(tfile, base, TTM_REF_USAGE, NULL, false); dma_buf_put(dma_buf); diff --git a/drivers/gpu/drm/virtio/virtgpu_ttm.c b/drivers/gpu/drm/virtio/virtgpu_ttm.c index 70ec8ca8d9b1..4e8e27d50922 100644 --- a/drivers/gpu/drm/virtio/virtgpu_ttm.c +++ b/drivers/gpu/drm/virtio/virtgpu_ttm.c @@ -431,6 +431,7 @@ static struct ttm_bo_driver virtio_gpu_bo_driver = { .verify_access = &virtio_gpu_verify_access, .io_mem_reserve = &virtio_gpu_ttm_io_mem_reserve, .io_mem_free = &virtio_gpu_ttm_io_mem_free, + .io_mem_pfn = ttm_bo_default_io_mem_pfn, .move_notify = &virtio_gpu_bo_move_notify, .swap_notify = &virtio_gpu_bo_swap_notify, }; diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_buffer.c b/drivers/gpu/drm/vmwgfx/vmwgfx_buffer.c index 4c7f24a67a2e..35bf781e418e 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_buffer.c +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_buffer.c @@ -859,4 +859,5 @@ struct ttm_bo_driver vmw_bo_driver = { .fault_reserve_notify = &vmw_ttm_fault_reserve_notify, .io_mem_reserve = &vmw_ttm_io_mem_reserve, .io_mem_free = &vmw_ttm_io_mem_free, + .io_mem_pfn = ttm_bo_default_io_mem_pfn, }; diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_fence.c b/drivers/gpu/drm/vmwgfx/vmwgfx_fence.c index b399f03a988d..6b2708b4eafe 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_fence.c +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_fence.c @@ -538,7 +538,7 @@ int vmw_fence_create(struct vmw_fence_manager *fman, struct vmw_fence_obj **p_fence) { struct vmw_fence_obj *fence; - int ret; + int ret; fence = kzalloc(sizeof(*fence), GFP_KERNEL); if (unlikely(fence == NULL)) @@ -701,6 +701,41 @@ void vmw_fence_fifo_up(struct vmw_fence_manager *fman) } +/** + * vmw_fence_obj_lookup - Look up a user-space fence object + * + * @tfile: A struct ttm_object_file identifying the caller. + * @handle: A handle identifying the fence object. + * @return: A struct vmw_user_fence base ttm object on success or + * an error pointer on failure. + * + * The fence object is looked up and type-checked. The caller needs + * to have opened the fence object first, but since that happens on + * creation and fence objects aren't shareable, that's not an + * issue currently. + */ +static struct ttm_base_object * +vmw_fence_obj_lookup(struct ttm_object_file *tfile, u32 handle) +{ + struct ttm_base_object *base = ttm_base_object_lookup(tfile, handle); + + if (!base) { + pr_err("Invalid fence object handle 0x%08lx.\n", + (unsigned long)handle); + return ERR_PTR(-EINVAL); + } + + if (base->refcount_release != vmw_user_fence_base_release) { + pr_err("Invalid fence object handle 0x%08lx.\n", + (unsigned long)handle); + ttm_base_object_unref(&base); + return ERR_PTR(-EINVAL); + } + + return base; +} + + int vmw_fence_obj_wait_ioctl(struct drm_device *dev, void *data, struct drm_file *file_priv) { @@ -726,12 +761,9 @@ int vmw_fence_obj_wait_ioctl(struct drm_device *dev, void *data, arg->kernel_cookie = jiffies + wait_timeout; } - base = ttm_base_object_lookup(tfile, arg->handle); - if (unlikely(base == NULL)) { - pr_err("Wait invalid fence object handle 0x%08lx\n", - (unsigned long)arg->handle); - return -EINVAL; - } + base = vmw_fence_obj_lookup(tfile, arg->handle); + if (IS_ERR(base)) + return PTR_ERR(base); fence = &(container_of(base, struct vmw_user_fence, base)->fence); @@ -770,12 +802,9 @@ int vmw_fence_obj_signaled_ioctl(struct drm_device *dev, void *data, struct ttm_object_file *tfile = vmw_fpriv(file_priv)->tfile; struct vmw_private *dev_priv = vmw_priv(dev); - base = ttm_base_object_lookup(tfile, arg->handle); - if (unlikely(base == NULL)) { - pr_err("Fence signaled invalid fence object handle 0x%08lx\n", - (unsigned long)arg->handle); - return -EINVAL; - } + base = vmw_fence_obj_lookup(tfile, arg->handle); + if (IS_ERR(base)) + return PTR_ERR(base); fence = &(container_of(base, struct vmw_user_fence, base)->fence); fman = fman_from_fence(fence); @@ -1022,6 +1051,7 @@ int vmw_fence_event_ioctl(struct drm_device *dev, void *data, (struct drm_vmw_fence_event_arg *) data; struct vmw_fence_obj *fence = NULL; struct vmw_fpriv *vmw_fp = vmw_fpriv(file_priv); + struct ttm_object_file *tfile = vmw_fp->tfile; struct drm_vmw_fence_rep __user *user_fence_rep = (struct drm_vmw_fence_rep __user *)(unsigned long) arg->fence_rep; @@ -1035,24 +1065,18 @@ int vmw_fence_event_ioctl(struct drm_device *dev, void *data, */ if (arg->handle) { struct ttm_base_object *base = - ttm_base_object_lookup_for_ref(dev_priv->tdev, - arg->handle); - - if (unlikely(base == NULL)) { - DRM_ERROR("Fence event invalid fence object handle " - "0x%08lx.\n", - (unsigned long)arg->handle); - return -EINVAL; - } + vmw_fence_obj_lookup(tfile, arg->handle); + + if (IS_ERR(base)) + return PTR_ERR(base); + fence = &(container_of(base, struct vmw_user_fence, base)->fence); (void) vmw_fence_obj_reference(fence); if (user_fence_rep != NULL) { - bool existed; - ret = ttm_ref_object_add(vmw_fp->tfile, base, - TTM_REF_USAGE, &existed); + TTM_REF_USAGE, NULL, false); if (unlikely(ret != 0)) { DRM_ERROR("Failed to reference a fence " "object.\n"); @@ -1095,8 +1119,7 @@ int vmw_fence_event_ioctl(struct drm_device *dev, void *data, return 0; out_no_create: if (user_fence_rep != NULL) - ttm_ref_object_base_unref(vmw_fpriv(file_priv)->tfile, - handle, TTM_REF_USAGE); + ttm_ref_object_base_unref(tfile, handle, TTM_REF_USAGE); out_no_ref_obj: vmw_fence_obj_unreference(&fence); return ret; diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_ioctl.c b/drivers/gpu/drm/vmwgfx/vmwgfx_ioctl.c index b8c6a03c8c54..5ec24fd801cd 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_ioctl.c +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_ioctl.c @@ -114,8 +114,6 @@ int vmw_getparam_ioctl(struct drm_device *dev, void *data, param->value = dev_priv->has_dx; break; default: - DRM_ERROR("Illegal vmwgfx get param request: %d\n", - param->param); return -EINVAL; } @@ -186,7 +184,7 @@ int vmw_get_cap_3d_ioctl(struct drm_device *dev, void *data, bool gb_objects = !!(dev_priv->capabilities & SVGA_CAP_GBOBJECTS); struct vmw_fpriv *vmw_fp = vmw_fpriv(file_priv); - if (unlikely(arg->pad64 != 0)) { + if (unlikely(arg->pad64 != 0 || arg->max_size == 0)) { DRM_ERROR("Illegal GET_3D_CAP argument.\n"); return -EINVAL; } diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_resource.c b/drivers/gpu/drm/vmwgfx/vmwgfx_resource.c index fa1037ec8e5f..7d591f653dfa 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_resource.c +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_resource.c @@ -546,7 +546,7 @@ static int vmw_user_dmabuf_synccpu_grab(struct vmw_user_dma_buffer *user_bo, return ret; ret = ttm_ref_object_add(tfile, &user_bo->prime.base, - TTM_REF_SYNCCPU_WRITE, &existed); + TTM_REF_SYNCCPU_WRITE, &existed, false); if (ret != 0 || existed) ttm_bo_synccpu_write_release(&user_bo->dma.base); @@ -730,7 +730,7 @@ int vmw_user_dmabuf_reference(struct ttm_object_file *tfile, *handle = user_bo->prime.base.hash.key; return ttm_ref_object_add(tfile, &user_bo->prime.base, - TTM_REF_USAGE, NULL); + TTM_REF_USAGE, NULL, false); } /** diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_surface.c b/drivers/gpu/drm/vmwgfx/vmwgfx_surface.c index 41b9d20d6ae7..7681341fe32b 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_surface.c +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_surface.c @@ -713,11 +713,14 @@ int vmw_surface_define_ioctl(struct drm_device *dev, void *data, 128; num_sizes = 0; - for (i = 0; i < DRM_VMW_MAX_SURFACE_FACES; ++i) + for (i = 0; i < DRM_VMW_MAX_SURFACE_FACES; ++i) { + if (req->mip_levels[i] > DRM_VMW_MAX_MIP_LEVELS) + return -EINVAL; num_sizes += req->mip_levels[i]; + } - if (num_sizes > DRM_VMW_MAX_SURFACE_FACES * - DRM_VMW_MAX_MIP_LEVELS) + if (num_sizes > DRM_VMW_MAX_SURFACE_FACES * DRM_VMW_MAX_MIP_LEVELS || + num_sizes == 0) return -EINVAL; size = vmw_user_surface_size + 128 + @@ -890,17 +893,16 @@ vmw_surface_handle_reference(struct vmw_private *dev_priv, uint32_t handle; struct ttm_base_object *base; int ret; + bool require_exist = false; if (handle_type == DRM_VMW_HANDLE_PRIME) { ret = ttm_prime_fd_to_handle(tfile, u_handle, &handle); if (unlikely(ret != 0)) return ret; } else { - if (unlikely(drm_is_render_client(file_priv))) { - DRM_ERROR("Render client refused legacy " - "surface reference.\n"); - return -EACCES; - } + if (unlikely(drm_is_render_client(file_priv))) + require_exist = true; + if (ACCESS_ONCE(vmw_fpriv(file_priv)->locked_master)) { DRM_ERROR("Locked master refused legacy " "surface reference.\n"); @@ -928,17 +930,14 @@ vmw_surface_handle_reference(struct vmw_private *dev_priv, /* * Make sure the surface creator has the same - * authenticating master. + * authenticating master, or is already registered with us. */ if (drm_is_primary_client(file_priv) && - user_srf->master != file_priv->master) { - DRM_ERROR("Trying to reference surface outside of" - " master domain.\n"); - ret = -EACCES; - goto out_bad_resource; - } + user_srf->master != file_priv->master) + require_exist = true; - ret = ttm_ref_object_add(tfile, base, TTM_REF_USAGE, NULL); + ret = ttm_ref_object_add(tfile, base, TTM_REF_USAGE, NULL, + require_exist); if (unlikely(ret != 0)) { DRM_ERROR("Could not add a reference to a surface.\n"); goto out_bad_resource; diff --git a/drivers/gpu/ipu-v3/Makefile b/drivers/gpu/ipu-v3/Makefile index 1ab9bceee755..8cdf9e4ae772 100644 --- a/drivers/gpu/ipu-v3/Makefile +++ b/drivers/gpu/ipu-v3/Makefile @@ -2,4 +2,8 @@ obj-$(CONFIG_IMX_IPUV3_CORE) += imx-ipu-v3.o imx-ipu-v3-objs := ipu-common.o ipu-cpmem.o ipu-csi.o ipu-dc.o ipu-di.o \ ipu-dp.o ipu-dmfc.o ipu-ic.o ipu-image-convert.o \ - ipu-pre.o ipu-prg.o ipu-smfc.o ipu-vdi.o + ipu-smfc.o ipu-vdi.o + +ifdef CONFIG_DRM + imx-ipu-v3-objs += ipu-pre.o ipu-prg.o +endif diff --git a/drivers/gpu/ipu-v3/ipu-common.c b/drivers/gpu/ipu-v3/ipu-common.c index 7aefccec31b1..16d556816b5f 100644 --- a/drivers/gpu/ipu-v3/ipu-common.c +++ b/drivers/gpu/ipu-v3/ipu-common.c @@ -1401,7 +1401,8 @@ static int ipu_probe(struct platform_device *pdev) ipu->id = of_alias_get_id(np, "ipu"); - if (of_device_is_compatible(np, "fsl,imx6qp-ipu")) { + if (of_device_is_compatible(np, "fsl,imx6qp-ipu") && + IS_ENABLED(CONFIG_DRM)) { ipu->prg_priv = ipu_prg_lookup_by_phandle(&pdev->dev, "fsl,prg", ipu->id); if (!ipu->prg_priv) @@ -1538,8 +1539,10 @@ static struct platform_driver imx_ipu_driver = { }; static struct platform_driver * const drivers[] = { +#if IS_ENABLED(CONFIG_DRM) &ipu_pre_drv, &ipu_prg_drv, +#endif &imx_ipu_driver, }; |