Merge tag 'amd-drm-fixes-5.13-2021-05-05' of https://gitlab.freedesktop.org/agd5f/linux into drm-next

amd-drm-fixes-5.13-2021-05-05:

amdgpu:
- MPO hang workaround
- Fix for concurrent VM flushes on vega/navi
- dcefclk is not adjustable on navi1x and newer
- MST HPD debugfs fix
- Suspend/resumes fixes
- Register VGA clients late in case driver fails to load
- Fix GEM leak in user framebuffer create
- Add support for polaris12 with 32 bit memory interface
- Fix duplicate cursor issue when using overlay
- Fix corruption with tiled surfaces on VCN3
- Add BO size and stride check to fix BO size verification

radeon:
- Fix off-by-one in power state parsing
- Fix possible memory leak in power state parsing

Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Alex Deucher <alexander.deucher@amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20210506033929.3875-1-alexander.deucher@amd.com

Author: airlied

drivers/gpu/drm/amd/amdgpu/amdgpu_device.c

@@ -3410,19 +3410,6 @@ int amdgpu_device_init(struct amdgpu_device *adev,
 	/* doorbell bar mapping and doorbell index init*/
 	amdgpu_device_doorbell_init(adev);
 
-	/* if we have > 1 VGA cards, then disable the amdgpu VGA resources */
-	/* this will fail for cards that aren't VGA class devices, just
-	 * ignore it */
-	if ((adev->pdev->class >> 8) == PCI_CLASS_DISPLAY_VGA)
-		vga_client_register(adev->pdev, adev, NULL, amdgpu_device_vga_set_decode);
-
-	if (amdgpu_device_supports_px(ddev)) {
-		px = true;
-		vga_switcheroo_register_client(adev->pdev,
-					       &amdgpu_switcheroo_ops, px);
-		vga_switcheroo_init_domain_pm_ops(adev->dev, &adev->vga_pm_domain);
-	}
-
 	if (amdgpu_emu_mode == 1) {
 		/* post the asic on emulation mode */
 		emu_soc_asic_init(adev);
@@ -3619,6 +3606,19 @@ int amdgpu_device_init(struct amdgpu_device *adev,
 	if (amdgpu_device_cache_pci_state(adev->pdev))
 		pci_restore_state(pdev);
 
+	/* if we have > 1 VGA cards, then disable the amdgpu VGA resources */
+	/* this will fail for cards that aren't VGA class devices, just
+	 * ignore it */
+	if ((adev->pdev->class >> 8) == PCI_CLASS_DISPLAY_VGA)
+		vga_client_register(adev->pdev, adev, NULL, amdgpu_device_vga_set_decode);
+
+	if (amdgpu_device_supports_px(ddev)) {
+		px = true;
+		vga_switcheroo_register_client(adev->pdev,
+					       &amdgpu_switcheroo_ops, px);
+		vga_switcheroo_init_domain_pm_ops(adev->dev, &adev->vga_pm_domain);
+	}
+
 	if (adev->gmc.xgmi.pending_reset)
 		queue_delayed_work(system_wq, &mgpu_info.delayed_reset_work,
 				   msecs_to_jiffies(AMDGPU_RESUME_MS));
@@ -3630,8 +3630,6 @@ int amdgpu_device_init(struct amdgpu_device *adev,
 
 failed:
 	amdgpu_vf_error_trans_all(adev);
-	if (px)
-		vga_switcheroo_fini_domain_pm_ops(adev->dev);
 
 failed_unmap:
 	iounmap(adev->rmmio);

drivers/gpu/drm/amd/amdgpu/amdgpu_display.c

@@ -837,6 +837,174 @@ static int convert_tiling_flags_to_modifier(struct amdgpu_framebuffer *afb)
 	return 0;
 }
 
+static void get_block_dimensions(unsigned int block_log2, unsigned int cpp,
+				 unsigned int *width, unsigned int *height)
+{
+	unsigned int cpp_log2 = ilog2(cpp);
+	unsigned int pixel_log2 = block_log2 - cpp_log2;
+	unsigned int width_log2 = (pixel_log2 + 1) / 2;
+	unsigned int height_log2 = pixel_log2 - width_log2;
+
+	*width = 1 << width_log2;
+	*height = 1 << height_log2;
+}
+
+static unsigned int get_dcc_block_size(uint64_t modifier, bool rb_aligned,
+				       bool pipe_aligned)
+{
+	unsigned int ver = AMD_FMT_MOD_GET(TILE_VERSION, modifier);
+
+	switch (ver) {
+	case AMD_FMT_MOD_TILE_VER_GFX9: {
+		/*
+		 * TODO: for pipe aligned we may need to check the alignment of the
+		 * total size of the surface, which may need to be bigger than the
+		 * natural alignment due to some HW workarounds
+		 */
+		return max(10 + (rb_aligned ? (int)AMD_FMT_MOD_GET(RB, modifier) : 0), 12);
+	}
+	case AMD_FMT_MOD_TILE_VER_GFX10:
+	case AMD_FMT_MOD_TILE_VER_GFX10_RBPLUS: {
+		int pipes_log2 = AMD_FMT_MOD_GET(PIPE_XOR_BITS, modifier);
+
+		if (ver == AMD_FMT_MOD_TILE_VER_GFX10_RBPLUS && pipes_log2 > 1 &&
+		    AMD_FMT_MOD_GET(PACKERS, modifier) == pipes_log2)
+			++pipes_log2;
+
+		return max(8 + (pipe_aligned ? pipes_log2 : 0), 12);
+	}
+	default:
+		return 0;
+	}
+}
+
+static int amdgpu_display_verify_plane(struct amdgpu_framebuffer *rfb, int plane,
+				       const struct drm_format_info *format,
+				       unsigned int block_width, unsigned int block_height,
+				       unsigned int block_size_log2)
+{
+	unsigned int width = rfb->base.width /
+		((plane && plane < format->num_planes) ? format->hsub : 1);
+	unsigned int height = rfb->base.height /
+		((plane && plane < format->num_planes) ? format->vsub : 1);
+	unsigned int cpp = plane < format->num_planes ? format->cpp[plane] : 1;
+	unsigned int block_pitch = block_width * cpp;
+	unsigned int min_pitch = ALIGN(width * cpp, block_pitch);
+	unsigned int block_size = 1 << block_size_log2;
+	uint64_t size;
+
+	if (rfb->base.pitches[plane] % block_pitch) {
+		drm_dbg_kms(rfb->base.dev,
+			    "pitch %d for plane %d is not a multiple of block pitch %d\n",
+			    rfb->base.pitches[plane], plane, block_pitch);
+		return -EINVAL;
+	}
+	if (rfb->base.pitches[plane] < min_pitch) {
+		drm_dbg_kms(rfb->base.dev,
+			    "pitch %d for plane %d is less than minimum pitch %d\n",
+			    rfb->base.pitches[plane], plane, min_pitch);
+		return -EINVAL;
+	}
+
+	/* Force at least natural alignment. */
+	if (rfb->base.offsets[plane] % block_size) {
+		drm_dbg_kms(rfb->base.dev,
+			    "offset 0x%x for plane %d is not a multiple of block pitch 0x%x\n",
+			    rfb->base.offsets[plane], plane, block_size);
+		return -EINVAL;
+	}
+
+	size = rfb->base.offsets[plane] +
+		(uint64_t)rfb->base.pitches[plane] / block_pitch *
+		block_size * DIV_ROUND_UP(height, block_height);
+
+	if (rfb->base.obj[0]->size < size) {
+		drm_dbg_kms(rfb->base.dev,
+			    "BO size 0x%zx is less than 0x%llx required for plane %d\n",
+			    rfb->base.obj[0]->size, size, plane);
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+
+static int amdgpu_display_verify_sizes(struct amdgpu_framebuffer *rfb)
+{
+	const struct drm_format_info *format_info = drm_format_info(rfb->base.format->format);
+	uint64_t modifier = rfb->base.modifier;
+	int ret;
+	unsigned int i, block_width, block_height, block_size_log2;
+
+	if (!rfb->base.dev->mode_config.allow_fb_modifiers)
+		return 0;
+
+	for (i = 0; i < format_info->num_planes; ++i) {
+		if (modifier == DRM_FORMAT_MOD_LINEAR) {
+			block_width = 256 / format_info->cpp[i];
+			block_height = 1;
+			block_size_log2 = 8;
+		} else {
+			int swizzle = AMD_FMT_MOD_GET(TILE, modifier);
+
+			switch ((swizzle & ~3) + 1) {
+			case DC_SW_256B_S:
+				block_size_log2 = 8;
+				break;
+			case DC_SW_4KB_S:
+			case DC_SW_4KB_S_X:
+				block_size_log2 = 12;
+				break;
+			case DC_SW_64KB_S:
+			case DC_SW_64KB_S_T:
+			case DC_SW_64KB_S_X:
+				block_size_log2 = 16;
+				break;
+			default:
+				drm_dbg_kms(rfb->base.dev,
+					    "Swizzle mode with unknown block size: %d\n", swizzle);
+				return -EINVAL;
+			}
+
+			get_block_dimensions(block_size_log2, format_info->cpp[i],
+					     &block_width, &block_height);
+		}
+
+		ret = amdgpu_display_verify_plane(rfb, i, format_info,
+						  block_width, block_height, block_size_log2);
+		if (ret)
+			return ret;
+	}
+
+	if (AMD_FMT_MOD_GET(DCC, modifier)) {
+		if (AMD_FMT_MOD_GET(DCC_RETILE, modifier)) {
+			block_size_log2 = get_dcc_block_size(modifier, false, false);
+			get_block_dimensions(block_size_log2 + 8, format_info->cpp[0],
+					     &block_width, &block_height);
+			ret = amdgpu_display_verify_plane(rfb, i, format_info,
+							  block_width, block_height,
+							  block_size_log2);
+			if (ret)
+				return ret;
+
+			++i;
+			block_size_log2 = get_dcc_block_size(modifier, true, true);
+		} else {
+			bool pipe_aligned = AMD_FMT_MOD_GET(DCC_PIPE_ALIGN, modifier);
+
+			block_size_log2 = get_dcc_block_size(modifier, true, pipe_aligned);
+		}
+		get_block_dimensions(block_size_log2 + 8, format_info->cpp[0],
+				     &block_width, &block_height);
+		ret = amdgpu_display_verify_plane(rfb, i, format_info,
+						  block_width, block_height, block_size_log2);
+		if (ret)
+			return ret;
+	}
+
+	return 0;
+}
+
 static int amdgpu_display_get_fb_info(const struct amdgpu_framebuffer *amdgpu_fb,
 				      uint64_t *tiling_flags, bool *tmz_surface)
 {
@@ -902,10 +1070,8 @@ int amdgpu_display_gem_fb_verify_and_init(
 	int ret;
 
 	rfb->base.obj[0] = obj;
-
-	/* Verify that bo size can fit the fb size. */
-	ret = drm_gem_fb_init_with_funcs(dev, &rfb->base, file_priv, mode_cmd,
-					 &amdgpu_fb_funcs);
+	drm_helper_mode_fill_fb_struct(dev, &rfb->base, mode_cmd);
+	ret = drm_framebuffer_init(dev, &rfb->base, &amdgpu_fb_funcs);
 	if (ret)
 		goto err;
 	/* Verify that the modifier is supported. */
@@ -967,9 +1133,12 @@ int amdgpu_display_framebuffer_init(struct drm_device *dev,
 		}
 	}
 
-	for (i = 1; i < rfb->base.format->num_planes; ++i) {
+	ret = amdgpu_display_verify_sizes(rfb);
+	if (ret)
+		return ret;
+
+	for (i = 0; i < rfb->base.format->num_planes; ++i) {
 		drm_gem_object_get(rfb->base.obj[0]);
-		drm_gem_object_put(rfb->base.obj[i]);
 		rfb->base.obj[i] = rfb->base.obj[0];
 	}
 
@@ -999,6 +1168,7 @@ amdgpu_display_user_framebuffer_create(struct drm_device *dev,
 	domains = amdgpu_display_supported_domains(drm_to_adev(dev), bo->flags);
 	if (obj->import_attach && !(domains & AMDGPU_GEM_DOMAIN_GTT)) {
 		drm_dbg_kms(dev, "Cannot create framebuffer from imported dma_buf\n");
+		drm_gem_object_put(obj);
 		return ERR_PTR(-EINVAL);
 	}
 
@@ -1412,7 +1582,7 @@ int amdgpu_display_suspend_helper(struct amdgpu_device *adev)
 			}
 		}
 	}
-	return r;
+	return 0;
 }
 
 int amdgpu_display_resume_helper(struct amdgpu_device *adev)

drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c

@@ -1573,6 +1573,9 @@ static int amdgpu_pmops_runtime_resume(struct device *dev)
 		amdgpu_device_baco_exit(drm_dev);
 	}
 	ret = amdgpu_device_resume(drm_dev, false);
+	if (ret)
+		return ret;
+
 	if (amdgpu_device_supports_px(drm_dev))
 		drm_dev->switch_power_state = DRM_SWITCH_POWER_ON;
 	adev->in_runpm = false;

drivers/gpu/drm/amd/amdgpu/amdgpu_ids.c

@@ -215,7 +215,11 @@ static int amdgpu_vmid_grab_idle(struct amdgpu_vm *vm,
 	/* Check if we have an idle VMID */
 	i = 0;
 	list_for_each_entry((*idle), &id_mgr->ids_lru, list) {
-		fences[i] = amdgpu_sync_peek_fence(&(*idle)->active, ring);
+		/* Don't use per engine and per process VMID at the same time */
+		struct amdgpu_ring *r = adev->vm_manager.concurrent_flush ?
+			NULL : ring;
+
+		fences[i] = amdgpu_sync_peek_fence(&(*idle)->active, r);
 		if (!fences[i])
 			break;
 		++i;
@@ -281,7 +285,7 @@ static int amdgpu_vmid_grab_reserved(struct amdgpu_vm *vm,
 	if (updates && (*id)->flushed_updates &&
 	    updates->context == (*id)->flushed_updates->context &&
 	    !dma_fence_is_later(updates, (*id)->flushed_updates))
-	    updates = NULL;
+		updates = NULL;
 
 	if ((*id)->owner != vm->immediate.fence_context ||
 	    job->vm_pd_addr != (*id)->pd_gpu_addr ||
@@ -290,6 +294,10 @@ static int amdgpu_vmid_grab_reserved(struct amdgpu_vm *vm,
 	     !dma_fence_is_signaled((*id)->last_flush))) {
 		struct dma_fence *tmp;
 
+		/* Don't use per engine and per process VMID at the same time */
+		if (adev->vm_manager.concurrent_flush)
+			ring = NULL;
+
 		/* to prevent one context starved by another context */
 		(*id)->pd_gpu_addr = 0;
 		tmp = amdgpu_sync_peek_fence(&(*id)->active, ring);
@@ -365,12 +373,7 @@ static int amdgpu_vmid_grab_used(struct amdgpu_vm *vm,
 		if (updates && (!flushed || dma_fence_is_later(updates, flushed)))
 			needs_flush = true;
 
-		/* Concurrent flushes are only possible starting with Vega10 and
-		 * are broken on Navi10 and Navi14.
-		 */
-		if (needs_flush && (adev->asic_type < CHIP_VEGA10 ||
-				    adev->asic_type == CHIP_NAVI10 ||
-				    adev->asic_type == CHIP_NAVI14))
+		if (needs_flush && !adev->vm_manager.concurrent_flush)
 			continue;
 
 		/* Good, we can use this VMID. Remember this submission as

drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c

@@ -3148,6 +3148,12 @@ void amdgpu_vm_manager_init(struct amdgpu_device *adev)
 {
 	unsigned i;
 
+	/* Concurrent flushes are only possible starting with Vega10 and
+	 * are broken on Navi10 and Navi14.
+	 */
+	adev->vm_manager.concurrent_flush = !(adev->asic_type < CHIP_VEGA10 ||
+					      adev->asic_type == CHIP_NAVI10 ||
+					      adev->asic_type == CHIP_NAVI14);
 	amdgpu_vmid_mgr_init(adev);
 
 	adev->vm_manager.fence_context =

drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h

@@ -331,6 +331,7 @@ struct amdgpu_vm_manager {
 	/* Handling of VMIDs */
 	struct amdgpu_vmid_mgr			id_mgr[AMDGPU_MAX_VMHUBS];
 	unsigned int				first_kfd_vmid;
+	bool					concurrent_flush;
 
 	/* Handling of VM fences */
 	u64					fence_context;

drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c

@@ -59,6 +59,7 @@ MODULE_FIRMWARE("amdgpu/tonga_mc.bin");
 MODULE_FIRMWARE("amdgpu/polaris11_mc.bin");
 MODULE_FIRMWARE("amdgpu/polaris10_mc.bin");
 MODULE_FIRMWARE("amdgpu/polaris12_mc.bin");
+MODULE_FIRMWARE("amdgpu/polaris12_32_mc.bin");
 MODULE_FIRMWARE("amdgpu/polaris11_k_mc.bin");
 MODULE_FIRMWARE("amdgpu/polaris10_k_mc.bin");
 MODULE_FIRMWARE("amdgpu/polaris12_k_mc.bin");
@@ -243,10 +244,16 @@ static int gmc_v8_0_init_microcode(struct amdgpu_device *adev)
 			chip_name = "polaris10";
 		break;
 	case CHIP_POLARIS12:
-		if (ASICID_IS_P23(adev->pdev->device, adev->pdev->revision))
+		if (ASICID_IS_P23(adev->pdev->device, adev->pdev->revision)) {
 			chip_name = "polaris12_k";
-		else
-			chip_name = "polaris12";
+		} else {
+			WREG32(mmMC_SEQ_IO_DEBUG_INDEX, ixMC_IO_DEBUG_UP_159);
+			/* Polaris12 32bit ASIC needs a special MC firmware */
+			if (RREG32(mmMC_SEQ_IO_DEBUG_DATA) == 0x05b4dc40)
+				chip_name = "polaris12_32";
+			else
+				chip_name = "polaris12";
+		}
 		break;
 	case CHIP_FIJI:
 	case CHIP_CARRIZO:

drivers/gpu/drm/amd/amdgpu/vcn_v3_0.c

@@ -589,6 +589,10 @@ static void vcn_v3_0_mc_resume_dpg_mode(struct amdgpu_device *adev, int inst_idx
 	WREG32_SOC15_DPG_MODE(inst_idx, SOC15_DPG_MODE_OFFSET(
 			VCN, inst_idx, mmUVD_VCPU_NONCACHE_SIZE0),
 			AMDGPU_GPU_PAGE_ALIGN(sizeof(struct amdgpu_fw_shared)), 0, indirect);
+
+	/* VCN global tiling registers */
+	WREG32_SOC15_DPG_MODE(0, SOC15_DPG_MODE_OFFSET(
+		UVD, 0, mmUVD_GFX10_ADDR_CONFIG), adev->gfx.config.gb_addr_config, 0, indirect);
 }
 
 static void vcn_v3_0_disable_static_power_gating(struct amdgpu_device *adev, int inst)