Edge 無法使用GPU進行硬體加速

如題，已將edge設定中的硬體加速打開，而且#ignore-gpu-blocklist也設定成enable

但還是無法使用顯卡加速
以下是Edge://GPU的部分內容(完整內容字數太多貼不上來)

Graphics Feature Status

• Canvas: Software only, hardware acceleration unavailable
• Canvas out-of-process rasterization: Disabled
• Direct Rendering Display Compositor: Disabled
• Compositing: Software only. Hardware acceleration disabled
• Multiple Raster Threads: Enabled
• OpenGL: Disabled
• Rasterization: Software only. Hardware acceleration disabled
• Raw Draw: Disabled
• Skia Renderer: Enabled
• Video Decode: Software only. Hardware acceleration disabled
• Video Encode: Software only. Hardware acceleration disabled
• Vulkan: Disabled
• WebGL: Software only, hardware acceleration unavailable
• WebGL2: Software only, hardware acceleration unavailable
• WebGPU: Disabled

Problems Detected

• WebGPU has been disabled via blocklist or the command line.
  Disabled Features: webgpu
• Gpu compositing has been disabled, either via blocklist, about:flags or the command line. The browser will fall back to software compositing and hardware acceleration will be unavailable.
  Disabled Features: gpu_compositing

ANGLE Features

• allowCompressedFormats (Frontend workarounds): Enabled: true
  Allow compressed formats
• disableAnisotropicFiltering (Frontend workarounds): Disabled
  Disable support for anisotropic filtering
• disableProgramBinary (Frontend features) anglebug:5007: Disabled
  Disable support for GL_OES_get_program_binary
• disableProgramCachingForTransformFeedback (Frontend workarounds): Disabled
  On some GPUs, program binaries don't contain transform feedback varyings
• enableCaptureLimits (Frontend features) anglebug:5750: Disabled
  Set the context limits like frame capturing was enabled
• enableCompressingPipelineCacheInThreadPool (Frontend workarounds) anglebug:4722: Disabled: false
  Enable compressing pipeline cache in thread pool.
• enableProgramBinaryForCapture (Frontend features) anglebug:5658: Disabled
  Even if FrameCapture is enabled, enable GL_OES_get_program_binary
• forceDepthAttachmentInitOnClear (Frontend workarounds) anglebug:7246: Disabled
  Force depth attachment initialization on clear ops
• forceInitShaderVariables (Frontend features): Disabled
  Force-enable shader variable initialization
• forceRobustResourceInit (Frontend features) anglebug:6041: Disabled
  Force-enable robust resource init
• loseContextOnOutOfMemory (Frontend workarounds): Enabled: true
  Some users rely on a lost context notification if a GL_OUT_OF_MEMORY error occurs
• scalarizeVecAndMatConstructorArgs (Frontend workarounds) 1165751: Disabled: false
  Always rewrite vec/mat constructors to be consistent
• allocateNonZeroMemory (Vulkan features) anglebug:4384: Disabled: false
  Fill new allocations with non-zero values to flush out errors.
• allowGenerateMipmapWithCompute (Vulkan features) anglebug:4551: Disabled: supportsSubgroupQuadOpsInComputeShader && mSubgroupExtendedTypesFeatures.shaderSubgroupExtendedTypes && maxComputeWorkGroupInvocations >= 256 && ((isAMD && !IsWindows()) || isNvidia || isSamsung)
  Use the compute path to generate mipmaps on devices that meet the minimum requirements, and the performance is better.
• asyncCommandQueue (Vulkan features) anglebug:4324: Disabled: false
  Use CommandQueue worker thread to dispatch work to GPU.
• basicGLLineRasterization (Vulkan features): Disabled
  Enable the use of pixel shader patching to implement OpenGL basic line rasterization rules
• bindEmptyForUnusedDescriptorSets (Vulkan workarounds) anglebug:2727: Disabled: IsAndroid() && isQualcomm
  Gaps in bound descriptor set indices causes the post-gap sets to misbehave
• bottomLeftOriginPresentRegionRectangles (Vulkan workarounds): Disabled: IsAndroid()
  On some platforms present region rectangles are expected to have a bottom-left origin, instead of top-left origin as from spec
• bresenhamLineRasterization (Vulkan features): Enabled: true
  Enable Bresenham line rasterization via VK_EXT_line_rasterization extension
• clampPointSize (Vulkan workarounds) anglebug:2970: Disabled: isNvidia && nvidiaVersion.major < uint32_t(IsWindows() ? 430 : 421)
  The point size range reported from the API is inconsistent with the actual behavior
• compressVertexData (Vulkan workarounds): Disabled
  Compress vertex data to smaller data types when possible. Using this feature makes ANGLE non-conformant.
• createPipelineDuringLink (Vulkan features) anglebug:7046: Disabled
  Create pipeline with default state during glLinkProgram
• deferFlushUntilEndRenderPass (Vulkan workarounds) https://issuetracker.google.com/issues/166475273: Enabled: !isQualcomm
  Allow glFlush to be deferred until renderpass ends
• depthClamping (Vulkan workarounds) anglebug:3970: Disabled: isNvidia && mPhysicalDeviceFeatures.depthClamp && ExtensionFound("VK_EXT_depth_clip_enable", deviceExtensionNames) && (!IsLinux() || nvidiaVersion.major > 418u)
  The depth value is not clamped to [0,1] for floating point depth buffers.
• disableFifoPresentMode (Vulkan workarounds) anglebug:3153: Disabled: IsLinux() && isIntel
  VK_PRESENT_MODE_FIFO_KHR causes random timeouts
• disableFlippingBlitWithCommand (Vulkan workarounds) anglebug:3498: Disabled: IsAndroid() && isQualcomm
  vkCmdBlitImage with flipped coordinates blits incorrectly.
• emulateAdvancedBlendEquations (Vulkan features) anglebug:3586: Disabled: !mFeatures.supportsBlendOperationAdvanced.enabled && !isIntel
  Emulate GL_KHR_blend_equation_advanced
• emulateDithering (Vulkan features) anglebug:6755: Disabled: IsAndroid()
  Emulate OpenGL dithering
• emulateR32fImageAtomicExchange (Vulkan workarounds) anglebug:5535: Enabled: true
  Emulate r32f images with r32ui to support imageAtomicExchange.
• emulateTransformFeedback (Vulkan features) anglebug:3205: Enabled: (!mFeatures.supportsTransformFeedbackExtension.enabled && mPhysicalDeviceFeatures.vertexPipelineStoresAndAtomics == 1U)
  Emulate transform feedback as the VK_EXT_transform_feedback is not present.
• emulatedPrerotation180 (Vulkan features) anglebug:4901: Disabled
  Emulate 180-degree prerotation.
• emulatedPrerotation270 (Vulkan features) anglebug:4901: Disabled
  Emulate 270-degree prerotation.
• emulatedPrerotation90 (Vulkan features) anglebug:4901: Disabled
  Emulate 90-degree prerotation.
• enableMultisampledRenderToTexture (Vulkan workarounds) anglebug:4937: Disabled: mFeatures.supportsMultisampledRenderToSingleSampled.enabled || (supportsIndependentDepthStencilResolve && !isSwiftShader && !(IsWindows() && (isIntel || isAMD)))
  Expose EXT_multisampled_render_to_texture
• enablePreRotateSurfaces (Vulkan features) anglebug:3502: Disabled: IsAndroid() && supportsNegativeViewport
  Enable Android pre-rotation for landscape applications
• enablePrecisionQualifiers (Vulkan features) anglebug:3078: Enabled: !(IsPixel2(mPhysicalDeviceProperties.vendorID, mPhysicalDeviceProperties.deviceID) && (mPhysicalDeviceProperties.driverVersion < kPixel2DriverWithRelaxedPrecision)) && !IsPixel4(mPhysicalDeviceProperties.vendorID, mPhysicalDeviceProperties.deviceID)
  Enable precision qualifiers in shaders
• explicitlyEnablePerSampleShading (Vulkan workarounds) anglebug:6876: Disabled: isARM
  Explicitly enable per-sample shading if the fragment shader contains the sample qualifier
• exposeNonConformantExtensionsAndVersions (Vulkan workarounds) anglebug:5375: Disabled: kExposeNonConformantExtensionsAndVersions
  Expose GLES versions and extensions that are not conformant.
• forceContinuousRefreshOnSharedPresent (Vulkan features) https://issuetracker.google.com/229267970: Disabled: false
  Force to create vulkan swapchain with continuous refresh on shared present
• forceD16TexFilter (Vulkan workarounds) anglebug:3452: Disabled: IsAndroid() && isQualcomm
  VK_FORMAT_D16_UNORM does not support VK_FORMAT_FEATURE_SAMPLED_IMAGE_FILTER_LINEAR_BIT, which prevents OES_depth_texture from being supported.
• forceDriverUniformOverSpecConst (Vulkan workarounds) http://issuetracker.google.com/173636783: Disabled: isQualcomm && mPhysicalDeviceProperties.driverVersion < kPixel4DriverWithWorkingSpecConstSupport
  Forces using driver uniforms instead of specialization constants.
• forceFallbackFormat (Vulkan workarounds): Disabled
  Force a fallback format for angle_end2end_tests
• forceFragmentShaderPrecisionHighpToMediump (Vulkan workarounds) https://issuetracker.google.com/184850002: Disabled: false
  Forces highp precision in fragment shader to mediump.
• forceMaxUniformBufferSize16KB (Vulkan workarounds) https://issuetracker.google.com/161903006: Disabled: isQualcomm && isAdreno540
  Force max uniform buffer size to 16K on some device due to bug
• forceNearestFiltering (Vulkan workarounds): Disabled
  Force nearest filtering when sampling.
• forceNearestMipFiltering (Vulkan workarounds): Disabled
  Force nearest mip filtering when sampling.
• forceSubmitImmutableTextureUpdates (Vulkan app workarounds) anglebug:6929: Disabled
  Force submit updates to immutable textures
• forceTextureLodOffset1 (Vulkan workarounds): Disabled
  Increase the minimum texture level-of-detail by 1 when sampling.
• forceTextureLodOffset2 (Vulkan workarounds): Disabled
  Increase the minimum texture level-of-detail by 2 when sampling.
• forceTextureLodOffset3 (Vulkan workarounds): Disabled
  Increase the minimum texture level-of-detail by 3 when sampling.
• forceTextureLodOffset4 (Vulkan workarounds): Disabled
  Increase the minimum texture level-of-detail by 4 when sampling.
• generateSPIRVThroughGlslang (Vulkan features) anglebug:4889: Disabled: kUseSpirvGenThroughGlslang
  Translate SPIR-V through glslang.
• logMemoryReportCallbacks (Vulkan features): Disabled: false
  Log each callback from VK_EXT_device_memory_report
• logMemoryReportStats (Vulkan features): Disabled: false
  Log stats from VK_EXT_device_memory_report each swap
• overrideSurfaceFormatRGB8ToRGBA8 (Vulkan workarounds) anglebug:6651: Enabled: true
  Override surface format GL_RGB8 to GL_RGBA8
• padBuffersToMaxVertexAttribStride (Vulkan workarounds) anglebug:4428: Disabled: isAMD || isSamsung
  Vulkan considers vertex attribute accesses to count up to the last multiple of the stride. This additional access supports AMD's robust buffer access implementation. AMDVLK in particular will return incorrect values when the vertex access extends into the range that would be the stride padding and the buffer is too small. This workaround limits GL_MAX_VERTEX_ATTRIB_STRIDE to a maximum value and pads up every buffer allocation size to be a multiple of the maximum stride.
• perFrameWindowSizeQuery (Vulkan workarounds) anglebug:3623: Disabled: isIntel || (IsWindows() && isAMD) || IsFuchsia() || isSamsung
  Vulkan swapchain is not returning VK_ERROR_OUT_OF_DATE when window resizing
• persistentlyMappedBuffers (Vulkan features) anglebug:2162: Enabled: true
  Persistently map buffer memory to reduce map/unmap IOCTL overhead.
• preferAggregateBarrierCalls (Vulkan workarounds) anglebug:4633: Enabled: isImmediateModeRenderer
  Single barrier call is preferred over multiple calls with fine grained pipeline stage dependency information
• preferCPUForBufferSubData (Vulkan features) http://issuetracker.google.com/200067929: Disabled: isARM
  Prefer use CPU to do bufferSubData instead of staged update.
• preferDeviceLocalMemoryHostVisible (Vulkan features) anglebug:7047: Enabled: !isDiscreteGPU
  Prefer adding HOST_VISIBLE flag for DEVICE_LOCAL memory when picking memory types
• preferDrawClearOverVkCmdClearAttachments (Vulkan workarounds) https://issuetracker.google.com/166809097: Disabled: isQualcomm
  On some hardware, clear using a draw call instead of vkCmdClearAttachments in the middle of render pass due to bugs
• preferSkippingInvalidateForEmulatedFormats (Vulkan workarounds) anglebug:6860: Enabled: isImmediateModeRenderer
  Skipping invalidate is preferred for emulated formats that have extra channels over re-clearing the image
• preferSubmitAtFBOBoundary (Vulkan workarounds) https://issuetracker.google.com/187425444: Disabled: isARM
  Submit commands to driver at each FBO boundary for performance improvements.
• provokingVertex (Vulkan features): Enabled: true
  Enable provoking vertex mode via VK_EXT_provoking_vertex extension
• retainSPIRVDebugInfo (Vulkan features) anglebug:5901: Disabled: getEnableValidationLayers()
  Retain debug info in SPIR-V blob.
• roundOutputAfterDithering (Vulkan workarounds) anglebug:6953: Disabled: isQualcomm
  Round output after dithering to workaround a driver bug that rounds the output up
• shadowBuffers (Vulkan features) anglebug:4339: Disabled: false
  Allocate a shadow buffer for GL buffer objects to reduce glMap* latency.
• supportsAndroidHardwareBuffer (Vulkan features): Disabled
  VkDevice supports the VK_ANDROID_external_memory_android_hardware_buffer extension
• supportsAndroidNativeFenceSync (Vulkan features) anglebug:2517: Disabled
  VkDevice supports the EGL_ANDROID_native_fence_sync extension
• supportsBlendOperationAdvanced (Vulkan features) anglebug:3586: Enabled: ExtensionFound("VK_EXT_blend_operation_advanced", deviceExtensionNames)
  VkDevice supports VK_EXT_blend_operation_advanced extension.
• supportsCustomBorderColor (Vulkan features) anglebug:3577: Disabled: mCustomBorderColorFeatures.customBorderColors == 1U && mCustomBorderColorFeatures.customBorderColorWithoutFormat == 1U && !isSwiftShader
  VkDevice supports the VK_EXT_custom_border_color extension
• supportsDepthClipControl (Vulkan features) anglebug:5421: Disabled: mDepthClipControlFeatures.depthClipControl == 1U
  VkDevice supports VK_EXT_depth_clip_control extension.
• supportsDepthStencilResolve (Vulkan features) anglebug:4836: Enabled: mFeatures.supportsRenderpass2.enabled && mDepthStencilResolveProperties.supportedDepthResolveModes != 0
  VkDevice supports the VK_KHR_depth_stencil_resolve extension with the independentResolveNone feature
• supportsExternalFenceCapabilities (Vulkan features): Enabled: true
  VkInstance supports the VK_KHR_external_fence_capabilities extension
• supportsExternalFenceFd (Vulkan features) anglebug:2517: Disabled: ExtensionFound("VK_KHR_external_fence_fd", deviceExtensionNames)
  VkDevice supports the VK_KHR_external_fence_fd extension
• supportsExternalMemoryDmaBufAndModifiers (Vulkan features) anglebug:6248: Disabled: ExtensionFound("VK_EXT_external_memory_dma_buf", deviceExtensionNames) && ExtensionFound("VK_EXT_image_drm_format_modifier", deviceExtensionNames)
  VkDevice supports the VK_EXT_external_memory_dma_buf and VK_EXT_image_drm_format_modifier extensions
• supportsExternalMemoryFd (Vulkan features): Disabled: ExtensionFound("VK_KHR_external_memory_fd", deviceExtensionNames)
  VkDevice supports the VK_KHR_external_memory_fd extension
• supportsExternalMemoryFuchsia (Vulkan features): Disabled: ExtensionFound("VK_FUCHSIA_external_memory", deviceExtensionNames)
  VkDevice supports the VK_FUCHSIA_external_memory extension
• supportsExternalMemoryHost (Vulkan features): Enabled: ExtensionFound("VK_EXT_external_memory_host", deviceExtensionNames)
  VkDevice supports the VK_EXT_external_memory_host extension
• supportsExternalSemaphoreCapabilities (Vulkan features): Enabled: true
  VkInstance supports the VK_KHR_external_semaphore_capabilities extension
• supportsExternalSemaphoreFd (Vulkan features): Disabled: ExtensionFound("VK_KHR_external_semaphore_fd", deviceExtensionNames)
  VkDevice supports the VK_KHR_external_semaphore_fd extension
• supportsExternalSemaphoreFuchsia (Vulkan features): Disabled: ExtensionFound("VK_FUCHSIA_external_semaphore", deviceExtensionNames)
  VkDevice supports the VK_FUCHSIA_external_semaphore extension
• supportsFilteringPrecision (Vulkan features): Enabled: ExtensionFound("VK_GOOGLE_sampler_filtering_precision", deviceExtensionNames)
  VkDevice supports the VK_GOOGLE_sampler_filtering_precision extension
• supportsFragmentShadingRate (Vulkan features) anglebug:7172: Disabled: canSupportFragmentShadingRate(deviceExtensionNames)
  VkDevice supports VK_KHR_fragment_shading_rate extension
• supportsGGPFrameToken (Vulkan features): Disabled
  VkDevice supports the VK_GGP_frame_token extension
• supportsGeometryStreamsCapability (Vulkan features) anglebug:3206: Disabled: mTransformFeedbackFeatures.geometryStreams == 1U
  Implementation supports the GeometryStreams SPIR-V capability.
• supportsHostQueryReset (Vulkan features) anglebug:6692: Enabled: (mHostQueryResetFeatures.hostQueryReset == 1U)
  VkDevice supports VK_EXT_host_query_reset extension
• supportsImageCubeArray (Vulkan features) anglebug:3584: Enabled: mPhysicalDeviceFeatures.imageCubeArray == 1U
  VkDevice supports the imageCubeArray feature properly
• supportsImageFormatList (Vulkan features) anglebug:5281: Enabled: ExtensionFound("VK_KHR_image_format_list", deviceExtensionNames)
  Enable VK_IMAGE_CREATE_MUTABLE_FORMAT_BIT by default for ICDs that support VK_KHR_image_format_list
• supportsIncrementalPresent (Vulkan features): Disabled: ExtensionFound("VK_KHR_incremental_present", deviceExtensionNames)
  VkDevice supports the VK_KHR_incremental_present extension
• supportsIndexTypeUint8 (Vulkan features) anglebug:4405: Disabled: mIndexTypeUint8Features.indexTypeUint8 == 1U
  VkDevice supports the VK_EXT_index_type_uint8 extension
• supportsLockSurfaceExtension (Vulkan features): Disabled: IsAndroid()
  Surface supports the EGL_KHR_lock_surface3 extension
• supportsMultiDrawIndirect (Vulkan features) anglebug:6439: Enabled: mPhysicalDeviceFeatures.multiDrawIndirect == 1U
  VkDevice supports the multiDrawIndirect extension
• supportsMultisampledRenderToSingleSampled (Vulkan features) anglebug:4836: Disabled: mFeatures.supportsRenderpass2.enabled && mFeatures.supportsDepthStencilResolve.enabled && mMultisampledRenderToSingleSampledFeatures.multisampledRenderToSingleSampled == 1U
  VkDevice supports the VK_EXT_multisampled_render_to_single_sampled extension
• supportsMultiview (Vulkan features) anglebug:6048: Enabled: mMultiviewFeatures.multiview == 1U
  VkDevice supports the VK_KHR_multiview extension
• supportsNegativeViewport (Vulkan features): Enabled: supportsNegativeViewport
  The driver supports inverting the viewport with a negative height.
• supportsPipelineStatisticsQuery (Vulkan features) anglebug:5430: Disabled: mPhysicalDeviceFeatures.pipelineStatisticsQuery == 1U
  VkDevice supports the pipelineStatisticsQuery feature
• supportsProtectedMemory (Vulkan features) anglebug:3965: Disabled: (mProtectedMemoryFeatures.protectedMemory == 1U) && !isARM
  VkDevice supports protected memory
• supportsRenderPassLoadStoreOpNone (Vulkan features) anglebug:5371: Enabled: ExtensionFound("VK_EXT_load_store_op_none", deviceExtensionNames)
  VkDevice supports VK_EXT_load_store_op_none extension.
• supportsRenderPassStoreOpNone (Vulkan features) anglebug:5055: Disabled: !mFeatures.supportsRenderPassLoadStoreOpNone.enabled && ExtensionFound("VK_QCOM_render_pass_store_ops", deviceExtensionNames)
  VkDevice supports VK_QCOM_render_pass_store_ops extension.
• supportsRenderpass2 (Vulkan features): Enabled: ExtensionFound("VK_KHR_create_renderpass2", deviceExtensionNames)
  VkDevice supports the VK_KHR_create_renderpass2 extension
• supportsShaderFloat16 (Vulkan features) anglebug:4551: Disabled: mShaderFloat16Int8Features.shaderFloat16 == 1U
  VkDevice supports the VK_KHR_shader_float16_int8 extension and has the shaderFloat16 feature
• supportsShaderFramebufferFetch (Vulkan features): Enabled: (IsAndroid() && isARM) || isSwiftShader
  Whether the Vulkan backend supports coherent framebuffer fetch
• supportsShaderFramebufferFetchNonCoherent (Vulkan features): Enabled: (IsAndroid() && !(isARM || isQualcomm)) || isSwiftShader
  Whether the Vulkan backend supports non-coherent framebuffer fetch
• supportsShaderStencilExport (Vulkan features): Enabled: ExtensionFound("VK_EXT_shader_stencil_export", deviceExtensionNames)
  VkDevice supports the VK_EXT_shader_stencil_export extension
• supportsSharedPresentableImageExtension (Vulkan features): Disabled
  VkSurface supports the VK_KHR_shared_presentable_images extension
• supportsSurfaceCapabilities2Extension (Vulkan features): Disabled
  VkInstance supports the VK_KHR_get_surface_capabilities2 extension
• supportsSurfaceProtectedCapabilitiesExtension (Vulkan features): Disabled
  VkInstance supports the VK_KHR_surface_protected_capabilities extension
• supportsSurfaceProtectedSwapchains (Vulkan features): Disabled: IsAndroid()
  VkSurface supportsProtected for protected swapchains
• supportsSurfacelessQueryExtension (Vulkan features): Disabled
  VkInstance supports the VK_GOOGLE_surfaceless_query extension
• supportsTransformFeedbackExtension (Vulkan features) anglebug:3206: Disabled: mTransformFeedbackFeatures.transformFeedback == 1U
  Transform feedback uses the VK_EXT_transform_feedback extension.
• supportsYUVSamplerConversion (Vulkan features): Enabled: mSamplerYcbcrConversionFeatures.samplerYcbcrConversion != 0U
  VkDevice supports the VK_KHR_sampler_ycbcr_conversion extension
• swapbuffersOnFlushOrFinishWithSingleBuffer (Vulkan features) anglebug:6878: Disabled: IsAndroid()
  Bypass deferredFlush with calling swapbuffers on flush or finish when in Shared Present mode
• useMultipleDescriptorsForExternalFormats (Vulkan workarounds) anglebug:6141: Enabled: true
  Return a default descriptor count for external formats.
• waitIdleBeforeSwapchainRecreation (Vulkan workarounds) anglebug:5061: Disabled: IsAndroid() && isARM
  Before passing an oldSwapchain to VkSwapchainCreateInfoKHR, wait for queue to be idle. Works around a bug on platforms which destroy oldSwapchain in vkCreateSwapchainKHR.

DAWN Info

• <Discrete GPU> D3D12 backend - NVIDIA GeForce RTX 2060 SUPER
  [Default Toggle Names]
• lazy_clear_resource_on_first_use:https://crbug.com/dawn/145: Clears resource to zero on first usage. This initializes the resource so that no dirty bits from recycled memory is present in the new resource.
• use_d3d12_resource_heap_tier2:https://crbug.com/dawn/27: Enable support for resource heap tier 2. Resource heap tier 2 allows mixing of texture and buffers in the same heap. This allows better heap re-use and reduces fragmentation.
• use_d3d12_render_pass:https://crbug.com/dawn/36: Use the D3D12 render pass API introduced in Windows build 1809 by default. On versions of Windows prior to build 1809, or when this toggle is turned off, Dawn will emulate a render pass.
• use_d3d12_residency_management:https://crbug.com/dawn/193: Enable residency management. This allows page-in and page-out of resource heaps in GPU memory. This component improves overcommitted performance by keeping the most recently used resources local to the GPU. Turning this component off can cause allocation failures when application memory exceeds physical device memory.
• disallow_unsafe_apis:http://crbug.com/1138528: Produces validation errors on API entry points or parameter combinations that aren't considered secure yet.
• d3d12_split_buffer_texture_copy_for_rows_per_image_paddings:https://crbug.com/dawn/1289: D3D12 requires more buffer storage than it should when rowsPerImage is greater than copyHeight, which means there are pure padding row(s) on each image. In this situation, the buffer used for B2T/T2B copy might be big enough according to WebGPU's spec but it doesn't meet D3D12's requirement, then we need to workaround it via split the copy operation into two copies, in order to make B2T/T2B copy being done correctly on D3D12.
• [WebGPU Forced Toggles - enabled]
• disallow_spirv:https://crbug.com/1214923: Disallow usage of SPIR-V completely so that only WGSL is used for shader modules. This is useful to prevent a Chromium renderer process from successfully sending SPIR-V code to be compiled in the GPU process.
• [Supported Features]
• texture-compression-bc
• pipeline-statistics-query
• timestamp-query
• depth24unorm-stencil8
• depth32float-stencil8
• dawn-internal-usages
• multiplanar-formats
• dawn-native
• <CPU> D3D12 backend - Microsoft Basic Render Driver
  [Default Toggle Names]
• lazy_clear_resource_on_first_use:https://crbug.com/dawn/145: Clears resource to zero on first usage. This initializes the resource so that no dirty bits from recycled memory is present in the new resource.
• use_d3d12_resource_heap_tier2:https://crbug.com/dawn/27: Enable support for resource heap tier 2. Resource heap tier 2 allows mixing of texture and buffers in the same heap. This allows better heap re-use and reduces fragmentation.
• use_d3d12_render_pass:https://crbug.com/dawn/36: Use the D3D12 render pass API introduced in Windows build 1809 by default. On versions of Windows prior to build 1809, or when this toggle is turned off, Dawn will emulate a render pass.
• use_d3d12_residency_management:https://crbug.com/dawn/193: Enable residency management. This allows page-in and page-out of resource heaps in GPU memory. This component improves overcommitted performance by keeping the most recently used resources local to the GPU. Turning this component off can cause allocation failures when application memory exceeds physical device memory.
• disallow_unsafe_apis:http://crbug.com/1138528: Produces validation errors on API entry points or parameter combinations that aren't considered secure yet.
• d3d12_split_buffer_texture_copy_for_rows_per_image_paddings:https://crbug.com/dawn/1289: D3D12 requires more buffer storage than it should when rowsPerImage is greater than copyHeight, which means there are pure padding row(s) on each image. In this situation, the buffer used for B2T/T2B copy might be big enough according to WebGPU's spec but it doesn't meet D3D12's requirement, then we need to workaround it via split the copy operation into two copies, in order to make B2T/T2B copy being done correctly on D3D12.
• [WebGPU Forced Toggles - enabled]
• disallow_spirv:https://crbug.com/1214923: Disallow usage of SPIR-V completely so that only WGSL is used for shader modules. This is useful to prevent a Chromium renderer process from successfully sending SPIR-V code to be compiled in the GPU process.
• [Supported Features]
• texture-compression-bc
• pipeline-statistics-query
• timestamp-query
• depth24unorm-stencil8
• depth32float-stencil8
• dawn-internal-usages
• multiplanar-formats
• dawn-native
• <Discrete GPU> Vulkan backend - NVIDIA GeForce RTX 2060 SUPER
  [Default Toggle Names]
• lazy_clear_resource_on_first_use:https://crbug.com/dawn/145: Clears resource to zero on first usage. This initializes the resource so that no dirty bits from recycled memory is present in the new resource.
• use_temporary_buffer_in_texture_to_texture_copy:https://crbug.com/dawn/42: Split texture-to-texture copy into two copies: copy from source texture into a temporary buffer, and copy from the temporary buffer into the destination texture when copying between compressed textures that don't have block-aligned sizes. This workaround is enabled by default on all Vulkan drivers to solve an issue in the Vulkan SPEC about the texture-to-texture copies with compressed formats. See #1005 (https://github.com/KhronosGroup/Vulkan-Docs/issues/1005) for more details.
• vulkan_use_d32s8:https://crbug.com/dawn/286: Vulkan mandates support of either D32_FLOAT_S8 or D24_UNORM_S8. When available the backend will use D32S8 (toggle to on) but setting the toggle to off will make it use the D24S8 format when possible.
• vulkan_use_s8:https://crbug.com/dawn/666: Vulkan has a pure stencil8 format but it is not universally available. When this toggle is on, the backend will use S8 for the stencil8 format, otherwise it will fallback to D32S8 or D24S8.
• disallow_unsafe_apis:http://crbug.com/1138528: Produces validation errors on API entry points or parameter combinations that aren't considered secure yet.
• use_vulkan_zero_initialize_workgroup_memory_extension:https://crbug.com/dawn/1302: Initialize workgroup memory with OpConstantNull on Vulkan when the Vulkan extension VK_KHR_zero_initialize_workgroup_memory is supported.
• [WebGPU Forced Toggles - enabled]
• disallow_spirv:https://crbug.com/1214923: Disallow usage of SPIR-V completely so that only WGSL is used for shader modules. This is useful to prevent a Chromium renderer process from successfully sending SPIR-V code to be compiled in the GPU process.
• [Supported Featur