encounter/dawn-cmake
1
0
Fork 0
mirror of https://github.com/encounter/dawn-cmake.git synced 2025-07-07 13:45:51 +00:00
Code Issues Packages Projects Releases Wiki Activity

Fix bugs about TextureViewDesc's default values

Browse Source 
TextureViewDescriptor's default values for dimension and
arrayLayerCount in Dawn are not correct according to WebGPU spec.
This change fixes these bugs.

Bug: dawn:760
Change-Id: Ic1d069838d6c0f7bb1afa1dceaf73e91bdfdb20a
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/58020
Commit-Queue: Yunchao He <yunchao.he@intel.com>
Reviewed-by: Kai Ninomiya <kainino@chromium.org>
This commit is contained in:
Yunchao He 2021-07-15 18:33:48 +00:00 committed by Dawn LUCI CQ
parent 5cbb168bbc
commit 37140e7c62
5 changed files with 75 additions and 27 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

19
src/dawn_native/Texture.cpp
View File

@ -362,11 +362,7 @@ namespace dawn_native {
break;
case wgpu::TextureDimension::e2D:
if (texture->GetArrayLayers() > 1u && desc.arrayLayerCount == 0) {
desc.dimension = wgpu::TextureViewDimension::e2DArray;
} else {
desc.dimension = wgpu::TextureViewDimension::e2D;
}
break;
case wgpu::TextureDimension::e3D:
@ -380,7 +376,22 @@ namespace dawn_native {
desc.format = texture->GetFormat().format;
}
if (desc.arrayLayerCount == 0) {
switch (desc.dimension) {
case wgpu::TextureViewDimension::e1D:
case wgpu::TextureViewDimension::e2D:
case wgpu::TextureViewDimension::e3D:
desc.arrayLayerCount = 1;
break;
case wgpu::TextureViewDimension::Cube:
desc.arrayLayerCount = 6;
break;
case wgpu::TextureViewDimension::e2DArray:
case wgpu::TextureViewDimension::CubeArray:
desc.arrayLayerCount = texture->GetArrayLayers() - desc.baseArrayLayer;
break;
default:
UNREACHABLE();
}
}
if (desc.mipLevelCount == 0) {
desc.mipLevelCount = texture->GetNumMipLevels() - desc.baseMipLevel;

49
src/tests/end2end/StorageTextureTests.cpp
View File

@ -583,8 +583,10 @@ fn IsEqualTo(pixel : vec4<f32>, expected : vec4<f32>) -> bool {
<< fragmentShader;
}
void CheckResultInStorageBuffer(wgpu::Texture readonlyStorageTexture,
const std::string& computeShader) {
void CheckResultInStorageBuffer(
wgpu::Texture readonlyStorageTexture,
const std::string& computeShader,
wgpu::TextureViewDimension dimension = wgpu::TextureViewDimension::e2D) {
wgpu::ComputePipeline pipeline = CreateComputePipeline(computeShader.c_str());
// Clear the content of the result buffer into 0.
@ -592,9 +594,11 @@ fn IsEqualTo(pixel : vec4<f32>, expected : vec4<f32>) -> bool {
wgpu::Buffer resultBuffer =
utils::CreateBufferFromData(device, &kInitialValue, sizeof(kInitialValue),
wgpu::BufferUsage::Storage | wgpu::BufferUsage::CopySrc);
wgpu::BindGroup bindGroup =
utils::MakeBindGroup(device, pipeline.GetBindGroupLayout(0),
{{0, readonlyStorageTexture.CreateView()}, {1, resultBuffer}});
wgpu::TextureViewDescriptor descriptor;
descriptor.dimension = dimension;
wgpu::BindGroup bindGroup = utils::MakeBindGroup(
device, pipeline.GetBindGroupLayout(0),
{{0, readonlyStorageTexture.CreateView(&descriptor)}, {1, resultBuffer}});
wgpu::CommandEncoder encoder = device.CreateCommandEncoder();
wgpu::ComputePassEncoder computeEncoder = encoder.BeginComputePass();
@ -635,12 +639,17 @@ fn IsEqualTo(pixel : vec4<f32>, expected : vec4<f32>) -> bool {
queue.Submit(1, &commandBuffer);
}
void WriteIntoStorageTextureInComputePass(wgpu::Texture writeonlyStorageTexture,
const char* computeShader) {
void WriteIntoStorageTextureInComputePass(
wgpu::Texture writeonlyStorageTexture,
const char* computeShader,
wgpu::TextureViewDimension dimension = wgpu::TextureViewDimension::e2D) {
// Create a compute pipeline that writes the expected pixel values into the storage texture.
wgpu::TextureViewDescriptor descriptor;
descriptor.dimension = dimension;
wgpu::ComputePipeline pipeline = CreateComputePipeline(computeShader);
wgpu::BindGroup bindGroup = utils::MakeBindGroup(
device, pipeline.GetBindGroupLayout(0), {{0, writeonlyStorageTexture.CreateView()}});
wgpu::BindGroup bindGroup =
utils::MakeBindGroup(device, pipeline.GetBindGroupLayout(0),
{{0, writeonlyStorageTexture.CreateView(&descriptor)}});
wgpu::CommandEncoder encoder = device.CreateCommandEncoder();
wgpu::ComputePassEncoder computePassEncoder = encoder.BeginComputePass();
@ -652,14 +661,19 @@ fn IsEqualTo(pixel : vec4<f32>, expected : vec4<f32>) -> bool {
queue.Submit(1, &commandBuffer);
}
void ReadWriteIntoStorageTextureInComputePass(wgpu::Texture readonlyStorageTexture,
void ReadWriteIntoStorageTextureInComputePass(
wgpu::Texture readonlyStorageTexture,
wgpu::Texture writeonlyStorageTexture,
const char* computeShader) {
const char* computeShader,
wgpu::TextureViewDimension dimension = wgpu::TextureViewDimension::e2D) {
// Create a compute pipeline that writes the expected pixel values into the storage texture.
wgpu::TextureViewDescriptor descriptor;
descriptor.dimension = dimension;
wgpu::ComputePipeline pipeline = CreateComputePipeline(computeShader);
wgpu::BindGroup bindGroup = utils::MakeBindGroup(
device, pipeline.GetBindGroupLayout(0),
{{0, writeonlyStorageTexture.CreateView()}, {1, readonlyStorageTexture.CreateView()}});
wgpu::BindGroup bindGroup =
utils::MakeBindGroup(device, pipeline.GetBindGroupLayout(0),
{{0, writeonlyStorageTexture.CreateView(&descriptor)},
{1, readonlyStorageTexture.CreateView(&descriptor)}});
wgpu::CommandEncoder encoder = device.CreateCommandEncoder();
wgpu::ComputePassEncoder computePassEncoder = encoder.BeginComputePass();
@ -991,7 +1005,7 @@ TEST_P(StorageTextureTests, Readonly2DArrayOr3DStorageTexture) {
}
})";
CheckResultInStorageBuffer(readonlyStorageTexture, csStream.str());
CheckResultInStorageBuffer(readonlyStorageTexture, csStream.str(), dimension);
}
}
@ -1018,7 +1032,8 @@ TEST_P(StorageTextureTests, Writeonly2DArrayOr3DStorageTexture) {
// Write the expected pixel values into the write-only storage texture.
const std::string computeShader =
CommonWriteOnlyTestCode("compute", kTextureFormat, dimension);
WriteIntoStorageTextureInComputePass(writeonlyStorageTexture, computeShader.c_str());
WriteIntoStorageTextureInComputePass(writeonlyStorageTexture, computeShader.c_str(),
dimension);
// Verify the pixel data in the write-only storage texture is expected.
CheckOutputStorageTexture(writeonlyStorageTexture, kTextureFormat, kSliceCount);
@ -1053,7 +1068,7 @@ TEST_P(StorageTextureTests, ReadWrite2DArrayOr3DStorageTexture) {
// Read values from read-only storage texture and write into the write-only storage texture.
const std::string computeShader = CommonReadWriteTestCode(kTextureFormat, dimension);
ReadWriteIntoStorageTextureInComputePass(readonlyStorageTexture, writeonlyStorageTexture,
computeShader.c_str());
computeShader.c_str(), dimension);
// Verify the data in the write-only storage texture is expected.
CheckOutputStorageTexture(writeonlyStorageTexture, kTextureFormat, kSliceCount);

4
src/tests/end2end/TextureViewTests.cpp
View File

@ -361,7 +361,9 @@ TEST_P(TextureViewSamplingTest, Default2DArrayTexture) {
constexpr uint32_t kMipLevels = 1;
initTexture(kLayers, kMipLevels);
wgpu::TextureView textureView = mTexture.CreateView();
wgpu::TextureViewDescriptor descriptor;
descriptor.dimension = wgpu::TextureViewDimension::e2DArray;
wgpu::TextureView textureView = mTexture.CreateView(&descriptor);
const char* fragmentShader = R"(
[[group(0), binding(0)]] var sampler0 : sampler;

4
src/tests/unittests/validation/RenderPassDescriptorValidationTests.cpp
View File

@ -559,7 +559,9 @@ namespace {
wgpu::Texture resolveTexture =
CreateTexture(device, wgpu::TextureDimension::e2D, kColorFormat, kSize, kSize,
kArrayLayers2, kLevelCount);
wgpu::TextureView resolveTextureView = resolveTexture.CreateView();
wgpu::TextureViewDescriptor viewDesc;
viewDesc.dimension = wgpu::TextureViewDimension::e2DArray;
wgpu::TextureView resolveTextureView = resolveTexture.CreateView(&viewDesc);
utils::ComboRenderPassDescriptor renderPass = CreateMultisampledRenderPass();
renderPass.cColorAttachments[0].resolveTarget = resolveTextureView;

20
src/tests/unittests/validation/TextureViewValidationTests.cpp
View File

@ -288,7 +288,7 @@ namespace {
// specifying the values they're supposed to default to.
// Variant for a 2D texture with more than 1 array layer.
TEST_F(TextureViewValidationTest, TextureViewDescriptorDefaults2DArray) {
constexpr uint32_t kDefaultArrayLayers = 6;
constexpr uint32_t kDefaultArrayLayers = 8;
wgpu::Texture texture = Create2DArrayTexture(device, kDefaultArrayLayers);
{ texture.CreateView(); }
@ -307,7 +307,25 @@ namespace {
texture.CreateView(&descriptor);
descriptor.dimension = wgpu::TextureViewDimension::e2DArray;
texture.CreateView(&descriptor);
// Setting view dimension to 2D, its arrayLayer will default to 1. And view creation
// will success.
descriptor.dimension = wgpu::TextureViewDimension::e2D;
texture.CreateView(&descriptor);
// Setting view dimension to Cube, its arrayLayer will default to 6.
descriptor.dimension = wgpu::TextureViewDimension::Cube;
texture.CreateView(&descriptor);
descriptor.baseArrayLayer = 2;
texture.CreateView(&descriptor);
descriptor.baseArrayLayer = 3;
ASSERT_DEVICE_ERROR(texture.CreateView(&descriptor));
// Setting view dimension to CubeArray, its arrayLayer will default to
// size.depthOrArrayLayers (kDefaultArrayLayers) - baseArrayLayer.
descriptor.dimension = wgpu::TextureViewDimension::CubeArray;
descriptor.baseArrayLayer = 0;
ASSERT_DEVICE_ERROR(texture.CreateView(&descriptor));
descriptor.baseArrayLayer = 2;
texture.CreateView(&descriptor);
descriptor.baseArrayLayer = 3;
ASSERT_DEVICE_ERROR(texture.CreateView(&descriptor));
}
{