Skip to content
Merged
Show file tree
Hide file tree
Changes from 14 commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
76 changes: 71 additions & 5 deletions docs/ReleaseNotes.md
Original file line number Diff line number Diff line change
Expand Up @@ -19,15 +19,81 @@ The included licenses apply to the following files:

### Upcoming Release

- Fixed regression: [#7510](https://github.com/microsoft/DirectXShaderCompiler/issues/7510) crash when calling `sizeof` on templated type.
- Fixed regression: [#7508](https://github.com/microsoft/DirectXShaderCompiler/issues/7508) crash when calling `Load` with `status`.
- Header file `dxcpix.h` was added to the release package.
#### Shader Model 6.9 Release

- Shader Model 6.9 is fully supported.
- See [the official blog
post](https://devblogs.microsoft.com/directx/shader-model-6-9-dxr-1-2-and-agilitysdk-1-619-release)
for more details.
- Execution tests for all intrinsics supporting DXIL vectors. See the [test
plan](https://github.com/microsoft/hlsl-specs/blob/main/proposals/infra/INF-0006-Long-Vector-ExecutionTest-Plan.md)
for more info.

#### Experimental Shader Model 6.10

- Moved Linear Algebra (Cooperative Vector) DXIL Opcodes to experimental Shader Model 6.10
- Added support for `long long` and `unsigned long long` compile-time constant evaluation, fixes [#7952](https://github.com/microsoft/DirectXShaderCompiler/issues/7952).
- Implement GetGroupWaveIndex and GetGroupWaveCount in experimental Shader Model 6.10
- Implement GetGroupWaveIndex and GetGroupWaveCount in experimental Shader Model 6.10.
- [proposal](https://github.com/microsoft/hlsl-specs/blob/main/proposals/0048-group-wave-index.md)
- GetGroupWaveIndex: New intrinsic for Compute, Mesh, Amplification and Node shaders which returns the index of the wave within the thread group that the the thread is executing.
- GetGroupWaveCount: New intrinsic for Compute, Mesh, Amplification and Node shaders which returns the total number of waves executing within the thread group.
- GetGroupWaveCount: New intrinsic for Compute, Mesh, Amplification and Node
shaders which returns the total number of waves executing within the thread
group.

#### Noteble SPIR-V updates

- Handle vector element assignment for asuint. [#8011](https://github.com/microsoft/DirectXShaderCompiler/issues/8011)
- Support sizeof(vk::BufferPointer). [#8010](https://github.com/microsoft/DirectXShaderCompiler/issues/8010)
- Scalar layout to follow C structure layout. [#7996](https://github.com/microsoft/DirectXShaderCompiler/issues/7996)
- Support asdouble() for uint3 argument type. [#7965](https://github.com/microsoft/DirectXShaderCompiler/issues/7965)
- Add const to many FeatureManager fns. [#7980](https://github.com/microsoft/DirectXShaderCompiler/issues/7980)
- Handle vk::BufferPointer in initializer list. [#7946](https://github.com/microsoft/DirectXShaderCompiler/issues/7946)
- Fix layout rule on ConstantBuffer alias. [#7960](https://github.com/microsoft/DirectXShaderCompiler/issues/7960)
- Fix layout rule for BufferPointer pointee type. [#7956](https://github.com/microsoft/DirectXShaderCompiler/issues/7956)
- Use desugared type when processing binary op. [#7948](https://github.com/microsoft/DirectXShaderCompiler/issues/7948)
- Handle associated counters for RWStructuredBuffer in base classes. [#7880](https://github.com/microsoft/DirectXShaderCompiler/issues/7880)
- Fix precision for dot2add. [#7861](https://github.com/microsoft/DirectXShaderCompiler/issues/7861)
- Support sign() intrinsics for unsigned integers. [#7845](https://github.com/microsoft/DirectXShaderCompiler/issues/7845)
- Fix resource heap & fvk-bind-register interactions. [#7858](https://github.com/microsoft/DirectXShaderCompiler/issues/7858)
- Fix incorrect branch gen on return. [#7834](https://github.com/microsoft/DirectXShaderCompiler/issues/7834)
- Add diagnostic for boolean bitfields. [#7722](https://github.com/microsoft/DirectXShaderCompiler/issues/7822)
- Implement WaveOpsIncludeHelperLanes. [#7806](https://github.com/microsoft/DirectXShaderCompiler/issues/7806)
- Fix spirv codegen for uabs intrinsic when argument is unsigned. [#7750](https://github.com/microsoft/DirectXShaderCompiler/issues/7750)
- Fix invalid codegen for empty cbuffers. [#7717](https://github.com/microsoft/DirectXShaderCompiler/issues/7717)
- Preserve NaN, Inf, and signed zeros w/ -Gis. [#7693](https://github.com/microsoft/DirectXShaderCompiler/issues/7693)
- Fix pcf with rich debug info enabled. [#7663](https://github.com/microsoft/DirectXShaderCompiler/issues/7663)
- Create only 1 DebugCompilationUnit per spirv module. [#7669](https://github.com/microsoft/DirectXShaderCompiler/issues/7669)
- Handle member traversal with template type. [#7674](https://github.com/microsoft/DirectXShaderCompiler/issues/7674)
- Handle partial template class specialization. [#7673](https://github.com/microsoft/DirectXShaderCompiler/issues/7673)
- Fix declaration order of values in decorations. [#7672](https://github.com/microsoft/DirectXShaderCompiler/issues/7672)
- Fix DebugSource for files which are not found. [#7662](https://github.com/microsoft/DirectXShaderCompiler/issues/7662)
- Fixed a crash if encounter constant buffer fields with overlapping register
assignments. [#7636](https://github.com/microsoft/DirectXShaderCompiler/issues/7636)
- Add option to use the Unknown image format. [#7632](https://github.com/microsoft/DirectXShaderCompiler/issues/7632)
- Add the derivative group execution mode only on shader types that allow it. [#7628](https://github.com/microsoft/DirectXShaderCompiler/issues/7628)
- Allow spirv type as template parameter. [#7626](https://github.com/microsoft/DirectXShaderCompiler/issues/7626)
- Explicitly state which layout rules require scalar block layout. [#7539](https://github.com/microsoft/DirectXShaderCompiler/issues/7539)
- Emit DebugScope in wrapper. [#77341](https://github.com/microsoft/DirectXShaderCompiler/issues/7341)
- Use unknown image format in vk1.3 and later. [#7528](https://github.com/microsoft/DirectXShaderCompiler/issues/7528)
- Use OpCopyLogical to reconstruct values. [#7530](https://github.com/microsoft/DirectXShaderCompiler/issues/7530)
- AMD work graphs extension. [#7353](https://github.com/microsoft/DirectXShaderCompiler/issues/7353)
- Get Alignment from pointee type for vk::BufferPointer store. [#7501](https://github.com/microsoft/DirectXShaderCompiler/issues/7501)
- Fix bool cast on buffers with swizzle. [#7497](https://github.com/microsoft/DirectXShaderCompiler/issues/7497)
- Add payload to OpEmitMeshTasksEXT. [#7485](https://github.com/microsoft/DirectXShaderCompiler/issues/7485)
- Fix r-value being used in mul intrinsic. [#7489](https://github.com/microsoft/DirectXShaderCompiler/issues/7489)
- Several small bug fixes.

#### Other Changes

- Fixed regression: [#7510](https://github.com/microsoft/DirectXShaderCompiler/issues/7510) crash when calling `sizeof` on templated type.
- Fixed regression: [#7508](https://github.com/microsoft/DirectXShaderCompiler/issues/7508) crash when calling `Load` with `status`.
- Header file `dxcpix.h` was added to the release package.
- C4146 (unary minus operatore applied to unsigned type) enabled as a build
break for MSVC builds.
- Implement isnormal function.
[#7720](https://github.com/microsoft/DirectXShaderCompiler/issues/7720)
- Disallow 64bit msad, ibfe, and ubfe. [#7774](https://github.com/microsoft/DirectXShaderCompiler/issues/7774)
- Various test fixes and project infrastructure improvements.

### Version 1.8.2505

Expand Down
4 changes: 2 additions & 2 deletions utils/version/latest-release.json
Original file line number Diff line number Diff line change
Expand Up @@ -2,7 +2,7 @@
"version": {
"major": "1",
"minor": "9",
"rev": "2601"
"rev": "2602"
},
"sha": "b06a58d960fb0c9313cda10d17cc984a610b880f"
"sha": "f255809c3dd408c32349d6b7b16019b8b24e1d2e"
}
6 changes: 3 additions & 3 deletions utils/version/version.inc
Original file line number Diff line number Diff line change
Expand Up @@ -13,12 +13,12 @@
#ifdef RC_VERSION_FIELD_2
#undef RC_VERSION_FIELD_2
#endif
#define RC_VERSION_FIELD_2 8
#define RC_VERSION_FIELD_2 9

#ifdef RC_VERSION_FIELD_3
#undef RC_VERSION_FIELD_3
#endif
#define RC_VERSION_FIELD_3 2505
#define RC_VERSION_FIELD_3 2602

#ifdef RC_VERSION_FIELD_4
#undef RC_VERSION_FIELD_4
Expand All @@ -28,7 +28,7 @@
#ifdef RC_FILE_VERSION
#undef RC_FILE_VERSION
#endif
#define RC_FILE_VERSION "1.8.2505.0"
#define RC_FILE_VERSION "1.9.2602.0"

#ifdef RC_FILE_DESCRIPTION
#undef RC_FILE_DESCRIPTION
Expand Down