Skip to content

Commit ea53cb5

Browse files
V-FEXrtdamyanp
andauthored
Update release notes (#8396)
Updates the release notes based on PRs since the last release --------- Co-authored-by: Damyan Pepper <[email protected]>
1 parent 0958b62 commit ea53cb5

1 file changed

Lines changed: 28 additions & 6 deletions

File tree

docs/ReleaseNotes.md

Lines changed: 28 additions & 6 deletions
Original file line numberDiff line numberDiff line change
@@ -17,15 +17,16 @@ The included licenses apply to the following files:
1717

1818
## Changelog
1919

20-
### Upcoming Release
21-
22-
Place release notes for the upcoming release below this line and remove this
23-
line upon naming the release. Refer to previous for appropriate section names.
20+
### Version 1.10.2605 (preview)
2421

2522
#### Experimental Shader Model 6.10
2623

27-
- Removed experimental Cooperative Vector, this has been replaced by LinAlg matrix.
28-
- Implement GetGroupWaveIndex and GetGroupWaveCount in experimental Shader Model 6.10.
24+
- Added DirectX LinAlg Matrix API in experimental Shader Model 6.10.
25+
- [proposal](https://github.com/microsoft/hlsl-specs/blob/main/proposals/0035-linalg-matrix.md)
26+
- New `Matrix` type that operates under three "scopes". ThreadScope, WaveScope, ThreadGroupScope
27+
- Supports a comprehensive set of Matrix operations including matrix-vector and matrix-matrix multiply
28+
- Removed experimental Cooperative Vector, this has been replaced by LinAlg Matrix.
29+
- Added GetGroupWaveIndex and GetGroupWaveCount in experimental Shader Model 6.10.
2930
- [proposal](https://github.com/microsoft/hlsl-specs/blob/main/proposals/0048-group-wave-index.md)
3031
- 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.
3132
- GetGroupWaveCount: New intrinsic for Compute, Mesh, Amplification and Node
@@ -35,6 +36,13 @@ line upon naming the release. Refer to previous for appropriate section names.
3536
- `DebugBreak()` triggers a breakpoint if a debugger is attached.
3637
- `dx::IsDebuggerPresent()` returns true if a debugger is attached.
3738
- SPIR-V: `DebugBreak()` emits `NonSemantic.DebugBreak` extended instruction; `IsDebuggerPresent()` is not supported.
39+
- Added TriangleObjectPositions in experimental Shader Model 6.10.
40+
- [proposal](https://github.com/microsoft/hlsl-specs/blob/main/proposals/0041-triangle-object-positions.md)
41+
- Added CluterID in experimental Shader Model 6.10.
42+
- [proposal](https://github.com/microsoft/hlsl-specs/blob/main/proposals/0045-clustered-geometry.md)
43+
- Added Variable GroupShared Limits in experimental Shader Model 6.10.
44+
- [proposal](https://github.com/microsoft/hlsl-specs/blob/main/proposals/0049-variable-groupshared-memory.md)
45+
3846

3947
#### Bug Fixes
4048

@@ -45,6 +53,18 @@ line upon naming the release. Refer to previous for appropriate section names.
4553
- User-defined conversion operators (e.g., `operator float4()`) now produce an
4654
error instead of being silently ignored.
4755
[#5103](https://github.com/microsoft/DirectXShaderCompiler/pull/8206)
56+
- Fixed SPIRV crash with out-of-line template decl
57+
[#8079](https://github.com/microsoft/DirectXShaderCompiler/pull/8079)
58+
- Fixed crash in CanConvert from incomplete type
59+
[#8110](https://github.com/microsoft/DirectXShaderCompiler/pull/8110)
60+
- Fixed crash in diagnostics when source contains embedded nulls
61+
[#8164](https://github.com/microsoft/DirectXShaderCompiler/pull/8164)
62+
- Fixed OOB in subscript indexing of col-major matrix in cbuffer
63+
[#7866](https://github.com/microsoft/DirectXShaderCompiler/pull/7866)
64+
- Allow groupshared args in templates
65+
[#8217](https://github.com/microsoft/DirectXShaderCompiler/pull/8217)
66+
- Fixed RawBufferVectorLoad/Store to use 32-bit element types for min precision types instead of 16-bit
67+
[#8274](https://github.com/microsoft/DirectXShaderCompiler/pull/8274)
4868
- DXIL validation: added validation for `CreateHandleFromBinding`.
4969
- DXIL validation now rejects non-standard integer bit widths (e.g. `i25`) in
5070
instructions.
@@ -55,6 +75,8 @@ line upon naming the release. Refer to previous for appropriate section names.
5575
default, with `/Fi` to override the output filename. The old FXC-style `/P
5676
<filename>` positional syntax has been renamed to `/Po`.
5777
[#4611](https://github.com/microsoft/DirectXShaderCompiler/issues/4611).
78+
- Added support for -Fre for Metal
79+
[#8159](https://github.com/microsoft/DirectXShaderCompiler/pull/8159)
5880

5981

6082
### Version 1.9.2602

0 commit comments

Comments
 (0)