Skip to content

Releases: shader-slang/slang

v0.24.47

02 Dec 14:53
92ae494
Compare
Choose a tag to compare
Slang v0.24.47

* 92ae4949 Inline functions with string param/return for GPU targets (#2544)
* 30b5335f Add helper functions to update DifferentialPair (#2547)
* 1562e73d Cleanup creation of `IDifferentiableType` in semantics checking. (#2548)
* f7d32a33 include `slang-rt` in windows release package. (#2542)
* 6ba4e6c4 Enable multithread build in premake (#2545)
* e7df8538 Allow `no_diff` on `this` parameter. (#2543)

v0.24.46

30 Nov 23:02
f51f69d
Compare
Choose a tag to compare
Slang v0.24.46

Changes:

f51f69d0 Fix missing semantic highlighting in attributes and ExtractExitentialValueExpr. (#2541)
976f5785 Cleanup DigestBuilder and MD5HashGen (#2539)
09684224 Support `no_diff` on existential typed params. (#2540)
f52b4de3 Allow `no_diff` modifier on parameters (#2538)
f5581786 Refactored reverse-mode implementation to use 4 separate passes. (#2531)
af7f4006 Complete removal of DifferentialBottom type. (#2537)
d85c7b80 FileStream-based implementation for updating cache index file (#2485)
d60c9252 Bug fix: partially specialized non-static generic invoke missing `this` argument. (#2536)
a54471a1 Add LockFile helper class (#2535)

v0.24.45

24 Nov 01:58
1b40fe5
Compare
Choose a tag to compare
Slang v0.24.45

Changes:
1b40fe56 Make differentiable data-flow pass recognize interface methods. (#2530)
d4787e92 Update comments in test. (#2529)
4ad04700 Fix issues around dynamic generic function and autodiff. (#2528)
97cb4851 Autodiff through simple dynamic dispatch. (#2527)
6178cb60 Refactor Auto-diff passes (#2526)

v0.24.44

21 Nov 20:50
d9b014c
Compare
Choose a tag to compare
Slang v0.24.44

Changes:
d9b014cb Added autodiff design doc (#2524)
b5626d4c Improve parser recovery around invalid function definitions. (#2525)
545de512 WIP: Fixed inout struct and added testing for calls to non-differentiable functions (#2505)
d58e08f8 Data flow validation pass for diagnosing derivative loss. (#2523)
0a050a43 Small fixes around SER on VK (#2522)
e13d38b6 Language server improvements for auto-diff. (#2521)
801aa3b4 Clean up type checking of higher order expressions. (#2519)
09d8e048 Squash warnings (#2515)

v0.24.43

16 Nov 16:11
8d50047
Compare
Choose a tag to compare
Slang v0.24.43

* 8d500475 (HEAD -> master, origin/master, origin/HEAD) `slang-llvm` 13.x-32 (#2518)
* 1643471d Mesh shader support (#2464)
* 4917d711 Shader Execution Reordering for VK (#2491)
* 7222b6cd Specify downstream compiler include paths (#2517)
* 069768d4 Add mesh and amplification shaders to slang-glslang (#2516)
* 25affe8e Adding some math functions and their derivatives (#2497)
* 368ec311 Minimum binary arithmetic reverse autodiff working. (#2514)
* 623f5c36 glslang: 11.10.0-16-g42398c93 -> 11.12.0-32-gf74f37c5 (#2507)
* 3e312b30 Improvements to NVRTC diagnostic parsing (#2504)
* 0b05fe33 Fix inlining pass. (#2506)
* 10834e69 Update nvapi-support.md
* 610851ab Fix document toc. (#2510)
* 66b5351b Update document toc (#2509)
* d5e4854b Update documentation on new features (#2508)
* 004f6e30 Add `[ForwardDerivativeOf]` attribute. (#2501)

v0.24.42

09 Nov 20:38
cedd936
Compare
Choose a tag to compare
Fix supported shader model check in D3D12 device (#2503)

* Fix supported shader model check in D3D12 device

* Add comment on shader model list order.

Co-authored-by: jsmall-nvidia <jsmall@nvidia.com>

v0.24.41

09 Nov 20:16
e743ddd
Compare
Choose a tag to compare
f32tof16 and f16tof32 support for CPU targets (#2500)

* #include an absolute path didn't work - because paths were taken to always be relative.

* Float16 support for C++/CPU based targets with f16tof32 and f32tof16.

* Small correction around INF/NAN handling for f32tof16

* Small improvement to f16tof32

* Disable CUDA test for now.

v0.24.40

09 Nov 19:58
e743ddd
Compare
Choose a tag to compare
f32tof16 and f16tof32 support for CPU targets (#2500)

* #include an absolute path didn't work - because paths were taken to always be relative.

* Float16 support for C++/CPU based targets with f16tof32 and f32tof16.

* Small correction around INF/NAN handling for f32tof16

* Small improvement to f16tof32

* Disable CUDA test for now.

v0.24.39

07 Nov 17:14
ea99c27
Compare
Choose a tag to compare
Initial version of DeviceLimits implemented in d3d12, d3d11, vulkan a…

v0.24.38

30 Oct 18:37
883d9f7
Compare
Choose a tag to compare
Slang v0.24.38

Changes:

883d9f7c Shader cache bugfixes and test additions (#2467)
5a510ab2 Fix  language server crash on incomplete higher order invoke expr. (#2476)
f9b1c565 More renaming in jvp pass. (#2475)