Google GitGoogle Git
Sign in
dawn/dawn/848d62464cf951be9a2bb66c3cb80f42e72f89b5
commit
848d62464cf951be9a2bb66c3cb80f42e72f89b5
[log]
author
Natalie Chouinard <chouinard@google.com>
Mon Sep 21 12:17:08 2026 -0700
committer
dawn-scoped@luci-project-accounts.iam.gserviceaccount.com <dawn-scoped@luci-project-accounts.iam.gserviceaccount.com>
Mon Sep 21 12:17:08 2026 -0700
tree
d7a50090d1ab07f2e2825cca4d8c903a664c59fc
parent
f2186f9c463f786e355c792d31edd00b5697ac45 [diff]
[litert-lm] Disable shader caches

The current LiteRT-LM benchmark exhibits bimodal performance in Init
Executor on the Linux GTX 1660, suspected to be due to host GPU driver
shader caches persisting between runs. Although testing the SPIR-V
driver compilation is not particularly relevant for us to measure, it's
so much larger than the tint compilation time that any change to tint's
compilation time affecting Init Executor time would be lost. Still
unlikely we would catch it with this perf test since it's <10% of Init,
but worth monitoring in a more deterministic way anyways. This change
should make a clean cold start to the full shader compilation for all
platforms currently tested, plus preemptively on Mesa.

Bug: 562993169
Change-Id: Ia4f2025adc3af9b19e36a76ab95ab511c6d39aee
Cq-Include-Trybots: luci.dawn.try:dawn-try-linux-x64-nvidia-gtx1660-rel,dawn-try-mac-arm64-apple-m2-rel
Include-Ci-Only-Tests: ci:dawn-linux-x64-nvidia-gtx1660-rel,ci:dawn-mac-arm64-apple-m2-rel|litert_lm_benchmark
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/343075
Commit-Queue: Natalie Chouinard <chouinard@google.com>
Reviewed-by: dan sinclair <dsinclair@chromium.org>
1 file changed
tree: d7a50090d1ab07f2e2825cca4d8c903a664c59fc
  1. .github/
  2. .vscode/
  3. agents/
  4. build_overrides/
  5. docs/
  6. generator/
  7. include/
  8. infra/
  9. scripts/
  10. src/
  11. test/
  12. third_party/
  13. tools/
  14. webgpu-cts/
  15. build
  16. buildtools
  17. testing
  18. .bazelignore
  19. .bazelrc
  20. .bazelversion
  21. .clang-format
  22. .clang-format-ignore
  23. .clang-tidy
  24. .git-blame-ignore-revs
  25. .gitattributes
  26. .gitignore
  27. .gitmodules
  28. .gn
  29. .style.yapf
  30. AUTHORS
  31. BUILD.bazel
  32. BUILD.gn
  33. CMakeLists.txt
  34. CMakeSettings.json
  35. CODE_OF_CONDUCT.md
  36. codereview.settings
  37. CONTRIBUTING.md
  38. CPPLINT.cfg
  39. DEPS
  40. DIR_METADATA
  41. go.mod
  42. go.sum
  43. go_presubmit_support.py
  44. LICENSE
  45. MODULE.bazel
  46. MODULE.bazel.lock
  47. OWNERS
  48. PRESUBMIT.py
  49. PRESUBMIT_test.py
  50. README.chromium
  51. README.md
  52. unsafe_buffers_paths.txt
  53. vpython.toml
  54. vpython.toml.uv.lock
  55. WATCHLISTS
  56. WORKSPACE.bazel
README.md

Build Status Matrix Space

Dawn, a WebGPU implementation

Dawn is an open-source and cross-platform implementation of the WebGPU standard. More precisely it implements webgpu.h that is a one-to-one mapping with the WebGPU IDL. Dawn is meant to be integrated as part of a larger system and is the underlying implementation of WebGPU in Chromium.

Dawn provides several WebGPU building blocks:

  • WebGPU C/C++ headers that applications and other building blocks use.
    • The webgpu.h version that Dawn implements.
    • A C++ wrapper for the webgpu.h.
  • A “native” implementation of WebGPU using platforms' GPU APIs: D3D12, Metal, Vulkan and OpenGL. See per API support for more details.
  • A client-server implementation of WebGPU for applications that are in a sandbox without access to native drivers
  • Tint is a compiler for the WebGPU Shader Language (WGSL) that can be used in standalone to convert shaders from and to WGSL.

Helpful links:

Documentation table of content

Developer documentation:

User documentation: (TODO, figure out what overlaps with the webgpu.h docs)

License

BSD 3-Clause License, please see LICENSE.

Disclaimer

This is not an officially supported Google product.