Skip to content

[ExecuTorch][Vulkan] SDK Integration 3/n: QueryPool With node_id - #3959

Closed
liuk22 wants to merge 1 commit into
gh/liuk22/3/basefrom
gh/liuk22/3/head
Closed

liuk22 wants to merge 1 commit into
gh/liuk22/3/basefrom
gh/liuk22/3/head

Conversation

@liuk22

@liuk22 liuk22 commented Jun 12, 2024 •

Copy link
Copy Markdown
Contributor

Stack from ghstack (oldest at bottom):

See design doc.

3/n: QueryPool with node_id

  • Now we have the full E2E linked of a coherent DDI, serialized AOT to the flatbuffer, deserialized Runtime to ShaderInfo.
  • We add node_id as default-valued argument (UINT32_MAX) to Context::submit_compute_job() which allowsShaderDuration to be constructed later with node_id
  • I modify QueryPool::shader_profile_begin() , Context::report_shader_dispatch_start(), to construct ShaderDuration with the node_id from ExecuteNode

Differential Revision: D57424317

See [design doc](https://docs.google.com/document/d/1KQ0vH0IhZK24sBbym7TYXfYssAgtPjI-htcxraoxEaM/edit).

__3/n: QueryPool with node_id__
- Now we have the full E2E linked of a coherent DDI, serialized AOT to the flatbuffer, deserialized Runtime to `ShaderInfo`.
- We add `node_id` as default-valued argument (`UINT32_MAX`) to `Context::submit_compute_job()` which allows`ShaderDuration` to be constructed later with `node_id`
- I modify `QueryPool::shader_profile_begin()` , `Context::report_shader_dispatch_start()`, to construct `ShaderDuration` with the `node_id` from `ExecuteNode`

Differential Revision: [D57424317](https://our.internmc.facebook.com/intern/diff/D57424317/)

[ghstack-poisoned]
@pytorch-bot

pytorch-bot Bot commented Jun 12, 2024 •

Copy link
Copy Markdown

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/3959

Note: Links to docs will display an error until the docs builds have been completed.

❌ 1 New Failure

As of commit ebc5caa with merge base 10de8b2 (image):

NEW FAILURE - The following job has failed:

This comment was automatically generated by Dr. CI and updates every 15 minutes.

@pytorch-bot pytorch-bot Bot added ciflow/periodic module: vulkan Issues related to the Vulkan delegate and code under backends/vulkan/ labels Jun 12, 2024
@pytorch-bot

pytorch-bot Bot commented Jun 12, 2024

Copy link
Copy Markdown

Warning: Unknown label ciflow/periodic.
Currently recognized labels are

  • ciflow/nightly
  • ciflow/trunk
  • ciflow/binaries
  • ciflow/binaries/all

Please add the new label to .github/pytorch-probot.yml

@facebook-github-bot facebook-github-bot added the CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. label Jun 12, 2024
@facebook-github-bot

Copy link
Copy Markdown
Contributor

This pull request was exported from Phabricator. Differential Revision: D57424317

@facebook-github-bot

Copy link
Copy Markdown
Contributor

This pull request has been merged in 8f3ae8d.

kedarnath03 pushed a commit to kedarnath03/executorch that referenced this pull request Jun 25, 2025
See [design doc](https://docs.google.com/document/d/1KQ0vH0IhZK24sBbym7TYXfYssAgtPjI-htcxraoxEaM/edit).

__3/n: QueryPool with node_id__
- Now we have the full E2E linked of a coherent DDI, serialized AOT to the flatbuffer, deserialized Runtime to `ShaderInfo`.
- We add `node_id` as default-valued argument (`UINT32_MAX`) to `Context::submit_compute_job()` which allows`ShaderDuration` to be constructed later with `node_id`
- I modify `QueryPool::shader_profile_begin()` , `Context::report_shader_dispatch_start()`, to construct `ShaderDuration` with the `node_id` from `ExecuteNode`

Differential Revision: [D57424317](https://our.internmc.facebook.com/intern/diff/D57424317/)

ghstack-source-id: 229964242
Pull Request resolved: pytorch/executorch#3959
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

ciflow/periodic CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. fb-exported Merged module: vulkan Issues related to the Vulkan delegate and code under backends/vulkan/

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants