[ET-VK] Introduce generalized shaders for transfer ops and use it for select and slice #11304
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Pull Request resolved: #11253 ## Changes * Call `encode_execute()` upon resize in `propagate_resize()` * Minor update to `DispatchNode` to store push constant data array as a persistent member of the class ## Motivation Passing in tensor metadata (i.e. sizes, strides) via push constants is typically more performant than passing them via a UBO (uniform buffer object). However, currently dynamic shapes do not work when push constants are used as I realized that the tensor metadata contained in the push constants do not get updated. It appears that that `vkCmdPushConstants` sets the push constants when encoding the command buffer, however the push constants will not be updated if the command buffer is submitted for execution multiple times. Therefore, to update push constant values **the command buffer needs to be re-encoded**. ## Performance Impact This may add a small performance overhead (i.e. re-encoding the command buffer) when executing models with dynamic shapes. Models that do not trigger tensor resizing will not be impacted. However, I measured the impact on a llama 3.2 1B model and the impact of re-encoding a command buffer appears to be negligible. In any case, re-encoding the command buffer is a "necessary evil" when working with dynamic shapes, otherwise the tensor metadata seen by shaders may never get updated. Furthermore, re-encoding the command buffer can allow an opportunity to adjust global work group sizing to match current tensor sizes, which may have a huge performance impact when maximum tensor sizes far exceeds what tensor sizes will realistically be during inference (one instance of this is for transformer models when the max sequence length is very long). ghstack-source-id: 287711101 @exported-using-ghexport Differential Revision: [D75686051](https://our.internmc.facebook.com/intern/diff/D75686051/)
Pull Request resolved: #11254 ## Changes For `DynamicDispatchNode`: * Pass in global work group size to the local work group size determination function * Add additional constructor for which the shader is not dynamic * During `encode`, check that pick functions are not `nullptr` ## Motivation Oftentimes it is useful to know what the global work group size is when determining what the local group group size should be. ## Performance Impact None. ghstack-source-id: 287711100 @exported-using-ghexport Differential Revision: [D75686047](https://our.internmc.facebook.com/intern/diff/D75686047/)
… select and slice Pull Request resolved: #11255 ## Changes * Introduce `transfer_buffer.glsl` and `transfer_texture.glsl`, and `Transfer.cpp` which generalizes shaders where each element of the output is copied from a unique element of the input. * Update `Slice.cpp` and `Select.cpp` to use `Transfer.cpp` * Remove old implementations of slice and select ## Motivation With this new implementation, the op can now support both buffers and textures of any packing. There are also benefits of code consolidation. ghstack-source-id: 287711104 @exported-using-ghexport Differential Revision: [D75686050](https://our.internmc.facebook.com/intern/diff/D75686050/)
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/11304
Note: Links to docs will display an error until the docs builds have been completed. This comment was automatically generated by Dr. CI and updates every 15 minutes. |
facebook-github-bot
added
the
CLA Signed
SS-JIA
approved these changes
Jun 3, 2025
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This PR was created by the merge bot to help merge the original PR into the main branch.
ghstack PR number: #11255 by @SS-JIA
^ Please use this as the source of truth for the PR details, comments, and reviews
ghstack PR base: https://github.com/pytorch/executorch/tree/gh/SS-JIA/231/base
ghstack PR head: https://github.com/pytorch/executorch/tree/gh/SS-JIA/231/head
Merge bot PR base: https://github.com/pytorch/executorch/tree/gh/SS-JIA/230/orig
Merge bot PR head: https://github.com/pytorch/executorch/tree/gh/SS-JIA/231/orig
@diff-train-skip-merge