[ET-VK] Replacing the use of uvec3 with WorkgroupSize class to reduce memory usage and improve processing speed #8634
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
… memory usage and improve processing speed This diff replaces the use of `uvec3` with `WorkgroupSize` class to reduce memory usage and improve processing speed in the Vulkan backend of Executorch. Differential Revision: [D70021032](https://our.internmc.facebook.com/intern/diff/D70021032/) [ghstack-poisoned]
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/8634
Note: Links to docs will display an error until the docs builds have been completed. ✅ No FailuresAs of commit ba10936 with merge base 728c255 ( This comment was automatically generated by Dr. CI and updates every 15 minutes. |
facebook-github-bot
added
the
CLA Signed
|
This pull request was exported from Phabricator. Differential Revision: D70021032 |
This was referenced Feb 22, 2025
trivedivivek
added a commit
that referenced
this pull request
Feb 22, 2025
… memory usage and improve processing speed This diff replaces the use of `uvec3` with `WorkgroupSize` class to reduce memory usage and improve processing speed in the Vulkan backend of Executorch. Differential Revision: [D70021032](https://our.internmc.facebook.com/intern/diff/D70021032/) ghstack-source-id: 267791087 Pull Request resolved: #8634
…s to reduce memory usage and improve processing speed" This diff replaces the use of `uvec3` with `WorkgroupSize` class to reduce memory usage and improve processing speed in the Vulkan backend of Executorch. Differential Revision: [D70021032](https://our.internmc.facebook.com/intern/diff/D70021032/) [ghstack-poisoned]
|
This pull request was exported from Phabricator. Differential Revision: D70021032 |
trivedivivek
added a commit
that referenced
this pull request
Feb 22, 2025
… memory usage and improve processing speed Pull Request resolved: #8634 This diff replaces the use of `uvec3` with `WorkgroupSize` class to reduce memory usage and improve processing speed in the Vulkan backend of Executorch. ghstack-source-id: 267792195 @exported-using-ghexport Differential Revision: [D70021032](https://our.internmc.facebook.com/intern/diff/D70021032/)
trivedivivek
changed the title
nathanaelsee
approved these changes
Feb 24, 2025
…s to reduce memory usage and improve processing speed" This diff replaces the use of `uvec3` with `WorkgroupSize` class to reduce memory usage and improve processing speed in the Vulkan backend of Executorch. Differential Revision: [D70021032](https://our.internmc.facebook.com/intern/diff/D70021032/) [ghstack-poisoned]
|
This pull request was exported from Phabricator. Differential Revision: D70021032 |
This was referenced Feb 24, 2025
…s to reduce memory usage and improve processing speed" This diff replaces the use of `uvec3` with `WorkgroupSize` class to reduce memory usage and improve processing speed in the Vulkan backend of Executorch. Differential Revision: [D70021032](https://our.internmc.facebook.com/intern/diff/D70021032/) [ghstack-poisoned]
|
This pull request was exported from Phabricator. Differential Revision: D70021032 |
…s to reduce memory usage and improve processing speed" This diff replaces the use of `uvec3` with `WorkgroupSize` class to reduce memory usage and improve processing speed in the Vulkan backend of Executorch. Differential Revision: [D70021032](https://our.internmc.facebook.com/intern/diff/D70021032/) [ghstack-poisoned]
|
This pull request was exported from Phabricator. Differential Revision: D70021032 |
trivedivivek
added a commit
that referenced
this pull request
Feb 25, 2025
… memory usage and improve processing speed Pull Request resolved: #8634 This diff replaces the use of `uvec3` with `WorkgroupSize` class to reduce memory usage and improve processing speed in the Vulkan backend of Executorch. ghstack-source-id: 268172660 @exported-using-ghexport Differential Revision: [D70021032](https://our.internmc.facebook.com/intern/diff/D70021032/)
facebook-github-bot
merged commit d9451ca
into
gh/trivedivivek/56/base
50 checks passed
facebook-github-bot
temporarily deployed
to
cherry-pick-bot
GitHub Actions
Inactive
trivedivivek
added a commit
that referenced
this pull request
Feb 25, 2025
… memory usage and improve processing speed (#8671) * [ET-VK] Adding a workgroup class to VecUtils Pull Request resolved: #8632 This diff adds a new class called `WorkgroupSize` to the `VecUtils` header file. The `WorkgroupSize` class takes three `uint32_t` values as parameters and stores them in a single `uint32_t` variable using bitwise operations. This class is used in the Vulkan backend to specify the size of a workgroup for a given operation. ghstack-source-id: 268172661 @exported-using-ghexport Differential Revision: [D70021019](https://our.internmc.facebook.com/intern/diff/D70021019/) * [ET-VK] Adding reserve and append functions to SpecVarList Pull Request resolved: #8633 This diff adds two new functions to the SpecVarList class in the Vulkan runtime library. The first function, reserve, allows the user to reserve a certain amount of space in the SpecVarList before adding any elements. The second function, append, allows the user to add a single SpecVar to the SpecVarList. These functions are useful for optimizing memory usage and improving performance in the Vulkan runtime. ghstack-source-id: 268172659 @exported-using-ghexport Differential Revision: [D70021782](https://our.internmc.facebook.com/intern/diff/D70021782/) * [ET-VK] Replacing the use of uvec3 with WorkgroupSize class to reduce memory usage and improve processing speed Pull Request resolved: #8634 This diff replaces the use of `uvec3` with `WorkgroupSize` class to reduce memory usage and improve processing speed in the Vulkan backend of Executorch. ghstack-source-id: 268172660 @exported-using-ghexport Differential Revision: [D70021032](https://our.internmc.facebook.com/intern/diff/D70021032/) --------- Co-authored-by: Vivek Trivedi <[email protected]>
SS-JIA
added a commit
that referenced
this pull request
Mar 19, 2025
…line` ## Context Fix changes introduced in #8634 (D70021032). This change decoupled local work group size from additional specialization constants. As part of this change, when creating a `VkComputePipeline` a temporary `SpecVarList` is created to merge the `WorkgroupSize` with additional specialization constants. However, this can be an issue with some Vulkan drivers because the `SpecVarList` is a temporary, and thus will be destroyed at the end of the function call. The pointer stored in the `VkSpecializationInfo` will be invalidated, leading to undefined behaviour. This diff fixes this by restoring the behaviour of `ComputePipeline::Descriptor` storing the `specialization_constants` directly. Also fix the fact that the `VkSpecializationMapEntry` vector was also a temporary when creating a `VkComputePipeline` by storing it in `ComputePipeline`. Differential Revision: [D71488015](https://our.internmc.facebook.com/intern/diff/D71488015/) [ghstack-poisoned]
SS-JIA
added a commit
that referenced
this pull request
Mar 19, 2025
…line` ## Context Fix changes introduced in #8634 (D70021032). This change decoupled local work group size from additional specialization constants. As part of this change, when creating a `VkComputePipeline` a temporary `SpecVarList` is created to merge the `WorkgroupSize` with additional specialization constants. However, this can be an issue with some Vulkan drivers because the `SpecVarList` is a temporary, and thus will be destroyed at the end of the function call. The pointer stored in the `VkSpecializationInfo` will be invalidated, leading to undefined behaviour. This diff fixes this by restoring the behaviour of `ComputePipeline::Descriptor` storing the `specialization_constants` directly. Also fix the fact that the `VkSpecializationMapEntry` vector was also a temporary when creating a `VkComputePipeline` by storing it in `ComputePipeline`. Differential Revision: [D71488015](https://our.internmc.facebook.com/intern/diff/D71488015/) ghstack-source-id: 272783852 Pull Request resolved: #9405
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.
Stack from ghstack (oldest at bottom):
This diff replaces the use of
uvec3withWorkgroupSizeclass to reduce memory usage and improve processing speed in the Vulkan backend of Executorch.Differential Revision: D70021032