[ET-VK][ez][Refactor] Re-order DispatchNode arguments to match shader layout spec
#10700
+316
−167
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
…er layout spec
## Context
As title. Note that this PR was written mainly with a meta internal coding agent. This PR re-orders the arguments to `DispatchNode` to be more intuitive.
I asked the agent:
---
I want to change the constructor of the class
```
explicit DispatchNode(
ComputeGraph& graph,
const vkapi::ShaderInfo& shader,
const utils::uvec3& global_workgroup_size,
const utils::uvec3& local_workgroup_size,
const std::vector<ArgGroup>& args,
const vkapi::ParamsBindList& params,
const vkapi::SpecVarList& spec_vars = {},
const ResizeFunction& resize_fn = nullptr,
const std::vector<ValueRef>& resize_args = {},
const std::vector<PushConstantDataInfo>& push_constants = {});
```
to instead be
```
explicit DispatchNode(
ComputeGraph& graph,
const vkapi::ShaderInfo& shader,
const utils::uvec3& global_workgroup_size,
const utils::uvec3& local_workgroup_size,
const std::vector<ArgGroup>& args,
const vkapi::ParamsBindList& params,,
const std::vector<PushConstantDataInfo>& push_constants = {},
const vkapi::SpecVarList& spec_vars = {},
const std::vector<ValueRef>& resize_args = {},
const ResizeFunction& resize_fn = nullptr);
```
Can you make this change and update the callsites as well?
---
The motivation is to have the arguments match the order in which parameter UBOs, push constant blocks, and specialization variables are declared in a GLSL shader.
The order of `resize`_args` and `resize_fn` was also swapped in the interest of having the function pointer be the last argument. It will also make more sense in a following diff where a `DynamicDispatchNode` class is introduced, which will allow selecting a different compute shader depending on input sizes.
As a small additional change, I also asked the agent to
---
Go through all the files under `xplat/executorch/backends/vulkan/runtime/graph/ops/impl`
Change `vkapi::MemoryAccessType::WRITE` to `vkapi::kWrite` and `vkapi::MemoryAccessType::READ` to `vkapi::kRead` for each file.
---
Differential Revision: [D74203482](https://our.internmc.facebook.com/intern/diff/D74203482/)
ghstack-source-id: 282122344
Pull Request resolved: #10693
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/10700
Note: Links to docs will display an error until the docs builds have been completed. ❗ 1 Active SEVsThere are 1 currently active SEVs. If your PR is affected, please view them below: ❌ 1 New Failure, 30 PendingAs of commit 3e87524 with merge base 578358b ( NEW FAILURE - The following job has failed:
This comment was automatically generated by Dr. CI and updates every 15 minutes. |
facebook-github-bot
added
the
CLA Signed
This PR needs a
|
SS-JIA
approved these changes
May 5, 2025
phaiting
pushed a commit
that referenced
this pull request
May 6, 2025
…er layout spec (#10700) ## Context As title. Note that this PR was written mainly with a meta internal coding agent. This PR re-orders the arguments to `DispatchNode` to be more intuitive. I asked the agent: --- I want to change the constructor of the class ``` explicit DispatchNode( ComputeGraph& graph, const vkapi::ShaderInfo& shader, const utils::uvec3& global_workgroup_size, const utils::uvec3& local_workgroup_size, const std::vector<ArgGroup>& args, const vkapi::ParamsBindList& params, const vkapi::SpecVarList& spec_vars = {}, const ResizeFunction& resize_fn = nullptr, const std::vector<ValueRef>& resize_args = {}, const std::vector<PushConstantDataInfo>& push_constants = {}); ``` to instead be ``` explicit DispatchNode( ComputeGraph& graph, const vkapi::ShaderInfo& shader, const utils::uvec3& global_workgroup_size, const utils::uvec3& local_workgroup_size, const std::vector<ArgGroup>& args, const vkapi::ParamsBindList& params,, const std::vector<PushConstantDataInfo>& push_constants = {}, const vkapi::SpecVarList& spec_vars = {}, const std::vector<ValueRef>& resize_args = {}, const ResizeFunction& resize_fn = nullptr); ``` Can you make this change and update the callsites as well? --- The motivation is to have the arguments match the order in which parameter UBOs, push constant blocks, and specialization variables are declared in a GLSL shader. The order of `resize`_args` and `resize_fn` was also swapped in the interest of having the function pointer be the last argument. It will also make more sense in a following diff where a `DynamicDispatchNode` class is introduced, which will allow selecting a different compute shader depending on input sizes. As a small additional change, I also asked the agent to --- Go through all the files under `xplat/executorch/backends/vulkan/runtime/graph/ops/impl` Change `vkapi::MemoryAccessType::WRITE` to `vkapi::kWrite` and `vkapi::MemoryAccessType::READ` to `vkapi::kRead` for each file. --- Differential Revision: [D74203482](https://our.internmc.facebook.com/intern/diff/D74203482/)
jhelsby
pushed a commit
to jhelsby/executorch
that referenced
this pull request
May 9, 2025
…er layout spec (pytorch#10700) ## Context As title. Note that this PR was written mainly with a meta internal coding agent. This PR re-orders the arguments to `DispatchNode` to be more intuitive. I asked the agent: --- I want to change the constructor of the class ``` explicit DispatchNode( ComputeGraph& graph, const vkapi::ShaderInfo& shader, const utils::uvec3& global_workgroup_size, const utils::uvec3& local_workgroup_size, const std::vector<ArgGroup>& args, const vkapi::ParamsBindList& params, const vkapi::SpecVarList& spec_vars = {}, const ResizeFunction& resize_fn = nullptr, const std::vector<ValueRef>& resize_args = {}, const std::vector<PushConstantDataInfo>& push_constants = {}); ``` to instead be ``` explicit DispatchNode( ComputeGraph& graph, const vkapi::ShaderInfo& shader, const utils::uvec3& global_workgroup_size, const utils::uvec3& local_workgroup_size, const std::vector<ArgGroup>& args, const vkapi::ParamsBindList& params,, const std::vector<PushConstantDataInfo>& push_constants = {}, const vkapi::SpecVarList& spec_vars = {}, const std::vector<ValueRef>& resize_args = {}, const ResizeFunction& resize_fn = nullptr); ``` Can you make this change and update the callsites as well? --- The motivation is to have the arguments match the order in which parameter UBOs, push constant blocks, and specialization variables are declared in a GLSL shader. The order of `resize`_args` and `resize_fn` was also swapped in the interest of having the function pointer be the last argument. It will also make more sense in a following diff where a `DynamicDispatchNode` class is introduced, which will allow selecting a different compute shader depending on input sizes. As a small additional change, I also asked the agent to --- Go through all the files under `xplat/executorch/backends/vulkan/runtime/graph/ops/impl` Change `vkapi::MemoryAccessType::WRITE` to `vkapi::kWrite` and `vkapi::MemoryAccessType::READ` to `vkapi::kRead` for each file. --- Differential Revision: [D74203482](https://our.internmc.facebook.com/intern/diff/D74203482/)
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: #10693 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/224/base
ghstack PR head: https://github.com/pytorch/executorch/tree/gh/SS-JIA/224/head
Merge bot PR base: https://github.com/pytorch/executorch/tree/main
Merge bot PR head: https://github.com/pytorch/executorch/tree/gh/SS-JIA/224/orig
@diff-train-skip-merge