[AOTI][refactor] Move a few util functions in atoi_torch #119987
Conversation
Summary: Move these util functions from an anonymous namespace to a common header so that later torchgen-ed files can use them. [ghstack-poisoned]
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/119987
Note: Links to docs will display an error until the docs builds have been completed. ✅ No FailuresAs of commit bb830b9 with merge base 9c597ff ( This comment was automatically generated by Dr. CI and updates every 15 minutes. |
Summary: Move these util functions from an anonymous namespace to a common header so that later torchgen-ed files can use them. [ghstack-poisoned]
Summary: Move these util functions from an anonymous namespace to a common header so that later torchgen-ed files can use them. [ghstack-poisoned]
Summary: Move these util functions from an anonymous namespace to a common header so that later torchgen-ed files can use them. [ghstack-poisoned]
|
@desertfire has imported this pull request. If you are a Meta employee, you can view this diff on Phabricator. |
Summary: Move these util functions from an anonymous namespace to a common header so that later torchgen-ed files can use them. Differential Revision: [D54258088](https://our.internmc.facebook.com/intern/diff/D54258088) [ghstack-poisoned]
|
@desertfire has imported this pull request. If you are a Meta employee, you can view this diff on Phabricator. |
Summary: Move these util functions from an anonymous namespace to a common header so that later torchgen-ed files can use them. Differential Revision: [D54258088](https://our.internmc.facebook.com/intern/diff/D54258088) [ghstack-poisoned]
|
@desertfire has imported this pull request. If you are a Meta employee, you can view this diff on Phabricator. |
|
@desertfire has imported this pull request. If you are a Meta employee, you can view this diff on Phabricator. |
2 similar comments
|
@desertfire has imported this pull request. If you are a Meta employee, you can view this diff on Phabricator. |
|
@desertfire has imported this pull request. If you are a Meta employee, you can view this diff on Phabricator. |
|
@pytorchbot merge -f 'Landed internally' (Initiating merge automatically since Phabricator Diff has merged, using force because this PR might not pass merge_rules.json but landed internally) |
Merge startedYour change will be merged immediately since you used the force (-f) flag, bypassing any CI checks (ETA: 1-5 minutes). Please use Learn more about merging in the wiki. Questions? Feedback? Please reach out to the PyTorch DevX Team |
Summary: Switch codegen for sdpa to always point to v2 in the C shim. Since aoti_torch__scaled_dot_product_flash_attention_v2 has been introduced for a while, there shouldn't be any FC issue in production. Differential Revision: [D54258090](https://our.internmc.facebook.com/intern/diff/D54258090) Pull Request resolved: #120592 Approved by: https://github.com/chenyang78 ghstack dependencies: #119987
Summary: Currently the logics for filling the default value for optional arguments are scattered in several places. By storing OpOverload in the base ExternKernel class, we can simplify codegen_kwargs, and this is a preparation step for enabling the torchgen-ed C shim. The default value filling logic for FallbackKernel can also be simplified, but that can come later. Differential Revision: [D54258089](https://our.internmc.facebook.com/intern/diff/D54258089) Pull Request resolved: #120629 Approved by: https://github.com/chenyang78 ghstack dependencies: #119987, #120592
Stack from ghstack (oldest at bottom):
Summary: Move these util functions from an anonymous namespace to a common header so that later torchgen-ed files can use them.
Differential Revision: D54258088