[MTIA ATen Backend][1/n] Migrate "_unsafe_view" and "view" ops from out-of-tree to pytorch in-tree #153670

andyanwang · 2025-05-15T22:54:14Z

Summary:

Context

The MTIA New Aten Backend work is essentially to move MTIA operators from pytorch out-of-tree to in-tree, with following benefits:

Avoid duplicate code copied from pytorch, e.g. view ops implementation, util functions.
Utilize TensorIterator and structured kernel codegen, avoid manual implementation of broadcasting, dtype casting, asserting, etc.
Eliminate MTIA's own codegen flow, which is unnecessary complexity.
Overall make MTIA's aten backend more pytorch native.

Differential Revision: D74672464

cc @egienvalue

pytorch-bot · 2025-05-15T22:54:16Z

This appears to be a diff that was exported from phabricator, but the PR author does not have sufficient permissions to run CI. @andyanwang, please do step 2 of internal wiki to get write access so you do not need to get CI approvals in the future. If you think this is a mistake, please contact the Pytorch Dev Infra team.

linux-foundation-easycla · 2025-05-15T22:54:18Z

The committers listed above are authorized under a signed CLA.

✅ login: andyanwang (7dbef6f)

pytorch-bot · 2025-05-15T22:54:20Z

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/153670

📄 Preview Python docs built from this PR
📄 Preview C++ docs built from this PR
❓ Need help or want to give feedback on the CI? Visit the bot commands wiki or our office hours

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

✅ No Failures

As of commit 7dbef6f with merge base 6f835a4 ():
💚 Looks good so far! There are no failures yet. 💚

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

facebook-github-bot · 2025-05-15T22:54:24Z

This pull request was exported from Phabricator. Differential Revision: D74672464

…-tree to pytorch in-tree (pytorch#153670) Summary: # Context The MTIA New Aten Backend work is essentially to move MTIA operators from pytorch out-of-tree to in-tree, with following benefits: 1. Avoid duplicate code copied from pytorch, e.g. view ops implementation, util functions. 2. Utilize TensorIterator and structured kernel codegen, avoid manual implementation of broadcasting, dtype casting, asserting, etc. 3. Eliminate MTIA's own codegen flow, which is unnecessary complexity. 4. Overall make MTIA's aten backend more pytorch native. Differential Revision: D74672464

facebook-github-bot · 2025-05-15T22:56:39Z

This pull request was exported from Phabricator. Differential Revision: D74672464

github-actions · 2025-05-15T22:58:00Z

Attention! native_functions.yaml was changed

If you are adding a new function or defaulted argument to native_functions.yaml, you cannot use it from pre-existing Python frontend code until our FC window passes (two weeks). Split your PR into two PRs, one which adds the new C++ functionality, and one that makes use of it from Python, and land them two weeks apart. See https://github.com/pytorch/pytorch/wiki/PyTorch's-Python-Frontend-Backward-and-Forward-Compatibility-Policy#forwards-compatibility-fc for more info.

Caused by:

aten/src/ATen/native/native_functions.yaml

pytorch-bot · 2025-05-16T17:15:31Z

❌ 🤖 pytorchbot command failed:

Got EOF while in a quoted string```
Try `@pytorchbot --help` for more info.

andyanwang · 2025-05-16T17:16:16Z

@pytorchbot label "topic: not user facing"

facebook-github-bot · 2025-05-16T17:25:01Z

This pull request was exported from Phabricator. Differential Revision: D74672464

…-tree to pytorch in-tree (pytorch#153670) Summary: # Context The MTIA New Aten Backend work is essentially to move MTIA operators from pytorch out-of-tree to in-tree, with following benefits: 1. Avoid duplicate code copied from pytorch, e.g. view ops implementation, util functions. 2. Utilize TensorIterator and structured kernel codegen, avoid manual implementation of broadcasting, dtype casting, asserting, etc. 3. Eliminate MTIA's own codegen flow, which is unnecessary complexity. 4. Overall make MTIA's aten backend more pytorch native. Differential Revision: D74672464

facebook-github-bot · 2025-05-19T04:26:18Z

This pull request was exported from Phabricator. Differential Revision: D74672464

…rward" # Context See the first PR #153670 # This diff Migrate log_sigmoid_forward to in-tree. Differential Revision: [D77191668](https://our.internmc.facebook.com/intern/diff/D77191668/) cc egienvalue [ghstack-poisoned]

# Context See the first PR #153670 # This diff Migrate log_sigmoid_forward to in-tree. Differential Revision: [D77191668](https://our.internmc.facebook.com/intern/diff/D77191668/) cc egienvalue [ghstack-poisoned]

… / eq.Scalar_out # Context See the first PR #153670 # This diff Migrate addcdiv.out / addcmul.out / eq.Tensor_out / eq.Scalar_out to in-tree. Differential Revision: [D78568103](https://our.internmc.facebook.com/intern/diff/D78568103/) [ghstack-poisoned]

# Context See the first PR #153670 # This diff Migrate addmm.out / baddbmm.out / bmm.out to in-tree. Differential Revision: [D78578483](https://our.internmc.facebook.com/intern/diff/D78578483/) [ghstack-poisoned]

# Context See the first PR #153670 # This diff Migrate addmm.out / baddbmm.out / bmm.out to in-tree. Differential Revision: [D78578483](https://our.internmc.facebook.com/intern/diff/D78578483/) ghstack-source-id: 297160102 Pull Request resolved: #158749

Pull Request resolved: #158749 # Context See the first PR #153670 # This diff Migrate addmm.out / baddbmm.out / bmm.out to in-tree. Differential Revision: [D78578483](https://our.internmc.facebook.com/intern/diff/D78578483/) ghstack-source-id: 297792774

…addcmul.out / eq.Tensor_out / eq.Scalar_out" # Context See the first PR #153670 # This diff Migrate addcdiv.out / addcmul.out / eq.Tensor_out / eq.Scalar_out to in-tree. Differential Revision: [D78568103](https://our.internmc.facebook.com/intern/diff/D78568103/) cc egienvalue [ghstack-poisoned]

….Tensor_out / eq.Scalar_out" # Context See the first PR #153670 # This diff Migrate addcdiv.out / addcmul.out / eq.Tensor_out / eq.Scalar_out to in-tree. Differential Revision: [D78568103](https://our.internmc.facebook.com/intern/diff/D78568103/) cc egienvalue [ghstack-poisoned]

…ddbmm.out / bmm.out" # Context See the first PR #153670 # This diff Migrate addmm.out / baddbmm.out / bmm.out to in-tree. Differential Revision: [D78578483](https://our.internmc.facebook.com/intern/diff/D78578483/) cc egienvalue [ghstack-poisoned]

…out" # Context See the first PR #153670 # This diff Migrate addmm.out / baddbmm.out / bmm.out to in-tree. Differential Revision: [D78578483](https://our.internmc.facebook.com/intern/diff/D78578483/) cc egienvalue [ghstack-poisoned]

… / eq.Scalar_out (#158748) # Context See the first PR #153670 # This diff Migrate addcdiv.out / addcmul.out / eq.Tensor_out / eq.Scalar_out to in-tree. Differential Revision: [D78568103](https://our.internmc.facebook.com/intern/diff/D78568103/) Pull Request resolved: #158748 Approved by: https://github.com/albanD, https://github.com/nautsimon

# Context See the first PR #153670 # This diff Migrate addmm.out / baddbmm.out / bmm.out to in-tree. Differential Revision: [D78578483](https://our.internmc.facebook.com/intern/diff/D78578483/) Pull Request resolved: #158749 Approved by: https://github.com/albanD, https://github.com/nautsimon ghstack dependencies: #158748

# Context See the first PR #153670 # This diff Migrate all foreach operators to in-tree, including: - _foreach_abs - _foreach_abs_ - _foreach_add.List - _foreach_add_.List - _foreach_add_.Scalar - _foreach_add_.Tensor - _foreach_addcmul.Scalar - _foreach_addcmul_.Scalar - _foreach_copy - _foreach_copy_ - _foreach_mul.List - _foreach_mul_.List - _foreach_mul_.Scalar - _foreach_mul.Tensor - _foreach_mul_.Tensor - _foreach_norm.Scalar - _foreach_sqrt_ Differential Revision: [D78913847](https://our.internmc.facebook.com/intern/diff/D78913847/) cc egienvalue [ghstack-poisoned]

Pull Request resolved: #159098 # Context See the first PR #153670 # This diff Migrate all foreach operators to in-tree, including: - _foreach_abs - _foreach_abs_ - _foreach_add.List - _foreach_add_.List - _foreach_add_.Scalar - _foreach_add_.Tensor - _foreach_addcmul.Scalar - _foreach_addcmul_.Scalar - _foreach_copy - _foreach_copy_ - _foreach_mul.List - _foreach_mul_.List - _foreach_mul_.Scalar - _foreach_mul.Tensor - _foreach_mul_.Tensor - _foreach_norm.Scalar - _foreach_sqrt_ ghstack-source-id: 298446017 @exported-using-ghexport Differential Revision: [D78913847](https://our.internmc.facebook.com/intern/diff/D78913847/)

# Context See the first PR #153670 # This diff Migrate all foreach operators to in-tree, including: - _foreach_abs - _foreach_abs_ - _foreach_add.List - _foreach_add_.List - _foreach_add_.Scalar - _foreach_add_.Tensor - _foreach_addcmul.Scalar - _foreach_addcmul_.Scalar - _foreach_copy - _foreach_copy_ - _foreach_mul.List - _foreach_mul_.List - _foreach_mul_.Scalar - _foreach_mul.Tensor - _foreach_mul_.Tensor - _foreach_norm.Scalar - _foreach_sqrt_ Differential Revision: [D78913847](https://our.internmc.facebook.com/intern/diff/D78913847/) cc egienvalue [ghstack-poisoned]

Pull Request resolved: #159098 # Context See the first PR #153670 # This diff Migrate all foreach operators to in-tree, including: - _foreach_abs - _foreach_abs_ - _foreach_add.List - _foreach_add_.List - _foreach_add_.Scalar - _foreach_add_.Tensor - _foreach_addcmul.Scalar - _foreach_addcmul_.Scalar - _foreach_copy - _foreach_copy_ - _foreach_mul.List - _foreach_mul_.List - _foreach_mul_.Scalar - _foreach_mul.Tensor - _foreach_mul_.Tensor - _foreach_norm.Scalar - _foreach_sqrt_ ghstack-source-id: 298481964 @exported-using-ghexport Differential Revision: [D78913847](https://our.internmc.facebook.com/intern/diff/D78913847/)

# Context See the first PR #153670 # This diff Migrate all foreach operators to in-tree, including: - _foreach_abs - _foreach_abs_ - _foreach_add.List - _foreach_add_.List - _foreach_add_.Scalar - _foreach_add_.Tensor - _foreach_addcmul.Scalar - _foreach_addcmul_.Scalar - _foreach_copy - _foreach_copy_ - _foreach_mul.List - _foreach_mul_.List - _foreach_mul_.Scalar - _foreach_mul.Tensor - _foreach_mul_.Tensor - _foreach_norm.Scalar - _foreach_sqrt_ Differential Revision: [D78913847](https://our.internmc.facebook.com/intern/diff/D78913847/) cc egienvalue [ghstack-poisoned]

Pull Request resolved: #159540 # Context See the first PR #153670 # This diff Migrate arange.start_out to in-tree. We can use this diff as an example for non-structured operator. ghstack-source-id: 299962562 @exported-using-ghexport Differential Revision: [D79317519](https://our.internmc.facebook.com/intern/diff/D79317519/)

# Context See the first PR #153670 # This diff Migrate all foreach operators to in-tree, including: - _foreach_abs - _foreach_abs_ - _foreach_add.List - _foreach_add_.List - _foreach_add_.Scalar - _foreach_add_.Tensor - _foreach_addcmul.Scalar - _foreach_addcmul_.Scalar - _foreach_copy - _foreach_copy_ - _foreach_mul.List - _foreach_mul_.List - _foreach_mul_.Scalar - _foreach_mul.Tensor - _foreach_mul_.Tensor - _foreach_norm.Scalar - _foreach_sqrt_ Differential Revision: [D78913847](https://our.internmc.facebook.com/intern/diff/D78913847/) cc egienvalue [ghstack-poisoned]

# Context See the first PR #153670 # This diff Migrate all foreach operators to in-tree, including: - _foreach_abs - _foreach_abs_ - _foreach_add.List - _foreach_add_.List - _foreach_add_.Scalar - _foreach_add_.Tensor - _foreach_addcmul.Scalar - _foreach_addcmul_.Scalar - _foreach_copy - _foreach_copy_ - _foreach_mul.List - _foreach_mul_.List - _foreach_mul_.Scalar - _foreach_mul.Tensor - _foreach_mul_.Tensor - _foreach_norm.Scalar - _foreach_sqrt_ Differential Revision: [D78913847](https://our.internmc.facebook.com/intern/diff/D78913847/) Pull Request resolved: #159098 Approved by: https://github.com/malfet

Pull Request resolved: #159540 # Context See the first PR #153670 # This diff Migrate arange.start_out to in-tree. We can use this diff as an example for non-structured operator. ghstack-source-id: 300416010 @exported-using-ghexport Differential Revision: [D79317519](https://our.internmc.facebook.com/intern/diff/D79317519/)

…xample diff for tutorial)" Differential Revision: [D79524919](https://our.internmc.facebook.com/intern/diff/D79524919/) # Context See the first PR #153670 # This diff Migrate any.all_out to in-tree. We can use this diff as an example for in-tree registration. [ghstack-poisoned]

Pull Request resolved: #159780 # Context See the first PR #153670 # This diff Migrate any.all_out to in-tree. We can use this diff as an example for in-tree registration. Differential Revision: [D79524919](https://our.internmc.facebook.com/intern/diff/D79524919/) ghstack-source-id: 300591646

…tutorial)" Differential Revision: [D79524919](https://our.internmc.facebook.com/intern/diff/D79524919/) # Context See the first PR #153670 # This diff Migrate any.all_out to in-tree. We can use this diff as an example for in-tree registration. [ghstack-poisoned]

facebook-github-bot added the fb-exported label May 15, 2025

andyanwang force-pushed the export-D74672464 branch from 5bb6aca to 8b59023 Compare May 15, 2025 22:56

pytorch-bot bot added the topic: not user facing topic category label May 16, 2025

andyanwang force-pushed the export-D74672464 branch from 8b59023 to 44086b6 Compare May 16, 2025 17:24

andyanwang force-pushed the export-D74672464 branch from 44086b6 to 7dbef6f Compare May 19, 2025 04:26

nautsimon requested review from malfet and nautsimon May 19, 2025 15:13

nautsimon added the module: mtia Device MTIA related issues label May 19, 2025

nautsimon requested review from egienvalue and albanD May 19, 2025 15:18

nautsimon changed the title ~~[MTIA Aten Backend] Migrate "_unsafe_view" and "view" ops from out-of-tree to pytorch in-tree~~ [MTIA ATen Backend] Migrate "_unsafe_view" and "view" ops from out-of-tree to pytorch in-tree May 19, 2025

pytorch-bot bot added the ciflow/trunk Trigger trunk jobs on your pull request label May 19, 2025

nautsimon requested review from leafs1, swolchok, hl475 and etaf and removed request for leafs1 May 20, 2025 21:22

andyanwang mentioned this pull request Jul 21, 2025

[MTIA Aten Backend] Migrate addcdiv.out / addcmul.out / eq.Tensor_out / eq.Scalar_out #158748

Closed

andyanwang mentioned this pull request Jul 21, 2025

[MTIA Aten Backend] Migrate addmm.out / baddbmm.out / bmm.out #158749

Closed

andyanwang mentioned this pull request Jul 24, 2025

[MTIA Aten Backend] Migrate all foreach ops #159098

Closed

andyanwang mentioned this pull request Aug 4, 2025

[MTIA Aten Backend] Migrate any.all_out (example diff for tutorial) #159780

Open

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

[MTIA ATen Backend][1/n] Migrate "_unsafe_view" and "view" ops from out-of-tree to pytorch in-tree #153670

[MTIA ATen Backend][1/n] Migrate "_unsafe_view" and "view" ops from out-of-tree to pytorch in-tree #153670

Uh oh!

andyanwang commented May 15, 2025 •

edited by pytorch-bot bot

Loading

Uh oh!

pytorch-bot bot commented May 15, 2025

Uh oh!

linux-foundation-easycla bot commented May 15, 2025 •

edited

Loading

Uh oh!

pytorch-bot bot commented May 15, 2025 •

edited

Loading

Uh oh!

facebook-github-bot commented May 15, 2025

Uh oh!

facebook-github-bot commented May 15, 2025

Uh oh!

github-actions bot commented May 15, 2025

Uh oh!

pytorch-bot bot commented May 16, 2025

Uh oh!

andyanwang commented May 16, 2025

Uh oh!

facebook-github-bot commented May 16, 2025

Uh oh!

facebook-github-bot commented May 19, 2025

Uh oh!

Uh oh!

[MTIA ATen Backend][1/n] Migrate "_unsafe_view" and "view" ops from out-of-tree to pytorch in-tree #153670

[MTIA ATen Backend][1/n] Migrate "_unsafe_view" and "view" ops from out-of-tree to pytorch in-tree #153670

Uh oh!

Conversation

andyanwang commented May 15, 2025 • edited by pytorch-bot bot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Context

Uh oh!

pytorch-bot bot commented May 15, 2025

Uh oh!

linux-foundation-easycla bot commented May 15, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

pytorch-bot bot commented May 15, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/153670

✅ No Failures

Uh oh!

facebook-github-bot commented May 15, 2025

Uh oh!

facebook-github-bot commented May 15, 2025

Uh oh!

github-actions bot commented May 15, 2025

Attention! native_functions.yaml was changed

Uh oh!

pytorch-bot bot commented May 16, 2025

Uh oh!

andyanwang commented May 16, 2025

Uh oh!

facebook-github-bot commented May 16, 2025

Uh oh!

facebook-github-bot commented May 19, 2025

Uh oh!

Uh oh!

andyanwang commented May 15, 2025 •

edited by pytorch-bot bot

Loading

linux-foundation-easycla bot commented May 15, 2025 •

edited

Loading

pytorch-bot bot commented May 15, 2025 •

edited

Loading