[c10] Move P2P access logic from ATen to c10#173571
Closed
minsii wants to merge 1 commit intopytorch:mainfrom
Closed
[c10] Move P2P access logic from ATen to c10#173571minsii wants to merge 1 commit intopytorch:mainfrom
minsii wants to merge 1 commit intopytorch:mainfrom
Conversation
This PR needs a
|
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/173571
Note: Links to docs will display an error until the docs builds have been completed. ✅ No FailuresAs of commit f342c43 with merge base 02a87d7 ( This comment was automatically generated by Dr. CI and updates every 15 minutes. |
8aba31c to
3495f7e
Compare
pytorch-bot bot
pushed a commit
that referenced
this pull request
Jan 27, 2026
Summary: Refactor PeerToPeerAccess by moving the core implementation from aten/src/ATen/cuda to c10/cuda. This makes P2P and fabric access queries available at the c10 layer without requiring ATen dependencies. The ATen layer now provides thin wrappers that ensure CUDA lazy initialization before forwarding to c10. This separation allows lower-level CUDA code to query P2P capabilities without pulling in ATen context machinery. Reviewed By: ngimel Differential Revision: D91506414
minsii
added a commit
to minsii/pytorch
that referenced
this pull request
Jan 27, 2026
Summary: Refactor PeerToPeerAccess by moving the core implementation from aten/src/ATen/cuda to c10/cuda. This makes P2P and fabric access queries available at the c10 layer without requiring ATen dependencies. The ATen layer now provides thin wrappers that ensure CUDA lazy initialization before forwarding to c10. This separation allows lower-level CUDA code to query P2P capabilities without pulling in ATen context machinery. Reviewed By: ngimel Differential Revision: D91506414
minsii
added a commit
to minsii/pytorch
that referenced
this pull request
Jan 28, 2026
Summary: Refactor PeerToPeerAccess by moving the core implementation from aten/src/ATen/cuda to c10/cuda. This makes P2P and fabric access queries available at the c10 layer without requiring ATen dependencies. The ATen layer now provides thin wrappers that ensure CUDA lazy initialization before forwarding to c10. This separation allows lower-level CUDA code to query P2P capabilities without pulling in ATen context machinery. Reviewed By: ngimel Differential Revision: D91506414
minsii
added a commit
to minsii/pytorch
that referenced
this pull request
Jan 28, 2026
Summary: Refactor PeerToPeerAccess by moving the core implementation from aten/src/ATen/cuda to c10/cuda. This makes P2P and fabric access queries available at the c10 layer without requiring ATen dependencies. The ATen layer now provides thin wrappers that ensure CUDA lazy initialization before forwarding to c10. This separation allows lower-level CUDA code to query P2P capabilities without pulling in ATen context machinery. Reviewed By: ngimel Differential Revision: D91506414
3495f7e to
3c46061
Compare
minsii
added a commit
to minsii/pytorch
that referenced
this pull request
Jan 28, 2026
Summary: Refactor PeerToPeerAccess by moving the core implementation from aten/src/ATen/cuda to c10/cuda. This makes P2P and fabric access queries available at the c10 layer without requiring ATen dependencies. The ATen layer now provides thin wrappers that ensure CUDA lazy initialization before forwarding to c10. This separation allows lower-level CUDA code to query P2P capabilities without pulling in ATen context machinery. Reviewed By: ngimel Differential Revision: D91506414
3c46061 to
1ef4deb
Compare
minsii
added a commit
to minsii/pytorch
that referenced
this pull request
Jan 28, 2026
Summary: Refactor PeerToPeerAccess by moving the core implementation from aten/src/ATen/cuda to c10/cuda. This makes P2P and fabric access queries available at the c10 layer without requiring ATen dependencies. The ATen layer now provides thin wrappers that ensure CUDA lazy initialization before forwarding to c10. This separation allows lower-level CUDA code to query P2P capabilities without pulling in ATen context machinery. Reviewed By: ngimel Differential Revision: D91506414
minsii
added a commit
to minsii/pytorch
that referenced
this pull request
Jan 28, 2026
Summary: Refactor PeerToPeerAccess by moving the core implementation from aten/src/ATen/cuda to c10/cuda. This makes P2P and fabric access queries available at the c10 layer without requiring ATen dependencies. The ATen layer now provides thin wrappers that ensure CUDA lazy initialization before forwarding to c10. This separation allows lower-level CUDA code to query P2P capabilities without pulling in ATen context machinery. Reviewed By: ngimel Differential Revision: D91506414
minsii
added a commit
to minsii/pytorch
that referenced
this pull request
Jan 28, 2026
Summary: Refactor PeerToPeerAccess by moving the core implementation from aten/src/ATen/cuda to c10/cuda. This makes P2P and fabric access queries available at the c10 layer without requiring ATen dependencies. The ATen layer now provides thin wrappers that ensure CUDA lazy initialization before forwarding to c10. This separation allows lower-level CUDA code to query P2P capabilities without pulling in ATen context machinery. Reviewed By: ngimel Differential Revision: D91506414
Summary: Refactor PeerToPeerAccess by moving the core implementation from aten/src/ATen/cuda to c10/cuda. This makes P2P and fabric access queries available at the c10 layer without requiring ATen dependencies. The ATen layer now provides thin wrappers that ensure CUDA lazy initialization before forwarding to c10. This separation allows lower-level CUDA code to query P2P capabilities without pulling in ATen context machinery. Reviewed By: ngimel Differential Revision: D91506414
1ef4deb to
f342c43
Compare
minsii
added a commit
to minsii/pytorch
that referenced
this pull request
Jan 28, 2026
Summary: Refactor PeerToPeerAccess by moving the core implementation from aten/src/ATen/cuda to c10/cuda. This makes P2P and fabric access queries available at the c10 layer without requiring ATen dependencies. The ATen layer now provides thin wrappers that ensure CUDA lazy initialization before forwarding to c10. This separation allows lower-level CUDA code to query P2P capabilities without pulling in ATen context machinery. Reviewed By: ngimel Differential Revision: D91506414
minsii
added a commit
to minsii/pytorch
that referenced
this pull request
Jan 28, 2026
Summary: Refactor PeerToPeerAccess by moving the core implementation from aten/src/ATen/cuda to c10/cuda. This makes P2P and fabric access queries available at the c10 layer without requiring ATen dependencies. The ATen layer now provides thin wrappers that ensure CUDA lazy initialization before forwarding to c10. This separation allows lower-level CUDA code to query P2P capabilities without pulling in ATen context machinery. Reviewed By: ngimel Differential Revision: D91506414
ngimel
approved these changes
Jan 28, 2026
minsii
added a commit
to minsii/pytorch
that referenced
this pull request
Jan 28, 2026
Summary: Refactor PeerToPeerAccess by moving the core implementation from aten/src/ATen/cuda to c10/cuda. This makes P2P and fabric access queries available at the c10 layer without requiring ATen dependencies. The ATen layer now provides thin wrappers that ensure CUDA lazy initialization before forwarding to c10. This separation allows lower-level CUDA code to query P2P capabilities without pulling in ATen context machinery. Reviewed By: ngimel Differential Revision: D91506414
minsii
added a commit
to minsii/pytorch
that referenced
this pull request
Jan 29, 2026
Summary: Refactor PeerToPeerAccess by moving the core implementation from aten/src/ATen/cuda to c10/cuda. This makes P2P and fabric access queries available at the c10 layer without requiring ATen dependencies. The ATen layer now provides thin wrappers that ensure CUDA lazy initialization before forwarding to c10. This separation allows lower-level CUDA code to query P2P capabilities without pulling in ATen context machinery. Reviewed By: ngimel Differential Revision: D91506414
minsii
added a commit
to minsii/pytorch
that referenced
this pull request
Jan 29, 2026
Summary: Refactor PeerToPeerAccess by moving the core implementation from aten/src/ATen/cuda to c10/cuda. This makes P2P and fabric access queries available at the c10 layer without requiring ATen dependencies. The ATen layer now provides thin wrappers that ensure CUDA lazy initialization before forwarding to c10. This separation allows lower-level CUDA code to query P2P capabilities without pulling in ATen context machinery. Reviewed By: ngimel Differential Revision: D91506414
minsii
added a commit
to minsii/pytorch
that referenced
this pull request
Jan 29, 2026
Summary: Refactor PeerToPeerAccess by moving the core implementation from aten/src/ATen/cuda to c10/cuda. This makes P2P and fabric access queries available at the c10 layer without requiring ATen dependencies. The ATen layer now provides thin wrappers that ensure CUDA lazy initialization before forwarding to c10. This separation allows lower-level CUDA code to query P2P capabilities without pulling in ATen context machinery. Reviewed By: ngimel Differential Revision: D91506414
minsii
added a commit
to minsii/pytorch
that referenced
this pull request
Jan 29, 2026
Summary: Refactor PeerToPeerAccess by moving the core implementation from aten/src/ATen/cuda to c10/cuda. This makes P2P and fabric access queries available at the c10 layer without requiring ATen dependencies. The ATen layer now provides thin wrappers that ensure CUDA lazy initialization before forwarding to c10. This separation allows lower-level CUDA code to query P2P capabilities without pulling in ATen context machinery. Reviewed By: ngimel Differential Revision: D91506414
minsii
added a commit
to minsii/pytorch
that referenced
this pull request
Jan 29, 2026
Summary: Refactor PeerToPeerAccess by moving the core implementation from aten/src/ATen/cuda to c10/cuda. This makes P2P and fabric access queries available at the c10 layer without requiring ATen dependencies. The ATen layer now provides thin wrappers that ensure CUDA lazy initialization before forwarding to c10. This separation allows lower-level CUDA code to query P2P capabilities without pulling in ATen context machinery. Reviewed By: ngimel Differential Revision: D91506414
minsii
added a commit
to minsii/pytorch
that referenced
this pull request
Jan 29, 2026
Summary: Refactor PeerToPeerAccess by moving the core implementation from aten/src/ATen/cuda to c10/cuda. This makes P2P and fabric access queries available at the c10 layer without requiring ATen dependencies. The ATen layer now provides thin wrappers that ensure CUDA lazy initialization before forwarding to c10. This separation allows lower-level CUDA code to query P2P capabilities without pulling in ATen context machinery. Reviewed By: ngimel Differential Revision: D91506414
Contributor
|
@pytorchbot merge (Initiating merge automatically since Phabricator Diff has merged) |
Collaborator
Merge failedReason: This PR needs a If not, please add the To add a label, you can comment to pytorchbot, for example For more information, see Details for Dev Infra teamRaised by workflow job |
This PR needs a
|
Contributor
|
@pytorchbot label "topic: not user facing" |
Contributor
|
closing for rebase |
pytorchmergebot
pushed a commit
that referenced
this pull request
Feb 10, 2026
Refactor PeerToPeerAccess by moving the core implementation from aten/src/ATen/cuda to c10/cuda. This makes P2P and fabric access queries available at the c10 layer without requiring ATen dependencies. The ATen layer now provides thin wrappers that ensure CUDA lazy initialization before forwarding to c10. This separation allows lower-level CUDA code to query P2P capabilities without pulling in ATen context machinery. Original diff by @minsii #173571 Bifferential Revision: [D92675476](https://our.internmc.facebook.com/intern/diff/D92675476/) Pull Request resolved: #174582 Approved by: https://github.com/Skylion007
radeksm
pushed a commit
to radeksm/pytorch
that referenced
this pull request
Feb 20, 2026
Refactor PeerToPeerAccess by moving the core implementation from aten/src/ATen/cuda to c10/cuda. This makes P2P and fabric access queries available at the c10 layer without requiring ATen dependencies. The ATen layer now provides thin wrappers that ensure CUDA lazy initialization before forwarding to c10. This separation allows lower-level CUDA code to query P2P capabilities without pulling in ATen context machinery. Original diff by @minsii pytorch#173571 Bifferential Revision: [D92675476](https://our.internmc.facebook.com/intern/diff/D92675476/) Pull Request resolved: pytorch#174582 Approved by: https://github.com/Skylion007
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
5 participants
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.
Summary:
Refactor PeerToPeerAccess by moving the core implementation from
aten/src/ATen/cuda to c10/cuda. This makes P2P and fabric access
queries available at the c10 layer without requiring ATen dependencies.
The ATen layer now provides thin wrappers that ensure CUDA lazy
initialization before forwarding to c10. This separation allows
lower-level CUDA code to query P2P capabilities without pulling in
ATen context machinery.
Reviewed By: ngimel
Differential Revision: D91506414