propagate debug handle from edge dialect graph back to exported graph#12337
Merged
Conversation
Pull Request resolved: #12288 Currently ExecuTorch debug handle infra traces operator transformation starting from export program. This diff makes etrecord support recording export program for further analysis use. ghstack-source-id: 295312924 @exported-using-ghexport Differential Revision: [D77965102](https://our.internmc.facebook.com/intern/diff/D77965102/)
Pull Request resolved: #12330 Using exported graph from torch.export as source of truth for aot intermediate output is our target. Once thing blocking us is the exported graph does not have debug handle, which will not show up in the export flow until DebugHandleGenerationPass, the last step of to_edge(). We need to equip the graph with same debug handle used in ExecuTorch flow. This diff creates a propagate_back_debug_handle function, which propagate debug handle from edge dialect program back to the exported program while maintain the correctness of operator tracing. ghstack-source-id: 295337600 @exported-using-ghexport Differential Revision: [D78051614](https://our.internmc.facebook.com/intern/diff/D78051614/)
pytorchbot
requested review from
Gasoonjia,
JacobSzwejbka and
larryliu0820
as code owners
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/12337
Note: Links to docs will display an error until the docs builds have been completed. ❌ 4 New Failures, 3 Unrelated FailuresAs of commit 1169788 with merge base dd06b3b ( NEW FAILURES - The following jobs have failed:
BROKEN TRUNK - The following jobs failed but were present on the merge base:👉 Rebase onto the `viable/strict` branch to avoid these failures
This comment was automatically generated by Dr. CI and updates every 15 minutes. |
facebook-github-bot
added
the
CLA Signed
Gasoonjia
approved these changes
Jul 10, 2025
Gasoonjia
left a comment
Gasoonjia
left a comment
Contributor
There was a problem hiding this comment.
it breaks oss cis because the core update the pr relies on haven't include into their nightly version yet. Should be auto-solved by the end of tmr
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
3 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.
This PR was created by the merge bot to help merge the original PR into the main branch.
ghstack PR number: #12330 by @Gasoonjia
^ 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/gasoonjia/19/base
ghstack PR head: https://github.com/pytorch/executorch/tree/gh/gasoonjia/19/head
Merge bot PR base: https://github.com/pytorch/executorch/tree/gh/gasoonjia/18/orig
Merge bot PR head: https://github.com/pytorch/executorch/tree/gh/gasoonjia/19/orig
@diff-train-skip-merge