convert inductor codecache to use getArtifactLogger #153766
Closed
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
[ghstack-poisoned]
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/153766
Note: Links to docs will display an error until the docs builds have been completed. ❌ 3 New Failures, 2 Unrelated FailuresAs of commit 88ef48c with merge base 76f182f ( NEW FAILURES - The following jobs have failed:
FLAKY - The following job failed but was likely due to flakiness present on trunk:
BROKEN TRUNK - The following job failed but was 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. |
oulgen
approved these changes
May 16, 2025
Skylion007
approved these changes
May 19, 2025
I'm not entirely sure of the background for why inductor codecache code uses default python logging instead of the new TORCH_LOGS-based artifact logging, but switching it over to artifact logging makes it easier to use nice testing utils in the next PR. cc voznesenskym penguinwu EikanWang jgong5 Guobing-Chen XiaobingSuper zhuhaozhe blzheng wenzhe-nrv jiayisunx ipiszy chenyang78 kadeng muchulee8 amjames chauhang aakhundov [ghstack-poisoned]
|
Starting merge as part of PR stack under #153672 |
pytorchmergebot
pushed a commit
that referenced
this pull request
May 30, 2025
…53672) Fixes pytorch/torchtitan#1185 It looks like inductor's logic to include inductor configs in the cache key skips configs with a leading underscore by default. This came up in torchtitan - there's an asyncTP pipelining pass in inductor gated by a private config, and by not caching on the config we were attempting to use asyncTP when we shouldn't be. I'm not sure how worried we should be on the blast radius of this change. On the one hand: (1) it technically fixes any silent correctness issues in the cache around any other private inductor configs (it looks like there are a few) (2) there is some risk that there are some "harmless" configs that we are now including in the key, which may increase false negatives. I do see that there is an explicit list for "configs we want to ignore for caching" (`_save_config_ignore`), so my hope is that all harmless configs are already encapsulated there. Pull Request resolved: #153672 Approved by: https://github.com/oulgen ghstack dependencies: #153766
|
@pytorchbot revert -m "I want to revert this change as I'm 90+% certain it somehow broke testing" -c weird |
|
@pytorchbot successfully started a revert job. Check the current status here. |
|
If I'm to install wheel generated for the top of the stack, it fails but if I re-run the same test with |
pytorchmergebot
added a commit
that referenced
this pull request
May 30, 2025
This reverts commit 5b6fd27. Reverted #153766 on behalf of https://github.com/malfet due to I want to revert this change as I'm 90+% certain it somehow broke testing ([comment](#153766 (comment)))
|
@bdhirsh your PR has been successfully reverted. |
nWEIdia
pushed a commit
to nWEIdia/pytorch
that referenced
this pull request
Jun 2, 2025
I'm not entirely sure of the background for why inductor codecache code uses default python logging instead of the new TORCH_LOGS-based artifact logging, but switching it over to artifact logging makes it easier to use nice testing utils in the next PR. Pull Request resolved: pytorch#153766 Approved by: https://github.com/oulgen, https://github.com/Skylion007
nWEIdia
pushed a commit
to nWEIdia/pytorch
that referenced
this pull request
Jun 2, 2025
…torch#153672) Fixes pytorch/torchtitan#1185 It looks like inductor's logic to include inductor configs in the cache key skips configs with a leading underscore by default. This came up in torchtitan - there's an asyncTP pipelining pass in inductor gated by a private config, and by not caching on the config we were attempting to use asyncTP when we shouldn't be. I'm not sure how worried we should be on the blast radius of this change. On the one hand: (1) it technically fixes any silent correctness issues in the cache around any other private inductor configs (it looks like there are a few) (2) there is some risk that there are some "harmless" configs that we are now including in the key, which may increase false negatives. I do see that there is an explicit list for "configs we want to ignore for caching" (`_save_config_ignore`), so my hope is that all harmless configs are already encapsulated there. Pull Request resolved: pytorch#153672 Approved by: https://github.com/oulgen ghstack dependencies: pytorch#153766
nWEIdia
pushed a commit
to nWEIdia/pytorch
that referenced
this pull request
Jun 2, 2025
…153766)" This reverts commit 5b6fd27. Reverted pytorch#153766 on behalf of https://github.com/malfet due to I want to revert this change as I'm 90+% certain it somehow broke testing ([comment](pytorch#153766 (comment)))
qingyi-yan
pushed a commit
to qingyi-yan/pytorch
that referenced
this pull request
Jun 3, 2025
I'm not entirely sure of the background for why inductor codecache code uses default python logging instead of the new TORCH_LOGS-based artifact logging, but switching it over to artifact logging makes it easier to use nice testing utils in the next PR. Pull Request resolved: pytorch#153766 Approved by: https://github.com/oulgen, https://github.com/Skylion007
qingyi-yan
pushed a commit
to qingyi-yan/pytorch
that referenced
this pull request
Jun 3, 2025
…torch#153672) Fixes pytorch/torchtitan#1185 It looks like inductor's logic to include inductor configs in the cache key skips configs with a leading underscore by default. This came up in torchtitan - there's an asyncTP pipelining pass in inductor gated by a private config, and by not caching on the config we were attempting to use asyncTP when we shouldn't be. I'm not sure how worried we should be on the blast radius of this change. On the one hand: (1) it technically fixes any silent correctness issues in the cache around any other private inductor configs (it looks like there are a few) (2) there is some risk that there are some "harmless" configs that we are now including in the key, which may increase false negatives. I do see that there is an explicit list for "configs we want to ignore for caching" (`_save_config_ignore`), so my hope is that all harmless configs are already encapsulated there. Pull Request resolved: pytorch#153672 Approved by: https://github.com/oulgen ghstack dependencies: pytorch#153766
qingyi-yan
pushed a commit
to qingyi-yan/pytorch
that referenced
this pull request
Jun 3, 2025
…153766)" This reverts commit 5b6fd27. Reverted pytorch#153766 on behalf of https://github.com/malfet due to I want to revert this change as I'm 90+% certain it somehow broke testing ([comment](pytorch#153766 (comment)))
iupaikov-amd
pushed a commit
to ROCm/pytorch
that referenced
this pull request
Jun 4, 2025
I'm not entirely sure of the background for why inductor codecache code uses default python logging instead of the new TORCH_LOGS-based artifact logging, but switching it over to artifact logging makes it easier to use nice testing utils in the next PR. Pull Request resolved: pytorch#153766 Approved by: https://github.com/oulgen, https://github.com/Skylion007
iupaikov-amd
pushed a commit
to ROCm/pytorch
that referenced
this pull request
Jun 4, 2025
…torch#153672) Fixes pytorch/torchtitan#1185 It looks like inductor's logic to include inductor configs in the cache key skips configs with a leading underscore by default. This came up in torchtitan - there's an asyncTP pipelining pass in inductor gated by a private config, and by not caching on the config we were attempting to use asyncTP when we shouldn't be. I'm not sure how worried we should be on the blast radius of this change. On the one hand: (1) it technically fixes any silent correctness issues in the cache around any other private inductor configs (it looks like there are a few) (2) there is some risk that there are some "harmless" configs that we are now including in the key, which may increase false negatives. I do see that there is an explicit list for "configs we want to ignore for caching" (`_save_config_ignore`), so my hope is that all harmless configs are already encapsulated there. Pull Request resolved: pytorch#153672 Approved by: https://github.com/oulgen ghstack dependencies: pytorch#153766
iupaikov-amd
pushed a commit
to ROCm/pytorch
that referenced
this pull request
Jun 4, 2025
…153766)" This reverts commit 5b6fd27. Reverted pytorch#153766 on behalf of https://github.com/malfet due to I want to revert this change as I'm 90+% certain it somehow broke testing ([comment](pytorch#153766 (comment)))
|
Looking into the error some more, it looks like inductor's handling for piping logging output between subprocesses doesn't have any knowledge of our I think I'm going to give up on this yak shave, and just land the PR above this one (the main reason I tried to move inductor's codecache logs to use artifact logging was to save ~20 lines of boilerplate on a test, which I'm going to add back). |
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.
I'm not entirely sure of the background for why inductor codecache code uses default python logging instead of the new TORCH_LOGS-based artifact logging, but switching it over to artifact logging makes it easier to use nice testing utils in the next PR.
Stack from ghstack (oldest at bottom):
cc @voznesenskym @penguinwu @EikanWang @jgong5 @Guobing-Chen @XiaobingSuper @zhuhaozhe @blzheng @wenzhe-nrv @jiayisunx @ipiszy @chenyang78 @kadeng @muchulee8 @amjames @chauhang @aakhundov