[CUDA] Use runtime driver API for cuStreamWriteValue32#156097
[CUDA] Use runtime driver API for cuStreamWriteValue32#156097eee4017 wants to merge 20 commits intopytorch:mainfrom
Conversation
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/156097
Note: Links to docs will display an error until the docs builds have been completed. ✅ You can merge normally! (1 Unrelated Failure)As of commit dabccdb with merge base 7f14b42 ( UNSTABLE - The following job is marked as unstable, possibly due to flakiness on trunk:
This comment was automatically generated by Dr. CI and updates every 15 minutes. |
96693fe to
ad432d9
Compare
|
@pytorchbot merge |
Merge startedYour change will be merged once all checks pass (ETA 0-4 Hours). Learn more about merging in the wiki. Questions? Feedback? Please reach out to the PyTorch DevX Team |
|
@pytorchmergebot revert -c nosiganl -m "break internal tests" |
|
❌ 🤖 pytorchbot command failed: Try |
|
@pytorchmergebot revert -c nosignal -m "break internal tests" |
|
@pytorchbot successfully started a revert job. Check the current status here. |
|
@eee4017 your PR has been successfully reverted. |
…#156097)" This reverts commit cf90c9f. Reverted #156097 on behalf of https://github.com/atalman due to break internal tests ([comment](#156097 (comment)))
This PR was reopened (likely due to being reverted), so your approval was removed. Please request another review.
This reverts commit f510f60.
| #undef LOOKUP_LIBCUDA_ENTRY | ||
| #define LOOKUP_LIBCUDA_ENTRY_WITH_VERSION_OPTIONAL(name, version) \ | ||
| r.name##_ = reinterpret_cast<decltype(&name)>(get_symbol(#name, version)); | ||
| C10_LIBCUDA_DRIVER_API_OPTIONAL(LOOKUP_LIBCUDA_ENTRY_WITH_VERSION_OPTIONAL) |
There was a problem hiding this comment.
The assertion was removed to allow compatibility with older drivers.
There was a problem hiding this comment.
Thank you for the update! Let me do a round of testing on Monday to confirm
There was a problem hiding this comment.
Rerun the manual test with the latest change from the PR https://github.com/pytorch/pytorch/actions/runs/16276707651/job/45957270665?pr=158181
There was a problem hiding this comment.
Although it's sill running, I think the test looks good now because:
- It correctly uses older NVIDIA driver https://github.com/pytorch/pytorch/actions/runs/16276707651/job/45959318683?pr=158181#step:13:764
- Tests are running ok instead of crashing
|
@pytorchbot merge |
Merge startedYour change will be merged once all checks pass (ETA 0-4 Hours). Learn more about merging in the wiki. Questions? Feedback? Please reach out to the PyTorch DevX Team |
Merge failedReason: 1 jobs have failed, first few of them are: Meta Internal-Only Changes Check Details for Dev Infra teamRaised by workflow job |
|
@eee4017 sorry can you please create a new pull request? |
|
Hi, it was created here: #158295 |
|
Looks like this PR hasn't been updated in a while so we're going to go ahead and mark this as |
|
Closed with #158585 |
Fixes #154073
Reference: NVIDIA/Fuser#4197
See PR #154097
@nWEIdia is currently out of the office, so I’ve temporarily taken over his work.
cc @H-Huang @awgu @wanchaol @fegin @fduwjj @wz337 @wconstab @d4l3k @pragupta @ezyang @msaroufim @dcci