New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Use unordered NEQ comparison for vec512 operator!= implementations #97466
Conversation
This is consistent with the vec256 operator!= implementations. _CMP_NEQ_UQ is the logical opposite of _CMP_EQ_OQ comparison used in the operator== implementations. Using the ordered NEQ operation results in nan != nan being false which is incorrect. Signed-off-by: BJ Hargrave <hargrave@us.ibm.com>
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/97466
Note: Links to docs will display an error until the docs builds have been completed. ✅ No FailuresAs of commit 7bb4c06: This comment was automatically generated by Dr. CI and updates every 15 minutes. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LGTM. Better we can cover vec512 cases in CI. @sanchitintel
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Thank you for catching & fixing this bug! :)
@jgong5, PyTorch CI had AVX512 ATen kernels enabled for a couple of months last year but this bug wasn't caught with the erstwhile CI jobs' workloads. We should probably add a UT to check |
@pytorchbot label "topic: bug fixes" "release notes: intel" |
@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: trunk / linux-focal-rocm5.4.2-py3.8 / test (default, 1, 3, linux.rocm.gpu) Details for Dev Infra teamRaised by workflow job |
@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 |
This is consistent with the vec256 operator!= implementations. _CMP_NEQ_UQ is the logical opposite of _CMP_EQ_OQ comparison used in the operator== implementations.
Using the ordered NEQ operation results in nan != nan being false which is incorrect.
cc @jgong5 @mingfeima @XiaobingSuper @sanchitintel @ashokei @jingxu10