Skip to content

Fix an issue in the Avx2 code of sparse matrix multiply #664

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

Closed
wants to merge 1 commit into from

Conversation

dskhudia
Copy link
Contributor

@dskhudia dskhudia commented Aug 6, 2021

Summary: There is an issue with respect to accessing memory outside bounds which is fixed with this change.

Reviewed By: dskhudia

Differential Revision: D30146214

Summary: There is an issue with respect to accessing memory outside bounds which is fixed with this change.

Reviewed By: dskhudia

Differential Revision: D30146214

fbshipit-source-id: 8599603372e9f5fc3f6b7cbc1ef768526e9d2f97
@facebook-github-bot
Copy link
Contributor

This pull request was exported from Phabricator. Differential Revision: D30146214

@facebook-github-bot
Copy link
Contributor

This pull request has been merged in a35fa9f.

q10 pushed a commit to q10/FBGEMM that referenced this pull request Apr 10, 2025
…da (pytorch#664)

Summary:
Pull Request resolved: facebookresearch/FBGEMM#664

X-link: pytorch#3578

`lengths` is a tensor with symbolic shapes. Calling `len` on it will force specialization on it which will cause data dependent failure as shown below:
 {F1974383976}

tlparse: https://fburl.com/74rjmr8e

The fix is to replace `len` with equivalent operations which support symbolic shapes.

Reviewed By: TroyGarden

Differential Revision: D67491452

fbshipit-source-id: ed2207b310697d774a284f296c8d34ca2da61adc
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants