Skip to content

Conversation

@cdaley
Copy link

@cdaley cdaley commented Oct 25, 2024

Add a modified version of the optimization mentioned in issue #4951. The code now avoids goto statements. The code now also only restricts the gemv forwarding on ARM64 systems since this is where we observed poor gemv performance.

Copy link
Contributor

@Mousius Mousius left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Looks good to me, thanks for removing the gotos 😸

@martin-frbg
Copy link
Collaborator

Thank you (both), merging this.

@martin-frbg martin-frbg merged commit ac73682 into OpenMathLib:develop Oct 25, 2024
80 of 83 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

Performance regression in version 0.3.28 on aarch64 because of GEMM to GEMV transformation

3 participants