Skip to content

Conversation

lewing
Copy link
Member

@lewing lewing commented Jun 20, 2024

Reverts #102670

I'm staging this because benchmarks show a regression in the NormalizeAscii benchmark, but I'm in no hurry to revert if we can think of a different resolution.

image

@kg
Copy link
Member

kg commented Jun 20, 2024

I think the cause is probably the use of MUL_P_IMM in the new intrinsic. My guess is that (after optimization) the old non-intrinsic path would have used a shift in some cases, which is maybe slightly faster? It should be easy to fix, so I'll prep a PR for that.

@lewing lewing merged commit 351926a into main Jul 18, 2024
@lewing lewing deleted the revert-102670-interp-blazor-intrinsics-1 branch July 18, 2024 16:08
@github-actions github-actions bot locked and limited conversation to collaborators Aug 18, 2024
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants