Skip to content

Use fabs intrinsic #95

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

Merged
merged 1 commit into from
Apr 19, 2021
Merged

Use fabs intrinsic #95

merged 1 commit into from
Apr 19, 2021

Conversation

workingjubilee
Copy link
Member

I noticed there's a simd_fabs even though there is no simd_abs.
Same story as #89.

@workingjubilee
Copy link
Member Author

...Why do these tests pass locally but fail on GH?

@Lokathor
Copy link
Contributor

dunno but the code itself looks good.

@calebzulawski
Copy link
Member

I just tried this on my machine (AMD) and it also passed.

@calebzulawski
Copy link
Member

Oh wait, I rebased and ran it in release mode and it failed. Anyway, I found the problem. Look at what asm is generated for f32x4::is_infinite():

 mov     rax, rdi
 xorps   xmm0, xmm0
 movaps  xmmword, ptr, [rdi], xmm0
 ret

@workingjubilee
Copy link
Member Author

wait
does that just... do nothing?

@calebzulawski
Copy link
Member

It always returns false :)

@workingjubilee
Copy link
Member Author

always finite but never normal around here, apparently. :^)

@calebzulawski
Copy link
Member

Opened rust-lang/rust#84268 and #97.

@calebzulawski
Copy link
Member

I posted it in the linked issue, but this looks like an LLVM 12 regression

@calebzulawski calebzulawski merged commit b2e25bc into master Apr 19, 2021
@calebzulawski calebzulawski deleted the intrinsic-fabs branch August 7, 2021 18:19
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.

3 participants