Skip to content
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.

Sign up for GitHub

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

Jump to bottom

[SYCL][CUDA] Add missing rintf #11561

Merged
merged 1 commit into from
Oct 19, 2023
Merged

[SYCL][CUDA] Add missing rintf #11561

merged 1 commit into from
Oct 19, 2023

Conversation

ph0b
Copy link
Contributor

@ph0b ph0b commented Oct 17, 2023

rintf was missing for NVPTX.

@ph0b ph0b requested a review from a team as a code owner October 17, 2023 12:19
@ph0b ph0b requested a review from bso-intel October 17, 2023 12:19
@ph0b ph0b temporarily deployed to WindowsCILock October 17, 2023 12:21 — with GitHub Actions Inactive
@ph0b ph0b temporarily deployed to WindowsCILock October 17, 2023 12:33 — with GitHub Actions Inactive
@ph0b
[SYCL][CUDA] Add missing rintf
4b1903a 
`rintf` was missing for NVPTX.
@ph0b ph0b temporarily deployed to WindowsCILock October 17, 2023 14:57 — with GitHub Actions Inactive
@ph0b ph0b temporarily deployed to WindowsCILock October 17, 2023 15:09 — with GitHub Actions Inactive
@ph0b
Copy link
Contributor Author

ph0b commented Oct 17, 2023

the cuda failure in CI tests tanh so it doesn't seem related, please advise.

@jinz2014
Copy link
Contributor

Do you find the missing function by calling std::rintf in a program ?
I have some questions. The modified file does not contain any functions for double-precision floating-point numbers and there are no HIP-specific functions. Thanks.

@ph0b
Copy link
Contributor Author

ph0b commented Oct 19, 2023 •
edited
Loading

@jinz2014 ptxas failed on this while compiling Blender for PTX on Windows,
The source code itself doesn't call std::rintf
I could get it to compile for HIP without such change.
I've also added rint to the fp64 variant of the file.

@jinz2014
Copy link
Contributor

I suppose that there is no unit/e2e test that reports any missing math functions.

@againull
Copy link
Contributor

This fail on CUDA is not related and covered by #11562
Failed Tests (1):
SYCL :: DeviceLib/built-ins/ext_native_math.cpp

@againull againull merged commit 3c327c7 into intel:sycl Oct 19, 2023
10 of 11 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@bso-intel bso-intel bso-intel approved these changes

Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
4 participants
@ph0b @jinz2014 @againull @bso-intel