Skip to content

[SYCL][CUDA] Add missing nearbyint #11177

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
Sep 14, 2023

Conversation

hdelan
Copy link
Contributor

@hdelan hdelan commented Sep 14, 2023

nearbyint was missing for NVPTX. nearbyint is not found in libclc, so instead of the usual libdevice->libclc->native_bc_lib it was necessary to just skip libclc and go straight to the __nv symbols which are found in nvvm libdevice.

@hdelan hdelan requested a review from a team as a code owner September 14, 2023 10:45
@steffenlarsen
Copy link
Contributor

AMD failure is known and reported in #10460.

@steffenlarsen steffenlarsen merged commit 0ef26d3 into intel:sycl Sep 14, 2023
againull pushed a commit that referenced this pull request Dec 30, 2024
Failed to compile OneDNN using DPCPP on AMD with sycl because the code
[here](https://github.com/oneapi-src/oneDNN/blob/c852fdc60073a1ae298618c4faf26487b4144755/src/gpu/generic/sycl/eltwise_kernels.hpp#L137)
that will run on the device actually calls `nearbyintf`
[here](https://github.com/oneapi-src/oneDNN/blob/c852fdc60073a1ae298618c4faf26487b4144755/src/common/math_utils.hpp#L120-L131)
, but `nearbyintf` lacks a symbolic definition.
Same solution as #11177 but go
straight to the `__ocml` symbols
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