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

[CUSOLVER] cuSOLVER handler does not support multiple streams #216

Closed
ericlars opened this issue Jul 29, 2022 · 1 comment · Fixed by #572
Closed

[CUSOLVER] cuSOLVER handler does not support multiple streams #216

ericlars opened this issue Jul 29, 2022 · 1 comment · Fixed by #572

Comments

@ericlars
Copy link
Contributor

Summary

Intel llvm introduced multiple CUDA streams per SYCL queue which broke the cuSOLVER scope handler that assumed one stream per device per thread. This broke asynchronous submissions, in the short term synchronous cuSOLVER submissions should be used. See #209 and #215 for further details.

Version

All versions built with intel llvm after dd418459868a976cd2eeae367fea6b92795ea611.

Environment

cuSOLVER backend
ericlars added a commit to ericlars/oneMKL that referenced this issue Jul 30, 2022
@ericlars
[CUSOLVER] workaround for issue uxlfoundation#216
1bc61ee
@JackAKirk JackAKirk mentioned this issue Aug 1, 2022
Closed
@mmeterel
Copy link
Contributor

@ericlars is this fixed and can be closed?

@JackAKirk JackAKirk mentioned this issue Oct 2, 2024
Merged
1 task
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

Successfully merging a pull request may close this issue.

[lapack][blas][cuda] Update host task impl to use enqueue_native_command JackAKirk/oneMKL
2 participants
@mmeterel @ericlars