Skip to content

fix clEnqueueNDRangeKernel for NULL work_global_size #760

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

Open
wants to merge 1 commit into
base: main
Choose a base branch
from

Conversation

rjodinchr
Copy link
Contributor

@rjodinchr rjodinchr commented Jan 24, 2025

According to the OpenCL specification, work_dim can be 0 and work_global_size can be NULL.

KhronosGroup/OpenCL-CTS#2240

@rjodinchr
Copy link
Contributor Author

switftshader artifact needs to be refreshed

@rjodinchr rjodinchr force-pushed the pr/null-work-global-size branch from 8766f6b to fcc8884 Compare January 24, 2025 15:13
@rjodinchr
Copy link
Contributor Author

The issue is shown by running: test_profiling profiling_timebase. It segfaults with llvmpipe on my end.

It passes with zero_sized_enqueue because that test never tries to pass NULL as the global_work_size, it always gives a valid structure with nothing to compute. While profiling_timebase calls clEnqueueNDRangeKernel with NULL as the global_work_size

@rjodinchr
Copy link
Contributor Author

rjodinchr commented Apr 14, 2025

TODO: change the patch to override global_work_size when it equals nullptr
Done

@rjodinchr rjodinchr force-pushed the pr/null-work-global-size branch from fcc8884 to b301a9e Compare April 15, 2025 08:20
According to the OpenCL specification, work_dim can be 0 and
work_global_size can be NULL.

KhronosGroup/OpenCL-CTS#2240
@rjodinchr rjodinchr force-pushed the pr/null-work-global-size branch from b301a9e to e4421e8 Compare April 15, 2025 08:45
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.

1 participant