Skip to content

Conversation

@Retribution98
Copy link
Contributor

@Retribution98 Retribution98 commented Oct 17, 2024

Closes intel#800

Signed-off-by: Kirill Suvorov <[email protected]>
Comment on lines -144 to -146
# Wait until driver complete all the jobs for the device_print, especially test_subprocess
# require this which captures stdout when child exits.
getattr(torch, device).synchronize()
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Is it necessary to move this?

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Yes, first of all the synchronize should be called.

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

this needs to be done in upstream as well

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Looks reasonable, I have created the upstream PR

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Is it necessary to move this?

even with the new driver?

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Is it necessary to move this?

even with the new driver?

yes

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Can you please check if these tests work with 2025.0.0 without moving these lines?
It was checked in Aug, and at that time, they work with 2025.0.0.

Comment on lines -1 to -8
# https://github.com/intel/intel-xpu-backend-for-triton/issues/800
test/unit/language/test_subprocess.py::test_print[device_print-float16]
test/unit/language/test_subprocess.py::test_print[device_print-float32]
test/unit/language/test_subprocess.py::test_print[device_print-float64]
test/unit/language/test_subprocess.py::test_print[device_print_scalar-float16]
test/unit/language/test_subprocess.py::test_print[device_print_scalar-float64]
test/unit/language/test_subprocess.py::test_print[device_print_scalar-float32]
# https://github.com/intel/intel-xpu-backend-for-triton/issues/1704
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

To make sure it works, you need to run corresponding workflow separately.

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

The no-basekit and the conda workflows don't work now. I can leave these skiplists, but I believe it should work as it works for other cases.

@etiotto
Copy link
Contributor

etiotto commented Nov 15, 2024

@Retribution98 still working on this PR ? It needs rebasing. Please close if you deem this unnecessary now.

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.

[UT] regression in test_subprocess.py with the PTDB 0.5.3

4 participants