Skip to content

[SYCL][L0] Fix BDF format on PCI query #6011

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

Conversation

steffenlarsen
Copy link
Contributor

The PCI address currently reported by the L0 backend does not correctly report domain, bus, device, and function as hexadecimal values. This commit changes the reported format to use convert the values hexadecimal representation. Additionally it also pads the values with leading zeros.

The PCI address currently reported by the L0 backend does not correctly
report domain, bus, device, and function as hexadecimal values. This
commit changes the reported format to use convert the values hexadecimal
representation. Additionally it also pads the values with leading zeros.

Signed-off-by: Larsen, Steffen <steffen.larsen@intel.com>
@steffenlarsen steffenlarsen requested a review from a team as a code owner April 13, 2022 15:47
@steffenlarsen
Copy link
Contributor Author

@smaslov-intel - To my knowledge the padding is not strictly required as part of the BDF format, but it makes the reported values more consistent and easier for users to parse. If you prefer not to include it, please let me know and I will remove it.

smaslov-intel
smaslov-intel previously approved these changes Apr 14, 2022
Copy link
Contributor

@smaslov-intel smaslov-intel left a comment

Choose a reason for hiding this comment

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

LGTM. Is it possible to add test for this, please?

Signed-off-by: Larsen, Steffen <steffen.larsen@intel.com>
@steffenlarsen
Copy link
Contributor Author

LGTM. Is it possible to add test for this, please?

New test is up here: intel/llvm-test-suite#991

It also showed that has(aspect::ext_intel_pci_address) did not correctly report support. Should be fixed with 0fe86c6.

@steffenlarsen
Copy link
Contributor Author

/verify with intel/llvm-test-suite#991

@steffenlarsen
Copy link
Contributor Author

Unexpected pass in KernelAndProgram/undefined-symbol.cpp on Windows. This is known and should be fixed with intel/llvm-test-suite#994 with GPU uplift.

@steffenlarsen steffenlarsen merged commit ad6253e into intel:sycl Apr 19, 2022
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.

2 participants