Skip to content

Cleanup OS provider config API #209

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 3 commits into from
Feb 13, 2024
Merged

Conversation

igchor
Copy link
Member

@igchor igchor commented Feb 5, 2024

@vinser52 @pbalcer please review. Each commit is an independent change.

They are only applicable for already allocated memory.
Right now, UMF provider does not expose any API for calling
mbind after allocation or moving specific pages.
@igchor igchor requested a review from a team as a code owner February 5, 2024 20:42
Copy link
Contributor

@lukaszstolarczuk lukaszstolarczuk left a comment

Choose a reason for hiding this comment

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

soo, you added DEFAULT flag (in commit: 360208e) and then removed it (in commit: ec8b068)...? 😉 It kinda makes sense only if you plan on reverting any of these commits some day

There was only one possible value (UMF_NUMA_FLAG_STRICT)
but it seems to mostly apply to page migration (i.e. when
MPOL_MF_MOVE is specified). Since UMF does not support this
for now, there's no point in keeping this flag. During allocation
time, we still have UMF_NUMA_MODE_PREFFERED that can be used
to request a 'preffered' binding. For any other mode, we require
mbind to succeed.
There is no relative node id concept on Windows.
Moreover, the provider API should be as simple as possible.
If we'd like to add support for relative nodes in the future,
we can always expose a function to translate node ids.
@igchor
Copy link
Member Author

igchor commented Feb 6, 2024

soo, you added DEFAULT flag (in commit: 360208e) and then removed it (in commit: ec8b068)...? 😉 It kinda makes sense only if you plan on reverting any of these commits some day

I guess I can remove it. Done.

Copy link
Contributor

@lukaszstolarczuk lukaszstolarczuk left a comment

Choose a reason for hiding this comment

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

👍

@bratpiorka bratpiorka merged commit 900e5aa into oneapi-src:main Feb 13, 2024
@igchor igchor deleted the os_provider branch February 13, 2024 15:51
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.

6 participants