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.

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

prov/efa: Fix the rnr cq read error test for efa-direct #10818

Merged
merged 2 commits into from
Feb 23, 2025

Conversation

shijin-aws
Copy link
Contributor

This PR contains 2 changes to fix the rnr cq read error test for efa-direct:

  1. The error code returned by efa-direct in the fi_send call is wrong when hitting resource exhaustion. It needs a conversion for ibv errno to fi errno
  2. For the cq read error test, efa-direct doesn't prepost any rx buffer internally, it posts whatever application posts, the total_send needs to be adjusted to match this behavior difference.

On failure, ibv_wr_complete() return positive errnno
Meanwhile, the function is supposed to return a negative errno.
So we need a conversion from the ibv errno to the fi errno,
especially for ENOMEM which corresponds to -FI_EAGAIN

Signed-off-by: Shi Jin <sjina@amazon.com>
efa-direct doesn't prepost any rx buffer internally, it posts
whatever application posts, the total_send needs to be adjusted
to match this behavior difference.

Signed-off-by: Shi Jin <sjina@amazon.com>
@shijin-aws
Copy link
Contributor Author

Tested by #10800

@@ -0,0 +1,19 @@
/* SPDX-License-Identifier: BSD-2-Clause OR GPL-2.0-only */
Copy link
Contributor Author

Choose a reason for hiding this comment

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

This is useful for general efa specific fabtests in the future. So I intend to keep it in a separate .h

@shijin-aws shijin-aws merged commit 5163e6a into ofiwg:main Feb 23, 2025
13 checks passed
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