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

xskdev:keep FQ as full as possible #177

Merged
merged 2 commits into from
Aug 30, 2022

Conversation

KeithWiles
Copy link
Contributor

It appears the FQ was not being kept as full as possible. The code would do a receive and
then put back 64 mbufs, but the rx_burst() request may have been larger then 64 and it
starts to deplete the FQ ring. It also appears when the FQ is low on packets the needs
wakeup flag is always set and forces a poll() call, which I think then attempts to refill
the NIC ring from the FQ and then increments the rx_fill_ring_emtpy counter.

Signed-off-by: Keith Wiles keith.wiles@intel.com

It appears the FQ was not being kept as full as possible. The code would do a receive and
then put back 64 mbufs, but the rx_burst() request may have been larger then 64 and it
starts to deplete the FQ ring. It also appears when the FQ is low on packets the needs
wakeup flag is always set and forces a poll() call, which I think then attempts to refill
the NIC ring from the FQ and then increments the rx_fill_ring_emtpy counter.

Signed-off-by: Keith Wiles <keith.wiles@intel.com>
@maryamtahhan maryamtahhan merged commit c8bea5c into CloudNativeDataPlane:main Aug 30, 2022
@KeithWiles KeithWiles deleted the xskdev-fq-full branch August 30, 2022 09:10
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