Page MenuHomeFreeBSD
Differential D58306

rtw89/core.c: Avoid re-queueing if 'ieee80211_start_tx_ba_session' fails with a non-recoverable error
AbandonedPublic
Actions

Authored by bnovkov on Fri, Jul 17, 4:37 PM.
Tags
None
Referenced Files
F163410925: D58306.diff
Wed, Jul 22, 11:39 PM
F163383464: D58306.id182145.diff
Wed, Jul 22, 5:30 PM
Unknown Object (File)
Tue, Jul 21, 7:14 AM
Unknown Object (File)
Tue, Jul 21, 3:32 AM
Unknown Object (File)
Sun, Jul 19, 6:16 PM
Unknown Object (File)
Sun, Jul 19, 11:14 AM
Unknown Object (File)
Sat, Jul 18, 10:43 PM
Subscribers
bz
imp

Details

Reviewers
None
Group Reviewers
wireless
Summary

ieee80211_start_tx_ba_session can return multiple error codes, only
one of which signals that the function experienced a transient error
(EAGAIN) and that the work should be restarted. Unfortunately,
the code only tests for EINVAL, leading to the same structure
being queued over and over again with no progress. On a Thinkpad
E14 laptop with an RTL8852CE card this manifests itself as an a very spammy
wlan0: linuxkpi_ieee80211_start_tx_ba_seesion: HT or later not supported
message that is printed every few seconds.

Fix this by explicitly testing for !EAGAIN to make sure we aren't
repeatedly re-queueing work that can never succeed.

Diff Detail

Repository
rG FreeBSD src repository
Lint
Lint Skipped
Unit
Tests Skipped
Build Status
Buildable 74892
Build 71775: arc lint + arc unit

Event Timeline

bnovkov created this revision.Fri, Jul 17, 4:37 PM
Herald added a subscriber: imp. · View Herald TranscriptFri, Jul 17, 4:37 PM
bnovkov requested review of this revision.Fri, Jul 17, 4:37 PM
Harbormaster completed remote builds in B74890: Diff 182143.Fri, Jul 17, 4:38 PM
bnovkov updated this revision to Diff 182145.Fri, Jul 17, 4:39 PM

Account for EINPROGRESS

Harbormaster completed remote builds in B74892: Diff 182145.Fri, Jul 17, 4:39 PM
bz added a subscriber: bz.Fri, Jul 17, 4:42 PM

Please see the wireless mailing list where I sent a patch for test and did not get feedback. Changing the driver is the wrong thing to do (not that it works well currently).

https://lists.freebsd.org/archives/freebsd-wireless/2026-July/004328.html

bnovkov abandoned this revision.Fri, Jul 17, 8:50 PM
In D58306#1337245, @bz wrote:

Please see the wireless mailing list where I sent a patch for test and did not get feedback. Changing the driver is the wrong thing to do (not that it works well currently).

https://lists.freebsd.org/archives/freebsd-wireless/2026-July/004328.html

Ah, I wasn't aware of your patch, thanks for pointing this out.
FWIW your patch fixes the message spam on my machine.
I'll abandon this revision so please feel free to commit the patch you proposed.

bz added a comment.Fri, Jul 17, 9:09 PM
In D58306#1337381, @bnovkov wrote:
In D58306#1337245, @bz wrote:

Please see the wireless mailing list where I sent a patch for test and did not get feedback. Changing the driver is the wrong thing to do (not that it works well currently).

https://lists.freebsd.org/archives/freebsd-wireless/2026-July/004328.html

Ah, I wasn't aware of your patch, thanks for pointing this out.
FWIW your patch fixes the message spam on my machine.
I'll abandon this revision so please feel free to commit the patch you proposed.

Thanks a lot for testing.
Just committed it.

PS: please feel free to let me know any new issues (on wireless@, by PR, or to bz@).
With rtw8x there's a couple of open problems, so chances are there is an open PR or a previous posting to the mailing list.

Revision Contents
Changeset List

PathSize
sys/
contrib/
dev/
rtw89/
2 lines

Diff 182145

View Options

sys/contrib/dev/rtw89/core.c

Loading...