misc/166357: [ath] 802.11n TX stall when the first _AND_ last frame in the BAW is in the software queue

[Adrian Chadd]

>Number:         166357
>Category:       misc
>Synopsis:       [ath] 802.11n TX stall when the first _AND_ last frame in the BAW is in the software queue
>Confidential:   no
>Severity:       non-critical
>Priority:       low
>Responsible:    freebsd-bugs
>State:          open
>Quarter:        
>Keywords:       
>Date-Required:
>Class:          sw-bug
>Submitter-Id:   current-users
>Arrival-Date:   Fri Mar 23 21:10:06 UTC 2012
>Closed-Date:
>Last-Modified:
>Originator:     Adrian Chadd
>Release:        10.0-CURRENT
>Organization:
>Environment:
mips24k (ar913x) (TP-WR1043ND)
>Description:
I'm seeing TX stalls after a bunch of bidirectional traffic in a lossy 802.11n environment.

Every single time, I see:

* the TID isn't scheduled;
* The software queue for that TID has the first and last seqno in the BAW in the software queue (along with other frames here and there);
* .. i _think_ the HW transmission has failed, so the frames appear again in the SWQ;
* .. but it's thus not being rescheduled.

I'll post further information soon.
>How-To-Repeat:

>Fix:


>Release-Note:
>Audit-Trail:
>Unformatted: