[Bug 281560] gve (4) uma deadlock during high tcp throughput

From: <bugzilla-noreply_at_freebsd.org>
Date: Thu, 26 Sep 2024 23:04:18 UTC
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=281560

--- Comment #13 from shailend@google.com ---
(In reply to Andrew Gallatin from comment #12)

Hmmm interesting. In this case though, I'm sure nothing is traversing the
networking stack, and no cpu is being consumed. The offending thread seems to
be in a weird state, almost like it went awol after it got that ipi.  

```
db> show thread 100719
Thread 100719 at 0xfffff800b87ca000:
 proc (pid 857): 0xfffffe018a340580
 name: iperf
 pcb: 0xfffff800b87ca520
 stack: 0xfffffe01a4e89000-0xfffffe01a4e8cfff
 flags: 0x5  pflags: 0x100
 state: RUNQ
 priority: 4
 container lock: sched lock 11 (0xfffffe001fd73440)
 last voluntary switch: 55821.710 s ago
 last involuntary switch: 55821.710 s ago
```

-- 
You are receiving this mail because:
You are the assignee for the bug.