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

https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=281560

--- Comment #9 from [email protected] ---
(In reply to Mark Johnston from comment #7)

Also the trace for the uma zone lock holding iperf thread:

```
db> trace 100719
Tracing pid 857 tid 100719 td 0xfffff800b87ca000
sched_switch() at sched_switch+0x5d2/frame 0xfffffe01a4e8c720
mi_switch() at mi_switch+0x171/frame 0xfffffe01a4e8c740
ipi_bitmap_handler() at ipi_bitmap_handler+0x86/frame 0xfffffe01a4e8c770
Xipi_intr_bitmap_handler() at Xipi_intr_bitmap_handler+0xb3/frame
0xfffffe01a4e8c770
--- interrupt, rip = 0xffffffff80ed8bed, rsp = 0xfffffe01a4e8c840, rbp =
0xfffffe01a4e8c870 ---
zone_fetch_bucket() at zone_fetch_bucket+0x8d/frame 0xfffffe01a4e8c870
cache_alloc() at cache_alloc+0x261/frame 0xfffffe01a4e8c8d0
cache_alloc_retry() at cache_alloc_retry+0x23/frame 0xfffffe01a4e8c910
tcp_m_copym() at tcp_m_copym+0x227/frame 0xfffffe01a4e8c9b0
tcp_default_output() at tcp_default_output+0x146c/frame 0xfffffe01a4e8cb80
tcp_usr_send() at tcp_usr_send+0x5a2/frame 0xfffffe01a4e8cc30
sosend_generic() at sosend_generic+0x5fe/frame 0xfffffe01a4e8cce0
sousrsend() at sousrsend+0x79/frame 0xfffffe01a4e8cd40
dofilewrite() at dofilewrite+0x7f/frame 0xfffffe01a4e8cd90
sys_write() at sys_write+0xb7/frame 0xfffffe01a4e8ce00
amd64_syscall() at amd64_syscall+0x147/frame 0xfffffe01a4e8cf30
fast_syscall_common() at fast_syscall_common+0xf8/frame 0xfffffe01a4e8cf30
```

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