kern/56925: Amd64 server randomly panics

[joel.bertrand%systella.fr@localhost]

>Number:         56925
>Category:       kern
>Synopsis:       Amd64 server randomly panics
>Confidential:   no
>Severity:       critical
>Priority:       high
>Responsible:    kern-bug-people
>State:          open
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Thu Jul 14 09:45:00 +0000 2022
>Originator:     BERTRAND Joël
>Release:        NetBSD 9.2_STABLE
>Organization:
>Environment:
NetBSD legendre.systella.fr 9.2_STABLE NetBSD 9.2_STABLE (CUSTOM) #12: Mon Jun 20 08:30:25 CEST 2022 root%legendre.systella.fr@localhost:/usr/src/netbsd-9/obj/sys/arch/amd64/compile/CUSTOM amd64

>Description:
Hello,

        My main server runs NetBSD 9.2 as rolling release. Since NetBSD 9.0, it randomly panics with traceback like :

[ 613880.906790] S-1C-1: ccb_timeout: num=1 total=1 disp=0
[ 613880.906790] uvm_fault(0xffffffff815862c0, 0x0, 2) -> e
[ 613880.906790] fatal page fault in supervisor mode
[ 613880.906790] trap type 6 code 0x2 rip 0xffffffff8022800c cs 0x8 rflags
0x10246 cr2 0x10 ilevel 0 rsp 0xffffdb813c0c9f48
[ 613880.906790] curlwp 0xfffffda14406ba60 pid 0.181 lowest kstack
0xffffdb813c0c72c0
[ 613880.906790] panic: trap
[ 613880.906790] cpu1: Begin traceback...
[ 613880.906790] vpanic() at netbsd:vpanic+0x160
[ 613880.906790] snprintf() at netbsd:snprintf
[ 613880.906790] startlwp() at netbsd:startlwp
[ 613880.906790] alltraps() at netbsd:alltraps+0xbb
[ 613880.916794] ccb_timeout() at netbsd:ccb_timeout+0xf0
[ 613880.916794] iscsi_cleanup_thread() at netbsd:iscsi_cleanup_thread+0x2b6
[ 613880.916794] cpu1: End traceback...
[ 613880.916794] uvm_fault(0xfffffda43aad15c0, 0x0, 2) -> e
[ 613880.916794] fatal page fault in supervisor mode
[ 613880.916794] trap type 6 code 0x2 rip 0xffffffff8097d518 cs
[ 613880.916794]  0x8 rflags 0x10246 cr2 0x84 ilevel 0x8 rsp
0xffffdb8141e16cd9u0m
[ 613880.916794] ping to dev 18,1 (offset=251919, size=4162814):
[ 613880.916794] curlwp 0xfffffda3854184c0 pid 25426.1 lowest kstacdku m0px
ffffdb8141e142c0

        It doesn't write dump file in /var/crash or enter in ddb. It only reboots. I'm pretty sure I haven't seen this kind of panic with NetBSD 8.x.

        This server acts as initiator and target iSCSI. Panic randomly occurs (last time was in may and I would believe this issue was fixed...) and can occur when server is idle.

>How-To-Repeat:
Issue seems to be related to iscsi initiator (target is provided by istgt daemon). But I don't know how trigger this bug.

>Fix:
No idea.