On Sun, May 15, 2022 at 12:06:33PM +0200, Stephan Mending wrote:
> Hi *,
> I've got a system running -current that keeps crashing on me every couple of
> days.
> Output of ddb:
>
> Connected to /dev/cuaU0 (speed 115200)
>
> ddb{0}> show panic
> the kernel did not panic
> ddb{0}> show uvm
> Current UVM status:
> pagesize=4096 (0x1000), pagemask=0xfff, pageshift=12
> 482451 VM pages: 43158 active, 132795 inactive, 35 wired, 192336 free
> (24054 z
> ero)
> min 10% (25) anon, 10% (25) vnode, 5% (12) vtext
> freemin=16081, free-target=21441, inactive-target=0, wired-max=160817
> faults=2487210, traps=2404140, intrs=211883, ctxswitch=1960560 fpuswitch=0
> softint=3499069, syscalls=2015497, kmapent=9
> fault counts:
> noram=0, noanon=0, noamap=0, pgwait=0, pgrele=0
> ok relocks(total)=192470(193514), anget(retries)=603205(0),
> amapcopy=177151
>
> neighbor anon/obj pg=82033/639788, gets(lock/unlock)=415897/193548
> cases: anon=570367, anoncow=32838, obj=347149, prcopy=67670,
> przero=1469152
>
> daemon and swap counts:
> woke=0, revs=0, scans=0, obscans=0, anscans=0
> busy=0, freed=0, reactivate=0, deactivate=0
> pageouts=0, pending=0, nswget=0
> nswapdev=1
> swpages=526020, swpginuse=0, swpgonly=0 paging=0
> kernel pointers:
> objs(kern)=0xffffffff8238a038
> ddb{0}> show trace
> No such command
> ddb{0}> trace
> icmp_mtudisc_timeout(fffffd807a50b070,0) at icmp_mtudisc_timeout+0x77
> rt_timer_timer(ffffffff8235d668) at rt_timer_timer+0x1cc
> softclock_thread(ffff8000fffff260) at softclock_thread+0x13b
> end trace frame: 0x0, count: -3
> ddb{0}>
>
This looks like some bad memory access. This is a fault and not really a
panic this is why 'show panic' returns 'the kernel did not panic'.
The crash in icmp_mtudisc_timeout() points to some error in the rttimer
code refactor I made. Please try a newer snapshot and include the dmesg.
If it happens again a call to db_show_rtentry in ddb may help better
understand what is going on:
call db_show_rtentry(fffffd807a50b070, 0, 0)
Where the first argument is derived from the first argument of
icmp_mtudisc_timeout from the trace.
--
:wq Claudio
