Date: Tue, 08 Feb 2005 21:55:06 +0100 From: Pawel Worach <pawel.worach@telia.com> To: Pawel Worach <pawel.worach@telia.com> Cc: current@freebsd.org Subject: Re: panic in recent 6-current Message-ID: <4209272A.5010101@telia.com> In-Reply-To: <42090E32.9020506@telia.com> References: <20050208140419.GA15236@stud.fit.vutbr.cz> <42090E32.9020506@telia.com>
next in thread | previous in thread | raw e-mail | index | archive | help
Pawel Worach wrote:
> Divacky Roman wrote:
>
>> #22 0xc0493087 in softclock (dummy=0x0) at atomic.h:365
>> #23 0xc04768a3 in ithread_loop (arg=0xc1577480)
>> at /usr/src/sys/kern/kern_intr.c:546
>> #24 0xc0475be9 in fork_exit (callout=0xc0476761 <ithread_loop>,
>> arg=0xc1577480, frame=0xd3fc3d48) at
>> /usr/src/sys/kern/kern_fork.c:790
>> #25 0xc057ef5c in fork_trampoline () at
>> /usr/src/sys/i386/i386/exception.s:208
>>
>
> I have seen the same thing but only these four frames in DDB, happened
> during
> boot. I'll see if I can get a dump with it, my kernel is built with -O.
>
It occurred right after enabling pf(4) on boot.
(kgdb) printf "%s", msgbufp->msg_ptr
...
<118>Feb 8 21:08:14 <kern.info> cookie kernel: pflog0: promiscuous mode enabled
pf: started
<118>pf enabled
altq: started
Fatal trap 12: page fault while in kernel mode
fault virtual address = 0x1c
---Type <return> to continue, or q <return> to quit---
fault code = supervisor write, page not present
instruction pointer = 0x8:0xc05222b7
stack pointer = 0x10:0xe32f4c94
frame pointer = 0x10:0xe32f4cd4
code segment = base 0x0, limit 0xfffff, type 0x1b
= DPL 0, pres 1, def32 1, gran 1
processor eflags = interrupt enabled, resume, IOPL = 0
current process = 28 (swi4: clock sio)
(kgdb) bt
#0 doadump () at pcpu.h:159
#1 0xc046dbf5 in db_fncall (dummy1=0, dummy2=0, dummy3=1999, dummy4=0xe32f4a84
" \uffffn\uffff\f") at /usr/src/sys/ddb/db_command.c:531
#2 0xc046d982 in db_command (last_cmdp=0xc06eb424, cmd_table=0x0,
aux_cmd_tablep=0xc06be500, aux_cmd_tablep_end=0xc06be504)
at /usr/src/sys/ddb/db_command.c:349
#3 0xc046da95 in db_command_loop () at /usr/src/sys/ddb/db_command.c:455
#4 0xc046fc15 in db_trap (type=12, code=0) at /usr/src/sys/ddb/db_main.c:221
#5 0xc0531a57 in kdb_trap (type=0, code=0, tf=0xe32f4c54) at
/usr/src/sys/kern/subr_kdb.c:421
#6 0xc0673b04 in trap_fatal (frame=0xe32f4c54, eva=0) at
/usr/src/sys/i386/i386/trap.c:801
#7 0xc06737f2 in trap_pfault (frame=0xe32f4c54, usermode=0, eva=28) at
/usr/src/sys/i386/i386/trap.c:724
#8 0xc0673350 in trap (frame=
{tf_fs = -483459048, tf_es = -1068367856, tf_ds = 16, tf_edi = 4, tf_esi
= 6, tf_ebp = -483439404, tf_isp = -483439488, tf_ebx = 0, tf_edx = -701736880,
tf_ecx = 21969, tf_eax = -1037546256, tf_trapno = 12, tf_err = 2, tf_eip =
-1068358985, tf_cs = 8, tf_eflags = 66118, tf_esp = 0, tf_ss = 800582919}) at
/usr/src/sys/i386/i386/trap.c:414
#9 0xc0663e4a in calltrap () at /usr/src/sys/i386/i386/exception.s:139
#10 0xe32f0018 in ?? ()
#11 0xc0520010 in thread_exit () at /usr/src/sys/kern/kern_thread.c:615
#12 0xc04f9978 in ithread_loop (arg=0xc227a480) at /usr/src/sys/kern/kern_intr.c:546
#13 0xc04f887f in fork_exit (callout=0xc04f97d0 <ithread_loop>, arg=0x0,
frame=0x0) at /usr/src/sys/kern/kern_fork.c:790
#14 0xc0663eac in fork_trampoline () at /usr/src/sys/i386/i386/exception.s:208
(kgdb) l *0xc05222b7
0xc05222b7 is in softclock (atomic.h:154).
149 atomic.h: No such file or directory.
in atomic.h
(kgdb) frame 12
#12 0xc04f9978 in ithread_loop (arg=0xc227a480) at /usr/src/sys/kern/kern_intr.c:546
546 ih->ih_handler(ih->ih_argument);
(kgdb) list
541 mtx_unlock(&ithd->it_lock);
542 goto restart;
543 }
544 if ((ih->ih_flags & IH_MPSAFE) == 0)
545 mtx_lock(&Giant);
546 ih->ih_handler(ih->ih_argument);
547 if ((ih->ih_flags & IH_MPSAFE) == 0)
548 mtx_unlock(&Giant);
549 }
550
(kgdb) print *ih
$3 = {ih_handler = 0xc0522000 <softclock>, ih_argument = 0x0, ih_flags =
-2147483648, ih_name = 0xc06b998b "clock",
ih_ithread = 0xc227a480, ih_need = 1, ih_next = {tqe_next = 0xc23e7940,
tqe_prev = 0xc227a4b0}, ih_pri = 52 '4'}
Frame #10 is strange, I'm fairly certain it was something like
"softclock(0,0,0,0,0)" in the ddb trace unfortunetly IBM decided to put a
parallel port instead of a serial one on my ThinkPad so I can't capture
that easily. Kernel was built with -O according to
0>me@cookie /usr/obj/usr/src/sys/IBMT41> make -V CFLAGS
-O -pipe -march=pentium2 -Wall ...
config has SCHED_4BSD, PREEMPTION and ADAPTIVE_GIANT, no SMP, WITNESS or
INVARIANTS. mpsafe{vfs,net,vm} is 1
What else is strange is that I have set
debug.debugger_on_panic=0
debug.trace_on_panic=1
in sysctl.conf but the box dropped to DDB on panic, shouldn't it dump
automagically with debug.debugger_on_panic set to 0 ?
--
Pawel
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?4209272A.5010101>