Hello...

* Patrick Welche (prlw1@newn.cam.ac.uk) [000530 00:32]:
| I think you want to try:
| target kcore netbsd.6.core
| (or better.. find the kernel file made with 'makeoptions DEBUG="-g"'...)

Yepp, both helped. Thanks.

| > I have no idea what triggered the panic.
| Hopefully the above will tell us..

Not really. But I hope someone else might make sense out of it. Here
is a backtrace from the core file:

(gdb) bt
#0  0xc028b9c8 in db_last_command ()
#1  0x46fa000 in ?? ()
#2  0xc01dff27 in cpu_reboot (howto=260, bootstr=0x0)
    at /usr/cvs.netbsd/src/sys/arch/i386/i386/machdep.c:1119
#3  0xc011cc3a in db_reboot_cmd ()
    at /usr/cvs.netbsd/src/sys/ddb/db_command.c:648
#4  0xc011c924 in db_command (last_cmdp=0xc028b9c8, cmd_table=0xc028b7e8)
    at /usr/cvs.netbsd/src/sys/ddb/db_command.c:317
#5  0xc011cac6 in db_command_loop ()
    at /usr/cvs.netbsd/src/sys/ddb/db_command.c:552
#6  0xc011f752 in db_trap (type=1, code=0)
    at /usr/cvs.netbsd/src/sys/ddb/db_trap.c:78
#7  0xc01ddbf0 in kdb_trap (type=1, code=0, regs=0xc9c28d68)
    at /usr/cvs.netbsd/src/sys/arch/i386/i386/db_interface.c:119
#8  0xc01e4ac4 in trap (frame={tf_es = 1660944400, tf_ds = 959709200,
      tf_edi = 256, tf_esi = -1071311202, tf_ebp = -909996640,
      tf_ebx = -909996608, tf_edx = 0, tf_ecx = 0, tf_eax = 0, tf_trapno = 1,
      tf_err = 0, tf_eip = -1071784800, tf_cs = 8, tf_eflags = 514,
      tf_esp = -909996620, tf_ss = -1072417163, tf_vm86_es = 0,
      tf_vm86_ds = 1, tf_vm86_fs = -910912460, tf_vm86_gs = -909996560})
    at /usr/cvs.netbsd/src/sys/arch/i386/i386/trap.c:296
#9  0xc0100d0d in calltrap ()
#10 0xc0143675 in panic (fmt=0xc025169e "vinvalbuf: dirty bufs, vp %p")
    at /usr/cvs.netbsd/src/sys/kern/subr_prf.c:219
#11 0xc015b51d in vinvalbuf (vp=0xc9b49434, flags=1, cred=0xc0672700,
    p=0xc9cdc7e0, slpflag=0, slptimeo=0)
    at /usr/cvs.netbsd/src/sys/kern/vfs_subr.c:620
#12 0xc016b295 in spec_close (v=0xc9c28e6c)
    at /usr/cvs.netbsd/src/sys/miscfs/specfs/spec_vnops.c:612
#13 0xc01cabc3 in ufsspec_close (v=0xc9c28e6c)
    at /usr/cvs.netbsd/src/sys/ufs/ufs/ufs_vnops.c:1752
#14 0xc01628d3 in vn_close (vp=0xc9b49434, flags=3, cred=0xc0672700,
    p=0xc9cdc7e0) at /usr/cvs.netbsd/src/sys/sys/vnode_if.h:171
#15 0xc01632cf in vn_closefile (fp=0xc9baef04, p=0xc9cdc7e0)
    at /usr/cvs.netbsd/src/sys/kern/vfs_vnops.c:605
#16 0xc0132594 in closef (fp=0xc9baef04, p=0xc9cdc7e0)
    at /usr/cvs.netbsd/src/sys/kern/kern_descrip.c:1090
#17 0xc013236e in fdfree (p=0xc9cdc7e0)
    at /usr/cvs.netbsd/src/sys/kern/kern_descrip.c:965
#18 0xc01338de in exit1 (p=0xc9cdc7e0, rv=0)
    at /usr/cvs.netbsd/src/sys/kern/kern_exit.c:185
#19 0xc01337e4 in sys_exit (p=0xc9cdc7e0, v=0xc9c28f88, retval=0xc9c28f80)
    at /usr/cvs.netbsd/src/sys/kern/kern_exit.c:138
#20 0xc01e513f in syscall (frame={tf_es = 31, tf_ds = 31, tf_edi = 134643176,
      tf_esi = 0, tf_ebp = -1077946588, tf_ebx = -1, tf_edx = 134595564,
      tf_ecx = 0, tf_eax = 1, tf_trapno = 3, tf_err = 2, tf_eip = 134602271,
      tf_cs = 23, tf_eflags = 598, tf_esp = -1077946608, tf_ss = 31,
      tf_vm86_es = 0, tf_vm86_ds = 0, tf_vm86_fs = 0, tf_vm86_gs = 0})
    at /usr/cvs.netbsd/src/sys/arch/i386/i386/trap.c:766
#21 0xc0100db9 in syscall1 ()
can not access 0xbfbfd724, invalid translation (invalid PDE)
can not access 0xbfbfd724, invalid translation (invalid PDE)
Cannot access memory at address 0xbfbfd724.


* Matt Thomas (matt@3am-software.com) [000530 00:33]:
| At 11:01 PM 5/29/2000 +0100, Patrick Welche wrote:
| >On Mon, May 29, 2000 at 11:32:25PM +0200, Bjoern Labitzke wrote:
| > > My X11 just hung. I did a reboot(0x104) and got a dump. I thought
| > > I should be able with "gdb netbsd.6 netbsd.6.core" to obtain a
| > > trace to provide for finding the problem. But gdb told me
| > >
| > > /var/crash/netbsd.6.core" is not a core dump: File format not recognized
| 
| BTW, I ran into this too and a bunch of fixes were committed to
| fix panics in this area.  They solved my panics.

As I explained in the mail, I am running -current build this very
morning with the sources updated from AnonCVS right before that. But
if the problems vanish (not that I had many of those before), I
wouldn't complain. ;-)


Thanks for the answers,

Bjoern

-- 
Bjoern Labitzke  <hermit@cs.tu-berlin.de>
   Use PGP! (Don't you use envelopes for your letters?)