Re: Panic in g_up (bg_fsck/cdparanoia interaction issues?)

On Thu, Jul 01, 2004 at 01:00:15PM -0400, Damian Gerow wrote:
> After a crash and reboot,  I start ripping the latter half of the CD,
> trying to reproduce the panic:
> 
>     % cdparanoia -Z -B "7-14[3:21.14]"
> 
> It got 3/4 of the way through track 13 before a panic.  Note that bg_fsck
> has finished processing ad1s1d (/home), and is now working on ad1s1e
> (/usr).  I am writing to ad1s1d:
> 
>     Fatal trap 12 = page fault while in kernel mode
>     cpuid = 0; apic id = 00
> 
>     fault virtual address    = 0xa00000a
>     fault code               = supervisor read, page not present
>     instruction pointer      = 0x8:0xc067b104
>     stack pointer            = 0x10:0xd4293b98
>     frame pointer            = 0x10:0xd4293ba8
>     code segment             = base 0xa0, limit 0xfffff, type 0x1b
>                              = DPL 0, pres 1, dev32 1, gran 1
>     processor eflags         = interrupt enabled, resume, IOPL = 0
>     current process          = 3 (g_up)
>     kernel: type 12 trap, code=0
>     Stopped at      handle_allocindir_partdone+0x40:        testb   $0x2,0xa(%ecx)
>     db> panic

It seems like aip->ai_indirdep became 0xa000000 somehow.  It would be
interesting to print out the whole *aip from gdb if you can get this
to happen again and get a crashdump.

(Referencing src/sys/ufs/ffs/ffs_softdep.c:handle_allocindir_partdone())

-- 
Brian Fundakowski Feldman                           \'[ FreeBSD ]''''''''''\
  <> green_at_FreeBSD.org                               \  The Power to Serve! \
 Opinions expressed are my own.                       \,,,,,,,,,,,,,,,,,,,,,,\