Re: btrfs zero divide (was: Re: Linux 3.10 problem reports (yes, plural))

[Josef Bacik <jbacik@fusionio.com>]

On Tue, Jul 30, 2013 at 11:07:30AM +0200, Geert Uytterhoeven wrote:
> On Tue, 30 Jul 2013, Thorsten Glaser wrote:
> > NEW problem: btrfs doesn’t work at all. I had to reboot my
> > buildd into 3.2 using echo s/u/s/o >/proc/sysrq-trigger as
> > the attempt to mount it left the system hanging there.
> 
> > [    0.000000] Linux version 3.10-1-m68k (debian-kernel@lists.debian.org) (gcc version 4.8.1 (Debian 4.8.1-7+m68k.1) ) #1 Debian 3.10.3-1 (2013-07-27)
> 
> > [    6.720000] bio: create slab <bio-1> at 1
> > [    6.740000] Btrfs loaded
> > [    6.830000] device label ara5-butter devid 1 transid 376178 /dev/nfhd8p3
> > [    7.150000] EXT4-fs (nfhd8p1): mounted filesystem with ordered data mode. Opts: (null)
> > [   14.520000] udevd[228]: starting version 175
> > [   17.820000] device label ara5-butter devid 1 transid 376178 /dev/nfhd8p3
> > [   20.850000] Adding 3670012k swap on /dev/nfhd8p2.  Priority:-1 extents:1 across:3670012k
> > [   21.380000] EXT4-fs (nfhd8p1): re-mounted. Opts: (null)
> > [   31.300000] EXT4-fs (nfhd8p1): re-mounted. Opts: errors=remount-ro
> > [   38.460000] device label ara5-butter devid 1 transid 376178 /dev/nfhd8p3
> > [   38.530000] btrfs: setting nodatacow, compression disabled
> > [   38.540000] btrfs: enabling auto recovery
> > [   38.570000] btrfs: disk space caching is enabled
> > [   38.600000] *** ZERO DIVIDE ***   FORMAT=2
> > [   38.630000] Current process id is 722
> > [   38.660000] BAD KERNEL TRAP: 00000000
> > [   38.680000] Modules linked in: evdev mac_hid ext4 crc16 jbd2 mbcache btrfs xor lzo_compress zlib_deflate raid6_pq crc32c libcrc32c
> > [   38.730000] PC: [<319535b2>] __btrfs_map_block+0x11c/0x119a [btrfs]
> 
> Woops, adding the btrfs devs to CC.
> 

Can you gdb btrfs.ko and do 

list *(__btrfs_map_block+0x11c)

so I can see where this is?  I've not seen this yet, just so I'm clear this is
blowing up because we're doing

blah / 0

right?  I've looked at all the places we do divides in this function and it
doesn't look like we're doing this anywhere but I could be blind.  Thanks,

Josef