Skip to contentSkip to conversation

"kernel BUG at fs/nfs/inode.c:152!"

FYI Ian Kumlien 2 participants 3 messages Sep 6, 2004

lore
Share & tools
lore
#0 lore raw

Hi,

When my networking card died yesterday (sundance based d-link card) i used dhclient to update the route. Unfortunately it caused it to change ip, thus today i had some NFS problems...

fstab entry:
blue:/mnt/large /mnt/large nfs
user,async,soft,rsize=32768,wsize=32768,tcp 0 0

Anyways, i did a bunch of umount -f's and so on and so forth as i have done a million times before, but when i tried remounting it, nothing happened... A quick ps aux |grep revealed:

root 26793 0.0 0.0 1740 696 ? D 14:23 0:00
/bin/umount /mnt/large
root 26800 0.0 0.0 1868 828 ? D 14:23 0:00
/bin/mount /mnt/large
root 26805 0.0 0.0 1868 828 ? D 14:23 0:00
/bin/mount /mnt/large
root 26807 0.0 0.0 1868 828 ? D 14:23 0:00
/bin/mount /mnt/large
root 26809 0.0 0.0 1868 828 ? D 14:23 0:00
/bin/mount /mnt/large

And to further my agony a dmesg revealed:

------------[ cut here ]------------
kernel BUG at fs/nfs/inode.c:152!
invalid operand: 0000 [#1]
Modules linked in: nls_cp437 vfat fat floppy nvidia twofish serpent blowfish sha256 crypto_null ip6table_filter ip6_tables forcedethCPU:    0
EIP:    0060:[<c01a8bb0>]    Tainted: P   VLI
EFLAGS: 00210286   (2.6.9-rc1-bk7)

EIP is at nfs_clear_inode+0x50/0x70
eax: ffffffff ebx: f70595a8 ecx: edc36c60 edx: f710cf60
esi: f7059490 edi: 000000ce ebp: c042792c esp: c05cbf04
ds: 007b es: 007b ss: 0068
Process umount (pid: 26773, threadinfo=c05ca000 task=df308bb0)

Stack: 00000002 f70595a8 c05cbf2c c016058d f70595a8 c01605dc 00000000 c1bff64c
       c1bff600 c0160736 c05cbf2c c05cbf2c c1bff600 c1bff64c c042a460 c05ca000
       c01506f5 ef95c000 00000015 c042a5a0 08050278 c0150dd9 f7e8cc00 c01ab2dc
Call Trace:
 [<c016058d>] clear_inode+0x8d/0xa0
 [<c01605dc>] dispose_list+0x3c/0x70
 [<c0160736>] invalidate_inodes+0x66/0x80
 [<c01506f5>] generic_shutdown_super+0x55/0x100
 [<c0150dd9>] kill_anon_super+0x9/0x20
 [<c01ab2dc>] nfs_kill_super+0xc/0x70
 [<c0150623>] deactivate_super+0x43/0x60
 [<c0162e4b>] sys_umount+0x3b/0x90
 [<c03a693d>] schedule+0x27d/0x450
 [<c0103db9>] sysenter_past_esp+0x52/0x71
Code: d4 39 43 d4 75 2f 8b 86 b4 00 00 00 85 c0 74 05 e8 f6 f8 1e 00 8b 86 ac 00 00 00 85 c0 75 0c 8b 5c 24 04 8b 74 24 08 83 c4 0c c3 <0f> 0b 98 00 7a 28 3c c0 eb ea 0f 0b 94 00 7a 28 3c c0 eb c7 8d

This is, as stated before, just FYI... Anyone feeling like making nvidia comments can read this and just skip mailing "/me no care".

CC, Since i'm not subbed.

Signature · 1 line
Ian Kumlien <pomac () vapor ! com> -- http://pomac.netswarm.net
#1 lore raw

Hello Ian,

This looks very much like the problem in iget() I described today in LKML. Hope I'll post a patch here soon.

Kirill

Quoted 55 lines

When my networking card died yesterday (sundance based d-link card) i used dhclient to update the route. Unfortunately it caused it to change ip, thus today i had some NFS problems...

fstab entry:
blue:/mnt/large /mnt/large nfs user,async,soft,rsize=32768,wsize=32768,tcp 0 0

Anyways, i did a bunch of umount -f's and so on and so forth as i have done a million times before, but when i tried remounting it, nothing happened... A quick ps aux |grep revealed:

root 26793 0.0 0.0 1740 696 ? D 14:23 0:00
/bin/umount /mnt/large
root 26800 0.0 0.0 1868 828 ? D 14:23 0:00
/bin/mount /mnt/large
root 26805 0.0 0.0 1868 828 ? D 14:23 0:00
/bin/mount /mnt/large
root 26807 0.0 0.0 1868 828 ? D 14:23 0:00
/bin/mount /mnt/large
root 26809 0.0 0.0 1868 828 ? D 14:23 0:00
/bin/mount /mnt/large

And to further my agony a dmesg revealed:

------------[ cut here ]------------
kernel BUG at fs/nfs/inode.c:152!
invalid operand: 0000 [#1]
Modules linked in: nls_cp437 vfat fat floppy nvidia twofish serpent blowfish sha256 crypto_null ip6table_filter ip6_tables forcedethCPU:    0
EIP:    0060:[<c01a8bb0>]    Tainted: P   VLI
EFLAGS: 00210286   (2.6.9-rc1-bk7)

EIP is at nfs_clear_inode+0x50/0x70
eax: ffffffff ebx: f70595a8 ecx: edc36c60 edx: f710cf60
esi: f7059490 edi: 000000ce ebp: c042792c esp: c05cbf04
ds: 007b es: 007b ss: 0068
Process umount (pid: 26773, threadinfo=c05ca000 task=df308bb0)

Stack: 00000002 f70595a8 c05cbf2c c016058d f70595a8 c01605dc 00000000 c1bff64c
       c1bff600 c0160736 c05cbf2c c05cbf2c c1bff600 c1bff64c c042a460 c05ca000
       c01506f5 ef95c000 00000015 c042a5a0 08050278 c0150dd9 f7e8cc00 c01ab2dc
Call Trace:
 [<c016058d>] clear_inode+0x8d/0xa0
 [<c01605dc>] dispose_list+0x3c/0x70
 [<c0160736>] invalidate_inodes+0x66/0x80
 [<c01506f5>] generic_shutdown_super+0x55/0x100
 [<c0150dd9>] kill_anon_super+0x9/0x20
 [<c01ab2dc>] nfs_kill_super+0xc/0x70
 [<c0150623>] deactivate_super+0x43/0x60
 [<c0162e4b>] sys_umount+0x3b/0x90
 [<c03a693d>] schedule+0x27d/0x450
 [<c0103db9>] sysenter_past_esp+0x52/0x71
Code: d4 39 43 d4 75 2f 8b 86 b4 00 00 00 85 c0 74 05 e8 f6 f8 1e 00 8b 86 ac 00 00 00 85 c0 75 0c 8b 5c 24 04 8b 74 24 08 83 c4 0c c3 <0f> 0b 98 00 7a 28 3c c0 eb ea 0f 0b 94 00 7a 28 3c c0 eb c7 8d

This is, as stated before, just FYI... Anyone feeling like making nvidia comments can read this and just skip mailing "/me no care".

CC, Since i'm not subbed.

#2 lore raw
Quoted 4 lines · Kirill Korotaev

Hello Ian,

This looks very much like the problem in iget() I described today in LKML. Hope I'll post a patch here soon.

ahhhh, ok, Must have missed it etc, though i only read lkml trough marc.theaimsgroup.com =P

I'm looking forward to seeing a patch included =)

Signature · 1 line
Ian Kumlien <pomac () vapor ! com> -- http://pomac.netswarm.net