Re: Null Dereference with the NSLU2 - NFS, XFS, JFS

[David Levitan <david@dlevitan.com>]

Martin Guy wrote:
> hi!
>
> 2008/1/28, David Levitan <david@dlevitan.com>:
> > Unable to handle kernel paging request at virtual address e2833d56
> > pgd = c1a08000
> > [e2833d56] *pgd=00000000
> > Internal error: Oops: f3 [#1]
>
> I'm beginning to wonder whether you have hardware problems, since
> every filesystem dies for you.
> If you can get root shell access to the box, you could try:
>    dd if=/dev/whatever of=/dev/null bs=1M
> to see whether it is simply intensive I/O that is killing the system,
> which might be an inadequate power supply.
>
>     M

No problems with direct access. I transferred 90M, 900M, and 9000M with 
no error messages.

I've also had no problems using the same USB2 enclosure, drive, and 
cable while accessing it through my laptop.

I tried uploading files through FTP onto the JFS partition. No problems 
until I tried to delete them (not through FTP) at which point JFS died 
(I've attached the message below). I have not had any problems like this 
with ext3.

To summarize:
JFS and XFS are unstable
ext3 seems to work fine except when paired with NFS4


Thanks,
David



Result of rm -fr * on JFS partition:
JFS: nTxBlock = 233, nTxLock = 1866
Unable to handle kernel NULL pointer dereference at virtual address 0000014c
pgd = c0c5c000
[0000014c] *pgd=009d6031, *pte=00000000, *ppte=00000000
Internal error: Oops: 17 [#1]
Modules linked in: jfs ledtrig_heartbeat nfs nfsd exportfs lockd nfs_acl 
sunrpc ipv6 fuse dm_snapshot dm_mirror dm_mod evdev ixp4xx_mac usblp 
ixp4xx_beeper ixp4xx_qmgr ixp4xx_npe firmware_class ext3 jbd mbcache 
sd_mod usb_storage scsi_mod ehci_hcd ohci_hcd usbcore
CPU: 0    Not tainted  (2.6.22-3-ixp4xx #1)
PC is at txCommit+0x1f4/0xf38 [jfs]
LR is at lmLog+0x184/0x18c [jfs]
pc : [<bf2149f4>]    lr : [<bf2133b0>]    psr: 20000013
sp : c1ffbe24  ip : c1ffbdfc  fp : c1ffbeac
r10: c0d338f4  r9 : c040ea44  r8 : c288f24c
r7 : 00000002  r6 : c2899180  r5 : c0cf1938  r4 : 00000000
r3 : 00000000  r2 : c0d30107  r1 : 00000106  r0 : 0000b7cc
Flags: nzCv  IRQs on  FIQs on  Mode SVC_32  Segment user
Control: 0000397f  Table: 00c5c000  DAC: 00000015
Process rm (pid: 2182, stack limit = 0xc1ffa258)
Stack: (0xc1ffbe24 to 0xc1ffc000)
be20:          c1ffbe78 00000000 c17292c0 c0088904 c155a000 00000010 
42d33137
be40: 0000002d c155a000 c02c0007 c1ffa000 c17292c0 c1f3e200 00000002 
c1ffbecc
be60: 0000005c 0000b7cc 01680800 00000000 00000010 00000002 00040080 
00000001
be80: 0000e000 c040ec5c 00000000 00000000 00000007 00000000 c040ea44 
c0d338f4
bea0: c1ffbf08 c1ffbeb0 bf1fa4cc bf21480c 00000003 c040eaa8 c0d3374c 
c040eab4
bec0: c0d336dc 479e6f85 1396f0de c040ec5c c0d338f4 0000002d c15c15a0 
00000006
bee0: 00000000 c0d338f4 c09078d4 c155a000 c0021f84 c1ffa000 00000000 
c1ffbf24
bf00: c1ffbf0c c00885dc bf1fa284 c040ec5c 00000000 c09078d4 c1ffbf94 
c1ffbf28
bf20: c008a8ac c0088508 c1847a68 c02fec20 42d33137 0000002d c155a000 
00000010
bf40: 00000000 00000000 c00439c4 c0043908 00000000 c1ffbf74 c1ffbf64 
c00439e8
bf60: 00000001 c0021d2c 40048304 c1ffa050 00000001 00000200 befd7ad8 
00000002
bf80: befd7c9a 0000000a c1ffbfa4 c1ffbf98 c008a968 c008a7ec 00000000 
c1ffbfa8
bfa0: c0021de0 c008a95c befd7ad8 00000002 befd7c9a befd7c9a 00000001 
00000002
bfc0: befd7ad8 00000002 befd7c9a 00000002 befd7c9a 00018080 00000000 
00000000
bfe0: 00017984 befd78ac 00009c38 400c8314 80000010 befd7c9a 41485352 
4f5f4c4c
Backtrace:
[<bf214800>] (txCommit+0x0/0xf38 [jfs]) from [<bf1fa4cc>] 
(jfs_unlink+0x254/0x3c4 [jfs])
[<bf1fa278>] (jfs_unlink+0x0/0x3c4 [jfs]) from [<c00885dc>] 
(vfs_unlink+0xe0/0x134)
[<c00884fc>] (vfs_unlink+0x0/0x134) from [<c008a8ac>] 
(do_unlinkat+0xcc/0x170)
 r6:c09078d4 r5:00000000 r4:c040ec5c
[<c008a7e0>] (do_unlinkat+0x0/0x170) from [<c008a968>] 
(sys_unlink+0x18/0x1c)
 r7:0000000a r6:befd7c9a r5:00000002 r4:befd7ad8
[<c008a950>] (sys_unlink+0x0/0x1c) from [<c0021de0>] 
(ret_fast_syscall+0x0/0x2c)
Code: e596200c e3833b02 e1c630b4 e592308c (e593314c)