Skip to contentSkip to conversation

Coda and Ext3

DISCUSSION Sujal Shah 1 participant 1 message Sep 19, 2001 1 message not in the archive

lore
Share & tools
lore

A message that is not in the archive (sent off-list or not yet received): [email protected]

Sujal Shah

author
#1 lore raw

Hi everyone,

The Linux Coda drivers and the ext3 patches don't seem to get along
very well, at least in Linux 2.4.7. I've got a stock 2.4.7 kernel with a patch applied to the USB drivers (for a sony digital camera; see
http://www.sujal.net/tech/linux/ just a change in unusual_devs.h).

After I applied the ext3 patches from
http://www.uow.edu.au/~andrewm/linux/ext3/ . Basically, when an application tries to write to a file system mounted via coda, the application terminates with "Memory Fault" returned to the terminal. THe file system still thinks it's busy (can't umount).

I'm using the ext3-2.4-0.9.5-247 patch. I have not yet tried a newer kernel.

THe funny thing is that mounting/unmount works fine as long as I don't try to write. Also, I can create files via touch or mkdir without causing problems. The following log snippet shows me mounting then immediately unmounting the filesystem, then remounting and trying a write operation (using cp).

Just to let everyone know, I am running my own replacement for venus (something similar to podfuk/uservfs), and I do have the NVidia drivers loaded. I was able to recreate this bug without the NVdriver module loaded, however. Also, I backed out the patches for ext3 and the problem went away.

THe following messages appear in my syslog:

Sep 18 19:02:00 pcsshah kernel: Coda Kernel/Venus communications, v5.3.14, [email protected]
Sep 18 19:02:16 pcsshah kernel: coda_read_super: Bad mount version
Sep 18 19:02:16 pcsshah kernel: coda_read_super: device index: 0
Sep 18 19:02:16 pcsshah kernel: coda_read_super: rootfid is (0x1234567,0xffffffff,0x1)
Sep 18 19:02:16 pcsshah kernel: coda_read_super: rootinode is 1450180609 dev 7
Sep 18 19:03:04 pcsshah kernel: Coda: Bye bye.
Sep 18 19:03:50 pcsshah gconfd (sujal-1184): 21 items remain in the cache after cleaning already-synced items older than 300 seconds
Sep 18 19:04:34 pcsshah kernel: coda_read_super: Bad mount version
Sep 18 19:04:34 pcsshah kernel: coda_read_super: device index: 0
Sep 18 19:04:34 pcsshah kernel: coda_read_super: rootfid is (0x1234567,0xffffffff,0x1)
Sep 18 19:04:34 pcsshah kernel: coda_read_super: rootinode is 1450180609 dev 7
Sep 18 19:05:02 pcsshah kernel: kernel BUG at file.c:45!
Sep 18 19:05:02 pcsshah kernel: invalid operand: 0000
Sep 18 19:05:02 pcsshah kernel: CPU:    0
Sep 18 19:05:02 pcsshah kernel: EIP:    0010:[<e59a73df>]
Sep 18 19:05:02 pcsshah kernel: EFLAGS: 00210282
Sep 18 19:05:02 pcsshah kernel: eax: 00000019   ebx: d403e1a0   ecx: 00000006   edx: 00000000
Sep 18 19:05:02 pcsshah kernel: esi: ffffffea   edi: d3eaf240   ebp: d3f71de0   esp: d36f3f60
Sep 18 19:05:02 pcsshah kernel: ds: 0018   es: 0018   ss: 0018
Sep 18 19:05:02 pcsshah kernel: Process cp (pid: 1650, stackpage=d36f3000)
Sep 18 19:05:02 pcsshah kernel: Stack: e59acb2c e59acc83 0000002d d3f71de0 ffffffea 00000000 00000400 c0130f96 
Sep 18 19:05:02 pcsshah kernel:        d3f71de0 bfffef10 00000400 d3f71e00 00000000 0009bce9 00000000 d3a78b60 
Sep 18 19:05:02 pcsshah kernel:        d383ede0 00000000 d3f719c0 bffff308 d36f2000 00000400 bfffef10 bfffeef8 
Sep 18 19:05:02 pcsshah kernel: Call Trace: [sys_write+150/208] [system_call+51/56] 
Sep 18 19:05:02 pcsshah kernel: 
Sep 18 19:05:02 pcsshah kernel: Code: 0f 0b 83 c4 0c 8b 43 08 8b 70 08 8d 5e 5c 89 d9 ff 4e 5c 0f 

Just FYI.

Thanks,

Sujal

Signature · 5 lines
---- Sujal Shah --- [email protected] ---

        http://www.sujal.net

Now Playing: George Michael - Freedom 90