FS corruption on Dell Inspiron 8k w/ IBM-DJSA-220

FS corruption on Dell Inspiron 8k w/ IBM-DJSA-220

[Emil Eifrem]

I've had quite a lot of problems with my Dell Inspiron 8000 laptop under
Linux, with random filesystem corruption and arbitrary kernel panics on (I
believe) high disk traffic. I have an IBM DJSA-220 20 gig harddisk:

/dev/hda:

 Model=IBM-DJSA-220, FwRev=JS4OAC3A, SerialNo=44K44SH8203
 Config={ HardSect NotMFM HdSw>15uSec Fixed DTR>10Mbs }
 RawCHS=16383/16/63, TrkSize=0, SectSize=0, ECCbytes=4
 BuffType=DualPortCache, BuffSize=1874kB, MaxMultSect=16, MultSect=16
 CurCHS=16383/16/63, CurSects=16514064, LBA=yes, LBAsects=39070080
 IORDY=on/off, tPIO={min:240,w/IORDY:120}, tDMA={min:120,rec:120}
 PIO modes: pio0 pio1 pio2 pio3 pio4 
 DMA modes: mdma0 mdma1 mdma2 udma0 udma1 udma2 
 AdvancedPM=yes: mode=0x80 (128) WriteCache=enabled
 Drive Supports : ATA/ATAPI-5 T13 1321D revision 1 : ATA-2 ATA-3 ATA-4 ATA-5 

I think other people have problems with similar setups, as recorded for
example in the thread starting here:

	http://www.uwsg.iu.edu/hypermail/linux/kernel/0110.0/0627.html

Anyway, I just got a "Kernel BUG" dump so I figured since this problem
seems to have stuck with us for a while I might as well report it here and
see if anyone might wanna make a go at it. I'm running the kernel that
ships with RH 7.3, reportedly 2.4.18-3. I was in runlevel 1 with basically
no other user processes running (it happened during a partition transfer),
which is good except that I couldn't copy and paste the dump to file. Thus,
the log below has been typed manually by me -- I was quite careful but
there may be typos.

--------------------------[ cut here ] ----------------------
Kernel BUG at page_alloc.c:117!
invalid operand: 0000
maestro3 ac97_codec soundcore agpgart NVdriver binfmt_misc autofs eepro100 ipc
CPU:	0
EIP:	0010:[<c01316e7>]	Tainted: P
EFLAGS:	00010282

EIP is at __free_pages_ok [kernel] 0x57 (2.4.18-3)
eax: 00000020	ebx: c12c4a38	ecx: 00000001	edx: 00002512
esi: 00000000	edi: c02c473c	ebp: 00000000	esp: c139bf58
ds: 0018	es: 0018	ss: 0018
Process kswapd (pid: 5, stackpage=c139b000)

Stack: [snipped]

Call Trace: [<c013d0e3>] try_to_free_buffers [kernel] 0xb3
[<c013b23a>] try_to_release_page [kernel] 0x3a
[<c012f164>] drop_page [kernel] 0x34
[<c0130726>] refill_inactive_zone [kernel] 0x206
[<c0131090>] kswapd [kernel] 0x280
[<c0105000>] stext [kernel] 0x0
[<c0107136>] kernel_thread [kernel] 0x26
[<c0130e10>] kswapd [kernel] 0x0

Code: 0f 0b 5d 58 8b 3d f0 e2 32 c0 89 d8 29 f8 69 c0 b7 6d db b6


Thanks in advance,

- - -
Emil Eifrem   [ emil@eifrem.com || www.eifrem.com || www.javamud.org ]

Re: FS corruption on Dell Inspiron 8k w/ IBM-DJSA-220

[Alan Cox]

On Thu, 2002-07-25 at 12:33, Emil Eifrem wrote:

> maestro3 ac97_codec soundcore agpgart NVdriver binfmt_misc autofs eepro100 ipc
> CPU:	0
> EIP:	0010:[<c01316e7>]	Tainted: P
> EFLAGS:	00010282

Please replicate the problem from a boot where you have never loaded
binary only drivers like NVdriver. That way we might actually be able to
try and debug it.

So I'd say get rid of NVdriver for a bit, force an fsck to check the
disk is clean then duplicate the corruption/crash and report it

Re: FS corruption on Dell Inspiron 8k w/ IBM-DJSA-220

[Emil Eifrem]

On 25 Jul 2002, Alan Cox wrote:
> Please replicate the problem from a boot where you have never loaded
> binary only drivers like NVdriver. That way we might actually be able to
> try and debug it.
> 
> So I'd say get rid of NVdriver for a bit, force an fsck to check the
> disk is clean then duplicate the corruption/crash and report it

I upgraded to the latest RH 7.3 kernel (as per Bill's suggestion), removed
all NVidia drivers, booted up in runlevel 1 and tried to reproduce the
problem. After a couple of hours of copying 1 gig files between
partitions, it hasn't crashed. This could be thanks to a bugfix that went
into 2.4.18-5 (as opposed to 2.4.18-3), but I wouldn't count on it.

The problems I've experienced so far (crashes, fs corruption, etc) have
been irregular at best and sometimes when I've made a supposed fix (such
as disabling DMA), ran some test scripts for several hours which passed
without a problem -- only to get a crash or corruption a week later.

Anyway, I'll let the laptop shuffle files overnight and drop you guys a 
report if it does crash.

-EE