Re: [2.2.13-pre15] null pointer exception in d_lookup

Jarno Paananen (jpaana@s2.org)
05 Oct 1999 18:29:07 +0300


[I'm Cc:ing this to Alan Cox too as he might know something
about this, see my question below]

Peter Hanecak <hanecak@megaloman.com> writes:

| i got an kernel oops in d_lookup

This is very interesting, because I got this same oops in
2.3.18ac10 (yes, in 2.3 kernel and yours is 2.2)

The routine looks about the same so this is very possibly the
same bug.

|
| [2.]
|
| when running image conversion (100% CPU usage, about 2-3 MB file copying
| from local CD-ROM to local ext2 fs on SCSI drive in about every 3-5 minutes
| then imagemagick's mogrify activity over copied file (again on same local
| ext2 SCSI filesystem)) i experienced kernel problems: OOPs in d_lookup
| (see below log entries).

I got it while compiling X with make -j3 (SMP system) so there
is also heavy disk activity)

There are some similarities too as I also have SCSI system with
Adaptec 2940U2W adapter and IBM U2W disk.

| no such error occured during usage of previous 2.2.13-preX kernels (tried
| pre8, pre9, pre11 and pre14). also nothing similar when using 2.2.12 and
| earlier (this do not means i did not encounter no errors at all :) - see my
| previous posts regarding instant freezes).

I am now running 2.2.13-pre14 (haven't had the time to compile
pre15) and it works just fine too. Alan, are there any backports
from 2.3 that might cause this?

| Linux version 2.2.13pre15 (root@m1.megaloman.sk) (gcc version egcs-2.91.66 \
| 19990314/Linux (egcs-1.1.2 release)) #3 Tue Oct 5 10:25:44 EDT 1999

Mine is 2.3.18ac10 compiled with gcc 2.7.2.3.

| Oct 5 11:47:03 m1 kernel: Unable to handle kernel paging request at virtual address fc478b00
[snip]
| Oct 5 11:47:03 m1 kernel: esi: 00000000 edi: c3b97026 ebp: fc478b00 esp: c1999f2c

I had ebp 0, but the same register caused the oops...

| 0: 8b 6d 00 movl 0x0(%ebp),%ebp

In this same instruction...

| image conversion program (probably) - i.e. intensive operations on ext2
| filesystem?
|
| when running connversion, error occured about 10 minutes after start of
| conversion. but when running conversion second time (after reboot), to see
| whether i can reproduce this crash, after about 4 hour is system still up
| (and converting) without any oops messages.

Yes, my oops also occured after some time of heavy activity
(compiling X takes about 20-25 minutes on my machine now and it
was about half done)

// Jarno

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/