[uClibc]e2fsck problem: (Bad address) while doing inode scan

Giulio Orsero giulioo at pobox.com
Fri Mar 28 18:58:44 UTC 2003


On Fri, 28 Mar 2003 10:55:04 -0700, Erik Andersen <andersen at codepoet.org>
wrote:
 
>> I verified the problem persists with e2fsprogs 1.27.
>
>How big is the partition you are running fsck.ext2 on?
I posted fdisk and tune2fs data at end of 1st email, about 3.3/3.5GB
Am I right that LFS should not matter in this case? uclibc is compiled
without it as you can see from 1st email.

>How you tried getting an strace of the problem?

copied by hand (strace e2fsck  /dev/hda3)

...
...
lseek(5, 2139351040, SEEK_SET)   = 2139351040
read(5, "\0\0\0\0\0 <cut >.... \0\0\0\0\0\0\0\0\0" ... 8192 = 8192

lseek(5, 2139359232, SEEK_SET)   = 2139359232
read(5, "\0\0\0\0\0 <cut >.... \0\0\0\0\0\0\0\0\0" ... 8192 = 8192

lseek(5, 2139367424, SEEK_SET)   = 2139367424
read(5, "\0\0\0\0\0 <cut >.... \0\0\0\0\0\0\0\0\0" ... 8192 = 8192

__lseek(5, 0, 0x80004800, SEEK_SET) = -1 EFAULT (Bad Address)
__lseek(5, 0, 0x80004800, SEEK_SET) = -1 EFAULT (Bad Address)

write(1 .. error reading block
write(1, "2097170", 72097170)...
...


Right now I don't have a real system (as opposed to vmware) to test it. If
necessary will do it next days.

Thanks

-- 
giulioo at pobox.com



More information about the uClibc mailing list