LinuxSA Mailing list archives

  From: Toby Corkindale <tjcorkin@steadycom.com.au>
  To  : linuxsa@linuxsa.org.au
  Date: Mon, 19 Apr 1999 01:56:16 +0000

Divide and seg fault..

Heyas
Lately I've been trying to put linux onto the last 2gb of a 10gb drive.
I've taken care to make sure the linux partition that i'm booting from
starts on a cylinder number lower than 1024, but even before I get to
booting I keep hitting serious errors.
Using a redhat or turbolinux install disk, I find that if I partition my
drive such that the last cylinder is in use, then when I save the
partition table, the partition table is trashed.
I have to resort to sfdisk to fix it. cfdisk, fdisk, diskdruid won't
load the corrupt table, and the kernel sometimes hangs on bootup when
trying to read it.

But I fixed all that, and made a partition set that works fine, albeit
by simpling ignoring the final cylinder, and the 8mb it represents. ;(
When trying to install from a Linux CD, onto the newly created
partition, i few packages install, and then one will randomly fail, and
the install program crashes dies...Sometimes I can reboot out, usually I
can't. No good reason is given.
(The partition is formatted fine, including the badblocks check. No
errors found.)

I put a 2gb drive into the same machine, and attempting to reformat the
partition and then mount it. Reformatting worked fine, as usual.
Attempting to mount the partition resulted in the following kernel dump!
(2.2.5 kernel version, on a pentium 2)
 
 [root@penfold /mnt]# mount /dev/hda5 new
 divide error: 0000
 CPU:    0
 EIP:    0010:[<c013ce82>]
 EFLAGS: 00010246
 eax: 0019bd07   ebx: 00002000   ecx: 0019bd07   edx: 00000000
 esi: c31f0c00   edi: c3190e00   ebp: 00000000   esp: c33a1ec4
 ds: 0018   es: 0018   ss: 0018
 Process mount (pid: 469, process nr: 35, stackpage=c33a1000)
 Stack: c01dd324 00000305 00000305 c01d0305 00000001 c31b9000 00000000
00000001
        c0126766 c3190e00 00000000 00000000 ffffffea c31d1060 00000305
00000305
        c0126bbb 00000305 c01bb31d 00000000 00000000 00000000 c3187650
c0ed0305
 Call Trace: [<c01d0305>] [<c0126766>] [<c0126bbb>] [<c01bb31d>]
[<c01bb31d>] [<c012708a>] <c01bb31d>]
        [<c0108848>]
 Code: f7 f5 89 c3 89 5f 68 8b 47 64 8d 4c 18 ff 89 c8 31 d2 f7 77
 Segmentation fault
 [root@penfold /mnt]#
-------------- 
Here's the output from fdisk for the curious:
 Output from fdisk /dev/hda:
 Command (m for help): p
 
 Disk /dev/hda: 255 heads, 63 sectors, 1245 cylinders
 Units = cylinders of 16065 * 512 bytes
 
    Device Boot    Start       End    Blocks   Id  System
 /dev/hda1   *         1      1020   8193118+   b  Win95 FAT32
 /dev/hda2          1021      1244   1799280    5  Extended
 /dev/hda5          1021      1230   1686793+  83  Linux native
 /dev/hda6          1231      1244    112423+  82  Linux swap


I'll try stuff again with kernel 2.2.6 now that it's out, but this
appears to be a pretty major die, since RH5.2 uses kernel 2.0.36, and
turbolinux was 2.2.3 or 4 I think...
Has anyone else seen this kind of problems?

Toby


-- 
..Veni, Vidi, VC++. (I came, I saw, I kludged)

-- 
Check out the LinuxSA web pages at http://www.linuxsa.org.au/
To unsubscribe from the LinuxSA list:
  mail linuxsa-request@linuxsa.org.au with "unsubscribe" as the subject