lkml.org 
[lkml]   [1998]   [Jul]   [7]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectAMD K5 trouble on 2.1.x

Hi,

Recently, when browsing through Alan's things-to-do-for-2.2 list, I
remembered the trouble I was having with my AMD K5 (@100MHz) and the
2.1.x kernels some months ago.

Perhaps this is related to the entry "AMD K5 chips arent being
correctly handled causing Oopses as the clock starting code isnt
executed".

The strange thing is that if I would boot a 2.1.x kernel and just let
it sit there, doing nothing, for about half an hour and if I would
then try to log in, the newly spawned shell would immediately Oops
with a divide error. This is exactly reproducible and happens after
about 20 or 25 minutes of idle time. It always looks like this:

Jun 21 13:22:38 the-box kernel: divide error: 0000
Jun 21 13:22:38 the-box kernel: CPU: 0
Jun 21 13:22:38 the-box kernel: EIP: 0010:[<c010cebd>]
Jun 21 13:22:38 the-box kernel: EFLAGS: 00010046
Jun 21 13:22:38 the-box kernel: eax: 00000000 ebx: 074b0da8 ecx: 0000179b edx: 00002710
Jun 21 13:22:38 the-box kernel: esi: 80c7cdf7 edi: 08049b50 ebp: 0808eec4 esp: c5787f8c
Jun 21 13:22:38 the-box kernel: ds: 0018 es: 0018 ss: 0018
Jun 21 13:22:38 the-box kernel: Process tcsh (pid: 93, process nr: 4, stackpage=c5787000)
Jun 21 13:22:38 the-box kernel: Stack: c5787fb8 00000202 c010cf9e c5787fb8 4000623c c0117309 c5787fb8 c5786000
Jun 21 13:22:38 the-box kernel: 4000623c 08049b50 bffff57c 358cecfe 000788e1 c01097f8 0808eec4 00000000
Jun 21 13:22:38 the-box kernel: 08096838 4000623c 08049b50 bffff57c 0000004e 0000002b 0000002b 0000004e
Jun 21 13:22:38 the-box kernel: Call Trace: [<c010cf9e>] [<c0117309>] [<c01097f8>]
Jun 21 13:22:38 the-box kernel: Code: f7 f1 a3 c4 de 1c c0 89 c3 0f 31 31 d2 29 f0 f7 e3 81 fa 0f

The funny thing is I can hide this bug by using the "no-hlt" kernel
option. All is fine then (I've been running Linux this way for some
months now).

Also interesting might be that the first kernel that showed this
behaviour was 2.1.75. 2.1.74 is perfectly fine, as far as these things
are concerned. This is not really astonishing, as the 2.1.75 patch
includes a lot of changes in the part of the kernel written in
Assembly.

Let me know if I can be of further help. I could even run that Oops
through ksymoops, but I don't have a recent one at hand, and I don't
want to reboot now :)

Other than that: good work all you guys did on 2.1.x; there's really a
lot of new interesting features! Thank you!

--
michael krause [aka raw style / lego] - http://ms.demo.org/rst/









-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu

\
 
 \ /
  Last update: 2005-03-22 13:43    [W:0.081 / U:0.740 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site