lkml.org 
[lkml]   [1998]   [Aug]   [12]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: ENOSYS, CD-ROM, and SMP oopses (from dcache)
On Wed, 12 Aug 1998, Erik Andersen wrote:

>I am assuming you built ide-cd as a module. Could you (under 2.1.115)
>add "options cdrom debug=1" to /etc/modules.conf so I can see where the
>uniform cdrom driver stops being uniform. This will cause it to spew
>out tons of stuff to the kernel logs when mounting a CD, and should let
>us see what went wrong.

With debug=1:

--
# modprobe ide-cd
/dev/hdc: No such device
post-install /lib/modules/2.1.115/block/ide-cd.o failed

cdrom: Device or resource busy
--

No logs.

Some bootup messages:

2.0.33 SMP says:
hda: Maxtor 71000 A, 958MB w/64kB Cache, CHS=1946/16/63, DMA
hdc: no response (status = 0xd0)

2.0.35 SMP says:
hda: Maxtor 71000 A, 958MB w/64kB Cache, CHS=1946/16/63, DMA
hdc: no response (status = 0xd0), resetting drive
hdc: CD820E.1v0789108, ATAPI CDROM drive

2.1.115 SMP says (with 'nosmp' option so it doesn't blow up on me):
hda: Maxtor 71000 A, ATA DISK drive
hdc: no response (status = 0xd0)

If I boot with 2.0.35, the eject button on the drive actually works. With
the other kernels, I hit the button and the drive just sits there taunting
me.


I'd still love to know why 2.1.115 SMP with one processor dies horribly
without the 'nosmp' option: It causes an Oops in the same spot every time.

Unable to handle kernel NULL pointer dereference at virtual address 00000008
current->tss.cr3 = 03b99000, %cr3 = 03b99000
*pde = 00000000
Oops: 0000
CPU: 0
EIP: 0010:[<c012f359>]
EFLAGS: 00010202
eax: 00000000 ebx: fffffffe ecx: c3f5d000 edx: 00000000
esi: c3f5d000 edi: c3f5d001 ebp: 00000000 esp: c35cddd4
ds: 0018 es: 0018 ss: 0018
Process klogd (pid: 1223, process nr: 19, stackpage=c35cd000)
Stack: 00000000 c01b9a10 00000000 00000000 00000000 00000000 c012f54f c3f5d000
00000000 00000001 00000020 c3f5d000 c3f5d000 c01b9a10 00000000 00008000
c012e597 c3f5d000 00000000 00000000 c3f5d000 c3f5d000 c35cc000 c35cdf8c
Call Trace: [<c012f54f>] [<c012e597>] [<c012eff4>] [<c0108a87>]
[<c0109d18>] [<c011c9e2>] [<c019dbec>]
[<c019dbef>] [<c011ca6a>]
Code: 8b 55 08 85 d2 0f 84 2c 01 00 00 bb ec ff ff ff 8b 42 58 85

Using `/System.map-2.1.115' to map addresses to symbols.

>>EIP: c012f359 <lookup_dentry+79/1c4>
Trace: c012f54f <open_namei+53/328>
Trace: c012e597 <do_execve+47/1d8>
Trace: c012eff4 <getname+58/90>
Trace: c0108a87 <sys_execve+4b/70>
Trace: c0109d18 <system_call+38/40>
Trace: c011c9e2 <exec_modprobe+152/190>
Trace: c019dbec <tvecs+a50/5a78>
Trace: c019dbef <tvecs+a53/5a78>
Trace: c011ca6a <request_module+4a/b0>
Code: c012f359 <lookup_dentry+79/1c4>
Code: c012f359 <lookup_dentry+79/1c4> 8b 55 08 movl 0x8(%ebp),%edx
Code: c012f35c <lookup_dentry+7c/1c4> 85 d2 testl %edx,%edx
Code: c012f35e <lookup_dentry+7e/1c4> 0f 84 2c 01 00 je c012f490 <lookup_dentry+1b0/1c4>
Code: c012f36a <lookup_dentry+8a/1c4> bb ec ff ff ff movl $0xffffffec,%ebx
Code: c012f36f <lookup_dentry+8f/1c4> 8b 42 58 movl 0x58(%edx),%eax
Code: c012f372 <lookup_dentry+92/1c4> 85 00 testl %eax,(%eax)
Code: c012f37a <lookup_dentry+9a/1c4> 90 nop
Code: c012f37b <lookup_dentry+9b/1c4> 90 nop
Code: c012f37c <lookup_dentry+9c/1c4> 90 nop

It's not 2.1.114 sched.c changes because 2.1.110 SMP does it also.

The AMD K6 arrived for my other computer so I can reclaim the Pentium
133 from this computer. If anyone is interested in the extremely easily
reproduced Oops, let me know and I'll leave the processor out for awhile.

-George


-
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.altern.org/andrebalsa/doc/lkml-faq.html

\
 
 \ /
  Last update: 2005-03-22 13:44    [W:0.049 / U:1.120 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site