Re: 2.1.86 + xmixer Oops + 2.1.86/gcc = death

Mark Seward (illuzion@paradise.net.nz)
Mon, 16 Feb 1998 23:21:12 +1300 (NZDT)


Follow up to previous posting:

At the recommendation of a fellow linuxer, I tried a vanilla gcc-2.7.2.3
compile of 2.1.86 (with all default Makefile settings bar commenting SMP =
1 out), to see if the problem was egcs-related.

With this new kernel booted, upon executing xmixer, instant death
occurs. X freezes, no keyboard, mouse, network, nothing. After hitting the
Big Red Button and booting up again, nothing appears in the logs, so I am
to assume this is something at least mildly worthy of attention :)

Perhaps it is something to do with the recent sound changes in the 2.1.x
tree?

I tried egcs-1.0.1 with more conservative optimisations (defaults for
CONFIG_M586), and this kernel Oopsed like the other egcsed one, however
it's ksymoops output seemed significantly different (perhaps more
helpful)...

Unable to handle kernel NULL pointer dereference at virtual address 00000000
current->tss.cr3 = 01341000, %cr3 = 01341000 *pde = 00000000
Oops: 0000
CPU: 0
EIP: 0010:[<c0109db8>]
EFLAGS: 00010046
eax: 00000010 ebx: 0000002b ecx: 00000000 edx: 00000010
esi: 00000000 edi: c1344000 ebp: c1343f40 esp: c1343ec4
ds: 0018 es: 0018 ss: 0018
Process xmixer (pid: 187, process nr: 35, stackpage=c1343000)
Stack: 0000002b c1343f40 00000000 00000000 00000000 02000000 c3000000 c2800000
001864d7 c0109e35 c1343f40 c01955d3 c019639e 00000000 c1343f40 c010ef9f
c019639e c1343f40 00000000 c1342000 bffffc30 80304d65 00000030 c1ff93c0
Call Trace: [<c3000000>] [<c2800000>] [<c0109e35>] [<c01955d3>] [<c019639e>] [<c010ef9f>] [<c019639e>]
[<c0109a52>] [<c2853000>] [<c2832f18>] [<c012b53c>] [<c01098da>]
Code: 64 8a 04 0e 0f a1 88 c2 31 c0 88 d0 89 44 24 10 50 68 cb 55

Using `./System.map-2.1.86-egcs' to map addresses to symbols.

>>EIP: c0109db8 <show_registers+20c/248>
Trace: c3000000
Trace: c2800000
Trace: c0109e35 <die_if_kernel+41/50>
Trace: c01955d3 <sprintf+17bb/1a18>
Trace: c019639e <bad_pmd_string+296/314>
Trace: c010ef9f <do_page_fault+317/328>
Trace: c019639e <bad_pmd_string+296/314>
Trace: c0109a52 <error_code+32/40>
Trace: c2853000
Trace: c2832f18
Trace: c012b53c <sys_ioctl+110/124>
Trace: c01098da <system_call+3a/40>
Code: c0109db8 <show_registers+20c/248>
Code: c0109db8 <show_registers+20c/248> 64 8a 04 0e movb %fs:(%esi,%ecx,1),%al
Code: c0109dbc <show_registers+210/248> 0f a1 popl %fs
Code: c0109dbe <show_registers+212/248> 88 c2 movb %al,%dl
Code: c0109dc0 <show_registers+214/248> 31 c0 xorl %eax,%eax
Code: c0109dc8 <show_registers+21c/248> 88 d0 movb %dl,%al
Code: c0109dca <show_registers+21e/248> 89 44 24 10 movl %eax,0x10(%esp,1)
Code: c0109dce <show_registers+222/248> 50 pushl %eax
Code: c0109dcf <show_registers+223/248> 68 cb 55 00 90 pushl $0x900055cb
Code: c0109dda <show_registers+22e/248> 90 nop
Code: c0109ddb <show_registers+22f/248> 90 nop

Once again, any help is greatly appreciated.

Thanks for your time,
Mark Seward. [illuzion@paradise.net.nz]

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