nV News Forums

 
 

nV News Forums (http://www.nvnews.net/vbulletin/index.php)
-   NVIDIA Linux (http://www.nvnews.net/vbulletin/forumdisplay.php?f=14)
-   -   5336 and kernel panics (http://www.nvnews.net/vbulletin/showthread.php?t=24298)

obi_wan 01-31-04 10:28 AM

5336 and kernel panics
 
Some apps cause a kernel panic when exiting. With 5328 I had it always when exiting apps like neverwinternights. It seems fixed with 5336 unless the app is killed.
For instance when running wine with the toolset and killing it:

Jan 31 15:59:59 teigetje kernel: kernel BUG at exit.c:515!
Jan 31 15:59:59 teigetje kernel: invalid operand: 0000
Jan 31 15:59:59 teigetje kernel: sg st sr_mod sd_mod scsi_mod binfmt_misc lp par
port_pc parport nvidia snd-seq-midi snd-emu10k1-synth snd-emux-synth snd-seq-mid
i-emul snd-seq-virmidi snd-seq-oss snd-seq-midi-event snd-seq snd-pcm-oss snd-mi
xer-oss snd-emu10k1 snd-pcm snd-timer snd-util-mem snd-rawmidi snd-page-alloc sn
d-seq-device snd-ac97-codec snd-hwdep snd soundcore af_packet ide-floppy ide-tap
e ide-cd cdrom floppy 8139too mii isofs loop aes xfs nls_iso8859-1 nls_cp850 vfa
t fat supermount usb-uhci usbcore rtc ext3 jbd
Jan 31 15:59:59 teigetje kernel: CPU: 0
Jan 31 15:59:59 teigetje kernel: EIP: 0010:[do_exit+166/736] Tainted: P
Jan 31 15:59:59 teigetje kernel: EIP: 0010:[<c01df496>] Tainted: P
Jan 31 15:59:59 teigetje kernel: EFLAGS: 00013246
Jan 31 15:59:59 teigetje kernel: EIP is at do_exit+0xa6/0x2e0 [kernel]
Jan 31 15:59:59 teigetje kernel: eax: 00000000 ebx: cffea200 ecx: 040023f3
edx: 040023f2
Jan 31 15:59:59 teigetje kernel: esi: 00000000 edi: cbebc000 ebp: cbebdfb0
esp: cbebdfa4
Jan 31 15:59:59 teigetje kernel: ds: 0018 es: 0018 ss: 0018
Jan 31 15:59:59 teigetje kernel: Process wine-pthread (pid: 0, stackpage=cbebd00
0)
Jan 31 15:59:59 teigetje kernel: Stack: cbebc000 401b68ac 401b68ac cbebdfbc c01d
f701 00000000 4002eb7c c01c5503
Jan 31 15:59:59 teigetje kernel: 00000000 00000001 00000001 401b68ac 401b
68ac 4002eb7c 00000001 0000002b
Jan 31 15:59:59 teigetje kernel: 0000002b 00000001 401350fa 00000023 0020
0246 4002eb60 0000002b
Jan 31 15:59:59 teigetje kernel: Call Trace:
Jan 31 15:59:59 teigetje kernel: [sys_exit+17/32] sys_exit+0x11/0x20 [kernel]
Jan 31 15:59:59 teigetje kernel: [<c01df701>] sys_exit+0x11/0x20 [kernel]
Jan 31 15:59:59 teigetje kernel: [system_call+51/64] system_call+0x33/0x40 [ker
nel]
Jan 31 15:59:59 teigetje kernel: [<c01c5503>] system_call+0x33/0x40 [kernel]
Jan 31 15:59:59 teigetje kernel:
Jan 31 15:59:59 teigetje kernel: Code: 0f 0b 03 02 f8 32 36 c0 8b 75 08 56 e8 29
46 00 00 5b 8b 5f
Jan 31 15:59:59 teigetje kernel: <0>Kernel panic: Attempted to kill the idle ta
sk!
Jan 31 15:59:59 teigetje kernel: In idle task - not syncing


It does not happen with 4496.

GForce2 MX400 on VIA KT266

obi_wan 01-31-04 10:35 AM

because linenumbers change between kernels here the part were it calls BUG in exit.c:

put_exec_domain(tsk->exec_domain);
if (tsk->binfmt && tsk->binfmt->module)
__MOD_DEC_USE_COUNT(tsk->binfmt->module);

tsk->exit_code = code;
exit_notify();
schedule();
BUG();
/*

obi_wan 02-01-04 11:44 AM

hmm..I now *think* it is due to preempt patch. Any comments on whether the driver is preempt save?

zander 02-01-04 11:52 AM

The driver should work just fine with CONFIG_PREEMPT; can you give more details on the kernel you're using and its configuration?

obi_wan 02-04-04 12:45 PM

kernel is 2.4.22-26mm.1mdk. Which is basically mandrakes 2.4.22-26mdk kernel plus Con Kolivas preempt, lowlatency and O(1) scheduler patch.
Since the 2.4.22-26mdk kernel works fine I suspect that 5536 cannot handle preemption or the new scheduler.

zander 02-04-04 01:00 PM

The driver is SMP safe and works just fine with O(1) and CONFIG_PREEMPT. I'd recommend you use a kernel that isn't as heavily patched.


All times are GMT -5. The time now is 04:52 PM.

Powered by vBulletin® Version 3.7.1
Copyright ©2000 - 2014, Jelsoft Enterprises Ltd.
Copyright 1998 - 2014, nV News.