2000-12-05 00:32:51

by Stefan Pfetzing

[permalink] [raw]
Subject: Posible Bug at page_alloc.c in the [2.4.0-test11] kernel

Hello,

I think i possibly foun a Bug in the Linux Kernel 2.4.0-test11.
I patched the kernel with the bttv driver 7.49 but nothing else.
My system was up for 5 1/2 days and than it crashed.

I have a Pentium 200MMX (overclocked to 225 but I REALLY don't think that's
the problem, because it worked fine with other Kernels.), a Asus TX97 Board,
64MB Ram, an Nvidia Riva 128 graphics card, a Hauppauge TV Card, a AVM Isdn Card,
a Symbios Logic SCSI Card and of course a D-Link network card (tulip based).
I also have several Usb devices attached (Monitor, Keyboard and Mouse)

At the moment when the kernel crashed, I was online (via ISDN) and I just closed
xawv (tv viewing program) and started playing a mp3 via mp3blaster.
(hope this information isn't too accurate)

/var/log/messages:
---snipp---
Dec 4 23:50:57 server kernel: kernel BUG at page_alloc.c:84!
Dec 4 23:50:57 server kernel: invalid operand: 0000
Dec 4 23:50:57 server kernel: CPU: 0
Dec 4 23:50:57 server kernel: EIP: 0010:[__free_pages_ok+73/892]
Dec 4 23:50:57 server kernel: EFLAGS: 00010286
Dec 4 23:50:57 server kernel: eax: 0000001f ebx: c1013210 ecx: 00000000 edx: 00000000
Dec 4 23:50:57 server kernel: esi: c1013238 edi: 00000000 ebp: 0000001e esp: c1165f64
Dec 4 23:50:57 server kernel: ds: 0018 es: 0018 ss: 0018
Dec 4 23:50:57 server kernel: Process kswapd (pid: 3, stackpage=c1165000)
Dec 4 23:50:57 server kernel: Stack: c01df9a5 c01dfb73 00000054 c1013210 c1013238 00000091 0000001e 00000091
Dec 4 23:50:57 server kernel: 0000001d 00000000 00000003 c0128882 c012a127 c0128a69 00000000 00000004
Dec 4 23:50:57 server kernel: 00000000 00000000 00000000 00000004 00000000 00000000 c01290d0 00000004
Dec 4 23:50:57 server kernel: Call Trace: [tvecs+7389/60504] [tvecs+7851/60504] [page_launder+670/1900] [__free_pa
Dec 4 23:50:57 server kernel: [kswapd+115/316] [kernel_thread+40/56]
Dec 4 23:50:57 server kernel: Code: 0f 0b 83 c4 0c 89 f6 89 da 2b 15 b8 fd 26 c0 89 d0 c1 e0 04
Dec 4 23:51:23 server kernel: kernel BUG at vmscan.c:533!
Dec 4 23:51:23 server kernel: invalid operand: 0000
Dec 4 23:51:23 server kernel: CPU: 0
Dec 4 23:51:23 server kernel: EIP: 0010:[reclaim_page+901/988]
Dec 4 23:51:23 server kernel: EFLAGS: 00010286
Dec 4 23:51:23 server kernel: eax: 0000001c ebx: c101322c ecx: 00000000 edx: 00000010
Dec 4 23:51:23 server kernel: esi: c1013210 edi: c06ab5bc ebp: 00000000 esp: c3491e10
Dec 4 23:51:23 server kernel: esi: c1013210 edi: c06ab5bc ebp: 00000000 esp: c3491e10
Dec 4 23:51:23 server kernel: ds: 0018 es: 0018 ss: 0018
Dec 4 23:51:23 server kernel: Process xglobe (pid: 14216, stackpage=c3491000)
Dec 4 23:51:23 server kernel: Stack: c01df405 c01df5c4 00000215 c020e6e0 c020e974 00000000 00000000 c0129dc8
Dec 4 23:51:23 server kernel: c020e6e0 00000000 c020e978 00000000 00000001 c0129f83 c020e96c 00000000
Dec 4 23:51:23 server kernel: 00000000 00000001 00000000 c1024188 000000a3 c1a79bbc 00000005 00000001
Dec 4 23:51:23 server kernel: Call Trace: [tvecs+5949/60504] [tvecs+6396/60504] [__alloc_pages_limit+124/172] [__a
Dec 4 23:51:23 server kernel: [handle_mm_fault+232/340] [do_page_fault+311/1004] [do_page_fault+0/1004] [do
Dec 4 23:51:23 server kernel: Code: 0f 0b 83 c4 0c 31 c0 0f b3 46 18 8d 5e 28 8d 46 2c 39 46 2c
Dec 4 23:55:00 server /USR/SBIN/CRON[14222]: (root) CMD ( ping http://www.heise.de -c1 >/dev/null )
Dec 4 23:55:43 server kernel: usb-uhci.c: interrupt, status 3, frame# 100
---snipp---

If i forgot anything (kernel config or anything else please mail me.

And please also mail to my address in replys because I am NOT suscribed to the mailing list.

thanx

dreamind



2000-12-05 01:22:46

by Christian Ehrhardt

[permalink] [raw]
Subject: Re: Posible Bug at page_alloc.c in the [2.4.0-test11] kernel

On Tue, Dec 05, 2000 at 01:04:05AM +0100, Stefan Pfetzing wrote:
> Hello,
>
> I think i possibly foun a Bug in the Linux Kernel 2.4.0-test11.
> I patched the kernel with the bttv driver 7.49 but nothing else.
> My system was up for 5 1/2 days and than it crashed.
> [ page_alloc: BUG at line 84 ]

That's interesting. Could you double check that you run 2.4.0-test11
and not 2.4.0-test12preX? There has been at least one reports of this
very bug and two others that look related. However, it looked like a
test12-pre[34] problem - until now.

> [ ... ]
> If i forgot anything (kernel config or anything else please mail me.

The compiler version you are using is always an interesting data point.
But I doubt that this is the problem here.

Gruesse Christian

--

THAT'S ALL FOLKS!

2000-12-05 01:32:37

by Stefan Pfetzing

[permalink] [raw]
Subject: Re: Posible Bug at page_alloc.c in the [2.4.0-test11] kernel



Am Tue, 5 Dec 2000 schrieb [email protected] :

> On Tue, Dec 05, 2000 at 01:04:05AM +0100, Stefan Pfetzing wrote:
> > Hello,
> >
> > I think i possibly foun a Bug in the Linux Kernel 2.4.0-test11.
> > I patched the kernel with the bttv driver 7.49 but nothing else.
> > My system was up for 5 1/2 days and than it crashed.
> > [ page_alloc: BUG at line 84 ]
>
> That's interesting. Could you double check that you run 2.4.0-test11
> and not 2.4.0-test12preX? There has been at least one reports of this
> very bug and two others that look related. However, it looked like a
> test12-pre[34] problem - until now.
>
no please believe me it IS 2.4.0-test11
uname -a:
Linux server 2.4.0-test11 #3 Tue Nov 28 21:15:01 CET 2000 i586 unknow

> > [ ... ]
> > If i forgot anything (kernel config or anything else please mail me.
>
> The compiler version you are using is always an interesting data point.
> But I doubt that this is the problem here.

the gcc is version 2.95.2
>
> Gruesse Christian
>
cu

dreamind

p.s.:please reply ALSO to my private address because I am NOT on the mailing-list.