2.6.26.x hangs on amd64/smp

2.6.26.x hangs on amd64/smp

[BERTRAND Joel]

	Hello,

System : debian/testing, tested kernels 2.6.26, 2.6.26.3, 2.6.26.5.
Hardware : core2duo, 4 GB, raid1 software, CFQ scheduler.

	I have written a program that work on cartographic data. This program 
is started as a daemon and does some fork() (and pthread_create()). I 
have seen that it requires 6 GB to work, each process takes 1,5 GB. The 
same program works fine under FreeBSD or Solaris (on of course the same 
hardware).

	When it starts, I can see disk activity (swap), and after 2 or 3 
minutes, kernel crashes without any trace (no more disk activity, sysrq 
does nothing...). I have reproduced this bug when I was logged on 
console. There was no messsage.

	If I introduce some nanosleep() syscalls in my code, crash is more 
difficult to reproduce.

cauchy:[~] > cat /proc/mdstat
Personalities : [raid1]
md1 : active raid1 sdb2[1] sda2[0]
       5855616 blocks [2/2] [UU]

md2 : active raid1 sdb3[1] sda3[0]
       48829440 blocks [2/2] [UU]

md3 : active raid1 sdb4[1] sda4[0]
       101474496 blocks [2/2] [UU]

md0 : active raid1 sdb1[1] sda1[0]
       128384 blocks [2/2] [UU]

unused devices: <none>

	swap in on /dev/md1.

cauchy:[~] > df -h
Sys. de fich.         Tail. Occ. Disp. %Occ. Monté sur
/dev/md2               46G   28G   16G  64% /
tmpfs                 2,0G     0  2,0G   0% /lib/init/rw
udev                   10M  124K  9,9M   2% /dev
tmpfs                 2,0G     0  2,0G   0% /dev/shm
/dev/md0              122M   60M   56M  52% /boot
/dev/md3               96G   56G   35G  62% /home
cauchy:[~] >

dmesg :
Linux version 2.6.26.5 (root@cauchy) (gcc version 4.3.1 (Debian 4.3.1-9) 
) #16 SMP PREEMPT Tue Sep 23 15:54:59 CEST 2008
...
ACPI: BIOS bug: multiple APIC/MADT found, using 0
ACPI: If "acpi_apic_instance=2" works better, notify 
linux-acpi@vger.kernel.org
ACPI: DMI detected: Toshiba
...

.config: see http://www.systella.fr/~bertrand/config.2.6.26.5

	Regards,

	JKB

Re: 2.6.26.x hangs on amd64/smp

[Oliver Pinter]

Hi!

I have similar problem, but the hw is asus p5q-e, c2q q9300, 4gb ram + rhd3450.

The X started very slowly ~1-2min, and randomly crashed the system,
with Freebsd all ok, but with linux not.
The other problem is, when I cp-ed 4gb<= file, this caused to file
corruption, when I booted with mem=512M kernel parameter, then this
file corrouption get off. I have memtest some hour long run, and with
freebsd i have ~1< uptime, then i turned off the pc. ...

On 9/26/08, BERTRAND Joel <bertrand@rayleigh.systella.fr> wrote:
> From: BERTRAND Joel <joel.bertrand@systella.fr>
> Newsgroups: linux.kernel
> Subject: 2.6.26.x hangs on amd64/smp
> Reply-To: mt1@systella.fr
>
> 	Some news.
>
> 	I have done some tests with the _same_ configuration, but without
> 	resident X (and console on a serial line). I'm not able to reproduce
> 	this bug anymore even with a load average
> 	greater than 30 (memory usage : 4GB + 6GB swap). Thus, this trouble
> 	seems to be X related.
>
> 	Graphic card is :
> 01:00.0 VGA compatible controller: ATI Technologies Inc M76 [Radeon
> Mobility HD 2600 Series]
>
> 	Xorg configuration :
> Section "Device"
>     Identifier  "ATI Technologies Inc ATI Default Card"
> 	Driver      "radeonhd"
> #   Driver      "fglrx"
> 	BusID       "PCI:1:0:0"
> EndSection
>
> cauchy:[~] > dpkg-query -l | grep xorg
> ii  xorg                                                    1:7.3+16
> ...
> ii  xserver-xorg-video-radeonhd                             1.2.1-2
>
>
> 	I'm writing a bug report on Xorg website.
>
> 	Regards,
>
> 	JKB
>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/
> --
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/
>