Message-ID: <EA16D384EF89C942B5948FEB8E5D2FF3061C5F@mailserver>
From: Tarkan Erimer <TARKANE@solmaz.com.tr>
To: "'Mikael Pettersson'" <mikpe@csd.uu.se>
Cc: "lkml (linux-kernel@vger.kernel.org)" <linux-kernel@vger.kernel.org>
Subject: RE: [BUG] Hard lock up exiting X on linux-2.6.0-test5/test5-mm4
Date: Fri, 26 Sep 2003 09:14:12 +0300
MIME-Version: 1.0
Content-Type: text/plain;
	charset="iso-8859-9"
Content-Transfer-Encoding: 8BIT
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 2562
Lines: 75

Hi again Mikael,

I tried last night with CONFIG_APM_DISPLAY_BLANK disabled and APM=y and that
worked fine for me.
Thanks for your very help. 

Regards

Tarkan Erimer

-----Original Message-----
From: Tarkan Erimer 
Sent: 25 Eyl?l 2003 Per?embe 12:26
To: 'Mikael Pettersson'
Cc: lkml (linux-kernel@vger.kernel.org)
Subject: RE: [BUG] Hard lock up exiting X on linux-2.6.0-test5/test5-mm4


Thanks for your reply, Mikael. I will try that, when I back to the home. 

Tarkan Erimer

-----Original Message-----
From: Mikael Pettersson [mailto:mikpe@csd.uu.se] 
Sent: 25 Eyl?l 2003 Per?embe 12:18
To: Tarkan Erimer
Cc: lkml (linux-kernel@vger.kernel.org)
Subject: Re: [BUG] Hard lock up exiting X on linux-2.6.0-test5/test5-mm4


Tarkan Erimer writes:
 > I compile and run linux-2.6.0-test5-mm4. It works wonderful, but when I
> switch to X windows it's OK. But, when I tried to exit X, it 
> completely  >
freezes the box. I tried this with open source nvidia (nv) driver and  >
proprietary nvidia (nvidia)driver. Results are always same. My hardware is:
> P-II 350, 384 RAM, BX board and Riva TNT gfx card. The same thing also  
> >
happened with vanilla linux-2.6.0-test5. I attached my .config and  >
version_linux outputs. Any idea ?

CONFIG_X86_UP_APIC=y
...
CONFIG_APM=m
# CONFIG_APM_IGNORE_USER_SUSPEND is not set CONFIG_APM_DO_ENABLE=y #
CONFIG_APM_CPU_IDLE is not set CONFIG_APM_DISPLAY_BLANK=y

You may suffer from a broken graphics card BIOS that hangs
if a local APIC timer interrupt arrives. I've seen this
happen on G400 and Radeons, and an old S3 I think.

Another issue, but not the one you're having, is that configuring APM=m is
known to hang some BIOSen at the point when the APM module is unloaded,
again due to BIOS code not handling local APIC interrupts.

Fix: Disable CONFIG_APM_DISPLAY_BLANK. You don't need it.
Also set APM=y if you're going to have it at all.

(An alternative fix is to disable UP_APIC, but
only do that as a last resort.)

I also noticed you had enabled a lot of ACPI stuff.
This is almost certainly pointless on a BX board.

/Mikael
-
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/