Re: Issues with 2.4.20/2.4.21-rc1[-ac3] and 2.5.68 on a Dell laptop

Re: Issues with 2.4.20/2.4.21-rc1[-ac3] and 2.5.68 on a Dell laptop

[Justin Pryzby]

Hi; FWIW, I've got an Inspiron 4000 running 2.5.68.  Neither
/proc/sys/cpu/0/speed nor /proc/cpufreq all me to change any values
(speed has min==max==0, `cat cpufreq` lists no parameters).  `cat i8k`
works, causing no such lock.  I don't use suspend/resume.  The screen
brightness buttons work and do not hang.  ACPI kind of works;
`cat BAT0/info` causes "Error: ul_allocate: Attempt to allocate zero bytes".
The same occurs when I reinsert the battery.

Justin Pryzby

PS: I can confirm that /proc/acpi/button/lid/LID/state works; I just
sshed in to check.

On Tue, Apr 29, 2003 at 05:35:00PM -0500, Ricardo Galli wrote:
> 
> 
> Hi Alan,
>        first of all, thanks for you -acX, it solved several issues in my Dell
> X200, the inclusion of XFS is great, and dri/drm with xfree 4.3 works 
> perfectly with the i830M (thanks).
> 
> But there still several problems (with APM enabled, no framebuffer):
> 
> - With cpufreq enabled, the kernel hangs if you change the CPU speed
> _after_ a suspend/resume via the old interface (/proc/sys/cpu/0/speed,
> (also repeatable in a vanilla or -rc1 kernel)). It doesn't happen if my
> governor program uses /proc/cpufreq instead. I saw this bug also with the
> original cpufreq patch and also with 2.5.68.
> 
> - cat /proc/i8k produces a long kernel lock, everything gets locked for a
> few seconds. If you are playing a music, you must restart the program in
> order the get alsa sync'd again.
> 
> - After suspend/resume, the kernel hangs during a shutdown (just like the
> infamous w98se shutdown bug :-), it happens after all processes have been
> TERMed. Sometimes the screen goes white (it happens also with Marcelo
> tree). I tried this with almost every different version and bios
> workaround in APM kernel options. I also happens with 2.5.68.
> 
> - The kernel hangs/lock hard if IO-APIC is enabled and you try to change
> the screen brightness (<Fn><[UP][DOWN]-Arrow>). It also happens with
> 2.5.68.
> 
> - ACPI doesn't see the battery, the shutdown buttons just turn down with
> notifying the kernel, suspend doesn't work. Also seen with ACPI original
> patches and 2.5.68.
> 
> - Only happens with -ac3 version: the poweroff button turns the machine
> off inmediately, it doesn't wait for a few seconds, as previous versions.
> 
> Regards,
> 
> 
> --
>  ricardo galli      GPG id C8114D34
> -
> 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/

Issues with 2.4.20/2.4.21-rc1[-ac3] and 2.5.68 on a Dell laptop

[Ricardo Galli]

Hi Alan,
	first of all, thanks for you -acX, it solved several issues in my Dell 
X200, the inclusion of XFS is great, and dri/drm with xfree 4.3 works  
perfectly with the i830M (thanks).

But there still several problems (with APM enabled, no framebuffer):

- With cpufreq enabled, the kernel hangs if you change the CPU speed 
_after_ a suspend/resume via the old interface (/proc/sys/cpu/0/speed, 
(also repeatable in a vanilla or -rc1 kernel)). It doesn't happen if my 
governor program uses /proc/cpufreq instead. I saw this bug also with the 
original cpufreq patch and also with 2.5.68.

- cat /proc/i8k produces a long kernel lock, everything gets locked for a 
few seconds. If you are playing a music, you must restart the program in 
order the get alsa sync'd again.

- After suspend/resume, the kernel hangs during a shutdown (just like the 
infamous w98se shutdown bug :-), it happens after all processes have been 
TERMed. Sometimes the screen goes white (it happens also with Marcelo 
tree). I tried this with almost every different version and bios 
workaround in APM kernel options. I also happens with 2.5.68.

- The kernel hangs/lock hard if IO-APIC is enabled and you try to change 
the screen brightness (<Fn><[UP][DOWN]-Arrow>). It also happens with 
2.5.68.

- ACPI doesn't see the battery, the shutdown buttons just turn down with 
notifying the kernel, suspend doesn't work. Also seen with ACPI original 
patches and 2.5.68.

- Only happens with -ac3 version: the poweroff button turns the machine 
off inmediately, it doesn't wait for a few seconds, as previous versions.

Regards,


-- 
  ricardo galli       GPG id C8114D34