S3 resume broken in 2.6.7

S3 resume broken in 2.6.7

[Adrian Yee]

Hi,

I've had S3 resume working with 2.6.5/6 (various mm and stock) and 2.6.7-rc3-mm1, but after trying 2.6.7-rc3-mm2 and 2.6.7 kernels, S3 resume has stopped working.  I'd rather not go through the process of modularizing all the drivers and figuring out what's causing the problem, so I wanted to know if anyone knows what change could have caused the problem.  Thanks.

Adrian


-------------------------------------------------------
This SF.Net email is sponsored by The 2004 JavaOne(SM) Conference
Learn from the experts at JavaOne(SM), Sun's Worldwide Java Developer
Conference, June 28 - July 1 at the Moscone Center in San Francisco, CA
REGISTER AND SAVE! http://java.sun.com/javaone/sf Priority Code NWMGYKND

Re: S3 resume broken in 2.6.7

[Andi Kleen]

On Mon, Jun 21, 2004 at 03:39:48PM +0000, Stefan D?singer wrote:
> Am Sonntag, 20. Juni 2004 02:53 schrieb Adrian Yee:
> > Hi,
> >
> > I've had S3 resume working with 2.6.5/6 (various mm and stock) and
> > 2.6.7-rc3-mm1, but after trying 2.6.7-rc3-mm2 and 2.6.7 kernels, S3 resume
> > has stopped working.  I'd rather not go through the process of modularizing
> > all the drivers and figuring out what's causing the problem, so I wanted to
> > know if anyone knows what change could have caused the problem.  Thanks.
> >
> On my notebook(Acer travelmate 803) I had the same problem. A look at the 
> changelog showed me, that from 2.6.7-rc2 a default resume function restores 
> the pci configuration of cards without a driver. On my system, any write to 
> the soundcard's pci configuration done by a kernel driver after resume causes 
> a crash.

Which driver does your sound card drive? Better would be to give
this driver an empty suspend/resume method that does nothing.

> I attached a path which undoes the change described above and should make S3 
> work again.

Overall the pci config space save/restore fixes much more resume 
problems than it causes new one, so this sledgehammer is the wrong
solution.

-Andi



-------------------------------------------------------
This SF.Net email is sponsored by The 2004 JavaOne(SM) Conference
Learn from the experts at JavaOne(SM), Sun's Worldwide Java Developer
Conference, June 28 - July 1 at the Moscone Center in San Francisco, CA
REGISTER AND SAVE! http://java.sun.com/javaone/sf Priority Code NWMGYKND

Re: S3 resume broken in 2.6.7

[Stefan Dösinger]

[-- Attachment #1: Type: text/plain, Size: 855 bytes --]

Am Sonntag, 20. Juni 2004 02:53 schrieb Adrian Yee:
> Hi,
>
> I've had S3 resume working with 2.6.5/6 (various mm and stock) and
> 2.6.7-rc3-mm1, but after trying 2.6.7-rc3-mm2 and 2.6.7 kernels, S3 resume
> has stopped working.  I'd rather not go through the process of modularizing
> all the drivers and figuring out what's causing the problem, so I wanted to
> know if anyone knows what change could have caused the problem.  Thanks.
>
On my notebook(Acer travelmate 803) I had the same problem. A look at the 
changelog showed me, that from 2.6.7-rc2 a default resume function restores 
the pci configuration of cards without a driver. On my system, any write to 
the soundcard's pci configuration done by a kernel driver after resume causes 
a crash.

I attached a path which undoes the change described above and should make S3 
work again.

Stefan

[-- Attachment #2: 2.6.7-fix --]
[-- Type: text/x-diff, Size: 314 bytes --]

--- linux-2.6.6/drivers/pci/pci-driver.c	2004-06-21 15:36:37.258692976 +0000
+++ drivers/pci/pci-driver.c	2004-06-16 21:16:14.000000000 +0000
@@ -332,8 +332,8 @@
 
 	if (drv && drv->resume)
 		drv->resume(pci_dev);
-	else
-		pci_default_resume(pci_dev);
+	//else
+	//	pci_default_resume(pci_dev);
 	return 0;
 }
 

Re: S3 resume broken in 2.6.

[Stefan Dösinger]

> Which driver does your sound card drive? Better would be to give
> this driver an empty suspend/resume method that does nothing.
i810_audio or snd_intel8x0.
In my system, there's at least one other card which causes the same problem.
> > I attached a path which undoes the change described above and should make
> > S3 work again.
>
> Overall the pci config space save/restore fixes much more resume
> problems than it causes new one, so this sledgehammer is the wrong
> solution.
I agree. I'd prefer a solution for the pci_write config fix.
The strange thing is that I can fix the PCI configuration with setpci after 
resume.

Disabling the default resume is a quick help for my system.

I can provide further information if necessary.

Stefan


-------------------------------------------------------
This SF.Net email sponsored by Black Hat Briefings & Training.
Attend Black Hat Briefings & Training, Las Vegas July 24-29 - 
digital self defense, top technical experts, no vendor pitches, 
unmatched networking opportunities. Visit www.blackhat.com