CONFIG_VIDEO_SELECT stole my will to live

Hey,

since 2.5.67-bk5. I have been unable to boot my laptop
tonight, I finally traced it back to

http://linux.bkbits.net:8080/linux-2.5/cset@1.1006

this changeset

more details on the problem can be found

http://bugzilla.mozilla.org/show_bug.cgi?id=677

there
and

http://bad-sports.com/~brett/bugreports/kernel/2.5.69-failboot/

there

the video card is a chips and tech 65545

i'd just like to know where to go from here, so that I can return to 
booting with CONFIG_VIDEO_SELECT set 

i'm not on fbdev-devel, so please just cc me to any response

thanks,

	/ Brett



-------------------------------------------------------
This SF.net email is sponsored by: ObjectStore.
If flattening out C++ or Java code to make your application fit in a
relational database is painful, don't do it! Check out ObjectStore.
Now part of Progress Software. http://www.objectstore.net/sourceforge

On Wed, May 21, 2003 at 12:14:50AM +1000, Brett wrote:

 > since 2.5.67-bk5. I have been unable to boot my laptop
 > tonight, I finally traced it back to
 > http://linux.bkbits.net:8080/linux-2.5/cset@1.1006

good work.

 > the video card is a chips and tech 65545
 > i'd just like to know where to go from here, so that I can return to 
 > booting with CONFIG_VIDEO_SELECT set 

Possibly the card's VGA BIOS has 'issues' with that call.

Wasn't the EDID stuff getting backed out anyways ?

		Dave

On Wed, May 21, 2003 at 12:14:50AM +1000, Brett wrote:

 > since 2.5.67-bk5. I have been unable to boot my laptop
 > tonight, I finally traced it back to
 > http://linux.bkbits.net:8080/linux-2.5/cset@1.1006

good work.

 > the video card is a chips and tech 65545
 > i'd just like to know where to go from here, so that I can return to 
 > booting with CONFIG_VIDEO_SELECT set 

Possibly the card's VGA BIOS has 'issues' with that call.

This is most likely the case. I just tested out the configuration he has 
and it worked for me. I'm running vga=5 right now. For teh majority it 
works but as usual there are some broken BIOS that cause issues. 

Wasn't the EDID stuff getting backed out anyways ?

Only in the VESA driver. Some people did have luck with the BIOS EDID info 
so I like to keep the BIOS call in there. 



-------------------------------------------------------
This SF.net email is sponsored by: ObjectStore.
If flattening out C++ or Java code to make your application fit in a
relational database is painful, don't do it! Check out ObjectStore.
Now part of Progress Software. http://www.objectstore.net/sourceforge

On Tue, May 20, 2003 at 05:55:41PM +0100, James Simmons wrote:

 > > Wasn't the EDID stuff getting backed out anyways ?
 > Only in the VESA driver. Some people did have luck with the BIOS EDID info 
 > so I like to keep the BIOS call in there. 

That code runs way earlier than before we get to do any PCI quirks,
so fixing that one up could be interesting.

		Dave



-------------------------------------------------------
This SF.net email is sponsored by: ObjectStore.
If flattening out C++ or Java code to make your application fit in a
relational database is painful, don't do it! Check out ObjectStore.
Now part of Progress Software. http://www.objectstore.net/sourceforge

On Tue, May 20, 2003 at 05:55:41PM +0100, James Simmons wrote:


 > > Possibly the card's VGA BIOS has 'issues' with that call.
 > This is most likely the case. I just tested out the configuration he has 
 > and it worked for me. I'm running vga=5 right now. For teh majority it 
 > works but as usual there are some broken BIOS that cause issues. 

Or possibly..
It looks to me that the store_edid function is unconditionally calling
the READ EDID function without first having called the installation
check function. (Int 10h ax=4f15 bx=0)

Ralf Brown's interrupt list (which is pretty much definitive afaik)
also notes that for the READ EDID function, dx should be 0, we are
using 1. Is there a reason for this ?

		Dave



-------------------------------------------------------
This SF.net email is sponsored by: ObjectStore.
If flattening out C++ or Java code to make your application fit in a
relational database is painful, don't do it! Check out ObjectStore.
Now part of Progress Software. http://www.objectstore.net/sourceforge