Intel 2.0.0 driver not seeing laptop screen resolutions properly
greg at kroah.com
Tue May 8 17:30:34 PDT 2007
On Tue, May 08, 2007 at 04:52:10PM -0700, Eric Anholt wrote:
> On Tue, 2007-05-08 at 15:30 -0700, Greg KH wrote:
> > On Tue, May 08, 2007 at 09:52:19AM -0700, Eric Anholt wrote:
> > > On Mon, 2007-05-07 at 18:17 -0700, Greg KH wrote:
> > > > So, I upgraded to 2.0.0 a few weeks ago, and it worked just fine,
> > > > without rebooting my laptop (quit X, restarted, and it was great). The
> > > > same resolution worked (1280x768) both with, and without any xorg.conf
> > > > file present.
> > > >
> > > > Then I rebooted to use a newer kernel, and I no longer can use 1280x768
> > > > anymore, the screen only runs at 1024x768, with a black band on the
> > > > right side where the unused pixels are (things aren't stretched to the
> > > > whole screen, like they had been in the past when switching to this
> > > > resolution.)
> > > >
> > > > Attached below is the log file, I don't seem to see any mention of the
> > > > 1280x768 resolution in it at all. I've also attached my xorg.conf file,
> > > > just in case I'm doing something really stupid there too.
> > > >
> > > > I can try the git version of the intel driver, but it didn't look like
> > > > anything had changed in there since the 2.0.0 release that would help
> > > > this out.
> > > >
> > > > My hardware for this laptop is:
> > > > 00:02.0 VGA compatible controller: Intel Corporation Mobile 915GM/GMS/910GML Express Graphics Controller (rev 04)
> > > >
> > > > Anyone have any ideas?
> > >
> > > Can you put Option "ModeDebug" "YES" in your Driver section and send a
> > > new log?
> > Sure, attached is the diff between the two logs without and with debug
> > enabled, and the whole new log file itself.
> > It looks like it sees the bios setting for the proper panel size, but
> > that might just be due to me still running the 915resolution program at
> > boot time, which used to be required under the old driver to get this
> > screen size working.
> > I have disabled 915resolution one time, and no difference in result, but
> > the log files might be different then, do you want me to try that to see
> > that log difference too?
> So, in your BIOS-initialized console, you've got a black band on the
> right side with a full screen of text, correct? It looks like your BIOS
> was too clever to use the panel fitter, and instead decided to achieve
> its goals by taking the 1280x768 panel mode and chopping it down to just
> 1024x768 active. Kind of.
> Here's a proposed hack to detect your situation and plug in better
> values for the detected panel timings. If it works, I'll add it. Note
> that in your case the BIOS gets the right answer (it looks like).
> However, since the BIOS sometimes likes to disappear on us, it's
> probably better to have this hack in place.
Sweet, it works!
Funny thing is, xrandr doesn't seem to see the new mode listed:
$ ./xrandr --prop
Screen 0: minimum 320 x 200, current 1280 x 768, maximum 1280 x 1280
VGA disconnected (normal left inverted right)
LVDS connected 1280x768+0+0 (normal left inverted right) 0mm x 0mm
BACKLIGHT: 31250 (0x00007a12) range: (0,31250)
1024x768 59.8*+ 85.0 75.0 70.1 60.0
800x600 85.1 72.2 75.0 60.3 56.2
640x480 85.0 72.8 75.0 59.9
TV disconnected (normal left inverted right)
BOTTOM: 37 (0x00000025) range: (0,100)
RIGHT: 46 (0x0000002e) range: (0,100)
TOP: 36 (0x00000024) range: (0,100)
LEFT: 54 (0x00000036) range: (0,100)
supported: NTSC-M NTSC-443 NTSC-J PAL-M
PAL-N PAL 480p at 59.94Hz 480p at 60Hz
576p 720p at 60Hz 720p at 59.94Hz 720p at 50Hz
1080i at 50Hz 1080i at 60Hz 1080i at 59.94H
If there's anything else I can do to test this, please let me know.
More information about the xorg