Re: [PATCH 2/2] drm/i915/lvds: Use i915.lvds_fixed_mode= as a last resort

[Mike Isely <isely@isely.net>]

On Thu, 21 Apr 2011, Mike Isely wrote:

> On Thu, 21 Apr 2011, Chris Wilson wrote:
> 
> > On Thu, 21 Apr 2011 17:30:21 -0500 (CDT), Mike Isely <isely@isely.net> wrote:
> > > It looks like the driver did notice the specified mode and used it - 
> > > because the display's refresh rate did get adjusted and the overall 
> > > resolution is correct.  However the display timings are slightly "off", 
> > > the image is stretched a few pixels off the right edge and a few lines 
> > > off the bottom edge.  Yet the same modeline given to "video=" in 
> > > combination with my controversial disable-fixed-mode patch however 
> > > produces a perfectly aligned image.
> > 
> > Adding a few printk will help. And perhaps comparing the resultant register
> > settings with intel_reg_dumper.
> 
> Oh absolutely.  But right now I'm not doing a very fair apples-to-apples 
> comparison.  I need to address that first and make sure nothing else is 
> influencing this.
> 
>   -Mike
> 

Chris:

Sorry this simple thing has taken 2 weeks, but I'm having to time-slice 
a lot of different tasks at the moment.  People keep pulling me away 
from this :-( However I've finally dug down and found the problem.  
There's a bug in the function drm_mode_parse_command_line_for_connector 
(the subject of this thread) that is leaving the margins field 
uninitialized.  This has been the cause all along for the differing 
results I was getting.

This bug didn't start with the underlying patch ("Use 
i915.lvds_fixed_mode= as a last resort").  I also looked at 
drm_fb_helper_connector_parse_command_line from before this patch and it 
looks like the same bug exists there.  Amazing that nobody has noticed 
this.

Patch to follow.

  -Mike


-- 

Mike Isely
isely @ isely (dot) net
PGP: 03 54 43 4D 75 E5 CC 92 71 16 01 E2 B5 F5 C1 E8