[Intel-gfx] [regression] i915: enable AVI infoframe for intel_hdmi.c [v4]

Alexey Fisher bug-track at fisher-privat.net
Sat Oct 23 15:06:22 CEST 2010


Am Samstag, den 23.10.2010, 12:38 +0100 schrieb Chris Wilson:
> On Sat, 23 Oct 2010 13:15:54 +0200, David Härdeman <david at hardeman.nu> wrote:
> > On Sat, Oct 23, 2010 at 10:03:38AM +0200, Alexey Fisher wrote:
> > >I have regression after commit "i915: enable AVI infoframe for
> > >intel_hdmi.c [v4]". After boot i get black screen with cursor on top
> > >left corner. I can access my pc per ssh.
> > >
> > >My HW: Intel DG45ID board, Monitor (iiyama prolite b2409hds) connected
> > >on DVI.
> > 
> > Perhaps the SDVOB and SDVOC handling is wrong? My hardware only has one 
> > HDMI port while yours has two (and HDMI2 is connected). I wonder why 
> > your DVI port is recognized as HDMI2 though? Chris?
> 
> In this case it is more likely that we are feeding an audio signal along a
> DVI cable to a HDMI-capable monitor that simply isn't expecting the extra
> data on its DVI connector. The reason why we don't recognise HDMI as DVI
> is that we simply ignore the VBT which tells us what connectors are
> actually attached to the HDMI encoders on the chip. (We have a few bugs
> open due to this already.)
> 
> Does 'xrandr --output HDMI2 --set force_audio -1' restore the signal?

ouch... i can't reproduce it anymore. I do not know how to explain it,
may be some site effect after bisection before and some configuration
bits was not reseted after restart. Now after cold start it just
working :/

Regrars,
Alexey.




More information about the Intel-gfx mailing list