Mailinglist Archive: radeonhd (333 mails)

[Christian König <deathsimple@xxxxxxxxxxx>]

Hi,

> I was trying to get this code tested and straigt into master yesterday, 
> but i got pulled away from that into opensuse 11.1 beta bug work.
>
> It's time this gets into master, and this is going to be a personal 
> priority for me now, as soon as the bugload is lessened that is :)
>
> Thanks for doing this work, and sticking to it, i think your code is 
> quite amazing and fuses in with the rest of the driver rather nicely.
Thanks for the flowers, and it's very nice to hear that this patch will
finally go into master.

Attached is a small patch fixing the VT switching bug.

Now to Anders testing Problem:
Rafał explained it just right. My patch enables the HDMI engine and
programs the audio timing, wich in turn depends on the selected video
resolution and so can't be calculated by alsa.

> Configure thinks I have some ansic.h thing installed and sets 
> HAVE_XF86_ANSIC_H in config.h. During compile we end up missing a
> prototype 
> for kill(). Reverting th definition makes <signal.h> included again.
> From your logs i can see that your X server is quite old (1.3), maybe
some kind of compatibly issue with the old server version.

> Without the patch. Everthing works (it seems. DRI is iffy).
> With the patch nothing works. black display. vt switching doesn't
> work. 
> Pressing ctrl/alt/del does work, so we're apparently outside of X.
> After a 
> while, we're back in bios boot (ctrl/alt/del properly rebooting the
> system)
Are you sure that your X server don't segfaults?
This sound like your X server gets killed in the startup process and
doesn't restores the correct display mode.

Try SSH from a remote machine and run:
sudo /usr/bin/X -logverbose 7
This should give a more verbose Xorg logfile and you can actually see
what's happening with the X server.

Bye, Christian.