Bug 5184

Created attachment 3923 [details]
Log file with 0MHz entries

Created attachment 3924 [details]
Config file

It looks like this warning has started appearing because the interpret_edid.c
code is now setting max_clock=0 when the monitor doesn't specify a value. The
old code used to set 2550MHz in this case. This trips up the code in xf86Mode.c
which still looks for the 2550MHz value. 

I've looked into the source and it is clear that this isn't a mach64 driver
problem, but i'm not sure which component it should be recorded against.

See
http://cvsweb.freedesktop.org/xorg/xc/programs/Xserver/hw/xfree86/ddc/interpret_edid.c?r1=1.2&r2=1.3

and programs/Xserver/hw/xfree86/common/xf86Mode.c:

768 : 	  	  	 if (mon_range->max_clock < 2550 &&
769 : 	  	  	mode->Clock / 1000.0 > mon_range->max_clock) {
770 : 	  	  	xf86Msg(X_WARNING,
771 : 	  	  	"(%s,%s) mode clock %gMHz exceeds DDC maximum %dMHz\n",
772 : 	  	  	mode->name, monitor->id,
773 : 	  	  	mode->Clock/1000.0, mon_range->max_clock);
774 : 	  	  	}

i think the change in 1.3 needs to be rolled back.

(In reply to comment #4)
> i think the change in 1.3 needs to be rolled back.

but i'm not going to do it right now; out to 7.2.

I have MGA and IA-32.
I tried reverting change in v1.3 and results were:
I got no "...exceeds DDC maximum 0MHz" warnings but
refresh rate was still 55 Hz (instead of 91Hz I have had since 1999).

I also nuked DDC out of the game, but 55 Hz was the max.
And why does Xorg ignore my VertRefresh and/or HorizSync settings?
Where does it pull the magic numbers 55 Hz and 60 kHz
and how can I make latest Xorg work?
Now I have to use v1.0.1.

https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=188770

ajax: still worth doing for 7.2?

This is solved in a better way by the autoconfig branch merge, and should no
longer happen in server 1.2.  Removing from the tracker.

Still a sensible thing to do for 1.1 though.  For anyone reading this: the chunk
from interpret_edid.c is correct, xf86Mode.c needs to be updated to not
interpret a max_clock of 0 as really meaning no signals allowed.

Sorry about the phenomenal bug spam, guys.  Adding xorg-team@ to the QA contact so bugs don't get lost in future.

No one's doing xserver 1.1 updates, ever.