This applies to g450/g550 only (AFAIK):
We currently set the M/N/P values for the video PLL before the video pll is actually powered up. This doesn't work, because VID_PLL_STAT will always say the PLL isn't locked if it's not powered up. Of course this doesn't work.
So the fix is to _first_ power up the video pll, and then setting its M/N/P values.
"signal out of range" weirdness for CRTC2.
Note to self: this one probably only applies to g550, since dual-head is working with the current code on g450 (with two vga outputs). Maybe verify that g450 uses the video pll, too.
Sorry about the phenomenal bug spam, guys. Adding xorg-team@ to the QA contact so bugs don't get lost in future.
what I am experiencing with my MGA G550 is that even though both screens get initalized to a correct refresh rate of 60 Hz:
(**) MGA(0): *Default mode "1280x1024": 108.0 MHz, 64.0 kHz, 60.0 Hz
(**) MGA(1): *Default mode "1280x1024": 108.0 MHz, 64.0 kHz, 60.0 Hz
the screen that is not connected to the DVI receives a 66Hz/71KHz signal instead, which it (being an LCD) cannot cope with.
Is this related to the incorrect setup of the PLL?
I'm running version 7.1.1 with the mga driver (no binary drivers).
this bug isn't about wrong setup of the VID_PLL values in general ;)
Please file a separate bug for your problem. I'm not sure if it's due to the video PLL though.
Created attachment 8977 [details] [review]
This patch fixes the problem on my G550 dual-dvi. It's pretty damn ugly, and we can fix it in a nicer way in the randr-1.2 branch eventually, so don't commit this please.
Fixed in the randr-1.2 branch.