Created attachment 63701 [details]
1) when is computer booted with VGA connected, ends with repeatly restarting GPU
2) when is monitor connected on-fly in running Xorg, it seems fine (except small artifacts, i guess EXA, on screen mostly on top of DVI and bottom VGA trigerred by switching KDE virtual desktops). No errors in dmesg
3) when is DPMS activated and monitors suspends, back to online is returned only DVI
4) randomly I found:
[drm:radeon_dp_link_train_cr] *ERROR* clock recovery reached max voltage
[drm:radeon_dp_link_train_cr] *ERROR* clock recovery failed
5) out-topic, exist even only on DVI: when is enabled OGL KWIN and OpenGL video render, color of rendered context goes to white (every frame is whitier, until it's #ffffff)
No HDMI or DP conneted to system.
DVI-D: LG IPS235 1920x1080
VGA: ASUS VW224 1680x1050
Kernel,libdrm,mesa,ddx: git, KMS
I'm able to test patches against anything.
1a) I tested switched (so Asus on DVI and LG on VGA - switched) and it booted
1b) secondary I tested only Asus on DVI and it had same problem...
only in this line numbers (13 => 12 etc.) is little bit changing:
radeon 0000:00:01.0: GPU lockup (waiting for 0x0000000000000016 last fence id 0x0000000000000013)
1c) tested only LG on VGA and it failed. (same errors)
2) of fly was tested: VGA only, DVI only, both vga+dvi, both dvi+vga, seems ok
sometimes it start rebooting, but it takes a some time.
Still valid with kernel 3.6.0-rc4
Created attachment 66971 [details]
after upgrade from 3.6.0_rc2 to rc4 second monitor works with clone (absolute 0 0 both), but extending desktop cause this (absolute 1920 0 and 0 0)
to picture - I tested rc2 and I accidantally discovered, that it's because OpenGL Kwin, after deactivating affects it's ok. So some git mesa update...
more can be founded here:
Although I didn't have display garbled that much (but I am not using an opengl-based DE).
Sorry I wanted to make it duplicate of #42490
*** This bug has been marked as a duplicate of bug 42490 ***