Created attachment 69282 [details] dmesg info System Environment: -------------------------- Platform: Ivybridge Kernel: (drm-intel-testing)b9960e75b5a348759c6e8c9ffb3f45e40ad702a5 Bug detailed description: ------------------------- On IVB platform, firstly I boot with VGA and 2 DVI monitors(we use HDMI->DVI and DP->DVI dongle)， they all can work well. But if I boot with one DVI, then I plug in another DVI, finally plug in the VGA, the VGA can't light up. This happened every time. Running testdisplay -i shows like this: CRTCs: id fb pos size 3 28 (0,0) (1680x1050) 1680x1050 60 1680 1784 1960 2240 1050 1053 1059 1089 0x6 0x48 146250 5 28 (0,0) (1920x1200) 1920x1200 60 1920 1968 2000 2080 1200 1203 1209 1235 0x9 0x40 154000 7 28 (0,0) (1920x1200) 1920x1200 60 1920 1968 2000 2080 1200 1203 1209 1235 0x6 0x40 154000
Hm, this could be and fdi b/c issue. Can you please retest with latest -queued, specifically commit f4fa49966d9db5fec7370a06812593865617335f Author: Daniel Vetter <daniel.vetter@ffwll.ch> Date: Sat Oct 27 15:58:40 2012 +0200 drm/i915: check fdi B/C lane sharing constraint
Created attachment 69289 [details] dmesg info with patch (In reply to comment #1) > Hm, this could be and fdi b/c issue. Can you please retest with latest > -queued, specifically > > commit f4fa49966d9db5fec7370a06812593865617335f > Author: Daniel Vetter <daniel.vetter@ffwll.ch> > Date: Sat Oct 27 15:58:40 2012 +0200 > > drm/i915: check fdi B/C lane sharing constraint I try the newest -queued kernel including your patch, the issue still occurs.
If you have all three screens plugged in (after boot) and then restart X, does it work?
(In reply to comment #3) > If you have all three screens plugged in (after boot) and then restart X, > does it work? Yeah, the issue described of this bug still occurs, but if I restart X, all three screens can light up.
Created attachment 70146 [details] dmesg info with restarting X (In reply to comment #4) > (In reply to comment #3) > > If you have all three screens plugged in (after boot) and then restart X, > > does it work? > Yeah, the issue described of this bug still occurs, but if I restart X, all > three screens can light up. Here is the dmesg.
Believed (Daniel!) to be working again.
Created attachment 71419 [details] dmesg info on 12.12 (In reply to comment #6) > Believed (Daniel!) to be working again. I try with the -nightly: Kernel: (drm-intel-nightly)539526b4137bc0e7a8806c38c8522f226814a0e6 now the issue still occurs, it likes this: If I boot with VGA and 2 DVI monitors(we use HDMI->DVI and DP->DVI dongle)， they all can work well. But if I boot with one DVI, then I plug in another DVI, finally plug in the VGA, the VGA can't light up. Then if I start X, the VGA still can't light up. I kill X and restart X, all screen can light up correctly. I attach the dmesg.
Can you please attach a reg dump for each of the three cases below? (In reply to comment #7) > If I boot with VGA and 2 DVI monitors(we use HDMI->DVI and DP->DVI dongle)， > they all can work well. After boot completes with this config and things work. > But if I boot with one DVI, then I plug in another DVI, finally plug in the > VGA, the VGA can't light up. After trying to light up VGA, but VGA not showing anything. > Then if I start X, the VGA still can't light up. I kill X and restart X, all > screen can light up correctly. After restarting X, so that all 3 pipes work again.
Created attachment 71479 [details] The reg dump after boot completes with this config and things work. The reg dump after boot completes with this config and things work.
Created attachment 71480 [details] The reg dump after trying to light up VGA, but VGA not showing anything. The reg dump after trying to light up VGA, but VGA not showing anything.
Created attachment 71481 [details] The reg dump after restarting X, so that all 3 pipes work again. The reg dump after restarting X, so that all 3 pipes work again. The above three reg dumps get with kernel: Kernel: (drm-intel-nightly)cdb96764a45f87e4614df1b16d68b2ccb9806f57
I try with the latest -testing: Kernel: (drm-intel-testing)5c43656cdac8062485ba54625bdc2edfc34ef0cf the issue is gone, so I think we can close this bug.
Awesome, and thanks for retesting.
Closing old verified.
Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.