Created attachment 81849 [details] Xorg.0.log of the failed X start Using xf86-video-intel 2.21.11, X does not come up anymore on my Haswell machine. 2.21.10 worked fine in that regard. $ lspci | grep VGA 00:02.0 VGA compatible controller: Intel Corporation Xeon E3-1200 v3/4th Gen Core Processor Integrated Graphics Controller (rev 06) On top of the errors in Xorg.0.log, there is a WARN in my dmesg. Just for the record, I have not tested UXA.
Created attachment 81850 [details] dmesg output
Today's wtf: Fatal server error: [ 12419.606] (EE) xf86OpenConsole: VT_ACTIVATE failed: Input/output error [ 12419.606] (EE) [ 12419.606] (EE) No idea.
Did anything else change in the meantime? Can you please bisect if 2.21.10 still works (my guess would be the refcounting of drmMaster is off, but I have no idea how that would result in that particular error).
(In reply to comment #3) > Did anything else change in the meantime? Nope. It works on 2.21.10, I upgrade, log out, it does not work, I downgrade xf86-v-intel via ssh, restart gdm via ssh (as I cannot change the vt at that point) and everything (X, vt changing) works again. > Can you please bisect if 2.21.10 still works? You mean bisect between 2.21.10 and 2.21.11? I can do that. I will report back later, then.
git bisect shows this: # first bad commit: [8a6a21bff86100144ba7960fc32a299ac54ada83] sna: Use the existing configuration for initial modes What I also noticed is that X does indeed come up (although with no display). I can blindly log in and do stuff. Don't know if this is relevant.
Spotted one bogosity in your Xorg.log, we tried to setup both HDMI/DP connectors on the same encoder. I've pushed commit c361b449cc3ec15819883afc220aad8823c0072d Author: Chris Wilson <chris@chris-wilson.co.uk> Date: Tue Jul 2 09:49:48 2013 +0100 sna: Include connector status in the initial probe We cannot simply rely on connector->encoder->crtc status as with the introduction of Haswell or SDVO we may have multiple connectors using the same encoder. So we need to explictly check the connector status first, before determining if the output is connected to an active encoder and CRTC. References: https://bugs.freedesktop.org/show_bug.cgi?id=66488 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> which should fix that, can you please build from git and attach the new Xorg.log?
Created attachment 81856 [details] Xorg.0.log at c361b44 which is working Yep, this is fixed with c361b44. Thanks
*** Bug 66491 has been marked as a duplicate of this bug. ***
Hmm, still not sure how that ended up with the screen being closed, but success is happiness...
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.