You are not logged in.
I'm not sure if this is releated to a recent update (xf86-video-intel 2.99.916-2) or not.
On my HP G60 laptop I have
$ inxi -Gx
Graphics: Card: Intel Mobile 4 Series Integrated Graphics Controller
bus-ID: 00:02.0
Display Server: X.Org 1.16.0 driver: intel
Resolution: 1366x768@60.00hz
GLX Renderer: Mesa DRI Mobile Intel GM45 Express
GLX Version: 2.1 Mesa 10.2.8 Direct Rendering: Yes
$ uname -a
Linux HpLappy 3.16.3-1-ARCH #1 SMP PREEMPT Wed Sep 17 21:54:13 CEST 2014 x86_64 GNU/Linux
$ pacman -Qs intel
local/intel-dri 10.2.8-1
Mesa drivers for Intel
local/ipw2100-fw 1.3-7
Intel Centrino Drivers firmware for IPW2100
local/ipw2200-fw 3.1-5
Firmware for the Intel PRO/Wireless 2200BG
local/lib32-intel-dri 10.2.7-2
Mesa DRI drivers for Intel (32-bit)
local/libva-intel-driver 1.3.2-1
VA-API implementation for Intel G45 and HD Graphics family
local/xf86-video-intel 2.99.916-1 (xorg-drivers xorg)
X.org Intel i810/i830/i915/945G/G965+ video drivers
Until recently things seemed to be working fine (although this is a fairly new installation)
Suddenly, Cinnamon crashes and Enlightenment falls back to software compositing.
xorg.log showing
intel_do_flush_locked failed: Input/output error
and system log shows
$ dmesg | grep -i fail
[ 8.820031] [drm:init_ring_common] *ERROR* render ring initialization failed ctl 0001f001 (valid? 1) head 00000298 tail 00000000 start 00303000 [expected 00303000]
[ 8.820047] [drm:i915_gem_init] *ERROR* Failed to initialize GPU, declaring it wedged
After some fiddlling I did get Cinnamon to use fallback as well.
THE FIX (for me, so far): Add
i915.i915_enable_rc6=0
to kernel boot parameters
In
/etc/default/grub
it looks like this:
GRUB_CMDLINE_LINUX="i915.i915_enable_rc6=0"
UPDATE: NEVER MIND ALL OF THE ABOVE. Somehow multilib was disable so I was running some old lib32 packages.
With a fully updated system the fix does not work (UXA does not work either).
I found the original issue being tracked here
Last edited by bartonc (2014-09-22 07:34:03)
Offline
I don't really know anything about that kernel parameter - but as of a couple months ago it seems a very wide range of intel gpus that were previously fine with SNA acceleration now show all sorts of issues. Switching to UXA (see intel wiki page for how) fixes it up nicely.
Perfomance drops a bit, but if this works for your issues it should be far better than software fallbacks. It also seems that there is even a full fix for this that may be in linux 3.17 - though I don't fully understand all chatter about the various current known bugs and patches that are coming in 3.17.
"UNIX is simple and coherent..." - Dennis Ritchie, "GNU's Not UNIX" - Richard Stallman
Offline
<snip> Switching to UXA (see intel wiki page for how) fixes it up nicely</snip>.
I tried that route first (got Cinnamon to the desktop in fallback), but no DE would use hardware compositing.
Thanks for the extra input on this issue.
Offline