kitkat i915 black screen

946 views
Skip to first unread message

Adriano Gradenigo

unread,
Feb 19, 2014, 11:38:51 AM2/19/14
to andro...@googlegroups.com
Hi all,
i'm trying to install android-x86 in an asus eeetop et1602. it has i915 video card.

when i boot the kitkat iso, logcat is telling me:

E/GRALLOC-KMS (5197): failed to set crtc (permission denied) (crtc_id 3, fb_id 28, com 8, mode 1368x768)

so i think that is something wrong with the kms module but i see that it's trying to set the resolution 1368x768 which is not the correct resolution for my screen. the correct one is 1366x768
Maybe is the wrong resolution that is messing up surfaceflinger?

Note. when i have the black screen i hear the sound of buttons when pressed and from log i can see that the system is runnig and it is in the configuration screen.
So it could be a simple fix.

Can someone give me a hint on where and what i have to change in the source code to compile a version with working graphics?  

Chih-Wei Huang

unread,
Feb 20, 2014, 8:29:37 PM2/20/14
to Android-x86
2014-02-20 0:38 GMT+08:00 Adriano Gradenigo <adriano....@gmail.com>:
> Hi all,
> i'm trying to install android-x86 in an asus eeetop et1602. it has i915
> video card.
>
> when i boot the kitkat iso, logcat is telling me:
>
> E/GRALLOC-KMS (5197): failed to set crtc (permission denied) (crtc_id 3,
> fb_id 28, com 8, mode 1368x768)
>
> so i think that is something wrong with the kms module but i see that it's
> trying to set the resolution 1368x768 which is not the correct resolution
> for my screen. the correct one is 1366x768
> Maybe is the wrong resolution that is messing up surfaceflinger?

You can try to change it in /system/etc/init.sh
(in source tree: device/generic/x86/init.sh)
Find the function:

function set_drm_mode()
{
case "$PRODUCT" in
ET1602*)
drm_mode=1366x768
;;
...
}


> Note. when i have the black screen i hear the sound of buttons when pressed
> and from log i can see that the system is runnig and it is in the
> configuration screen.
> So it could be a simple fix.
>
> Can someone give me a hint on where and what i have to change in the source
> code to compile a version with working graphics?

You may also try the property:

setprop debug.drm.mode 1366x768


--
Chih-Wei
Android-x86 project
http://www.android-x86.org

Greg McGee

unread,
Feb 22, 2014, 12:42:18 AM2/22/14
to andro...@googlegroups.com
On i915 lenovo thinkpad x200 and x300 (gm45 for x200)
same issue.

Forced resolution to 1280x800 using desribed method,  on my x200 and pulled a clean debug log.
this does not hang like jb did at first, i can still switch to a vt and do things and reboot.

there were some issues with i915 that were recently fixed in 3.12 kernel, headed off to see what patches Mageia4 uses.
gzipped log attached.
log.txt.gz

Adriano Gradenigo

unread,
Feb 22, 2014, 3:16:41 AM2/22/14
to andro...@googlegroups.com

Finally thanks to the tips of chih-wei i have found a solution.
The problem was in the init.rc file in which is used the fucntion debug.drm.mode.force to set the resolution but this function is not working, instead i used the function debug.drm.mode (without force) and now is workig very well.
It boots normally without a problem and have opengl enabeled.
I’m very happy now this kitkat version is working very very well. I love it.
Another big thanks to chih-wei huang

Greg McGee

unread,
Feb 23, 2014, 1:55:38 AM2/23/14
to andro...@googlegroups.com
Also tried debug,drm.mode (with .force removed) in /etc/init.rc to no effect on my X200 on the rc1 and a local build.
Kernel 3.12.9 does seem to work ~splendidly otherwise with nomodeset

Chih-Wei Huang

unread,
Feb 23, 2014, 8:35:06 PM2/23/14
to Android-x86
2014-02-22 16:16 GMT+08:00 Adriano Gradenigo <adriano....@gmail.com>:
>
> Finally thanks to the tips of chih-wei i have found a solution.
> The problem was in the init.rc file in which is used the fucntion
> debug.drm.mode.force to set the resolution but this function is not working,
> instead i used the function debug.drm.mode (without force) and now is workig
> very well.

OK. ET1602 has several skus that their graphic part have a little different.
(I don't know what differs exactly -- just know they are different)
Actually debug.drm.mode.force is designed for my sku
which can't detect the native resolution correctly.
If your sku can detect the resolution correctly, you don't need the
.force property.
(check logcat for all the available modes)

Could you show me the output of the command?
cat /sys/class/dmi/id/uevent

Probably setting both properties in the init.sh is better.
But I'm not sure if it has any side effect.
Could you please try it?

Adriano Gradenigo

unread,
Feb 26, 2014, 4:29:20 AM2/26/14
to andro...@googlegroups.com
Here is my logcat and uevent.
tell me if you notice something starnge.
thanks
uevent.txt
log.txt

Chih-Wei Huang

unread,
Feb 26, 2014, 6:40:09 AM2/26/14
to Android-x86
2014-02-26 17:29 GMT+08:00 Adriano Gradenigo <adriano....@gmail.com>:
> Here is my logcat and uevent.
> tell me if you notice something starnge.

The log.txt seems fine.
It boots with debug.drm.mode=1366x768, right?
You should show me the log without debug.drm.mode.
Reply all
Reply to author
Forward
0 new messages