Intel integrated graphics boots in low graphics mode

[Adam Bruce]

After successfully installing 13.04 on my Dell XPS 15z, I restarted the laptop, successfully booting into Ubuntu with no issues.

Restarting after that then gives the 'System is running in low graphics mode' message.

I can only boot in recovery mode

As mentioned in bug 982889 comment#5

"A workaround is to add some sleep to /etc/init/lightdm.conf:

    sleep 10
    exec lightdm"

This works around the issue. However, with further testing it was found that bug 982889 was not being bit and the fix does not fix this bug.

The crash report is here: bug 1127023

Extra info:

This is an repost of bug 1124661 as I submitted that in recovery mode

Other dell users seem to be experiencing similar issues: http://ubuntuforums.org/showthread.php?t=2115355

I installed and use the parameters mentioned here: https://wiki.ubuntu.com/HardwareSupport/Machines/Laptops/Dell/XPS/15z

I have an optimus card, which the Nvidia card should be disabled as it does not work

ProblemType: Bug
DistroRelease: Ubuntu 13.04
Package: xorg 1:7.7+1ubuntu4
ProcVersionSignature: Ubuntu 3.8.0-6.13-generic 3.8.0-rc7
Uname: Linux 3.8.0-6-generic x86_64
.tmp.unity.support.test.0:

ApportVersion: 2.8-0ubuntu4
Architecture: amd64
CompizPlugins: No value set for `/apps/compiz-1/general/screen0/options/active_plugins'
CompositorRunning: compiz
CompositorUnredirectDriverBlacklist: '(nouveau|Intel).*Mesa 8.0'
CompositorUnredirectFSW: true
Date: Fri Feb 15 01:50:36 2013
DistUpgraded: Fresh install
DistroCodename: raring
DistroVariant: ubuntu
ExtraDebuggingInterest: Yes, including running git bisection searches
GraphicsCard:
 Intel Corporation 2nd Generation Core Processor Family Integrated Graphics Controller [8086:0126] (rev 09) (prog-if 00 [VGA controller])
   Subsystem: Dell Device [1028:0446]
 NVIDIA Corporation GF108 [GeForce GT 540M] [10de:0df5] (rev a1) (prog-if 00 [VGA controller])
   Subsystem: Dell Device [1028:0446]
InstallationDate: Installed on 2013-02-11 (3 days ago)
InstallationMedia: Ubuntu 13.04 "Raring Ringtail" - Alpha amd64 (20130210)
MachineType: Dell Inc. Dell System XPS 15Z
MarkForUpload: True
ProcEnviron:
 LANGUAGE=en_GB:en
 PATH=(custom, no user)
 XDG_RUNTIME_DIR=<set>
 LANG=en_GB.UTF-8
 SHELL=/bin/bash
ProcKernelCmdLine: BOOT_IMAGE=/boot/vmlinuz-3.8.0-6-generic root=UUID=93acc5a7-d531-4af3-a215-472a7bd065d6 ro acpi_backlight=vendor dell_laptop.backlight=0 quiet splash vt.handoff=7
SourcePackage: xorg
Symptom: display
UpgradeStatus: No upgrade log present (probably fresh install)
XorgConf:
 Section "Device"
    Identifier "Intel Graphics"
    Driver "intel"
    Option "AccelMethod" "uxa"
 EndSection
dmi.bios.date: 07/27/2011
dmi.bios.vendor: Dell Inc.
dmi.bios.version: A04
dmi.board.name: 060G42
dmi.board.vendor: Dell Inc.
dmi.board.version: A00
dmi.chassis.type: 8
dmi.chassis.vendor: Dell Inc.
dmi.chassis.version: 0.1
dmi.modalias: dmi:bvnDellInc.:bvrA04:bd07/27/2011:svnDellInc.:pnDellSystemXPS15Z:pvr:rvnDellInc.:rn060G42:rvrA00:cvnDellInc.:ct8:cvr0.1:
dmi.product.name: Dell System XPS 15Z
dmi.sys.vendor: Dell Inc.
version.compiz: compiz 1:0.9.9~daily13.02.08-0ubuntu1
version.ia32-libs: ia32-libs N/A
version.libdrm2: libdrm2 2.4.42-0ubuntu1
version.libgl1-mesa-dri: libgl1-mesa-dri 9.0.2-0ubuntu1
version.libgl1-mesa-dri-experimental: libgl1-mesa-dri-experimental N/A
version.libgl1-mesa-glx: libgl1-mesa-glx 9.0.2-0ubuntu1
version.xserver-xorg-core: xserver-xorg-core 2:1.13.2-0ubuntu2
version.xserver-xorg-input-evdev: xserver-xorg-input-evdev 1:2.7.3-0ubuntu2
version.xserver-xorg-video-ati: xserver-xorg-video-ati 1:7.1.0-0ubuntu1
version.xserver-xorg-video-intel: xserver-xorg-video-intel 2:2.21.2-0ubuntu1
version.xserver-xorg-video-nouveau: xserver-xorg-video-nouveau 1:1.0.6-0ubuntu2

[AG Restringere (ag-restringere-deactivatedaccount)]

@Adam Bruce: I suggest you read these guides on how to use Bumblebee to enable Optimus support as that seems to be the issue at hand...

https://wiki.ubuntu.com/HardwareSupport/Machines/Laptops/Dell/XPS/15z
https://wiki.ubuntu.com/Bumblebee

Let me know if this helps...

[Adam Bruce (brucey-99-deactivatedaccount)]

I'm hesitant on doing this, as I did try it for 12.04 and it only caused more issues.

Having the Nvidia disabled has worked fine for 11.10, 12.04 and 12.10

Although if you're sure it's the issue, I will try and find another hard drive to test it on?

[Chris Wilson (ickle)]

Still not an SNA bug :-p

When you specify UXA manually through the xorg.conf snippet that overrides the GPU autoconfiguration, and disables activation of the nvidia GPU and sharing with nouveau through PRIME. When you leave it at its defaults, then it explodes inside nouveau.

Also it appears that you are experiencing the plymouthd race.

[Chris Wilson (ickle)]

If you were to change UXA to SNA in your xorg.conf snippet, it should also work. Can you please test that?

[Chris Wilson (ickle)]

"Also it appears that you are experiencing the plymouthd race." which only causes the fbdev fallback with UXA...

[Adam Bruce (brucey-99-deactivatedaccount)]

"If you were to change UXA to SNA in your xorg.conf snippet, it should also work. Can you please test that?"

You are correct!

Let me know if there are any further tests you want me to do

[Timo Aaltonen (tjaalton)]

[ 4.377] setversion 1.4 failed

yes, this is the race bug. Does it boot up properly more often if you add 'plymouth.enable=0' to the kernel boot options?

[Chris Wilson (ickle)]

Also note in the fbdev fallback case the error is inside RandR, specifically it fails to check that the root window is not NULL before calling xf86RandR12TellChanged() in hw/xfree86/modes/xf86Crtc.c::xf86SetSingleMode(). See sna_set_fallback_mode() for a comparison.

[Chris Wilson (ickle)]

As for the nouveau crash, it looks like it does not correctly uncouple itself from the BlockHandler chain upon removal - but I can't see anything obvious there. (Time to look for a less obvious cause.) As for why nouveau is removed and replaced with the modesetting driver???

[Adam Bruce (brucey-99-deactivatedaccount)]

"Does it boot up properly more often if you add 'plymouth.enable=0' to the kernel boot options?"

So I removed the /etc/X11/xorg.conf file and added plymouth.enable=0 to the boot options (after quiet splash - correct?)

And no, it does not boot up properly, it still goes to low graphics mode

[Chris Wilson (ickle)]

Can you attach the crashing Xorg.0.log from that boot, as I'd like to see if the nouveau failure is related to that race? I don't think it is, but it will be interesting nevertheless.

[Adam Bruce (brucey-99-deactivatedaccount)]

Can you explain how to find this? I have searched my system for that file name and nothing is found

I have 3 files in /var/crash/ if that helps?

See attached

[Chris Wilson (ickle)]

Hmm, they should make sense to the bot, but I don't know what tools I need to use on ubuntu to decode them. As it stands it didn't have the information I was looking for.

There should be /var/log/lightdm/:0.log (and friends). See if you can find one that corresponds to the crash upon boot.

[Adam Bruce (brucey-99-deactivatedaccount)]

OK, I have attached the files from that folder, there are only one of each, so I don't know how to tell which crash it's from?

[Adam Bruce (brucey-99-deactivatedaccount)]

I just noticed that Xorg.0.log file in /var/log/ if it helps

Again, I don't know which boot it's from

Should I delete those files and then repeat the boot?

[Bryce Harrington (bryce)]

apport-unpack can be used to decompress the .crash file to review things locally, however note this doesn't run the apport hook itself, so it contains a subset of the stuff usually collected into bug reports.

apport-collect <bug-number> run from the troubled machine will collect the appropriate data files, and post them to the bug report. Run it only after reproducing the error (e.g. while system is in failsafe, or locked up or whatever.)

[Bryce Harrington (bryce)]

The xserver crash we'll follow up on as bug #1127023 which has full stack trace info. The plymouthd race (actually we're not sure plymouth is really the root cause) is effectively dupe of #982889, which is where we're trying to centralize development discussion for that bug. I'll add this as a dupe of the latter issue.

[Adam Bruce (brucey-99-deactivatedaccount)]

Bryce, I have added sleep to lightdm.conf as mentioned in comment #5 in bug 982889

This boots up perfectly everytime, even without the /etc/X11/xorg.conf file I needed before

So I guess we can assume plymouth is the issue here, as detailed in bug 982889

I'll keep an eye on that report and let you know if it fixes this bug when that one is fixed

Thanks for your efforts!

[Ernie 07 (ernestboyd)]

64-bit 3.8.0-9-generic #18-Ubuntu SMP Thu Feb 28 17:02:06 UTC 2013
Desktop with Intel e8500 cpu, Nvidia GT9500 and default (non-Nvidia) video driver.

I get this problem frequently following a restart from 12.04.

Pressing the reset button or executing a FULL power shutdown prior to completion of booting 13.04 will usually eliminate this problem. This would imply that a hardware state that is properly initialized following a reset or full power shutdown is NOT being properly initialized during a warm reboot.

On occasion, I have also had to update/install grub, either from 12.04 or a via fix a broken system from a server CD.

[Adam Bruce (brucey-99-deactivatedaccount)]

After testing in bug 982889 I think we found I was never hitting the same bug there, so I will unduplicate this for now

[Adam Bruce (brucey-99-deactivatedaccount)]

Let me know what I need to do to change this from Incomplete

[Chris Wilson (ickle)]

I don't understand: the fix for 982889 is to delay X starting (technically lightdm starting) until plymouth finishes - which is exactly the symptom you have.

[Adam Bruce (brucey-99-deactivatedaccount)]

Actually since, bug 1157614 was fixed I no longer get put into Low Graphics mode. So I guess it is best to mark as a dup of that?

Not sure why, but bug 982889 never fixed it for me

[Chris Wilson (ickle)]

I don't mind so long as we don't have a mysterious unresolved bug on our hands! Thanks for your feedback.