Bug #1277343 “XMir starts with black screen ("[xmir] Failed to s...” : Bugs : mir package : Ubuntu

Revision history for this message

Stephen Cradock (s-cradock) wrote on 2014-02-07:

#1

BootDmesg.txt Edit (63.2 KiB, text/plain; charset="utf-8")
BootLog.txt Edit (1.8 KiB, text/plain; charset="utf-8")
CurrentDmesg.txt Edit (4.2 KiB, text/plain; charset="utf-8")
Dependencies.txt Edit (3.9 KiB, text/plain; charset="utf-8")
DpkgLog.txt Edit (338.5 KiB, text/plain; charset="utf-8")
GconfCompiz.txt Edit (45.5 KiB, text/plain; charset="utf-8")
LightdmDisplayLog.txt Edit (2.2 KiB, text/plain; charset="utf-8")
LightdmGreeterLog.txt Edit (24.1 KiB, text/plain; charset="utf-8")
LightdmGreeterLogOld.txt Edit (8.8 KiB, text/plain; charset="utf-8")
LightdmLog.txt Edit (3.7 KiB, text/plain; charset="utf-8")
Lspci.txt Edit (15.1 KiB, text/plain; charset="utf-8")
Lsusb.txt Edit (807 bytes, text/plain; charset="utf-8")
MonitorsUser.xml.txt Edit (2.1 KiB, text/plain; charset="utf-8")
ProcCpuinfo.txt Edit (1.5 KiB, text/plain; charset="utf-8")
ProcEnviron.txt Edit (103 bytes, text/plain; charset="utf-8")
ProcInterrupts.txt Edit (1.6 KiB, text/plain; charset="utf-8")
ProcModules.txt Edit (3.4 KiB, text/plain; charset="utf-8")
UdevDb.txt Edit (174.9 KiB, text/plain; charset="utf-8")
UdevLog.txt Edit (306.0 KiB, text/plain; charset="utf-8")
UnitySupportTest.txt Edit (685 bytes, text/plain; charset="utf-8")
XorgLog.txt Edit (76.3 KiB, text/plain; charset="utf-8")
XorgLogOld.txt Edit (29.4 KiB, text/plain; charset="utf-8")
Xrandr.txt Edit (1009 bytes, text/plain; charset="utf-8")
xdpyinfo.txt Edit (30.7 KiB, text/plain; charset="utf-8")
xserver.devices.txt Edit (427 bytes, text/plain; charset="utf-8")

Revision history for this message

Chris Wilson (ickle) wrote on 2014-02-07:

#2

It's a mir issue.

affects:

xserver-xorg-video-intel (Ubuntu) → xorg-server (Ubuntu)

Revision history for this message

Timo Aaltonen (tjaalton) wrote on 2014-02-07:

#3

upgrade again, mir abi had changed

Changed in xorg-server (Ubuntu):
status:	New → Incomplete

Revision history for this message

Timo Aaltonen (tjaalton) wrote on 2014-02-07:

#4

hmm no you have the latest versions..

Revision history for this message

Timo Aaltonen (tjaalton) wrote on 2014-02-07:

#5

[ 41.737] (EE) [xmir] Failed to set new display config: /build/buildd/mir-0.1.4+14.04.20140204/src/platform/graphics/mesa/real_kms_display_configuration.cpp(130): Throw in function virtual void mir::graphics::mesa::RealKMSDisplayConfiguration::configure_output(mir::graphics::DisplayConfigurationOutputId, bool, mir::geometry::Point, size_t, MirPixelFormat, MirPowerMode, MirOrientation)
Dynamic exception type: boost::exception_detail::clone_impl<boost::exception_detail::error_info_injector<std::runtime_error> >
std::exception::what: Invalid format for used output

Changed in xorg-server (Ubuntu):
status:	Incomplete → New
summary:	- Trusty boots to incorrect display size/ratio + xmir crashes on startup

Revision history for this message

Stephen Cradock (s-cradock) wrote on 2014-02-08: Re: xmir crashes on startup

#6

OK - sounds as if there is a problem - hope it gets fixed soon.

I tried downgrading to the previous versions of xserver-xorg-xmir and xserver-xorg-core (2:1.15.0-1ubuntu2), but no help.

Then I removed ubuntu-desktop-xmir (and xserver-xorg-xmir) and all is well again.

Revision history for this message

Launchpad Janitor (janitor) wrote on 2014-02-08:

#7

Status changed to 'Confirmed' because the bug affects multiple users.

Changed in mir (Ubuntu):
status:	New → Confirmed
Changed in xorg-server (Ubuntu):
status:	New → Confirmed

Javier López (javier-lopez) on 2014-02-09

Changed in xorg-server (Ubuntu):
importance:	Undecided → Critical

Revision history for this message

Alberto Salvia Novella (es20490446e) wrote on 2014-02-09:

#9

Renders the system temporarily or permanently unusable.

Changed in mir (Ubuntu):
importance:	Undecided → Critical

Daniel van Vugt (vanvugt) on 2014-02-10

Changed in mir:
status:	New → Confirmed
importance:	Undecided → Critical
assignee:	nobody → Daniel van Vugt (vanvugt)
summary:	- xmir crashes on startup + xmir crashes on startup ("std::exception::what: Invalid format for used + output")

Revision history for this message

Daniel van Vugt (vanvugt) wrote on 2014-02-10: Re: xmir crashes on startup ("std::exception::what: Invalid format for used output")

#10

Although I didn't write it, this exception is in a region of code I am actively refactoring. And the placement of this validity check is something I was going to have to modify as part of that.

Revision history for this message

Daniel van Vugt (vanvugt) wrote on 2014-02-10:

#11

Unfortunately bug 1237332 prevents us from finding the root cause of the crash.

Revision history for this message

Rick Harris (rickfharris) wrote on 2014-02-10:

#12

It would seem the offending code comes from https://code.launchpad.net/~andreas-pokorny/mir/add-pixel-format-to-display-configuration/+merge/200294 ?

Revision history for this message

Daniel van Vugt (vanvugt) wrote on 2014-02-10:

#13

Yeah, I'm not convinced the validation check is useful or even correct. But I will have to review it in detail soon.

Revision history for this message

Daniel van Vugt (vanvugt) wrote on 2014-02-14:

#14

The offending validation check has been removed, moved and changed in this proposal:
lp:~vanvugt/mir/eliminate-configure-output

Unfortunately, without details of what the invalid format being used was, I can't tell if the issue is fixed or could simply reoccur in a new location.

Revision history for this message

Stephen Cradock (s-cradock) wrote on 2014-02-14:

#15

After updating to xserver-xorg-xmir 2:1.15.0-1ubuntu6 on Feb 13th the system reboots without problems. A whole set of mir packages went through at the same time; don't know if Daniel's fix was included, but something fixed my problem at least.

Thanks for the quick solution.

Revision history for this message

Daniel van Vugt (vanvugt) wrote on 2014-02-14:

#16

No the proposal in comment #14 has not landed in any release yet.

Changed in mir:
assignee:	Daniel van Vugt (vanvugt) → nobody
status:	Confirmed → Incomplete
Changed in mir (Ubuntu):
status:	Confirmed → Incomplete
Changed in xorg-server (Ubuntu):
status:	Confirmed → Incomplete
Changed in mir:
status:	Incomplete → Confirmed
importance:	Critical → High
Changed in mir (Ubuntu):
status:	Incomplete → Confirmed
importance:	Critical → High
Changed in xorg-server (Ubuntu):
status:	Incomplete → Confirmed
importance:	Critical → High

Revision history for this message

Rick Harris (rickfharris) wrote on 2014-02-15:

#17

Big thanks, applying the preview.diff from lp:~vanvugt/mir/eliminate-configure-output fixes the crash for me :)

Revision history for this message

Rick Harris (rickfharris) wrote on 2014-02-16:

#18

Only problem the elimination presents now is that unity-system-compositor will fail to build when it doesn't see 'configure_output'.

unity-system-compositor-0.0.2+14.04.20140212.1/src/dbus_screen.cpp:66:28: error: 'class mir::graphics::DisplayConfiguration' has no member named 'configure_output'
displayConfig->configure_output(
^
src/CMakeFiles/unity-system-compositor.dir/build.make:97: recipe for target 'src/CMakeFiles/unity-system-compositor.dir/dbus_screen.cpp.o' failed
make[2]: *** [src/CMakeFiles/unity-system-compositor.dir/dbus_screen.cpp.o] Error 1

Revision history for this message

Daniel van Vugt (vanvugt) wrote on 2014-02-17:

#19

Thanks. When/if that branch lands in Mir, I'll be sure to fix unity-system-compositor et al.

Daniel van Vugt (vanvugt) on 2014-02-17

Changed in mir:
status:	Confirmed → Triaged
Changed in mir (Ubuntu):
status:	Confirmed → Triaged
Changed in xorg-server (Ubuntu):
status:	Confirmed → Triaged
Changed in xmir:
status:	New → Triaged
importance:	Undecided → High

Daniel van Vugt (vanvugt) on 2014-02-17

summary:	- xmir crashes on startup ("std::exception::what: Invalid format for used - output") + XMir starts with black screen ("std::exception::what: Invalid format for + used output")
summary:	- XMir starts with black screen ("std::exception::what: Invalid format for - used output") + XMir starts with black screen ("[xmir] Failed to set new display config: + ... std::exception::what: Invalid format for used output")

Daniel van Vugt (vanvugt) on 2014-02-17

Changed in mir:
assignee:	nobody → Daniel van Vugt (vanvugt)
status:	Triaged → In Progress
milestone:	none → 0.1.6

Revision history for this message

Daniel van Vugt (vanvugt) wrote on 2014-02-17:

#20

OK, I just realized "eliminate-configure-output" doesn't fix the bug. The branch itself has a bug which prevents this bug from manifesting (which it still should).

Daniel van Vugt (vanvugt) on 2014-02-17

Changed in xmir:
status:	Triaged → Invalid
Changed in xorg-server (Ubuntu):
status:	Triaged → Invalid

Revision history for this message

Edwin Pujols (edwinpm5) wrote on 2014-02-17:

#21

Tried u-s-c (with ubuntu-desktop-mir). At first I got a blank screen, but after running `sudo restart lightdm` I was able to log into Unity. The fonts looks a a bit weird and some apps (Firefox, VLC, file selector) are not themed. Otherwise, usable... Nonetheless, when I removed these packages with `remove --auto-remove` and rebooting, lightdm wouldn't start (and I was presented with the "Low graphics" dialog). So for now I'll have to use XMir.

Revision history for this message

Edwin Pujols (edwinpm5) wrote on 2014-02-17:

#22

After starting `gnome-settings-daemon` manually, Firefox (and the other apps) have recovered their theme. The fonts still look funny, though.

Revision history for this message

Edwin Pujols (edwinpm5) wrote on 2014-02-17:

#23

Just did a dist-upgrad (maybe the last one got interrupted, or this is a new update), and the fonts look good again and gnome-settings-daemon started fine (no theming issue). The blank screen persists, and I still need to restart lightdm as a workaround.

Revision history for this message

Daniel van Vugt (vanvugt) wrote on 2014-02-18:

#24

Edwin - That's an X bug. Mir and unity-system-compositor only deal the the screen as a whole. Not the pixels/fonts etc. Please log the bug elsewhere.

Revision history for this message

PS Jenkins bot (ps-jenkins) wrote on 2014-02-18:

#25

Fix committed into lp:mir/devel at revision None, scheduled for release in mir, milestone Unknown

Changed in mir:
status:	In Progress → Fix Committed

Revision history for this message

Daniel van Vugt (vanvugt) wrote on 2014-02-19:

#26

A fix is ready for Mir 0.1.6. Maybe we can try to get it into Ubuntu before 0.1.6 even...

kevin gunn (kgunn72) on 2014-02-20

tags:	added: xmir
Changed in mir (Ubuntu):
status:	Triaged → Fix Committed

Revision history for this message

Daniel van Vugt (vanvugt) wrote on 2014-02-21:

#27

Not Fix Committed to Ubuntu just yet...
https://code.launchpad.net/~mir-team/mir/trusty

Changed in mir (Ubuntu):
status:	Fix Committed → Triaged

Daniel van Vugt (vanvugt) on 2014-02-28

Changed in mir:
status:	Fix Committed → Fix Released

Revision history for this message

Rick Harris (rickfharris) wrote on 2014-02-28:

#28

Thanks again Daniel.
http://bazaar.launchpad.net/~mir-team/mir/development-branch/revision/1405 fixes nicely, look forward to 0.1.6.

Revision history for this message

Joseph Wakeling (webdrake) wrote on 2014-03-01:

#29

Is there an ETA for when this will land in Trusty ... ?

Revision history for this message

Daniel van Vugt (vanvugt) wrote on 2014-03-04:

#30

We're trying to land Mir 0.1.6 in trusty right now. There's no predicting how long that can take. Sometimes a couple of hours, sometimes a week. Keep an eye out: https://launchpad.net/ubuntu/+source/mir

Revision history for this message

Launchpad Janitor (janitor) wrote on 2014-03-11:

#31

Download full text (3.9 KiB)

This bug was fixed in the package mir - 0.1.6+14.04.20140310-0ubuntu1

---------------
mir (0.1.6+14.04.20140310-0ubuntu1) trusty; urgency=medium

  [ Kevin Gunn ]
  * Cherry-picked from future release 0.1.7:
    - Enhancement: options: Make it easier to customize DefaultConfiguration
    - Bug fix: Don't pass a parameter to bash's popd command. It doesn't expect
      one and will return an error from cross-compile-chroot.sh. (LP: #1287600)
  * New upstream release 0.1.6 (https://launchpad.net/mir/+milestone/0.1.6)
    - mirserver ABI bumped to 16.
    - Server API changes:
      . Report classes have moved to new namespaces/components. Please
        consult the server headers.
      . General clean-ups and virtual function changes affecting the ABI
        (a rebuild is required for all shells).
    - mirclient ABI unchanged, still at 7. Clients do not need rebuilding.
    - Add new feature: Screen recording ("screencasting") as demonstrated by
      the new mir-util "mirscreencast".
    - debian: Provide platform packages managed with dpkg alternatives.
    - Add new valgrind suppressions (mostly for armhf right now)
    - tools: Add new libmirclientplatform.so to install_on_android.sh.
    - Add "mir_demo_standalone_render_overlays", which is a free-standing
      example of overlay support.
    - Proper DisplayConfiguration for the AndroidDisplay to be better
      prepared for multi-monitor Android.
    - Commenced/continued simplifying the surface class hierarchy: BasicSurface
      has now absorbed SurfaceData.
    - Dynamically load the client platform library.
    - Bugs fixed:
      . Exceptions thrown from within compositing threads were untraceable,
        resulting in a bunch of undebuggable crash reports (LP: #1237332)
      . Ensure we close drm device file descriptors on exec. (LP: #1284081)
      . nested render_surfaces fails on N4 [std::exception::what: Failed to
        compile vertex shader:] (LP: #1284597)
      . examples: correct "fullscreen-surfaces" option. (LP: #1284554)
      . Fix mutex data race reported by helgrind in integration test:
        SwapperSwappingStress (LP: #1282886)
      . scene: Return null snapshots for sessions without surfaces
        (LP: #1281728)
      . tests: Address some raciness in
        MesaDisplayTest.drm_device_change_event_triggers_handler.
        (LP: #1283085)
      . tests: Suppress spurious memory errors occuring when running the
        unit tests on armhf with valgrind. (LP: #1279438)
      . Mir shows an old frame on client startup (for Mesa GL clients)
        (LP: #1281938)
      . tests: Add expectation to validate second SessionAuthorizer method
        is called wit correct parameter. (LP: #1218436)
      . Set field "current_format" when sending a display config to the
        server. (LP: #1277343)
      . Fix crash in android devices by working around a subtle threading
        bug, use a dummy thead_local array to push the gl/egl context TLS
        into a region where the future wait code does not overwrite it.
        (LP: #1280086)
      . compositor: calling SwitchingBundle::force_requests_to_complete()
        can be a no-op when there are no requests. (LP...

This bug was fixed in the package mir - 0.1.6+14.04.20140310-0ubuntu1

---------------
mir (0.1.6+14.04.20140310-0ubuntu1) trusty; urgency=medium

[ Kevin Gunn ]
  * Cherry-picked from future release 0.1.7:
    - Enhancement: options: Make it easier to customize DefaultConfiguration
    - Bug fix: Don't pass a parameter to bash's popd command. It doesn't expect
      one and will return an error from cross-compile-chroot.sh. (LP: #1287600)
  * New upstream release 0.1.6 (https://launchpad.net/mir/+milestone/0.1.6)
    - mirserver ABI bumped to 16.
    - Server API changes:
      . Report classes have moved to new namespaces/components. Please
        consult the server headers.
      . General clean-ups and virtual function changes affecting the ABI
        (a rebuild is required for all shells).
    - mirclient ABI unchanged, still at 7. Clients do not need rebuilding.
    - Add new feature: Screen recording ("screencasting") as demonstrated by
      the new mir-util "mirscreencast".
    - debian: Provide platform packages managed with dpkg alternatives.
    - Add new valgrind suppressions (mostly for armhf right now)
    - tools: Add new libmirclientplatform.so to install_on_android.sh.
    - Add "mir_demo_standalone_render_overlays", which is a free-standing
      example of overlay support.
    - Proper DisplayConfiguration for the AndroidDisplay to be better
      prepared for multi-monitor Android.
    - Commenced/continued simplifying the surface class hierarchy: BasicSurface
      has now absorbed SurfaceData.
    - Dynamically load the client platform library.
    - Bugs fixed:
      . Exceptions thrown from within compositing threads were untraceable,
        resulting in a bunch of undebuggable crash reports (LP: #1237332)
      . Ensure we close drm device file descriptors on exec. (LP: #1284081)
      . nested render_surfaces fails on N4 [std::exception::what: Failed to
        compile vertex shader:] (LP: #1284597)
      . examples: correct "fullscreen-surfaces" option. (LP: #1284554)
      . Fix mutex data race reported by helgrind in integration test:
        SwapperSwappingStress (LP: #1282886)
      . scene: Return null snapshots for sessions without surfaces
        (LP: #1281728)
      . tests: Address some raciness in
        MesaDisplayTest.drm_device_change_event_triggers_handler.
        (LP: #1283085)
      . tests: Suppress spurious memory errors occuring when running the
        unit tests on armhf with valgrind. (LP: #1279438)
      . Mir shows an old frame on client startup (for Mesa GL clients)
        (LP: #1281938)
      . tests: Add expectation to validate second SessionAuthorizer method
        is called wit correct parameter. (LP: #1218436)
      . Set field "current_format" when sending a display config to the
        server. (LP: #1277343)
      . Fix crash in android devices by working around a subtle threading
        bug, use a dummy thead_local array to push the gl/egl context TLS
        into a region where the future wait code does not overwrite it.
        (LP: #1280086)
      . compositor: calling SwitchingBundle::force_requests_to_complete()
        can be a no-op when there are no requests. (LP: #1281145)
      . tests: Don't emit an "add" uevent manually when adding a device
        (LP: #1281146)
      . config: reduce size of default RPC thread pool. (LP: #1233001)
      . Clients freeze on startup if 10 or more are already running
        (LP: #1267323)
      . Fix Nexus 10 leaking FDs in the hwc prepare/set loop. (LP: #1278658)
      . Parallelize page flipping with rendering of the next frame, fixing
        stuttering observed in multi-monitor clone mode (LP: #1213801)
      . Ubuntu trusty update "glm 0.9.5.1-1" broke Mir builds. Work around the
        change. (LP: #1285955)
      . Screencast feature by compositing to offscreen buffer (LP: #1207312)

[ Ubuntu daily release ]
  * New rebuild forced
 -- Ubuntu daily release <ps-jenkins@lists.canonical.com>   Mon, 10 Mar 2014 19:28:46 +0000

Changed in mir (Ubuntu):
status:	Triaged → Fix Released

Ubuntu
mir package

XMir starts with black screen ("[xmir] Failed to set new display config: ... std::exception::what: Invalid format for used output")

Bug Description

Related branches

Duplicates of this bug

Other bug subscribers

Bug attachments

Remote bug watches

	Status	Importance	Assigned to	Milestone
Mir	Fix Released	High	Daniel van Vugt	Mir 0.1.6
mir (Ubuntu)	Fix Released	High	Unassigned
xorg-server (Ubuntu)	Invalid	High	Unassigned

Ubuntumir package

XMir starts with black screen ("[xmir] Failed to set new display config: ... std::exception::what: Invalid format for used output")

Bug Description

Related branches

Duplicates of this bug

Other bug subscribers

Bug attachments

Remote bug watches

Ubuntu
mir package