New issue
Advanced search Search tips
Note: Color blocks (like or ) mean that a user may not be available. Tooltip shows the reason.

Issue 648205 link

Starred by 2 users

Issue metadata

Status: Duplicate
Merged: issue 647855
Owner:
OOO (till 1/28)
Closed: Oct 2016
Cc:
Components:
EstimatedDays: ----
NextAction: ----
OS: Chrome
Pri: 1
Type: Bug



Sign in to add a comment

Crash on fast-user-switch to personal profile after last weeks update

Project Member Reported by mglb@google.com, Sep 19 2016

Issue description

Chrome Version       : 53.0.2785.103
OS Version: 8530.81.0
URLs (if applicable) :
Other browsers tested:
  Add OK or FAIL after other browsers where you have tested this issue:
     Safari 5:
  Firefox 4.x:
     IE 7/8/9:

What steps will reproduce the problem?
1. Configure chromeos with a secondary user allowed for fast user switching
2. From the primary profile switch to the secondary profile (I'm using the Ctrl-Alt-> key combo)

What is the expected result?

Desktop switches to the secondary profile.


What happens instead of that?

Since rebooting for a ChromeOS update sometime late last week, I've observed this switch to trigger a crash (screen black before restoring) roughly 30-50% of the times when I've attempted to switch to the secondary profile. 

After the black screen restores, I end up in the profile I was trying to switch to, with an empty browser tab in the secondary profile (no crash notification, no tab restoration) - only way to recover is to select "N closed tabs" from the recent history menu.

If I switch back to the primary profile the browser there is displayign the yellow chrome crashed butter bar, with a restore button which when clicked restores the state of the primary profile. 


Please provide any additional information below. Attach a screenshot if
possible.

UserAgentString: Mozilla/5.0 (X11; CrOS x86_64 8530.81.0) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/53.0.2785.103 Safari/537.36


I'm not really sure how else to debug or report this. I've turned on debug logs, so hopefully I might catch somethign useful that you can use - I can trigger this pretty repeatably if I just switch between profiles a few times.

 

Comment 1 by 111...@phsdmail.org, Sep 19 2016

happens to me on my lenovo

Comment 2 by 111...@phsdmail.org, Sep 19 2016

are you getting aa personal sync problem?
Project Member

Comment 3 by sheriffbot@chromium.org, Sep 19 2016

Labels: Hotlist-Google

Comment 4 by ka...@chromium.org, Sep 20 2016

Components: UI>Browser>Profiles
Labels: Stability-Crash
Can you file a feedback report(Shift+Alt+i)?

Comment 5 by mglb@google.com, Sep 22 2016

The crash rate has declined significantly since I first experienced this, it took me over a day to reproduce it and file a feedback report. I'm not sure how to link it to this bug - I didn't get any sort of reference or link that I could see, but I made sure my email was included, and referenced this bug in the description text, so hopefully you can find it via that? 

Thanks
Are there any IDs in chrome://crashes?

Comment 7 by mglb@google.com, Sep 27 2016

Crash ID Chrome (Server ID: 5a7842cd00000000)

Automatically reported Saturday, September 24, 2016 at 3:54:58 PM

Provide additional details

Crash ID Chrome (Server ID: cd6ecf4d00000000)

Automatically reported Saturday, September 24, 2016 at 1:26:18 AM

Provide additional details

Crash ID Chrome (Server ID: 87bf194d00000000)

Automatically reported Friday, September 23, 2016 at 12:12:35 AM

Provide additional details

Crash ID Chrome (Server ID: a9b43d1e00000000)

Automatically reported Monday, September 19, 2016 at 11:59:06 PM

Provide additional details

Crash ID Chrome (Server ID: 3f4be60d00000000)

Automatically reported Sunday, September 18, 2016 at 12:29:04 AM

Provide additional details

Crash ID Chrome (Server ID: 41b467bd00000000)

Automatically reported Saturday, September 17, 2016 at 4:18:35 PM

Provide additional details

Crash ID Chrome (Server ID: a07bb20d00000000)

Automatically reported Saturday, September 17, 2016 at 2:07:01 PM

Provide additional details

Crash ID Chrome (Server ID: a410721e00000000)

Automatically reported Saturday, September 17, 2016 at 1:04:42 PM

Comment 8 by mglb@google.com, Oct 2 2016

Were the above Crash IDs helpful? Anything else I can provide to help track this down? 

It's not happening as frequently as I noted in the initial report, but it's still happening sufficiently often (e.g. once every day or two) to be significantly impacting to my workflow and annoying especially given the "restore" after crash feature is not reliable at bringing tabs back.

Comment 9 by mglb@google.com, Oct 2 2016

And on cue, tonight we're back to crashing on every switch.

Crash ID Chrome (Server ID: 456f29de00000000)

Automatically reported Sunday, October 2, 2016 at 11:24:44 PM

Provide additional details

Crash ID Chrome (Server ID: e6fb29de00000000)

Automatically reported Sunday, October 2, 2016 at 11:20:52 PM

Provide additional details

Crash ID Chrome (Server ID: 3d57b04300000000)

Automatically reported Sunday, October 2, 2016 at 12:22:20 AM

Provide additional details

Crash ID Chrome (Server ID: dc6c226d00000000)

Automatically reported Saturday, October 1, 2016 at 12:50:43 AM

Provide additional details

Crash ID Chrome (Server ID: ee75c26d00000000)

Automatically reported Saturday, October 1, 2016 at 12:43:13 AM

Provide additional details
Components: UI>Shell>Shelf
Labels: -Pri-3 Pri-1
Owner: mfomitchev@chromium.org
Status: Assigned (was: Unconfirmed)
Thread 0 CRASHED [SIGABRT @ 0x000003e80000121f ] MAGIC SIGNATURE THREAD
0x00007f0cf258bb82	(libc-2.19.so -raise.c:56 )	raise
0x00007f0cf258d89f	(libc-2.19.so -abort.c:89 )	abort
0x00007f0cf2e8a624	(libstdc++.so.6.0.20 -vterminate.cc:95 )	__gnu_cxx::__verbose_terminate_handler
0x00007f0cf2e88365	(libstdc++.so.6.0.20 -eh_terminate.cc:47 )	__cxxabiv1::__terminate
0x00007f0cf2e883b0	(libstdc++.so.6.0.20 -eh_terminate.cc:57 )	std::terminate
0x00007f0cf2e88620	(libstdc++.so.6.0.20 -eh_throw.cc:87 )	__cxxabiv1::__cxa_throw
0x00007f0cf2ee55d0	(libstdc++.so.6.0.20 -functexcept.cc:101 )	std::__throw_out_of_range_fmt
0x00007f0cf51c4dc1	(chrome -stl_vector.h:898 )	std::vector<CJS_Value, std::allocator<CJS_Value> >::_M_range_check
0x00007f0cf775b528	(chrome -stl_vector.h:864 )	ash::ShelfModel::Move
0x00007f0cf76428b0	(chrome -chrome_launcher_controller_impl.cc:1074 )	ChromeLauncherControllerImpl::RestoreUnpinnedRunningApplicationOrder
0x00007f0cf76456bf	(chrome -chrome_launcher_controller_impl.cc:704 )	ChromeLauncherControllerImpl::ActiveUserChanged
0x00007f0cf751bd69	(chrome -user_switch_animator_chromeos.cc:230 )	chrome::UserSwitchAnimatorChromeOS::TransitionUserShelf
0x00007f0cf751c41e	(chrome -user_switch_animator_chromeos.cc:156 )	chrome::UserSwitchAnimatorChromeOS::AdvanceUserTransitionAnimation
0x00007f0cf550d558	(chrome -callback.h:389 )	base::Timer::RunScheduledTask
0x00007f0cf4ca07dc	(chrome -callback.h:389 )	base::debug::TaskAnnotator::RunTask
0x00007f0cf4c8e7b8	(chrome -message_loop.cc:493 )	base::MessageLoop::RunTask
0x00007f0cf4c8f372	(chrome -message_loop.cc:502 )	base::MessageLoop::DoDelayedWork
0x00007f0cf4c896ee	(chrome -message_pump_libevent.cc:227 )	base::MessagePumpLibevent::Run
0x00007f0cf54f6497	(chrome -run_loop.cc:35 )	base::RunLoop::Run
0x00007f0cf8c4ecc4	(chrome -chrome_browser_main.cc:1943 )	ChromeBrowserMainParts::MainMessageLoopRun
0x00007f0cf7a13cba	(chrome -browser_main_loop.cc:942 )	content::BrowserMainLoop::RunMainMessageLoopParts
0x00007f0cf78483d4	(chrome -browser_main_runner.cc:155 )	content::BrowserMainRunnerImpl::Run
0x00007f0cf78482b2	(chrome -browser_main.cc:46 )	content::BrowserMain
0x00007f0cf54a3660	(chrome -content_main_runner.cc:785 )	content::ContentMainRunnerImpl::Run
0x00007f0cf54a213a	(chrome -content_main.cc:20 )	content::ContentMain
0x00007f0cf5130c7e	(chrome -chrome_main.cc:84 )	ChromeMain
0x00007f0cf2576fb5	(libc-2.19.so -libc-start.c:292 )	__libc_start_main
0x00007f0cf5130ade	(chrome + 0x01164ade )	_start
Cc: mfomitchev@chromium.org
Owner: skuhne@chromium.org
I doubt this is me, I moved the code around, but haven't changed much other than files/class names: https://codereview.chromium.org/2052013002

That, said, looking at RestoreUnpinnedRunningApplicationOrder: what if some apps get closed as a result of the user switch? If this happens, there will be less items corresponding to running apps in the shelf than there were when last_used_running_application_order_ was saved. This means that running_index could get out of bounds, right? Stefan, WDYT? 

Note also that some crash items quoted in comment 9 have an OnNoMemory error, which probably needs a separate bug.
Mergedinto: 647855
Status: Duplicate (was: Assigned)

Sign in to add a comment