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

Issue 614191 link

Starred by 1 user

Issue metadata

Status: WontFix
Owner: ----
Closed: Jun 2016
Cc:
Components:
EstimatedDays: ----
NextAction: ----
OS: Chrome
Pri: 3
Type: Bug



Sign in to add a comment

ChromeOS crashes once or twice a day, but crashes don't show up in chrome://crashes

Project Member Reported by rdb@chromium.org, May 23 2016

Issue description

Version 50.0.2661.104 (64-bit)
Platform 7978.76.0 (Official Build) stable-channel samus
Firmware Google_Samus.6300.174.0


Chrome OS is up to date.


How frequently does this problem reproduce? (Always, sometimes, hard to
reproduce?)

Once or twice a day.

What is the impact to the user, and is there a workaround? If so, what is
it?

All windows must be restored, in both logged-in users (dual-user login enabled).

If text is entered in an app and isn't saved, it is lost.

Please provide any additional information below. Attach a screen shot or
log if possible.


 
Components: Internals>CrashReporting

Comment 2 by rdb@chromium.org, May 27 2016

Happened again today. what can I do?

Comment 3 by rdb@chromium.org, Jun 2 2016

Happened again today. If I was on a slow Internet connection I would be really upset.

I recall a similar bug being filed and fixed in M50 - maybe this is related?

Comment 4 by rdb@chromium.org, Jun 4 2016

Once again today.
Cc: vapier@chromium.org
+vapier to help triage.

Comment 6 by vapier@chromium.org, Jun 14 2016

is there nothing in chrome://crashes even after clicking the upload link ?

if so, look at file:///var/log/messages and see if there's anything related in there.

Comment 7 by rdb@chromium.org, Jun 14 2016

> is there nothing in chrome://crashes even after clicking the upload link ?

Correct. I opened chrome://crashes and the most recent crash was "Automatically reported Monday, June 13, 2016 at 5:07:34 PM".  I clicked "Start uploading crashes" and nothing changed. My most recent crash was approximately 2 minutes ago (Tuesday June 14 2016 10:06 am pacific).

> if so, look at file:///var/log/messages and see if there's anything related in there.

I think this starts immediately before the crash/restart (attached)

chromeos crash 2016-06-14.txt
154 KB View Download

Comment 8 by vapier@chromium.org, Jun 14 2016

can you look at the log after the crash ?  that log doesn't show any crashes happening on the system at all.

Comment 9 by rdb@chromium.org, Jun 14 2016

Just to clarify my terminology: 

When I say crash, I mean that the computer randomly reboots (shows me the white chrome screen), and then goes to the login screen. After I enter my password it brings up my previously open windows and applications, though everything has to reload.

I have no idea if the kernel is panicking or what's going on behind the scenes.
Cc: sonnyrao@chromium.org bleung@chromium.org
if the system reboots, then that's a kernel or hardware issue.  if it were a panic or watchdog, there should be a kernel log waiting in the buffer on reboot.

the log shows the crash init starting up:
2016-06-14T10:04:57.490802-07:00 INFO crash_reporter[9944]: Enabling kernel crash handling

but it doesn't list any crashes being found or the watchdog going off.  let's add a few kernel guys, and they can perhaps bounce to someone to investigate further.
Cc: snanda@chromium.org
+Sameer

next time it happens, log in and open chrome://system (takes a long time to load)
and search for kernel-crashes

and copy/paste that into the bug if it's there

If that is empty, then  look for "eventlog" and copy paste that into this bug

Comment 12 by rdb@chromium.org, Jun 15 2016

Nothing in kernel-crashes, but something interesting in eventlog (see #181 which corresponds with the 'crash' this morning).

170 | 2016-06-14 00:56:13 | ACPI Enter | S3
171 | 2016-06-14 09:42:32 | ACPI Wake | S3
172 | 2016-06-14 09:42:32 | Wake Source | GPIO | 27
173 | 2016-06-14 09:42:32 | EC Event | Lid Open
174 | 2016-06-14 09:55:50 | ACPI Enter | S3
175 | 2016-06-14 10:02:41 | ACPI Wake | S3
176 | 2016-06-14 10:02:41 | Wake Source | GPIO | 27
177 | 2016-06-14 10:02:41 | EC Event | Lid Open
178 | 2016-06-14 10:04:42 | Kernel Event | Clean Shutdown
179 | 2016-06-14 10:04:42 | ACPI Enter | S5
180 | 2016-06-14 10:04:50 | System boot | 28
181 | 2016-06-14 10:04:50 | SUS Power Fail
182 | 2016-06-14 10:04:50 | System Reset
183 | 2016-06-14 10:04:50 | ACPI Wake | S5
184 | 2016-06-14 10:04:50 | Wake Source | Power Button | 0
185 | 2016-06-14 10:04:50 | Log area cleared | 1033
186 | 2016-06-14 10:04:50 | Wake Source | GPIO | 27
187 | 2016-06-14 10:04:50 | EC Event | Lid Open
188 | 2016-06-14 10:13:50 | ACPI Enter | S3
189 | 2016-06-14 10:14:46 | ACPI Wake | S3
190 | 2016-06-14 10:14:46 | Wake Source | GPIO | 27
191 | 2016-06-14 10:14:46 | EC Event | Lid Open
192 | 2016-06-14 10:52:24 | ACPI Enter | S3
193 | 2016-06-14 11:12:32 | ACPI Wake | S3
194 | 2016-06-14 11:12:32 | Wake Source | GPIO | 27
195 | 2016-06-14 11:12:32 | EC Event | Key Pressed
196 | 2016-06-14 12:25:09 | ACPI Enter | S3
197 | 2016-06-14 13:16:43 | ACPI Wake | S3
198 | 2016-06-14 13:16:43 | Wake Source | GPIO | 27
199 | 2016-06-14 13:16:43 | EC Event | Lid Open

Cc: sha...@chromium.org adurbin@chromium.org dlaurie@chromium.org
Hmm that actually looks like the kernel shut the system down cleanly and then the system decided to boot again -- I think "SUS power Fail" might be a red herring 

I added a few firmware folks who might know better what is happening there
We're logging a clean shutdown in the event log. The powerd logs (or other logs) may be helpful in discovering what is happening here. When this issue happens again, please boot and immediately file feedback (alt+shift+i). Please include the text "crbug 614191" in the feedback report so we can find it.

Comment 15 by rdb@chromium.org, Jun 15 2016

Will do, thanks.

Comment 16 by rdb@chromium.org, Jun 22 2016

Since the update to 

Version 51.0.2704.103 (64-bit)
Platform 8172.56.0 (Official Build) stable-channel samus
Firmware Google_Samus.6300.174.0


I haven't had this crash/power-off issue.  I'll keep this bug open for a few more days and close if I can't repro anymore.

Comment 17 by rdb@chromium.org, Jun 24 2016

Status: WontFix (was: Unconfirmed)

Sign in to add a comment