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

Issue 684917 link

Starred by 1 user
Status: Archived
Owner: ----
Closed: Apr 2018
Cc:
derat@chromium.org
█ igo@chromium.org
█ ejcaruso@chromium.org
tbroch@chromium.org
Components:
EstimatedDays: ----
NextAction: ----
OS: Chrome
Pri: 3
Type: Bug



Sign in to add a comment

segfaults in __epoll_wait_nocancel

Project Member Reported by kirtika@google.com, Jan 25 2017 
OS: ChromeOS M57

Seeing a recently introduced crash in powerd. 
No repro steps since I found this in a feedback report and then
on crash.corp.


crash/ link:

https://crash.corp.google.com/browse?q=product.name%3D%27ChromeOS%27%20%20AND%20stable_signature%3D%27epoll_dispatch-6cacb3eb%27&ignore_case=false&enable_rewrite=false&omit_field_name=&omit_field_value=&omit_field_opt=#samplereports:5,cpuarchitecture,hardwareclass:20


Stack trace:

	0x00007fabb451dc93	(libc-2.19.so + 0x000f9c93 )	__epoll_wait_nocancel
0x00007fabb40f8ea3	(libevent_core-2.0.so.5.1.9 -epoll.c:407 )	epoll_dispatch
0x00007fabb40e9911	(libevent_core-2.0.so.5.1.9 -event.c:1625 )	event_base_loop
0x00007fabb53a4d64	(libbase-core-395517.so -message_pump_libevent.cc:256 )	base::MessagePumpLibevent::Run(base::MessagePump::Delegate*)
0x00007fabb53c6684	(libbase-core-395517.so -run_loop.cc:35 )	base::RunLoop::Run()
0x00007fabb53a1119	(libbase-core-395517.so -message_loop.cc:294 )	base::MessageLoop::Run()
0x00007fabb54d29ea	(powerd -main.cc:322 )	main
0x00007fabb4443fb5	(libc-2.19.so -libc-start.c:292 )	__libc_start_main
0x00007fabb54d1fad	(powerd + 0x00013fad )	_start
0x00007fffdecb0757

Comment 1 by derat@chromium.org, Jan 25 2017

Cc: ejcaruso@chromium.org
Summary: powerd segfault in __epoll_wait_nocancel (was: powerd crash since 8936.0.0)
I don't think this is new; I've seen sporadic crash reports about epoll 0x0 segfaults for years. I assume it's a libevent bug. It's within Chrome's message loop code and probably unrelated to anything that powerd is doing, in any case.

Comment 2 by ejcaruso@chromium.org, Jan 25 2017 

I've seen this in other daemons as well. It's not a powerd issue and I'm not sure what we can do about it other than upreving libevent and crossing our fingers.

Comment 3 by derat@chromium.org, Mar 16 2017

Cc: derat@chromium.org
Components: OS>Systems
Labels: -Pri-2 -M-57 Pri-3
Owner: ----
Status: Available (was: Untriaged)
Summary: segfaults in __epoll_wait_nocancel (was: powerd segfault in __epoll_wait_nocancel)
I'll leave this open so we can dupe other bugs and crash reports against it.

Comment 4 by benchan@chromium.org, Mar 16 2017 

fyi, I updated libevent to 2.1.8 (chromium:692757), so we may no longer see the same crash signature (either the underlying bug goes away or the crash signature changes)
Project Member

Comment 5 by sheriffbot@chromium.org, Apr 6 2018

Labels: Hotlist-Recharge-Cold
Status: Untriaged (was: Available)
This issue has been Available for over a year. If it's no longer important or seems unlikely to be fixed, please consider closing it out. If it is important, please re-triage the issue.

Sorry for the inconvenience if the bug really should have been left as Available.

For more details visit https://www.chromium.org/issue-tracking/autotriage - Your friendly Sheriffbot

Comment 6 by ejcaruso@chromium.org, Apr 6 2018

Status: Archived (was: Untriaged)
►

Sign in to add a comment