881027 - 100% cpu when on new maps.google.com (may be unable to kill, requiring reboot)

Status: RESOLVED WORKSFORME

(Firefox :: Untriaged, defect)

Description

[u123541]

User Agent: Mozilla/5.0 (X11; Linux i686; rv:23.0) Gecko/20130506 Firefox/23.0 (Nightly/Aurora)
Build ID: 20130506030925

Steps to reproduce:

was starting to use new version of maps.google.com. This was slow, so I tried to file a report via google dropdown at top-right of page.  


Actual results:

KDE, mouse, keyboard all locked up.  Accessing the system from another laptop, Linux is still running fine. 

top reports:  
 PID USER      PR  NI  VIRT  RES  SHR S %CPU %MEM    TIME+  COMMAND
 8250 pfortin   30  10 1926m 610m  57m R  100  7.6 200:52.13 firefox
 4058 root      20   0 1446m 1.0g 416m D    0 12.7   1585:19 X

So it appears X is hung on I/O.  Attempting to attach gdb gives startup messages, including:
This GDB was configured as "x86_64-mageia-linux-gnu".
For bug reporting instructions, please see:
<http://www.gnu.org/software/gdb/bugs/>...
Reading symbols from /home/usrlocal/bin/firefox24.0a1/firefox...(no debugging symbols found)...done.
Attaching to program: /home/usrlocal/bin/firefox24.0a1/firefox, process 8250
[but no prompt]

It's possible to remotely kill gdb (not via Ctrl+c); but never FF or X.

Tried killing FF via every possible signal, to no avail.  After over 3 hours trying different options, giving up and will reboot...  :(


Expected results:

No super-tight loop that can't be kill'd, gdb'd, strace'd...

Comment 1

[u123541]

Nightly just went into 100% CPU again while on the new maps.google.com...

Not so tight this time, so got a gdb trace; but looks like a corrupt stack...

Comment 2

[u123541]

Attachment: gdb output

Comment 3

[Paul Silaghi, QA [:pauly]]

(In reply to Pierre Fortin from comment #0)
> was starting to use new version of maps.google.com
Do you mean the new maps that are currently in preview? https://www.google.com/maps/about/explore/

Comment 4

[u123541]

Yes. Been avoiding it since then. (It's also quite slow and not fun to use as a result.)

Comment 5

[Paul Silaghi, QA [:pauly]]

All I can confirm is slowness in FF from the new google maps.

Comment 6

[Ioana (away)]

Can you please try to reproduce this issue with a clean profile?
http://support.mozilla.org/en-US/kb/profile-manager-create-and-remove-firefox-profiles

Do you still get the 100% cpu that way or just the slowness Paul is talking about?

Comment 7

[u123541]

Lately, it's just been the slowness...  so much that Google's new effort may be a waste of time if their implementation doesn't make better use of server/network/client resources...

Comment 8

[Adrian Florinescu [:aflorinescu]]

I can see no apparent issue on a leisure test on Version 46.0a1 Build ID 20160111030207 | Mozilla/5.0 (X11; Linux i686; rv:46.0) Gecko/20100101 Firefox/46.0.
@reporter Please reopen if this is still reproducible on a specific section of gmaps.

Comment 9

[marccollin7379]

same issue with firefox 45

Comment 10

[Adrian Florinescu [:aflorinescu]]

Hey Marc. Can you add some steps in order to help us reproduce the problem? Maybe you're using the gmaps in some particular way. Also please post the operating system as well? 

I don't see any issue on the 45 release:
45.0.1 	20160315153207 User Agent 	Mozilla/5.0 (X11; Linux i686; rv:45.0) Gecko/20100101 Firefox/45.0 - Ubuntu x32 14.04

Comment 11

[marccollin7379]

i start firefox, go to google map and type an address.

i use firefox 25 with opensuse, kernel 4.5.0.


When i start to move the map with the mouse, firefox use all the cpu.

I have a lenovo s430, i5, 16 of ram and 2 ssd, intel 4000 video card.

Comment 12

[Adrian Florinescu [:aflorinescu]]

Marc, I think you should consider updating Firefox to a more recent version. That would probably fix your problem.
I don't know for sure what FF version you distro has as canonical version, but it should be possible to get the latest available FF.

https://software.opensuse.org/ymp/openSUSE:Leap:42.1:Update/standard/MozillaFirefox.ymp?base=openSUSE%3ALeap%3A42.1&query=MozillaFirefox

Please let me know if upgrading fixes the issue.