Bug 16799 - Latest Flash update packages kill firefox and a whole other kde ( mageia control center ) applications
Summary: Latest Flash update packages kill firefox and a whole other kde ( mageia cont...
Status: RESOLVED WORKSFORME
Alias: None
Product: Mageia
Classification: Unclassified
Component: RPM Packages (show other bugs)
Version: 5
Hardware: x86_64 Linux
Priority: High critical
Target Milestone: ---
Assignee: Mageia Bug Squad
QA Contact:
URL:
Whiteboard:
Keywords: NEEDINFO
Depends on:
Blocks:
 
Reported: 2015-09-22 01:39 CEST by Sam Khangyi
Modified: 2017-06-10 22:28 CEST (History)
7 users (show)

See Also:
Source RPM: flash-player-plugin-11.2.202.521-1.mga5.nonfree.x86_64
CVE:
Status comment:


Attachments

Description Sam Khangyi 2015-09-22 01:39:54 CEST
Description of problem:


Version-Release number of selected component (if applicable):


How reproducible:


Steps to Reproduce:
1. install flash-player-plugin-11.2.202.521-1.mga5.nonfree.x86_64 and flash-player-plugin-kde-11.2.202.521-1.mga5.nonfree.x86_64 as recommended by the update.

2. Firefox will start and either crash immediately or when you press a few menu items
3. Firefox will crash even after firefox reset, even on basic mode.

Mageia control center shows splash then crashes.

Had to uninstall package to be able to use system again.

urpme flash-player-plugin-11.2.202.521-1.mga5.nonfree.x86_64
To satisfy dependencies, the following 2 packages will be removed (40KB):
  flash-player-plugin-11.2.202.521-1.mga5.nonfree.x86_64
  flash-player-plugin-kde-11.2.202.521-1.mga5.nonfree.x86_64
   (due to unsatisfied flash-player-plugin == 11.2.202.521-1.mga5.nonfree,
    due to unsatisfied flash-player-plugin == 11.2.202.521-1.mga5.nonfree)
Remove 2 packages? (y/N) y
removing flash-player-plugin-11.2.202.521-1.mga5.nonfree.x86_64 flash-player-plugin-kde-11.2.202.521-1.mga5.nonfree.x86_64
removing package flash-player-plugin-kde-11.2.202.521-1.mga5.nonfree.x86_64
      1/2: removing flash-player-plugin-kde-11.2.202.521-1.mga5.nonfree.x86_64
                                 ########################################################################################################################################
removing package flash-player-plugin-11.2.202.521-1.mga5.nonfree.x86_64
      2/2: removing flash-player-plugin-11.2.202.521-1.mga5.nonfree.x86_64
                                 ########################################################################################################################################
[root@bogi bogi]# 


Now i can use firefox and mageia control centere again. 


Reproducible: 

Steps to Reproduce:
Comment 1 Samuel Verschelde 2015-09-22 09:17:38 CEST
Does anyone confirm this bug?

Priority: Normal => High
CC: (none) => anssi.hannula, luigiwalser, qa-bugs, thierry.vignaud

Comment 2 claire robinson 2015-09-22 09:55:25 CEST
It's ok here. Not sure how flash and MCC are connected.

Could you give some system details please and logs if available too.
Comment 3 claire robinson 2015-09-22 09:58:53 CEST
Also please check if the same symptoms exist if you create a new user on the system. The flash update was a critical security update which prevents many remote code execution issues.
Comment 4 Sam Khangyi 2015-09-22 12:09:50 CEST
Sep 21 16:54:01 bogi kernel: Web Content[31512]: segfault at 0 ip 00007f249bbea0bd sp 00007ffd556c76f0 error 6 in libmozalloc.so[7f249bbe9000+2000]
Comment 5 Sam Khangyi 2015-09-22 12:24:43 CEST
i don't think any log entries are there, firefox crashed at least 6 -7 times without a trace. and it did stop crashing after I uninstalled the 2 update packages. 

Sep 21 16:46:30 bogi kernel: sd_festival[27784]: segfault at 2d0 ip 00007f80e444e860 sp 00007fff83c711c8 error 4 in libpthread-2.20.so[7f80e4441000+17000]
Sep 21 16:54:01 bogi kernel: Web Content[31512]: segfault at 0 ip 00007f249bbea0bd sp 00007ffd556c76f0 error 6 in libmozalloc.so[7f249bbe9000+2000]
Sep 21 17:01:40 bogi kernel: sd_festival[27772]: segfault at 2d0 ip 00007f684e04a860 sp 00007ffea49db8c8 error 4 in libpthread-2.20.so[7f684e03d000+17000]
Sep 21 17:10:06 bogi kernel: sd_festival[24613]: segfault at 2d0 ip 00007f863144b860 sp 00007ffd1712b818 error 4 in libpthread-2.20.so[7f863143e000+17000]
Sep 21 17:17:23 bogi kernel: sd_festival[2443]: segfault at 2d0 ip 00007f6029aba860 sp 00007ffce48b6698 error 4 in libpthread-2.20.so[7f6029aad000+17000]


all the segfaults i could find ...
Comment 6 Samuel Verschelde 2015-09-22 12:30:43 CEST
I think this might be a bug that appears only in some locales, with some input methods or encodings.
Comment 7 Samuel Verschelde 2015-09-22 12:31:34 CEST
Does MCC started from console give useful logs?
Comment 8 Samuel Verschelde 2015-09-22 12:32:59 CEST
I would also advise to check your hardware memory with memcheck or similar tools, and also your disk. This could be hardware related but would have been triggered only with the new packages. Does it still segfault if you install the update again?
Comment 9 Sam Khangyi 2015-09-22 12:47:47 CEST
i can try installing them and see if it happens again ... see you on the other side :-)
Comment 10 David Walser 2015-09-22 14:30:45 CEST
If it's a real bug, it could also be that Adobe changed how they compile it again.  They already changed it a few years ago to use CPU extensions not available on every CPU, so it stopped working on a lot of computers.  If the problem persists, I'd recommend uninstalling this and installing Google Chrome and freshplayerplugin and see if Flash in Firefox behaves any better.

CC: luigiwalser => (none)

Comment 11 Rémi Verschelde 2015-09-22 14:36:59 CEST
(In reply to David Walser from comment #10)
> If it's a real bug, it could also be that Adobe changed how they compile it
> again.  They already changed it a few years ago to use CPU extensions not
> available on every CPU, so it stopped working on a lot of computers.

I'm puzzled that this would make the MCC crash, it should have nothing to do with flash player.
Comment 12 Sam Khangyi 2015-09-22 16:43:43 CEST
Ok guys,
Back from the other side.
installed the 2 packages, and everything is working fine. 

Flash is running, and mcc is running fine.

I have no idea what went wrong when the 2 packages came in as updates.

erasing them and re installing them seems to have fixed the issue.

Tho i would care to note, firefox was working just fine without any flash installed on the system, including youtube and many news sites ....
Comment 13 Rémi Verschelde 2015-09-22 16:50:23 CEST
(In reply to Sam Khangyi from comment #12)
> Tho i would care to note, firefox was working just fine without any flash
> installed on the system, including youtube and many news sites ....

<offtopic> Indeed, one can very well live without Flash nowadays, and I can only say good riddance. Of course there are some specific applications that you might need Flash for, but I personally deinstalled it some months ago and I don't miss it. http://occupyflash.org/ </offtopic>
Comment 14 Marja Van Waes 2015-09-22 17:22:54 CEST
setting to unconfirmed for now, only leaving it open in case someone else hit(s) this

Status: NEW => UNCONFIRMED
CC: (none) => marja11
Ever confirmed: 1 => 0

Comment 15 Sam Khangyi 2015-09-22 18:44:36 CEST
(In reply to Marja van Waes from comment #14)
> setting to unconfirmed for now, only leaving it open in case someone else
> hit(s) this

agreed.
Comment 16 Bit Twister 2015-09-22 20:16:04 CEST
So, what site is used to verify the rpm/package works?

Is the bouncing ball enough at https://www.adobe.com/software/flash/about  :)
Or animation/mouse at http://www.chemgapedia.de/vsengine/help/en/flash/     :)
or animation  at http://www.itma.vt.edu/tech/shockwave.htm   :)
or movie at http://www.itma.vt.edu/tech/shockwave.htm   :(

Currently running Vendor install firefox on mga5 64 bit

CC: (none) => bittwister2

Comment 17 Ezequiel Partida 2015-12-03 06:36:19 CET
I had a different problem today so I don't know if I should open a new report or not... I didn't had that problem before.

I used freshplayerplugin-0.3.1-1.mga5.x86_64

This is what a I get.

[root@localhost ~]# mcc
No bp log location saved, using default.
[000:000] Cpu: 6.42.7, x4, 3200Mhz, 3833MB
[000:000] Computer model: Not available
[000:000] Browser XEmbed support present: 1
[000:000] Browser toolkit is Gtk2.
[000:000] Using Gtk2 toolkit
NOT SANDBOXED
[ERROR:flash/platform/pepper/common/pep_process.cpp(19)] CHECK(g_process_type == PROCESS_TYPE_UNKNOWN) failed


MCC closes immediately

Regards

CC: (none) => ezequiel_partida

Comment 18 claire robinson 2015-12-03 09:35:14 CET
Yes, new bug report please. Pepperflash is a completely different package.
Angelo Naselli 2015-12-15 13:48:51 CET

CC: (none) => anaselli

Comment 19 Sandro kensan 2016-03-27 01:38:25 CET
I have make a new installation of mageia 5 in my HD (previously I had installed Mageia 4 in my SSD) and now I have MCC that randomly don't open it. If I reboot it works and then stop to open it. Firefox had some problem and one times a day, randomly crash.

So my problem is similar to Sam's problem.

urpmi works. In the future i make a new fresh installation in my SSD of mageia 5.

This bug is not related to flash player plugin.

CC: (none) => kensan

claire robinson 2016-03-27 18:55:17 CEST

CC: qa-bugs => (none)

Comment 20 Marja Van Waes 2017-03-12 09:57:22 CET
@ Sandro

Did you have similar segfaults to Sam's when you reported having a similar problem?

If so, do you still have this issue now, in Mageia 5 and/or 6sta2?

@ Sam

Do you still see this in fully updated Mga5 or 6sta2?

@ Sam & Sandro

If so, please run, after seeing the bug again, as root:

   journalctl -ab > journal.txt

And attach journal.txt to this bug report, while telling at which time(s) the bug reoccurred.

Thanks :-)

Keywords: (none) => NEEDINFO

Comment 21 Marja Van Waes 2017-03-12 10:01:53 CET
Btw, with "the bug" or "the problem" I do _not_ mean seeing the MCC or Firefox issue after getting a flash update, but having the MCC or Firefox issue regardless of what was updated, and seeing segfaults in libmozalloc.so and/or sd_festival segfaults.
Comment 22 Sandro kensan 2017-03-12 10:53:48 CET
No more problem with MCC after the re-installation. Other glitch with Firefox but this is another bug. For me this bug is closed. TY.
Comment 23 Sam Khangyi 2017-06-10 01:00:23 CEST
I agree with Sandoro here. This bug should be closed ...
Comment 24 Rémi Verschelde 2017-06-10 22:28:13 CEST
Thanks, closing.

Status: UNCONFIRMED => RESOLVED
Resolution: (none) => WORKSFORME


Note You need to log in before you can comment on or make changes to this bug.