903993 - Firefox 25 spike in Android crash @ libflashplayer.so@0x75...

Status: RESOLVED FIXED

(Firefox for Android Graveyard :: Plugins, defect)

Description

[Scoobidiver (away)]

With combined signatures, it's #9 crasher in 23.0, #7 in 24.0b1, #1 in 25.0a2, and #1 in 26.0a1.

It started spiking in 25.0a1/20130717. The regression range for the spike is:
http://hg.mozilla.org/mozilla-central/pushloghtml?fromchange=5976b9c673f8&tochange=0888e29c83a3

Here is the breakdown per device:
+ Aurora:
*libflashplayer.so@0x75be5e 	10
Samsung GT-N8000 	10

+ Nightly:
* libflashplayer.so@0x759e5e 	8
ASUS Transformer Pad TF300T 	7
unknown cm_tenderloin 	1
* libflashplayer.so@0x75c0a6 	8
Samsung GT-I9300 	3
Samsung GT-P5110 	3
Samsung Galaxy Nexus 	2
* libflashplayer.so@0x75be5e 	7
Amazon KFTT 	4
Samsung Nexus 10 	2
Samsung GT-N7000 	1

Frame 	Module 	Signature 	Source
0 	libc.so 	libc.so@0x18a9c 	
1 	libflashplayer.so 	libflashplayer.so@0x75be5e 	
2 	libflashplayer.so 	libflashplayer.so@0x40524d 	
3 	libflashplayer.so 	libflashplayer.so@0x6dbfd1 	
4 	libflashplayer.so 	libflashplayer.so@0x75be5e 	
5 	libflashplayer.so 	libflashplayer.so@0x145c55 	
6 	libflashplayer.so 	libflashplayer.so@0x75be5e
...

More reports at:
https://crash-stats.mozilla.com/query/?product=FennecAndroid&query_search=signature&query_type=contains&query=libflashplayer.so%400x75

Comment 1

[Mathieu Marquer]

Reproducible when clicking to activate the flash plug-in on http://losangeles.cbslocal.com/2013/08/15/sea-snail-egg-grew-and-hatched-inside-aliso-veijo-boys-knee/

Comment 2

[Kevin Brosnan [Ex-Mozilla]]

QA wanted to investigate comment 1.

Comment 3

[Aaron Train [:aaronmt]]

Crashed on first-attempt at loading the URL (Samsung Galaxy SIV, Android 4.3) Flash (11.1.115.69).

d57dc6ac-f6c1-4a9f-b40c-8b7b42130819

https://crash-stats.mozilla.com/report/index/d57dc6ac-f6c1-4a9f-b40c-8b7b42130819

Comment 4

[Aaron Train [:aaronmt]]

Attachment: Nightly (Galaxy SIV Android 4.3) Log

Comment 5

[Kevin Brosnan [Ex-Mozilla]]

Looks like the inbound builds are gone. Will manually bisect this soon

Comment 6

[Kevin Brosnan [Ex-Mozilla]]

user:        George Wright <[e-mail]>
date:        Thu Jul 11 16:41:15 2013 -0400
summary:     Bug 890272 - Use SkFontHost_linux to provide SkFontHost factories on Android for use by skia-npapi r=mattwoodrow

Comment 7

[Kevin Brosnan [Ex-Mozilla]]

Leaving a need info to George so there will be a flag outside the vacation bugmail.

Comment 8

[George Wright (:gw280) (needinfo me!)]

Attachment: 0001-Bug-903993-null-check-the-SkStream-s-getFileName-bef.patch

This fixes this crash for now and the flash works fine, but we should investigate further as to why we aren't getting an SkFILEStream. This will be affected by a rebase of Skia as SkStream::getFileName() no longer exists upstream.

Comment 9

[George Wright (:gw280) (needinfo me!)]

https://hg.mozilla.org/integration/mozilla-inbound/rev/c04bbc49869f

Comment 10

[Mark Finkle (:mfinkle) (use needinfo?)]

Ask for uplift approval when ready

Comment 11

[Ryan VanderMeulen [:RyanVM]]

https://hg.mozilla.org/mozilla-central/rev/c04bbc49869f

Comment 12

[Kevin Brosnan [Ex-Mozilla]]

Poking this for Firefox 25 Beta uplift. This is already fixed in Firefox 26.

Comment 13

[Alex Keybl [:akeybl]]

gw280/snorp - uplift request for Beta please

Comment 14

[James Willcox (:snorp) ([e-mail]) (he/him)]

Comment on attachment 797307 [details] [diff] [review]
0001-Bug-903993-null-check-the-SkStream-s-getFileName-bef.patch

[Approval Request Comment]
Bug caused by (feature/regressing bug #): bug 848491
User impact if declined: Crashes.
Testing completed (on m-c, etc.): m-c and aurora for a few weeks, no issues
Risk to taking this patch (and alternatives if risky): low
String or IDL/UUID changes made by this patch: none

Comment 15

[bhavana bajaj [:bajaj]]

Comment on attachment 797307 [details] [diff] [review]
0001-Bug-903993-null-check-the-SkStream-s-getFileName-bef.patch

Approving as this is low risk and help avoid a reproducible crash.

Request to land asap so this can get into beta which is going to build today.

Comment 16

[James Willcox (:snorp) ([e-mail]) (he/him)]

https://hg.mozilla.org/releases/mozilla-beta/rev/8015563e32fe

Comment 17

[Kevin Brosnan [Ex-Mozilla]]

This is showing up on 29 beta again did this code change during the release?

Comment 18

[James Willcox (:snorp) ([e-mail]) (he/him)]

(In reply to Kevin Brosnan [:kbrosnan] from comment #17)
> This is showing up on 29 beta again did this code change during the release?

Nope.

Comment 19

[James Willcox (:snorp) ([e-mail]) (he/him)]

(In reply to James Willcox (:snorp) ([e-mail]) from comment #18)
> (In reply to Kevin Brosnan [:kbrosnan] from comment #17)
> > This is showing up on 29 beta again did this code change during the release?
> 
> Nope.

Wait, maybe. George?

Comment 20

[James Willcox (:snorp) ([e-mail]) (he/him)]

This patch is commented out on trunk, presumably because latest Skia doesn't even support the required API. I would say the new crash must be different than this one?

Comment 21

[James Willcox (:snorp) ([e-mail]) (he/him)]

Attachment: Don't use __fork, it's gone in newer bionic

Comment 22

[James Willcox (:snorp) ([e-mail]) (he/him)]

Comment on attachment 8412859 [details] [diff] [review]
Don't use __fork, it's gone in newer bionic

Wrong bug.

Comment 23

[Kevin Brosnan [Ex-Mozilla]]

Filed 1001624.

Comment 24

[George Wright (:gw280) (needinfo me!)]

Yeah, the code is commented out. Is there a stacktrace anywhere?