657100 - crash everytime accessing some url [@ libxul.so@0xeb6335 ]

Status: RESOLVED WORKSFORME

(Firefox :: General, defect)

Description

[Domfe]

User-Agent:       Mozilla/5.0 (X11; Linux i686; rv:2.0.1) Gecko/20100101 Firefox/4.0.1
Build Identifier: Mozilla/5.0 (X11; Linux i686; rv:2.0.1) Gecko/20100101 Firefox/4.0.1

Firefox crash everytime I access some pages. E.g. www.fineco.it
Also some other url that I didn't remember.
I've runned gdb and founded the crash happen with libfreebl3.so.
I'm using Ubuntu Natty with all updates.
Now I've removed the libfreebl provided with firefox and rewrited
it with /usr/lib/nss/libfreebl3.so. Now I can browse without crashes!
Hope this helps!

Reproducible: Always

Steps to Reproduce:
1. launch firefox
2. go to www.fineco.it

Actual Results:  
Firefox crashes everytime.

Expected Results:  
The site load correctly

[New Thread 0xa64ffb70 (LWP 1750)]
[New Thread 0xa59ffb70 (LWP 1751)]

Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread 0xa88fbb70 (LWP 1744)]
0xabb3b17f in ?? () from /usr/lib/firefox-4.0.1/libfreebl3.so

Comment 1

[Thomas Ahlblom]

Please post the related crash IDs from about:crashes.
https://developer.mozilla.org/en/How_to_get_a_stacktrace_for_a_bug_report

Are you running a Firefox that has been downloaded directly from Mozilla or has it been installed from some Ubuntu repository?

Does the issue still occur if you start Firefox in Safe Mode?
https://support.mozilla.com/en-US/kb/Safe+Mode

How about with a new, empty profile?
https://support.mozilla.com/en-US/kb/Basic+Troubleshooting#w_8-make-a-new-profile

Comment 2

[Domfe]

Crash ID bp-8f9d66a4-da2b-4d42-92da-589b02110515

I'm running the firefox shipped with Ubuntu

Comment 3

[Domfe]

Tested with -safe-mode: crash.
After restart it crashed with my gmail page.
Here are the last crash ids:
bp-8b3e8872-f569-491c-9ddd-10dec2110515
69e3cb04-591d-ee92-4b8e61df-01c52e49

Comment 4

[Tyler Downer [He/Him]]

If you download the officialo Mozilla.com build, does it still crash?

Comment 5

[Domfe]

A new profile doesn't resolve the issue.
The official build is currently working!

Comment 6

[Domfe]

Sorry, the official build crashed: I was watching mail on gmail.
The terminal reported:
"Rilevato trace/breakpoint"
It's the italian version.

Comment 7

[Domfe]

Other crashes happening.
E.g.
5f109078-df5e-076d-63ec9ce9-152738ad

Comment 8

[Chris Coulson]

I'm a bit concerned that these crash reports from our builds in Ubuntu aren't providing any useful information. For https://crash-stats.mozilla.com/report/index/8b3e8872-f569-491c-9ddd-10dec2110515, I've checked that the symbols for the libxul.so with that debug ID (7F9ED70DAEFBD131466E4E97CD6B06D60) are uploaded, but it still hasn't worked

Comment 9

[Chris Coulson]

Ted - any idea what might be going wrong there? :/

Comment 10

[(not currently active) Ted Mielczarek]

Since the Socorro installation moved to our datacenter in Phoenix, symbols are copied from the netapp in San Jose to Phoenix via a set of scripts doing rsyncs. I think they're keyed off of the -symbols.txt index files, but I don't know a whole lot more about them.

Jabba: can you check that:
a) symbols_ubuntu is being rsynced to PHX
b) the symbol syncing scripts don't have a problem with the index.txt files in symbols_ubuntu?

In this particular crash report, the index is firefox-4.0.1-Linux-20110422203143-110422201246-i686-symbols.txt and the symbol file in question is libxul.so/7F9ED70DAEFBD131466E4E97CD6B06D60/libxul.so.sym

Comment 11

[Justin Dow [:jabba]]

I don't see any problems with the syncing. The output of `df` on dm-symbolpush01 and on one of the boxes in phx with the phx symbol store mounted agree to the very byte that the same amount of space is being used. I've verified that the contents of symbols_ubuntu/ in sjc and phx are identical, i.e. what you see on dm-symbolpush01 is what is available in phx.

Comment 12

[(not currently active) Ted Mielczarek]

Okay, thanks! Can you check the processor config and ensure that symbols_ubuntu is being passed to minidump_stackwalk in production?

Comment 13

[Chris Coulson]

Just to make everything more confusing, this report is from one of our builds, and it appears to have worked: https://crash-stats.mozilla.com/report/index/bp-22e23234-387d-43d0-ad3e-a11c12110613

Comment 14

[Justin Dow [:jabba]]

export processorSymbolsPathnameList="/mnt/socorro/symbols/symbols_ffx,/mnt/socorro/symbols/symbols_sea,/mnt/socorro/symbols/symbols_tbrd,/mnt/socorro/symbols/symbols_mob,/mnt/socorro/symbols/symbols_penelope,/mnt/socorro/symbols/symbols_sbrd,/mnt/socorro/symbols/symbols_camino,/mnt/socorro/symbols/symbols_os,/mnt/socorro/symbols/symbols_solaris,/mnt/socorro/symbols/symbols_opensuse,/mnt/socorro/symbols/symbols_ubuntu,/mnt/socorro/symbols/symbols_fedora"


That is from the config override in use. I'll let rhelmer or lars comment as to whether that looks syntactically correct, but I imagine if it were wrong, we'd have more problems than this?

Comment 15

[(not currently active) Ted Mielczarek]

That looks correct. Chris: do you have a delay between when your builds go live and your symbols get uploaded, or anything like that?

Comment 16

[Chris Coulson]

There can be a delay of up to an hour between the builds being initially published on our main server and the symbols being pushed (and the symbols are probably pushed before the builds get mirrored to local servers). If that turns out to be an issue, I can try and figure out another way to automate it (I literally just have a cron job which scans for newly published packages every hour, as that's the only way i can do it without actually hooking it in to Launchpad)

Comment 17

[Domfe]

Also with ff 5.0.
bp-8736788e-b138-4829-8234-642a02110702
bp-2e5705c1-299a-45ed-9a94-df91c2110702

Comment 18

[Domfe]

Crashes many times a day. What can I do? Can I help with something?
It seems retaled with Javascript.

Comment 19

[Domfe]

I'm having always problems (now FF 6): bp-892c3bda-b4d1-4f87-8315-6fb1d2110905.
It happened when GMail has received a new email. It's very annoying, what other can I do to help you?

Comment 20

[Domfe]

It's about five days that the error doesn't show up. My current version is now 7.0.1.
I'll wait some days more and if nothing happens I'll close the bug.

Thank you all.

Comment 21

[Leonard Camacho [:lcamacho]]

I can't reproduce this with nightly or firefox release

Mozilla/5.0 (X11; Ubuntu; Linux x86_64; rv:13.0) Gecko/20100101 Firefox/13.0
Mozilla/5.0 (X11; Linux x86_64; rv:16.0) Gecko/16.0 Firefox/16.0a1

Comment 22

[Wayne Mery (:wsmwk)]

WFM per comment 20