930827 - Hang when visiting icu-project.org

Status: RESOLVED DUPLICATE of bug 927901

(Core :: Security, defect)

Description

[(no longer active)]

Attachment: Profile

STR:
1. Go to http://site.icu-project.org
2. Click on the User Guide link on the left side menu.
3. Firefox hangs.

There's a js engine thread consuming all of the CPU and the main thread gets stuck.

Comment 1

[Alice0775 White]

Regression window(m-c)
Good:
http://hg.mozilla.org/mozilla-central/rev/54434a926c5f
Mozilla/5.0 (Windows NT 6.1; WOW64; rv:25.0) Gecko/20100101 Firefox/25.0 ID:20130805022018
Bad:
http://hg.mozilla.org/mozilla-central/rev/ad0ae007aa9e
Mozilla/5.0 (Windows NT 6.1; WOW64; rv:25.0) Gecko/20100101 Firefox/25.0 ID:20130805022418
Pusgloh
http://hg.mozilla.org/mozilla-central/pushloghtml?fromchange=54434a926c5f&tochange=ad0ae007aa9e

In local build:
Last good: 81fada98a394
First bad: 8aafb671f1e9
Regressed by: 
	8aafb671f1e9	Yoshi Huang — Bug 883741 - Part 1: WebCrypto: Move Crypto to WebIDL. r=bz

Comment 2

[Kyle Huey (Exited; not receiving bugmail, old account, do not use)]

Alice is this fixed with the patch from Bug 927901?

Comment 3

[Olli Pettay [:smaug][[e-mail]]]

ehsan, do you really get hang and not unresponsive alert thing, since that is what I get
without the patch for bug 927901?
With the patch the page loads fine.

Comment 4

[Olli Pettay [:smaug][[e-mail]]]

Actually, I'm not yet sure bug 927901 helped here, since I don't get the hang always even without
the patch.

Comment 5

[Yoshi Cheng-Hao Huang [:allstars.chh][:allstarschh][:yoshi]]

Before applying Bug 927901, I do get the hang,
after applying, the hang is gone, I can browse the user guide page smoothy.

Comment 6

[Alice0775 White]

Yes, I can confirm the hang is gone in Latest Nightly tinderbox build which is including the patch of Bug 927901.

http://hg.mozilla.org/mozilla-central/rev/186e834d87dc
Mozilla/5.0 (Windows NT 6.1; WOW64; rv:27.0) Gecko/20100101 Firefox/27.0 ID:20131025011312

Comment 7

[Olli Pettay [:smaug][[e-mail]]]

So, Google API thing calls window.crypto in the first page
) ["https://apis.google.com/_/scs/apps-static/_/js/k=oz.gapi.fi.SuvlWLYuwc8.O/m=card/rt=j/sv=1/d=1/ed=1/am=AQ/rs=AItRSTOLE2_G8FGie4Kh56Sq24HLeJinjg/cb=gapi.loaded_0":58]
    this = [object Window]
1 anonymous(    <Failed to get 'arguments' while inspecting stack frame>
) ["https://apis.google.com/js/api.js":4]
    this = [object Window]
2 anonymous(    <Failed to get 'arguments' while inspecting stack frame>
) ["https://apis.google.com/js/api.js":6]
    this = [object Window]
3 anonymous(    <Failed to get 'arguments' while inspecting stack frame>
) ["https://apis.google.com/js/api.js":4]
    this = [object Window]
4 anonymous(    <Failed to get 'arguments' while inspecting stack frame>
) ["https://apis.google.com/js/api.js":5]
    this = [object Window]
5 anonymous(    <Failed to get 'arguments' while inspecting stack frame>
) ["https://apis.google.com/js/api.js":6]
    this = [object Window]
6 anonymous(    <Failed to get 'arguments' while inspecting stack frame>
) ["https://apis.google.com/js/api.js":6]
    this = 
7 anonymous(    <Failed to get 'arguments' while inspecting stack frame>
) ["https://apis.google.com/js/api.js":6]
    this = [object Object]
8 <TOP LEVEL> ["https://apis.google.com/_/scs/apps-static/_/js/k=oz.gapi.fi.SuvlWLYuwc8.O/m=card/rt=j/sv=1/d=1/ed=1/am=AQ/rs=AItRSTOLE2_G8FGie4Kh56Sq24HLeJinjg/cb=gapi.loaded_0":1]
    this = [object Window]

Comment 8

[Olli Pettay [:smaug][[e-mail]]]

And the same thing is then used in the next page.

Comment 9

[Olli Pettay [:smaug][[e-mail]]]

The google script tries to use getRandomValues

Comment 10

[Olli Pettay [:smaug][[e-mail]]]

This looks indeed rather bad and we should perhaps get this to FF25

Comment 11

[Olli Pettay [:smaug][[e-mail]]]

I mean, to get the patch for Bug 927901 to FF25.

Comment 12

[Jan de Mooij [:jandem]]

Not a JS bug.

Comment 14

[Alex Keybl [:akeybl]]

We tracked and landed bug 927901 instead.

Comment 15

[Ioana (away)]

Verified as fixed on Windows 7, Ubuntu 13.04, Mac OS X 10.8.4:
Firefox 25b12 - 20131025150754
Firefox 25.0 RC b#3 - 20131025151332
Firefox 27.0a1 - 20131025100746

Still reproducible on:
Firefox 26.0a2 - 20131025004002

Comment 16

[u279076]

(In reply to Ioana Budnar, QA [:ioana] from comment #15)
> Still reproducible on:
> Firefox 26.0a2 - 20131025004002

I don't think the patch landed in this build, it's fixed in today's Aurora though (2013-10-26). Thanks to Tracy for confirming that for me. Marking this verified fixed across all branches as per Ioana and Tracy's testing here and in bug 927901.