805692 - 64-bit Flash hang in F799232668 when in full screen

Status: RESOLVED FIXED

(Core Graveyard :: Plug-ins, defect, P4)

Description

[jfn4th]

User Agent: Mozilla/5.0 (Windows NT 6.1; Win64; x64; rv:19.0) Gecko/19.0 Firefox/19.0
Build ID: 20121025030620

Steps to reproduce:

I went to youtube, and opened a video in full screen.


Actual results:

The Nightly Plugin Container crashed. To get it working, I had to let it crash and refresh the page. Fullscreen then worked. Each video I go to has same process of initial crash when entering fullscreen, but then fullscreen works fine after refreshing the page post crash.


Expected results:

Video should have gone fullscreen without crash.

Comment 1

[jfn4th]

forgot to mention, I am running Nightly's 64 bit build.

Comment 2

[Scoobidiver (away)]

Can you provide the crash ID from about:crashes?
Does it happen in Safe Mode (see https://support.mozilla.org/kb/troubleshoot-firefox-issues-using-safe-mode)?
Does it happen with a new profile (see https://support.mozilla.org/kb/profile-manager-create-and-remove-firefox-profiles)?
Does it happen with 32-bit Nightly?

Comment 3

[jfn4th]

Where can I find about:crashes?

The problem still occurs in safe mode

The problem still occurs in new profiles

The problem does NOT occur in 32 bit Nightly

Comment 4

[Scoobidiver (away)]

(In reply to jfn4th from comment #3)
> Where can I find about:crashes?
Type about:crashes in the location bar.

I can reproduce in 64-bit builds: https://crash-analysis.mozilla.com/hang-reports/2012/10-26/hr-20121026-e073f611-8b75-4e2a-84e2-ddc1bca9a73d.html

Comment 5

[Benjamin Smedberg]

I'm pretty sure we don't hook windowprocs in 64-bit builds, so there are going to be plenty of things broken and I don't think we're going to fix them. I don't know why we bother to ship them.

Comment 7

[Alice0775 White]

Regression window(m-i)
Good:
http://hg.mozilla.org/integration/mozilla-inbound/rev/d400b4e402f3
Mozilla/5.0 (Windows NT 6.1; Win64; x64; rv:19.0) Gecko/19.0 Firefox/19.0 ID:20121024124702
hang:
http://hg.mozilla.org/integration/mozilla-inbound/rev/b6089a8b78d3
Mozilla/5.0 (Windows NT 6.1; Win64; x64; rv:19.0) Gecko/19.0 Firefox/19.0 ID:20121024130901
Pushlog:
http://hg.mozilla.org/integration/mozilla-inbound/pushloghtml?fromchange=d400b4e402f3&tochange=b6089a8b78d3

Regressed by: 
b6089a8b78d3	Neil Deakin — Bug 782547, allow killfocus and deactivate messages to arrive in any order, remove blur suppression code, r=jmathies

Comment 8

[Marek Raida]

I can confirm the same, hanging out...
Benjamin, are you telling us to use rather 32 than 64 bit versions??

Comment 9

[Benjamin Smedberg]

Yes. The 64-bit versions are experimental and are not supported.

Comment 10

[u52928]

(In reply to Benjamin Smedberg  [:bsmedberg] from comment #9)
> Yes. The 64-bit versions are experimental and are not supported.

@Mr. Smedberg
Is this comment also addressed to the rather long-standing group of beta-testers?

Or asked in another way: What is your vision for the 64-bit version?

Comment 11

[Scoobidiver (away)]

64-bit builds are only for Nightly-testers as there is no 64-bit Beta version.

Joe, you should read: https://groups.google.com/forum/?fromgroups=#!topic/mozilla.dev.planning/aeTXSZ_WFAs

Comment 12

[u52928]

(In reply to Scoobidiver from comment #11)
> 64-bit builds are only for Nightly-testers as there is no 64-bit Beta
> version.
> 
> Joe, you should read:
> https://groups.google.com/forum/?fromgroups=#!topic/mozilla.dev.planning/
> aeTXSZ_WFAs

Thank you Scoobdiver. I've read it, but I have to confess I didn't see it as being conclusive in any way, shape or form. 

I saw eight comments from various devs made in Nov 2011. It's now almost a year later, and I'd be quite surprised if anyone were to say that in the time that's passed since those posts were made, 64-bit computing has proven to be a dead end and is on its way out. 

Or have I missed something?

And by the way, I apologize for being off the track of what this bug is about. I certainly didn't start the general discussion here about 64-but computing and hope I'm still within the acceptable range of commenting on already-existing comments.

Comment 13

[u52928]

(In reply to Joe Greenman from comment #12)

> And by the way, I apologize for being off the track of what this bug is
> about. I certainly didn't start the general discussion here about 64-but
> computing and hope I'm still within the acceptable range of commenting on
> already-existing comments.

In the paragraph above, I meant, of course, "64-bit computing".

Comment 14

[Virtual_ManPL [:Virtual] 🇵🇱 - (please needinfo? me - so I will see your comment/reply/question/etc.)]

FYI...

Win 64bit builds ARE SUPPORTED. Of course as Tier2 builds.

>Breakage or regressions in these platforms does not immediately close the tree;

>Developers who break these platforms are expected
>to work with platform maintainers to fix problems,
>and may be required to back out if a fix cannot be found

Comment 15

[Ryan VanderMeulen [:RyanVM]]

I've pushed a backout of bug 782547 to m-c. It should be in the next nightly.
https://hg.mozilla.org/mozilla-central/rev/3621795c03e1

Comment 16

[Virtual_ManPL [:Virtual] 🇵🇱 - (please needinfo? me - so I will see your comment/reply/question/etc.)]

(In reply to Ryan VanderMeulen from comment #15)
> I've pushed a backout of bug 782547 to m-c. It should be in the next nightly.
> https://hg.mozilla.org/mozilla-central/rev/3621795c03e1

Thank you very much!