Closed Bug 879022 Opened 11 years ago Closed 3 years ago

Whe a secondary page is loaded in an iframe with additional flash objects then all flash objects stop working.

Categories

(Core Graveyard :: Plug-ins, defect, P3)

Product:
Core Graveyard
Component:
Plug-ins
Version:
17 Branch
Platform:
x86_64
Linux
Type:
defect

Tracking

(Not tracked)

Status:
RESOLVED WONTFIX

People

(Reporter: rbarajas, Unassigned)

Details

User Agent: Mozilla/5.0 (compatible; MSIE 9.0; Windows NT 6.1; WOW64; Trident/5.0; SLCC2; .NET CLR 2.0.50727; .NET CLR 3.5.30729; .NET CLR 3.0.30729; Media Center PC 6.0; .NET4.0C; .NET4.0E; InfoPath.3; Creative AutoUpdate v1.41.07)

Steps to reproduce:

All that is required is to use our product XPOCAST which uses the Flash Player to deliver rich media. You open the main XPOCAST web console and start a broadcast. Everything is fine until you load a secondary page in an iframe. If the secondary page contains more flash objects then all flash objects stop working.


Actual results:

 Flash objects then all flash objects stop working.


Expected results:

Well the flash objects should not stop working. It works correctly in 32 bit but not in 64 bit. System is a Red Hat Enterprise Linux workstation Release 6.3 (Santiago) Kernel Linux 2.6.32-333.el6.x86-64 and Firefox version is Firefox ESR 17.0.2 and the Flash Player version is 11.2.202.258
Severity: normal → major
OS: Windows 7 → Linux
What does 'stop working' mean here? Crashes, audio/video breakage, ...?
Do you have a stand-alone test-case or a publicly accessible URL?
Flags: needinfo?(rbarajas)
(In reply to Georg Fritzsche [:gfritzsche] from comment #1)
> What does 'stop working' mean here? Crashes, audio/video breakage, ...?
> Do you have a stand-alone test-case or a publicly accessible URL?

Here is a url that you can use to recreate the issue.

https://vts.inxpo.com/Launch/Event.htm?ShowKey=14564

As stated above this only happens with 64bit version of 17 ESR in RHEL.
Flags: needinfo?(rbarajas)
(In reply to rbarajas from comment #2)
> (In reply to Georg Fritzsche [:gfritzsche] from comment #1)
> > What does 'stop working' mean here? Crashes, audio/video breakage, ...?
> > Do you have a stand-alone test-case or a publicly accessible URL?
> 
> Here is a url that you can use to recreate the issue.
> 
> https://vts.inxpo.com/Launch/Event.htm?ShowKey=14564
> 
> As stated above this only happens with 64bit version of 17 ESR in RHEL.

Note: you may need to enter 14564 in the "Login ID" prompt, the ShowKey parameter wasn't working for me

I was unable to reproduce this on 17esr, either nightly or 17.0.2 release. I'm on a 64bit host, and tried both 32bit and 64bit flash/firefox. My system is significantly newer than a RHEL release, so there are likely a number of differences. You mention 17.0.2 and a slightly out of date flash, did you double check that this isn't solved by 17.0.6 or flash 11.2.202.285?

Are you using RHEL builds of firefox? Does using a mozilla build make any difference? 
( https://ftp.mozilla.org/pub/mozilla.org/firefox/releases/17.0.6esr/linux-x86_64/en-US/ )

Does this work in a nightly build (non-ESR)? In that case, we have a pretty easy to use tool that could track down the nightly that fixed it, which could shed light on the issue:
http://mozilla.github.io/mozregression/
Flags: needinfo?(rbarajas)
(In reply to John Schoenick [:johns] from comment #3)
> (In reply to rbarajas from comment #2)
> > (In reply to Georg Fritzsche [:gfritzsche] from comment #1)
> > > What does 'stop working' mean here? Crashes, audio/video breakage, ...?
> > > Do you have a stand-alone test-case or a publicly accessible URL?
> > 
> > Here is a url that you can use to recreate the issue.
> > 
> > https://vts.inxpo.com/Launch/Event.htm?ShowKey=14564
> > 
> > As stated above this only happens with 64bit version of 17 ESR in RHEL.
> 
> Note: you may need to enter 14564 in the "Login ID" prompt, the ShowKey
> parameter wasn't working for me
> 
> I was unable to reproduce this on 17esr, either nightly or 17.0.2 release.
> I'm on a 64bit host, and tried both 32bit and 64bit flash/firefox. My system
> is significantly newer than a RHEL release, so there are likely a number of
> differences. You mention 17.0.2 and a slightly out of date flash, did you
> double check that this isn't solved by 17.0.6 or flash 11.2.202.285?
> 
> Are you using RHEL builds of firefox? Does using a mozilla build make any
> difference? 
> (
> https://ftp.mozilla.org/pub/mozilla.org/firefox/releases/17.0.6esr/linux-
> x86_64/en-US/ )
> 
> Does this work in a nightly build (non-ESR)? In that case, we have a pretty
> easy to use tool that could track down the nightly that fixed it, which
> could shed light on the issue:
> http://mozilla.github.io/mozregression/

Yes, the Firefox build I'm testing with is a RHEL build. I'm using a machine that they provided me for testing. I will try using the updated version and updating flash. I will post back my results.
Flags: needinfo?(rbarajas)
(In reply to rbarajas from comment #4)
> (In reply to John Schoenick [:johns] from comment #3)
> > (In reply to rbarajas from comment #2)
> > > (In reply to Georg Fritzsche [:gfritzsche] from comment #1)
> > > > What does 'stop working' mean here? Crashes, audio/video breakage, ...?
> > > > Do you have a stand-alone test-case or a publicly accessible URL?
> > > 
> > > Here is a url that you can use to recreate the issue.
> > > 
> > > https://vts.inxpo.com/Launch/Event.htm?ShowKey=14564
> > > 
> > > As stated above this only happens with 64bit version of 17 ESR in RHEL.
> > 
> > Note: you may need to enter 14564 in the "Login ID" prompt, the ShowKey
> > parameter wasn't working for me
> > 
> > I was unable to reproduce this on 17esr, either nightly or 17.0.2 release.
> > I'm on a 64bit host, and tried both 32bit and 64bit flash/firefox. My system
> > is significantly newer than a RHEL release, so there are likely a number of
> > differences. You mention 17.0.2 and a slightly out of date flash, did you
> > double check that this isn't solved by 17.0.6 or flash 11.2.202.285?
> > 
> > Are you using RHEL builds of firefox? Does using a mozilla build make any
> > difference? 
> > (
> > https://ftp.mozilla.org/pub/mozilla.org/firefox/releases/17.0.6esr/linux-
> > x86_64/en-US/ )
> > 
> > Does this work in a nightly build (non-ESR)? In that case, we have a pretty
> > easy to use tool that could track down the nightly that fixed it, which
> > could shed light on the issue:
> > http://mozilla.github.io/mozregression/
> 
> Yes, the Firefox build I'm testing with is a RHEL build. I'm using a machine
> that they provided me for testing. I will try using the updated version and
> updating flash. I will post back my results.

We've updated the Firefox build to 17.0.6 and as for the flash player version 11.2.202.285 we are unable to located the download for it. Every page from Adobe states we have the latest Flash Player version. We still are experiencing the issue with 17.0.6.
(In reply to rbarajas from comment #5)
> We've updated the Firefox build to 17.0.6 and as for the flash player
> version 11.2.202.285 we are unable to located the download for it. Every
> page from Adobe states we have the latest Flash Player version. We still are
> experiencing the issue with 17.0.6.

The most recent version that can be downloaded on [1]/[2] is 11.2.202.285 - maybe you you already have that version installed and not 11.2.202.258?

Did you have a chance to test this in a Nightly build? [3]

[1] http://get.adobe.com/flashplayer/
[2] http://get.adobe.com/flashplayer/otherversions/
[3] http://nightly.mozilla.org/
Flags: needinfo?(rbarajas)
(In reply to Georg Fritzsche [:gfritzsche] from comment #6)
> (In reply to rbarajas from comment #5)
> > We've updated the Firefox build to 17.0.6 and as for the flash player
> > version 11.2.202.285 we are unable to located the download for it. Every
> > page from Adobe states we have the latest Flash Player version. We still are
> > experiencing the issue with 17.0.6.
> 
> The most recent version that can be downloaded on [1]/[2] is 11.2.202.285 -
> maybe you you already have that version installed and not 11.2.202.258?
> 
> Did you have a chance to test this in a Nightly build? [3]
> 
> [1] http://get.adobe.com/flashplayer/
> [2] http://get.adobe.com/flashplayer/otherversions/
> [3] http://nightly.mozilla.org/

We have upgrade the Flash player and still we saw the issue. We then proceeded to use the nightly build and still the issue was present. Do you think there is merit in trying a newer version of the RHEL build, or trying to recreate the issue on a different Linux build.
Flags: needinfo?(rbarajas)
(In reply to rbarajas from comment #8)
> Do you
> think there is merit in trying a newer version of the RHEL build, or trying
> to recreate the issue on a different Linux build.

Per comment 3, definitely try with a newer RHEL build if upgrading that is an option.

You could also still try mozregression [1] for more insight into this.

[1] http://mozilla.github.io/mozregression/
(In reply to Georg Fritzsche [:gfritzsche] from comment #9)
> (In reply to rbarajas from comment #8)
> > Do you
> > think there is merit in trying a newer version of the RHEL build, or trying
> > to recreate the issue on a different Linux build.
> 
> Per comment 3, definitely try with a newer RHEL build if upgrading that is
> an option.
> 
> You could also still try mozregression [1] for more insight into this.
> 
> [1] http://mozilla.github.io/mozregression/

According to our contact at Red Hat trying a new version is not option. We were wondering if we could screen sharing from the Red Hat loaned machine with you guys. We figure this would help in figuring out the issue.
What exact behavior are you seeing when flash stops working? Does the entire browser crash or hang, or does just flash hang or crash (the "Adobe flash has stopped working" dialog)?

I fired up a RHEL6 VM on AWS and tried to reproduce this in a VNC session. With nightly, esr 17.0.6 (mozilla) and esr 17.0.6 (RHEL), I can play the presentation successfully. However, with ESR 17.0.2, I was able to get the "Adobe Flash has stopped working" dialog when the second instance spawns, though not every time.

You mentioned you tested this in a nightly build, was that a mozilla-central (firefox 24) nightly build, or a ESR17 nightly?

Another thing to verify, RHEL6 x64 has both 32-bit and 64-bit builds of firefox available, are you using both 64bit firefox and 64bit flash? (i.e. not the 32bit flash via the "npwrapper" tool)

You can also try disabling hardware acceleration in flash (right click -> settings), to rule out driver issues.
With ESR 17.0.2 the flash objects hang but with the central build the browser crashes.

With the nightly build tried te central build, we weren't aware that there was also a nightly build of ESR17. Yes we are using the 64bit version of flash. Where can we get the nightly build ESR17?
(In reply to rbarajas from comment #12)
> With the nightly build tried te central build, we weren't aware that there
> was also a nightly build of ESR17. Yes we are using the 64bit version of
> flash. Where can we get the nightly build ESR17?

You can get ESR17 nightlies at:
https://ftp.mozilla.org/pub/mozilla.org/firefox/nightly/latest-mozilla-esr17/

But if this is still broken on central, they're not likely to be helpful.

> With ESR 17.0.2 the flash objects hang but with the central build the
> browser crashes.

When the browser crashes you should be given the option to submit a crash report -- Can you go to about:crashes and link us to one of the reports you get from this?

Does this still occur if you open the process in a VNC session, as opposed to a real X11 server?
(In reply to John Schoenick [:johns] from comment #13)
> (In reply to rbarajas from comment #12)
> > With the nightly build tried te central build, we weren't aware that there
> > was also a nightly build of ESR17. Yes we are using the 64bit version of
> > flash. Where can we get the nightly build ESR17?
> 
> You can get ESR17 nightlies at:
> https://ftp.mozilla.org/pub/mozilla.org/firefox/nightly/latest-mozilla-esr17/
> 
> But if this is still broken on central, they're not likely to be helpful.
> 
> > With ESR 17.0.2 the flash objects hang but with the central build the
> > browser crashes.
> 
> When the browser crashes you should be given the option to submit a crash
> report -- Can you go to about:crashes and link us to one of the reports you
> get from this?
> 
> Does this still occur if you open the process in a VNC session, as opposed
> to a real X11 server?

My IT department is currently in the process of trying to install the nightly build of ESR. It appeared that it still did not work but now they are trying to re-install flash because after nightly was installed the flash player is no longer found. 

As for the VNC vs X11 question, the only way we are engaging with RHEL is thru the official machine that was provided by Red Hat themselves.

Once I hear back from my IT department I will update the post.
We are able to test it with Firefox ESR 17.0.7 but the browser crashed. I'm including the link to the crash report https://crash-stats.mozilla.com/report/index/7f819095-d9c8-4312-a2de-c13202130627.
(In reply to rbarajas from comment #15)
> We are able to test it with Firefox ESR 17.0.7 but the browser crashed. I'm
> including the link to the crash report
> https://crash-stats.mozilla.com/report/index/7f819095-d9c8-4312-a2de-
> c13202130627.

This looks like bug 733322.
Please don't run Flash in-process (i.e. with no overrides for the dom.ipc.plugins.* settings in about:config), otherwise issues with Flash can crash the browser.

If you have Flash running out-of-process again, what problem do you end up with now? Flash crashing? If yes, can you submit and link crash reports for that?
Flags: needinfo?(rbarajas)
(In reply to Georg Fritzsche [:gfritzsche] from comment #16)
> (In reply to rbarajas from comment #15)
> > We are able to test it with Firefox ESR 17.0.7 but the browser crashed. I'm
> > including the link to the crash report
> > https://crash-stats.mozilla.com/report/index/7f819095-d9c8-4312-a2de-
> > c13202130627.
> 
> This looks like bug 733322.
> Please don't run Flash in-process (i.e. with no overrides for the
> dom.ipc.plugins.* settings in about:config), otherwise issues with Flash can
> crash the browser.
> 
> If you have Flash running out-of-process again, what problem do you end up
> with now? Flash crashing? If yes, can you submit and link crash reports for
> that?

We tried this and it appears to work when we run our product in VOD mode but when we try it in Live mode the flash player crashes. Here is one of the crash reports.

https://crash-stats.mozilla.com/report/index/1e473135-b818-46ba-890f-288262130701
Flags: needinfo?(rbarajas)
I was wondering if I could an update on my last post.
Sorry, I meant to say I was wondering if I could get an update on my last post. Thanks

(In reply to rbarajas from comment #18)
> I was wondering if I could an update on my last post.
There is no update to give. Nobody has volunteered to do further diagnosis and this is not a high-priority item.
Priority: -- → P3
(In reply to rbarajas from comment #17)
> (In reply to Georg Fritzsche [:gfritzsche] from comment #16)
> > (In reply to rbarajas from comment #15)
> > > We are able to test it with Firefox ESR 17.0.7 but the browser crashed. I'm
> > > including the link to the crash report
> > > https://crash-stats.mozilla.com/report/index/7f819095-d9c8-4312-a2de-
> > > c13202130627.
> > 
> > This looks like bug 733322.
> > Please don't run Flash in-process (i.e. with no overrides for the
> > dom.ipc.plugins.* settings in about:config), otherwise issues with Flash can
> > crash the browser.
> > 
> > If you have Flash running out-of-process again, what problem do you end up
> > with now? Flash crashing? If yes, can you submit and link crash reports for
> > that?
> 
> We tried this and it appears to work when we run our product in VOD mode but
> when we try it in Live mode the flash player crashes. Here is one of the
> crash reports.
> 
> https://crash-stats.mozilla.com/report/index/1e473135-b818-46ba-890f-
> 288262130701

This looks like a pure flash crash -- something that you would need to take up with adobe. One common debugging step is disabling flash's hardware acceleration.

For the issues you're seeing in in-process mode, it honestly is likely also an adobe bug -- but in-process flash is not generally supported by adobe and is increasingly unsupported by us, so quirks in that configuration are unfortunately not likely to get looked into :(
Resolving as wont fix, plugin support deprecated in Firefox 85.
Status: UNCONFIRMED → RESOLVED
Closed: 3 years ago
Resolution: --- → WONTFIX
Product: Core → Core Graveyard
You need to log in before you can comment on or make changes to this bug.