Flash Player continuously crashing in Mozilla and Chrome

Report · Feb 15, 2017

Since 3 days I am facing the problem with FP, as it crashes immediately after loading the web page

Report · Feb 15, 2017

Report · Feb 16, 2017

Hi,

Please share the URLs for which you can reproduce the issue. Also, you can try after disabling Hardware acceleration and share the details.

Thanks!

Report · Feb 15, 2017

I am using Mozilla 51.0.1 (32-bit)

Report · Feb 17, 2017

In Firefox, please try disabling hardware acceleration.

Instructions are included here:

https://helpx.adobe.com/flash-player/kb/video-playback-issues.html

If the problem goes away, you need to update your graphics drivers. If none are available (or updated drivers do not resolve the problem), you can leave hardware acceleration disabled until drivers become available that fix the problem.

If the issue persists, you can re-enable hardware acceleration (since that's not the problem) and then please provide a link to the Mozilla crash report for further analysis.

In the address bar, type:

about:plugins

You should see a list of crashes. Click the first few to submit them to Mozilla, then paste the links to those pages in a reply here. I'll be happy to take a look at them.

Thanks!

Report · Feb 18, 2017

Hello,

we are facing the same problem with the latest flash plugin.

The 24.0.0.194 version worked just fine and now with the 24.0.0.221 the application always crashes. The application has worked months and years before without any issues. Unfortunately I can not provide a link because the application is in a protected area, but I enclosed the crash reports below.

Here are some related crash reports from Firefox (all browsers are affected as soon as the latest 24.0.0.221 is installed).

It seems to be a common problem, here is another example Flash 24.0.0.221 crashing issues discussion on Kongregate

https://crash-stats.mozilla.com/report/index/c814f03a-3fa2-4d0e-981c-7a80c2170218#tab-details

https://crash-stats.mozilla.com/report/index/918455bf-918e-493b-a1fc-e6c632170218

Regards,

Ramin

Report · Feb 21, 2017

I've opened a bug for these two (they're the same crash), but I'm skeptical that they're reflective of the new issue being reported here.

https://crash-stats.mozilla.com/report/index/c814f03a-3fa2-4d0e-981c-7a80c2170218#tab-deta ils

https://crash-stats.mozilla.com/report/index/918455bf-918e-493b-a1fc-e6c632170218

[Update: these are also the same signature]

https://crash-stats.mozilla.com/report/index/c336b559-3692-4294-a737-7d8d12170215

https://crash-stats.mozilla.com/report/index/365a60ee-36d6-4055-856d-6ad3b2170220

According to the aggregate stats in Mozilla's crash reporter, they're relatively low volume, and seem to affect several previous versions. Looking at the stacks themselves, there's not an obvious smoking gun. A set of repro steps would greatly improve our chances of fixing it. In this instance, it's quite possible that the bad allocation happened hundreds or thousands of steps before the actual crash (i.e. way before the logs here start), and an interactive debugging session would turn this from something we're unlikely to solve to a straightforward debugging exercise.

I'm still working through the other crash reports and will post updates as I go.

[Update]

If this crash really is the issue, it's definitely not related to hardware acceleration. It's really strange to me that you're suddenly just starting to see it with the latest versions, though.

Based on the callstack, moving to 64-bit Firefox might avoid this problem entirely. Mozilla serves 32-bit Firefox by default, so you have to jump through a couple hoops to get the 64-bit Firefox at the moment. That said, that allows us to run 64-bit Flash Player, which is *far* more secure and stable, and the native NPAPI sandbox implementation means that we can ditch our bolt-on sandbox for 32-bit Firefox (this is architecturally more efficient, and solves a number of IPC-related problems that commonly affect 32-bit Windows Firefox users (the classic hangs and crashes that have plagued Flash in Firefox for years).

Thanks!

Report · Feb 20, 2017

Hi,

Here is two other crash reports :

- https://crash-stats.mozilla.com/report/index/c336b559-3692-4294-a737-7d8d12170215

- https://crash-stats.mozilla.com/report/index/365a60ee-36d6-4055-856d-6ad3b2170220

Same as Ramin, i can't provide a link because the app is on a private network.

Tests have been done with & without hardware acceleration, without any differences

Report · Mar 10, 2017

Your crash signatures are the same as Ramin's. We were able to work with him and his team to reproduce, diagnose and resolve the issue. (We actually had work that fixed this in the pipeline, but it was slated for the June major version update.) We were able to prioritize that fix and confirm using Ramin's content that the issue was fixed, and provided preview drops to his team for additional testing, which looks good.

Report · Feb 20, 2017

Hi,

What Adobe is failing to acknowledge is that there is a problem with Flash Player. The issue is that FP worked fine prior to version 24.0.0.221. You (Adobe) have broken FP with version 24.0.0211 and you need to fix it, not tell customers to update their graphics drivers.

Here is a link to a page where FP crashes: Flash Player video playback issues. You'll notice that this is one of your own pages. For what it is worth, the suggestion contained on that page to disable hardware acceleration to prevent FP crashing does not work and it continues to crash.

If Adobe bothered to check, they would see that the problem of version 24.0.0.211 crashing is ubiquitous on the Internet. Adobe needs to make the necessary effort to solve the problem. Even on this page Adobe have stated that the problem is "assumed answered" when it clearly is not. Come on Adobe, get your act together!

Report · Feb 21, 2017

Unfortunately, the forums don't have an "In Process" status. I'm subscribed to email notifications from this thread, and they land in my inbox when people reply. I marked the thread Assumed Answered so that the other folks answering questions can differentiate the threads actively being worked from the threads that don't yet have a response.

That said, "it crashes" isn't going to lead us to a solution, which is all I really care about. While you're more than welcome to vent, "get it together" isn't going to help me get your problem solved faster, and with ~2 billion users and 2.5 million lines of code, not all crashes share the same root cause. Since the only people that submit posts to the forums are people with problems, there's significant selection bias in the sample set.

This particular thread is interesting because there's a cluster of people seeing a similar symptom in the same timeframe, but I don't have much in the way of actionable data yet.

The absolute fastest and most consistent way for us to fix a crash is to reproduce it under a debugger. To do that, I need to find a reasonably similar machine, affected content, and a set of steps that results in the crash. That's why I'm asking for those details. I believe that you guys see a problem.

The trick (since the player has already gone through tens of thousands of test hours across >30K+ automated and manual tests, and real-world content before release across a wide variety of hardware, and was also distributed to a beta audience of ~5 million users) is figuring out what the common denominator is, so that I can narrow the search to something that's going to show up in lab conditions. That's why I ask for the data I ask for.

To that end, the following information is extremely useful:

https://forums.adobe.com/thread/1195540

As are actual crash reports:

https://helpx.adobe.com/flash-player/kb/report-flash-player-crash.html

Report · Feb 22, 2017

[@ FlashPlayerDebugger-10.6@0x91398e] - Firefox 51.0.1 Crash Report - Report ID: 9f67a3c3-621d-4cb4-...

Chrome: Absturz-ID 683c725c-daa5-4fe9-820a-5bfb9cefdb94 (Server-ID: e3ccf1b940000000)

We are running into crashing issues with the new version as well.

Customers have reported it, and we have verified it, on multiple systems and browsers.

I've enabled the debugger version, if that helps at all.

Report · Mar 10, 2017

Thanks for providing the crash dumps from the Debugger.

For what it's worth, the ActionScript debugger doesn't include C++ symbols, so it doesn't really help us investigate crashes. It's actually less helpful than crash dumps from the release player. I can't reliably compare crash signatures that Mozilla generates that are encountered in the debugger to the aggregate body of signatures coming in from the release clients. The sample size of the debugger audience is tiny compared to the release player, so I can't confidently correlate the issue you're seeing in the debugger to one or more crash signatures that may already be on my radar, so it's hard to differentiate a crash that only affects the debugger from one that has a meaningful impact on the larger audience. Since the debugger is explicitly a developer tool and not something that you should necessarily browse the web with, our standards there are also different.

In terms of things that help us to quickly fix bugs like this, the following pieces of info are useful (in order of helpfulness):

A link and reproducible steps that demonstrate the crash, and a detailed description of the browser/os combination(s) where you see the problem.
A Firefox crash report with the regular release Flash Player
A minidump (or crash report from Mac) from other browsers (Mozilla makes their crash reporting data public, which is useful. The other crash reports go to the browser vendors, but we don't get the same kind of broad access for investigation, so the crash dumps themselves are more useful).

Report · Feb 20, 2017

Our in-house enterprise Flex/Flash-based software is now crashing across the board since the update. I watched this spread on a day-to-day basis as automatic updates went into our franchise system. Flash code that has been stable since 2006 is now crashing every 10-15 minutes, even with no other browser windows open and nothing running in the background. Unfortunately we have had to roll every computer in the company back, manually, to a previous version of Flash. This is happening in ALL BROWSERS ON MAC AND PC. It happens at random and has nothing to do with the stability of the Flash code. In fact, it even happens when only running our legacy AS2 software!

We have dozens of computers all experiencing the same problem, including both PCs and Macs, many of them brand new. We have also confirmed that Firefox shows slightly less crashing, but still crashes. Pepper is a disaster. Turning off hardware acceleration does NOT fix the problem, so this is not a graphics card issue as far as I can see. And if it were one, then there's literally no modern graphics card that doesn't have the issue.

I may be the only person in a position to have dozens of people running Flash player as part of their front desk business software 24/7/365, so I've been in a position to test a lot of different fixes and so far the only one is to roll back to the previous version.

24.0.0.221 is very broken. We have rolled our whole organization back and hope Adobe addresses the issue.

Report · Feb 20, 2017

Yes this happened to us as well, as mentioned inthis forked conversation:

Intermittent Crash with Video and Webcam Content

We were able to strip the flex code out of our player (as it really wasnt needed) and that code appears to be stable on ...221

However the original was totally screwed by that last update. Ill post crash dumps when i get back to work, but yea it does crash other places as well as Digby suggests.

Report · Feb 21, 2017

https://bugs.chromium.org/p/chromium/issues/detail?id=693362

(prob didnt need to actualyl submit that) but wasnt sure where to link the crash repts:

Crash ID bd70377c-b41f-4295-91a9-0c39f841afaf (Server ID: 3cce1a8440000000)

Crash ID 1a65bc0a-cde7-4cd8-a33a-3faabb6c83dc (Server ID: c190e73580000000)

Report · Feb 21, 2017

I'm trying to come up to speed on this, but it's unclear to me that we've been given reproducible steps or a site.

I've tried the Flash Video help site mentioned in post #7 and I've played the anti-idle game on Kongregate on both Chrome and Firefox on Windows 10. Both browsers are running Flash 24.0.0.221. I haven't crashed yet, but I'll keep trying.

If anyone is willing to contact me privately with detailed steps or access to a test site, please email me at ccampbel@adobe.com.

Thanks,

Chris

Report · Feb 21, 2017

https://bugs.chromium.org/p/chromium/issues/detail?id=693362

(prob didnt need to actualyl submit that) but wasnt sure where to link the crash repts:

Crash ID bd70377c-b41f-4295-91a9-0c39f841afaf (Server ID: 3cce1a8440000000)

Crash ID 1a65bc0a-cde7-4cd8-a33a-3faabb6c83dc (Server ID: c190e73580000000)

Unfortunately for Chrome, we don't have access to Chrome crash data in the same way that we do with Mozilla's crash reporting system. We can analyze Chrome stacks, but you actually have to send them to us.

There are details here on how to retrieve and share those crash dumps:

https://helpx.adobe.com/flash-player/kb/report-flash-player-crash.html

If you have questions or run into problems, please let me know. I'm happy to help.

Thanks!

Report · Feb 21, 2017

Maybe a wild goose chase, but is 24.0.0.221 the first version where uncaught error events in the Pepper debug player don't trigger the "Dismiss All / Continue" browser dialog, but instead show that little Flex-type dialog box in the movie itself? I can't remember seeing that before. And it's not consistent with the behavior of the Firefox debug player.

Interestingly I have not yet seen this crash happen in the 221 debug player. Has anyone? I guess I'm wondering whether there could be some code in the non-debug players for suppressing errors that got switched off when this dialog box change was made.

Report · Feb 23, 2017

I was experiencing the same issue with my code. I managed to track down the point in my code where it was breaking. It only crashed after Transition.OUT was called. I removed this from my code yesterday and it has not seen a crash since.

Hope this information will be of some help.

Report · Mar 09, 2017

I have FINALLY found the fix for this issue. Open your flash player settings under in your control panel. Add the site you want to allow...then choose...BLOCK ALL SITES FROM STORING INFORMATION TO THIS COMPUTER> ...Problem solved. I was having problems playing any facebook games and adobe plugin kept crashing. Now this no longer happens.

Report · Mar 10, 2017

This is obviously a REAL bug and all intranet users of our software have been forced back to previous versions of Flash Player.

So no, it has nothing to do with the Transition library.

I'm questioning neils response but we will try it. If it has to do with local storage, then the question is whether it's a SharedObject problem or whether maybe malicious code is being executed on some machines. I have seen machines with the same version of Flash that are not showing the problem, and it occurred to me that there might be a zero day exploit involved for someone who encountered or continues to encounter nasty SO's.

Adobe: Please take this seriously, and don't wait for an aftermarket AS3 coder to provide proof. You know pretty well that most people using Flash to run arbitrary stuff on the internet have no idea what they're doing, and that even the better ones are not qualified to determine what structural flaws you've introduced to the runtime. Not just that: many of us are dealing with legacy software and we are not getting full crash reports from people using it. But we're hearing that it suddenly started breaking every 15 minutes since the 221 update. There is definitely a major problem with 24.0.0.221, it's crashing constantly for most users PC and Mac, Firefox, Chrome and Safari... anyone using it in daily business for longer than a couple minutes. But it's also conceivable that something in the wild is triggering these crashes.

We will inform our users to turn off local storage and see how that goes, if it helps.

Report · Mar 10, 2017

A couple of the early reporters were able to give us reproducible examples, and we've been working with them to identify the issue and confirm the fix with them. The timing was such that there was no window for a beta release, but the fix should land in Tuesday's release. Assuming that everyone is seeing the same crash and not conflating distinct issues, the fix will land shortly.

If you have an issue that persists after applying Tuesday's update, if you can tell us how to reproduce it (e.g. provide a clear description of the affected OS and Browser, steps to reproduce, a link to affected content, etc.), debugging and fixing the issue becomes a straightforward exercise.

Thanks!

Report · Mar 10, 2017

I have a strong suspicion that the local filesystem storage issue that you're describing is distinct from the primary issue being discussed on this thread. "It crashes" isn't a problem, it's an outcome -- very much in the same way that "they died" isn't a helpful medical diagnosis. There's a timeline and list of failures that happened in a specific order that led up to the outcome, and in order fix those problems, we need to be able to differentiate (bucket) crashes (typically, by signature), and then understand the timeline to identify the pathological behavior that ultimately leads to the crash.

In general, if you were experiencing a crash on accessing files on the local system (which is what you blew away when clearing the local storage), I'd be interested in those crash logs -- just to definitely rule out the correlation to the broader issue here, but my suspicion is that the files that store that data became corrupted on the local system. I suspect that you would benefit from running your system's disk utilities, to scan for and repair any underlying problems with the filesystem.

The bug that we fixed has nothing to do with local storage, but we also haven't really changed anything about local storage lately that would have caused an uptick in crashes in timeframe that most people are talking about.

Report · Mar 15, 2017

The same thing is happening to me when I go here: http://mmanuts.com