Participating Frequently

Question

Flash Player 9,0,28,0 Causes IE 7.0 to die.

Forum|Forum|19 years ago
April 6, 2007
39 replies
5862 views

If I happen to have an instance of a debugger (for example Dev Studio 2005) run, I get a chance to debug the exception. EVERY time this has happened, the stack was in the Flash.OCX and the problem was an invalid memory access.

You guys are accessing freed memory.

This topic has been closed for replies.

Show previous replies

M

mtnntwks

Participant

I have a customer in San Jose, Ca with the same problem. I have called and left a message for Bentley Wolfe.

As I read through this thread, Adobe's problem is they can't duplicate this problem on any of their computers on their end. The problem is intermittent on the computers that are affected. And the problem is not consistent across all computers (I can't make it crash on mine no matter what I do).

So I've called my customer and have received permission from them to give Adobe physical access to my customer's computer who is affected by the issue.

If there is an engineer in the San Jose campus, call me on my cell phone and I will make an appointment to give you physical access to a computer with the problem.

I realize that a site visit is probably far outside the scope of support, but if you don't get access to a computer with the problem, you're doomed to wait for someone who can give you a log with the "magic key" to the problem. Let's be a little more proactive about resolving this issue. This issue has already been going on for over a month and based on what I'm reading in this thread, there are a lot of smart people who are no closer to finding a problem then when this thread was first started.

Thanks...
Dave

B

Bbbent

Inspiring

IDisposable, open a web case and get a case number. Then we can attach the logs to the case (and have tracking...) Reply back with the case # when you get it and I'll have a level2 take it over..

I

IDisposableAuthor

Participating Frequently

Can't you use the one I originally opened and you guys closed for no good reason? 172583030

I

IDisposableAuthor

Participating Frequently

Just FYI, I reopened the case and pointed it back to here... and you might tell the people that run this site that it should be possible to attach DMP and ZIP files to the cases...

B

Bbbent

Inspiring

We're escalating to Engineering on any crash for which we have a crash dump log. I can't say how or when it might be resolved. Sorry.

M

mrhcky20

Participating Frequently

Is this problem ever gonna be resolved?? It's really frustrating when all you're trying to do is surf the web and I can't even do that for very long before I have to close my IE. VERY DISAPPOINTING AT THIS POINT GUYS!!!!

C

CalvinoHobbes

Participant

i tested flash 9.0.45 on my Mac, same problem

A

Anonymous

Hi etolman,

Have you tested this with the latest version of Flash Player 9.0.45.0?

E

etolman

Participant

Yes, we have tested 9.0.45 as well and have the same issue.

We will also make sure to upgrade to 9.0.r45, which is the current version of the debug player and test with that/

B

Bbbent

Inspiring

"I think Adobe needs to step up here and make this a priority to be fixed."

This is exactly what we're doing, internally.

E

etolman

Participant

I have a similar issue, we are developing a flash based app, and get frequent crashes where the flash player crashes while attempting to dereference a null pointer. I tried to get the crash dump log, but the instructions in the thread didn't work.

The current crash we get is the flash player crashing when moving the mouse over the contents of a Flex application's VBox control with a bunch of Flex CheckBox buttons in the container

This is using the 9.0.28 debug player:

In the debugger the call stack is as follows:

> FlDbg9b.ocx!3006c772()
[Frames below may be incorrect and/or missing, no symbols loaded for FlDbg9b.ocx]
FlDbg9b.ocx!3008ee9f()
FlDbg9b.ocx!3009d9c6()
FlDbg9b.ocx!300c08b8()
FlDbg9b.ocx!300c103d()
FlDbg9b.ocx!300ad5f9()
FlDbg9b.ocx!300ad994()
user32.dll!77d48734()
user32.dll!77d48816()
user32.dll!77d4c63f()
user32.dll!77d4e905()
App.exe!CWnd::DefWindowProcA(unsigned int nMsg=512, unsigned int wParam=0, long lParam=24904274) Line 1035 + 0x13 bytes C++
App.exe!CWnd::WindowProc(unsigned int message=512, unsigned int wParam=0, long lParam=24904274) Line 1742 + 0x13 bytes C++
App.exe!AfxCallWndProc(CWnd * pWnd=0x00000000, HWND__ * hWnd=0x001206a8, unsigned int nMsg=512, unsigned int wParam=0, long lParam=24904274) Line 243 C++
App.exe!AfxWndProc(HWND__ * hWnd=0x001206a8, unsigned int nMsg=512, unsigned int wParam=0, long lParam=24904274) Line 388 + 0x10 bytes C++
user32.dll!77d48734()
user32.dll!77d48816()
user32.dll!77d489cd()
App.exe!CWnd::FromHandle(HWND__ * hWnd=0x00185d70) Line 311 C++
user32.dll!77d496c7()
App.exe!AfxInternalPumpMessage() Line 183 C++
App.exe!CWnd::RunModalLoop(unsigned long dwFlags=5) Line 4322 + 0x5 bytes C++
App.exe!CDialog::DoModal() Line 544 + 0x8 bytes C++
App.exe!CreatorApp::InitInstance() Line 65 C++
App.exe!AfxWinMain(HINSTANCE__ * hInstance=0x00400000, HINSTANCE__ * hPrevInstance=0x00000000, char * lpCmdLine=0x00161f06, int nCmdShow=1) Line 37 + 0x7 bytes C++
App.exe!__tmainCRTStartup() Line 315 + 0x1c bytes C

The message is WM_MOUSEMOVE, and the assembly and exception indicate dereferencing a NULL pointer:

Assembly:
3006C742 FF 74 24 04 push dword ptr [esp+4]
3006C746 8B 4C 24 0C mov ecx,dword ptr [esp+0Ch]
3006C74A E8 F5 6B FC FF call 30033344
3006C74F FF 74 24 04 push dword ptr [esp+4]
3006C753 E8 01 87 14 00 call 301B4E59
3006C758 59 pop ecx
3006C759 C3 ret
3006C75A 81 E1 00 F0 FF FF and ecx,0FFFFF000h
3006C760 8B 01 mov eax,dword ptr [ecx]
3006C762 8B 80 28 01 00 00 mov eax,dword ptr [eax+128h]
3006C768 C3 ret
3006C769 F6 41 57 02 test byte ptr [ecx+57h],2
3006C76D 8B 41 78 mov eax,dword ptr [ecx+78h]
3006C770 74 02 je 3006C774
-----> 3006C772 8B 00 mov eax,dword ptr [eax] ; eax is 0, so it crashes
3006C774 C3 ret

I can't really think of anything we could be doing that would cause a valid crash inside of flash in response to moving the mouse.

Is there any way to get support for this? We can't really release our application like this.

O

OceanSize-orMcRw

Participant

Okay, I followed the instructions to create the supposed error log files using "mms.cfg".
No ".mdmp" files are on my hard drive at all. I've searched the entire drive.

There were multiple URLs open when the crash happened, however I was watching a video on youtube.com when the crash took place.
That was the only flash being played on any of the URLs that were open.
The error message indicated that it was the flash plugin that crashed.
Therefore, the remaining URLs I had open have absolutely no bearing on this issue.

I think Adobe needs to step up here and make this a priority to be fixed.
I find it impossible to believe that Adobe cannot recreate a crash scenario in-house.
As everyone can plainly see - this is a problem for people. And we're supposed to think that miraculously, it happens for everyone in the world except Adobe employees?

Further, it's ridiculous to expect your users to do all the trouble shooting for you. I might be more inclined to do so if Adobe were paying me to do that. As it happens, they already pay people who should be trouble shooting this.

A

adobeisdumb

Participating Frequently

Has adobe ever looked into the flash player having a memory leak? The reason i ask is because I have to put up 8 or 10 flash sites and then it usually crashes but just one and my computer is fine. Bloomberg.com, if you open up a bunch of news stories their flash is at the top of every page and it will crash in about an hour for me.

M

mrhcky20

Participating Frequently

The guy talking about WWW.WEATHER.COM wasn't kidding. First time ever there, went to the interact page, and BOOM!! Dude ain't lying!!

Show more replies