4027 replies to this topic

[jsm174]

@DuMichAuch - Any chance for a copy of that mini dump file for the debug version of the stack trace. I'd like to see if I can reverse with WinDbg. 

[jsm174]

Also, I just put another debug version that includes the .pdb file. This might make the stack trace more readable.

 

https://github.com/j...runs/1069003121

[DuMichAuch]

VPinballX-10.7.0-287-077370d-win-x86

Crash report VPX rev287 (077370d)
============
Process: VPinballX.exe
Reason: 0xC0000005 - EXCEPTION_ACCESS_VIOLATION at 0023:0116F8E3
Attempt to read from 0x00000000
Thread ID: 0x38F8 [14584]

Call stack
==========
0116F8E3 VPinballX.exe (0x00BF2280 0x00AFE014 0x00AFDDA8 0xffffffffcccccccc)
0116F015 VPinballX.exe (0x00BF1A64 0x5BF7B992 0x00AFE3C0 0x00afe514)
0121527A VPinballX.exe (0x5BF7B9F2 0x00AFE514 0x00BF19C0 0x00afe080)
0125D116 VPinballX.exe (0x00000000 0x5BF7B93E 0x00AFE3C0 0x00afe514)
01264B7C VPinballX.exe (0x00BD650C 0x5BF7B972 0x00AFE514 0xffffffffcccccccc)
0120E6B0 VPinballX.exe (0x00AFEB3C 0x5BF7BA7E 0x00AFE4D8 0x00afe514)
0133B9B7 VPinballX.exe (0x00BDC218 0x00000001 0x00000005 0x00afeb3c)
01338DFF VPinballX.exe (0x0000138D 0x5BF7BD56 0x00AFE76C 0x00afe514)
0133ABA3 VPinballX.exe (0x0000138D 0x00000000 0x00AFE514 0x00cccccc)
0133E6D5 VPinballX.exe (0x0000138D 0x00000000 0x5BF7BECA 0x00000111)
00D79CA5 VPinballX.exe (0x00000111 0x0000138D 0x00000000 0x00afe7f0)
01367305 VPinballX.exe (0x00000111 0x0000138D 0x00000000 0x00afe7f0)
013677CC VPinballX.exe (0x00000111 0x0000138D 0x00000000 0x00afe7f0)
0136761C VPinballX.exe (0x00000111 0x0000138D 0x00000000 0x00afe81c)
0133E914 VPinballX.exe (0x00000111 0x0000138D 0x00000000 0x005810ea)
00D7963D VPinballX.exe (0x005810EA 0x00000111 0x0000138D 0x00000000)
76B5EEBB USER32.dll AddClipboardFormatListener + 0x4B (0x00CE3EBC 0x005810EA 0x00000111 0x0000138d)
76B55E7A USER32.dll GetClassLongW + 0x7AA (0x00CE3EBC 0x00000000 0x00000111 0x0000138d)
76B53BEA USER32.dll DispatchMessageW + 0x24A (0x00AFE9F8 0x0133C999 0x00AFE9D4 0x00aff6f8)
76B45D70 USER32.dll DispatchMessageA + 0x10 (0x00AFE9D4 0x00AFF6F8 0x00CD15A0 0xffffffffcccccccc)
0133C999 VPinballX.exe (0x00CD15A0 0x00AFEA3C 0x00AFF708 0x01316cca)
01334174 VPinballX.exe (0x5BF7AEBA 0x00CD15A0 0x00CD15A0 0x009f7000)
01316CCA VPinballX.exe (0x00CD0000 0x00000000 0x00B24803 0x00000001)
013B054D VPinballX.exe (0x5BF7AE36 0x00CD15A0 0x00CD15A0 0x009f7000)
013B03D7 VPinballX.exe (0x00AFF794 0x013B05C8 0x00AFF7A4 0x7513fa29)
013B026D VPinballX.exe (0x00AFF7A4 0x7513FA29 0x009F7000 0x7513fa10)
013B05C8 VPinballX.exe (0x009F7000 0x7513FA10 0x00AFF800 0x772d7a7e)
7513FA29 KERNEL32.DLL BaseThreadInitThunk + 0x1772D7A7E ntdll.ntdl
Environment
===========
Date/time: 27/7/2021, 00:48:23:89
Number of CPUs: 4
Processor type: 586
System: Windows 10 (10.0 19043)

Memory status
=============
Total Reserved: 49604K (48M) bytes
Total Commited: 166312K (162M) bytes
Total Free: 3978324K (3885M) bytes
Largest Free: 2086224K (2037M) bytes

Registers
=========
EAX=00000000 EBX=00CE3EBC ECX=00BF2280 EDX=00000000
ESI=00AFDCB4 EDI=00AFDCF4 EBP=00AFF810 ESP=00AFF808 EIP=772D7A4E
FLG=00010282 CS=0023 DS=002B SS=002B ES=002B FS=0053 GS=002B


Mini dump saved successfully.

[jsm174]

I think you ran the first one again. The sha would be different.

It should be 5a933de instead of 077370d.

[clyde]

Here is what I got from your latest version.  Crash report VPX rev287 (5a933de)

============

Process: VPinballX.exe

Reason: 0xC0000005 - EXCEPTION_ACCESS_VIOLATION at 0023:00F6F8E3

Attempt to read from 0x00000000

Thread ID: 0x2754 [10068]

 

Call stack

==========

00F6F8E3 VPinballX.exe (0x0074B730 0x0053E498 0x0053E22C 0xffffffffcccccccc)

00F6F015 VPinballX.exe (0x0B245EC4 0xFFFFFFFFF8023D74 0x0053E844 0x0053eac0)

0101527A VPinballX.exe (0xFFFFFFFFF8023D14 0x0053EAC0 0x0B245E20 0x0053e504)

0105D116 VPinballX.exe (0x00000000 0xFFFFFFFFF8023CC0 0x0053E844 0x0053eac0)

01064B7C VPinballX.exe (0x0B3545CC 0xFFFFFFFFF8023C94 0x0053EAC0 0xffffffffcccccccc)

0100E6B0 VPinballX.exe (0x0053F0E8 0xFFFFFFFFF8023180 0x0053E984 0x0053eac0)

0113B9B7 VPinballX.exe (0x0B3E7250 0x00000001 0xFFFFFFFFF8023040 0x0053ea84)

0113B83A VPinballX.exe (0x00000001 0xFFFFFFFFF8023340 0x0053ED18 0x0053eac0)

0113ABB4 VPinballX.exe (0x00008001 0x00000000 0x0053EAC0 0x00cccccc)

0113E6D5 VPinballX.exe (0x00008001 0x00000000 0xFFFFFFFFF80234F4 0x00000111)

00B79CA5 VPinballX.exe (0x00000111 0x00008001 0x00000000 0x0053ed9c)

01167305 VPinballX.exe (0x00000111 0x00008001 0x00000000 0x0053ed9c)

011677CC VPinballX.exe (0x00000111 0x00008001 0x00000000 0x0053ed9c)

0116761C VPinballX.exe (0x00000111 0x00008001 0x00000000 0x0053edc8)

0113E914 VPinballX.exe (0x00000111 0x00008001 0x00000000 0x001c03a6)

00B7963D VPinballX.exe (0x001C03A6 0x00000111 0x00008001 0x00000000)

75DEEEBB USER32.dll AddClipboardFormatListener + 0x4B (0x00AE3EBC 0x001C03A6 0x00000111 0x00008001)

75DE5E7A USER32.dll GetClassLongW + 0x7AA (0x00AE3EBC 0x00000000 0x00000111 0x00008001)

75DE3BEA USER32.dll DispatchMessageW + 0x24A (0x0053EFA4 0x0113C999 0x0053EF80 0x0053fca4)

75DD5D70 USER32.dll DispatchMessageA + 0x10 (0x0053EF80 0x0053FCA4 0x00AD15A0 0xffffffffcccccccc)

0113C999 VPinballX.exe (0x00AD15A0 0x0053EFE8 0x0053FCB4 0x01116cca)

01134174 VPinballX.exe (0xFFFFFFFFF8022564 0x00AD15A0 0x00AD15A0 0x00396000)

01116CCA VPinballX.exe (0x00AD0000 0x00000000 0x006727B7 0x00000001)

011B054D VPinballX.exe (0xFFFFFFFFF80224E0 0x00AD15A0 0x00AD15A0 0x00396000)

011B03D7 VPinballX.exe (0x0053FD40 0x011B05C8 0x0053FD50 0x7670fa29)

011B026D VPinballX.exe (0x0053FD50 0x7670FA29 0x00396000 0x7670fa10)

011B05C8 VPinballX.ex7670FA29 KERNEL32.D77AE7A7E ntdll.

Environment

===========

Date/time: 26/7/2021, 19:12:47:453

Number of CPUs: 6

Processor type: 586

System: Windows 10 (10.0 19042)

 

Memory status

=============

Total Reserved: 63784K (62M) bytes

Total Commited: 233012K (227M) bytes

Total Free: 3897444K (3806M) bytes

Largest Free: 2086008K (2037M) bytes

 

Registers

=========

EAX=00000000 EBX=00AE3EBC ECX=0074B730 EDX=00000000

ESI=0053E138 EDI=0053E178 EBP=0053FDAC ESP=0053FD58 EIP=77AE7A7E

FLG=00210282 CS=0023 DS=002B SS=002B ES=002B FS=0053 GS=002B

 

 

Mini dump saved successfully.

[DuMichAuch]

I think you ran the first one again. The sha would be different.

It should be 5a933de instead of 077370d.

 

Ups, I've sorted the test folders by name, not by time...

Crash report VPX rev287 (5a933de)
============
Process: VPinballX.exe
Reason: 0xC0000005 - EXCEPTION_ACCESS_VIOLATION at 0023:00B6F8E3
Attempt to read from 0x00000000
Thread ID: 0x36A0 [13984]

Call stack
==========
00B6F8E3 VPinballX.exe CodeViewer::InitializeScriptEngine + 0xC3 codeview.cpp(398+0018 byte(s)) (0x04A4C0E0 0x0473E6EC 0x0473E480 0xffffffffcccccccc)
00B6F015 VPinballX.exe CodeViewer::Init + 0x115 codeview.cpp(84+0019 byte(s)) (0x04A4B8C4 0xFFFFFFFFDADC8479 0x0473EA98 0x0473ebec)
00C1527A VPinballX.exe PinTable::PinTable + 0x8AA pintable.cpp(1099) (0xFFFFFFFFDADC8599 0x0473EBEC 0x04A4B820 0x0473e758)
00C5D116 VPinballX.exe ATL::CComObject<PinTable>::CComObject<PinTable> + 0x36 atlcom.h(2905+0008 byte(s)) (0x00000000 0xFFFFFFFFDADC85E5 0x0473EA98 0x0473ebec)
00C64B7C VPinballX.exe ATL::CComObject<PinTable>::CreateInstance + 0xCC atlcom.h(2964+0048 byte(s)) (0x04A1C88C 0xFFFFFFFFDADC8519 0x0473EBEC 0xffffffffcccccccc)
00C0E6B0 VPinballX.exe PinTableMDI::PinTableMDI + 0x70 pintable.cpp(10070) (0x0473F214 0xFFFFFFFFDADC8825 0x0473EBB0 0x0473ebec)
00D3B9B7 VPinballX.exe VPinball::LoadFileName + 0xD7 vpinball.cpp(948+0047 byte(s)) (0x04A208E8 0x00000001 0x00000005 0x0473f214)
00D38DFF VPinballX.exe VPinball::OpenRecentFile + 0x3F vpinball.cpp(2378) (0x0000138D 0xFFFFFFFFDADC893D 0x0473EE44 0x0473ebec)
00D3ABA3 VPinballX.exe VPinball::ParseCommand + 0x423 vpinball.cpp(686) (0x0000138D 0x00000000 0x0473EBEC 0x00cccccc)
00D3E6D5 VPinballX.exe VPinball::OnCommand + 0x55 vpinball.cpp(1594+0054 byte(s)) (0x0000138D 0x00000000 0xFFFFFFFFDADC8CD1 0x00000111)
00779CA5 VPinballX.exe Win32xx::CWnd::WndProcDefault + 0x1E5 wxx_wincore.h(1135+0023 byte(s)) (0x00000111 0x0000138D 0x00000000 0x0473eec8)
00D67305 VPinballX.exe Win32xx::CFrameT<Win32xx::CDocker>::WndProcDefault + 0x3F5 wxx_frame.h(3955) (0x00000111 0x0000138D 0x00000000 0x0473eec8)
00D677CC VPinballX.exe Win32xx::CDockFrame::WndProcDefault + 0x14C wxx_dockframe.h(214) (0x00000111 0x0000138D 0x00000000 0x0473eec8)
00D6761C VPinballX.exe Win32xx::CMDIFrameT<Win32xx::CDockFrame>::WndProcDefault + 0x11C wxx_mdi.h(714) (0x00000111 0x0000138D 0x00000000 0x0473eef4)
00D3E914 VPinballX.exe VPinball::WndProc + 0x1E4 vpinball.cpp(1656+0030 byte(s)0077963D VPinballX.76B5EEBB US
Environment
===========
Date/time: 27/7/2021, 01:34:28:55
Number of CPUs: 4
Processor type: 586
System: Windows 10 (10.0 19043)

Memory status
=============
Total Reserved: 60244K (58M) bytes
Total Commited: 180668K (176M) bytes
Total Free: 3953328K (3860M) bytes
Largest Free: 2094352K (2045M) bytes

Registers
=========
EAX=00000000 EBX=006E3EBC ECX=04A4C0E0 EDX=00000000
ESI=00790DE6 EDI=00000111 EBP=0473EF2C ESP=0473EF08 EIP=76B5EEBB
FLG=00010282 CS=0023 DS=002B SS=002B ES=002B FS=0053 GS=002B


Mini dump saved successfully.

[jsm174]

Awesome. That is exactly what we need!    

 

00B6F8E3 VPinballX.exe CodeViewer::InitializeScriptEngine + 0xC3 codeview.cpp(398+0018 byte(s)) (0x04A4C0E0 0x0473E6EC 0x0473E480 0xffffffffcccccccc)

 

https://github.com/v...deview.cpp#L398

 

 

 

I just added a check around that. Can you see if this helps:

 

 

https://github.com/j...runs/3166896569

Edited by jsm174, 26 July 2021 - 11:58 PM.

[Bendo]

I have noticed that most tables load a little slower in 10.7 than in 10.6. It hasn't been an issue, I just chalk it up to the more features being added. But today I ran into a very long load. Surf'n Safari took so long to load I thought it had crashed. After a lot of back and forth testing with both 10.6 and 10.7 I came up with this data.  In 10.6 the table loads in 4 seconds.  In 10.7 the table loads in 1 minute 3 seconds. I tested this multiple times with the same results.

 

Does anyone have an explanation or a fix for this?  For now I am just using the alternate launcher in Popper to launch the table using 10.6.

 

Thanks

[clyde]

Just tried 10.7 rev.288 (d48d82a) & It Is working now.    

[wiesshund]

Awesome. That is exactly what we need!    

 

00B6F8E3 VPinballX.exe CodeViewer::InitializeScriptEngine + 0xC3 codeview.cpp(398+0018 byte(s)) (0x04A4C0E0 0x0473E6EC 0x0473E480 0xffffffffcccccccc)

 

https://github.com/v...deview.cpp#L398

 

 

 

I just added a check around that. Can you see if this helps:

 

 

https://github.com/j...runs/3166896569

 

What was the cause?
What is the common factor?
As i was not able to reproduce it, i kinda wonder if i am lacking something that might later turn up to be an unexpected issue instead of a blessing later?
 

I notice a new file, VpinballX.pdb
also noticed VpinballX.exe is 64MB as opposed to 17MB

 

Have not tried it yet, as i had no issue with the build that everyone was crashing in

I have noticed that most tables load a little slower in 10.7 than in 10.6. It hasn't been an issue, I just chalk it up to the more features being added. But today I ran into a very long load. Surf'n Safari took so long to load I thought it had crashed. After a lot of back and forth testing with both 10.6 and 10.7 I came up with this data.  In 10.6 the table loads in 4 seconds.  In 10.7 the table loads in 1 minute 3 seconds. I tested this multiple times with the same results.

 

Does anyone have an explanation or a fix for this?  For now I am just using the alternate launcher in Popper to launch the table using 10.6.

 

Thanks

 

 

Check this thread back some pages
Someone else had that issue, with popper, quite a few revisions back, and i think the solution was the new launch arguments 10.7 will take
i think it had to do with the argument to limit threads created at launch.

 

Since i do not have PinUP Popper, i can not reproduce or test the issue you are having, but i think they fixed it by changing the launch parameters

[jpsalas]

I just put a debug build here if anyone wants to try..

 

https://github.com/j...runs/1068707264

 

Maybe we will get a little better stack trace.

 

Yes, 288 fixed the problem

Edited by jpsalas, 27 July 2021 - 01:04 AM.

[DuMichAuch]

 

I just added a check around that. Can you see if this helps:

 

 

https://github.com/j...runs/3166896569

 

Thank you. No chrashes wile opening a table.

[jsm174]

@DuMichAuch, @clyde, @jpsalas, Thanks for testing. I just committed that fix to the official repo, so the official artifacts would be at:

 

 

@wiesshund,  PR #19 added some script debugger enhancements. Maybe your machine has something installed that the others do not, or vice-versa. 

So you were never hitting the line that was causing the crash. https://github.com/vpinball/vpinball/commit/1c2376e8d9095941d29d8c2f9ccd6c6e200c9c42

 

As for the file size differences you noticed, the version they were testing was a special debug build in my fork which had all the extra information in it to show us the line number where the app was crashing.

 

You'll see the official artifacts are much smaller. 

[wiesshund]

 

@wiesshund,  PR #19 added some script debugger enhancements. Maybe your machine has something installed that the others do not, or vice-versa. 

So you were never hitting the line that was causing the crash. https://github.com/v...ccd6c6e200c9c42

 

 

 

Hmm
Well, if you find something that points at a dependency missing, or conflict with something
might be worth making note of

 

Of the 3 PC's i have it on

One is an old desktop
Windows 10 Pro, not sure if it has 21H1 update as it is not connected to the internet much.
Old hardware Athlon X2 250 and a Radeon 4850, VIA chipset audio, forgot which chipset

Legacy drivers on all hardware

Only other software beside the OS installed on it is VPX, and a couple old games

 

Laptop is Windows 10 Pro
It is always online, so it has all updates

Older Laptop
I5 3230m and only the on CPU Intel HD4000 for graphics, typical Intel IDT audio chip

Legacy Driver on audio and video

Has MS Office 2016 pro on it, VPX and an assortment of MAME and console emulators.
No system altering software though

 

 

Desktop is also Windows 10 Pro

AMD Ryzen 5, AMD RX580 x Graphics, HDMI Audio out and Realtech audio
And it also has all OS updates, and AMD driver/chipset updates 
It has 10,000 things installed on it.

 

Only thing consistent with them is they are all windows 10 pro

[clyde]

I am running windows 10 home edition,wonder If that could be the difference? Hmm

[jpsalas]

I'm using Windows 10 Pro 21H1, but the difference between mine and wiesshund computers is that I'm running an Nvidia Card (GTX 3080)

[clyde]

I am also using an Nvidia Card (GTX 1660 Super)

[TouchMaster97]

Just a question: I'm considering getting this Version of VPX (though not sure where), but I see that there are constant new Revisions. Does this Version of VPX come with some sort of Automatic Updater or something else to quickly update the program in case a new Revision comes in?

Edited by TouchMaster97, 27 July 2021 - 10:20 AM.

[jsm174]

The official releases (including beta pre-releases) are at:

 

https://github.com/v...inball/releases

 

 

Pre-releases get added there when @toxie and @fuzzel decide the code is ready for testing.

 

 

Many people in this thread are grabbing builds from the CI after each commit. While this is great that it helps find issues like above, it may not be the right choice for most users as you'll occasionally have issues like above.  

Edited by jsm174, 27 July 2021 - 11:56 AM.

[jsm174]

@DuMichAuch, @clyde, @jpsalas, @wiesshund - I don't think it is hardware related. 

 

When the scripting engine was getting initialized, it executes this chunk of code:

	// This can fail on some systems (I tested with wine 6.9 and this fails)
	// In that case, m_pProcessDebugManager will remain as nullptr
	CoCreateInstance(
		CLSID_ProcessDebugManager,
		0,
		CLSCTX_ALL,
		IID_IProcessDebugManager,
		(LPVOID*)&m_pProcessDebugManager
	);

So `m_pProcessDebugManager` was successfully set for @wiesshund. 

 

 

After Googling "Process Debug Manager", I found this article:

 

https://community.sp...prise-architect

 

 

Various software could have installed PDM.DLL  somewhere along the way.  So maybe @wiesshund you have that dll installed and the other's don't?