Graph control causes general protection fault when changing axis range (64bit)

Hello,
i wrote a 32bit application which i'm porting to 64bit now and stumbled across a problem with the graph control.
The Graph axsis are set to editable so that the user can change the axis range manually at runtime. The 32bit version has no problem here, but the 64bit version is crashing with a general protection fault at the source line with RunUserInterface().
Is there a compiler option/value or something i should be aware of when swithing to 64bit that should be set?
Or is this just a problem with the code of the control?
I tried it with LabWindows v13.0.2 (278) and v13.0.1, running under Windows 7 64bit.
Thanks.

Hello, danouj!
Thank you for reporting this issue to us!
However, information you provided is still unsufficient for us to determine the background of this bug.
Please also provide the following details, so that we can identify the issue and investigate whether a workaround for this problem exists:
From my understanding, when porting the application to 64-bit, you also performed some code changes to your CVI application. What are these code changes?
If you run the 64-bit version of your application without these code changes, does the application also crash?
What is the address of this crash?
It would also be very useful, if you can send us a copy of a minimal snapshot of your project application, in which the problem still occurs, for us to investigate the issue.
You can upload the sources at ftp://ftp.natinst.com/incoming
Best regards,
- Johannes

Similar Messages

General protection fault with alt-tabbing and pop ups

Hi,
This is probably a case of me doing something I shouldn't, but I figured out a while ago that I could get resizable modal pop ups by doing something like this:
SetPanelAttribute(panelHandle2, ATTR_SIZABLE, 0);
InstallPopup(panelHandle2);
SetPanelAttribute(panelHandle2, ATTR_SIZABLE, 1);
This is almost certainly a hack, but it seems to work and I get modal pop ups that the user can resize and maximize. However, the run-time engine does not seem to like this, and I am getting general protection faults when alt-tabbing between the main application window and the pop up (which both have taskbar buttons). The address where it crashes, at least on my computer, is 0x684F5E34 which is in cvirte.dll. This also seems to crash in the release configuation.
I've attached a small sample application. In order to see the problem, launch the application, open the Pop Up, and alt-tab a few times between the two windows. On my development PC this will generate a general protection fault after a few times (often the first time). I'm running CVI 2012SP1 on a Windows XP Pro PC.
Any insight would be helpful. Thanks.
Attachments:
Pop Up Test.zip ‏5 KB

What I don't understand is how you can switch between main and popup panel since the latter is displayed with InstallPopup: I never succeed in getting focus to the main panel when the popup is active, and it is not even listed among windows in alt-tab procedure. If I select the main window in taskbar, the popup immediately comes up and gets the focus.
Proud to use LW/CVI from 3.1 on.
My contributions to the Developer Zone Community
If I have helped you, why not giving me a kudos?

Savepanelstate general protection fault

I'm once in a while receiving a "General Protection Fault" when using SavePanelState in CVI 2010 running Windows XP.
This a program that runs 24/7, and that function call is probably executed 20k times a day, and runs on number of benches with the same configuration and no problems.
It sounds similiar to the problem that was found a while back for 64 bit applications, the CVI Runtime engine is 10.0.1.419.
Any updates that would correct this available? Any other ideas?
Terry

Hi Terry,
There was a bug in 64-bit for SavePanelState, but it was related to the creation of the save file. It could only occur for 64-bit applications and SavePanelState would always fail. So I don't think that is related to your issue.
Am I correct that this issue only occurs on 1 system? How often does it occur?
Are you able to reproduce this with the debugger attached? Is there an address attached to the fault?
Could you generate a crash dump and send it to us?
How To Create and Submit a Crash Dump to NI
National Instruments
Product Support Engineer

General Protection Fault in module WIN87EM.DLL while Running Windows XP Mode

I am attempting to run a legacy, 16-bit program, P3, on a Windows XP Mode Virtual Machine, hosted by a Windows 7 x64 box. When performing various activities in the program, I receive "P3 caused a General Protection Fault in module WIN87EM.DLL
at 0001:02C6." This occurs on multiple platforms (Core i5, Core 2 Duo).
Based on my research, I have found "The Windows 80x87 emulator library,
WIN87EM.DLL, works at the 16-bit-Windows level
to virtualize the coprocessor among multiple Windows-based applications that run inside the system VM."
It appears to me that, with the Virtual Machine using a Virtualized Processor (not related to the above reference to "virtualized coprocessor") as opposed to an emulated processor, the Virtual Math Coprocessor Device (VMCPD) and/or WIN87EM.DLL
generate an error from the virtualized processor (perhaps they do no recognize it?)
In addition "When the kernel loads an application, it checks to see if floating-point hardware is present. If a coprocessor is not present, the kernel uses the
fixup records to replace the actual floating-point instructions with
Interrupt 3x calls to emulation code in WIN87EM.DLL."
I have found some information that points to "hiding" the math coprocessor from the Virtual Machine, so it doesn't go to the processor. This apparently can be achieved by using a program called WinFloat which includes a tool called HIDE87, which is suppose
to hide the math coprocessor from the kernal.
The issue I'm running into is, to properly achieve this, HIDE87 must be loaded before Windows. I have tried a couple techniques, but I think I'm doing it wrong. I have tried adding the HIDE87 program to the System32 directory and then adding
HIDE87 (just before autocheck autochk *) to BootExecute under the HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\Session Manager
key.
For a little more info, when you run HIDE87 from the command prompt it reads "HIDE87: Int 11h hook installed (must reboot to remove)." The documentation says this is what needs to occur before Windows loads. You can find WinFloat here:
http://support.microsoft.com/kb/97265.
Any help or insight would be greatly appreciated.

This link helped point me to winfloat to solve the very same issue with WIN87EM.dll crashing under Windows XP on VMware 5 for an old legacy Windows 3.1. I thought I would mention that the original approach taken by tstock_sjc should work fine. First
off, winfloat is a self extracting executable and we really don't want to run winfloat, but rather hide87.com which is a tiny TSR that is inside the winfloat archive. When you run any DOS or Windows 3.1 executable, Windows XP first boots up (if one isn't
already running) a mini-MSDOS 5 environment called NTVDM.exe. Just as with the original MSDOS, when the shell loads, it loads files similar to config.sys and autoexec.bat. Of course, there are more restrictions on drivers and not everything
under MSDOS is emulated but nevertheless, it is a fairly good approximation of an MSDOS 5 environment. The issue that tstock_sjc had was NTVDM.exe looks for %WINDIR%\system32\autoexec.nt and %WINDIR%\system32\config.nt instead of autoexec.bat
and config.sys. If you extract hide87.com out of winfloat.exe and place it into the %WINDIR%\system32 directory and add a line like %SystemRoot%\system32\hide87.com to the end of autoexec.nt , NTVDM will indeed load this TSR before launching its emulated
copy of of Windows 3.1 (wowexec.exe). When the 16 bit application is finally launched after wowexec.exe, the coprocessor is then masked. It worked fine for me anyway (after applying changes and rebooting). Under process monitor I could clearly
see NTVDM.exe reading autoexec.nt and hide87.com and the WIN87EM.dll GP fault went away.

General protection fault windows 3.1

I am using a real time devices pc104 computer stack running windows 3.1, trying to install labwindows 5.1. program installs, but when I try to run I get
"CVI caused a general protection fault in module cvi.exe at 0001:07A4." Any one have an idea why? I have this program installed on an identical computer stack with no problems. I have also tried a third stack with the same result as the first.

Hi scottharris,
It seems as if this is a known issue of Windows 3.x. I've looked through our records, and it doesn't seem as if this particular behavior has been reported when opening CVI. So, unfortunately, I cannot give you a direct solution to this issue. You may want to take a look at some of the many online troubleshooting guides for General Protection Faults, like this one.
Regards,
Sara Lewandroski
Applications Engineer | National Instruments
http://www.ni.com/support

CVI2009 general protection fault in user interface library commands

I have a program which we have been using and updating since 1999. We have updated it in various versions of LabWindows/CVI and had no problems. We recently updated it and compiled it in CVI2009 and now we get general protection faults. The faults are occurring in commands from the user interface library and not always in the same command. Sometimes it will process the commands properly and sometimes not. PlotXY, PrintCtrl, SetMouseCursor, and DiscardPanel have each caused the faults. Any suggestions?

Hello -
My initial guess is that you are running into the issue described in this KB. What is the version of cvirte.dll in c:\windows\system32? If it is 9.1.0.427, you need to upgrade to version 9.1.0.428, which is linked in the KB.
Let us know if that does not resolve it for you!
NickB
National Instruments

FATAL RUN-TIME ERROR: The program has caused a 'General Protection' fault at 0x6846680A.

Dear all,
I have some troubles with LabWindows CVI. Sometimes (maybe 2 or 3 times by week) my application crash with General protection error in cvirte adress. This crash never happens in the same part of the code is always happens in random way. The problem is that in debug mode we don't have any information about where the problem occurs in cvirte. This bugs can happens with differents application from very simple application to very complex applications.
Do you have some ideas that how can I solve these problems?
Thanks

Hello J. Romero!
Thank you for the crash information you provided!
Hover you managed to collaborate with NI France more in order to determine the cause of this crash?
Unfortunately, the information you provided is still not sufficient for us to determine the exact cause and circumstance of this crash:
The provided event logs only contain relative addresses of the locations. An absolute crash address can better suggest the exact location. Please provide a screenshot of the crash window dialog, which contains the exact absolute address of this crash.
In your original post, you mentioned that you can reproduce the crash with a "very simple application". Is it possible for you to send us the application on our incoming FTP server, so that we can build and run the program on our side?
Do you have a suspicion for any particular libraries, components or UI controls that you are using, which reproduce the crash. E.g. "whenever I perform this action in a Numeric Control, the crash occurs...". This will further help us to narrow down on the problem!
If #2 is not possible, one other solution would be to send us a crash dump of the application. Please follow the steps in this Developer Community post, in order to generate and send us a crash dump: https://decibel.ni.com/content/docs/DOC-1496
Best regards!
- Johannes

RT DLL general protection fault associated with serial I/O

Hello,
I'm experiencing Generatl Protection Faults on my real-time PC running LabView Real Time 9.1.0. The faults occur when oppening/reading/writing from a serial rs-232/422 port, regardless of whether I use the Serial library or VISA library.
How do I debug this? Is there a stack dump stored somewhere on the RT system when a fault occurs?
I am using LabWindows/CVI 2009 with the corresponding RT module to do my development.
Regards,
ndo

Michael,
I was able to get a report of the crash just as the KB instructed.
(Although no report was generated when I ran my DLL in debug mode, the report was only generated when I deployed the DLL to the RT system and a crash occurred. )
I've attached the report here, can you tell me why it's crashing?
Regards,
ndo
Attachments:
lvrt_err_log.txt ‏51 KB

Collect causes segmentation fault when profiling L2 data cache misses on T1

I am attempting to profile L2 data cache misses on a SUN T1000 server, but collect causes a segmentation fault when I choose to profile L2 data cache misses.
Is it possible to profile L2 data misses on the T1 processor?
Thanks!
-John
A simple hello world application and the failed profiling:
%] collect -p +on -h ecdm hello
Segmentation fault
I compiled with
CC -xhwcprof -xdebugformat=dwarf -g -03 -o hello hello.c

No it is not possible to profile L2 cache misses -- the chip does not generate
interrupts on performance register 0 (only register 1, which is hard-wired
to instruction count).
That said, we should not generate a segmentation fault. Exactly
which version of collect are you using? We did patch Studio 11
to correctly handle the US-T1, and Studio 12 express also
correctly handles it.
Is the SEGV on the collect command, or on the target it launches?
Marty Itzkowitz,
Project Lead, Sun Studio Analyzer

Protection Fault related to printing in CS5

When I went to set printer properties in InDesign CS5, I got a message "Function address...caused a protection fault. Some or all property page(s) may not be displayed." Expanding on advice in a previous forum posting, I renamed my printer by going to Control Panel > Hardware and Sound > View Devices and Printers. I clicked on my printer icon and selected properties. The printer dialogue box came up and I renamed the printer in the "general" slot from "HP Photosmart D7500 series" to "HP Photosmart D7560" (though it probably could be any name). PRESTO! I am now able to print!. This was a new problem not encountered with a previous installation of CS5. It relates to CS5 and not other programs such as WORD. I don't need a response, since the problem is apparently solved, but I thought you'd like to know. Don't know why the problem arose in the first place.

Hi Terry,
Thanks for the reply.
I fixed the DAQ error by removing it, problem remained;
I commented out all the code in the timer, problem remained;
I changed timer to be 1 second, problem remained.
I did one more thing: try my project on CVI 8.0, and got the same GPF problem.
Looks to me it has something to do with line 2670 which has sprintf there. Because if I put another new line before this line like "i=i+1;", I could run through this line without error;
I also tried to replace "s" on line 2670 with another string var "ss", and got the same error.
Pls let me know. Thanks.
Jason

Adobe form Error "Memory protection fault in OLE container control"

Hi,
I am getting the error "Memory protection fault in OLE container control" while trying to open the layout tab of adobe forms.
I am getting this error for all the adobe forms.Please let me know the solution.
Thanks in advance.
Thanks,
Nuthan.

I can reproduce this error consistently.. This is how: Install Microsoft SQL Server 2008.
As soon as I remove SQL 2008 the error went away. Adobe LiveCycle Designer was invoked via OLE2 interface. Apparently SQL Server 2008 (as in my case or some other applications in your case) modified the way OLE2 behaves and therefore causes this error.

Why the 3D graph control shortcut menu is disabled?And how to set 3D graph to default view when running?

1).Controls have shortcut menu,but the 3d graph's shortcut menu cannt be able. why? how to make it able?
2).3D graph can be rotated free by dragging, and when it is not running,right click ,select default view, the graph go back to initial view. But when the VI run, the shortmenu is disabled, how to make it to defaul tview. the property node and method seem have none about default view

The 3D Graph control is actually an ActiveX control (in Windows), so it's not a "standard" LabVIEW control with the "standard" run-time popup menu. Consequently, to access most of the properties and actions you actually have to use the Property Node and Invoke Node functions in the Application Control palette:
Message Edited by smercurio_fc on 05-30-2008 09:55 AM
Attachments:
Example_VI_BD6.png ‏1 KB

When i use my TTY on my Iphone 4s the volume is really loud and i cant control it. How do i change this?

When i use TTY on my Iphone 4s the volume is really loud and i cant control it. How do i change this?

That is probably because you have one or more bookmarks with a long title.<br />
The width of a column in the Bookmarks menu list is in current Firefox versions automatically adjusted to fit the bookmark with the longest name.<br />
You can check the names of the bookmarks and make the longer ones shorter.<br />
<br />

General protection faul

HI,
I'm using the AmpPhaseSpectrum function and i don't know how it works exactly. For instance, i don't understand what i have to pass in the 'df' parametre (expected to be a double in the recall function panel but a pointer to double in function help), is this an output? I'm not really sure... :s
Moreover, when i run the code, at the line of this function, the code stops and it is said : The program has caused a 'Genereal protection' fault at... What does it mean?
Supplementary question with no link with above :
I want to compute the module of a complex number. I use two local variables : a and b. Respectively the real and imaginary part. Then I compute : sqrt(a*a+b*b). The error seems to be that a*a+b*b is not a positive number (error==33(0x21) EDOM domain error). Did i miss something?
Thanks,
_trent_

Hi,
In the df parameter you have to pass a double by its reference, this way you will change the value of df in the AmpPhaseSpectrum function, it's like an output yes.
May be you have the error 'General protection' because you didn't check the number of elements in the arrays you pass in parameters "Amplitude Spectrum" and "Phase Spectrum" that should be numberOfElements/2 at least as described in the help of each parameter.
c = sqrt(a*a+b*b); is correct, what are the values of a & b ?
Regards
Jean-Luc D. NI France
National Instruments France
#adMrkt{text-align: center;font-size:11px; font-weight: bold;} #adMrkt a {text-decoration: none;} #adMrkt a:hover{font-size: 9px;} #adMrkt a span{display: none;} #adMrkt a:hover span{display: block;}
Journées techniques : du contrôle d'instruments aux bancs de test
Du 4 au 13 juin : Grenoble - Pari...

RefreshGraph General Protection Error

Hi,
I would like to report 2 strange behaviors with Graphs that occur when I draw a bitmap image to a graph and then a rectangle.
1/ I would like to change the size of the rectangle by deleting all and then redraw all. I set the attribute ATTR_REFRESH_GRAPH to 0, then I delete all with the attribute VAL_DELAYED_DRAW, I draw all and then call RefreshGraph ().
The problem is that even with the attribute ATTR_REFRESH_GRAPH at 0 the graph is blinking when I change the size of the rectangle and redraw all.
2/ When I draw the bitmap with PlotBitmap and do a RefreshGraph before setting the image with SetCtrlBitmap I get a General Protection Error.
You will find attached a small projetc that illustrates these behaviors.
Thank you
Attachments:
Image On Graph issue.7z ‏1487 KB

Hi
I think the attached 7z file is corrupted.
Here is a new one.
By the way I made the same tests in CVI 2015 Beta and got nearly the same behaviors.
The only difference is that the programm stops without giving the reason, you don't get the message General protection error.
Best regards
Attachments:
Image On Graph issue.7z ‏1487 KB

Graph control causes general protection fault when changing axis range (64bit)

Similar Messages

Maybe you are looking for