My application which is built on VC++ 2008 , when run on a freshly installed XP shows an error that asks dotnet.fx to be installed. And also i faced similar problem when i tried to run it on another xp system after i installed that, it started to run.
Is there any way that i can make my application independent of all these additional installs ?
Are you sure you didn't compile managed C++ (targetting CLR) instead of real C++. In the former case dotnetfx is the .NET runtime you absolutely need and not some optional addon. In the latter case you need only the VC_redist in either x86 or 64bit variant. That installs the C and C++ dlls only. These are side-by-side assemblies and must match the version you compiled for.
Related
I am trying to build a few different 32-bit C++ applications with Visual Studio 2015 that use CEF. To use CEF, I am currently acquiring the CEF prebuilts from Spotify. The dll wrapper for CEF is built using Visual Studio 2015 with some modifications to its CMake files to force it to build with MD and MDd mode regardless of other settings. This was sufficient to make these C++ applications run on some machines. Any machine on which Visual Studio 2015 is installed on before anything else they can run on, however some machines seem to exist in a state such that the program will produce an error on starting when lacking the MSVC 2015 runtime (as expected), but when adding the MSVC 2015 runtime the program simply crashes; however, it only crashes after CEF is used. These programs work fine when they don't link to libcef.dll and don't include the browser functionality.
Upon investigating, I found that libcef.dll, as built by spotify, links to MSVCP110_WIN.DLL, which is from the Visual C++ 2012 Redistributable package. Naturally, the application I am building links to MSVCP140.DLL, which is from the Visual C++ 2015 Redistributable package. This means that the application ultimately links to two runtimes simultaneously. I do not know if this is an issue, but so far it seems to be my best lead. Installing the Visual C++ 2012 Redistributable does not change the outcome and it continues to crash when CEF is used.
This issue has been witnessed on both Windows 7 and Windows 10 and the application works without CEF on both of those operating systems as well, so the operating system is not likely to be the cause of the failure on these systems specifically.
Has anyone else encountered this issue and does anyone know a workaround? Also, does anybody know if it is okay to mix these two runtimes and what the limitations are? It seems that the installation history of a given machine affects the success of running the application, so any hints into what combination of things leads to this failure would be helpful as well.
You have some options:
Use the lastest version of VS that will allow a selection of Platform Toolset that matches libdef.dll. For example, VS2013 might allow the selection of the 2012 CRT.
Or convince Spotify to rebuild libcef.dll such that it matches your version of CRT
Or convince Spotify to not release libraries that depend on the CRT (yes that's probably a bit of work).
Or make a small app built against the older CRT, this app can then successfully use libcef.dll. Then you get to use any IPC technique so that your main VS2015 app can talk to this wrapper. Running out-of-process is one way to segregate the unruly third party libraries.
EDIT:
This is open source? Well good news, you can fix this yourself, built the CEF against your favorite version of VS.
I have written an application that uses CR for Visual Studio. It deploys and runs fine on 32 bit systems. I want it to work on 64 bit systems too, so I ran CRRuntime_64bit_13_0_7.msi to install the redistributable.
When I select my prerequisites in VS 2012, should I see an entry for CR 64 bit? I don't.
In VS, I set the target CPU to x64 and publish (Clickonce). When I try to install it on my Win 7 64 bit system, I get an error:
Unable to install or run the application. The application requires that assembly CrystalDecisions.ReportAppServer.CommonObjectMode Version 13.0.2000.0 be installed in the Global Assembly Cache (GAC) first.
How can I deploy the 64 bit CR runtime with my application?
Thanks
Maybe there's a bug with the Service Pack 7 version of the files? Have you tried with the Service Pack 8 version that was recently released?
SAP Crystal Reports, Developer version for Microsoft Visual Studio
Is your development environment 32-bit or 64-bit? Because that might have a bit to do with it also. Installing and deploying Crystal has always been a pain (I've been working with it since version 7.0) and to introduce the complexity of 32/64-bit makes it even more fun!
Another thing to point out on that link was that you have to run the "Install executable" to get it to integrate into VS properly. When I did so, it asked me if I wanted to install the 64-bit runtime as my PC was 64-bit, did you get to this point and if so, which version did you choose?
In the app we're working on, we have a routine that inspects all assemblies in the dependencies folder and auto-loads classes with a certain attribute, but when I went to run the app targeting a 64-bit CPU (or even Any CPU) it would give me exceptions when trying to load the Crystal Reports assemblies on start-up, but it would work fine when targeting a 32-bit CPU.
Which leads me to think that perhaps they either don't have their 32/64-bitness sorted properly, or their installer isn't doing what it says it is.
I am using JNI to interface to a CAN driver I wrote using MSVC++ 2012. Everything compiles and runs fine - but only on my computer. Whenever I try to run on any other computer, I get the JNI "UnsatisfiedLinkError" - can't find dependent libraries. I've implemented JNI before, and typically this issue is resolved by simply installing the Microsoft Visual C++ Redistributable Package before running my program. The Redistributable does not solve my issue, however, installing the entire MSVC++ Express 2012 IDE on the computer in question seems to make everything work just fine. Thus, I assume this means that there is some dependency that gets installed and added to PATH when the IDE is installed but not when the Redistributable is installed. Just for verification, I uninstalled the IDE and my JNI driver failed to load once again. The primary difference between this and my other implementations of JNI is that I use Windows.h since it is required for the CAN API I use in my driver. Any ideas on any libraries installed with the IDE but not the Redistributable and whose prototypes are included with Windows.h? (Note: I am aware that Windows.h includes a number of headers itself but I imagine the other criteria make the issue a bit more specific.)
First guess: You are installing a debug build intead of a release build. Debug builds depend on debug versions of the CRT, which cannnot be redistributed.
Check that you are using the right "Microsoft Visual C++ Redistributable Package". There are 2012, 2012 Update 1 and 2012 Update 2 (as well as many for 2010, 2008, ...).
Assuming your dependencies are load-time DLLs, you can use Depends to find out what's missing. (Note: you'll typically see some delay-load DLLs that are "missing". Ignore thoses; See the FAQ.)
Of course, once you find out the name and path of the missing DLL, you'll have to find out what redist package installs it.
I'm trying to investigate why would my app, written in VC 2008, crash on bare Windows XP with c0000005. It works great on any windows vista and 7. On XP it would crash with no additional info. Below are the details and the result of a few days worth of my headaches, eliminating one reason after another. Here's the details:
I have Win7 ultimate with XP Mode with iexplore 6 and presumably no patches whatsoever.
I install my program and nothing else, since it is supposed to be install-and-run thing with no additional packages needed.
My app has compiled-in VC2008 runtime (using /MT, as opposed to /MD compiler switch)
I'm using firebird embedded, which needs VC2005 (edited, i wrote VC2008 before) libs for itself anyway, so i put them in my app's working dir
when i install VC++ 2008 on this XP Mode machine, it doesn't crash anymore, so debugging this way is impossible
I've narrowed this error's occurence down to the VC2005 runtime - right after i install it my app stops crashing. But i don't know what could be using it's components. When running the release version under VC2008 IDE, nothing shows any vc2005 libs being loaded (weird, if i say so myself).
My app's linker's dependencies lists these (i added only the first two, the rest was there courtesy of VisualStudio): jpeg.lib, gdiplus.lib, kernel32.lib, user32.lib, gdi32.lib, winspool.lib, comdlg32.lib, advapi32.lib, shell32.lib, ole32.lib, oleaut32.lib, uuid.lib, odbc32.lib, odbccp32.lib, comctl32.lib, %(AdditionalDependencies)
I am also using IBPP firebird interface, which loads these dynamically, so they are present in my app's dir (they need msvc?80.dlls):
fbembed.dll, ib_util.dll, icudt30.dll, icuin30.dll, icuuc30.dll, msvcp80.dll, msvcr80.dll
Please advise why would vc2005 runtime be a remedy for my app to not crash on a bare Win XP. Or how could i debug that one without installing vc 2008 on target machine. I'm starting to blame windows xp's components. Ultimately i want my app to not require any additional packages, such as VC2008 or 2005, especially when no component of my app requires the latter.
I am also using IBPP firebird interface ... msvcp80.dll, msvcr80.dll
The answer is in your question, these DLLs are VS2005 runtime support DLLs. They cannot be stored in your exe directory, they must be registered in the side-by-side cache. This does not typically cause an AV, but it isn't impossible if the code doesn't check the return value of LoadLibrary(). You can download an installer for them from Microsoft. It doesn't otherwise have anything to do with Windows XP, you just happened to try to run this program on machines that already had the DLLs installed. Very common on a dev machine for example.
If you don't want a dependency on them then you'll have to rebuild those DLLs that require it. Which is a good idea, having a dependency on more than one version of the CRT is pretty unhealthy.
If you are wondering why you are having to deal with this: it was Microsoft's attempt to put DLL Hell in the developer's lap instead of the user's. They've since withdrawn this, VS2010 again makes it a user problem. That was a pretty big party in Bangalore, I'd imagine.
If I built my app against D3DX June2007_d3dx9_34 and the target system has a newer version Nov2007_d3dx9_36 should that be a problem?
I distribute D3DX DLLs using MS' redist-installer tool but I noticed one one 'clean' PC (no D3DX previously installed) I got an error about missing D3DX DLL. Before trying to figure out if the installer itself is not working, I wondered if maybe supplying a too-new version of the DLL is the problem - will the app be looking for a DLL with the exact name?
D3DX libraries are not backwards-compatible, and the API may change between versions. So the target system must have the same version of D3DX that you build your application with - a newer version won't work.
If you use Dependency Walker on your application, it should show you that there's a dependency on a DLL with the exact version number.
The redistributable in the DirectX SDK includes the whole history of the D3DX dlls. Installing the latest version of the DirectX redistributable should have your system end up with every version of these dlls. It is possible and allowed for applications to only install the files they need by deleting cab files out of the redistributable's directory structure, in order to reduce the size of the redistributable components, and this is how you can end up with having 43 of the dll but not 37 for instance.
If you remove the dependency of D3DX and D3DCOMPILER dll's from your application you can ensure the application will work as the d3d9 d3d10 and d3d11 core dlls are version-agnostic and are also tied to the windows service pack releases (i.e. D3D11 for Vista comes from Vista SP2 or Win7 stock installs for example).
Another possible way to remove the dependency and let the app startup is to make the dll a delay load module, so it is only loaded when needed (such as compiling shaders on a development environment). As long as you are sure the shaders exist on end-user machines properly it won't need to compile shaders and won't crash. Calling LoadLibrary and invoking the functions manually would be a safer choice with a way better error message of your choosing, but is a lot more work.