I am trying to add a CUDA file to my existing C++ Visual Studio project. I have the CUDA 5.0 SDK installed, I have created a new .cu file and also set its Item Type to CUDA/C++ in the CUDA file properties. But it looks like it just does not compile giving errors that say that the compiler does not recognize the CUDA keywords. One of the errors I get is:
error C2065: 'threadIdx' : undeclared identifier
Any suggestions?
I found that the best way to do this is to perform the following in the existing CPU project
1) Build Dependencies -> Build Customisations
click the Cuda checkbox
2) Create a new simple CUDA project using the wizard (you anyway probably want to test your CUDA project builds ok firstly), load both projects into the IDE and then compare settings between the two projects, you will need to add
the following in project settings
$(CudaToolkitLibDir) to additional libraries settings (linker tab)
$(CudaToolkitIncludeDir) to additional include directories (c++ tab)
cudart.lib to additional dependencies (linker tab)
Then compare the CUDA tabs
I found that 32 bit had been pre-selected for the target machine architecture for some reason so I changed that to 64 bit.
After this I added a define _CUDA_CODE_COMPILE_ to preprocessor definitions to switch between CUDA or CPU compilation.
#ifdef _CUDA_CODE_COMPILE_
cudaCodeFunction();
#else
cpuCodeFunction();
#endif
Not ideal but necessary since there seem to be no defines set to indicate that NVCC is installed (other than performing a shell command!)
I can't go through it all at the moment but I think those steps are necessary:
Right click on your Project in the Project Explorer Build...(customization?) [my Version is German. I can't tell the keyword exactly but it's something about "Build...". You need to check "CUDA 5.0" here.
Set up the "Additional Include Directories" for Cuda in the "General" Tab of the Compiler options (Project Properties).
Add the cuda libfile to "Additional Dependencies" in the "Input" Tab of the Linker.
Mark the File as Cuda file (you've done that).
You have to select the right Compiler for the .cu files
Are you following any of the tutorial on how to setup it on visual studio ?
http://blog.norture.com/2012/10/gpu-parallel-programming-in-vs2012-with-nvidia-cuda/
Related
I'm modifying a simple example project (blinky) from the nRF SDK. I added a header file and a .c file in a new folder inside the project directory then added that path ./lib to the common preprocessor user include directories. I then included the header to main.c.
I can compile the new library on its own but when I build the whole project, I get Build failed error with no stated reason to follow up.
Here is an image of that:
Does anyone here know how to beat this?
I haven't used Segger Studio specifically, but it seems to be the CrossWorks IDE underneath.
In CrossWorks, you have to do the following:
Download & install all relevant libs from inside the IDE, under Tools -> Packages -> Install packages. Grab your specific target MCU as well as any specific boards or libraries you'll be using. In case some needed lib is missing here, you will get very weird errors.
In the project, click on the project name itself in "project explorer". Then in the properties window, check settings (this is a bit different in different versions of CrossWorks, might have to right click and pick properties in older versions). Under "user include directories" you should have something like this:
$(DeviceIncludePath)
$(TargetsDir)/NameOfMCU/Include
$(PackagesDir)/CMSIS_4/CMSIS/Include
$(ProjectDir)/NameOfDirectory
Where "NameOfMCU" is the name of the MCU family used, CMSIS should be there in case you are using any ARM, "NameOfDirectory" is the name of your custom directory (you can add several).
Also, get the debug build working first, before switching to release build.
I am running Visual Studio Community 2017 and Windows 10. I have a project for which I have added an include-file directory under Project > Configuration Properties > VC++ Directories > Include Directories, but Visual Studio simply ignores the directory and will not find the include files. It finds the files if I hard-code the directory into the #include statement. I checked the .vcxproj file and the directory shows up there. I just updated/corrected Visual Studio to the most recent version and have since re-booted, but the problem remains. Help, please!
First, ensure that MFC is enabled in project properties > Configuration Properties > General
use of mfc must be set to static or dynamic library.
then to ensure that cl.exe is using all of the include directories that you specify, you need to go to project properties > c/c++ > general and change suppress startup banner to no /nologo.this will give you the full cl command for each source file, showing exactly what visual studio is attempting to do with the code and configuration options that you give to it.
Additional ideas for troubleshooting this kind of issue:
When using vs2015 you can try to add '#include <somenoneexistingfile.h>' to the top of any .c/.cpp and right-click on it to open a context menu and select "Open Document <somenoneexistingfile.h>. This will open a dialog showing you the actual paths being used for including. Note: This specific approach doesn't work with vs2019 or later.
Besides this, you can check if:
you edited the actual configuration (Debug/Release) which you are starting, i.e. maybe you only edited the additional-include-path for the debug-build but are trying to compile a release-build.
you edited the actual platform (x64 vs x86) which you are starting, i.e. maybe you only configured the additional-include-path for x86 but are trying to compile the x64 platform.
This may be a very simple question but I haven't been able to figure it out so any help is appreciated.
I have a header that is located in a general folder because I want to use it in several projects for example:
C:\user\geninclude\program\header.h
I created a new empty project with a very simple main, in the main I put
#include <program/header.h>
I then went to the project properties and in VC++ in include directories added C:\user\geninclude\
but when I tried to build the program the program tells me it cannot find header.h because is not in the current directory or in the build system path.
I also tried in the project properties in C/C++ general Additional Include Directories adding C:\user\geninclude\ but still the same error.
I know is something simple I am missing, but I don't know what, I am very new to this just learning.
For reference I am using Visual Studio 2013.
Thank you in advance for your help.
UPDATE: Thank you all for your kind responses, I have tried everything you have told me (check release vs debug in both instances, change / for \ and <> for "", and double checking the header and still the system does not see it. It really is very weird. I'll keep trying...
Please check if your file is really an header file otherwise it won't appear on include.
What you can also do (as a workaround if you need that method fast) is to put your header file (or folder with header files) on the visual studio "include" folder. The path should look like this "C:\Program Files (x86)\Microsoft Visual Studio 14.0\VC\include"
PS: You should also check the properties configuration when you're adding the path to VC++ include directories. You could be adding the path to debug configuration and trying to run it in release mode.
You do indeed want
Project Properties -> Configuration Properties -> C/C++ -> Additional Include Directories
(or something close to that; I'm using VS 2008). Make sure the configuration you're editing in the top left (debug/release) matches the configuration you're building with (typically visible up top in the main window). So it sounds like you may have done it correctly; I'd double-check for the file's existence in that location. You could also try program\header.h instead of program/header.h. If none of those work, try adding C:\user\geninclude\program to the include directories (no \ at the end) and change it to #include "header.h". If that doesn't work either, you've almost surely got the header file in the wrong spot.
Another thing that can cause include files not being picked up is a difference between the platform set in your c++ project's Property Pages and your "Active Solution Platform" in configuration manager. Can just check if one is set to x64 and the other x86
check if you have specified the path correctly. for example I had written cpp instead of c++ and therefore suffered a lot and wasted like an hour searching here and there.
For Visual Studio 2019 users:
Project(P) > yours_project_name properties(P) > Platform Toolset Visual Studio 2019(V###)
Reasoning: You might download the project from Online and they used other version of Visual Studio as Platform.
Project(P) > yours_project_name properties(P) > Windows SDK Version ##.#(latest installed version).
Reasoning: You might download the project from Online and they used version SDK 8.0 while you have SDK 10.0
ntucvhw
In my Visual Studio 2010 project I have files with .mm file extension, that need to be compiled as normal C++ files. Is there a way to make a build rule for new extensions or something like that? In VS 2008 there were options for that, but they are gone in 2010.
For an individual file: Right click on the file > Properties > Configuration Properties - General > Item Type : C/C++ compiler.
In general for a project: How to: Select the Files to Build, Walkthrough: Using MSBuild, Walkthrough: Creating an MSBuild Project File from Scratch.
You can set how the file will be compiled for any unknown file type by:
Right click on the file in your solution explorer
Click on properties on the file
Go to the General tab
Set exclude from build to: No
Set Item type: to C/C++ compiler
in vs2008 you can set in project properties (simply select extensions that should be treated as C++ via menu)
in vs2010 and vs2012 I've never been able to configure this properly, I tried all these walkthroughs, tutorials and tricks. They simply do not work as expected and my *.mm files are not compiled.
This is an incredibly basic question, but how do I start a new CUDA project in Visual Studio 2008? I have found tons and tons of documentation about CUDA related matters, but nothing about how to start a new project. I am working with Windows 7 x64 Visual Studio 2008 C++. I would really like to find some sort of really really basic Hello World app to just get a basic program compiling and running.
Edit:
I tried your steps Tom. I setup a console app. I then deleted the default .cpp it drops in and copied over the three files from the template project just to have something to compile. When I compile that, template_gold.cpp complained about not having stdafx.h included, so i included that. Now the build fails with this:
1>------ Build started: Project: CUDASandbox, Configuration: Debug x64 ------
1>Compiling...
1>template_gold.cpp
1>Linking...
1>LIBCMT.lib(crt0.obj) : error LNK2019: unresolved external symbol main referenced in function __tmainCRTStartup
1>D:\Stuff\Programming\Visual Studio 2008\Projects\CUDASandbox\x64\Debug\CUDASandbox.exe : fatal error LNK1120: 1 unresolved externals
1>Build log was saved at "file://d:\Stuff\Programming\Visual Studio 2008\Projects\CUDASandbox\CUDASandbox\x64\Debug\BuildLog.htm"
1>CUDASandbox - 2 error(s), 0 warning(s)
========== Build: 0 succeeded, 1 failed, 0 up-to-date, 0 skipped ==========
NOTE With the release of version 3.2 of the CUDA Toolkit, NVIDIA now includes the rules file with the Toolkit as opposed to the SDK. Therefore I've split this answer into two halves, use the correct instructions for your version of the Toolkit.
NOTE These instructions are valid for Visual Studio 2005 and 2008. For Visual Studio 2010 see this answer.
CUDA TOOLKIT 3.2 and later
I recommend using the NvCudaRuntimeApi.rules file (or NvCudaDriverApi.rules if using the driver API) provided by NVIDIA, this is released with the toolkit and supports the latest compiler flags in a friendly manner. Personally I would advise against using the VS wizard, but only because I really don't think you need it.
The rules file (installed into the Program Files\Microsoft Visual Studio 9.0\VC\VCProjectDefaults directory) "teaches" Visual Studio how to compile and link any .cu files in your project into your application.
Create a new project using the standard MS wizards (e.g. an empty console project)
Implement your host (serial) code in .c or .cpp files
Implement your wrappers and kernels in .cu files
Add the NvCudaRuntimeApi.rules (right click on the project, Custom Build Rules, tick the relevant box), see note 1
Add the CUDA runtime library (right click on the project and choose Properties, then in Linker -> General add $(CUDA_PATH)\lib\$(PlatformName) to the Additional Library Directories and in Linker -> Input add cudart.lib to the Additional Dependencies), see notes [2] and [3]
Optionally add the CUDA include files to the search path, required if you include any CUDA files in your .cpp files (as opposed to .cu files) (right click on the project and choose Properties, then in C/C++ -> General add $(CUDA_PATH)\include to the Additional Include Directories), see note [3]
Then just build your project and the .cu files will be compiled to .obj and added to the link automatically
Some other tips:
Change the code generation to use statically loaded C runtime to match the CUDA runtime; right click on the project and choose Properties, then in C/C++ -> Code Generation change the Runtime Library to /MT (or /MTd for debug, in which case you will need to mirror this in Runtime API -> Host -> Runtime Library), see note [4]
Enable syntax highlighting using the usertype.dat file included with the SDK, see the readme.txt in <sdk_install_dir>\C\doc\syntax_highlighting\visual_studio_8
I'd also recommend enabling Intellisense support with the following registry entry (replace 9.0 with 8.0 for VS2005 instead of VS2008):
[HKEY_CURRENT_USER\Software\Microsoft\VisualStudio\9.0\Languages\Language Services\C/C++]
"NCB Default C/C++ Extensions"=".cpp;.cxx;.c;.cc;.h;.hh;.hxx;.hpp;.inl;.tlh;.tli;.cu;.cuh;.cl"
Incidentally I would advocate avoiding cutil if possible, instead roll your own checking. Cutil is not supported by NVIDIA, it's just used to try to keep the examples in the SDK focussed on the actual program and algorithm design and avoid repeating the same things in every example (e.g. command line parsing). If you write your own then you will have much better control and will know what is happening. For example, the cutilSafeCall wrapper calls exit() if the function fails - a real application (as opposed to a sample) should probably handle the failure more elegantly!
CUDA TOOLKIT 3.1 and earlier
I would use the Cuda.rules file provided by NVIDIA with the SDK, this is released alongside the toolkit and supports the latest compiler flags in a friendly manner. Personally I would advise against using the VS wizard, but only because I really don't think you need it.
The rules file (in the C\common directory of the SDK) "teaches" Visual Studio how to compile and link any .cu files in your project into your application.
Create a new project using the standard MS wizards (e.g. an empty console project)
Implement your host (serial) code in .c or .cpp files
Implement your wrappers and kernels in .cu files
Add the Cuda.rules (right click on the project, Custom Build Rules, browse for the rules file and ensure it is ticked)
Add the CUDA runtime library (right click on the project and choose Properties, then in Linker -> General add $(CUDA_LIB_PATH) to the Additional Library Directories and in Linker -> Input add cudart.lib to the Additional Dependencies), see note [2] below
Optionally add the CUDA include files to the search path, required if you include any CUDA files in your .cpp files (as opposed to .cu files) (right click on the project and choose Properties, then in C/C++ -> General add $(CUDA_INC_PATH) to the Additional Include Directories)
Then just build your project and the .cu files will be compiled to .obj and added to the link automatically
Some other tips:
Change the code generation to use statically loaded C runtime to match the CUDA runtime, right click on the project and choose Properties, then in C/C++ -> Code Generation change the Runtime Library to /MT (or /MTd for debug, in which case you will need to mirror this in CUDA Build Rule -> Hybrid CUDA/C++ Options), see note [4]
Enable syntax highlighting using the usertype.dat file included with the SDK, see the readme.txt in <sdk_install_dir>\C\doc\syntax_highlighting\visual_studio_8
I'd also recommend enabling Intellisense support with the following registry entry (replace 9.0 with 8.0 for VS2005 instead of VS2008):
[HKEY_CURRENT_USER\Software\Microsoft\VisualStudio\9.0\Languages\Language Services\C/C++]
"NCB Default C/C++ Extensions"=".cpp;.cxx;.c;.cc;.h;.hh;.hxx;.hpp;.inl;.tlh;.tli;.cu;.cuh;.cl"
Incidentally I would advocate avoiding cutil if possible, instead roll your own checking. Cutil is not supported by NVIDIA, it's just used to try to keep the examples in the SDK focussed on the actual program and algorithm design and avoid repeating the same things in every example (e.g. command line parsing). If you write your own then you will have much better control and will know what is happening. For example, the cutilSafeCall wrapper calls exit() if the function fails - a real application (as opposed to a sample) should probably handle the failure more elegantly!
NOTE
You can also use a Toolkit-version-specific rules fule e.g. NvCudaRuntimeApi.v3.2.rules. This means that instead of looking for the CUDA Toolkit in %CUDA_PATH% it will look in %CUDA_PATH_V3_2%, which in turn means that you can have multiple versions of the CUDA Toolkit installed on your system and different projects can target different versions. See also note [3].
The rules file cannot modify the C/C++ compilation and linker settings, since it is simply adding compilation settings for the CUDA code. Therefore you need to do this step manually. Remember to do it for all configurations!
If you want to stabilise on a specific CUDA Toolkit version then you should replace CUDA_PATH with CUDA_PATH_V3_2. See also note 1.
Having mismatched version of the C runtime can cause a variety of problems; in particular if you have any errors regarding LIBCMT (e.g. LNK4098: defaultlib 'LIBCMT' conflicts with use of other libs) or multiply defined symbols for standard library functions, then this should be your first suspect.
What a great question!! For all the CUDA documentation out there, this is something that I've always thought was an obvious omission... In fact, I'm really glad I found this post, because after using CUDA for quite a while, I still hadn't found an official, correct way to get VS to produce a CUDA program from scratch.
When I've needed to start a new CUDA program, I've always just copied and modified the "template" example from the SDK directory. This may not be exactly what you're looking for, because it doesn't start fresh, but it is a quick way to get a CUDA-capable project working in VS with all the correct project/file names...
Make a copy of the "template" example from the SDK, and rename the directory -- the only necessary contents in the directory are source code and VS .sln and .vcproj files
Rename both .sln and .vcproj files
Open the .sln file in a text editor, and rename the Project variable and .vcproj filename in the 3rd line of the file
Open the .vcproj file in a text editor, and rename the Name and RootNamespace variables in the first few lines of the file
Open the project with VS, and open the Property Pages (right click on the project name in the solution explorer pane, select "Properties")
Change the Output File name in the Property Pages (under Configuration Properties -> Linker -> General) ... Before I change the filename, I select "All Configurations" from the Configuration pull-down and "x64" from the Platform pull-down, since I'm on a 64-bit system
Change the Program Database File name in the Property Pages (under Configuration Properties -> Linker -> Debugging) ... Before I change the filename, I select "Debug" and "x64" in the pull-downs.
Install CUDA VS wizard. It will setup VS and add CUDA Project to the "new project" menu.
Make sure that you have x64 compiler installed (must be checked during VS install).
Check if you have x64 libs, includes, nvcc dir and in the search path.
Create new project using CUDA template.
Change project type to x64 and CUDA setting to Native (if you have nv cuda-enabled card) or emulation otherwise.
The template will create custom build rules that compile .cu files with nvcc and other files with default compiler.
if, vs is trying to compile .cu files with C/C++ compiler, click on that file in solution explorer and disable compilation for that files (red dot on file's icon)
Additional info about installing CUDA wizard on VS2008 can be found here and here
[edit]
If you don't want to use wizard you have to setup CUDA lib/include/nvcc paths manually and add custom build rules to each new CUDA program. For additional info how to do it take a look at Tom's Answer.
You may want to take a look at this guide: http://www.programmerfish.com/how-to-run-cuda-on-visual-studio-2008-vs08/