Error while Compiling on Win10

[jakobm789]

Hi Guys,

i get an error at following step from the installation-guide:

_After you execute cmake .. -DBUILD_SHAREDLIBS=TRUE -G "Visual Studio 14 2015 Win64", you can find a project “thundersvm.sln“ under build directory. You can double click it to open it with Visual Studio. Then you should build the solution inside Visual Studio. You can refer to this this link. After you build the project, you should be able to use the python interface.

Ive tried multiple VS Versions (2019 and 2022) and I have installed the Win8.1 SDK. I get the following Error:

Die Erstellung wurde um 15:00 gestartet...
1>------ Erstellen gestartet: Projekt: ZERO_CHECK, Konfiguration: Debug x64 ------
1>Checking Build System
2>------ Erstellen gestartet: Projekt: thundersvm, Konfiguration: Debug x64 ------
2>Building NVCC (Device) object src/thundersvm/CMakeFiles/thundersvm.dir/kernel/Debug/thundersvm_generated_kernelmatrix_kernel.cu.obj
2>kernelmatrix_kernel.cu
2>C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.8\include\cub/device/dispatch/dispatch_segmented_sort.cuh(338): error : invalid combination of type specifiers
2>
2>C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.8\include\cub/device/dispatch/dispatch_segmented_sort.cuh(338): error : expected an identifier
2>
2>C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.8\include\cub/device/dispatch/dispatch_segmented_sort.cuh(379): error : expected a member name
2>
2>3 errors detected in the compilation of "C:/SVM/thundersvm/src/thundersvm/kernel/kernelmatrix_kernel.cu".

Maybe someone had the same Issue and knows how to fix this.

Thank you for your help!

[thegreatesthoneybee]

Use latest cuda version, today I solved this problem using the latest cuda 12.5 version.