Building issues [no CUDA toolset found] #103

ph57fe · 2020-04-22T14:13:10Z

I have building issues with the following system settings:

Windows 10
Visual Studio 2019 (Community Edition) Version 16.5.4
cmake 3.17.1
python 3.7
CUDA 10.1

Even though CUDA is installed, but cmake is not able to locate it. The VCTargets is not created in the VC folder:
C:/Program Files (x86)/Microsoft Visual Studio/2017/Community/Common7/IDE/VC/

Ankit@DESKTOP-KVF0V7L MINGW64 ~/localgitrepository/mitsuba2 (master)
$ cmake -G "Visual Studio 16 2019" -A x64
CMake Warning:
  No source or binary directory provided.  Both will be assumed to be the
  same as the current working directory, but note that this warning will
  become a fatal error in future CMake releases.


-- Selecting Windows SDK version 10.0.18362.0 to target Windows 10.0.18363.
-- The C compiler identification is MSVC 19.25.28614.0
-- The CXX compiler identification is MSVC 19.25.28614.0
-- Check for working C compiler: C:/Program Files (x86)/Microsoft Visual Studio/2019/Community/VC/Tools/MSVC/14.25.28610/bin/Hostx64/x64/cl.exe
-- Check for working C compiler: C:/Program Files (x86)/Microsoft Visual Studio/2019/Community/VC/Tools/MSVC/14.25.28610/bin/Hostx64/x64/cl.exe - works
-- Detecting C compiler ABI info
-- Detecting C compiler ABI info - done
-- Detecting C compile features
-- Detecting C compile features - done
-- Check for working CXX compiler: C:/Program Files (x86)/Microsoft Visual Studio/2019/Community/VC/Tools/MSVC/14.25.28610/bin/Hostx64/x64/cl.exe
-- Check for working CXX compiler: C:/Program Files (x86)/Microsoft Visual Studio/2019/Community/VC/Tools/MSVC/14.25.28610/bin/Hostx64/x64/cl.exe - works
-- Detecting CXX compiler ABI info
-- Detecting CXX compiler ABI info - done
-- Detecting CXX compile features
-- Detecting CXX compile features - done
-- Found PythonInterp: C:/Program Files (x86)/Microsoft Visual Studio/Shared/Python37_64/python.exe (found version "3.7.5")
-- Found PythonLibs: C:/Program Files (x86)/Microsoft Visual Studio/Shared/Python37_64/libs/Python37.lib
-- Building the following variants of Mitsuba:
--  * scalar_rgb
--  * scalar_spectral
--  * gpu_rgb
--  * gpu_autodiff_rgb
-- Mitsuba: building the Python plugin.
CMake Error at C:/Program Files/CMake/share/cmake-3.17/Modules/CMakeDetermineCompilerId.cmake:367 (message):
  No CUDA toolset found.
Call Stack (most recent call first):
  C:/Program Files/CMake/share/cmake-3.17/Modules/CMakeDetermineCompilerId.cmake:32 (CMAKE_DETERMINE_COMPILER_ID_BUILD)
  C:/Program Files/CMake/share/cmake-3.17/Modules/CMakeDetermineCUDACompiler.cmake:72 (CMAKE_DETERMINE_COMPILER_ID)
  ext/enoki/CMakeLists.txt:20 (enable_language)

-- Configuring incomplete, errors occurred!
See also "C:/Users/Ankit/localgitrepository/mitsuba2/CMakeFiles/CMakeOutput.log".

Do you have an idea? Thanks!

The text was updated successfully, but these errors were encountered:

Speierers · 2020-04-23T07:13:45Z

Hi,

is nvcc in your PATH?

E.g. if you try to run:

nvcc --version

ph57fe · 2020-04-23T12:40:01Z

This is what I get with nvcc --version :

Ankit@DESKTOP-KVF0V7L MINGW64 ~/localgitrepository3/mitsuba2 (master)
$ nvcc --version
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2019 NVIDIA Corporation
Built on Fri_Feb__8_19:08:26_Pacific_Standard_Time_2019
Cuda compilation tools, release 10.1, V10.1.105

Got the issue solved by manually copying the files from C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v10.1\extras\visual_studio_integration\MSBuildExtensions to C:\Program Files (x86)\Microsoft Visual Studio\2019\Community\MSBuild\Microsoft\VC\v160\BuildCustomizations.

Speierers · 2020-04-23T13:48:50Z

Looking at the environment variables in my system. Do you have something similar?

ph57fe · 2020-04-23T14:23:45Z

Yes, I have these.

Speierers · 2020-04-24T05:58:16Z

Could you try to pull latest master and update the submodule. We fixed something in enoki recently which might be related to this.

git pull --rebase origin master
git submodule update --recursive

hjoonpark · 2020-04-25T04:57:05Z

Is my problem similar to this?
I get this CMAKE configure error:

CMake Warning:
No source or binary directory provided. Both will be assumed to be the
same as the current working directory, but note that this warning will
become a fatal error in future CMake releases.

-- Building the following variants of Mitsuba:
-- * scalar_rgb
-- * gpu_autodiff_rgb
-- Mitsuba: building the Python plugin.
CMake Error at C:/Users/hjoon/Documents/cmake-3.16.0-win64-x64/share/cmake-3.16/Modules/CMakeDetermineCompilerId.cmake:351 (message):
No CUDA toolset found.
Call Stack (most recent call first):
C:/Users/hjoon/Documents/cmake-3.16.0-win64-x64/share/cmake-3.16/Modules/CMakeDetermineCompilerId.cmake:32 (CMAKE_DETERMINE_COMPILER_ID_BUILD)
C:/Users/hjoon/Documents/cmake-3.16.0-win64-x64/share/cmake-3.16/Modules/CMakeDetermineCUDACompiler.cmake:72 (CMAKE_DETERMINE_COMPILER_ID)
ext/enoki/CMakeLists.txt:28 (enable_language)

-- Configuring incomplete, errors occurred!

Here are what I've done so far.
I pulled the latest master using:

git pull --rebase origin master
git submodule update --recursive

I checked nvcc is installed:

nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2019 NVIDIA Corporation
Built on Sun_Jul_28_19:12:52_Pacific_Daylight_Time_2019
Cuda compilation tools, release 10.1, V10.1.243

and I set the environment variables:

CMAKE_CUDA_COMPILER = C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v10.1\bin\nvcc
CUDA_PATH = C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v10.1
CUDA_PATH_V10_1 = C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v10.1

I also followed this:

Got the issue solved by manually copying the files from C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v10.1\extras\visual_studio_integration\MSBuildExtensions to C:\Program Files (x86)\Microsoft Visual Studio\2019\Community\MSBuild\Microsoft\VC\v160\BuildCustomizations.

but still not resolved..

Speierers · 2020-04-27T06:20:58Z

Would you be able to update CUDA on your system? (e.g. v10.2)

Not sure this would solve your problem, but it is worth a try!

ph57fe · 2020-04-28T17:19:39Z

Hi @Speierers

Thanks, I got all the previous issues resolved by updating to CUDA 10.2. I'm facing now facing an error when integrating with PyTorch.

I have the following:
Python 3.7.6
Pytorch 1.5.0
conda 4.8.0
cudatoolkit 10.2.89
torchvision 0.6.0

I am facing the following error.

Traceback (most recent call last):
  File "C:/Users/Ankit/localgitrepository/mitsuba2/dist/scene_render.py", line 13, in <module>
    import torch
  File "C:\Users\Ankit\Anaconda3\lib\site-packages\torch\__init__.py", line 81, in <module>
    ctypes.CDLL(dll)
  File "C:\Users\Ankit\Anaconda3\lib\ctypes\__init__.py", line 364, in __init__
    self._handle = _dlopen(self._name, mode)
OSError: [WinError 127] The specified procedure could not be found

Do you have an idea? Thanks!

Speierers · 2020-04-29T05:45:56Z

This seems to be related to Pytorch and not Mitsuba:
See this: pytorch/pytorch#37209

KnightYao · 2021-01-06T11:48:44Z

mark

Theo-Wu · 2021-07-21T06:20:58Z

Is my problem similar to this?
I get this CMAKE configure error:

CMake Warning:
No source or binary directory provided. Both will be assumed to be the
same as the current working directory, but note that this warning will
become a fatal error in future CMake releases.
-- Building the following variants of Mitsuba:
-- * scalar_rgb
-- * gpu_autodiff_rgb
-- Mitsuba: building the Python plugin.
CMake Error at C:/Users/hjoon/Documents/cmake-3.16.0-win64-x64/share/cmake-3.16/Modules/CMakeDetermineCompilerId.cmake:351 (message):
No CUDA toolset found.
Call Stack (most recent call first):
C:/Users/hjoon/Documents/cmake-3.16.0-win64-x64/share/cmake-3.16/Modules/CMakeDetermineCompilerId.cmake:32 (CMAKE_DETERMINE_COMPILER_ID_BUILD)
C:/Users/hjoon/Documents/cmake-3.16.0-win64-x64/share/cmake-3.16/Modules/CMakeDetermineCUDACompiler.cmake:72 (CMAKE_DETERMINE_COMPILER_ID)
ext/enoki/CMakeLists.txt:28 (enable_language)
-- Configuring incomplete, errors occurred!

Here are what I've done so far.
I pulled the latest master using:

git pull --rebase origin master
git submodule update --recursive

I checked nvcc is installed:

nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2019 NVIDIA Corporation
Built on Sun_Jul_28_19:12:52_Pacific_Daylight_Time_2019
Cuda compilation tools, release 10.1, V10.1.243

and I set the environment variables:

CMAKE_CUDA_COMPILER = C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v10.1\bin\nvcc
CUDA_PATH = C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v10.1
CUDA_PATH_V10_1 = C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v10.1

I also followed this:

Got the issue solved by manually copying the files from C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v10.1\extras\visual_studio_integration\MSBuildExtensions to C:\Program Files (x86)\Microsoft Visual Studio\2019\Community\MSBuild\Microsoft\VC\v160\BuildCustomizations.

but still not resolved..

I am using CUDA 10.2 but get the same problem too. Can you help me with this?

Solved by reinstalling CUDA 10.2. Then deleted CmakeFiles folder and CmakeCache.txt, and built successfully.

Chenguoz · 2022-01-23T13:06:28Z

60\BuildCustomizations。

Thanks, I tossed and turned all day until I saw your reply and I solved the problem!

giodude12 · 2022-05-29T15:16:28Z

Hey, i'm still getting the issue. My path seems right and i have the files manually copied into visua studio. Does anybody have a fix?

SHPEYOLO · 2022-06-18T07:52:15Z

Hey, i'm getting the issue.

nvcc fatal : Could not set up the environment for Microsoft Visual Studio using 'D:/VS2019/Community/VC/Tools/MSVC/14.29.30133/bin/HostX64/x64/../../../
../../../../VC/Auxiliary/Build/vcvars64.bat'
D:\VS2019\Community\MSBuild\Microsoft\VC\v160\BuildCustomizations\CUDA 11.6.targets(879,9): error MSB3721: 命令“"G:\CUDA_manger\CUDA11.6\CUDA2\bin\nvcc.exe" -
dlink -o ngp.dir\RelWithDebInfo\ngp.device-link.obj -Xcompiler "/EHsc /W1 /nologo /O2 /FdF:\Tutorial\ngp\instant-ngp\build\RelWithDebInfo\ngp.pdb /Zi /MD /
GR" -LG:\CUDA_manger\CUDA11.6\CUDA2\bin/crt -LG:\CUDA_manger\CUDA11.6\CUDA2\lib\x64 opengl32.lib dependencies\glfw\src\glfw_objects.dir\RelWithDebInfo\conte
xt.obj dependencies\glfw\src\glfw_objects.dir\RelWithDebInfo\init.obj dependencies\glfw\src\glfw_objects.dir\RelWithDebInfo\input.obj dependencies\glfw\src
glfw_objects.dir\RelWithDebInfo\monitor.obj dependencies\glfw\src\glfw_objects.dir\RelWithDebInfo\vulkan.obj dependencies\glfw\src\glfw_objects.dir\RelWithD
ebInfo\window.obj dependencies\glfw\src\glfw_objects.dir\RelWithDebInfo\win32_init.obj dependencies\glfw\src\glfw_objects.dir\RelWithDebInfo\win32_joystick.
obj dependencies\glfw\src\glfw_objects.dir\RelWithDebInfo\win32_monitor.obj dependencies\glfw\src\glfw_objects.dir\RelWithDebInfo\win32_time.obj dependencie
s\glfw\src\glfw_objects.dir\RelWithDebInfo\win32_thread.obj dependencies\glfw\src\glfw_objects.dir\RelWithDebInfo\win32_window.obj dependencies\glfw\src\glf
w_objects.dir\RelWithDebInfo\wgl_context.obj dependencies\glfw\src\glfw_objects.dir\RelWithDebInfo\egl_context.obj dependencies\glfw\src\glfw_objects.dir\Re
lWithDebInfo\osmesa_context.obj dependencies\glfw\src\glfw_objects.dir\RelWithDebInfo\glfw.res "dependencies\tiny-cuda-nn\src\RelWithDebInfo\tiny-cuda-nn.li
b" cuda.lib cublas.lib cudadevrt.lib cudart_static.lib kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut
32.lib uuid.lib odbc32.lib odbccp32.lib cudart.lib -forward-unknown-to-host-compiler -Wno-deprecated-gpu-targets -gencode=arch=compute_86,code=compute_86 -
gencode=arch=compute_86,code=sm_86 --machine 64 ngp.dir\RelWithDebInfo\camera_path.obj ngp.dir\RelWithDebInfo\common_device.obj ngp.dir\RelWithDebInfo\marc
hing_cubes.obj ngp.dir\RelWithDebInfo\nerf_loader.obj ngp.dir\RelWithDebInfo\render_buffer.obj ngp.dir\RelWithDebInfo\testbed.obj ngp.dir\RelWithDebInfo\tes
tbed_image.obj ngp.dir\RelWithDebInfo\testbed_nerf.obj ngp.dir\RelWithDebInfo\testbed_sdf.obj ngp.dir\RelWithDebInfo\testbed_volume.obj ngp.dir\RelWithDebIn
fo\tinyexr_wrapper.obj ngp.dir\RelWithDebInfo\triangle_bvh.obj”已退出，返回代码为 1。 [F:\Tutorial\ngp\instant-ngp\build\ngp.vcxproj]

merlinND · 2022-06-18T09:46:00Z

Hi @SHPEYOLO,

Your error log seems to refer to instant-ngp rather than Mitsuba, did you post in the right repo?

Shy-fawn · 2022-11-12T12:41:34Z

您好!!!!這是我以下碰到的問題:
我需要在命令提示字元輸入:
cmake . -B build
cmake --build build --config RelWithDebInfo -j 16

但是一直顯示這樣的錯誤:
D:\Tutorial\ngp\instant-ngp>cmake . -B build
-- Selecting Windows SDK version 10.0.19041.0 to target Windows 10.0.22000.
-- Automatic GPU detection failed. Building for common architectures.
-- Targeting GPU architectures: 53;60;61;70;75;80;86
CMake Warning at dependencies/tiny-cuda-nn/CMakeLists.txt:140 (message):
Fully fused MLPs do not support GPU architectures of 70 or less. Falling
back to CUTLASS MLPs. Remove GPU architectures 70 and lower to allow
maximum performance

CMake Error at dependencies/tiny-cuda-nn/CMakeLists.txt:188 (add_subdirectory):
The source directory

D:/Tutorial/ngp/instant-ngp/dependencies/tiny-cuda-nn/dependencies/fmt

does not contain a CMakeLists.txt file.

-- Could NOT find Vulkan (missing: Vulkan_LIBRARY Vulkan_INCLUDE_DIR) (found version "")
CMake Warning at CMakeLists.txt:118 (message):
Vulkan was not found. Neural graphics primitives will still compile and
run correctly, but DLSS will not be supported.

-- Using Win32 for window creation
-- OptiX_INSTALL_DIR value: C:\ProgramData\NVIDIA Corporation\OptiX SDK 7.6.0
-- pybind11 v2.7.1
-- Configuring incomplete, errors occurred!
See also "D:/Tutorial/ngp/instant-ngp/build/CMakeFiles/CMakeOutput.log".
See also "D:/Tutorial/ngp/instant-ngp/build/CMakeFiles/CMakeError.log".

D:\Tutorial\ngp\instant-ngp>cmake --build build --config RelWithDebInfo -j 16
Microsoft (R) Build Engine for .NET Framework 16.11.2+f32259642 版
Copyright (C) Microsoft Corporation. 著作權所有，並保留一切權利。

MSBUILD : error MSB1009: 專案檔不存在。
參數: ALL_BUILD.vcxproj

請問我是有什麼沒安裝好，還是此步驟其實是可以忽略的?

merlinND · 2022-11-12T17:33:51Z

Hello @Shy-fawn,

Same as above: your error log seems to refer to instant-ngp rather than Mitsuba, did you post in the right repo?

Shy-fawn · 2022-11-14T13:10:36Z

@merlinND 不好意思再打擾您
請問您是指這部分嗎?
請問可以查看到我付上的圖片嗎?

wjakob · 2022-11-14T13:16:12Z

@Shy-fawn 您在錯誤的存儲庫上發布。請用英文發帖，否則我們很遺憾需要封鎖您的帳戶。

zeynepiskenderoglu · 2023-01-04T08:42:34Z

'cmake' is not recognized as an internal or external command, operable program or batch file. i am having the problem.
Got the issue solved by manually copying the files from C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v10.1\extras\visual_studio_integration\MSBuildExtensions to C:\Program Files (x86)\Microsoft Visual Studio\2019\Community\ MSBuild\Microsoft\VC\v160\BuildCustomizations
I am trying to implement the process, but there is no such path on my computer; C:\Program Files (x86)\Microsoft Visual Studio\2019\Community\MSBuild\Microsoft\VC\v160\BuildCustomizations
instead there are these

what should I do?

wjakob · 2023-01-04T12:32:12Z

@zeynepiskenderoglu -- please use Mitsuba 3, this project is no longer supported.

aakash-chaddha · 2023-11-19T23:31:32Z

If still facing the issue on windows:

Make sure cmake is path is in windows global env-variables.
Steps:

Search "Developer Command Prompt" in all-apps, I get this "Developer Command Prompt for VS 2019"
Now type "where cmake" copy the path i.e "C:\Program Files (x86)\Microsoft Visual Studio\2019\BuildTools\Common7\IDE\CommonExtensions\Microsoft\CMake\CMake\bin"
Append it in Path in System Variables
Make sure to remove "build" folder from your working dir where cmake build command failed.
run cmake build command

Speierers closed this as completed Apr 29, 2020

This was referenced Jan 17, 2022

Build problem: No CUDA toolset found NVlabs/instant-ngp#18

Closed

Build problem on Windows: No known features for CXX compiler "MSVC" NVlabs/instant-ngp#21

Closed

greenbrettmichael mentioned this issue May 2, 2022

Cuda not found on windows 10 NVlabs/instant-ngp#461

Closed

herrpedro mentioned this issue Dec 18, 2022

CMake error at CMakelists.txt:11 The system cannot find the file specified NVlabs/instant-ngp#1083

Open

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Building issues [no CUDA toolset found] #103

Building issues [no CUDA toolset found] #103

ph57fe commented Apr 22, 2020

Speierers commented Apr 23, 2020

ph57fe commented Apr 23, 2020 •

edited

Loading

Speierers commented Apr 23, 2020

ph57fe commented Apr 23, 2020

Speierers commented Apr 24, 2020

hjoonpark commented Apr 25, 2020 •

edited

Loading

Speierers commented Apr 27, 2020

ph57fe commented Apr 28, 2020

Speierers commented Apr 29, 2020 •

edited

Loading

KnightYao commented Jan 6, 2021

Theo-Wu commented Jul 21, 2021 •

edited

Loading

Chenguoz commented Jan 23, 2022

giodude12 commented May 29, 2022

SHPEYOLO commented Jun 18, 2022

merlinND commented Jun 18, 2022

Shy-fawn commented Nov 12, 2022

merlinND commented Nov 12, 2022

Shy-fawn commented Nov 14, 2022

wjakob commented Nov 14, 2022

zeynepiskenderoglu commented Jan 4, 2023

wjakob commented Jan 4, 2023

aakash-chaddha commented Nov 19, 2023

Building issues [no CUDA toolset found] #103

Building issues [no CUDA toolset found] #103

Comments

ph57fe commented Apr 22, 2020

Speierers commented Apr 23, 2020

ph57fe commented Apr 23, 2020 • edited Loading

Speierers commented Apr 23, 2020

ph57fe commented Apr 23, 2020

Speierers commented Apr 24, 2020

hjoonpark commented Apr 25, 2020 • edited Loading

Speierers commented Apr 27, 2020

ph57fe commented Apr 28, 2020

Speierers commented Apr 29, 2020 • edited Loading

KnightYao commented Jan 6, 2021

Theo-Wu commented Jul 21, 2021 • edited Loading

Chenguoz commented Jan 23, 2022

giodude12 commented May 29, 2022

SHPEYOLO commented Jun 18, 2022

merlinND commented Jun 18, 2022

Shy-fawn commented Nov 12, 2022

merlinND commented Nov 12, 2022

Shy-fawn commented Nov 14, 2022

wjakob commented Nov 14, 2022

zeynepiskenderoglu commented Jan 4, 2023

wjakob commented Jan 4, 2023

aakash-chaddha commented Nov 19, 2023

ph57fe commented Apr 23, 2020 •

edited

Loading

hjoonpark commented Apr 25, 2020 •

edited

Loading

Speierers commented Apr 29, 2020 •

edited

Loading

Theo-Wu commented Jul 21, 2021 •

edited

Loading