CMake Error: math cannot parse the expression

[tuannv-github]

Hello,

I got the following error when building nvblox release-0.0.5:

➜  build git:(release-0.0.5) cmake -DCMAKE_CUDA_COMPILER:PATH=/usr/local/cuda/bin/nvcc ..
-- The CUDA compiler identification is NVIDIA 12.5.82
-- Detecting CUDA compiler ABI info
-- Detecting CUDA compiler ABI info - done
-- Check for working CUDA compiler: /usr/local/cuda/bin/nvcc - skipped
-- Detecting CUDA compile features
-- Detecting CUDA compile features - done
-- Downloading 3rdparty dependencies
-- Downloading Eigen
-- Downloading STDGPU
CMake Error at build/_deps/ext_stdgpu-src/cmake/Findthrust.cmake:17 (math):
  math cannot parse the expression: "200400 // macro expansion with ##
  requires this to be a single value / 100000": syntax error, unexpected
  exp_DIVIDE (9).
Call Stack (most recent call first):
  build/_deps/ext_stdgpu-src/src/stdgpu/CMakeLists.txt:7 (find_package)


CMake Error at build/_deps/ext_stdgpu-src/cmake/Findthrust.cmake:18 (math):
  math cannot parse the expression: "(200400 // macro expansion with ##
  requires this to be a single value / 100) % 1000": syntax error, unexpected
  exp_DIVIDE (10).
Call Stack (most recent call first):
  build/_deps/ext_stdgpu-src/src/stdgpu/CMakeLists.txt:7 (find_package)


CMake Error at build/_deps/ext_stdgpu-src/cmake/Findthrust.cmake:19 (math):
  math cannot parse the expression: "200400 // macro expansion with ##
  requires this to be a single value % 100": syntax error, unexpected
  exp_DIVIDE (9).
Call Stack (most recent call first):
  build/_deps/ext_stdgpu-src/src/stdgpu/CMakeLists.txt:7 (find_package)


CMake Error at /usr/share/cmake-3.22/Modules/FindPackageHandleStandardArgs.cmake:230 (message):
  Could NOT find thrust: Found unsuitable version "ERROR.ERROR.ERROR", but
  required is at least "1.9.2" (found /usr/local/cuda/include)
Call Stack (most recent call first):
  /usr/share/cmake-3.22/Modules/FindPackageHandleStandardArgs.cmake:592 (_FPHSA_FAILURE_MESSAGE)
  build/_deps/ext_stdgpu-src/cmake/Findthrust.cmake:26 (find_package_handle_standard_args)
  build/_deps/ext_stdgpu-src/src/stdgpu/CMakeLists.txt:7 (find_package)


-- Configuring incomplete, errors occurred!
See also "/home/vantuan_ngo/nvblox/nvblox/build/CMakeFiles/CMakeOutput.log".
See also "/home/vantuan_ngo/nvblox/nvblox/build/CMakeFiles/CMakeError.log".
➜  build git:(release-0.0.5) 

[nyne-nyne]

Hello! Did you manage to resolve it?
I was able to resolve this by using an older version of CUDA (12.2). It seems that, (see here: stotko/stdgpu#407) that from CUDA 12.4 and up there's a comment after THRUST_VERSION in thrust_version.h that messes up the regex, and causes the issue there.

Another fix, if you required CUDA 12.5, is to change the commit hash here stdgpu.cmake to commit 1f0b2d51718692ec9046fe1b36173a591c611bdb or later.

[josyulavt]

I did change the commit hash to the latest version, but I'm still getting a cmake error with cuda 12.6

CMake Error at tests/CMakeLists.txt:147 (add_executable):
  add_executable cannot create target "benchmark" because another target with
  the same name already exists.  The existing target is a static library
  created in source directory
  "/home/josyulak/Projects/nvblox/nvblox/build/_deps/benchmark-src/src".  See
  documentation for policy CMP0002 for more details.


CMake Error at tests/CMakeLists.txt:149 (target_link_libraries):
  The keyword signature for target_link_libraries has already been used with
  the target "benchmark".  All uses of target_link_libraries with a target
  must be either all-keyword or all-plain.

  The uses of the keyword signature are here:

   * build/_deps/benchmark-src/src/CMakeLists.txt:43 (target_link_libraries)
   * build/_deps/benchmark-src/src/CMakeLists.txt:48 (target_link_libraries)



CMake Error at CMakeLists.txt:272 (install):
  install TARGETS target stdgpu is exported but not all of its interface file
  sets are installed

[nyne-nyne]

This seems to be a different issue. Could you try seeing where the that target was already defined?