Ch21_migrating_cuda_code needs to be qualified

[jaynu2]

samples/CMakeLists.txt should be qualified
Otherwise the build just fails...

Otherwise, nice job on the cmake.

if(WITHCUDA)
add_subdirectory( Ch21_migrating_cuda_code )
endif()

although a suggestion ...
mkdir build
cd build
cmake -S .. -D .
make
would keep all the build data from cluttering up the source tree.

Also, I added this to .bashrc
if [[ -z "$SETVARS_COMPLETED" ]]; then
source /opt/intel/oneapi/setvars.sh
fi

Otherwise vscode complains when you create a new term.

[bashbaug]

Hi, thanks for the feedback!

The code for Chapter 21 has been migrated to SYCL so it shouldn't need to be qualified. Can you describe the build errors you are seeing without the qualification?

Note that the suggested build steps in the README suggests building in a separate "build" directory. The exact commands are a bit different than what you've suggested above, but they should have the same effect, and they shouldn't clutter your source tree with build files.

[jaynu2]

[ 99%] Building CXX object samples/Ch21_migrating_cuda_code/CMakeFiles/fig_21_13-14_reve rse_migrated.dir/fig_21_13-14_reverse_migrated.cpp.o /home/jayn/dpcpp/data-parallel-CPP/samples/Ch21_migrating_cuda_code/fig_21_13-14_reverse _migrated.cpp:5:10: fatal error: 'dpct/dpct.hpp' file not found 5 | #include <dpct/dpct.hpp> | ^~~~~~~~~~~~~~~ I presume it is looking for some CUDA related include files.

…

On Mon, Jun 30, 2025 at 1:42 PM Ben Ashbaugh ***@***.***> wrote: *bashbaug* left a comment (Apress/data-parallel-CPP#142) <#142 (comment)> Hi, thanks for the feedback! The code for Chapter 21 has been migrated to SYCL so it shouldn't need to be qualified. Can you describe the build errors you are seeing without the qualification? Note that the suggested build steps in the README <https://github.com/Apress/data-parallel-CPP?tab=readme-ov-file#building-the-samples> suggests building in a separate "build" directory. The exact commands are a bit different than what you've suggested above, but they should have the same effect, and they shouldn't clutter your source tree with build files. — Reply to this email directly, view it on GitHub <#142 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/BS3MCXANNT3LWNVGAGLADKT3GGAIFAVCNFSM6AAAAACALOPDIKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTAMRQGMZDANZTGM> . You are receiving this because you authored the thread.Message ID: ***@***.***>

[jaynu2]

Maybe dpct has moved. I see it has been renamed to syclomatic. Anyway, it looks like not included in whatever toolkit I installed. The HPC download looked pretty heavy. The initialization shell script reports this: :: initializing oneAPI environment ... bash: BASH_VERSION = 5.2.21(1)-release args: Using "$@" for setvars.sh arguments: :: ccl -- latest :: compiler -- latest :: debugger -- latest :: dev-utilities -- latest :: dnnl -- latest :: dpl -- latest :: mkl -- latest :: mpi -- latest :: pti -- latest :: tbb -- latest :: umf -- latest :: oneAPI environment initialized ::

…

On Mon, Jun 30, 2025 at 11:06 PM jay Norwood ***@***.***> wrote: [ 99%] Building CXX object samples/Ch21_migrating_cuda_code/CMakeFiles/fig_21_13-14_reve rse_migrated.dir/fig_21_13-14_reverse_migrated.cpp.o /home/jayn/dpcpp/data-parallel-CPP/samples/Ch21_migrating_cuda_code/fig_21_13-14_reverse _migrated.cpp:5:10: fatal error: 'dpct/dpct.hpp' file not found 5 | #include <dpct/dpct.hpp> | ^~~~~~~~~~~~~~~ I presume it is looking for some CUDA related include files. On Mon, Jun 30, 2025 at 1:42 PM Ben Ashbaugh ***@***.***> wrote: > *bashbaug* left a comment (Apress/data-parallel-CPP#142) > <#142 (comment)> > > Hi, thanks for the feedback! > > The code for Chapter 21 has been migrated to SYCL so it shouldn't need to > be qualified. Can you describe the build errors you are seeing without the > qualification? > > Note that the suggested build steps in the README > <https://github.com/Apress/data-parallel-CPP?tab=readme-ov-file#building-the-samples> > suggests building in a separate "build" directory. The exact commands are a > bit different than what you've suggested above, but they should have the > same effect, and they shouldn't clutter your source tree with build files. > > — > Reply to this email directly, view it on GitHub > <#142 (comment)>, > or unsubscribe > <https://github.com/notifications/unsubscribe-auth/BS3MCXANNT3LWNVGAGLADKT3GGAIFAVCNFSM6AAAAACALOPDIKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTAMRQGMZDANZTGM> > . > You are receiving this because you authored the thread.Message ID: > ***@***.***> >