Fix precompiled headers build

[albestro]

I accidentally end up building with DLAF_WITH_PRECOMPILED_HEADERS when I faced this error

ialberto@daint105:~/workspace/dla-future.master/build-gpu> ninja dlaf.pch_exe
[1/1] Linking CXX executable src/dlaf.pch_exe
FAILED: src/dlaf.pch_exe
: && /project/csstaff/ialberto/opt/spack-daint/lib/spack/env/case-insensitive/CC -O2 -g -DNDEBUG -pthread -Wl,-lstdc++fs -Wl,-z,defs -fvisibility=hidden -gz src/CMakeFiles/dlaf.pch_exe.dir/dummy.cpp.o src/CMakeFiles/dlaf.pch_exe.dir/dummy.c.o -o src/dlaf.pch_exe  -Wl,-rpath,/project/csstaff/ialberto/opt/spack-daint/opt/spack/cray-cnl7-haswell/gcc-11.2.0/pika-0.24.0-3ffp63gse4arykov5lqfvtxlbcizxurt/lib64:/project/csstaff/ialberto/opt/spack-daint/opt/spack/cray-cnl7-haswell/gcc-11.2.0/mimalloc-2.1.2-jnuakfdhri3vtpomqfvjudomcqb5vs7b/lib64:/project/csstaff/ialberto/opt/spack-daint/opt/spack/cray-cnl7-haswell/gcc-11.2.0/hwloc-2.9.1-umciffc276gxr5xjg63uh2kcwtp42fvx/lib:/project/csstaff/ialberto/opt/spack-daint/opt/spack/cray-cnl7-haswell/gcc-11.2.0/cuda-11.8.0-kffo4e4lxlqpkycnwnzp3pezipuxpql7/lib64:/project/csstaff/ialberto/opt/spack-daint/opt/spack/cray-cnl7-haswell/gcc-11.2.0/lapackpp-2023.11.05-iwqitwa5pepgyvpagytdvbaow7wh3lra/lib64:/project/csstaff/ialberto/opt/spack-daint/opt/spack/cray-cnl7-haswell/gcc-11.2.0/blaspp-2023.11.05-hcccnba6uimx6scyck7tujifktta43nv/lib64:/project/csstaff/ialberto/opt/spack-daint/opt/spack/cray-cnl7-haswell/gcc-11.2.0/hdf5-1.14.3-xms4am33tuov57rfsidnebnseicdr5up/lib  -m64 -L/project/csstaff/ialberto/opt/spack-daint/opt/spack/cray-cnl7-haswell/gcc-11.2.0/intel-oneapi-mkl-2024.0.0-5ibuwehkykfoicua3gqfxh56dtrx7uyt/lib -Wl,--no-as-needed -lmkl_intel_lp64 -lmkl_sequential -lmkl_core -lpthread -lm -ldl  /project/csstaff/ialberto/opt/spack-daint/opt/spack/cray-cnl7-haswell/gcc-11.2.0/pika-0.24.0-3ffp63gse4arykov5lqfvtxlbcizxurt/lib64/libpika.so.0.24.0  /project/csstaff/ialberto/opt/spack-daint/opt/spack/cray-cnl7-haswell/gcc-11.2.0/mimalloc-2.1.2-jnuakfdhri3vtpomqfvjudomcqb5vs7b/lib64/libmimalloc.so.2.1  /project/csstaff/ialberto/opt/spack-daint/opt/spack/cray-cnl7-haswell/gcc-11.2.0/hwloc-2.9.1-umciffc276gxr5xjg63uh2kcwtp42fvx/lib/libhwloc.so  /project/csstaff/ialberto/opt/spack-daint/opt/spack/cray-cnl7-haswell/gcc-11.2.0/cuda-11.8.0-kffo4e4lxlqpkycnwnzp3pezipuxpql7/lib64/libcusolver.so  /project/csstaff/ialberto/opt/spack-daint/opt/spack/cray-cnl7-haswell/gcc-11.2.0/cuda-11.8.0-kffo4e4lxlqpkycnwnzp3pezipuxpql7/lib64/libcublas.so  /project/csstaff/ialberto/opt/spack-daint/opt/spack/cray-cnl7-haswell/gcc-11.2.0/cuda-11.8.0-kffo4e4lxlqpkycnwnzp3pezipuxpql7/lib64/libculibos.a  /project/csstaff/ialberto/opt/spack-daint/opt/spack/cray-cnl7-haswell/gcc-11.2.0/cuda-11.8.0-kffo4e4lxlqpkycnwnzp3pezipuxpql7/lib64/libcublasLt.so  /project/csstaff/ialberto/opt/spack-daint/opt/spack/cray-cnl7-haswell/gcc-11.2.0/cuda-11.8.0-kffo4e4lxlqpkycnwnzp3pezipuxpql7/lib64/libcusparse.so  -latomic  -lrt  /project/csstaff/ialberto/opt/spack-daint/opt/spack/cray-cnl7-haswell/gcc-11.2.0/cuda-11.8.0-kffo4e4lxlqpkycnwnzp3pezipuxpql7/lib64/libcudart.so  /project/csstaff/ialberto/opt/spack-daint/opt/spack/cray-cnl7-haswell/gcc-11.2.0/lapackpp-2023.11.05-iwqitwa5pepgyvpagytdvbaow7wh3lra/lib64/liblapackpp.so  /project/csstaff/ialberto/opt/spack-daint/opt/spack/cray-cnl7-haswell/gcc-11.2.0/blaspp-2023.11.05-hcccnba6uimx6scyck7tujifktta43nv/lib64/libblaspp.so  /project/csstaff/ialberto/opt/spack-daint/opt/spack/cray-cnl7-haswell/gcc-11.2.0/intel-oneapi-mkl-2024.0.0-5ibuwehkykfoicua3gqfxh56dtrx7uyt/mkl/2024.0/lib/libmkl_scalapack_lp64.so  /project/csstaff/ialberto/opt/spack-daint/opt/spack/cray-cnl7-haswell/gcc-11.2.0/intel-oneapi-mkl-2024.0.0-5ibuwehkykfoicua3gqfxh56dtrx7uyt/mkl/2024.0/lib/libmkl_cdft_core.so  /project/csstaff/ialberto/opt/spack-daint/opt/spack/cray-cnl7-haswell/gcc-11.2.0/intel-oneapi-mkl-2024.0.0-5ibuwehkykfoicua3gqfxh56dtrx7uyt/mkl/2024.0/lib/libmkl_gf_lp64.so  /project/csstaff/ialberto/opt/spack-daint/opt/spack/cray-cnl7-haswell/gcc-11.2.0/intel-oneapi-mkl-2024.0.0-5ibuwehkykfoicua3gqfxh56dtrx7uyt/mkl/2024.0/lib/libmkl_sequential.so  /project/csstaff/ialberto/opt/spack-daint/opt/spack/cray-cnl7-haswell/gcc-11.2.0/intel-oneapi-mkl-2024.0.0-5ibuwehkykfoicua3gqfxh56dtrx7uyt/mkl/2024.0/lib/libmkl_core.so  /project/csstaff/ialberto/opt/spack-daint/opt/spack/cray-cnl7-haswell/gcc-11.2.0/intel-oneapi-mkl-2024.0.0-5ibuwehkykfoicua3gqfxh56dtrx7uyt/mkl/2024.0/lib/libmkl_blacs_intelmpi_lp64.so  /opt/cray/pe/mpt/7.7.18/gni/mpich-gnu/8.2/lib/libmpich.so  -lpthread  -lm  /usr/lib64/libdl.so  /opt/gcc/11.2.0/snos/lib64/libgomp.so  /project/csstaff/ialberto/opt/spack-daint/opt/spack/cray-cnl7-haswell/gcc-11.2.0/umpire-2024.02.0-bqph3c62c3btxrubssa2q2vzukdak7pm/lib64/libumpire.a  /project/csstaff/ialberto/opt/spack-daint/opt/spack/cray-cnl7-haswell/gcc-11.2.0/camp-2024.02.0-taonagxrzo2x4tpcdxrynz222ewoxv2w/lib/libcamp.a  /project/csstaff/ialberto/opt/spack-daint/opt/spack/cray-cnl7-haswell/gcc-11.2.0/fmt-10.2.1-4zxoeuyyoqlhxumbnli2dbj6npuc7ems/lib64/libfmt.a  /project/csstaff/ialberto/opt/spack-daint/opt/spack/cray-cnl7-haswell/gcc-11.2.0/cuda-11.8.0-kffo4e4lxlqpkycnwnzp3pezipuxpql7/lib64/libcudart_static.a  /usr/lib64/librt.so  /project/csstaff/ialberto/opt/spack-daint/opt/spack/cray-cnl7-haswell/gcc-11.2.0/hdf5-1.14.3-xms4am33tuov57rfsidnebnseicdr5up/lib/libhdf5_cpp.so.310.0.3  /project/csstaff/ialberto/opt/spack-daint/opt/spack/cray-cnl7-haswell/gcc-11.2.0/hdf5-1.14.3-xms4am33tuov57rfsidnebnseicdr5up/lib/libhdf5.so.310.3.0  -ldl && :
/usr/bin/ld: /usr/lib/../lib64/crt1.o: in function `_start':
/home/abuild/rpmbuild/BUILD/glibc-2.26/csu/../sysdeps/x86_64/start.S:110: undefined reference to `main'
collect2: error: ld returned 1 exit status
ninja: build stopped: subcommand failed.

This is the first thing that came to my mind to fix it, but I'm open to other suggestion of better/correct solution (I didn't investigate why I never faced it before but now it shows up)

[albestro]

I added +pch variant and enabled it for "CUDA ScaLAPACK debug" CI configuration.

Waiting the CI infrastructure to be back so that I can test it.

[albestro]

cscs-ci run

[albestro]

CI passed correctly except for one of the ROCm build that timed out after 2hours.

I changed the spack package so it rebuilds all dependencies, and the machine where the CI is currently running on might not be powerful enough.

I'm confident it is not a problem related to this PR, so I think we can consider it a successful CI run.

[rasolca]

ROCM CI is not related to rebuilding the dependencies stack. It's just resources not powerful enough. Happened with other PRs as well.