At least on Debian 10 the standard c++ compiler is g++-8,
but CUDA only supports g++-7 and our test for CUDA in cmake
did send an error in that case/combination which is quite
annoying.
The reason was that check_language(CUDA) did not honor
the CMAKE_CUDA_FLAGS variable and always used the default g++7,
but enable_language(CUDA) did honor it. As we do set the underlying
host compiler the fromer reported that CUDA is available while the
latter marked the CUDA compiler as broken.
With this commit we work around this by setting the environment
variable ENV{CUDAHOSTCXX} which nvcc will use. Hence now we only try
to enable CUDA if it is compatible with the C++ compiler
It seems like the VERSION_GREATER_EQUAL operator for boolean
expressions was introduced after CMake 3.6 and hence the current
check whether to activate CUDA or not is broken in version 3.6 and
below.
This PR fixes this by using VERSION_GREATER.
Closes#2375.
We experienced weired linker errors when using host compiler version for
compilation that were not supported by the nvcc used to compile the
cuda code:
```
[ 15%] Linking CXX executable bin/test_timer
/usr/bin/ld: /home/mblatt/src/dune/opm-2.6/opm-common/opm-seq/lib/libopmcommon.a(Parser.cpp.o): in function `Opm::(anonymous namespace)::file& std::vector<Opm::(anonymous namespace)::file, std::allocator<Opm::(anonymous namespace)::file> >::emplace_back<std::filesystem::__cxx11::path&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >&>(std::filesystem::__cxx11::path&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >&)':
Parser.cpp:(.text+0x1096): undefined reference to `std::filesystem::__cxx11::path::_M_split_cmpts()'
/usr/bin/ld: Parser.cpp:(.text+0x10ad): undefined reference to `std::filesystem::__cxx11::path::_M_split_cmpts()'
/usr/bin/ld: /home/mblatt/src/dune/opm-2.6/opm-common/opm-seq/lib/libopmcommon.a(Parser.cpp.o): in function `Opm::(anonymous namespace)::ParserState::loadFile(std::filesystem::__cxx11::path const&)':
Parser.cpp:(.text+0x23a1): undefined reference to `std::filesystem::canonical(std::filesystem::__cxx11::path const&)'
/usr/bin/ld: Parser.cpp:(.text+0x24e0): undefined reference to
`std::filesystem::__cxx11::path::_M_split_cmpts()'
```
The reason turned out to be that the library path was build up by
paths of the old (g++-7) compiler used by nvcc and the actual (newer) compiler
g++-8. This completely messed up the linker paths for CMake.
To detect this situation already when running cmake we have resorted
to first setting the CMAKE_CUDA_FLAGS to force cmake to make nvcc use
the host compiler and to activate CUDA (if available) before calling
`find_package(CUDA)`. If the host compiler is not supported CMake will
error out during `enable(CUDA)`
Note that we still use (deprecated) FindCUDA later to determine the
libraries to link to.
The users has the option to either deactivate CUDA by setting
`-DCMAKE_DISABLE_FIND_PACKAGE_CUDA=ON` or to use a compiler supported
by nvcc (setting `-DCMAKE_CXX_COMPILER=compiler`).
Additionally we do not try to activate CUDA the CMake version is <
3.8. Please note that previously CMake would have errored out here
anyway since we used the unsupported `enable_language(CUDA)` even in
this case.
Closes#2363.
`mebos` works similarly as `flow`, but in contrast to `flow`, `mebos`
only creates the deck in the common code path whilst the
'EclipseState' and the other higher-level parser objects are always
created internally by the vanguard. this approach avoids code
duplication and the worst effects of parser API creep.
to avoid having to compile non-trivial compile units multiple times,
the actual code of the variants is moved into `ebos_$VARIANT.{hh,cc}`
files and the respective compile units are each put into a small
static library whilst the main function of said libraries are invoked
by either the multiplexed or the respective specialized simulator's
`main()`. This is also somewhat similar of how `flow` works, with the
difference that `mebos` uses the blackoil variant to determine the
parameters it needs to know for parsing the deck instead of
introducing a "fake" type tag for this. The rationale is to reduce
compile time compared to the "fake type tag" approach and -- to a
lesser extend -- avoid unnecessary copy-and-pasting of code. In
particular, this means that for the vast majority of cases, only one
place needs changed in the code for all `ebos` variants if, for
example, the parser API requires further objects in the future.
