Thanks for you help.
They are really shared lib, not static. I know it sounds weird: it's a
python extension which must not link the ${PYTHON_LIBRARY} (according to
official docs; in fact If I do so I get a segmentation fault) but finally
in my executable,a sort of custom/home made ${PYTHON_EXEC
You have order dependent static libraries which can be solved by
constructing a cycle. As Bo stated by constructing the cycle
B;A;ImportedExtLib;B;A each library can see each other.
Looking at your original code example it looks like you are importing
the libraries as SHARED, but I think these are
I read the docs about LINK_INTERFACE_MULTIPLICITYtoo.
But I'm not in the situation of a "cyclic dependency" so I 'm not sure if
it solve my problem.
In my case (with B A and ImpLib) what would be the syntax?
Il giorno mar 6 nov 2018 alle ore 06:30 Robert Maynard <
robert.mayn...@kitware.com>
The target_link_libraries has a property called
LINK_INTERFACE_MULTIPLICITY that should help you out.
https://cmake.org/cmake/help/v3.13/command/target_link_libraries.html#cyclic-dependencies-of-static-libraries
On Tue, Nov 6, 2018 at 8:24 AM scrgiorgio wrote:
>
> Thanks for the help,
>
> trying
Thanks for the help,
trying this (or any combination):
target_link_libraries(MyExe B)
target_link_libraries(MyExe A )
target_link_libraries(MyExe ImpLib)
I get this order:
'B;...whatever...;A;ImpLib"
and the last past is causing the problem. Apparently there is no way to
change the right part
Hi All,
Like many, I've been having trouble understanding the RPATH control settings.
In particular, I can't seem to get external libraries (e.g. boost) to link in
*any* way that allows it to find dependent libraries without DYLD_LIBRARY_PATH.
Eventually I'd like to refer to them with relative pa
