Instead of the DynamicLinker returning a hardcoded value like
`-Wl,-foo`, it now is passed a value that could be '-Wl,', or could be
something '-Xlinker='
This makes a few things cleaner, and will make it possible to fix using
clang (not clang-cl) on windows, where it invokes either link.exe or
lld-link.exe instead of a gnu-ld compatible linker.
Before python 3.5.4 typing didn't have the NoReturn type, use a string
to forward declare it.
aside: I can't wait to be able to use python 3.7's
__future__.annotations and not have to deal wit hthis anymore.
Fixes#5822
This is a mixin class for DynamicLinkers that aims to provide only the
most basic posix-like linker arguments, in other words bits that gnu ld,
apple's ld, solaris's ld, and other completely disjoint linkers that are
used on Unix-like OSes share.
I started out just annotating, but realized there is a lot of really
silly duplication of methods in this module. Nearly all of the methods
return [], but each class implemented the same methods over and over!
I've fixed this by moving the default implementation down into the
`StaticLinker` class, and then overriding as necessary in the derived
classes.
It appears that LIB/LINK default to the host architecture if they can't
guess it from the first object. With the MSVC toolchain, resource files
are (usually) compiled to an arch-neutral .res format. Always
explicitly provide a '/MACHINE:' argument to avoid it guessing
incorrectly when cross-compiling.
ninja chokes when building FFmpeg's static libraries, as the
command line can be larger than 32000.
This was disabled on purpose in #1649, but the rsp syntax was
different: this commit makes it so the options and output file
are still passed on the command line, gcc-ar didn't work
otherwise.
Aleksey Gurtovoy