FreeBSD. Idea to tweak the configure script and WANTLIB changes from jca@.
Thanks to phessler@ for getting the gears moving on this.
ok jca@ jasper@ (MAINTAINER), makes sense to jsg@
This switches libffi to the use of two different mappings per closure,
one with RW access, the other with RX access. Even if not perfect it is
already an improvement. Soon RWX memory will be unavailable by default,
so fixing libffi means that we don't have to tweak an unknown percentage
of the 565 ports that link against it...
Initial diff from sthen@ who also did most of the tests.
ok sthen@ jasper@ (maintainer)
functions returning structs of 5 to 15 bytes, inclusive; making me wonder
what I had actually tested. Now all the testcase passes, for real. Noticed
while working on cleaning these changes to push them upstream.
Bump revision.
SysV/m68k, which NetBSD and OpenBSD follow).
Add m88k/ELF support.
Add vax/a.out support.
All tests from the libffi.call dejagnu testsuite pass on these three platforms.
ok jasper@
FFI stands for Foreign Function Interface. A foreign function interface
is the popular name for the interface that allows code written in one
language to call code written in another language. The libffi library
really only provides the lowest, machine dependent layer of a fully
featured foreign function interface. A layer must exist above libffi
that handles type conversions for values passed between the two
languages.
successfully tested on seven arches by me. other arches are disabled for now as
they need more work (mips64) or haven't been tested yet.