This is the mail archive of the
libffi-discuss@sourceware.org
mailing list for the libffi project.
Re: Status?
- From: Bill Northcott <w dot northcott at unsw dot edu dot au>
- To: Evan Lavelle <eml-libffi at cyconix dot com>
- Cc: libffi-discuss at sourceware dot org
- Date: Wed, 12 Jul 2006 11:11:48 +1000
- Subject: Re: Status?
- References: <44B3EAB3.6000306@cyconix.com>
On 12/07/2006, at 4:15 AM, Evan Lavelle wrote:
Can anyone please tell me what the status of libffi is? What's the
current version, and where's the best place to download stand-alone
sources?
I'm also confused about why libffi is now part of gcc - does this
mean that gcc has some sort of native support for dynamic function
calls? I can't find anything in the online docs about this.
Good question. As far as I can see, it is all disintegrating into
one of those unfortunate open source messes, which happen when the
project is not managed.
I have spent some time on this and discovered the following:
There is a version of libffi distributed with gcc. However, it is
not officially part of gcc because of problems over copyright. All
official gcc code must be copyright FSF and covered by GPL. This
requires the original authors and copyright holders to assign their
rights to the FSF. In this case I think there is uncertainty about
who they are.
As a result little work is done on it in the gcc code base. I think
it is used to as part of gcc Java support.
There is another version distributed by the PyObjc folks, who have
added ppc64 and i386 support for Darwin, but these have not been
added back into gcc because of the copyright issues.
GNUstep developers also pitch in some times.
There are probably other forks out there produced by people who
needed them for one reason or another.
There is always ffcall which has similar functionality but again has
possible copyright issues.
Bill Northcott