This is the mail archive of the
libffi-discuss@sourceware.cygnus.com
mailing list for the libffi project.
Re: Emotional judgments (was: ANNOUNCE: Ffidl 0.1 Alpha)
- Subject: Re: Emotional judgments (was: ANNOUNCE: Ffidl 0.1 Alpha)
- From: Anthony Green <green at cygnus dot com>
- Date: 27 Oct 1999 12:49:59 -0700
- CC: libffi-discuss at sourceware dot cygnus dot com, krab at daimi dot au dot dk
- Newsgroups: comp.lang.tcl
- Posted-To: comp.lang.tcl
- References: <m37lk9rdse.fsf@elf.org> <38162FFD.44DCD36A@scriptics.com><845587F822B2936A.842F460DD6B60C8A.319C7D34A79A2BB0@lp.airnews.net><Rz91vGA9huF4Eww1@jessikat.demon.co.uk> <m33duwvjji.fsf@elf.org>
The following message is a courtesy copy of an article
that has been posted as well.
Roger E Critchlow Jr <rec@elf.org> writes:
> Libffi compiled, after much trickery involving config.status, with
> only two modifications to sysv.S, deleting the lines:
>
> .type ffi_call_SYSV,@function
> and
> .size ffi_call_SYSV,.ffi_call_SYSV_end-ffi_call_SYSV
>
> whose purpose I don't know. I'm rebooting into Windows today for
> another try.
It's ok to delete these two lines. I know someone had it working on
Windows with mingw32 once before.
> The question of timing Tcl code has me, you'll excuse the expression,
> really ticked off.
FYI - development snapshots of libffi 2 have an interface that is
slightly faster on some architectures (x86 in particular), and a
closure mechanism for callbacks - both of which we are using in the
gcj runtime (http://sourceware.cygnus.com/java).
AG
--
Anthony Green Cygnus Solutions
Sunnyvale, California