>I'm not that familiar with OSX and will have to ask someone how to
>determine the binutils version used natively.
>However, I see there's a port of this binutils:
>
>binutils devel/binutils 2.17 FSF Binutils for native development.
>
>I installed this, but the build fails in the same way; full log below.
Are you sure it's actually using the newly-installed binutils? Try adding
a -v to the compilation command and see what version "as" reports (this
should all be on one line):
cd aclib && gcc -DHAVE_CONFIG_H -I. -I. -I.. -D_REENTRANT -DLINUX -I.. -Wall
-D_LARGEFILE_SOURCE -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -g -O2
-no-cpp-precomp -D_INTL_REDIRECT_MACROS -c tcmemcpy.c -fno-common -DPIC -o
.libs/tcmemcpy.o -v
Either way, I guess I'll just bite the bullet and use more compatible
syntax. I've committed patches for both the "0b111" problem and the
".align" problem to CVS HEAD; try compiling that and see if you have any
better luck.
>Interesting -- because adding the --disable-mmx doesn't solve the
>problem -- build fails in the same way. I'm happy to send the full log
>of that too.
I can't reproduce this with CVS HEAD; the configure system was considerably
less refined in 1.0.x, so there may be some problems there. Repeat after
me: "I want to use CVS HEAD... I want to use CVS HEAD..." (:
--Andrew Church
[EMAIL PROTECTED]
http://achurch.org/
