Mail Archives: djgpp/2004/09/10/09:09:56
> >I doubt this will get better with gcc 3.4.x... :) My bcmp.c is
> >identical to current CVS version(!).
> .. Good to see that you have checked, but what about the rest of the
> files?
There is at least one similar "cast-reject" in swab.c (CVS HEAD
version as well).
Thanks for your lonely reply! Probably other developers became afraid
to have to help me compile the whole suite, which may not be a task
for everyone...
> BTW: I haven't done a complete DJGPP build for a long time so I cannot
> give any info as to shat occures with GCC 3.4
I will have to study the thing a lot before I could even think of
contributing. I really don't want to ask beginner's questions here...
Yesterday I finally remained stuck unable to link something because of
the recent (_)environ changes--the symbol should be provided through a
change in "djgpp.djl" linker script, but the linker is instructed to
use a lib/gcc-lib/.../djgpp-x.djl instead, which doesn't have the
needed PROVIDE directive, but when I replace it with the djgpp.djl CVS
version, the linker chokes on it (parse error on new "LONG(0)" line),
so... It's a black magic. Now I see the truth: don't help me, I'm
not worth it! :-))
Egon
- Raw text -