Camm Maguire wrote:
> Greetings! Could you please outline your build problems?
>
> Take care,
>
> "Vadim V. Zhytnikov" <vvzhy@mail.ru> writes:
>
> >
> > Certainly GCL should not be damped on purpose, but I worry that this will
> > happen "automatically" unless we'll find a person(s) who is willing
> > to support GCL (I already can't rebuild GCL after recent Linux upgrades).
> > I'm not so optimistic about this. Personally I'm interested in developing
> > Maxima not any particular Lisp system.
> >
> >
Let me first describe symptoms. During standard build
procedure make stage fails. raw_gcl builds and works OK.
saved_gcl also builds but with suspicious error messages
and result is not operational. saved_gcl can be
started and even performs some lisp operations but dumps
core on others. For example you can crash it by typing
any unbound variable. Bad saved_gcl later fails to compile
some additional modules (gcl-tk) and total make process
fails.
I build GCL on ALT Linux distributive (Mandrake derived
Russian distro) and everything was OK with GCL and Maxima
until I made some large system upgrade. The interesting point is that
I still can build GCL and Maxima on old unupgraded system
and run resulting binaries on new one without any trouble
I just can't build GCL on it. I do not know which part of the system
causes the conflict. I only know so far that this is not Linux kernel
and compiler (problem remains with both gcc 2.95.3 and 2.96).
My primary suspect is new glibc 2.2.4.
I also positive that problem is not due to my personal
occasional Linux misconfiguration since same trouble
is observed in Debian
Vadim
--
[ Vadim V. Zhytnikov <vvzhy@mail.ru> <vvzhy@td.lpi.ac.ru> ]