Re: [Maxima-discuss] Compiling pregexp.lisp fails with GCL?

[Camm Maguire]

Greetings!  If you'd like to post your GCL 2.6.13pre119 build failure
with CentOS, I'd be happy to take a look.  FYI, I'm using Debian
unstable/testing as the primary ongoing development environment for GCL,
where all is building successfully at the moment.  I'll also be checking
Mingw/macosx right before release.

Take care,

"Viktor T. Toth" <vttoth@vttoth.com> writes:

> I was able to compile Maxima on CentOS7 using GCL 2.6.13pre50.
>
> A more recent version, GCL 2.6.13pre119, itself failed to build on CentOS 7 
> which is why I picked an earlier version.
>
> The testsuite yields 6 failures, all in share. Haven't investigated these yet 
> or checked how this version of GCL fares against other LISPs.
>
>
> Viktor
>
>
>
>> -----Original Message-----
>> From: Viktor T. Toth <vttoth@vttoth.com>
>> Sent: Wednesday, August 3, 2022 2:54 AM
>> To: 'Robert Dodier' <robert.dodier@gmail.com>
>> Cc: maxima-discuss@lists.sourceforge.net
>> Subject: Re: [Maxima-discuss] Compiling pregexp.lisp fails with GCL?
>> 
>> This is exactly what I ran into.
>> 
>> FWIW, my gcl is compiled from source, but otherwise it is a very vanilla 
>> CentOS 7 setup.
>> 
>> I'll check to see if perhaps rebuilding gcl helps.
>> 
>> 
>> Viktor
>> 
>> 
>> > -----Original Message-----
>> > From: Robert Dodier <robert.dodier@gmail.com>
>> > Sent: Tuesday, August 2, 2022 11:49 AM
>> > To: Viktor T. Toth <vttoth@vttoth.com>
>> > Cc: <maxima-discuss@lists.sourceforge.net> 
>> > <maxima-discuss@lists.sourceforge.net>
>> > Subject: Re: [Maxima-discuss] Compiling pregexp.lisp fails with GCL?
>> >
>> > I just tried compiling current Git version w/ GCL (advertising itself
>> > as 2.6.12, but the version number hasn't changed in a long time, so
>> > it's hard to say exactly what version it is; I compiled it from a
>> > tarball 2015-08-12), and I see an error compiling src/pregexp.lisp.
>> >
>> > Reported to stdout:
>> >
>> > SIMPLE-ERROR: (SYSTEM "gcc -c -Wall -fsigned-char
>> > -Wno-unused-but-set-variable -pipe
>> > -I/usr/local/lib/gcl-2.6.12/unixport/../h  -O3 -fomit-frame-pointer -c
>> > \"./binary-gcl/pregexp.c\" -o \"./binary-gcl/pregexp.o\" ") returned a
>> > non-zero value 0.
>> >
>> >
>> > Reported to stderr:
>> >
>> > ./binary-gcl/pregexp.c: In function ‘L45’:
>> > ./binary-gcl/pregexp.c:4287:7: error: ‘fun’ undeclared (first use in
>> > this function)
>> >   if(((fun->cc.cc_turbo[1]->c.c_car))==Cnil){
>> >        ^
>> > ./binary-gcl/pregexp.c:4287:7: note: each undeclared identifier is
>> > reported only once for each function it appears in
>> > ./binary-gcl/pregexp.c: In function ‘L46’:
>> > ./binary-gcl/pregexp.c:4449:11: error: ‘fun’ undeclared (first use in
>> > this function)
>> >   funcall((fun->cc.cc_turbo[0]->c.c_car));
>> >            ^
>> > ./binary-gcl/pregexp.c: In function ‘L44’:
>> > ./binary-gcl/pregexp.c:4625:11: error: ‘fun’ undeclared (first use in
>> > this function)
>> >   funcall((fun->cc.cc_turbo[0]->c.c_car));
>> >            ^
>> >
>> > etc., there are several more errors of the same kind.
>> >
>> > For good measure, I've attached the generated pregexp.c. Indeed it
>> > looks like variable `fun` is referenced without a definition in some
>> > functions.
>> >
>> > Dunno what it is in pregexp.lisp which is triggering the problem. My
>> > first guess is that there might be a lot of closures in the Lisp code,
>> > and maybe that's somehow tickling a bug, but that's just a guess.
>> >
>> > best,
>> >
>> > Robert
>> 
>> 
>> 
>> _______________________________________________
>> Maxima-discuss mailing list
>> Maxima-discuss@lists.sourceforge.net
>> https://lists.sourceforge.net/lists/listinfo/maxima-discuss
>
>
>
> _______________________________________________
> Maxima-discuss mailing list
> Maxima-discuss@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/maxima-discuss

-- 
Camm Maguire                                        camm@maguirefamily.org
==========================================================================
"The earth is but one country, and mankind its citizens."  --  Baha'u'llah