[Gcl-devel] Re: FriCAS and gcl-2.7.0

[Camm Maguire]

Greetings!

OK, with the current cvs, I'm at:
cp BINARY.pht 
/fix/t1/camm/fricas/dist-155_2.7/target/i686-pc-linux/share/hypertex/pages/BINARY.pht
(AXIOM=/fix/t1/camm/fricas/dist-155_2.7/target/i686-pc-linux; 
LC_CTYPE=ISO-8859-1; export AXIOM LC_CTYPE; \
         unset DAASE; 
/fix/t1/camm/fricas/dist-155_2.7/target/i686-pc-linux/bin/sman -noihere -paste 
/fix/t1/camm/fricas/dist-155_2.7/target/i686-pc-linux/share/hypertex/pages/BOP.ht)
GCL (GNU Common Lisp)  2.7.0 CLtL1    Jan  3 2008 01:35:06
Source License: LGPL(gcl,gmp,pargcl), GPL(unexec,bfd,xgcl)
Binary License:  GPL due to GPL'ed components: (XGCL READLINE BFD UNEXEC)
Modifications of this banner must retain notice of a compatible license
Dedicated to the memory of W. Schelter

Use (help) to get some basic information on how to use GCL.

Temporary directory for compiler files set to /tmp/
                 FriCAS (AXIOM fork) Computer Algebra System 
                         Version: FriCAS 2007-12-14
              Timestamp: Tuesday January 15, 2008 at 15:07:36 
-----------------------------------------------------------------------------
   Issue )copyright to view copyright notices.
   Issue )summary for a summary of useful system commands.
   Issue )quit to leave FriCAS and return to shell.
-----------------------------------------------------------------------------
 
(1) -> 
making BOP.input
parsing: BasicOperatorXmpPage
writing:        y := operator 'y
writing:        deq := D(y x, x, 2) + D(y x, x) + y x = 0
writing:        solve(deq, y, x)



I'm hoping this is past the point to which you refer.  In any case,
the build seems to have stopped here with no current cpu usage.
Hmm...

Take care,

Waldek Hebisch <address@hidden> writes:

> Camm Maguire wrote:
> > Greetings, and thanks so much for the feedback!
> > 
> > The standard way to debug this is to compile gcl with --enable-debug
> > and run under gdb, 'watch'ing for changes to the address listed.  I'd
> > be happy to look at this, as it is likely more efficient than
> > explaining how to do it.
> > 
> > However, I cannot seem to get the ./configure to see gclcvs if both
> > this and gcl are installed.  (Tried ln -s /usr/bin/gclcvs ./gcl;export
> > PATH=$(pwd):$PATH; ./configure --with-list=$(pwd)/gcl ....)  Always
> > picks up /usr/bin/gcl for me.
> >
>                                    ^^^^^^^^^
> 
> 'with-list' is a misspelling, the correct option is 'with-lisp'.  
> To build with gcl-2.7.0 one still needs a few patches compared
> to distributed version.  Also, ATM for testing it is better
> to use version containig precompiled Lisp files (compiling
> Spad to Lisp is likely to fail in simlar way, but during 
> compilation it is harder to reproduce problems).  I have put
> the source tree + precompiled algebra (with 2.7 patch applied)
> at:
> 
> http://www.math.uni.wroc.pl/~hebisch/fricas/dist-155_2.7.tar.bz2
> 
> > Likewise, it is likely more efficient to build at first with
> > si::*disable-recompile* set to t.
> 
> my current 2.7 patch is doing this -- otherwise I get failure
> earlier.
> 
> > This is likely something simple, if as you report the problem is after
> > the first gc.
> > 
> 
> I have some more data:  on my machine there is a call to object
> at 0x01e30c70.  This object (probably hashtable) is a part of
> AXIOMsys image and garbage collector marks it as free.  It seems
> that this object really is unused: setting hadware watch at
> 0x01e30c70 I see that the first access is from garbage collector.
> When I disabled garbage collector (replacing GBC by 'ret' instruction)
> FriCAS goes further and I see no access to 0x01e30c70...
> 
> -- 
>                               Waldek Hebisch
> address@hidden 
> 
> 
> 

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