June 2002 - PPL-devel - BUGSENG Mailman

ppl/src Constraint.inlines.hh
by Enea Zaffanella 18 Jun '02

18 Jun '02

CVSROOT: /cvs/ppl Module name: ppl Changes by: zaffanella(a)cs.unipr.it 2002-06-18 10:16:14 Modified files: src : Constraint.inlines.hh Log message: Avoiding the creation of temporaries in Constraint operator>(const LinExpression& e1, const LinExpression& e2) Constraint operator>(const Integer& n, const LinExpression& e) Constraint operator>(const LinExpression& e, const Integer& n). Patches: http://www.cs.unipr.it/cgi-bin/cvsweb.cgi/ppl/src/Constraint.inlines.hh.dif…

1 0

ppl/doc definitions.dox
by Patricia Hill 18 Jun '02

18 Jun '02

CVSROOT: /cvs/ppl Module name: ppl Changes by: hill(a)cs.unipr.it 2002-06-18 08:57:22 Modified files: doc : definitions.dox Log message: Paragraph for widening improved. Paragraph for time-elapse added (just a reference to the 1997 Halbwachs paper). Patches: http://www.cs.unipr.it/cgi-bin/cvsweb.cgi/ppl/doc/definitions.dox.diff?cvsr…

1 0

Throwing Prolog exceptions from C code
by Roberto Bagnara 18 Jun '02

18 Jun '02

Hi there, I don't know if I am doing something stupid but, in order to throw a Prolog exception with term t from C code, I create the Term `throw(t)' and then I pass it to YapCallProlog(). Everything fails horribly and I have no idea why. I append a stack trace hoping it may be enlightening for someone. All the best Roberto -- Prof. Roberto Bagnara Computer Science Group Department of Mathematics, University of Parma, Italy http://www.cs.unipr.it/~bagnara/ mailto:bagnara@cs.unipr.it Program received signal SIGSEGV, Segmentation fault. do_goal (CodeAdr=0x81a5ac0 "", arity=1869768820, pt=0x81afe44, args_to_save=0, top=0) at C/exec.c:908 908 XREGS[i+1] = *pt++; (gdb) info stack #0 do_goal (CodeAdr=0x81a5ac0 "", arity=1869768820, pt=0x81afe44, args_to_save=0, top=0) at C/exec.c:908 #1 0x0806ce69 in execute_goal (t=135986753, nargs=0, mod=1) at C/exec.c:996 #2 0x0805025b in YapCallProlog (t=135986753) at C/c_interface.c:587 #3 0x400b631e in Prolog_raise_exception(unsigned long) (t=135986733) at /home/roberto/ppl/ppl/interfaces/Prolog/YAP/ppl_yap.cc:188 #4 0x400a6b0c in handle_exception(not_a_polyhedron_kind const&) ( e=@0x100e6b40) at /home/roberto/ppl/ppl/interfaces/Prolog/ppl_prolog.icc:139 #5 0x400a9303 in ppl_new_Polyhedron_from_dimension (t_tp=1686330, t_nd=54, t_ph=135986672) at /home/roberto/ppl/ppl/interfaces/Prolog/ppl_prolog.icc:681

2 6

ppl/interfaces/Prolog/YAP ppl_yap.cc
by Roberto Bagnara 18 Jun '02

18 Jun '02

CVSROOT: /cvs/ppl Module name: ppl Changes by: roberto(a)cs.unipr.it 2002-06-18 07:35:54 Modified files: interfaces/Prolog/YAP: ppl_yap.cc Log message: Code for testing YAP exceptions temporarily included. Patches: http://www.cs.unipr.it/cgi-bin/cvsweb.cgi/ppl/interfaces/Prolog/YAP/ppl_yap…

1 0

Re: [Doxygen-users] Dogygen + Automake
by Roberto Bagnara 18 Jun '02

18 Jun '02

Braden McDaniel wrote: > Anyone using Automake to build/install Doxygen docs for a project? I'm > looking for a good way to do this that will hopefully avoid explicitly > listing in my Makefile.am the 600-odd files Doxygen generates. Right now > I just generate the files directly to the installation location; but > this is not interacting well with RPM. Hi there, I believe the `dist-hook' target is what you are looking for. You can see some examples of how we are using this and other hooks automake provides at http://www.cs.unipr.it/cgi-bin/cvsweb.cgi/ppl/ Please let me know if you come up with better solutions. All the best, Roberto -- Prof. Roberto Bagnara Computer Science Group Department of Mathematics, University of Parma, Italy http://www.cs.unipr.it/~bagnara/ mailto:bagnara@cs.unipr.it

2 1

ppl/src Polyhedron.cc Polyhedron.defs.hh
by Enea Zaffanella 18 Jun '02

18 Jun '02

CVSROOT: /cvs/ppl Module name: ppl Changes by: zaffanella(a)cs.unipr.it 2002-06-18 00:06:25 Modified files: src : Polyhedron.cc Polyhedron.defs.hh Log message: Private methods strongly_minimize* now return a boolean which is false if the result is an empty polyhedron (this is consistent with the behavior of private method minimize()). Patched method widening_CC92_assign() to also work for NNC polyhedra. Patches: http://www.cs.unipr.it/cgi-bin/cvsweb.cgi/ppl/src/Polyhedron.cc.diff?cvsroo… http://www.cs.unipr.it/cgi-bin/cvsweb.cgi/ppl/src/Polyhedron.defs.hh.diff?c…

1 0

ppl/interfaces/Prolog/YAP Makefile.am ppl_yap. ...
by Roberto Bagnara 17 Jun '02

17 Jun '02

CVSROOT: /cvs/ppl Module name: ppl Changes by: roberto(a)cs.unipr.it 2002-06-17 23:28:40 Modified files: interfaces/Prolog/YAP: Makefile.am ppl_yap.pl yap_clpq.pl Log message: Some progress in the support for YAP. Patches: http://www.cs.unipr.it/cgi-bin/cvsweb.cgi/ppl/interfaces/Prolog/YAP/Makefil… http://www.cs.unipr.it/cgi-bin/cvsweb.cgi/ppl/interfaces/Prolog/YAP/ppl_yap… http://www.cs.unipr.it/cgi-bin/cvsweb.cgi/ppl/interfaces/Prolog/YAP/yap_clp…

1 0

ppl/src Row.cc Row.defs.hh Row.inlines.hh
by Roberto Bagnara 17 Jun '02

17 Jun '02

CVSROOT: /cvs/ppl Module name: ppl Changes by: roberto(a)cs.unipr.it 2002-06-17 22:29:38 Modified files: src : Row.cc Row.defs.hh Row.inlines.hh Log message: The NDEBUG symbol no longer changes data layout. Now, to obtain a version of Row where each object remembers its own capacity, the macro EXTRA_ROW_DEBUG must evaluate to true (whether this capacity is tested in run-time assertions or not depends, as usual, on the NDEBUG macro not being or being defined). Patches: http://www.cs.unipr.it/cgi-bin/cvsweb.cgi/ppl/src/Row.cc.diff?cvsroot=ppl&r… http://www.cs.unipr.it/cgi-bin/cvsweb.cgi/ppl/src/Row.defs.hh.diff?cvsroot=… http://www.cs.unipr.it/cgi-bin/cvsweb.cgi/ppl/src/Row.inlines.hh.diff?cvsro…

1 0

#if #if vs #ifdef
by Roberto Bagnara 17 Jun '02

17 Jun '02

There is a consolidated tendency to rely onto #if instead of #ifdef, and I believe it would be a nice thing if we follow the common practices. Currently, the STRONG_NORMALIZATION symbol is the only one that needs to be cured. The symbol definition should also be commented so that the comment shows up in the developer's manual. Ciao Roberto -- Prof. Roberto Bagnara Computer Science Group Department of Mathematics, University of Parma, Italy http://www.cs.unipr.it/~bagnara/ mailto:bagnara@cs.unipr.it

1 0

ppl/tests Makefile.am wideningcc921.cc widenin ...
by Enea Zaffanella 17 Jun '02

17 Jun '02

CVSROOT: /cvs/ppl Module name: ppl Changes by: zaffanella(a)cs.unipr.it 2002-06-17 18:28:32 Modified files: tests : Makefile.am wideningcc921.cc wideningcc922.cc Added files: tests : wideningcc923.cc Log message: Added a new tests showing a bug in the current implementation of widening_CC92_assign, when the arguments are NNC polyhedra. Corrected the output of the other widening tests. Patches: http://www.cs.unipr.it/cgi-bin/cvsweb.cgi/ppl/tests/wideningcc923.cc.diff?c… http://www.cs.unipr.it/cgi-bin/cvsweb.cgi/ppl/tests/Makefile.am.diff?cvsroo… http://www.cs.unipr.it/cgi-bin/cvsweb.cgi/ppl/tests/wideningcc921.cc.diff?c… http://www.cs.unipr.it/cgi-bin/cvsweb.cgi/ppl/tests/wideningcc922.cc.diff?c…

1 0