August 2009 - PPL-devel - BUGSENG Mailman

Internal compiler error building ppl on Linux...
by Tovrea, George W (US SSA) 16 Aug '09

16 Aug '09

Has anyone seen this problem? If so, is there a work around? Do I need to build this directory? Platform: uname -m = x86_64 uname -r = 2.6.9-67.ELxenU uname -s = Linux uname -v = #1 SMP Wed Nov 7 14:17:17 EST 2007 /usr/bin/uname -p = unknown /bin/uname -X = unknown /bin/arch = x86_64 /usr/bin/arch -k = unknown /usr/convex/getsysinfo = unknown /usr/bin/hostinfo = unknown /bin/machine = unknown /usr/bin/oslevel = unknown /bin/universe = unknown Error: make[7]: Entering directory `/home/users/tovrea/GCC_LINUX/build_linux/ppl/demos/ppl_lcdd/examples' make[7]: Nothing to be done for `all'. make[7]: Leaving directory `/home/users/tovrea/GCC_LINUX/build_linux/ppl/demos/ppl_lcdd/examples' make[7]: Entering directory `/home/users/tovrea/GCC_LINUX/build_linux/ppl/demos/ppl_lcdd' g++ -DHAVE_CONFIG_H -I. -I/home/users/tovrea/GCC_LINUX/gcc-4.4.0/ppl/demos/ppl_lcdd -I../.. -I../.. -I../../src -I/home/users/tovrea/GCC_LINUX/gcc-4.4.0/ppl/utils -g -fkeep-inline-functions -MT ppl_lcdd.o -MD -MP -MF .deps/ppl_lcdd.Tpo -c -o ppl_lcdd.o /home/users/tovrea/GCC_LINUX/gcc-4.4.0/ppl/demos/ppl_lcdd/ppl_lcdd.cc In file included from /home/users/tovrea/GCC_LINUX/gcc-4.4.0/ppl/demos/ppl_lcdd/ppl_lcdd.cc:37 : ../../src/ppl.hh: In function `Parma_Polyhedra_Library::Result Parma_Polyhedra_Library::Checked::input_generic(Type&, std::istream&, Parma_Polyhedra_Library::Rounding_Dir)': ../../src/ppl.hh:4678: internal compiler error: in invert_truthvalue, at fold-const.c:2700 Please submit a full bug report, with preprocessed source if appropriate. See <URL:http://bugzilla.redhat.com/bugzilla> for instructions. Preprocessed source stored into /tmp/ccdHbmnk.out file, please attach this to your bugreport. make[7]: *** [ppl_lcdd.o] Error 1 make[7]: Leaving directory `/home/users/tovrea/GCC_LINUX/build_linux/ppl/demos/ppl_lcdd' Bill Tovrea BAE Systems 16550 West Bernardo Drive San Diego, CA 92127 Telephone: (858) 592-5292 Pager: (888) 971-4964 E-mail: George.Tovrea(a)baesystems.com

2 1

Re: [PPL-devel] Undefined PPL references attempting to build cloog-ppl (fwd)
by P M Hill 15 Aug '09

15 Aug '09

---------- Forwarded message ---------- Date: Tue, 11 Aug 2009 21:44:39 +0100 From: "Tovrea, George W (US SSA)" <george.tovrea(a)baesystems.com> To: P M Hill <hill(a)comp.leeds.ac.uk> Subject: RE: [PPL-devel] Undefined PPL references attempting to build cloog-ppl Yes, they are also messed up there: required_instantiations_c_cxx_headers = ppl_c_Polyhedron.hh ppl_c_Grid.hh ppl_c_Rational_Box.hh ppl_c_BD_Shape_mpz_class.hh ppl_c__BD_Shape_mpq_class.hh ppl_c__Octagonal_Shape_mpz_class.hh ppl_c__Octagonal_Shape_mpq_class.hh ppl_c__Constraints_Product_C_Polyhedron_Grid.hh ppl_c__Pointset_Powerset_C_Polyhedron.hh ppl_c__Pointset_Powerset_NNC_Polyhedron.hh required_instantiations_c_cxx_objects = ppl_c_Polyhedron.lo ppl_c_Grid.lo ppl_c_Rational_Box.lo ppl_c_BD_Shape_mpz_class.lo ppl_c__BD_Shape_mpq_class.lo ppl_c__Octagonal_Shape_mpz_class.lo ppl_c__Octagonal_Shape_mpq_class.lo ppl_c__Constraints_Product_C_Polyhedron_Grid.lo ppl_c__Pointset_Powerset_C_Polyhedron.lo ppl_c__Pointset_Powerset_NNC_Polyhedron.lo required_instantiations_c_cxx_sources = ppl_c_Polyhedron.cc ppl_c_Grid.cc ppl_c_Rational_Box.cc ppl_c_BD_Shape_mpz_class.cc ppl_c__BD_Shape_mpq_class.cc ppl_c__Octagonal_Shape_mpz_class.cc ppl_c__Octagonal_Shape_mpq_class.cc ppl_c__Constraints_Product_C_Polyhedron_Grid.cc ppl_c__Pointset_Powerset_C_Polyhedron.cc ppl_c__Pointset_Powerset_NNC_Polyhedron.cc required_instantiations_canonical_names = Polyhedron@Grid@Rational_Box@BD_Shape_mpz_class@_BD_Shape_mpq_class@_Oct agonal_Shape_mpz_class@_Octagonal_Shape_mpq_class@_Constraints_Product_C _Polyhedron_Grid@_Pointset_Powerset_C_Polyhedron@_Pointset_Powerset_NNC_ Polyhedron -----Original Message----- From: P M Hill [mailto:hill@comp.leeds.ac.uk] Sent: Tuesday, August 11, 2009 1:42 PM To: Tovrea, George W (US SSA) Cc: Sebastian Pop; The Parma Polyhedra Library developers' list; Roberto Bagnara Subject: Re: [PPL-devel] Undefined PPL references attempting to build cloog-ppl Also check the Makefile in the same directory. If they are messed up there, then the problem appears to be with autoreconf. Pat On Tue, 11 Aug 2009, Tovrea, George W (US SSA) wrote: > Yes, I have looked at that file (ppl_interface_instantiations.m4) and > the names are messed up in there, as you can see: > > m4_define(`m4_interface_classes_names', > `Polyhedron@Grid@Rational_Box@BD_Shape_mpz_class@_BD_Shape_mpq_class@_Oc > tagonal_Shape_mpz_class@_Octagonal_Shape_mpq_class@_Constraints_Product_ > C_Polyhedron_Grid@_Pointset_Powerset_C_Polyhedron@_Pointset_Powerset_NNC > _Polyhedron') > m4_define(`m4_cplusplus_classes_names', > `Polyhedron@Grid@Rational_Box@BD_Shape<mpz_class>@ BD_Shape<mpq_class>@ > Octagonal_Shape<mpz_class>@ Octagonal_Shape<mpq_class>@ > Constraints_Product<C_Polyhedron,Grid>@ Pointset_Powerset<C_Polyhedron>@ > Pointset_Powerset<NNC_Polyhedron>') > > -----Original Message----- > From: P M Hill [mailto:hill@comp.leeds.ac.uk] > Sent: Tuesday, August 11, 2009 1:31 PM > To: Tovrea, George W (US SSA) > Cc: Roberto Bagnara; The Parma Polyhedra Library developers' list; > Sebastian Pop > Subject: Re: [PPL-devel] Undefined PPL references attempting to build > cloog-ppl > > Here are a few observations wrt the code you have sent and maybe they > might help us find the bug. > > I have noticed in your snippets of code that the interface names with > the > extra "_" are always the same - and it is added before the main root. > (Either "_" precedes the name or "ppl_c__" precedes the name where the > double "__" should be a single "_".) A list of these names is built by > the > configure as the default interface classes and are placed in the > Makefile > in the interfaces directory. When make is executed, the Makefile > generates > a file called ppl_interface_instantiations.m4 in the interfaces > directory. > > Assuming the default configuration is used, this should contain the > following: > > m4_define(`m4_interface_classes_names', > `Polyhedron@Grid@Rational_Box@BD_Shape_mpz_class@BD_Shape_mpq_class@Octa > gonal_Shape_mpz_class@Octagonal_Shape_mpq_class@Constraints_Product_C_Po > lyhedron_Grid@Pointset_Powerset_C_Polyhedron@Pointset_Powerset_NNC_Polyh > edron@Double_Box@BD_Shape_double@Octagonal_Shape_double') > m4_define(`m4_cplusplus_classes_names', > `Polyhedron@Grid@Rational_Box@BD_Shape<mpz_class>@BD_Shape<mpq_class>@Oc > tagonal_Shape<mpz_class>@Octagonal_Shape<mpq_class>@Constraints_Product< > C_Polyhedron,Grid>@Pointset_Powerset<C_Polyhedron>@Pointset_Powerset<NNC > _Polyhedron>@Double_Box@BD_Shape<double>@Octagonal_Shape<double>') > > and is then used by the m4 to generate the C code for each of these > classes. > > So it may be helpful to check that if one or both of these generated > files > have the names correctly (ie as above). I suspect that some names may > already be buggy and have the extra "_". In which case, it could be > autoreconf or even make that could be the problem and not m4. > > HTH > > Pat > > On Tue, 11 Aug 2009, Tovrea, George W (US SSA) wrote: > >> In a previous mail, you wrote: >> >> Strange: this is the same version I use. I don't know what to say: >> in ppl_c.h you should end up with >> >> PPL_TYPE_DECLARATION(Polyhedron) >> PPL_TYPE_DECLARATION(Grid) >> PPL_TYPE_DECLARATION(Rational_Box) >> PPL_TYPE_DECLARATION(BD_Shape_mpz_class) >> PPL_TYPE_DECLARATION(BD_Shape_mpq_class) >> PPL_TYPE_DECLARATION(Octagonal_Shape_mpz_class) >> PPL_TYPE_DECLARATION(Octagonal_Shape_mpq_class) >> PPL_TYPE_DECLARATION(Constraints_Product_C_Polyhedron_Grid) >> PPL_TYPE_DECLARATION(Pointset_Powerset_C_Polyhedron) >> PPL_TYPE_DECLARATION(Pointset_Powerset_NNC_Polyhedron) >> >> whereas you have >> >> PPL_TYPE_DECLARATION(Polyhedron) >> PPL_TYPE_DECLARATION(Grid) >> PPL_TYPE_DECLARATION(Rational_Box) >> PPL_TYPE_DECLARATION(BD_Shape_mpz_class) >> PPL_TYPE_DECLARATION(_BD_Shape_mpq_class) >> PPL_TYPE_DECLARATION(_Octagonal_Shape_mpz_class) >> PPL_TYPE_DECLARATION(_Octagonal_Shape_mpq_class) >> PPL_TYPE_DECLARATION(_Constraints_Product_C_Polyhedron_Grid) >> PPL_TYPE_DECLARATION(_Pointset_Powerset_C_Polyhedron) >> PPL_TYPE_DECLARATION(_Pointset_Powerset_NNC_Polyhedron) >> >> Note that all the declarations that are in error above have a double >> underscore after ppl_c in the snippet below. All the ones that are >> correct have a single underscore after ppl_c. For example, >> BD_Shape_mpz_class above relates to ppl_c_BD_Shape_mpz_class whereas >> _BD_Shape_mpq_class above relates to ppl_c__BD_Shape_mpq_class below. >> Seems more than coincidental. However, I modified config.status > changing >> all ppl_c__ tp ppl_c_. This did not build (still looking for some >> ppl_c__ files). I am currently trying to debug >> >> /net/whistler/auto/users/tovrea/local/sol2/bin/m4 --prefix-builtin > -I.. >> -I../../../interfaces/C -I../../../interfaces/C/.. \ >> ../../../interfaces/C/ppl_interface_generator_c_h.m4 > >> ppl_c_domains.h >> >> >> ppl_c_domains.h contains all the ppl_c__ names >> >> >> -----Original Message----- >> From: Roberto Bagnara [mailto:bagnara@cs.unipr.it] >> Sent: Tuesday, August 11, 2009 12:54 AM >> To: Tovrea, George W (US SSA) >> Cc: Sebastian Pop; The Parma Polyhedra Library developers' list >> Subject: Re: [PPL-devel] Undefined PPL references attempting to build >> cloog-ppl >> >> Tovrea, George W (US SSA) wrote: >>> The names are already messed up after running configure. From >> config.status: >>> >>> S["required_instantiations_c_cxx_objects"]="ppl_c_Polyhedron.lo >> ppl_c_Grid.lo ppl_c_Rational_Box.lo ppl_c_BD_Shape_mpz_class.lo >> ppl_c__BD_Shape_mpq_class.lo ppl_c__Octagonal_Shape_mpz_class.lo"\ >>> " ppl_c__Octagonal_Shape_mpq_class.lo >> ppl_c__Constraints_Product_C_Polyhedron_Grid.lo >> ppl_c__Pointset_Powerset_C_Polyhedron.lo ppl_c__Pointset_Powers"\ >>> "et_NNC_Polyhedron.lo" >>> S["required_instantiations_c_cxx_sources"]="ppl_c_Polyhedron.cc >> ppl_c_Grid.cc ppl_c_Rational_Box.cc ppl_c_BD_Shape_mpz_class.cc >> ppl_c__BD_Shape_mpq_class.cc ppl_c__Octagonal_Shape_mpz_class.cc"\ >>> " ppl_c__Octagonal_Shape_mpq_class.cc >> ppl_c__Constraints_Product_C_Polyhedron_Grid.cc >> ppl_c__Pointset_Powerset_C_Polyhedron.cc ppl_c__Pointset_Powers"\ >>> "et_NNC_Polyhedron.cc" >>> S["required_instantiations_c_cxx_headers"]="ppl_c_Polyhedron.hh >> ppl_c_Grid.hh ppl_c_Rational_Box.hh ppl_c_BD_Shape_mpz_class.hh >> ppl_c__BD_Shape_mpq_class.hh ppl_c__Octagonal_Shape_mpz_class.hh"\ >>> " ppl_c__Octagonal_Shape_mpq_class.hh >> ppl_c__Constraints_Product_C_Polyhedron_Grid.hh >> ppl_c__Pointset_Powerset_C_Polyhedron.hh ppl_c__Pointset_Powers"\ >>> "et_NNC_Polyhedron.hh" >> >> Hi Bill, >> >> I see nothing messed up in this snippet. >> Can you explain? >> All the best, >> >> Roberto >> >> > > _______________________________________________ > PPL-devel mailing list > PPL-devel(a)cs.unipr.it > http://www.cs.unipr.it/mailman/listinfo/ppl-devel >

1 0

[GIT] ppl/ppl(master): Mention correction of bug in Pointset_Powerset maximize/minimize methods.
by Enea Zaffanella 15 Aug '09

15 Aug '09

Module: ppl/ppl Branch: master Commit: 6a96e861272517118cfef7806f7b696dccd4d7e6 URL: http://www.cs.unipr.it/git/gitweb.cgi?p=ppl/ppl.git;a=commit;h=6a96e8612725… Author: Enea Zaffanella <zaffanella(a)cs.unipr.it> Date: Sat Aug 15 10:05:48 2009 +0200 Mention correction of bug in Pointset_Powerset maximize/minimize methods. --- NEWS | 4 ++++ 1 files changed, 4 insertions(+), 0 deletions(-) diff --git a/NEWS b/NEWS index fabf3a4..1c71776 100644 --- a/NEWS +++ b/NEWS @@ -50,6 +50,10 @@ o All methods having a name ending in `_and_minimize' (e.g., Bugfixes ======== +o Corrected a bug in maximize and mininimize optimization methods of + class template Pointset_Powerset, whereby the Boolean value true + (indicating successful optimization) was returned for empty powersets. + o Corrected a bug in method bool NNC_Polyhedron::poly_hull_assign_if_exact(const NNC_Polyhedron&); whereby some inexact NNC hulls were incorrectly flagged as exact.

1 0

[GIT] ppl/w3ppl(master): Mention correction of bug affecting Pointset_Powerset max/min-imize.
by Enea Zaffanella 15 Aug '09

15 Aug '09

Module: ppl/w3ppl Branch: master Commit: 7946eb43c6139601014facc5f860105754f834e8 URL: http://www.cs.unipr.it/git/gitweb.cgi?p=ppl/w3ppl.git;a=commit;h=7946eb43c6… Author: Enea Zaffanella <zaffanella(a)cs.unipr.it> Date: Sat Aug 15 09:51:34 2009 +0200 Mention correction of bug affecting Pointset_Powerset max/min-imize. --- htdocs/Bugs/Bugs.raw | 10 ++++++++++ htdocs/Bugs/archive.raw | 10 ++++++++++ 2 files changed, 20 insertions(+), 0 deletions(-) diff --git a/htdocs/Bugs/Bugs.raw b/htdocs/Bugs/Bugs.raw index 526d7bf..4b2bd94 100644 --- a/htdocs/Bugs/Bugs.raw +++ b/htdocs/Bugs/Bugs.raw @@ -50,6 +50,16 @@ affecting various releases of the PPL. <DL> +<DT>August 15, 2009</DT> +<DD> +Fixed a bug whereby <CODE>maximize</CODE> and <CODE>minimize</CODE> +optimization methods of the class template <CODE>Pointset_Powerset</CODE> +were returning the Boolean value <CODE>true</CODE> (indicating a successful +optimization) even when the pointset powerset was empty, thereby +contradicting their specification. Relevant commit is +<A HREF="http://www.cs.unipr.it/git/gitweb.cgi?p=ppl/ppl.git;a=commit;h=b52a2d273f4f…">b52a2d273f4f96eec4a2ac5bb453756906e52a88</A>. +</DD> + <DT>August 7, 2009</DT> <DD> Fixed a bug whereby method diff --git a/htdocs/Bugs/archive.raw b/htdocs/Bugs/archive.raw index f63538f..56643f2 100644 --- a/htdocs/Bugs/archive.raw +++ b/htdocs/Bugs/archive.raw @@ -10,6 +10,16 @@ for various versions of the PPL. <DL> +<DT>August 15, 2009</DT> +<DD> +Fixed a bug whereby <CODE>maximize</CODE> and <CODE>minimize</CODE> +optimization methods of the class template <CODE>Pointset_Powerset</CODE> +were returning the Boolean value <CODE>true</CODE> (indicating a successful +optimization) even when the pointset powerset was empty, thereby +contradicting their specification. Relevant commit is +<A HREF="http://www.cs.unipr.it/git/gitweb.cgi?p=ppl/ppl.git;a=commit;h=b52a2d273f4f…">b52a2d273f4f96eec4a2ac5bb453756906e52a88</A>. +</DD> + <DT>August 7, 2009</DT> <DD> Fixed a bug whereby method

1 0

[GIT] ppl/ppl(master): Indentation fixed.
by Enea Zaffanella 15 Aug '09

15 Aug '09

Module: ppl/ppl Branch: master Commit: ba692c7b20024a266852f51e98fe709a4e02eb3b URL: http://www.cs.unipr.it/git/gitweb.cgi?p=ppl/ppl.git;a=commit;h=ba692c7b2002… Author: Enea Zaffanella <zaffanella(a)cs.unipr.it> Date: Sat Aug 15 08:18:50 2009 +0200 Indentation fixed. --- src/Pointset_Powerset.templates.hh | 57 ++++++++++++++++++----------------- 1 files changed, 29 insertions(+), 28 deletions(-) diff --git a/src/Pointset_Powerset.templates.hh b/src/Pointset_Powerset.templates.hh index ff79c36..daeba2d 100644 --- a/src/Pointset_Powerset.templates.hh +++ b/src/Pointset_Powerset.templates.hh @@ -286,8 +286,8 @@ Pointset_Powerset<PSET>::remove_space_dimensions(const Variables_Set& vars) { template <typename PSET> void -Pointset_Powerset<PSET>::remove_higher_space_dimensions(dimension_type - new_dimension) { +Pointset_Powerset<PSET> +::remove_higher_space_dimensions(dimension_type new_dimension) { Pointset_Powerset& x = *this; if (new_dimension < x.space_dim) { for (Sequence_iterator si = x.sequence.begin(), @@ -328,7 +328,7 @@ Pointset_Powerset<PSET>::map_space_dimensions(const Partial_Function& pfunc) { template <typename PSET> void Pointset_Powerset<PSET>::expand_space_dimension(Variable var, - dimension_type m) { + dimension_type m) { Pointset_Powerset& x = *this; for (Sequence_iterator si = x.sequence.begin(), s_end = x.sequence.end(); si != s_end; ++si) @@ -340,7 +340,7 @@ Pointset_Powerset<PSET>::expand_space_dimension(Variable var, template <typename PSET> void Pointset_Powerset<PSET>::fold_space_dimensions(const Variables_Set& vars, - Variable dest) { + Variable dest) { Pointset_Powerset& x = *this; Variables_Set::size_type num_folded = vars.size(); if (num_folded > 0) { @@ -355,9 +355,9 @@ Pointset_Powerset<PSET>::fold_space_dimensions(const Variables_Set& vars, template <typename PSET> void Pointset_Powerset<PSET>::affine_image(Variable var, - const Linear_Expression& expr, - Coefficient_traits::const_reference - denominator) { + const Linear_Expression& expr, + Coefficient_traits::const_reference + denominator) { Pointset_Powerset& x = *this; for (Sequence_iterator si = x.sequence.begin(), s_end = x.sequence.end(); si != s_end; ++si) { @@ -373,9 +373,9 @@ Pointset_Powerset<PSET>::affine_image(Variable var, template <typename PSET> void Pointset_Powerset<PSET>::affine_preimage(Variable var, - const Linear_Expression& expr, - Coefficient_traits::const_reference - denominator) { + const Linear_Expression& expr, + Coefficient_traits::const_reference + denominator) { Pointset_Powerset& x = *this; for (Sequence_iterator si = x.sequence.begin(), s_end = x.sequence.end(); si != s_end; ++si) { @@ -441,7 +441,9 @@ Pointset_Powerset<PSET> ::generalized_affine_preimage(Variable var, const Relation_Symbol relsym, const Linear_Expression& expr, - Coefficient_traits::const_reference denominator) { Pointset_Powerset& x = *this; + Coefficient_traits::const_reference + denominator) { + Pointset_Powerset& x = *this; for (Sequence_iterator si = x.sequence.begin(), s_end = x.sequence.end(); si != s_end; ++si) { si->pointset().generalized_affine_preimage(var, relsym, expr, denominator); @@ -882,9 +884,9 @@ Pointset_Powerset<PSET> template <typename PSET> bool Pointset_Powerset<PSET>::maximize(const Linear_Expression& expr, - Coefficient& sup_n, - Coefficient& sup_d, - bool& maximum) const { + Coefficient& sup_n, + Coefficient& sup_d, + bool& maximum) const { const Pointset_Powerset& x = *this; x.omega_reduce(); if (x.is_empty()) @@ -937,10 +939,10 @@ Pointset_Powerset<PSET>::maximize(const Linear_Expression& expr, template <typename PSET> bool Pointset_Powerset<PSET>::maximize(const Linear_Expression& expr, - Coefficient& sup_n, - Coefficient& sup_d, - bool& maximum, - Generator& g) const { + Coefficient& sup_n, + Coefficient& sup_d, + bool& maximum, + Generator& g) const { const Pointset_Powerset& x = *this; x.omega_reduce(); if (x.is_empty()) @@ -1000,9 +1002,9 @@ Pointset_Powerset<PSET>::maximize(const Linear_Expression& expr, template <typename PSET> bool Pointset_Powerset<PSET>::minimize(const Linear_Expression& expr, - Coefficient& inf_n, - Coefficient& inf_d, - bool& minimum) const { + Coefficient& inf_n, + Coefficient& inf_d, + bool& minimum) const { const Pointset_Powerset& x = *this; x.omega_reduce(); if (x.is_empty()) @@ -1055,10 +1057,10 @@ Pointset_Powerset<PSET>::minimize(const Linear_Expression& expr, template <typename PSET> bool Pointset_Powerset<PSET>::minimize(const Linear_Expression& expr, - Coefficient& inf_n, - Coefficient& inf_d, - bool& minimum, - Generator& g) const { + Coefficient& inf_n, + Coefficient& inf_d, + bool& minimum, + Generator& g) const { const Pointset_Powerset& x = *this; x.omega_reduce(); if (x.is_empty()) @@ -1270,8 +1272,7 @@ template <typename Cert> bool Pointset_Powerset<PSET>:: is_cert_multiset_stabilizing(const std::map<Cert, size_type, - typename Cert::Compare>& y_cert_ms - ) const { + typename Cert::Compare>& y_cert_ms) const { typedef std::map<Cert, size_type, typename Cert::Compare> Cert_Multiset; Cert_Multiset x_cert_ms; collect_certificates(x_cert_ms); @@ -1317,7 +1318,7 @@ template <typename PSET> template <typename Cert, typename Widening> void Pointset_Powerset<PSET>::BHZ03_widening_assign(const Pointset_Powerset& y, - Widening wf) { + Widening wf) { // `x' is the current iteration value. Pointset_Powerset& x = *this;

1 0

[GIT] ppl/ppl(master): Corrected a bug in Pointset_Powerset methods minimize() and maximize().
by Enea Zaffanella 15 Aug '09

15 Aug '09

Module: ppl/ppl Branch: master Commit: b52a2d273f4f96eec4a2ac5bb453756906e52a88 URL: http://www.cs.unipr.it/git/gitweb.cgi?p=ppl/ppl.git;a=commit;h=b52a2d273f4f… Author: Enea Zaffanella <zaffanella(a)cs.unipr.it> Date: Sat Aug 15 08:15:14 2009 +0200 Corrected a bug in Pointset_Powerset methods minimize() and maximize(). We now return false if the powerset is found to be empty. --- src/Pointset_Powerset.templates.hh | 12 ++++++++++++ 1 files changed, 12 insertions(+), 0 deletions(-) diff --git a/src/Pointset_Powerset.templates.hh b/src/Pointset_Powerset.templates.hh index ea10566..ff79c36 100644 --- a/src/Pointset_Powerset.templates.hh +++ b/src/Pointset_Powerset.templates.hh @@ -887,6 +887,9 @@ Pointset_Powerset<PSET>::maximize(const Linear_Expression& expr, bool& maximum) const { const Pointset_Powerset& x = *this; x.omega_reduce(); + if (x.is_empty()) + return false; + bool first = true; PPL_DIRTY_TEMP_COEFFICIENT(supt_n); @@ -940,6 +943,9 @@ Pointset_Powerset<PSET>::maximize(const Linear_Expression& expr, Generator& g) const { const Pointset_Powerset& x = *this; x.omega_reduce(); + if (x.is_empty()) + return false; + bool first = true; PPL_DIRTY_TEMP_COEFFICIENT(supt_n); @@ -999,6 +1005,9 @@ Pointset_Powerset<PSET>::minimize(const Linear_Expression& expr, bool& minimum) const { const Pointset_Powerset& x = *this; x.omega_reduce(); + if (x.is_empty()) + return false; + bool first = true; PPL_DIRTY_TEMP_COEFFICIENT(inft_n); @@ -1052,6 +1061,9 @@ Pointset_Powerset<PSET>::minimize(const Linear_Expression& expr, Generator& g) const { const Pointset_Powerset& x = *this; x.omega_reduce(); + if (x.is_empty()) + return false; + bool first = true; PPL_DIRTY_TEMP_COEFFICIENT(inft_n);

1 0

[GIT] ppl/ppl(master): Corrected a few expected results in Powerset/maxmin1 .cc.
by Enea Zaffanella 15 Aug '09

15 Aug '09

Module: ppl/ppl Branch: master Commit: dfff7e13e8817970647c38289b1f4e8d94999ea2 URL: http://www.cs.unipr.it/git/gitweb.cgi?p=ppl/ppl.git;a=commit;h=dfff7e13e881… Author: Enea Zaffanella <zaffanella(a)cs.unipr.it> Date: Sat Aug 15 08:10:19 2009 +0200 Corrected a few expected results in Powerset/maxmin1.cc. This uncovered a bug in Pointset_Powerset methods maximize() and minimize(), whereby we are returning true even for empty powersets. --- tests/Octagonal_Shape/max_min1.cc | 1 - tests/Polyhedron/max_min1.cc | 25 ++++++++++- tests/Powerset/maxmin1.cc | 93 +++++++++++++++++++------------------ 3 files changed, 72 insertions(+), 47 deletions(-) diff --git a/tests/Octagonal_Shape/max_min1.cc b/tests/Octagonal_Shape/max_min1.cc index c0915ac..4d87172 100644 --- a/tests/Octagonal_Shape/max_min1.cc +++ b/tests/Octagonal_Shape/max_min1.cc @@ -326,7 +326,6 @@ test09() { bool test10() { - TOctagonal_Shape oct(0); print_constraints(oct, "*** oct ***"); diff --git a/tests/Polyhedron/max_min1.cc b/tests/Polyhedron/max_min1.cc index c8da74b..8a73b99 100644 --- a/tests/Polyhedron/max_min1.cc +++ b/tests/Polyhedron/max_min1.cc @@ -132,7 +132,6 @@ test02() { bool test03() { - C_Polyhedron ph(0); print_constraints(ph, "*** ph ***"); @@ -172,10 +171,34 @@ test03() { return ok; } +bool +test04() { + C_Polyhedron ph(2, EMPTY); + + print_constraints(ph, "*** ph ***"); + + Coefficient num = 0; + Coefficient den = 0; + bool included = false; + Generator g(point()); + Linear_Expression LE; + bool ok = !ph.maximize(LE, num, den, included, g) + && num == 0 && den == 0 && !included + && g.is_point() + && g.divisor() == 1; + ok = ok && !ph.minimize(LE, num, den, included, g) + && num == 0 && den == 0 && !included + && g.is_point() + && g.divisor() == 1; + + return ok; +} + } // namespace BEGIN_MAIN DO_TEST(test01); DO_TEST_F8(test02); DO_TEST(test03); + DO_TEST(test04); END_MAIN diff --git a/tests/Powerset/maxmin1.cc b/tests/Powerset/maxmin1.cc index b1a351f..0973762 100644 --- a/tests/Powerset/maxmin1.cc +++ b/tests/Powerset/maxmin1.cc @@ -29,15 +29,15 @@ namespace { bool test01() { Pointset_Powerset<C_Polyhedron> ps(0, EMPTY); - Coefficient num; - Coefficient den; - bool included; + Coefficient num = 0; + Coefficient den = 0; + bool included = false; Generator g(point()); Linear_Expression LE; - bool ok = ps.maximize(LE, num, den, included) - && num == 0 && den == 1 && !included; - ok = ok && ps.maximize(LE, num, den, included, g) - && num == 0 && den == 1 && !included + bool ok = !ps.maximize(LE, num, den, included) + && num == 0 && den == 0 && !included; + ok = ok && !ps.maximize(LE, num, den, included, g) + && num == 0 && den == 0 && !included && g.is_point() && g.divisor() == 1; @@ -69,9 +69,9 @@ test02() { Constraint_System cs; Linear_Expression LE = x; - Coefficient max_n; - Coefficient max_d; - bool max_included; + Coefficient max_n = 0; + Coefficient max_d = 0; + bool max_included = false; Generator max_g(point()); Pointset_Powerset<NNC_Polyhedron> ps(1, EMPTY); @@ -123,9 +123,9 @@ test03() { Constraint_System cs; Linear_Expression LE = 9*x + y; - Coefficient num; - Coefficient den; - bool included; + Coefficient num = 0; + Coefficient den = 0; + bool included = false; Generator g(point()); Pointset_Powerset<C_Polyhedron> ps(2, EMPTY); @@ -148,7 +148,6 @@ test03() { print_generator(g); nout << endl; - if (!ok) return false; @@ -180,16 +179,16 @@ test04() { Variable x(0); Linear_Expression LE = x; - Coefficient num; - Coefficient den; - bool included; + Coefficient num = 0; + Coefficient den = 0; + bool included = false; Generator g(point()); Pointset_Powerset<C_Polyhedron> ps(1, EMPTY); - bool ok = ps.maximize(LE, num, den, included) - && num == 0 && den == 1 && !included; - ok = ok && ps.maximize(LE, num, den, included, g) - && num == 0 && den == 1 && !included + bool ok = !ps.maximize(LE, num, den, included) + && num == 0 && den == 0 && !included; + ok = ok && !ps.maximize(LE, num, den, included, g) + && num == 0 && den == 0 && !included && g.is_point() && g.divisor() == 1; @@ -197,7 +196,10 @@ test04() { return false; ps.add_disjunct(C_Polyhedron(1)); - ok = !ps.maximize(LE, num, den, included, g); + ok = !ps.maximize(LE, num, den, included, g) + && num == 0 && den == 0 && !included + && g.is_point() + && g.divisor() == 1; return ok; } @@ -206,16 +208,16 @@ test04() { bool test05() { Linear_Expression LE; - Coefficient num; - Coefficient den; - bool included; + Coefficient num = 0; + Coefficient den = 0; + bool included = false; Generator g(point()); Pointset_Powerset<C_Polyhedron> ps(0, EMPTY); - bool ok = ps.minimize(LE, num, den, included) - && num == 0 && den == 1 && !included; - ok = ok && ps.minimize(LE, num, den, included, g) - && num == 0 && den == 1 && !included + bool ok = !ps.minimize(LE, num, den, included) + && num == 0 && den == 0 && !included; + ok = ok && !ps.minimize(LE, num, den, included, g) + && num == 0 && den == 0 && !included && g.is_point() && g.divisor() == 1; @@ -247,9 +249,9 @@ test06() { Constraint_System cs; Linear_Expression LE = x; - Coefficient min_n; - Coefficient min_d; - bool min_included; + Coefficient min_n = 0; + Coefficient min_d = 0; + bool min_included = false; Generator min_g(point()); Pointset_Powerset<NNC_Polyhedron> ps(1, EMPTY); @@ -302,9 +304,9 @@ test07() { Constraint_System cs; Linear_Expression LE = x + y; - Coefficient num; - Coefficient den; - bool included; + Coefficient num = 0; + Coefficient den = 0; + bool included = false; Generator g(point()); Pointset_Powerset<C_Polyhedron> ps(2, EMPTY); @@ -327,7 +329,6 @@ test07() { print_generator(g); nout << endl; - if (!ok) return false; @@ -359,25 +360,27 @@ test08() { Variable x(0); Linear_Expression LE = x; - Coefficient num; - Coefficient den; - bool included; + Coefficient num = 0; + Coefficient den = 0; + bool included = false; Generator g(point()); Pointset_Powerset<C_Polyhedron> ps(1, EMPTY); - bool ok = ps.minimize(LE, num, den, included) - && num == 0 && den == 1 && !included; - ok = ok && ps.minimize(LE, num, den, included, g) - && num == 0 && den == 1 && !included + bool ok = !ps.minimize(LE, num, den, included) + && num == 0 && den == 0 && !included; + ok = ok && !ps.minimize(LE, num, den, included, g) + && num == 0 && den == 0 && !included && g.is_point() && g.divisor() == 1; - if (!ok) return false; ps.add_disjunct(C_Polyhedron(1)); - ok = !ps.minimize(LE, num, den, included, g); + ok = !ps.minimize(LE, num, den, included, g) + && num == 0 && den == 0 && !included + && g.is_point() + && g.divisor() == 1; return ok; }

1 0

Re: [PPL-devel] [Bug middle-end/40981] aermod.f90 ICEs on -O2 -fgraphite-identity -floop-strip-mine
by Sebastian Pop 15 Aug '09

15 Aug '09

> Actually the error in gdb has changed with 1677_max.diff... > As expected: see the gcc_assert in the patch +/* Return in RES the maximum of the linear expression LE on polyhedron PS. */ + +void +ppl_max_for_le (ppl_Pointset_Powerset_C_Polyhedron_t ps, + ppl_Linear_Expression_t le, Value res) +{ + ppl_Coefficient_t num, denom; + Value dv, nv; + int maximum; + + value_init (nv); + value_init (dv); + ppl_new_Coefficient (&num); + ppl_new_Coefficient (&denom); + ppl_Pointset_Powerset_C_Polyhedron_maximize (ps, le, num, denom, &maximum); + + if (maximum) + { + ppl_Coefficient_to_mpz_t (num, nv); + ppl_Coefficient_to_mpz_t (denom, dv); + gcc_assert (value_notzero_p (dv)); + value_division (res, nv, dv); + } + + value_clear (nv); + value_clear (dv); + ppl_delete_Coefficient (num); + ppl_delete_Coefficient (denom); +} What worries me is that PPL finds a maximum, but that is not a valid max, as the denominator is zero. Roberto, could you please look at the bug http://gcc.gnu.org/PR40981 ? Thanks, Sebastian > Program exited with code 04. > (gdb) bt > No stack. > > so I am no longer able to get a back trace. That's because you do not load the gdbinit.in from the gcc dir, where you have the following breakpoints: # Put breakpoints at exit and fancy_abort in case abort is mapped # to either fprintf/exit or fancy_abort. b fancy_abort # Put a breakpoint on internal_error to help with debugging ICEs. b internal_error

3 2

Re: [PPL-devel] Undefined PPL references attempting to build cloog-ppl
by Tovrea, George W (US SSA) 11 Aug '09

11 Aug '09

Yes, I have looked at that file (ppl_interface_instantiations.m4) and the names are messed up in there, as you can see: m4_define(`m4_interface_classes_names', `Polyhedron@Grid@Rational_Box@BD_Shape_mpz_class@_BD_Shape_mpq_class@_Oc tagonal_Shape_mpz_class@_Octagonal_Shape_mpq_class@_Constraints_Product_ C_Polyhedron_Grid@_Pointset_Powerset_C_Polyhedron@_Pointset_Powerset_NNC _Polyhedron') m4_define(`m4_cplusplus_classes_names', `Polyhedron@Grid@Rational_Box@BD_Shape<mpz_class>@ BD_Shape<mpq_class>@ Octagonal_Shape<mpz_class>@ Octagonal_Shape<mpq_class>@ Constraints_Product<C_Polyhedron,Grid>@ Pointset_Powerset<C_Polyhedron>@ Pointset_Powerset<NNC_Polyhedron>') -----Original Message----- From: P M Hill [mailto:hill@comp.leeds.ac.uk] Sent: Tuesday, August 11, 2009 1:31 PM To: Tovrea, George W (US SSA) Cc: Roberto Bagnara; The Parma Polyhedra Library developers' list; Sebastian Pop Subject: Re: [PPL-devel] Undefined PPL references attempting to build cloog-ppl Here are a few observations wrt the code you have sent and maybe they might help us find the bug. I have noticed in your snippets of code that the interface names with the extra "_" are always the same - and it is added before the main root. (Either "_" precedes the name or "ppl_c__" precedes the name where the double "__" should be a single "_".) A list of these names is built by the configure as the default interface classes and are placed in the Makefile in the interfaces directory. When make is executed, the Makefile generates a file called ppl_interface_instantiations.m4 in the interfaces directory. Assuming the default configuration is used, this should contain the following: m4_define(`m4_interface_classes_names', `Polyhedron@Grid@Rational_Box@BD_Shape_mpz_class@BD_Shape_mpq_class@Octa gonal_Shape_mpz_class@Octagonal_Shape_mpq_class@Constraints_Product_C_Po lyhedron_Grid@Pointset_Powerset_C_Polyhedron@Pointset_Powerset_NNC_Polyh edron@Double_Box@BD_Shape_double@Octagonal_Shape_double') m4_define(`m4_cplusplus_classes_names', `Polyhedron@Grid@Rational_Box@BD_Shape<mpz_class>@BD_Shape<mpq_class>@Oc tagonal_Shape<mpz_class>@Octagonal_Shape<mpq_class>@Constraints_Product< C_Polyhedron,Grid>@Pointset_Powerset<C_Polyhedron>@Pointset_Powerset<NNC _Polyhedron>@Double_Box@BD_Shape<double>@Octagonal_Shape<double>') and is then used by the m4 to generate the C code for each of these classes. So it may be helpful to check that if one or both of these generated files have the names correctly (ie as above). I suspect that some names may already be buggy and have the extra "_". In which case, it could be autoreconf or even make that could be the problem and not m4. HTH Pat On Tue, 11 Aug 2009, Tovrea, George W (US SSA) wrote: > In a previous mail, you wrote: > > Strange: this is the same version I use. I don't know what to say: > in ppl_c.h you should end up with > > PPL_TYPE_DECLARATION(Polyhedron) > PPL_TYPE_DECLARATION(Grid) > PPL_TYPE_DECLARATION(Rational_Box) > PPL_TYPE_DECLARATION(BD_Shape_mpz_class) > PPL_TYPE_DECLARATION(BD_Shape_mpq_class) > PPL_TYPE_DECLARATION(Octagonal_Shape_mpz_class) > PPL_TYPE_DECLARATION(Octagonal_Shape_mpq_class) > PPL_TYPE_DECLARATION(Constraints_Product_C_Polyhedron_Grid) > PPL_TYPE_DECLARATION(Pointset_Powerset_C_Polyhedron) > PPL_TYPE_DECLARATION(Pointset_Powerset_NNC_Polyhedron) > > whereas you have > > PPL_TYPE_DECLARATION(Polyhedron) > PPL_TYPE_DECLARATION(Grid) > PPL_TYPE_DECLARATION(Rational_Box) > PPL_TYPE_DECLARATION(BD_Shape_mpz_class) > PPL_TYPE_DECLARATION(_BD_Shape_mpq_class) > PPL_TYPE_DECLARATION(_Octagonal_Shape_mpz_class) > PPL_TYPE_DECLARATION(_Octagonal_Shape_mpq_class) > PPL_TYPE_DECLARATION(_Constraints_Product_C_Polyhedron_Grid) > PPL_TYPE_DECLARATION(_Pointset_Powerset_C_Polyhedron) > PPL_TYPE_DECLARATION(_Pointset_Powerset_NNC_Polyhedron) > > Note that all the declarations that are in error above have a double > underscore after ppl_c in the snippet below. All the ones that are > correct have a single underscore after ppl_c. For example, > BD_Shape_mpz_class above relates to ppl_c_BD_Shape_mpz_class whereas > _BD_Shape_mpq_class above relates to ppl_c__BD_Shape_mpq_class below. > Seems more than coincidental. However, I modified config.status changing > all ppl_c__ tp ppl_c_. This did not build (still looking for some > ppl_c__ files). I am currently trying to debug > > /net/whistler/auto/users/tovrea/local/sol2/bin/m4 --prefix-builtin -I.. > -I../../../interfaces/C -I../../../interfaces/C/.. \ > ../../../interfaces/C/ppl_interface_generator_c_h.m4 > > ppl_c_domains.h > > > ppl_c_domains.h contains all the ppl_c__ names > > > -----Original Message----- > From: Roberto Bagnara [mailto:bagnara@cs.unipr.it] > Sent: Tuesday, August 11, 2009 12:54 AM > To: Tovrea, George W (US SSA) > Cc: Sebastian Pop; The Parma Polyhedra Library developers' list > Subject: Re: [PPL-devel] Undefined PPL references attempting to build > cloog-ppl > > Tovrea, George W (US SSA) wrote: >> The names are already messed up after running configure. From > config.status: >> >> S["required_instantiations_c_cxx_objects"]="ppl_c_Polyhedron.lo > ppl_c_Grid.lo ppl_c_Rational_Box.lo ppl_c_BD_Shape_mpz_class.lo > ppl_c__BD_Shape_mpq_class.lo ppl_c__Octagonal_Shape_mpz_class.lo"\ >> " ppl_c__Octagonal_Shape_mpq_class.lo > ppl_c__Constraints_Product_C_Polyhedron_Grid.lo > ppl_c__Pointset_Powerset_C_Polyhedron.lo ppl_c__Pointset_Powers"\ >> "et_NNC_Polyhedron.lo" >> S["required_instantiations_c_cxx_sources"]="ppl_c_Polyhedron.cc > ppl_c_Grid.cc ppl_c_Rational_Box.cc ppl_c_BD_Shape_mpz_class.cc > ppl_c__BD_Shape_mpq_class.cc ppl_c__Octagonal_Shape_mpz_class.cc"\ >> " ppl_c__Octagonal_Shape_mpq_class.cc > ppl_c__Constraints_Product_C_Polyhedron_Grid.cc > ppl_c__Pointset_Powerset_C_Polyhedron.cc ppl_c__Pointset_Powers"\ >> "et_NNC_Polyhedron.cc" >> S["required_instantiations_c_cxx_headers"]="ppl_c_Polyhedron.hh > ppl_c_Grid.hh ppl_c_Rational_Box.hh ppl_c_BD_Shape_mpz_class.hh > ppl_c__BD_Shape_mpq_class.hh ppl_c__Octagonal_Shape_mpz_class.hh"\ >> " ppl_c__Octagonal_Shape_mpq_class.hh > ppl_c__Constraints_Product_C_Polyhedron_Grid.hh > ppl_c__Pointset_Powerset_C_Polyhedron.hh ppl_c__Pointset_Powers"\ >> "et_NNC_Polyhedron.hh" > > Hi Bill, > > I see nothing messed up in this snippet. > Can you explain? > All the best, > > Roberto > >

2 1

Re: [PPL-devel] Undefined PPL references attempting to build cloog-ppl
by Tovrea, George W (US SSA) 11 Aug '09

11 Aug '09

In a previous mail, you wrote: Strange: this is the same version I use. I don't know what to say: in ppl_c.h you should end up with PPL_TYPE_DECLARATION(Polyhedron) PPL_TYPE_DECLARATION(Grid) PPL_TYPE_DECLARATION(Rational_Box) PPL_TYPE_DECLARATION(BD_Shape_mpz_class) PPL_TYPE_DECLARATION(BD_Shape_mpq_class) PPL_TYPE_DECLARATION(Octagonal_Shape_mpz_class) PPL_TYPE_DECLARATION(Octagonal_Shape_mpq_class) PPL_TYPE_DECLARATION(Constraints_Product_C_Polyhedron_Grid) PPL_TYPE_DECLARATION(Pointset_Powerset_C_Polyhedron) PPL_TYPE_DECLARATION(Pointset_Powerset_NNC_Polyhedron) whereas you have PPL_TYPE_DECLARATION(Polyhedron) PPL_TYPE_DECLARATION(Grid) PPL_TYPE_DECLARATION(Rational_Box) PPL_TYPE_DECLARATION(BD_Shape_mpz_class) PPL_TYPE_DECLARATION(_BD_Shape_mpq_class) PPL_TYPE_DECLARATION(_Octagonal_Shape_mpz_class) PPL_TYPE_DECLARATION(_Octagonal_Shape_mpq_class) PPL_TYPE_DECLARATION(_Constraints_Product_C_Polyhedron_Grid) PPL_TYPE_DECLARATION(_Pointset_Powerset_C_Polyhedron) PPL_TYPE_DECLARATION(_Pointset_Powerset_NNC_Polyhedron) Note that all the declarations that are in error above have a double underscore after ppl_c in the snippet below. All the ones that are correct have a single underscore after ppl_c. For example, BD_Shape_mpz_class above relates to ppl_c_BD_Shape_mpz_class whereas _BD_Shape_mpq_class above relates to ppl_c__BD_Shape_mpq_class below. Seems more than coincidental. However, I modified config.status changing all ppl_c__ tp ppl_c_. This did not build (still looking for some ppl_c__ files). I am currently trying to debug /net/whistler/auto/users/tovrea/local/sol2/bin/m4 --prefix-builtin -I.. -I../../../interfaces/C -I../../../interfaces/C/.. \ ../../../interfaces/C/ppl_interface_generator_c_h.m4 > ppl_c_domains.h ppl_c_domains.h contains all the ppl_c__ names -----Original Message----- From: Roberto Bagnara [mailto:bagnara@cs.unipr.it] Sent: Tuesday, August 11, 2009 12:54 AM To: Tovrea, George W (US SSA) Cc: Sebastian Pop; The Parma Polyhedra Library developers' list Subject: Re: [PPL-devel] Undefined PPL references attempting to build cloog-ppl Tovrea, George W (US SSA) wrote: > The names are already messed up after running configure. From config.status: > > S["required_instantiations_c_cxx_objects"]="ppl_c_Polyhedron.lo ppl_c_Grid.lo ppl_c_Rational_Box.lo ppl_c_BD_Shape_mpz_class.lo ppl_c__BD_Shape_mpq_class.lo ppl_c__Octagonal_Shape_mpz_class.lo"\ > " ppl_c__Octagonal_Shape_mpq_class.lo ppl_c__Constraints_Product_C_Polyhedron_Grid.lo ppl_c__Pointset_Powerset_C_Polyhedron.lo ppl_c__Pointset_Powers"\ > "et_NNC_Polyhedron.lo" > S["required_instantiations_c_cxx_sources"]="ppl_c_Polyhedron.cc ppl_c_Grid.cc ppl_c_Rational_Box.cc ppl_c_BD_Shape_mpz_class.cc ppl_c__BD_Shape_mpq_class.cc ppl_c__Octagonal_Shape_mpz_class.cc"\ > " ppl_c__Octagonal_Shape_mpq_class.cc ppl_c__Constraints_Product_C_Polyhedron_Grid.cc ppl_c__Pointset_Powerset_C_Polyhedron.cc ppl_c__Pointset_Powers"\ > "et_NNC_Polyhedron.cc" > S["required_instantiations_c_cxx_headers"]="ppl_c_Polyhedron.hh ppl_c_Grid.hh ppl_c_Rational_Box.hh ppl_c_BD_Shape_mpz_class.hh ppl_c__BD_Shape_mpq_class.hh ppl_c__Octagonal_Shape_mpz_class.hh"\ > " ppl_c__Octagonal_Shape_mpq_class.hh ppl_c__Constraints_Product_C_Polyhedron_Grid.hh ppl_c__Pointset_Powerset_C_Polyhedron.hh ppl_c__Pointset_Powers"\ > "et_NNC_Polyhedron.hh" Hi Bill, I see nothing messed up in this snippet. Can you explain? 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