April 2010 - PPL-devel - BUGSENG Mailman

[GIT] ppl/ppl(master): Tentative release date is May 8, 2010. Reorganized and improved some news items.
by Roberto Bagnara 10 Apr '10

10 Apr '10

Module: ppl/ppl Branch: master Commit: 31f6a807bebbc51442f62c1bd1a0c2d71eb3285a URL: http://www.cs.unipr.it/git/gitweb.cgi?p=ppl/ppl.git;a=commit;h=31f6a807bebb… Author: Roberto Bagnara <bagnara(a)cs.unipr.it> Date: Sat Apr 10 11:33:41 2010 +0200 Tentative release date is May 8, 2010. Reorganized and improved some news items. --- NEWS | 78 +++++++++++++++++++++++++++++++++-------------------------------- 1 files changed, 40 insertions(+), 38 deletions(-) diff --git a/NEWS b/NEWS index 74781bc..88347c0 100644 --- a/NEWS +++ b/NEWS @@ -4,45 +4,24 @@ Parma Polyhedra Library NEWS -- history of user-visible changes -------------------------------------------------------------------------- -NEWS for version 0.11 (released date to be decided) +NEWS for version 0.11 (to be released on May 8, 2010) -------------------------------------------------------------------------- New and Changed Features ======================== -o Implementation of general (C and NNC) polyhedra speeded up. - -o When the PPL has been configured with - CPPFLAGS="-DPPL_ARM_CAN_CONTROL_FPU=1", the library initialization - procedure checks that the FPU can indeed be controlled and fails if - that is not the case. - -o New configure option --with-gmp-prefix supersedes the (now removed) - options --with-libgmp-prefix and --with-libgmpxx-prefix. +o New class PIP_Problem provides a Parametric Integer Programming + (PIP) problem solver (mainly based on P. Feautrier's + specification). The implementation combines a parametric dual + simplex algorithm using exact arithmetic with Gomory's cut + generation. + Still under beta testing. o New "deterministic" timeout computation facilities: it is now possible to set computational bounds (on the library calls taking exponential time) that do not depend on the actual elapsed time and hence are independent from the actual computation environment (CPU, operating system, etc.). - Still under alpha testing. - -o The Java interface has to be explicitly initialized before use by - calling static method Parma_Polyhedra_Library.initialize_library(). - Initialization makes more explicit the exact point where PPL - floating point rounding mode is set; it also allows for the caching - of Java classes and field/method IDs, thereby reducing the overhead - of native method callbacks. - -o The C and Java interfaces now support timeout computation facilities. - -o New configuration option `--with-gmp-build=DIR' allows to use a - non-installed build of GMP in DIR. - -o New class PIP_Problem provides a Parametric Integer Programming (PIP) - problem solver mainly based on P. Feautrier's specification. - The implementation combines a parametric dual simplex algorithm using - exact arithmetic with Gomory's cut generation. - Still under beta testing. + Still under beta testing. o New Linear_Expression methods @@ -71,14 +50,37 @@ o For the PPL semantic objects, other than the Pointset_Powerset and and a value (v = val_n/val_d) closest to zero so that every point in the object satisfies the congruence (expr %= v) / f. -o New reduction operator "Shape_Preserving_Reduction" has been added to - the Partially_Reduced_Product abstraction. - This operator is aimed at the product of a grid and a shape domain; - allowing the bounds of the shape to shrink to the touch the points - of the grid, such that the new bounds are parallel to the old bounds. +o New reduction operator "Shape_Preserving_Reduction" has been added + to the Partially_Reduced_Product abstraction. This operator is + aimed at the product of a grid and a shape domain, allowing the + bounds of the shape to shrink to touch the points of the grid, + such that the new bounds are parallel to the old bounds. + +o The Java interface has to be explicitly initialized before use by + calling static method Parma_Polyhedra_Library.initialize_library(). + Initialization makes more explicit the exact point where PPL + floating point rounding mode is set; it also allows for the caching + of Java classes and field/method IDs, thereby reducing the overhead + of native method callbacks. + +o The C and Java interfaces now support timeout computation facilities. + +o Implementation of general (C and NNC) polyhedra speeded up. + +o When the PPL has been configured with + CPPFLAGS="-DPPL_ARM_CAN_CONTROL_FPU=1", the library initialization + procedure checks that the FPU can indeed be controlled and fails if + that is not the case. + +o New configure option --with-gmp-prefix supersedes the (now removed) + options --with-libgmp-prefix and --with-libgmpxx-prefix. + +o New configuration option `--with-gmp-build=DIR' allows to use a + non-installed build of GMP in DIR. + -Deprecated and removed methods ------------------------------- +Deprecated and Removed Methods +============================== o All methods having a name ending in `_and_minimize' (e.g., add_constraints_and_minimize, poly_hull_assign_and_minimize, ...) @@ -441,8 +443,8 @@ o The methods (i.e., it has no constraints, generators, congruences or grid-generators, respectively). -Deprecated and removed methods ------------------------------- +Deprecated and Removed Methods +============================== o As all PPL semantic objects can now be constructed from boxes, the constructors

1 0

[GIT] ppl/ppl(master): Man pages updated.
by Roberto Bagnara 10 Apr '10

10 Apr '10

Module: ppl/ppl Branch: master Commit: e77b2da3be6919dd3e18d85ecc28b7d825e0fb98 URL: http://www.cs.unipr.it/git/gitweb.cgi?p=ppl/ppl.git;a=commit;h=e77b2da3be69… Author: Roberto Bagnara <bagnara(a)cs.unipr.it> Date: Sat Apr 10 11:25:01 2010 +0200 Man pages updated. --- TODO | 6 ------ demos/ppl_lcdd/ppl_lcdd.1 | 2 +- demos/ppl_lpsol/ppl_lpsol.1 | 2 +- demos/ppl_pips/ppl_pips.1 | 20 ++++++++++---------- doc/ppl-config.1 | 2 +- 5 files changed, 13 insertions(+), 19 deletions(-) diff --git a/TODO b/TODO index a617803..76d679e 100644 --- a/TODO +++ b/TODO @@ -1,10 +1,4 @@ -Enhancements for PPL 0.11 -========================= - -- Complete the ppl_pips man page. - - Enhancements for PPL 0.12 or later versions =========================================== diff --git a/demos/ppl_lcdd/ppl_lcdd.1 b/demos/ppl_lcdd/ppl_lcdd.1 index e22d7e6..f3737cb 100644 --- a/demos/ppl_lcdd/ppl_lcdd.1 +++ b/demos/ppl_lcdd/ppl_lcdd.1 @@ -1,5 +1,5 @@ .\" DO NOT MODIFY THIS FILE! It was generated by help2man 1.36. -.TH PPL_LCDD "1" "April 2010" "ppl_lcdd 0.11" "User Commands" +.TH PPL_LCDD "1" "May 2010" "ppl_lcdd 0.11" "User Commands" .SH NAME ppl_lcdd \- a PPL-based program for vertex/facet enumeration of convex polyhedra .SH SYNOPSIS diff --git a/demos/ppl_lpsol/ppl_lpsol.1 b/demos/ppl_lpsol/ppl_lpsol.1 index a2d6174..db6b5fd 100644 --- a/demos/ppl_lpsol/ppl_lpsol.1 +++ b/demos/ppl_lpsol/ppl_lpsol.1 @@ -1,5 +1,5 @@ .\" DO NOT MODIFY THIS FILE! It was generated by help2man 1.36. -.TH PPL_LPSOL "1" "April 2010" "ppl_lpsol 0.11" "User Commands" +.TH PPL_LPSOL "1" "May 2010" "ppl_lpsol 0.11" "User Commands" .SH NAME ppl_lpsol \- a PPL-based mixed integer programming problem solver .SH SYNOPSIS diff --git a/demos/ppl_pips/ppl_pips.1 b/demos/ppl_pips/ppl_pips.1 index 58b6fee..6b8806b 100644 --- a/demos/ppl_pips/ppl_pips.1 +++ b/demos/ppl_pips/ppl_pips.1 @@ -1,5 +1,5 @@ .\" DO NOT MODIFY THIS FILE! It was generated by help2man 1.36. -.TH PPL_PIPS "1" "April 2010" "ppl_pips 0.11" "User Commands" +.TH PPL_PIPS "1" "May 2010" "ppl_pips 0.11" "User Commands" .SH NAME ppl_pips \- a PPL-based parametric integer programming problem solver .SH SYNOPSIS @@ -21,10 +21,10 @@ prints this help text to stdout appends output to PATH .TP \fB\-P\fR, \fB\-\-polylib\fR -read problem in Polylib format (default) +reads problem in PolyLib format (default) .TP \fB\-p\fR, \fB\-\-piplib\fR -read problem in PIPlib format +reads problem in PipLib format .TP \fB\-t\fR, \fB\-\-timings\fR prints timings to stderr @@ -33,7 +33,7 @@ prints timings to stderr produces lots of output .TP \fB\-i\fR, \fB\-\-iterations\fR=\fIN\fR -execute the resolution N times (default=1) +executes the resolution N times (default=1) .TP \fB\-V\fR, \fB\-\-version\fR prints version information to stdout @@ -43,20 +43,20 @@ checks if the result is equal to what is in PATH .SS "Cut generation options:" .TP \fB\-f\fR, \fB\-\-cut\-first\fR -use the first non\-integer row (default) +uses the first non\-integer row (default) .TP \fB\-d\fR, \fB\-\-cut\-deepest\fR -try to generate the deepest cut +tries to generate the deepest cut .TP \fB\-a\fR, \fB\-\-cut\-all\fR -always generate all possible cuts +always generates all possible cuts .SS "Pivot row strategy options:" .TP \fB\-F\fR, \fB\-\-row\-first\fR -use the first row with negative parameter (default) +uses the first row with negative parameter (default) .TP \fB\-M\fR, \fB\-\-row\-max\fR -choose row generating the lexico\-maximal pivot column +chooses row generating the lexico\-maximal pivot column .SH AVAILABILITY The latest version of the Parma Polyhedra Library and all the documentation is available at \fBhttp://www.cs.unipr.it/ppl/\fR . @@ -93,7 +93,7 @@ The following row contains the parameter number for the so-called If no big parameter is used, the value is -1. .TP - -The following rows encode the problem inequality matrix.. +The following rows encode the problem inequality matrix. As for the context matrix, the first two values are the dimensions of the matrix. The number of variables is the number of columns in the matrix minus the number of parameters minus 2. diff --git a/doc/ppl-config.1 b/doc/ppl-config.1 index e08658a..5e54525 100644 --- a/doc/ppl-config.1 +++ b/doc/ppl-config.1 @@ -1,5 +1,5 @@ .\" DO NOT MODIFY THIS FILE! It was generated by help2man 1.36. -.TH PPL-CONFIG "1" "April 2010" "ppl-config 0.11" "User Commands" +.TH PPL-CONFIG "1" "May 2010" "ppl-config 0.11" "User Commands" .SH NAME ppl-config \- obtain information about an installation of the Parma Polyhedra Library .SH SYNOPSIS

1 0

[GIT] ppl/ppl(floating_point): Fixed Documentation.
by Fabio Biselli 09 Apr '10

09 Apr '10

Module: ppl/ppl Branch: floating_point Commit: b22579cc1c086702eb713365ea49e5c8783f3ace URL: http://www.cs.unipr.it/git/gitweb.cgi?p=ppl/ppl.git;a=commit;h=b22579cc1c08… Author: Fabio Biselli <fabio.biselli(a)studenti.unipr.it> Date: Fri Apr 9 10:25:46 2010 +0200 Fixed Documentation. --- src/Cast_Floating_Point_Expression.defs.hh | 2 +- src/Constant_Floating_Point_Expression.defs.hh | 2 +- src/Difference_Floating_Point_Expression.defs.hh | 2 +- src/Division_Floating_Point_Expression.defs.hh | 2 +- src/Floating_Point_Expression.defs.hh | 8 ++++---- ...ultiplication_Floating_Point_Expression.defs.hh | 2 +- src/Sum_Floating_Point_Expression.defs.hh | 2 +- src/Variable_Floating_Point_Expression.defs.hh | 2 +- 8 files changed, 11 insertions(+), 11 deletions(-) diff --git a/src/Cast_Floating_Point_Expression.defs.hh b/src/Cast_Floating_Point_Expression.defs.hh index 5dacccc..c02f3a7 100644 --- a/src/Cast_Floating_Point_Expression.defs.hh +++ b/src/Cast_Floating_Point_Expression.defs.hh @@ -105,7 +105,7 @@ public: FP_Linear_Form FP_Linear_Form; /*! \brief - Alias for the std::map<dimension_type, FP_Interval_Type> from + Alias for the Box<FP_Interval_Type> from Floating_Point_Expression. */ typedef typename diff --git a/src/Constant_Floating_Point_Expression.defs.hh b/src/Constant_Floating_Point_Expression.defs.hh index 5036e75..190f86d 100644 --- a/src/Constant_Floating_Point_Expression.defs.hh +++ b/src/Constant_Floating_Point_Expression.defs.hh @@ -77,7 +77,7 @@ public: FP_Linear_Form FP_Linear_Form; /*! \brief - Alias for the std::map<dimension_type, FP_Interval_Type> from + Alias for the Box<FP_Interval_Type> from Floating_Point_Expression. */ typedef typename diff --git a/src/Difference_Floating_Point_Expression.defs.hh b/src/Difference_Floating_Point_Expression.defs.hh index 1d1ab02..5fb74c6 100644 --- a/src/Difference_Floating_Point_Expression.defs.hh +++ b/src/Difference_Floating_Point_Expression.defs.hh @@ -118,7 +118,7 @@ public: ::FP_Linear_Form FP_Linear_Form; /*! \brief - Alias for the std::map<dimension_type, FP_Interval_Type> from + Alias for the Box<FP_Interval_Type> from Floating_Point_Expression. */ typedef typename diff --git a/src/Division_Floating_Point_Expression.defs.hh b/src/Division_Floating_Point_Expression.defs.hh index ea4ebae..56f64f6 100644 --- a/src/Division_Floating_Point_Expression.defs.hh +++ b/src/Division_Floating_Point_Expression.defs.hh @@ -134,7 +134,7 @@ public: ::FP_Linear_Form FP_Linear_Form; /*! \brief - Alias for the std::map<dimension_type, FP_Interval_Type> from + Alias for the Box<FP_Interval_Type> from Floating_Point_Expression. */ typedef typename diff --git a/src/Floating_Point_Expression.defs.hh b/src/Floating_Point_Expression.defs.hh index 5a25e9f..3f608d8 100644 --- a/src/Floating_Point_Expression.defs.hh +++ b/src/Floating_Point_Expression.defs.hh @@ -130,8 +130,8 @@ public: Verifies if a given linear form overflows. \param lf The linear form to verify. \return - Returns <CODE>true</CODE> if all coefficients in \p lf are bounded, - <CODE>false</CODE> otherwise. + Returns <CODE>false</CODE> if all coefficients in \p lf are bounded, + <CODE>true</CODE> otherwise. */ static bool overflows(const FP_Linear_Form& lf); @@ -146,7 +146,7 @@ public: This method makes <CODE>result</CODE> become a linear form obtained by evaluating the function \f$\varepsilon_{\mathbf{f}}(l)\f$ - on the linear form \p lf. This function is defined as: + on the linear form \p lf. This function is defined such as: \f[ \varepsilon_{\mathbf{f}}\left([a;b]+\sum_{v \in \cV}[a_{v};b_{v}]v\right) @@ -192,7 +192,7 @@ private: Static helper method that is used to compute the value of the public static field <CODE>absolute_error</CODE>. - \return the interval \f$[-\omega; \omega]\f$ corresponding to the value + \return The interval \f$[-\omega; \omega]\f$ corresponding to the value of <CODE>absolute_error</CODE> */ static FP_Interval_Type compute_absolute_error(); diff --git a/src/Multiplication_Floating_Point_Expression.defs.hh b/src/Multiplication_Floating_Point_Expression.defs.hh index e2d994a..9630bed 100644 --- a/src/Multiplication_Floating_Point_Expression.defs.hh +++ b/src/Multiplication_Floating_Point_Expression.defs.hh @@ -148,7 +148,7 @@ public: ::FP_Linear_Form FP_Linear_Form; /*! \brief - Alias for the std::map<dimension_type, FP_Interval_Type> from + Alias for the Box<FP_Interval_Type> from Floating_Point_Expression. */ typedef typename diff --git a/src/Sum_Floating_Point_Expression.defs.hh b/src/Sum_Floating_Point_Expression.defs.hh index f9bb6a7..4cd1009 100644 --- a/src/Sum_Floating_Point_Expression.defs.hh +++ b/src/Sum_Floating_Point_Expression.defs.hh @@ -111,7 +111,7 @@ public: ::FP_Linear_Form FP_Linear_Form; /*! \brief - Alias for the std::map<dimension_type, FP_Interval_Type> from + Alias for the Box<FP_Interval_Type> from Floating_Point_Expression. */ typedef typename diff --git a/src/Variable_Floating_Point_Expression.defs.hh b/src/Variable_Floating_Point_Expression.defs.hh index 8c7eac7..1d0025b 100644 --- a/src/Variable_Floating_Point_Expression.defs.hh +++ b/src/Variable_Floating_Point_Expression.defs.hh @@ -80,7 +80,7 @@ public: FP_Linear_Form FP_Linear_Form; /*! \brief - Alias for the std::map<dimension_type, FP_Interval_Type> from + Alias for the Box<FP_Interval_Type> from Floating_Point_Expression. */ typedef typename

1 0

Re: [PPL-devel] Linear ranking functions/relations in PPL
by Roberto Bagnara 08 Apr '10

08 Apr '10

On 04/08/10 13:50, Aziem Chawdhary wrote: > Dear Prof. Roberto Bagnara and Prof. Enea Zaffanella, > > I have just seen your very interesting paper on linear ranking functions > on Arxiv. > The paper mentions that your algorithms are implemented in PPL, but I > cannot find a reference to this in the documentation for PPL. > > Are the implementations in the git version of PPL and is there any > documentation for using the functions to generate ranking relations? > > Any pointers would be very much appreciated. > > Thanks, > Aziem Chawdhary Dear Aziem, yes, the implementation is in the Git repository. From time to time we also post a release snapshot at http://www.cs.unipr.it/ppl/Download/ftp/snapshots/ The paper together with the PPL documentation contain all the information you need. However, please do not hesitate to ask on ppl-devel(a)cs.unipr.it if you need help. All the best, Roberto P.S. I will post an updated snapshot later today. I will mail you when this is available. -- Prof. Roberto Bagnara Applied Formal Methods Laboratory Department of Mathematics, University of Parma, Italy http://www.cs.unipr.it/~bagnara/ mailto:bagnara@cs.unipr.it

1 1

[GIT] ppl/ppl(master): Renamed.
by Roberto Bagnara 08 Apr '10

08 Apr '10

Module: ppl/ppl Branch: master Commit: 6d23fe5df7c16842e70e349886de0d9faf70affa URL: http://www.cs.unipr.it/git/gitweb.cgi?p=ppl/ppl.git;a=commit;h=6d23fe5df7c1… Author: Roberto Bagnara <bagnara(a)cs.unipr.it> Date: Thu Apr 8 22:42:20 2010 +0200 Renamed. --- README.Solaris => README.solaris | 0 1 files changed, 0 insertions(+), 0 deletions(-) diff --git a/README.Solaris b/README.solaris similarity index 100% rename from README.Solaris rename to README.solaris

1 0

[GIT] ppl/ppl(master): Pushed one item to PPL 0.12. We will not have 0.10. 3 release.
by Roberto Bagnara 08 Apr '10

08 Apr '10

Module: ppl/ppl Branch: master Commit: b8ca985398d213b28930c8dc29a1aca72158da54 URL: http://www.cs.unipr.it/git/gitweb.cgi?p=ppl/ppl.git;a=commit;h=b8ca985398d2… Author: Roberto Bagnara <bagnara(a)cs.unipr.it> Date: Thu Apr 8 22:41:02 2010 +0200 Pushed one item to PPL 0.12. We will not have 0.10.3 release. --- TODO | 24 ++---------------------- 1 files changed, 2 insertions(+), 22 deletions(-) diff --git a/TODO b/TODO index 3ade302..a617803 100644 --- a/TODO +++ b/TODO @@ -2,34 +2,14 @@ Enhancements for PPL 0.11 ========================= -- Make all the *affine*image() methods uniform as far as the - specification is concerned. - Complete the ppl_pips man page. -Commits to consider should we release PPL 0.10.3 -================================================ - -b4dcb2959e3cc7b74aaa9f20c865b848e206818b -(Fixed PPL_SPECIALIZE_LIMITS_INT.) - -9a08d1d3a98ca1af8ff0cb94017888528c6875a7 -(No longer use CS and PS as template parameter names.) - -e3460918400e04cf24fda33ff0cf9d18a95840b8 -(Updated from Gnulib.) - -b94dc7951129d065d858b6dbd932eac59a643f96 -(We cannot test the Java interface unless ENABLE_SHARED.) - -1c8a74aea14bce95dc1916df66995dd5a163eb2b -(New configuration option `--with-gmp-build=DIR'. -This allows to use a non-installed build of GMP in DIR.) - - Enhancements for PPL 0.12 or later versions =========================================== +- Make all the *affine*image() methods uniform as far as the + specification is concerned. - Add the domain of circular linear progressions. - Finish the domain of intervals. - Intervals are best instantiated with checked numbers with

1 0

[GIT] ppl/ppl(master): Merge remote branch 'origin/master'
by Roberto Bagnara 08 Apr '10

08 Apr '10

Module: ppl/ppl Branch: master Commit: e3d2caf268af5d73af10e6d2a2ab1f7d9a9b58ea URL: http://www.cs.unipr.it/git/gitweb.cgi?p=ppl/ppl.git;a=commit;h=e3d2caf268af… Author: Roberto Bagnara <bagnara(a)cs.unipr.it> Date: Thu Apr 8 22:39:46 2010 +0200 Merge remote branch 'origin/master' Conflicts: TODO ---

1 0

[GIT] ppl/ppl(master): Perl is no longer a strict requirement.
by Roberto Bagnara 08 Apr '10

08 Apr '10

Module: ppl/ppl Branch: master Commit: be262c561929432c22960e53026853d520396d7b URL: http://www.cs.unipr.it/git/gitweb.cgi?p=ppl/ppl.git;a=commit;h=be262c561929… Author: Roberto Bagnara <bagnara(a)cs.unipr.it> Date: Thu Apr 8 22:34:23 2010 +0200 Perl is no longer a strict requirement. --- TODO | 3 --- configure.ac | 26 ++++++++------------------ interfaces/C/Makefile.am | 16 +++++++++++++++- src/Makefile.am | 16 ++++++++++++++-- 4 files changed, 37 insertions(+), 24 deletions(-) diff --git a/TODO b/TODO index ddebb7f..b32e567 100644 --- a/TODO +++ b/TODO @@ -7,9 +7,6 @@ Enhancements for PPL 0.11 - Make all the *affine*image() methods uniform as far as the specification is concerned. - Look carefully at http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42115 -- Lack of perl can cause more problems than explained in the warning - message given by our configuration procedure. See, e.g., - http://www.cs.unipr.it/pipermail/ppl-devel/2009-November/015758.html - Complete the ppl_pips man page. diff --git a/configure.ac b/configure.ac index d56c885..3d0c31c 100644 --- a/configure.ac +++ b/configure.ac @@ -23,7 +23,7 @@ # Process this file with Autoconf to produce a configure script. # Every other copy of the package version number gets its value from here. -AC_INIT([the Parma Polyhedra Library], [0.11pre18], [ppl-devel(a)cs.unipr.it], [ppl]) +AC_INIT([the Parma Polyhedra Library], [0.11pre19], [ppl-devel(a)cs.unipr.it], [ppl]) # Minimum Autoconf version required. AC_PREREQ(2.61) @@ -78,6 +78,8 @@ ISODATE=`date +%Y-%m-%d` AC_SUBST(ISODATE) AH_TOP([ +/* BEGIN ppl-config.h */ + /* Unique (nonzero) code for the IEEE 754 Single Precision floating point format. */ #define PPL_FLOAT_IEEE754_SINGLE 1 @@ -116,6 +118,8 @@ AH_BOTTOM([ #ifdef PPL_HAVE_INTTYPES_H # include <inttypes.h> #endif + +/* END ppl-config.h */ ]) @@ -908,17 +912,13 @@ xall) ;; esac -# Perl is needed to build the C++ interface. -if test x${cxx_interface_enabled} = xyes \ -&& test x"$PERL" != xno +if test x${cxx_interface_enabled} = xyes then build_cxx_interface=yes fi AM_CONDITIONAL(BUILD_CXX_INTERFACE, test x$build_cxx_interface = xyes) -# Perl is needed to build the C interface. -if test x${c_interface_enabled} = xyes \ -&& test x"$PERL" != xno +if test x${c_interface_enabled} = xyes then build_c_interface=yes fi @@ -1592,23 +1592,13 @@ fi if test x"$PERL" = xno then - if test x${c_interface_enabled} = xyes - then - AC_MSG_ERROR([PERL UNAVAILABLE: -*** CANNOT BUILD THE C_INTERFACE. -*** Perl could not be found on your system, so that the C interface cannot -*** be built. You should either disable the generation of the C interface -*** (with the `--enable-interfaces' configure option) or add Perl to your -*** development environment: see http://www.perl.org/ for more information.]) - else - AC_MSG_WARN([PERL UNAVAILABLE: + AC_MSG_WARN([PERL UNAVAILABLE: *** CANNOT REBUILD SOME FILES IF SOURCES ARE CHANGED. *** Perl could not be found on your system. This is OK, if you only plan *** to build the PPL without modifying the sources and without trying to *** build a distribution out of a CVS working copy. If this is not the case, *** then you need to add Perl to your development environment: *** see http://www.perl.org/ for more information.]) - fi fi if test x"$coefficient_kind" = xnative diff --git a/interfaces/C/Makefile.am b/interfaces/C/Makefile.am index 33dfc7c..ea27234 100644 --- a/interfaces/C/Makefile.am +++ b/interfaces/C/Makefile.am @@ -84,16 +84,30 @@ WATCHDOG_INCLUDE_OPTIONS = -I$(top_builddir)/Watchdog/src endif BUILD_WATCHDOG_LIBRARY +if HAVE_PERL + ppl_c.h: ppl_c_header.h ppl_c_version.h ppl_c_domains.h Makefile $(top_builddir)/utils/build_header $(top_builddir)/utils/build_header \ -I $(top_builddir)/interfaces/C -I $(top_srcdir)/src \ $(top_srcdir)/interfaces/C/ppl_c_header.h >$@ +else !HAVE_PERL + +ppl_c.h: ppl_c.h.dist + cp -f $< $@ + +endif !HAVE_PERL + BUILT_SOURCES = \ ppl_c.h \ ppl_c_implementation_domains.hh \ ppl_c_implementation_domains.cc.stamp +# ppl_c.h is not distributed. +# ppl_c.h.dist, which is distributed, is a copy of ppl_c.h. +dist-hook: + mv -f $(distdir)/ppl_c.h $(distdir)/ppl_c.h.dist + interface_generator_dependencies = \ $(top_builddir)/interfaces/ppl_interface_instantiations.m4 \ ../ppl_interface_generator_common.m4 \ @@ -231,7 +245,7 @@ libppl_c_la_LDFLAGS = \ $(NO_UNDEFINED_FLAG) \ -version-info $(LIBPPL_C_LT_CURRENT):$(LIBPPL_C_LT_REVISION):$(LIBPPL_C_LT_AGE) -nodist_include_HEADERS = ppl_c.h +include_HEADERS = ppl_c.h nodist_noinst_HEADERS = \ ppl_c_domains.h \ diff --git a/src/Makefile.am b/src/Makefile.am index d76fbfe..5d20cf7 100644 --- a/src/Makefile.am +++ b/src/Makefile.am @@ -523,8 +523,20 @@ CREDITS.cc: $(top_builddir)/utils/text2cxxarray $(top_srcdir)/CREDITS else !HAVE_PERL -ppl.hh: ppl.hh.dist - cp -f $< $@ +ppl.hh.sed: Makefile + @echo "/^\/\* BEGIN ppl-config.h \*\/$$/ {" >$@ + @echo " r $(top_builddir)/ppl-config.h" >>$@ + @echo " : loop" >>$@ + @echo " /\n\/\* END ppl-config.h \*\/$$/ !{" >>$@ + @echo " N" >>$@ + @echo " b loop" >>$@ + @echo " }" >>$@ + @echo " d" >>$@ + @echo "}" >>$@ + +ppl.hh: ppl.hh.dist ppl.hh.sed + $(SED) -f ppl.hh.sed $< >$@ + rm -f ppl.hh.sed BUGS.cc: BUGS.cc.dist cp -f $< $@

1 0

[GIT] ppl/ppl(sparse_matrices): Unlimited_Sparse_Row: fix combine_needs_second() for backends that define PPL_SPARSE_BACKEND_INVALIDATES_REFERENCES.
by Marco Poletti 08 Apr '10

08 Apr '10

Module: ppl/ppl Branch: sparse_matrices Commit: ba97824c0307ae2b5534a04ccb0196444f3e400c URL: http://www.cs.unipr.it/git/gitweb.cgi?p=ppl/ppl.git;a=commit;h=ba97824c0307… Author: Marco Poletti <poletti.marco(a)gmail.com> Date: Thu Apr 8 19:00:01 2010 +0200 Unlimited_Sparse_Row: fix combine_needs_second() for backends that define PPL_SPARSE_BACKEND_INVALIDATES_REFERENCES. --- src/Unlimited_Sparse_Row.templates.hh | 30 ++++++++++++------------------ 1 files changed, 12 insertions(+), 18 deletions(-) diff --git a/src/Unlimited_Sparse_Row.templates.hh b/src/Unlimited_Sparse_Row.templates.hh index d990c55..8f15962 100644 --- a/src/Unlimited_Sparse_Row.templates.hh +++ b/src/Unlimited_Sparse_Row.templates.hh @@ -104,9 +104,8 @@ Unlimited_Sparse_Row::combine_needs_second(const Unlimited_Sparse_Row& y, #ifdef PPL_SPARSE_BACKEND_INVALIDATES_REFERENCES i = last_i; ++i; - i_end = x.end(); - if (& static_cast<PIP_Tree_Node:: - matrix_row_const_reference_type>(x) == &y) { + i_end = end(); + if (this == &y) { j = last_i; j_end = y.end(); } @@ -124,9 +123,8 @@ Unlimited_Sparse_Row::combine_needs_second(const Unlimited_Sparse_Row& y, #ifdef PPL_SPARSE_BACKEND_INVALIDATES_REFERENCES i = last_i; ++i; - i_end = x.end(); - if (& static_cast<PIP_Tree_Node:: - matrix_row_const_reference_type>(x) == &y) { + i_end = end(); + if (this == &y) { j = last_i; j_end = y.end(); } @@ -156,9 +154,8 @@ Unlimited_Sparse_Row::combine_needs_second(const Unlimited_Sparse_Row& y, #ifdef PPL_SPARSE_BACKEND_INVALIDATES_REFERENCES i = last_i; ++i; - i_end = x.end(); - if (& static_cast<PIP_Tree_Node:: - matrix_row_const_reference_type>(x) == &y) { + i_end = end(); + if (this == &y) { j = last_i; j_end = y.end(); } @@ -272,9 +269,8 @@ Unlimited_Sparse_Row::combine(const Unlimited_Sparse_Row& y, const Func1& f, #ifdef PPL_SPARSE_BACKEND_INVALIDATES_REFERENCES i = last_i; ++i; - i_end = x.end(); - if (& static_cast<PIP_Tree_Node:: - matrix_row_const_reference_type>(x) == &y) { + i_end = end(); + if (this == &y) { j = last_i; j_end = y.end(); } @@ -293,9 +289,8 @@ Unlimited_Sparse_Row::combine(const Unlimited_Sparse_Row& y, const Func1& f, #ifdef PPL_SPARSE_BACKEND_INVALIDATES_REFERENCES i = last_i; ++i; - i_end = x.end(); - if (& static_cast<PIP_Tree_Node:: - matrix_row_const_reference_type>(x) == &y) { + i_end = end(); + if (this == &y) { j = last_i; j_end = y.end(); } @@ -326,9 +321,8 @@ Unlimited_Sparse_Row::combine(const Unlimited_Sparse_Row& y, const Func1& f, #ifdef PPL_SPARSE_BACKEND_INVALIDATES_REFERENCES i = last_i; ++i; - i_end = x.end(); - if (& static_cast<PIP_Tree_Node:: - matrix_row_const_reference_type>(x) == &y) { + i_end = end(); + if (this == &y) { j = last_i; j_end = y.end(); }

1 0

[GIT] ppl/ppl(sparse_matrices): Unlimited_Sparse_Row: update the backend requirements.
by Marco Poletti 08 Apr '10

08 Apr '10

Module: ppl/ppl Branch: sparse_matrices Commit: d4d4042ded72da571a7984a5800d4de84176aaf6 URL: http://www.cs.unipr.it/git/gitweb.cgi?p=ppl/ppl.git;a=commit;h=d4d4042ded72… Author: Marco Poletti <poletti.marco(a)gmail.com> Date: Thu Apr 8 18:59:11 2010 +0200 Unlimited_Sparse_Row: update the backend requirements. --- src/Unlimited_Sparse_Row.defs.hh | 73 ++++++++++++++++++++++++++++++++++++++ 1 files changed, 73 insertions(+), 0 deletions(-) diff --git a/src/Unlimited_Sparse_Row.defs.hh b/src/Unlimited_Sparse_Row.defs.hh index 5aeaec0..d53a552 100644 --- a/src/Unlimited_Sparse_Row.defs.hh +++ b/src/Unlimited_Sparse_Row.defs.hh @@ -97,6 +97,9 @@ namespace Parma_Polyhedra_Library { * C::value_type should be \p std::pair<dimension_type,Coefficient> . * C must have a swap() method. * C must have a const OK() method, returning bool. + * C must have a const external_memory_in_bytes() method, returning a + memory_size_type. This method returns the size in bytes of the memory + managed by \p *this. * C::const_iterator must have a default constructor * C::const_iterator must define operator*(), operator->(), and operator == and !=. @@ -109,6 +112,73 @@ namespace Parma_Polyhedra_Library { * C::dangerous_iterator must follow all the requirements for C::iterator. * C::dangerous_iterator must have a static method next(C::iterator i) that returns a dangerous_iterator pointing to the element after i. + * C must have a method find_dangerous(dimension_type i) returning a + C::dangerous_iterator. This method returns a dangerous_iterator pointing + to the element with index \p i, or end_dangerous() if there is no such + element. + * C must have a method + find_dangerous(dimension_type i, dangerous_iterator itr) returning a + C::dangerous_iterator. This method returns a dangerous_iterator pointing + to the element with index \p i, or end_dangerous() if there is no such + element. \p itr must be an iterator pointing to an element with index + less than i. + * C must have a method find(dimension_type i) returning a C::iterator. + This method returns a dangerous_iterator pointing to the element with + index \p i, or end() if there is no such element. + * C must have a method find(dimension_type i, iterator itr) returning a + C::iterator. This method returns an iterator pointing to the element with + index \p i, or end() if there is no such element. \p itr must be an + iterator pointing to an element with index less than i. + * C must have a const method find(dimension_type i) returning a + C::const_iterator. This method returns a const_iterator pointing to + the element with index \p i, or end() if there is no such element. + * C must have a const method find(dimension_type i, const_iterator itr) + returning a C::const_iterator. This method returns an iterator pointing to + the element with index \p i, or end() if there is no such element. + \p itr must be an iterator pointing to an element with index less than i. + * C must have a method lower_bound_dangerous(dimension_type i) returning a + C::dangerous_iterator. This method returns a dangerous_iterator pointing + to the element with index \p i, or to the last element with index less + than \p i, if there is no such element. + * C must have a method + lower_bound_dangerous(dimension_type i, dangerous_iterator itr) returning + a C::dangerous_iterator. This method returns a dangerous_iterator pointing + to the element with index \p i, or to the last element with index less + than \p i, if there is no such element. + \p itr must be an iterator pointing to an element with index less than i. + * C must have a method lower_bound(dimension_type i) returning a C::iterator. + This method returns a dangerous_iterator pointing to the element with + index \p i, or to the last element with index less than \p i, if there is + no such element. + * C must have a method lower_bound(dimension_type i, iterator itr) returning a + C::iterator. This method returns an iterator pointing to the element with + index \p i, or to the last element with index less than \p i, if there is + no such element. + \p itr must be an iterator pointing to an element with index less than i. + * C must have a const method lower_bound(dimension_type i) returning a + C::const_iterator. This method returns a const_iterator pointing to + the element with index \p i, or to the last element with index less than + \p i, if there is no such element. + * C must have a const method lower_bound(dimension_type i, const_iterator itr) + returning a C::const_iterator. This method returns an iterator pointing to + the element with index \p i, or to the last element with index less than + \p i, if there is no such element. + \p itr must be an iterator pointing to an element with index less than i. + * C must have a method + find2_dangerous(dimension_type c1, dimension_type c2, + C::dangerous_iterator& itr1, C::dangerous_iterator& itr2) + returning void, equivalent to<BR> + itr1=find_dangerous(c1); itr2=find_dangerous(c2); + * C must have a method + find2(dimension_type c1, dimension_type c2, + iterator& itr1, iterator& itr2); + returning void, equivalent to<BR> + itr1=find(c1); itr2=find(c2); + * C must have a const method + find2(dimension_type c1, dimension_type c2, + C::const_iterator& itr1, C::const_iterator& itr2) + returning void, equivalent to<BR> + itr1=find(c1); itr2=find(c2); * C must have a method insert(C::dangerous_iterator pos, const C::value_type& x) returning a C::dangerous_iterator. This method inserts x before pos and @@ -119,6 +189,9 @@ namespace Parma_Polyhedra_Library { returning a C::dangerous_iterator. This method inserts the pair (i, x) before pos and returns an iterator to the inserted element. This operation invalidates all C::dangerous_iterator objects equal to pos. + * C must have a method push_back(const C::value_type& x) returning void, + equivalent to<BR> + insert(end_dangerous(), x); * C must have a method erase(C::dangerous_iterator pos) returning a C::dangerous_iterator, that erases the element pointed to by pos. This operation invalidates all C::dangerous_iterators objects equal to

1 0