December 2010 - PPL-devel - BUGSENG Mailman

[GIT] ppl/ppl(master): Avoid using identifier PS, since Solaris has a macro with that name.
by Enea Zaffanella 11 Dec '10

11 Dec '10

Module: ppl/ppl Branch: master Commit: 533d88da39b83836bd2b4bf4ec69a6f0c9143f0c URL: http://www.cs.unipr.it/git/gitweb.cgi?p=ppl/ppl.git;a=commit;h=533d88da39b8… Author: Enea Zaffanella <zaffanella(a)cs.unipr.it> Date: Sat Dec 11 18:27:07 2010 +0100 Avoid using identifier PS, since Solaris has a macro with that name. --- doc/definitions.dox | 8 ++++---- instchk.hh | 4 ++-- tests/Powerset/powerset1.cc | 18 +++++++++--------- 3 files changed, 15 insertions(+), 15 deletions(-) diff --git a/doc/definitions.dox b/doc/definitions.dox index b94d2e8..fd1ceb4 100644 --- a/doc/definitions.dox +++ b/doc/definitions.dox @@ -103,10 +103,10 @@ where: Other semantic GDs, the <EM>compound classes</EM>, can be constructed (also recursively) from all the GDs classes. These include: - - \link Parma_Polyhedra_Library::Pointset_Powerset \c Pointset_Powerset\<PS\> \endlink, + - \link Parma_Polyhedra_Library::Pointset_Powerset \c Pointset_Powerset\<PSET\> \endlink, - \link Parma_Polyhedra_Library::Partially_Reduced_Product <CODE>Partially_Reduced_Product\<D1, D2, R\></CODE> \endlink, . -where \c PS, \c D1 and \c D2 can be any semantic GD classes and \c R is the +where \c PSET, \c D1 and \c D2 can be any semantic GD classes and \c R is the reduction operation to be applied to the component domains of the product class. @@ -2689,10 +2689,10 @@ not themselves occur as one of their argument domains. Therefore their use comes with the following warning. \warning -The <CODE>Pointset_Powerset<PS></CODE> and +The <CODE>Pointset_Powerset<PSET></CODE> and <CODE>Partially_Reduced_Product<D1, D2, R></CODE> should only be used with the following instantiations -for the disjunct domain template \p PS and component domain +for the disjunct domain template \p PSET and component domain templates \p D1 and \p D2: <CODE>C_Polyhedron</CODE>, <CODE>NNC_Polyhedron</CODE>, diff --git a/instchk.hh b/instchk.hh index 0a07a08..e848172 100644 --- a/instchk.hh +++ b/instchk.hh @@ -247,11 +247,11 @@ public: } }; -template <typename PS> +template <typename PSET> class Pointset_Powerset { public: static bool valid_instantiation() { - return PS::valid_Pointset_Powerset_argument(); + return PSET::valid_Pointset_Powerset_argument(); } }; diff --git a/tests/Powerset/powerset1.cc b/tests/Powerset/powerset1.cc index 09ab286..2c1e3f6 100644 --- a/tests/Powerset/powerset1.cc +++ b/tests/Powerset/powerset1.cc @@ -27,14 +27,14 @@ namespace { // Uses every public Powerset method. bool test01() { - typedef Powerset<FCAIBVP> PS; + typedef Powerset<FCAIBVP> PSET; Variable A(0); - PS ps1; + PSET ps1; ps1.add_disjunct(FCAIBVP(A)); - PS ps2 = ps1; + PSET ps2 = ps1; if (ps2 != ps1 || !(ps2 == ps1)) return false; @@ -43,7 +43,7 @@ test01() { nout << "ps1:" << endl << ps1 << endl; FCAIBVP d(A); - PS ps3(d); + PSET ps3(d); if (!ps1.definitely_entails(ps3)) return false; @@ -67,35 +67,35 @@ test01() { // Iterator. dimension_type count = 0; - for (PS::iterator i = ps3.begin(); i != ps3.end(); ++i) + for (PSET::iterator i = ps3.begin(); i != ps3.end(); ++i) ++count; if (count != 1) return false; // Constant iterator. count = 0; - for (PS::const_iterator i = ps3.begin(); i != ps3.end(); ++i) + for (PSET::const_iterator i = ps3.begin(); i != ps3.end(); ++i) ++count; if (count != 1) return false; // Reverse iterator. count = 0; - for (PS::reverse_iterator i = ps3.rbegin(); i != ps3.rend(); ++i) + for (PSET::reverse_iterator i = ps3.rbegin(); i != ps3.rend(); ++i) ++count; if (count != 1) return false; // Constant reverse iterator. count = 0; - for (PS::const_reverse_iterator i = ps3.rbegin(), + for (PSET::const_reverse_iterator i = ps3.rbegin(), ps3_rend = ps3.rend(); i != ps3_rend; ++i) ++count; if (count != 1) return false; ps2 = ps3; - PS ps_empty; + PSET ps_empty; ps2.drop_disjunct(ps2.begin()); if (ps2 != ps_empty) return false;

1 0

if the compiler has the remainder bug... yes ??
by Dennis Clarke 11 Dec '10

11 Dec '10

Could someone explain what this "remainder bug" is ? . . . checking the binary format of C++ long doubles... IEEE754 Quad Precision checking whether C++ provides exact output for long doubles... no checking whether the C++ compiler supports flexible arrays... yes checking whether the IEEE inexact flag is supported in C++... no checking if the compiler has the remainder bug... yes checking whether the C++ compiler supports __attribute__ ((weak))... yes checking for fenv.h... (cached) yes checking for ieeefp.h... (cached) yes checking getopt.h usability... no checking getopt.h presence... no checking for getopt.h... no . . . -- Dennis Clarke dclarke(a)opensolaris.ca <- Email related to the open source Solaris dclarke(a)blastwave.org <- Email related to open source for Solaris

2 1

Re: [PPL-devel] powerset1.cc: In function 'bool<unnamed>::test01()':
by Dennis Clarke 11 Dec '10

11 Dec '10

> On 12/10/10 17:38, Dennis Clarke wrote: >> >> While trying to build ppl-0.10.2 on Solaris 8 with GCC 4.4.4 I have seen >> the following error in the 'make check' stage : >> >> [...] >> >> The sources were taken from >> http://www.cs.unipr.it/ppl/Download/ftp/ppl/releases/LATEST/solaris/ >> >> Not sure if ppl can be compiled on Solaris 8. Certainly gmp, mpfr and >> mpc >> all work flawlessly. I would certainly like to incorporate ppl into the >> next GCC release if it could pass the testsuite. > > Hi Dennis. > > The PPL 0.10 series is no longer supported. We plan to release PPL 0.11.1 > really soon now. You can download the current release candidate from > > http://www.cs.unipr.it/ppl/Download/ftp/snapshots/ > > Please let us know how it goes: we will be glad to incorporate > improvements for Solaris 8 (or any other platform) into PPL 0.11.1. > All the best, > Thank you Sir ! I will begin to work with that RC right away and let you know what I see. -- Dennis Clarke dclarke(a)opensolaris.ca <- Email related to the open source Solaris dclarke(a)blastwave.org <- Email related to open source for Solaris

1 0

powerset1.cc: In function 'bool<unnamed>::test01()':
by Dennis Clarke 10 Dec '10

10 Dec '10

While trying to build ppl-0.10.2 on Solaris 8 with GCC 4.4.4 I have seen the following error in the 'make check' stage : /bin/bash ../../libtool --tag=CXX --mode=link /usr/local/gcc4/bin/g++ -g -O2 -march=i386 -frounding-math -march=i386 -malign-double -mpreferred-stack-boundary=8 -m32 -O3 -Qy -fexceptions -D_TS_ERRNO -march=i386 -malign-double -mpreferred-stack-boundary=8 -m32 -O3 -Qy -fexceptions -D_TS_ERRNO -W -Wall -o membytes1 membytes1.o ../../utils/libppl_utils.a ../../tests/libppl_tests.a ../../src/libppl.la -lm -L/usr/local/lib -lgmpxx -L/usr/local/lib -lgmp -R/usr/local/lib -R/usr/local/lib libtool: link: /usr/local/gcc4/bin/g++ -g -O2 -march=i386 -frounding-math -march=i386 -malign-double -mpreferred-stack-boundary=8 -m32 -O3 -Qy -fexceptions -D_TS_ERRNO -march=i386 -malign-double -mpreferred-stack-boundary=8 -m32 -O3 -Qy -fexceptions -D_TS_ERRNO -W -Wall -o .libs/membytes1 membytes1.o ../../utils/libppl_utils.a ../../tests/libppl_tests.a ../../src/.libs/libppl.so -L/usr/local/lib /usr/local/lib/libgmpxx.so /usr/local/gcc4/lib/libstdc++.so -lm /usr/local/lib/libgmp.so -Wl,-R -Wl,/usr/local/lib -Wl,-R -Wl,/usr/local/gcc4/lib ld: warning: file /usr/local/gcc4/lib/gcc/i386-pc-solaris2.8/4.4.4/../../../libstdc++.so: linked to /usr/local/gcc4/lib/libstdc++.so: attempted multiple inclusion of file /usr/local/gcc4/bin/g++ -DHAVE_CONFIG_H -I. -I../.. -I../../src -I../../src -I../../tests -I../../utils -DNDEBUG=1 -I/usr/local/include -g -O2 -march=i386 -frounding-math -march=i386 -malign-double -mpreferred-stack-boundary=8 -m32 -O3 -Qy -fexceptions -D_TS_ERRNO -march=i386 -malign-double -mpreferred-stack-boundary=8 -m32 -O3 -Qy -fexceptions -D_TS_ERRNO -W -Wall -MT powerset1.o -MD -MP -MF .deps/powerset1.Tpo -c -o powerset1.o powerset1.cc powerset1.cc: In function 'bool<unnamed>::test01()': powerset1.cc:30: error: expected unqualified-id before numeric constant powerset1.cc:34: error: expected ';' before 'ps1' powerset1.cc:34: warning: statement has no effect powerset1.cc:35: error: 'ps1' was not declared in this scope powerset1.cc:37: error: expected ';' before 'ps2' powerset1.cc:37: warning: statement has no effect powerset1.cc:39: error: 'ps2' was not declared in this scope powerset1.cc:46: error: expected ';' before 'ps3' powerset1.cc:46: warning: statement has no effect powerset1.cc:48: error: 'ps3' was not declared in this scope powerset1.cc:51: error: 'ps3' was not declared in this scope powerset1.cc:57: error: 'ps3' was not declared in this scope powerset1.cc:70: error: expected ';' before '::' token powerset1.cc:70: warning: statement has no effect powerset1.cc:70: error: 'i' was not declared in this scope powerset1.cc:77: error: expected ';' before '::' token powerset1.cc:77: warning: statement has no effect powerset1.cc:77: error: 'i' was not declared in this scope powerset1.cc:84: error: expected ';' before '::' token powerset1.cc:84: warning: statement has no effect powerset1.cc:84: error: 'i' was not declared in this scope powerset1.cc:91: error: expected ';' before '::' token powerset1.cc:92: warning: statement has no effect powerset1.cc:92: error: 'i' was not declared in this scope powerset1.cc:92: error: 'ps3_rend' was not declared in this scope powerset1.cc:97: error: 'ps2' was not declared in this scope powerset1.cc:98: error: expected ';' before 'ps_empty' powerset1.cc:98: warning: statement has no effect powerset1.cc:100: error: 'ps_empty' was not declared in this scope powerset1.cc:105: error: 'ps_empty' was not declared in this scope powerset1.cc:110: error: 'ps_empty' was not declared in this scope powerset1.cc:115: error: 'ps_empty' was not declared in this scope powerset1.cc:118: error: 'ps_empty' was not declared in this scope make[3]: *** [powerset1.o] Error 1 make[3]: Leaving directory `/export/medusa/dclarke/build/GCC/4.5.2-RC/ppl/i386/ppl-0.10.2.002/tests/Powerset' make[2]: *** [check-am] Error 2 make[2]: Leaving directory `/export/medusa/dclarke/build/GCC/4.5.2-RC/ppl/i386/ppl-0.10.2.002/tests/Powerset' make[1]: *** [check-recursive] Error 1 make[1]: Leaving directory `/export/medusa/dclarke/build/GCC/4.5.2-RC/ppl/i386/ppl-0.10.2.002/tests' make: *** [check-recursive] Error 1 [ titan ] $ The sources were taken from http://www.cs.unipr.it/ppl/Download/ftp/ppl/releases/LATEST/solaris/ Not sure if ppl can be compiled on Solaris 8. Certainly gmp, mpfr and mpc all work flawlessly. I would certainly like to incorporate ppl into the next GCC release if it could pass the testsuite. -- Dennis Clarke dclarke(a)opensolaris.ca <- Email related to the open source Solaris dclarke(a)blastwave.org <- Email related to open source for Solaris

3 2

ppl 0.10.2 fails on Solaris 8 Sparc
by Dennis Clarke 10 Dec '10

10 Dec '10

During the make test phase I see this : ../../src/OR_Matrix.inlines.hh:522: instantiated from 'Parma_Polyhedra_Library::OR_Matrix<T>::OR_Matrix(const Parma_Polyhedra_Library::OR_Matrix<U>&) [with U = Parma_Polyhedra_Library::Checked_Number<char, Parma_Polyhedra_Library::WRD_Extended_Number_Policy>, T = Parma_Polyhedra_Library::Checked_Number<__gmp_expr<__mpq_struct [1], __mpq_struct [1]>, Parma_Polyhedra_Library::WRD_Extended_Number_Policy>]' ../../src/Octagonal_Shape.inlines.hh:122: instantiated from 'Parma_Polyhedra_Library::Octagonal_Shape<T>::Octagonal_Shape(const Parma_Polyhedra_Library::Octagonal_Shape<T>&, Parma_Polyhedra_Library::Complexity_Class) [with U = char, T = __gmp_expr<__mpq_struct [1], __mpq_struct [1]>]' addspacedims1.cc:47: instantiated from here ../../src/checked.defs.hh:436: error: incomplete type 'Parma_Polyhedra_Library::Checked::assign_function_struct<Parma_Polyhedra_Library::WRD_Extended_Number_Policy, Parma_Polyhedra_Library::WRD_Extended_Number_Policy, __gmp_expr<__mpq_struct [1], __mpq_struct [1]>, char>' used in nested name specifier make[5]: *** [addspacedims1.o] Error 1 make[5]: Leaving directory `/export/medusa/dclarke/build/GCC/4.5.2-RC/ppl/sparc/ppl-0.10.2.002/tests/Octagonal_Shape' make[4]: *** [check-am] Error 2 make[4]: Leaving directory `/export/medusa/dclarke/build/GCC/4.5.2-RC/ppl/sparc/ppl-0.10.2.002/tests/Octagonal_Shape' make[3]: *** [run_tests.stamp] Error 1 make[3]: Leaving directory `/export/medusa/dclarke/build/GCC/4.5.2-RC/ppl/sparc/ppl-0.10.2.002/tests/Octagonal_Shape' make[2]: *** [check-am] Error 2 make[2]: Leaving directory `/export/medusa/dclarke/build/GCC/4.5.2-RC/ppl/sparc/ppl-0.10.2.002/tests/Octagonal_Shape' make[1]: *** [check-recursive] Error 1 make[1]: Leaving directory `/export/medusa/dclarke/build/GCC/4.5.2-RC/ppl/sparc/ppl-0.10.2.002/tests' make: *** [check-recursive] Error 1 -- Dennis Clarke dclarke(a)opensolaris.ca <- Email related to the open source Solaris dclarke(a)blastwave.org <- Email related to open source for Solaris

1 0

Re: [PPL-devel] powerset1.cc: In function 'bool<unnamed>::test01()':
by Dennis Clarke 10 Dec '10

10 Dec '10

> Il 10/12/2010 19:04, Dennis Clarke ha scritto: >> >>> Il 10/12/2010 17:38, Dennis Clarke ha scritto: >>>> >>>> While trying to build ppl-0.10.2 on Solaris 8 with GCC 4.4.4 I have >>>> seen >>>> the following error in the 'make check' stage : >>>> >>>> /bin/bash ../../libtool --tag=CXX --mode=link >>>> /usr/local/gcc4/bin/g++ >>>> -g -O2 -march=i386 -frounding-math -march=i386 -malign-double >>>> -mpreferred-stack-boundary=8 -m32 -O3 -Qy -fexceptions -D_TS_ERRNO >>>> -march=i386 -malign-double -mpreferred-stack-boundary=8 -m32 -O3 -Qy >>>> -fexceptions -D_TS_ERRNO -W -Wall -o membytes1 membytes1.o >>>> ../../utils/libppl_utils.a ../../tests/libppl_tests.a >>>> ../../src/libppl.la >>>> -lm -L/usr/local/lib -lgmpxx -L/usr/local/lib -lgmp -R/usr/local/lib >>>> -R/usr/local/lib >>>> libtool: link: /usr/local/gcc4/bin/g++ -g -O2 -march=i386 >>>> -frounding-math >>>> -march=i386 -malign-double -mpreferred-stack-boundary=8 -m32 -O3 -Qy >>>> -fexceptions -D_TS_ERRNO -march=i386 -malign-double >>>> -mpreferred-stack-boundary=8 -m32 -O3 -Qy -fexceptions -D_TS_ERRNO -W >>>> -Wall -o .libs/membytes1 membytes1.o ../../utils/libppl_utils.a >>>> ../../tests/libppl_tests.a ../../src/.libs/libppl.so -L/usr/local/lib >>>> /usr/local/lib/libgmpxx.so /usr/local/gcc4/lib/libstdc++.so -lm >>>> /usr/local/lib/libgmp.so -Wl,-R -Wl,/usr/local/lib -Wl,-R >>>> -Wl,/usr/local/gcc4/lib >>>> ld: warning: file >>>> /usr/local/gcc4/lib/gcc/i386-pc-solaris2.8/4.4.4/../../../libstdc++.so: >>>> linked to /usr/local/gcc4/lib/libstdc++.so: attempted multiple >>>> inclusion >>>> of file >>>> /usr/local/gcc4/bin/g++ -DHAVE_CONFIG_H -I. -I../.. -I../../src >>>> -I../../src -I../../tests -I../../utils -DNDEBUG=1 >>>> -I/usr/local/include >>>> -g -O2 -march=i386 -frounding-math -march=i386 -malign-double >>>> -mpreferred-stack-boundary=8 -m32 -O3 -Qy -fexceptions -D_TS_ERRNO >>>> -march=i386 -malign-double -mpreferred-stack-boundary=8 -m32 -O3 -Qy >>>> -fexceptions -D_TS_ERRNO -W -Wall -MT powerset1.o -MD -MP -MF >>>> .deps/powerset1.Tpo -c -o powerset1.o powerset1.cc >>> >>> Can you try to replace PS in this file with something other (e.g. Q_PS) >>> and recompile? >>> >> >> I'm sorry but I don't understand your instructions. >> >> Is there a diff you want me to apply or some change to powerset1.cc you >> want ? > > In tests/Powerset/powerset1.cc there are 9 occurrences of string PS, you > should replace them with another string (e.g. Q_PS) and run the > compilation command you wrote above. > > If it compiles ok this means that solaris header files defines the macro > PS. > OKay .. easily done thus : [ titan ] $ diff tests/Powerset/powerset1.cc tests/Powerset/powerset1.cc.backup 30c30 < typedef Powerset<FCAIBVP> Q_PS; --- > typedef Powerset<FCAIBVP> PS; 34c34 < Q_PS ps1; --- > PS ps1; 37c37 < Q_PS ps2 = ps1; --- > PS ps2 = ps1; 46c46 < Q_PS ps3(d); --- > PS ps3(d); 70c70 < for (Q_PS::iterator i = ps3.begin(); i != ps3.end(); ++i) --- > for (PS::iterator i = ps3.begin(); i != ps3.end(); ++i) 77c77 < for (Q_PS::const_iterator i = ps3.begin(); i != ps3.end(); ++i) --- > for (PS::const_iterator i = ps3.begin(); i != ps3.end(); ++i) 84c84 < for (Q_PS::reverse_iterator i = ps3.rbegin(); i != ps3.rend(); ++i) --- > for (PS::reverse_iterator i = ps3.rbegin(); i != ps3.rend(); ++i) 91c91 < for (Q_PS::const_reverse_iterator i = ps3.rbegin(), --- > for (PS::const_reverse_iterator i = ps3.rbegin(), 98c98 < Q_PS ps_empty; --- > PS ps_empty; [ titan ] $ cd tests/Powerset/ [ titan ] $ /usr/local/gcc4/bin/g++ -DHAVE_CONFIG_H -I. -I../.. \ > -I../../src -I../../tests -I../../utils -DNDEBUG=1 -I/usr/local/include \ > -g -O2 -march=i386 -frounding-math -malign-double -mpreferred-stack-boundary=8 \ > -m32 -O3 -Qy -fexceptions -D_TS_ERRNO -W -Wall -MT powerset1.o -MD \ > -MP -MF .deps/powerset1.Tpo -v -c -o powerset1.o powerset1.cc Using built-in specs. Target: i386-pc-solaris2.8 Configured with: ../../../gcc-4.4.4/configure --build=i386-pc-solaris2.8 --with-gnu-as --with-as=/usr/local/bin/as --without-gnu-ld --with-ld=/usr/ccs/bin/ld --with-cpu-32=i386 --with-arch-32=i386 --enable-stage1-languages=c --enable-nls --with-libiconv-prefix=/usr/local --enable-threads=posix --prefix=/usr/local/gcc4 --with-local-prefix=/usr/local --enable-shared --enable-multilib --with-system-zlib --with-pkgversion='Blastwave.org Inc. Tue May 4 18:49:38 GMT 2010' --with-bugurl=http://wwww.blastwave.org/support --enable-languages=c,c++,objc,fortran,ada --enable-bootstrap Thread model: posix gcc version 4.4.4 (Blastwave.org Inc. Tue May 4 18:49:38 GMT 2010) COLLECT_GCC_OPTIONS='-DHAVE_CONFIG_H' '-I.' '-I../..' '-I../../src' '-I../../tests' '-I../../utils' '-DNDEBUG=1' '-I/usr/local/include' '-g' '-O2' '-march=i386' '-frounding-math' '-malign-double' '-mpreferred-stack-boundary=8' '-m32' '-O3' '-Qy' '-fexceptions' '-D_TS_ERRNO' '-W' '-Wall' '-MT' 'powerset1.o' '-MD' '-MP' '-MF' '.deps/powerset1.Tpo' '-v' '-c' '-o' 'powerset1.o' '-shared-libgcc' /usr/local/gcc4/libexec/gcc/i386-pc-solaris2.8/4.4.4/cc1plus -quiet -v -I. -I../.. -I../../src -I../../tests -I../../utils -I/usr/local/include -MD powerset1.d -MF .deps/powerset1.Tpo -MP -MT powerset1.o -DHAVE_CONFIG_H -DNDEBUG=1 -D_TS_ERRNO powerset1.cc -quiet -dumpbase powerset1.cc -march=i386 -malign-double -mpreferred-stack-boundary=8 -m32 -auxbase-strip powerset1.o -g -O2 -O3 -W -Wall -version -frounding-math -fexceptions -o /var/tmp//ccdg9w9B.s ignoring nonexistent directory "/usr/local/gcc4/lib/gcc/i386-pc-solaris2.8/4.4.4/../../../../i386-pc-solaris2.8/include" ignoring duplicate directory "/usr/local/include" as it is a non-system directory that duplicates a system directory #include "..." search starts here: #include <...> search starts here: . ../.. ../../src ../../tests ../../utils /usr/local/gcc4/lib/gcc/i386-pc-solaris2.8/4.4.4/../../../../include/c++/4.4.4 /usr/local/gcc4/lib/gcc/i386-pc-solaris2.8/4.4.4/../../../../include/c++/4.4.4/i386-pc-solaris2.8 /usr/local/gcc4/lib/gcc/i386-pc-solaris2.8/4.4.4/../../../../include/c++/4.4.4/backward /usr/local/include /usr/local/gcc4/include /usr/local/gcc4/lib/gcc/i386-pc-solaris2.8/4.4.4/include /usr/local/gcc4/lib/gcc/i386-pc-solaris2.8/4.4.4/include-fixed /usr/include End of search list. GNU C++ (Blastwave.org Inc. Tue May 4 18:49:38 GMT 2010) version 4.4.4 (i386-pc-solaris2.8) compiled by GNU C version 4.4.4, GMP version 5.0.1, MPFR version 2.4.2. GGC heuristics: --param ggc-min-expand=82 --param ggc-min-heapsize=98255 Compiler executable checksum: 4d5a5900091ff6e382c3c06dea0e82ea COLLECT_GCC_OPTIONS='-DHAVE_CONFIG_H' '-I.' '-I../..' '-I../../src' '-I../../tests' '-I../../utils' '-DNDEBUG=1' '-I/usr/local/include' '-g' '-O2' '-march=i386' '-frounding-math' '-malign-double' '-mpreferred-stack-boundary=8' '-m32' '-O3' '-Qy' '-fexceptions' '-D_TS_ERRNO' '-W' '-Wall' '-MT' 'powerset1.o' '-MD' '-MP' '-MF' '.deps/powerset1.Tpo' '-v' '-c' '-o' 'powerset1.o' '-shared-libgcc' /usr/local/bin/as -v -I. -I../.. -I../../src -I../../tests -I../../utils -I/usr/local/include -V -Qy -s -o powerset1.o /var/tmp//ccdg9w9B.s GNU assembler version 2.20.1 (i386-pc-solaris2.8) using BFD version (GNU Binutils) 2.20.1.20100303 COMPILER_PATH=/usr/local/gcc4/libexec/gcc/i386-pc-solaris2.8/4.4.4/:/usr/local/gcc4/libexec/gcc/i386-pc-solaris2.8/4.4.4/:/usr/local/gcc4/libexec/gcc/i386-pc-solaris2.8/:/usr/local/gcc4/lib/gcc/i386-pc-solaris2.8/4.4.4/:/usr/local/gcc4/lib/gcc/i386-pc-solaris2.8/:/usr/ccs/bin/ LIBRARY_PATH=/usr/local/gcc4/lib/gcc/i386-pc-solaris2.8/4.4.4/:/usr/ccs/lib/:/usr/local/gcc4/lib/gcc/i386-pc-solaris2.8/4.4.4/../../../:/lib/:/usr/lib/ COLLECT_GCC_OPTIONS='-DHAVE_CONFIG_H' '-I.' '-I../..' '-I../../src' '-I../../tests' '-I../../utils' '-DNDEBUG=1' '-I/usr/local/include' '-g' '-O2' '-march=i386' '-frounding-math' '-malign-double' '-mpreferred-stack-boundary=8' '-m32' '-O3' '-Qy' '-fexceptions' '-D_TS_ERRNO' '-W' '-Wall' '-MT' 'powerset1.o' '-MD' '-MP' '-MF' '.deps/powerset1.Tpo' '-v' '-c' '-o' 'powerset1.o' '-shared-libgcc' [ titan ] $ [ titan ] $ ls -l powerset1.o -rw-r--r-- 1 dclarke csw 479944 Dec 10 18:21 powerset1.o [ titan ] $ file powerset1.o powerset1.o: ELF 32-bit LSB relocatable 80386 Version 1 That looks better thus far. I'll go back and reconfigure, make and then run make check again from scratch and see what I get. Thank you and I'll let you know what I see. -- Dennis Clarke dclarke(a)opensolaris.ca <- Email related to the open source Solaris dclarke(a)blastwave.org <- Email related to open source for Solaris

1 0

Re: [PPL-devel] powerset1.cc: In function 'bool<unnamed>::test01()':
by Dennis Clarke 10 Dec '10

10 Dec '10

> Il 10/12/2010 17:38, Dennis Clarke ha scritto: >> >> While trying to build ppl-0.10.2 on Solaris 8 with GCC 4.4.4 I have seen >> the following error in the 'make check' stage : >> >> /bin/bash ../../libtool --tag=CXX --mode=link /usr/local/gcc4/bin/g++ >> -g -O2 -march=i386 -frounding-math -march=i386 -malign-double >> -mpreferred-stack-boundary=8 -m32 -O3 -Qy -fexceptions -D_TS_ERRNO >> -march=i386 -malign-double -mpreferred-stack-boundary=8 -m32 -O3 -Qy >> -fexceptions -D_TS_ERRNO -W -Wall -o membytes1 membytes1.o >> ../../utils/libppl_utils.a ../../tests/libppl_tests.a >> ../../src/libppl.la >> -lm -L/usr/local/lib -lgmpxx -L/usr/local/lib -lgmp -R/usr/local/lib >> -R/usr/local/lib >> libtool: link: /usr/local/gcc4/bin/g++ -g -O2 -march=i386 >> -frounding-math >> -march=i386 -malign-double -mpreferred-stack-boundary=8 -m32 -O3 -Qy >> -fexceptions -D_TS_ERRNO -march=i386 -malign-double >> -mpreferred-stack-boundary=8 -m32 -O3 -Qy -fexceptions -D_TS_ERRNO -W >> -Wall -o .libs/membytes1 membytes1.o ../../utils/libppl_utils.a >> ../../tests/libppl_tests.a ../../src/.libs/libppl.so -L/usr/local/lib >> /usr/local/lib/libgmpxx.so /usr/local/gcc4/lib/libstdc++.so -lm >> /usr/local/lib/libgmp.so -Wl,-R -Wl,/usr/local/lib -Wl,-R >> -Wl,/usr/local/gcc4/lib >> ld: warning: file >> /usr/local/gcc4/lib/gcc/i386-pc-solaris2.8/4.4.4/../../../libstdc++.so: >> linked to /usr/local/gcc4/lib/libstdc++.so: attempted multiple inclusion >> of file >> /usr/local/gcc4/bin/g++ -DHAVE_CONFIG_H -I. -I../.. -I../../src >> -I../../src -I../../tests -I../../utils -DNDEBUG=1 -I/usr/local/include >> -g -O2 -march=i386 -frounding-math -march=i386 -malign-double >> -mpreferred-stack-boundary=8 -m32 -O3 -Qy -fexceptions -D_TS_ERRNO >> -march=i386 -malign-double -mpreferred-stack-boundary=8 -m32 -O3 -Qy >> -fexceptions -D_TS_ERRNO -W -Wall -MT powerset1.o -MD -MP -MF >> .deps/powerset1.Tpo -c -o powerset1.o powerset1.cc > > Can you try to replace PS in this file with something other (e.g. Q_PS) > and recompile? > I'm sorry but I don't understand your instructions. Is there a diff you want me to apply or some change to powerset1.cc you want ? -- Dennis Clarke dclarke(a)opensolaris.ca <- Email related to the open source Solaris dclarke(a)blastwave.org <- Email related to open source for Solaris

2 1

Fwd: Re: kern/133583: [libm] fma(3) does not respect rounding mode using extended precision
by Abramo Bagnara 03 Dec '10

03 Dec '10

Too much worrisome to be true?!!? -------- Messaggio originale -------- Oggetto: Re: kern/133583: [libm] fma(3) does not respect rounding mode using extended precision Data: Fri, 3 Dec 2010 07:02:04 GMT Mittente: das(a)FreeBSD.org A: abramo.bagnara(a)gmail.com, das(a)FreeBSD.org, freebsd-bugs(a)FreeBSD.org Synopsis: [libm] fma(3) does not respect rounding mode using extended precision State-Changed-From-To: open->closed State-Changed-By: das State-Changed-When: Fri Dec 3 07:01:28 UTC 2010 State-Changed-Why: Thanks for the report! This limitation is described in the source for fma(), and unfortunately, it is unlikely to ever change. There are several reasons: - We are a long way from having the necessary compiler support to make dynamic precision changes work as expected. - Dynamic FPU precision changes aren't officially supported, and fpsetprec() has been documented as deprecated for many years. - The only supported architecture that can have this problem due to dynamic precision changes is i386, and even then only for non-SSE2 builds. - The cost and complexity associated with making every function in libm detect and adapt to dynamic precision changes is prohibitive. I have updated the manpage for fpsetprec() to explain that changing the FPU precision isn't supported by the compiler or libraries. http://www.freebsd.org/cgi/query-pr.cgi?pr=133583

1 0

[GIT] ppl/ppl(floating_point): Fixed a problem involving a missing iostream
by Fabio Bossi 02 Dec '10

02 Dec '10

Module: ppl/ppl Branch: floating_point Commit: dc027679371676490ea74a3356c5c5aa6e9b735f URL: http://www.cs.unipr.it/git/gitweb.cgi?p=ppl/ppl.git;a=commit;h=dc0276793716… Author: Fabio Bossi <bossi(a)cs.unipr.it> Date: Thu Dec 2 13:33:58 2010 +0100 Fixed a problem involving a missing iostream inclusion. --- src/Linear_Form.inlines.hh | 9 ++------- 1 files changed, 2 insertions(+), 7 deletions(-) diff --git a/src/Linear_Form.inlines.hh b/src/Linear_Form.inlines.hh index f799fc4..022f54e 100644 --- a/src/Linear_Form.inlines.hh +++ b/src/Linear_Form.inlines.hh @@ -24,6 +24,7 @@ site: http://www.cs.unipr.it/ppl/ . */ #define PPL_Linear_Form_inlines_hh 1 #include "Variable.defs.hh" +#include <iostream> #include <stdexcept> namespace Parma_Polyhedra_Library { @@ -194,8 +195,6 @@ Linear_Form<C>::swap(Linear_Form& y) { template <typename C> inline void Linear_Form<C>::ascii_dump(std::ostream& s) const { - // FIXME: the following commented code does not compile. - /* using namespace IO_Operators; const char separator = ' '; dimension_type space = space_dimension(); @@ -203,18 +202,16 @@ Linear_Form<C>::ascii_dump(std::ostream& s) const { for (int i = 0; i <= space; ++i) s << vec[i] << separator; s << "\n"; - */ } template <typename C> inline bool Linear_Form<C>::ascii_load(std::istream& s) { - // FIXME: the following commented code does not compile. - /* using namespace IO_Operators; dimension_type new_dim; if (!(s >> new_dim)) return false; + vec.resize(new_dim + 1, zero); for (int i = 0; i <= new_dim; ++i) { if (!(s >> vec[i])) @@ -223,8 +220,6 @@ Linear_Form<C>::ascii_load(std::istream& s) { PPL_ASSERT(OK()); return true; - */ - return false; } // Floating point analysis related methods.

1 0

[GIT] ppl/ppl(floating_point): Added a proposed implementation for ascii_dump()
by Fabio Bossi 02 Dec '10

02 Dec '10

Module: ppl/ppl Branch: floating_point Commit: 31db7495cb05e632726d12b881a4d7ed5b062a7a URL: http://www.cs.unipr.it/git/gitweb.cgi?p=ppl/ppl.git;a=commit;h=31db7495cb05… Author: Fabio Bossi <bossi(a)cs.unipr.it> Date: Thu Dec 2 12:18:22 2010 +0100 Added a proposed implementation for ascii_dump() and ascii_load(). --- src/Linear_Form.inlines.hh | 27 +++++++++++++++++++++++++-- 1 files changed, 25 insertions(+), 2 deletions(-) diff --git a/src/Linear_Form.inlines.hh b/src/Linear_Form.inlines.hh index 43ca4e6..f799fc4 100644 --- a/src/Linear_Form.inlines.hh +++ b/src/Linear_Form.inlines.hh @@ -194,13 +194,36 @@ Linear_Form<C>::swap(Linear_Form& y) { template <typename C> inline void Linear_Form<C>::ascii_dump(std::ostream& s) const { - // FIXME: to be written. + // FIXME: the following commented code does not compile. + /* + using namespace IO_Operators; + const char separator = ' '; + dimension_type space = space_dimension(); + s << space << "\n"; + for (int i = 0; i <= space; ++i) + s << vec[i] << separator; + s << "\n"; + */ } template <typename C> inline bool Linear_Form<C>::ascii_load(std::istream& s) { - // FIXME: to be written. + // FIXME: the following commented code does not compile. + /* + using namespace IO_Operators; + dimension_type new_dim; + if (!(s >> new_dim)) + return false; + vec.resize(new_dim + 1, zero); + for (int i = 0; i <= new_dim; ++i) { + if (!(s >> vec[i])) + return false; + } + + PPL_ASSERT(OK()); + return true; + */ return false; }

1 0