February 2011 - PPL-devel - BUGSENG Mailman

[GIT] ppl/ppl(ppl-0_11-branch): Work around a bug showing up under mingw-w64.
by Roberto Bagnara 23 Feb '11

23 Feb '11

Module: ppl/ppl Branch: ppl-0_11-branch Commit: 5e9b50198b500bfb8a827882ebaa40f1f6305dd4 URL: http://www.cs.unipr.it/git/gitweb.cgi?p=ppl/ppl.git;a=commit;h=5e9b50198b50… Author: Roberto Bagnara <bagnara(a)cs.unipr.it> Date: Wed Feb 23 13:21:12 2011 +0100 Work around a bug showing up under mingw-w64. --- src/fpu-ia32.cc | 4 ++++ 1 files changed, 4 insertions(+), 0 deletions(-) diff --git a/src/fpu-ia32.cc b/src/fpu-ia32.cc index 21aa6e3..16e9aa1 100644 --- a/src/fpu-ia32.cc +++ b/src/fpu-ia32.cc @@ -30,6 +30,10 @@ site: http://www.cs.unipr.it/ppl/ . */ #include "fpu.defs.hh" #include <csetjmp> #include <csignal> +// This inclusion is to work around a bug present in some versions +// of GCC under mingw-w64. +// See http://www.cs.unipr.it/pipermail/ppl-devel/2011-February/017342.html +#include <cstddef> namespace {

1 0

[GIT] ppl/ppl(ppl-0_11-branch): Set INLINE_SOURCES = NO when producing LaTeX documentation.
by Enea Zaffanella 23 Feb '11

23 Feb '11

Module: ppl/ppl Branch: ppl-0_11-branch Commit: c60bb10f06f90c9faa7bf59889deabb678f3ce38 URL: http://www.cs.unipr.it/git/gitweb.cgi?p=ppl/ppl.git;a=commit;h=c60bb10f06f9… Author: Enea Zaffanella <zaffanella(a)cs.unipr.it> Date: Wed Feb 23 10:19:21 2011 +0100 Set INLINE_SOURCES = NO when producing LaTeX documentation. This change is a workaround to a pdftex + hyperref issue that may result in the following error \pdfendlink ended up in different nesting level than \pdfstartlink when using some versions of Doxygen (e.g., 1.7.3). According to http://tug.org/errors.html the error is triggered when a hyperlink is split across page boundaries and it can be generated even by valid LaTeX code. Disabling source inlining is a way to greatly decrease its likelihood. --- Watchdog/doc/devref.doxyconf-latex.in | 2 +- doc/devref-language-interface.doxyconf.in | 2 +- doc/devref.doxyconf-latex.in | 2 +- 3 files changed, 3 insertions(+), 3 deletions(-) diff --git a/Watchdog/doc/devref.doxyconf-latex.in b/Watchdog/doc/devref.doxyconf-latex.in index e5d8bfe..ddc0d08 100644 --- a/Watchdog/doc/devref.doxyconf-latex.in +++ b/Watchdog/doc/devref.doxyconf-latex.in @@ -137,7 +137,7 @@ FILTER_SOURCE_FILES = NO # configuration options related to source browsing #--------------------------------------------------------------------------- SOURCE_BROWSER = YES -INLINE_SOURCES = YES +INLINE_SOURCES = NO STRIP_CODE_COMMENTS = YES REFERENCED_BY_RELATION = YES REFERENCES_RELATION = YES diff --git a/doc/devref-language-interface.doxyconf.in b/doc/devref-language-interface.doxyconf.in index a324118..e277689 100644 --- a/doc/devref-language-interface.doxyconf.in +++ b/doc/devref-language-interface.doxyconf.in @@ -106,7 +106,7 @@ FILTER_SOURCE_FILES = NO # configuration options related to source browsing #--------------------------------------------------------------------------- SOURCE_BROWSER = YES -INLINE_SOURCES = YES +INLINE_SOURCES = NO STRIP_CODE_COMMENTS = YES REFERENCED_BY_RELATION = YES REFERENCES_RELATION = YES diff --git a/doc/devref.doxyconf-latex.in b/doc/devref.doxyconf-latex.in index b947949..e4dd017 100644 --- a/doc/devref.doxyconf-latex.in +++ b/doc/devref.doxyconf-latex.in @@ -352,7 +352,7 @@ FILTER_SOURCE_FILES = NO # configuration options related to source browsing #--------------------------------------------------------------------------- SOURCE_BROWSER = YES -INLINE_SOURCES = YES +INLINE_SOURCES = NO STRIP_CODE_COMMENTS = YES REFERENCED_BY_RELATION = YES REFERENCES_RELATION = YES

1 0

[GIT] ppl/ppl(ppl-0_11-branch): Fixed a couple of Doxygen related problems.
by Enea Zaffanella 22 Feb '11

22 Feb '11

Module: ppl/ppl Branch: ppl-0_11-branch Commit: 9f1f95afbe1f5fa50432952f197f4fbb3dbfafe1 URL: http://www.cs.unipr.it/git/gitweb.cgi?p=ppl/ppl.git;a=commit;h=9f1f95afbe1f… Author: Enea Zaffanella <zaffanella(a)cs.unipr.it> Date: Tue Feb 22 17:07:35 2011 +0100 Fixed a couple of Doxygen related problems. --- src/meta_programming.hh | 13 +++++++------ 1 files changed, 7 insertions(+), 6 deletions(-) diff --git a/src/meta_programming.hh b/src/meta_programming.hh index 25f4648..62de118 100644 --- a/src/meta_programming.hh +++ b/src/meta_programming.hh @@ -207,8 +207,8 @@ struct Is_Same<T, T> : public True { template <typename T> struct B; template <typename T> struct D : public B<T>; \endcode - Of course, we cannot test if, for some type variable <CODE>U</CODE>, - we have <CODE>Is_Same_Or_Derived<B<U>, Type>::value == true</CODE>. + Of course, we cannot test if, for some type variable <CODE>U</CODE>, we have + <CODE>Is_Same_Or_Derived<B<U>, Type>:: anonymous_enum:: value == true</CODE>. But we can do as follows: \code struct B_Base { @@ -217,7 +217,7 @@ struct Is_Same<T, T> : public True { template <typename T> struct B : public B_Base; \endcode This enables us to enquire - <CODE>Is_Same_Or_Derived<B_Base, Type>::value</CODE>. + <CODE>Is_Same_Or_Derived<B_Base, Type>:: anonymous_enum:: value</CODE>. */ #endif // defined(PPL_DOXYGEN_INCLUDE_IMPLEMENTATION_DETAILS) template <typename Base, typename Derived> @@ -248,9 +248,10 @@ struct Is_Same_Or_Derived { enum anonymous_enum { /*! Assuming <CODE>sizeof(char) != sizeof(double)</CODE>, the C++ - overload resolution mechanism guarantees that \p value evaluates - to <CODE>true</CODE> if and only if \p Base is the same type - as \p Derived or \p Derived is a class derived from \p Base. + overload resolution mechanism guarantees that <CODE>value</CODE> + evaluates to <CODE>true</CODE> if and only if <CODE>Base</CODE> + is the same type as <CODE>Derived</CODE> or <CODE>Derived</CODE> + is a class derived from <CODE>Base</CODE>. */ value = (sizeof(func(derived_object())) == sizeof(char)) };

1 0

A problem with hypertarget.
by Enea Zaffanella 22 Feb '11

22 Feb '11

When producing a PDF file with PDF_HYPERLINKS = yes, the latex generated by a template class documentation such as //! A template class. template <typename T> class Templ { T a; }; generates the following LaTeX: \hypertarget{classTempl}{ \section{Templ$<$ T $>$ Class Template Reference} \label{classTempl}\index{Templ@{Templ}} } which results in the following warning: Package hyperref Warning: Token not allowed in a PDF string (Unicode): (hyperref) removing `math shift' on input line 4. The problem is in the use of `$' inside a hypertarget. The attached patch solves this issue by adding a boolean status flag (insideHypertarget) to classes LatexGenerator and LatexDocVisitor. NOTE: the patch only addresses the cases of $<$ and $>$. It might be worth to consider whether or not other mathematical symbols produced by function LatexDocVisitor::visit(DocSymbol *s) could occur inside a hypertarget (and hence should be guarded using the same technique). Regards, Enea.

1 1

Re: [PPL-devel] Parma Polyhedra Library 0.11.1
by Roberto Bagnara 22 Feb '11

22 Feb '11

On 02/22/2011 06:04 AM, Dongsheng Song wrote: > When I build on i686-w64-mingw32 target: > > libtool: compile: i686-w64-mingw32-g++ -DHAVE_CONFIG_H -I. > -I/home/oracle/src/ppl-0.11.1/src -I.. -I.. > -I/home/oracle/src/ppl-0.11.1/src > -I/home/oracle/tmp/gcc-4.5-windows-obj/misc//include -g -O2 > -frounding-math -march=x86-64 -O2 -flto -pipe -D_WIN32 -W -Wall -MT > fpu-ia32.lo -MD -MP -MF .deps/fpu-ia32.Tpo -c > /home/oracle/src/ppl-0.11.1/src/fpu-ia32.cc -DDLL_EXPORT -DPIC -o > .libs/fpu-ia32.o > /home/oracle/src/ppl-0.11.1/src/fpu-ia32.cc: In function 'void > Parma_Polyhedra_Library::detect_sse_unit()': > /home/oracle/src/ppl-0.11.1/src/fpu-ia32.cc:52:7: error: 'NULL' was > not declared in this scope > make[3]: *** [fpu-ia32.lo] Error 1 > make[3]: Leaving directory `/tmp/x/src' > make[2]: *** [all] Error 2 > make[2]: Leaving directory `/tmp/x/src' > make[1]: *** [all-recursive] Error 1 > make[1]: Leaving directory `/tmp/x' > make: *** [all] Error 2 > > Here is my patch: > > $ git diff src/fpu-ia32.cc > diff --git a/src/fpu-ia32.cc b/src/fpu-ia32.cc > index d361411..8a2a6a2 100644 > --- a/src/fpu-ia32.cc > +++ b/src/fpu-ia32.cc > @@ -30,6 +30,7 @@ site: http://www.cs.unipr.it/ppl/ . */ > #include "fpu.defs.hh" > #include<csetjmp> > #include<csignal> > +#include<cstddef> > > namespace { > Hi Dongsheng, I don't see any occurrences of NULL in fpu-ia32.cc. Perhaps this is a bug in the <csetjmp> or related header file in your system? > And I'm doubt the assumption GMP does not support exception when cross > compiling: > > $ git diff m4/ac_check_gmp.m4 > diff --git a/m4/ac_check_gmp.m4 b/m4/ac_check_gmp.m4 > index c5dd1c9..8c2af74 100644 > --- a/m4/ac_check_gmp.m4 > +++ b/m4/ac_check_gmp.m4 > @@ -181,8 +181,8 @@ int main() { > ac_cv_gmp_supports_exceptions=yes, > AC_MSG_RESULT(no) > ac_cv_gmp_supports_exceptions=no, > - AC_MSG_RESULT([assuming not]) > - ac_cv_gmp_supports_exceptions=no) > + AC_MSG_RESULT([assuming yes]) > + ac_cv_gmp_supports_exceptions=yes) > > gmp_supports_exceptions=${ac_cv_gmp_supports_exceptions} > if test x"$gmp_supports_exceptions" = xyes How does this affect you? I mean, that setting only affects the PPL testsuite and if you are cross-compiling you are not using it. All the best, Roberto -- Prof. Roberto Bagnara CEO & CTO Applied Formal Methods Laboratory BUGSENG srl Department of Mathematics Parco Area delle Scienze 53/A University of Parma, Italy I-43124 Parma, Italy http://www.cs.unipr.it/~bagnara/ http://bugseng.com/ mailto:bagnara@cs.unipr.it mailto:roberto.bagnara@bugseng.com

2 2

[GIT] ppl/ppl(master): Updated.
by Roberto Bagnara 22 Feb '11

22 Feb '11

Module: ppl/ppl Branch: master Commit: 225b29f1ab3b83fc3ca6b3d68007919ad3036954 URL: http://www.cs.unipr.it/git/gitweb.cgi?p=ppl/ppl.git;a=commit;h=225b29f1ab3b… Author: Roberto Bagnara <bagnara(a)cs.unipr.it> Date: Tue Feb 22 10:13:29 2011 +0100 Updated. --- NEWS | 4 +++- 1 files changed, 3 insertions(+), 1 deletions(-) diff --git a/NEWS b/NEWS index 2db6ec8..7556a73 100644 --- a/NEWS +++ b/NEWS @@ -42,11 +42,13 @@ NEWS for version 0.11.2 (released date to be decided) Bugfixes ======== -o Fixed the semantics of the `--disable-fpmath' configure option +o Fixed the semantics of the `--disable-fpmath' configure option (which is equivalent to `--enable-fpmath=no'). It now disables all floating point computations and, consequently, all numerical abstractions based on floating point numbers. +o The PPL no longer overwrites the SIGILL signal handler. + -------------------------------------------------------------------------- NEWS for version 0.11.1 (released on February 20, 2011)

1 0

[GIT] ppl/ppl(master): No longer overwrite the SIGILL handler.
by Roberto Bagnara 22 Feb '11

22 Feb '11

Module: ppl/ppl Branch: master Commit: 3c526a80d5c063d3f66c2fd2ed438d9d5fe91274 URL: http://www.cs.unipr.it/git/gitweb.cgi?p=ppl/ppl.git;a=commit;h=3c526a80d5c0… Author: Roberto Bagnara <bagnara(a)cs.unipr.it> Date: Tue Feb 22 09:51:45 2011 +0100 No longer overwrite the SIGILL handler. --- src/fpu-ia32.cc | 8 ++++---- 1 files changed, 4 insertions(+), 4 deletions(-) diff --git a/src/fpu-ia32.cc b/src/fpu-ia32.cc index d361411..21aa6e3 100644 --- a/src/fpu-ia32.cc +++ b/src/fpu-ia32.cc @@ -37,7 +37,6 @@ jmp_buf env; void illegal_instruction_catcher(int) { - signal(SIGILL, SIG_DFL); longjmp(env, 1); } @@ -49,6 +48,7 @@ bool have_sse_unit = true; void detect_sse_unit() { + void (*old)(int); if (setjmp(env)) { // We will end up here if sse_get_control() raises SIGILL. have_sse_unit = false; @@ -56,14 +56,14 @@ detect_sse_unit() { } // Install our own signal handler for SIGILL. - signal(SIGILL, illegal_instruction_catcher); + old = signal(SIGILL, illegal_instruction_catcher); (void) sse_get_control(); // sse_get_control() did not raise SIGILL: we have an SSE unit. have_sse_unit = true; restore_sigill_handler: - // Restore the default signal handler for SIGILL. - signal(SIGILL, SIG_DFL); + // Restore the previous signal handler for SIGILL. + signal(SIGILL, old); } } // namespace Parma_Polyhedra_Library

1 0

[GIT] ppl/ppl(master): Never use `\c'; use <CODE></CODE> instead (yes, uppercase).
by Roberto Bagnara 22 Feb '11

22 Feb '11

Module: ppl/ppl Branch: master Commit: d6c9105aa28e84fdc0c7283c839a017e6d3949aa URL: http://www.cs.unipr.it/git/gitweb.cgi?p=ppl/ppl.git;a=commit;h=d6c9105aa28e… Author: Roberto Bagnara <bagnara(a)cs.unipr.it> Date: Mon Feb 21 12:19:22 2011 +0100 Never use `\c'; use <CODE></CODE> instead (yes, uppercase). --- .../Parma_Polyhedra_Library.java | 33 ++++++++++--------- 1 files changed, 17 insertions(+), 16 deletions(-) diff --git a/interfaces/Java/parma_polyhedra_library/Parma_Polyhedra_Library.java b/interfaces/Java/parma_polyhedra_library/Parma_Polyhedra_Library.java index 8c1a6e7..c692d57 100644 --- a/interfaces/Java/parma_polyhedra_library/Parma_Polyhedra_Library.java +++ b/interfaces/Java/parma_polyhedra_library/Parma_Polyhedra_Library.java @@ -38,8 +38,9 @@ site: http://www.cs.unipr.it/ppl/ . */ their representation in the PPL and the operations provided by the PPL is given in the main \extref{preamble, PPL user manual}. Here we just describe those aspects that are specific to the Java interface. - In the sequel, \c prefix is the path prefix under which - the library has been installed (typically \c /usr or \c /usr/local). + In the sequel, <CODE>prefix</CODE> is the path prefix under which + the library has been installed (typically <CODE>/usr</CODE> or + <CODE>/usr/local</CODE>). <H2>Overview</H2> @@ -49,29 +50,29 @@ site: http://www.cs.unipr.it/ppl/ . */ - When the Parma Polyhedra Library is configured, it will automatically test for the existence of the Java system (unless configuration options are passed to disable the build of the Java interface; - see configuration option <code>--enable-interfaces</code>). + see configuration option <CODE>--enable-interfaces</CODE>). If Java is correctly installed in a standard location, things will be arranged so that the Java interface is built and installed - (see configuration option <code>--with-java</code> if you need to + (see configuration option <CODE>--with-java</CODE> if you need to specify a non-standard location for the Java system). - The Java interface files are all installed in the directory - \c prefix/lib/ppl. Since this includes shared and + <CODE>prefix/lib/ppl</CODE>. Since this includes shared and dynamically loaded libraries, you must make your dynamic linker/loader aware of this fact. If you use a GNU/Linux system, try the commands <CODE>man ld.so</CODE> and <CODE>man ldconfig</CODE> for more information. - Any application using the PPL should: - - Load the PPL interface library by calling <code>System.load</code> + - Load the PPL interface library by calling <CODE>System.load</CODE> and passing the full path of the dynamic shared object; - Make sure that only the intended version(s) of the library has - been loaded, e.g., by calling static method <code>version()</code> - in class \c parma_polyhedra_library.Parma_Polyhedra_Library; + been loaded, e.g., by calling static method <CODE>version()</CODE> + in class <CODE>parma_polyhedra_library.Parma_Polyhedra_Library</CODE>; - Starting from version 0.11, initialize the interface by calling - static method <code>initialize_library()</code>; + static method <CODE>initialize_library()</CODE>; when all library work is done, finalize the interface by calling - <code>finalize_library()</code>. + <CODE>finalize_library()</CODE>. - The numerical abstract domains available to the Java user as Java classes consist of the <EM>simple</EM> domains, @@ -79,7 +80,7 @@ site: http://www.cs.unipr.it/ppl/ . */ of simple domains. Note that the default configuration will only enable a subset of these domains (if you need a different set of domains, see configuration option - <code>--enable-instantiations</code>). + <CODE>--enable-instantiations</CODE>). - The simple domains are: - convex polyhedra, which consist of C_Polyhedron and NNC_Polyhedron;<BR> @@ -159,7 +160,7 @@ public class Parma_Polyhedra_Library { This method must be called when work with the library is done. After finalization, no other library method can be called (except those in class Parma_Polyhedra_Library), unless the library - is re-initialized by calling <code>initialize_library()</code>. + is re-initialized by calling <CODE>initialize_library()</CODE>. */ public static native void finalize_library(); @@ -246,10 +247,10 @@ public class Parma_Polyhedra_Library { Computations taking exponential time will be interrupted some time after \p hsecs hundreths of seconds have elapsed since the call to the timeout setting function, by throwing a - <code>Timeout_Exception</code> object. + <CODE>Timeout_Exception</CODE> object. Otherwise, if the computation completes without being interrupted, then the timeout should be reset by calling - <code>reset_timeout()</code>. + <CODE>reset_timeout()</CODE>. */ public static native void set_timeout(int hsecs); @@ -268,10 +269,10 @@ public class Parma_Polyhedra_Library { Computations taking exponential time will be interrupted some time after reaching the \p weight complexity threshold, by throwing a - <code>Timeout_Exception</code> object. + <CODE>Timeout_Exception</CODE> object. Otherwise, if the computation completes without being interrupted, then the deterministic timeout should be reset by calling - <code>reset_deterministic_timeout()</code>. + <CODE>reset_deterministic_timeout()</CODE>. \note This "timeout" checking functionality is said to be \e deterministic

1 0

[GIT] ppl/ppl(master): The `--disable-fpmath' configure option now disables all floating point computations.
by Roberto Bagnara 22 Feb '11

22 Feb '11

Module: ppl/ppl Branch: master Commit: 07f8c154d6e0bef0e99284e1597c327beddf6b45 URL: http://www.cs.unipr.it/git/gitweb.cgi?p=ppl/ppl.git;a=commit;h=07f8c154d6e0… Author: Roberto Bagnara <bagnara(a)cs.unipr.it> Date: Mon Feb 21 09:56:13 2011 +0100 The `--disable-fpmath' configure option now disables all floating point computations. --- NEWS | 13 +++++++++++++ README.configure | 5 ++++- configure.ac | 30 ++++++++++++++++++++---------- m4/ac_check_fpu_control.m4 | 3 --- 4 files changed, 37 insertions(+), 14 deletions(-) diff --git a/NEWS b/NEWS index 8979ab9..2db6ec8 100644 --- a/NEWS +++ b/NEWS @@ -36,6 +36,19 @@ o The input routine for PPL numeric datatypes has been extended to -------------------------------------------------------------------------- +NEWS for version 0.11.2 (released date to be decided) +-------------------------------------------------------------------------- + +Bugfixes +======== + +o Fixed the semantics of the `--disable-fpmath' configure option + (which is equivalent to `--enable-fpmath=no'). It now disables all + floating point computations and, consequently, all numerical + abstractions based on floating point numbers. + + +-------------------------------------------------------------------------- NEWS for version 0.11.1 (released on February 20, 2011) -------------------------------------------------------------------------- diff --git a/README.configure b/README.configure index 7f87184..d25ac3f 100644 --- a/README.configure +++ b/README.configure @@ -357,7 +357,10 @@ For floating point computations, the option allows for selecting, on the IA32 and x86_64, the floating point instruction set. The allowed values for INSTRUCTION_SET are `sse', `sse2', `387, -`sse+387', and `sse2+387'. +`sse+387', and `sse2+387', `default', and `no'. The latter option, which +is equivalent to specifying `--disable-fpmath', has the effect of disabling +all floating point computation and, consequently, all the numerical +abstractions based on floating point numbers. On the other hand, there are configure options to request lesser degrees of optimization for the sake of debugging. These are, diff --git a/configure.ac b/configure.ac index 6fd60e8..117e8b1 100644 --- a/configure.ac +++ b/configure.ac @@ -378,23 +378,24 @@ no) esac enableval=default +use_fpmath=yes fpmath_may_use_387=yes fpmath_may_use_sse=yes -AC_MSG_CHECKING([whether to select specific floating point arithmetics]) +AC_MSG_CHECKING([whether to use (a specific) floating point arithmetic]) AC_ARG_ENABLE(fpmath, AS_HELP_STRING([--enable-fpmath=INSTRUCTION_SET], - [select floating point arithmetics])) + [select floating point arithmetic])) case "${enableval}" in sse) AC_MSG_RESULT(sse) OPT_FLAGS="$OPT_FLAGS -msse -mfpmath=sse" - # The SSE instruction set only supports single precision arithmetics: - # double and extended precision arithmetics is still done using 387. + # The SSE instruction set only supports single precision arithmetic: + # double and extended precision arithmetic is still done using 387. ;; sse2) AC_MSG_RESULT(sse2) OPT_FLAGS="$OPT_FLAGS -msse2 -mfpmath=sse" - # SSE2 still does not support extended precision arithmetics. + # SSE2 still does not support extended precision arithmetic. ;; 387) AC_MSG_RESULT(387) @@ -418,7 +419,8 @@ default) AC_MSG_RESULT(default) ;; no) - AC_MSG_RESULT(default) + AC_MSG_RESULT(no) + use_fpmath=no ;; *) AC_MSG_ERROR([bad value ${enableval} for --enable-fpmath, needs sse, sse2, 387, sse+387, sse2+387, default or no]) @@ -427,12 +429,12 @@ esac if test x"$fpmath_may_use_387" = xyes then - AC_DEFINE(PPL_FPMATH_MAY_USE_387, 1, [Defined if floating point arithmetics may use the 387 unit.]) + AC_DEFINE(PPL_FPMATH_MAY_USE_387, 1, [Defined if floating point arithmetic may use the 387 unit.]) fi if test x"$fpmath_may_use_sse" = xyes then - AC_DEFINE(PPL_FPMATH_MAY_USE_SSE, 1, [Defined if floating point arithmetics may use the SSE instruction set.]) + AC_DEFINE(PPL_FPMATH_MAY_USE_SSE, 1, [Defined if floating point arithmetic may use the SSE instruction set.]) fi # Disable all transformations and optimizations that assume default @@ -663,8 +665,16 @@ AC_LANG(C++) # The PPL uses <cmath> for sqrt and other functions in several places. AC_SEARCH_LIBS([sqrt], [m]) -# Check for the possibility to control the FPU. -AC_CHECK_FPU_CONTROL +if test x"$use_fpmath" = xyes +then + # Check for the possibility to control the FPU. + AC_CHECK_FPU_CONTROL +else + ac_cv_can_control_fpu=0 +fi +AM_CONDITIONAL(CAN_CONTROL_FPU, test $ac_cv_can_control_fpu = 1) +AC_DEFINE_UNQUOTED(PPL_CAN_CONTROL_FPU, $ac_cv_can_control_fpu, + [Not zero if the FPU can be controlled.]) # Check whether for the C++ compiler the plain char type is signed. AC_CXX_PLAIN_CHAR_IS_SIGNED diff --git a/m4/ac_check_fpu_control.m4 b/m4/ac_check_fpu_control.m4 index e285f77..cac3a32 100644 --- a/m4/ac_check_fpu_control.m4 +++ b/m4/ac_check_fpu_control.m4 @@ -141,8 +141,5 @@ main() { ac_cv_can_control_fpu=0 ) ) -AM_CONDITIONAL(CAN_CONTROL_FPU, test $ac_cv_can_control_fpu = 1) -AC_DEFINE_UNQUOTED(PPL_CAN_CONTROL_FPU, $ac_cv_can_control_fpu, - [Not zero if the FPU can be controlled.]) AC_LANG_POP(C++) ])

1 0

[GIT] ppl/ppl(ppl-0_11-branch): Updated.
by Roberto Bagnara 22 Feb '11

22 Feb '11

Module: ppl/ppl Branch: ppl-0_11-branch Commit: 3146c72703c71d604ef95f4f00f752f00e33cb69 URL: http://www.cs.unipr.it/git/gitweb.cgi?p=ppl/ppl.git;a=commit;h=3146c72703c7… Author: Roberto Bagnara <bagnara(a)cs.unipr.it> Date: Tue Feb 22 10:13:29 2011 +0100 Updated. --- NEWS | 4 +++- 1 files changed, 3 insertions(+), 1 deletions(-) diff --git a/NEWS b/NEWS index 12cc4a7..b0ed095 100644 --- a/NEWS +++ b/NEWS @@ -16,11 +16,13 @@ NEWS for version 0.11.2 (released date to be decided) Bugfixes ======== -o Fixed the semantics of the `--disable-fpmath' configure option +o Fixed the semantics of the `--disable-fpmath' configure option (which is equivalent to `--enable-fpmath=no'). It now disables all floating point computations and, consequently, all numerical abstractions based on floating point numbers. +o The PPL no longer overwrites the SIGILL signal handler. + -------------------------------------------------------------------------- NEWS for version 0.11.1 (released on February 20, 2011)

1 0