December 2009 - PPL-devel - BUGSENG Mailman

[GIT] ppl/ppl(floating_point): Implemented Box<ITV>:: affine_image on Interval Linear Forms.
by Roberto Amadini 26 Dec '09

26 Dec '09

Module: ppl/ppl Branch: floating_point Commit: 9b8c9752949ee6b96518f7009a290b64c86b35cc URL: http://www.cs.unipr.it/git/gitweb.cgi?p=ppl/ppl.git;a=commit;h=9b8c9752949e… Author: Roberto Amadini <r.amadini(a)virgilio.it> Date: Sat Dec 26 18:20:50 2009 +0100 Implemented Box<ITV>::affine_image on Interval Linear Forms. Tests in tests/Floating_Point_Expression actually use only IEEE754 floating point format. --- src/Box.defs.hh | 30 ++++++ src/Box.templates.hh | 47 ++++++++++ src/Octagonal_Shape.templates.hh | 2 +- tests/Floating_Point_Expression/Makefile.am | 7 +- tests/Floating_Point_Expression/digitalfilters1.cc | 98 +++++++------------- .../floatingpointexpr1.cc | 2 +- 6 files changed, 115 insertions(+), 71 deletions(-) Diff: http://www.cs.unipr.it/git/gitweb.cgi?p=ppl/ppl.git;a=commitdiff;h=9b8c9752…

1 0

[GIT] ppl/ppl(master): Version number bumped.
by Roberto Bagnara 22 Dec '09

22 Dec '09

Module: ppl/ppl Branch: master Commit: e8fa0e1fbc060260e754373bf5a8b45b87174786 URL: http://www.cs.unipr.it/git/gitweb.cgi?p=ppl/ppl.git;a=commit;h=e8fa0e1fbc06… Author: Roberto Bagnara <bagnara(a)cs.unipr.it> Date: Tue Dec 22 19:25:39 2009 +0100 Version number bumped. --- configure.ac | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/configure.ac b/configure.ac index 7e166a1..5ffb490 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], [0.11pre6], [ppl-devel(a)cs.unipr.it], [ppl]) +AC_INIT([the Parma Polyhedra Library], [0.11pre7], [ppl-devel(a)cs.unipr.it], [ppl]) # Minimum Autoconf version required. AC_PREREQ(2.61)

1 0

[GIT] ppl/ppl(master): Updated.
by Roberto Bagnara 22 Dec '09

22 Dec '09

Module: ppl/ppl Branch: master Commit: 9e4dbdd8255a6e30e904e193884640d0e6874842 URL: http://www.cs.unipr.it/git/gitweb.cgi?p=ppl/ppl.git;a=commit;h=9e4dbdd8255a… Author: Roberto Bagnara <bagnara(a)cs.unipr.it> Date: Tue Dec 22 19:20:54 2009 +0100 Updated. --- INSTALL | 37 +++++----- Watchdog/INSTALL | 37 +++++----- Watchdog/compile | 6 +- Watchdog/config.guess | 196 +++++++++++++++++-------------------------------- Watchdog/config.sub | 43 ++++++++--- compile | 6 +- config.guess | 196 +++++++++++++++++-------------------------------- config.sub | 43 ++++++++--- 8 files changed, 242 insertions(+), 322 deletions(-) Diff: http://www.cs.unipr.it/git/gitweb.cgi?p=ppl/ppl.git;a=commitdiff;h=9e4dbdd8…

1 0

[GIT] ppl/w3ppl(master): Added Roberto Amadini, Fabio Biselli and Fabio Bossi to the list of contributors.
by Roberto Bagnara 22 Dec '09

22 Dec '09

Module: ppl/w3ppl Branch: master Commit: fe2975c0f3f706a5d77fd24b39cf694adfe4bb54 URL: http://www.cs.unipr.it/git/gitweb.cgi?p=ppl/w3ppl.git;a=commit;h=fe2975c0f3… Author: Roberto Bagnara <bagnara(a)cs.unipr.it> Date: Tue Dec 22 10:25:15 2009 +0100 Added Roberto Amadini, Fabio Biselli and Fabio Bossi to the list of contributors. --- htdocs/Credits/Credits.raw | 19 +++++++++++++++++++ 1 files changed, 19 insertions(+), 0 deletions(-) diff --git a/htdocs/Credits/Credits.raw b/htdocs/Credits/Credits.raw index 016c81d..8365976 100644 --- a/htdocs/Credits/Credits.raw +++ b/htdocs/Credits/Credits.raw @@ -44,6 +44,12 @@ maintained and improved by the following people: <UL> <LI> + Roberto Amadini + (student of the University of Parma) + is working on the PPL support for the approximation + of floating point computations. + </LI> + <LI> Abramo Bagnara (Opera Unica) rewrote and generalized the support for checked coefficients. He also @@ -53,6 +59,19 @@ maintained and improved by the following people: design and implementation issues. </LI> <LI> + Fabio Biselli + (student of the University of Parma) + is working on the PPL support for the approximation + of floating point computations. + + </LI> + <LI> + Fabio Bossi + (student of the University of Parma) + is working on the PPL support for the approximation + of floating point computations. + </LI> + <LI> Andrea Cimino (former student of the University of Parma) wrote most of the mixed

1 0

[GIT] ppl/ppl(master): Merge branch 'floating_point'
by Roberto Bagnara 22 Dec '09

22 Dec '09

Module: ppl/ppl Branch: master Commit: 2cda93e86a8ba9567265ca6a6a5b026b2b5e78dd URL: http://www.cs.unipr.it/git/gitweb.cgi?p=ppl/ppl.git;a=commit;h=2cda93e86a8b… Author: Roberto Bagnara <bagnara(a)cs.unipr.it> Date: Tue Dec 22 09:33:21 2009 +0100 Merge branch 'floating_point' ---

1 0

[GIT] ppl/ppl(master): Capture the expected failure due to overflows.
by Roberto Bagnara 22 Dec '09

22 Dec '09

Module: ppl/ppl Branch: master Commit: 4e2145f8d498bc3ea055a06a1244de677b7bbab5 URL: http://www.cs.unipr.it/git/gitweb.cgi?p=ppl/ppl.git;a=commit;h=4e2145f8d498… Author: Roberto Bagnara <bagnara(a)cs.unipr.it> Date: Tue Dec 22 09:30:55 2009 +0100 Capture the expected failure due to overflows. --- tests/Floating_Point_Expression/digitalfilters1.cc | 36 +++++++++---------- tests/Floating_Point_Expression/polyhedron1.cc | 4 +- tests/Floating_Point_Expression/polyhedron2.cc | 4 +- 3 files changed, 21 insertions(+), 23 deletions(-) diff --git a/tests/Floating_Point_Expression/digitalfilters1.cc b/tests/Floating_Point_Expression/digitalfilters1.cc index 01d54f6..ffd5c11 100644 --- a/tests/Floating_Point_Expression/digitalfilters1.cc +++ b/tests/Floating_Point_Expression/digitalfilters1.cc @@ -46,8 +46,7 @@ for (n = 0; n < N; ++n) { } */ -// Tests rate limiter using intervals abstract domain -// and ignoring rounding errors. +// Tests rate limiter using boxes and ignoring rounding errors. bool test01() { // Input signal. @@ -139,8 +138,8 @@ test01() { return !y.is_bounded(); } -// Tests rate limiter using bounded differences abstract domain -// and ignoring rounding errors. +// Tests rate limiter using bounded differences and ignoring rounding +// errors. bool test02() { // Input signal. @@ -229,8 +228,8 @@ test02() { return !y.is_bounded(); } -// Tests rate limiter using octagons abstract domain -// and ignoring rounding errors. +// Tests rate limiter using octagonal shapes and ignoring rounding +// errors. bool test03() { // Input signal. @@ -324,8 +323,7 @@ test03() { return y.is_bounded(); } -// Tests rate limiter using polyhedra abstract domain -// and ignoring rounding errors. +// Tests rate limiter using polyhedra and ignoring rounding errors. bool test04() { // Input signal. @@ -419,8 +417,8 @@ test04() { return y.is_bounded(); } -// Tests rate limiter using bounded differences abstract domain and -// linearization of floating point expressions. +// Tests rate limiter using bounded differences and linearization of +// floating point expressions. // In order to improve the analysis, the interval domain is used // in parallel with bounded differences domain. bool @@ -598,8 +596,8 @@ test05() { return y.is_bounded(); } -// Tests rate limiter using octagons abstract domain and -// linearization of floating point expressions. +// Tests rate limiter using octagonal shapes and linearization of +// floating point expressions. // In order to improve the analysis, the interval domain is used // in parallel with octagons domain. bool @@ -784,8 +782,8 @@ test06() { return y.is_bounded(); } -// Tests rate limiter using polyhedra abstract domain and -// linearization of floating point expressions. +// Tests rate limiter using polyhedra domain and linearization of +// floating point expressions. // In order to improve the analysis, the interval domain is used // in parallel with poyhedra domain. bool @@ -968,9 +966,9 @@ test07() { BEGIN_MAIN DO_TEST(test01); DO_TEST(test02); - DO_TEST(test03); - DO_TEST(test04); - DO_TEST(test05); - DO_TEST(test06); - DO_TEST(test07); + DO_TEST_F8(test03); + DO_TEST_F64(test04); + DO_TEST_F8(test05); + DO_TEST_F8(test06); + DO_TEST_F64(test07); END_MAIN diff --git a/tests/Floating_Point_Expression/polyhedron1.cc b/tests/Floating_Point_Expression/polyhedron1.cc index f853ba3..7b3804e 100644 --- a/tests/Floating_Point_Expression/polyhedron1.cc +++ b/tests/Floating_Point_Expression/polyhedron1.cc @@ -260,8 +260,8 @@ BEGIN_MAIN DO_TEST(test01); DO_TEST(test02); DO_TEST(test03); - DO_TEST(test04); - DO_TEST(test05); + DO_TEST_F16(test04); + DO_TEST_F32(test05); DO_TEST(test06); DO_TEST(test07); END_MAIN diff --git a/tests/Floating_Point_Expression/polyhedron2.cc b/tests/Floating_Point_Expression/polyhedron2.cc index b3c7dd3..f5dffeb 100644 --- a/tests/Floating_Point_Expression/polyhedron2.cc +++ b/tests/Floating_Point_Expression/polyhedron2.cc @@ -222,6 +222,6 @@ BEGIN_MAIN DO_TEST(test01); DO_TEST(test02); DO_TEST(test03); - DO_TEST(test04); - DO_TEST(test05); + DO_TEST_F32(test04); + DO_TEST_F16(test05); END_MAIN

1 0

[GIT] ppl/ppl(floating_point): Capture the expected failure due to overflows.
by Roberto Bagnara 22 Dec '09

22 Dec '09

Module: ppl/ppl Branch: floating_point Commit: 4e2145f8d498bc3ea055a06a1244de677b7bbab5 URL: http://www.cs.unipr.it/git/gitweb.cgi?p=ppl/ppl.git;a=commit;h=4e2145f8d498… Author: Roberto Bagnara <bagnara(a)cs.unipr.it> Date: Tue Dec 22 09:30:55 2009 +0100 Capture the expected failure due to overflows. --- tests/Floating_Point_Expression/digitalfilters1.cc | 36 +++++++++---------- tests/Floating_Point_Expression/polyhedron1.cc | 4 +- tests/Floating_Point_Expression/polyhedron2.cc | 4 +- 3 files changed, 21 insertions(+), 23 deletions(-) diff --git a/tests/Floating_Point_Expression/digitalfilters1.cc b/tests/Floating_Point_Expression/digitalfilters1.cc index 01d54f6..ffd5c11 100644 --- a/tests/Floating_Point_Expression/digitalfilters1.cc +++ b/tests/Floating_Point_Expression/digitalfilters1.cc @@ -46,8 +46,7 @@ for (n = 0; n < N; ++n) { } */ -// Tests rate limiter using intervals abstract domain -// and ignoring rounding errors. +// Tests rate limiter using boxes and ignoring rounding errors. bool test01() { // Input signal. @@ -139,8 +138,8 @@ test01() { return !y.is_bounded(); } -// Tests rate limiter using bounded differences abstract domain -// and ignoring rounding errors. +// Tests rate limiter using bounded differences and ignoring rounding +// errors. bool test02() { // Input signal. @@ -229,8 +228,8 @@ test02() { return !y.is_bounded(); } -// Tests rate limiter using octagons abstract domain -// and ignoring rounding errors. +// Tests rate limiter using octagonal shapes and ignoring rounding +// errors. bool test03() { // Input signal. @@ -324,8 +323,7 @@ test03() { return y.is_bounded(); } -// Tests rate limiter using polyhedra abstract domain -// and ignoring rounding errors. +// Tests rate limiter using polyhedra and ignoring rounding errors. bool test04() { // Input signal. @@ -419,8 +417,8 @@ test04() { return y.is_bounded(); } -// Tests rate limiter using bounded differences abstract domain and -// linearization of floating point expressions. +// Tests rate limiter using bounded differences and linearization of +// floating point expressions. // In order to improve the analysis, the interval domain is used // in parallel with bounded differences domain. bool @@ -598,8 +596,8 @@ test05() { return y.is_bounded(); } -// Tests rate limiter using octagons abstract domain and -// linearization of floating point expressions. +// Tests rate limiter using octagonal shapes and linearization of +// floating point expressions. // In order to improve the analysis, the interval domain is used // in parallel with octagons domain. bool @@ -784,8 +782,8 @@ test06() { return y.is_bounded(); } -// Tests rate limiter using polyhedra abstract domain and -// linearization of floating point expressions. +// Tests rate limiter using polyhedra domain and linearization of +// floating point expressions. // In order to improve the analysis, the interval domain is used // in parallel with poyhedra domain. bool @@ -968,9 +966,9 @@ test07() { BEGIN_MAIN DO_TEST(test01); DO_TEST(test02); - DO_TEST(test03); - DO_TEST(test04); - DO_TEST(test05); - DO_TEST(test06); - DO_TEST(test07); + DO_TEST_F8(test03); + DO_TEST_F64(test04); + DO_TEST_F8(test05); + DO_TEST_F8(test06); + DO_TEST_F64(test07); END_MAIN diff --git a/tests/Floating_Point_Expression/polyhedron1.cc b/tests/Floating_Point_Expression/polyhedron1.cc index f853ba3..7b3804e 100644 --- a/tests/Floating_Point_Expression/polyhedron1.cc +++ b/tests/Floating_Point_Expression/polyhedron1.cc @@ -260,8 +260,8 @@ BEGIN_MAIN DO_TEST(test01); DO_TEST(test02); DO_TEST(test03); - DO_TEST(test04); - DO_TEST(test05); + DO_TEST_F16(test04); + DO_TEST_F32(test05); DO_TEST(test06); DO_TEST(test07); END_MAIN diff --git a/tests/Floating_Point_Expression/polyhedron2.cc b/tests/Floating_Point_Expression/polyhedron2.cc index b3c7dd3..f5dffeb 100644 --- a/tests/Floating_Point_Expression/polyhedron2.cc +++ b/tests/Floating_Point_Expression/polyhedron2.cc @@ -222,6 +222,6 @@ BEGIN_MAIN DO_TEST(test01); DO_TEST(test02); DO_TEST(test03); - DO_TEST(test04); - DO_TEST(test05); + DO_TEST_F32(test04); + DO_TEST_F16(test05); END_MAIN

1 0

[GIT] ppl/ppl(master): Never leave operators at end of line.
by Roberto Bagnara 21 Dec '09

21 Dec '09

Module: ppl/ppl Branch: master Commit: 1242e420c15da5fa9f671f4d93227ced7272f035 URL: http://www.cs.unipr.it/git/gitweb.cgi?p=ppl/ppl.git;a=commit;h=1242e420c15d… Author: Roberto Bagnara <bagnara(a)cs.unipr.it> Date: Mon Dec 21 18:28:30 2009 +0100 Never leave operators at end of line. --- src/BD_Shape.templates.hh | 3 ++- src/Boundary.defs.hh | 8 ++++---- src/Circular_Interval.defs.hh | 23 ++++++++++++----------- src/Interval.defs.hh | 4 ++-- src/Interval.inlines.hh | 6 ++++-- src/Octagonal_Shape.templates.hh | 4 ++-- src/checked_float.inlines.hh | 9 ++++++--- src/globals.defs.hh | 8 ++++---- 8 files changed, 36 insertions(+), 29 deletions(-) diff --git a/src/BD_Shape.templates.hh b/src/BD_Shape.templates.hh index 202135b..1368871 100644 --- a/src/BD_Shape.templates.hh +++ b/src/BD_Shape.templates.hh @@ -4461,7 +4461,8 @@ BD_Shape<T> bool is_right_coeff_one = (right_w_coeff == 1); bool is_right_coeff_minus_one = (right_w_coeff == -1); if (left_w_id == right_w_id) { - if ((is_left_coeff_one && is_right_coeff_one) || + if ((is_left_coeff_one && is_right_coeff_one) + || (is_left_coeff_minus_one && is_right_coeff_minus_one)) { // Here we have an identity or a constants-only constraint. return; diff --git a/src/Boundary.defs.hh b/src/Boundary.defs.hh index c4e25fc..ab7965a 100644 --- a/src/Boundary.defs.hh +++ b/src/Boundary.defs.hh @@ -316,8 +316,8 @@ is_boundary_infinity_closed(Boundary_Type type, const T& x, const Info& info) { template <typename Info> inline bool boundary_infinity_is_open(Boundary_Type type, const Info& info) { - return !Info::may_contain_infinity || - info.get_boundary_property(type, OPEN); + return !Info::may_contain_infinity + || info.get_boundary_property(type, OPEN); } template <typename T, typename Info> @@ -357,8 +357,8 @@ eq(Boundary_Type type1, const T1& x1, const Info1& info1, return is_minus_infinity(type2, x2, info2); else if (is_plus_infinity(type1, x1, info1)) return is_plus_infinity(type2, x2, info2); - else if (is_minus_infinity(type2, x2, info2) || - is_plus_infinity(type2, x2, info2)) + else if (is_minus_infinity(type2, x2, info2) + || is_plus_infinity(type2, x2, info2)) return false; else return equal(x1, x2); diff --git a/src/Circular_Interval.defs.hh b/src/Circular_Interval.defs.hh index ecf451f..f363aaa 100644 --- a/src/Circular_Interval.defs.hh +++ b/src/Circular_Interval.defs.hh @@ -217,8 +217,8 @@ private: } public: bool OK() const { - return delta_ < delta_max() || - (delta_ == delta_max() && start_ <= 1); + return delta_ < delta_max() + || (delta_ == delta_max() && start_ <= 1); } Circular_Interval() { } @@ -609,8 +609,8 @@ public: I_Result add_assign(const Circular_Interval& x, const Circular_Interval& y) { if (x.maybe_check_empty() || y.maybe_check_empty()) return assign(EMPTY); - if (x.is_universe() || y.is_universe() || - add_saturates_unsigned(delta_, x.delta_, y.delta_, Policy::modulo)) + if (x.is_universe() || y.is_universe() + || add_saturates_unsigned(delta_, x.delta_, y.delta_, Policy::modulo)) return assign(UNIVERSE); add_wrap_unsigned(start_, x.start_, y.start_, Policy::modulo); return I_NOT_DEGENERATE | I_EXACT; @@ -618,8 +618,8 @@ public: I_Result sub_assign(const Circular_Interval& x, const Circular_Interval& y) { if (x.maybe_check_empty() || y.maybe_check_empty()) return assign(EMPTY); - if (x.is_universe() || y.is_universe() || - add_saturates_unsigned(delta_, x.delta_, y.delta_, Policy::modulo)) + if (x.is_universe() || y.is_universe() + || add_saturates_unsigned(delta_, x.delta_, y.delta_, Policy::modulo)) return assign(UNIVERSE); sub_wrap_unsigned(start_, x.start_, y.end(), Policy::modulo); return I_NOT_DEGENERATE | I_EXACT; @@ -629,11 +629,12 @@ public: return assign(EMPTY); I_Result r; delta_type d; - if (y.delta_ == 0 ? - mul_saturates_unsigned(d, y.start_, x.delta_, Policy::modulo) : - (add_saturates_unsigned(d, x.start_, x.delta_, Policy::modulo) || - mul_saturates_unsigned(d, d, y.delta_, Policy::modulo) || - add_mul_saturates_unsigned(d, y.start_, x.delta_, Policy::modulo))) { + if (y.delta_ == 0 + ? mul_saturates_unsigned(d, y.start_, x.delta_, Policy::modulo) + : (add_saturates_unsigned(d, x.start_, x.delta_, Policy::modulo) + || mul_saturates_unsigned(d, d, y.delta_, Policy::modulo) + || add_mul_saturates_unsigned(d, y.start_, x.delta_, + Policy::modulo))) { assign(UNIVERSE); r = I_UNIVERSE; } diff --git a/src/Interval.defs.hh b/src/Interval.defs.hh index a40b971..1eba8ca 100644 --- a/src/Interval.defs.hh +++ b/src/Interval.defs.hh @@ -239,8 +239,8 @@ public: } bool check_empty(I_Result r) const { - return (r & I_ANY) == I_EMPTY || - ((r & I_ANY) != I_NOT_EMPTY && is_empty()); + return (r & I_ANY) == I_EMPTY + || ((r & I_ANY) != I_NOT_EMPTY && is_empty()); } bool is_singleton() const { diff --git a/src/Interval.inlines.hh b/src/Interval.inlines.hh index 69ae867..94212b8 100644 --- a/src/Interval.inlines.hh +++ b/src/Interval.inlines.hh @@ -381,7 +381,8 @@ inline typename Enable_If<Is_Singleton<From>::value Interval<To_Boundary, To_Info>::difference_assign(const From& x) { PPL_ASSERT(f_OK(x)); // FIXME: restrictions - if (lt(UPPER, upper(), info(), LOWER, f_lower(x), f_info(x)) || + if (lt(UPPER, upper(), info(), LOWER, f_lower(x), f_info(x)) + || gt(LOWER, lower(), info(), UPPER, f_upper(x), f_info(x))) return combine(V_EQ, V_EQ); bool nl = ge(LOWER, lower(), info(), LOWER, f_lower(x), f_info(x)); @@ -416,7 +417,8 @@ Interval<To_Boundary, To_Info>::difference_assign(const From1& x, PPL_DIRTY_TEMP(To_Info, to_info); to_info.clear(); // FIXME: restrictions - if (lt(UPPER, f_upper(x), f_info(x), LOWER, f_lower(y), f_info(y)) || + if (lt(UPPER, f_upper(x), f_info(x), LOWER, f_lower(y), f_info(y)) + || gt(LOWER, f_lower(x), f_info(x), UPPER, f_upper(y), f_info(y))) return assign(x); bool nl = ge(LOWER, f_lower(x), f_info(x), LOWER, f_lower(y), f_info(y)); diff --git a/src/Octagonal_Shape.templates.hh b/src/Octagonal_Shape.templates.hh index 19a0463..4838527 100644 --- a/src/Octagonal_Shape.templates.hh +++ b/src/Octagonal_Shape.templates.hh @@ -634,8 +634,8 @@ Octagonal_Shape<T>::refine_with_linear_form_inequality( bool is_right_coeff_one = (right_w_coeff == 1); bool is_right_coeff_minus_one = (right_w_coeff == -1); if (left_w_id == right_w_id) { - if ((is_left_coeff_one && is_right_coeff_one) || - (is_left_coeff_minus_one && is_right_coeff_minus_one)) { + if ((is_left_coeff_one && is_right_coeff_one) + || (is_left_coeff_minus_one && is_right_coeff_minus_one)) { // Here we have an identity or a constants-only constraint. PPL_ASSERT(OK()); return; diff --git a/src/checked_float.inlines.hh b/src/checked_float.inlines.hh index 3655786..e01ebed 100644 --- a/src/checked_float.inlines.hh +++ b/src/checked_float.inlines.hh @@ -495,7 +495,8 @@ template <typename To_Policy, typename From1_Policy, typename From2_Policy, inline Result mul_float(Type& to, const Type x, const Type y, Rounding_Dir dir) { if (To_Policy::check_inf_mul_zero - && ((x == 0 && is_inf_float<From2_Policy>(y)) || + && ((x == 0 && is_inf_float<From2_Policy>(y)) + || (y == 0 && is_inf_float<From1_Policy>(x)))) { return assign_nan<To_Policy>(to, V_INF_MUL_ZERO); } @@ -911,7 +912,8 @@ template <typename To_Policy, typename From1_Policy, typename From2_Policy, inline Result add_mul_float(Type& to, const Type x, const Type y, Rounding_Dir dir) { if (To_Policy::check_inf_mul_zero - && ((x == 0 && is_inf_float<From2_Policy>(y)) || + && ((x == 0 && is_inf_float<From2_Policy>(y)) + || (y == 0 && is_inf_float<From1_Policy>(x)))) { return assign_nan<To_Policy>(to, V_INF_MUL_ZERO); } @@ -943,7 +945,8 @@ template <typename To_Policy, typename From1_Policy, typename From2_Policy, type inline Result sub_mul_float(Type& to, const Type x, const Type y, Rounding_Dir dir) { if (To_Policy::check_inf_mul_zero - && ((x == 0 && is_inf_float<From2_Policy>(y)) || + && ((x == 0 && is_inf_float<From2_Policy>(y)) + || (y == 0 && is_inf_float<From1_Policy>(x)))) { return assign_nan<To_Policy>(to, V_INF_MUL_ZERO); } diff --git a/src/globals.defs.hh b/src/globals.defs.hh index 65d3804..52a7edf 100644 --- a/src/globals.defs.hh +++ b/src/globals.defs.hh @@ -363,16 +363,16 @@ struct Constant_ : public TConstant<T, v> { template <typename T, long long v, bool prefer_signed> struct Constant_<T, v, prefer_signed, typename Enable_If<(Fit<typename C_Integer<T>::smaller_signed_type, v>::value - && (prefer_signed || - !Fit<typename C_Integer<T>::smaller_unsigned_type, v>::value))>::type> + && (prefer_signed + || !Fit<typename C_Integer<T>::smaller_unsigned_type, v>::value))>::type> : public Constant_<typename C_Integer<T>::smaller_signed_type, v, prefer_signed> { }; template <typename T, long long v, bool prefer_signed> struct Constant_<T, v, prefer_signed, typename Enable_If<(Fit<typename C_Integer<T>::smaller_unsigned_type, v>::value - && (!prefer_signed || - !Fit<typename C_Integer<T>::smaller_signed_type, v>::value))>::type> + && (!prefer_signed + || !Fit<typename C_Integer<T>::smaller_signed_type, v>::value))>::type> : public Constant_<typename C_Integer<T>::smaller_unsigned_type, v, prefer_signed> { };

1 0

[GIT] ppl/ppl(master): Never leave operators at end of line.
by Roberto Bagnara 21 Dec '09

21 Dec '09

Module: ppl/ppl Branch: master Commit: 6b88ad1f70258ac43e9775852dbaf2fe1c5509b8 URL: http://www.cs.unipr.it/git/gitweb.cgi?p=ppl/ppl.git;a=commit;h=6b88ad1f7025… Author: Roberto Bagnara <bagnara(a)cs.unipr.it> Date: Mon Dec 21 18:24:08 2009 +0100 Never leave operators at end of line. --- src/Boundary.defs.hh | 24 ++++++++++---------- src/Circular_Interval.defs.hh | 7 +++-- src/Difference_Floating_Point_Expression.defs.hh | 8 +++--- src/Division_Floating_Point_Expression.defs.hh | 8 +++--- ...Division_Floating_Point_Expression.templates.hh | 4 +- src/Floating_Point_Expression.templates.hh | 15 +++++------ src/Interval.defs.hh | 6 +++- src/Interval.inlines.hh | 8 +++--- src/MIP_Problem.cc | 6 ++-- ...ultiplication_Floating_Point_Expression.defs.hh | 8 +++--- ...lication_Floating_Point_Expression.templates.hh | 12 +++++----- src/Opposite_Floating_Point_Expression.defs.hh | 4 +- src/Sum_Floating_Point_Expression.defs.hh | 4 +- src/checked_float.inlines.hh | 8 +++--- src/wrap_assign.hh | 6 ++-- 15 files changed, 65 insertions(+), 63 deletions(-) Diff: http://www.cs.unipr.it/git/gitweb.cgi?p=ppl/ppl.git;a=commitdiff;h=6b88ad1f…

1 0

[GIT] ppl/ppl(floating_point): Never leave operators at end of line.
by Roberto Bagnara 21 Dec '09

21 Dec '09

Module: ppl/ppl Branch: floating_point Commit: 6b88ad1f70258ac43e9775852dbaf2fe1c5509b8 URL: http://www.cs.unipr.it/git/gitweb.cgi?p=ppl/ppl.git;a=commit;h=6b88ad1f7025… Author: Roberto Bagnara <bagnara(a)cs.unipr.it> Date: Mon Dec 21 18:24:08 2009 +0100 Never leave operators at end of line. --- src/Boundary.defs.hh | 24 ++++++++++---------- src/Circular_Interval.defs.hh | 7 +++-- src/Difference_Floating_Point_Expression.defs.hh | 8 +++--- src/Division_Floating_Point_Expression.defs.hh | 8 +++--- ...Division_Floating_Point_Expression.templates.hh | 4 +- src/Floating_Point_Expression.templates.hh | 15 +++++------ src/Interval.defs.hh | 6 +++- src/Interval.inlines.hh | 8 +++--- src/MIP_Problem.cc | 6 ++-- ...ultiplication_Floating_Point_Expression.defs.hh | 8 +++--- ...lication_Floating_Point_Expression.templates.hh | 12 +++++----- src/Opposite_Floating_Point_Expression.defs.hh | 4 +- src/Sum_Floating_Point_Expression.defs.hh | 4 +- src/checked_float.inlines.hh | 8 +++--- src/wrap_assign.hh | 6 ++-- 15 files changed, 65 insertions(+), 63 deletions(-) Diff: http://www.cs.unipr.it/git/gitweb.cgi?p=ppl/ppl.git;a=commitdiff;h=6b88ad1f…

1 0