From patchwork Mon Aug 21 21:04:02 2017 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Gabriel F T Gomes X-Patchwork-Id: 22293 Received: (qmail 84419 invoked by alias); 21 Aug 2017 21:04:20 -0000 Mailing-List: contact libc-alpha-help@sourceware.org; run by ezmlm Precedence: bulk List-Id: List-Unsubscribe: List-Subscribe: List-Archive: List-Post: List-Help: , Sender: libc-alpha-owner@sourceware.org Delivered-To: mailing list libc-alpha@sourceware.org Received: (qmail 84385 invoked by uid 89); 21 Aug 2017 21:04:19 -0000 Authentication-Results: sourceware.org; auth=none X-Virus-Found: No X-Spam-SWARE-Status: No, score=-24.3 required=5.0 tests=AWL, BAYES_00, GIT_PATCH_0, GIT_PATCH_1, GIT_PATCH_2, GIT_PATCH_3, KAM_LAZY_DOMAIN_SECURITY, RCVD_IN_DNSWL_LOW autolearn=ham version=3.3.2 spammy= X-HELO: mx0a-001b2d01.pphosted.com From: "Gabriel F. T. Gomes" To: libc-alpha@sourceware.org Subject: [PATCH v3] Provide a C++ version of issignaling that does not use __MATH_TG Date: Mon, 21 Aug 2017 18:04:02 -0300 X-TM-AS-GCONF: 00 x-cbid: 17082121-0024-0000-0000-000002C2CFA8 X-IBM-SpamModules-Scores: X-IBM-SpamModules-Versions: BY=3.00007587; HX=3.00000241; KW=3.00000007; PH=3.00000004; SC=3.00000223; SDB=6.00905697; UDB=6.00453871; IPR=6.00685900; BA=6.00005545; NDR=6.00000001; ZLA=6.00000005; ZF=6.00000009; ZB=6.00000000; ZP=6.00000000; ZH=6.00000000; ZU=6.00000002; MB=3.00016799; XFM=3.00000015; UTC=2017-08-21 21:04:05 X-IBM-AV-DETECTION: SAVI=unused REMOTE=unused XFE=unused x-cbparentid: 17082121-0025-0000-0000-0000452B5F8E Message-Id: <20170821210402.12835-1-gftg@linux.vnet.ibm.com> X-Proofpoint-Virus-Version: vendor=fsecure engine=2.50.10432:, , definitions=2017-08-21_14:, , signatures=0 X-Proofpoint-Spam-Details: rule=outbound_notspam policy=outbound score=0 spamscore=0 suspectscore=4 malwarescore=0 phishscore=0 adultscore=0 bulkscore=0 classifier=spam adjust=0 reason=mlx scancount=1 engine=8.0.1-1707230000 definitions=main-1708210329 Changes since v2: - Added inline keyword to function definitions. - Replace __HAVE_FLOAT128 with __HAVE_DISTINCT_FLOAT128 in the test case. - Fixed various white-space errors. Changes since v1: - Added test case. -- 8< -- The macro __MATH_TG contains the logic to select between long double and _Float128, when these types are ABI-distinct. This logic relies on __builtin_types_compatible_p, which is not available in C++ mode. On the other hand, C++ function overloading provides the means to distinguish between the floating-point types. The overloading resolution will match the correct parameter regardless of type qualifiers, i.e.: const and volatile. Tested for powerpc64le, s390x, and x86_64. * math/math.h [defined __cplusplus] (issignaling): Provide a C++ definition for issignaling that does not rely on __MATH_TG, since __MATH_TG uses __builtin_types_compatible_p, which is only available in C mode. (CFLAGS-test-math-issignaling.cc): New variable. * math/Makefile [CXX] (tests): Add test-math-issignaling. * math/test-math-issignaling.cc: New test for C++ implementation of type-generic issignaling. * sysdeps/powerpc/powerpc64le/Makefile [subdir == math] (CXXFLAGS-test-math-issignaling.cc): Add -mfloat128 to the build options of test-math-issignaling on powerpc64le. --- math/Makefile | 3 +- math/math.h | 19 +++++- math/test-math-issignaling.cc | 113 +++++++++++++++++++++++++++++++++++ sysdeps/powerpc/powerpc64le/Makefile | 1 + 4 files changed, 134 insertions(+), 2 deletions(-) create mode 100644 math/test-math-issignaling.cc diff --git a/math/Makefile b/math/Makefile index e09b0c0545..0130fcf38b 100644 --- a/math/Makefile +++ b/math/Makefile @@ -203,7 +203,7 @@ tests-static = test-fpucw-static test-fpucw-ieee-static \ test-signgam-ullong-static test-signgam-ullong-init-static ifneq (,$(CXX)) -tests += test-math-isinff test-math-iszero +tests += test-math-isinff test-math-iszero test-math-issignaling endif ifneq (no,$(PERL)) @@ -350,6 +350,7 @@ CFLAGS-test-signgam-ullong-init-static.c = -std=c99 CFLAGS-test-math-isinff.cc = -std=gnu++11 CFLAGS-test-math-iszero.cc = -std=gnu++11 +CFLAGS-test-math-issignaling.cc = -std=gnu++11 CFLAGS-test-iszero-excess-precision.c = -fexcess-precision=standard CFLAGS-test-iseqsig-excess-precision.c = -fexcess-precision=standard diff --git a/math/math.h b/math/math.h index dea8dbe1ae..add86af724 100644 --- a/math/math.h +++ b/math/math.h @@ -474,7 +474,24 @@ enum # include /* Return nonzero value if X is a signaling NaN. */ -# define issignaling(x) __MATH_TG ((x), __issignaling, (x)) +# ifndef __cplusplus +# define issignaling(x) __MATH_TG ((x), __issignaling, (x)) +# else + /* In C++ mode, __MATH_TG cannot be used, because it relies on + __builtin_types_compatible_p, which is a C-only builtin. On the + other hand, overloading provides the means to distinguish between + the floating-point types. The overloading resolution will match + the correct parameter (regardless of type qualifiers (i.e.: const + and volatile). */ +extern "C++" { +inline int issignaling (float __val) { return __issignalingf (__val); } +inline int issignaling (double __val) { return __issignaling (__val); } +inline int issignaling (long double __val) { return __issignalingl (__val); } +# if __HAVE_DISTINCT_FLOAT128 +inline int issignaling (_Float128 __val) { return __issignalingf128 (__val); } +# endif +} /* extern C++ */ +# endif /* Return nonzero value if X is subnormal. */ # define issubnormal(x) (fpclassify (x) == FP_SUBNORMAL) diff --git a/math/test-math-issignaling.cc b/math/test-math-issignaling.cc new file mode 100644 index 0000000000..72b2a92ebe --- /dev/null +++ b/math/test-math-issignaling.cc @@ -0,0 +1,113 @@ +/* Test for the C++ implementation of issignaling. + Copyright (C) 2017 Free Software Foundation, Inc. + This file is part of the GNU C Library. + + The GNU C Library is free software; you can redistribute it and/or + modify it under the terms of the GNU Lesser General Public + License as published by the Free Software Foundation; either + version 2.1 of the License, or (at your option) any later version. + + The GNU C Library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public + License along with the GNU C Library; if not, see + . */ + +#define _GNU_SOURCE 1 +#include +#include + +#include + +/* There is no signaling_NaN for _Float128 in std::numeric_limits. + Include ieee754_float128.h and use the bitfields in the union + ieee854_float128.ieee_nan to build a signaling NaN. */ +#if __HAVE_DISTINCT_FLOAT128 +# include +#endif + +static bool errors; + +static void +check (int actual, int expected, const char *actual_expr, int line) +{ + if (actual != expected) + { + errors = true; + printf ("%s:%d: error: %s\n", __FILE__, line, actual_expr); + printf ("%s:%d: expected: %d\n", __FILE__, line, expected); + printf ("%s:%d: actual: %d\n", __FILE__, line, actual); + } +} + +#define CHECK(actual, expected) \ + check ((actual), (expected), #actual, __LINE__) + +template +static void +check_type () +{ + typedef std::numeric_limits limits; + CHECK (issignaling (T{0}), 0); + if (limits::has_infinity) + { + CHECK (issignaling (limits::infinity ()), 0); + CHECK (issignaling (-limits::infinity ()), 0); + } + if (limits::has_quiet_NaN) + CHECK (issignaling (limits::quiet_NaN ()), 0); + if (limits::has_signaling_NaN) + CHECK (issignaling (limits::signaling_NaN ()), 1); +} + +#if __HAVE_DISTINCT_FLOAT128 +static void +check_float128 () +{ + ieee854_float128 q; + + q.d = 0; + CHECK (issignaling (q.d), 0); + + /* Infinity. */ + q.ieee.negative = 0; + q.ieee.exponent = 0x7FFF; + q.ieee.mantissa0 = 0x0000; + q.ieee.mantissa1 = 0x00000000; + q.ieee.mantissa2 = 0x00000000; + q.ieee.mantissa3 = 0x00000000; + CHECK (issignaling (q.d), 0); + + /* Quite NaN. */ + q.ieee_nan.quiet_nan = 1; + q.ieee_nan.mantissa0 = 0x0000; + CHECK (issignaling (q.d), 0); + + /* Still a quite NaN. */ + q.ieee_nan.quiet_nan = 1; + q.ieee_nan.mantissa0 = 0x4000; + CHECK (issignaling (q.d), 0); + + /* Signaling NaN. */ + q.ieee_nan.quiet_nan = 0; + q.ieee_nan.mantissa0 = 0x4000; + CHECK (issignaling (q.d), 1); +} +#endif + +static int +do_test (void) +{ + check_type (); + check_type (); + check_type (); +#if __HAVE_DISTINCT_FLOAT128 + check_float128 (); +#endif + return errors; +} + +#include diff --git a/sysdeps/powerpc/powerpc64le/Makefile b/sysdeps/powerpc/powerpc64le/Makefile index 77617b670a..19adbfa1c1 100644 --- a/sysdeps/powerpc/powerpc64le/Makefile +++ b/sysdeps/powerpc/powerpc64le/Makefile @@ -16,6 +16,7 @@ $(foreach suf,$(all-object-suffixes),%f128_r$(suf)): CFLAGS += -mfloat128 $(foreach suf,$(all-object-suffixes),$(objpfx)test-float128%$(suf)): CFLAGS += -mfloat128 $(foreach suf,$(all-object-suffixes),$(objpfx)test-ifloat128%$(suf)): CFLAGS += -mfloat128 CFLAGS-libm-test-support-float128.c += -mfloat128 +CFLAGS-test-math-issignaling.cc += -mfloat128 $(objpfx)test-float128% $(objpfx)test-ifloat128%: \ gnulib-tests += $(f128-loader-link) endif