float128: Add test-{float128,ifloat128,float128-finite}

This adds test support for float128, and lays some groundwork for future
_FloatN types.

	* math/gen-libm-test.pl (@all_floats): Add ifloat128 and float128.
	(%all_floats_pfx): Add macro prefix for float128 (FLT128).
	* math/libm-test-exp10.inc (pow10_test): Do not test for _FloatN,
	* math/libm-test-isfinite.inc (finite_test): Likewise.
	* math/libm-test-lgamma.inc (gamma_test): Likewise.
	* math/libm-test-nexttoward.inc (nexttoward_test): Likewise.
	(nexttoward_test_data}: Likewise.
	* math/libm-test-remainder.inc (drem_test): Likewise.
	* math/libm-test-scalb.inc (scalb_test): Likewise.
	(scalb_test_data): Likewise.
	* math/libm-test-significand.inc (significand_test): Likewise.
	(significand_test_data): Likewise.
	* math/libm-test-support.c (check_complex): Replace __complex__
	FLOAT with CFLOAT to get the support for old compiler.
	* math/libm-test-support.h (check_complex): Likewise.
	* math/test-double.h (CFLOAT, TEST_FLOATN): New macros.
	* math/test-float.h (CFLOAT, TEST_FLOATN): Likewise.
	* math/test-ldouble.h (CFLOAT, TEST_FLOATN): Likewise.
	* math/test-float128.h: New file.
	* math/test-math-floatn.h: New file.

ChangeLog

@@ -1,3 +1,28 @@
+2017-06-12  Paul E. Murphy  <murphyp@linux.vnet.ibm.com>
+	    Gabriel F. T. Gomes  <gftg@linux.vnet.ibm.com>
+	    Tulio Magno Quites Machado Filho  <tuliom@linux.vnet.ibm.com>
+
+	* math/gen-libm-test.pl (@all_floats): Add ifloat128 and float128.
+	(%all_floats_pfx): Add macro prefix for float128 (FLT128).
+	* math/libm-test-exp10.inc (pow10_test): Do not test for _FloatN,
+	* math/libm-test-isfinite.inc (finite_test): Likewise.
+	* math/libm-test-lgamma.inc (gamma_test): Likewise.
+	* math/libm-test-nexttoward.inc (nexttoward_test): Likewise.
+	(nexttoward_test_data}: Likewise.
+	* math/libm-test-remainder.inc (drem_test): Likewise.
+	* math/libm-test-scalb.inc (scalb_test): Likewise.
+	(scalb_test_data): Likewise.
+	* math/libm-test-significand.inc (significand_test): Likewise.
+	(significand_test_data): Likewise.
+	* math/libm-test-support.c (check_complex): Replace __complex__
+	FLOAT with CFLOAT to get the support for old compiler.
+	* math/libm-test-support.h (check_complex): Likewise.
+	* math/test-double.h (CFLOAT, TEST_FLOATN): New macros.
+	* math/test-float.h (CFLOAT, TEST_FLOATN): Likewise.
+	* math/test-ldouble.h (CFLOAT, TEST_FLOATN): Likewise.
+	* math/test-float128.h: New file.
+	* math/test-math-floatn.h: New file.
+
 2017-06-12  Gabriel F. T. Gomes  <gftg@linux.vnet.ibm.com>
 
 	* include/float.h: Allow the definition of macros prefixed with

math/gen-libm-test.pl

@@ -44,14 +44,15 @@ use vars qw ($ulps_file);
 use vars qw (%auto_tests);
 
 # all_floats is sorted and contains all recognised float types
-@all_floats = ('double', 'float', 'idouble',
+@all_floats = ('double', 'float', 'float128', 'idouble',
-	       'ifloat', 'ildouble', 'ldouble');
+	       'ifloat', 'ifloat128', 'ildouble', 'ldouble');
 
 # all_floats_pfx maps C types to their C like prefix for macros.
 %all_floats_pfx =
   ( "double" => "DBL",
     "ldouble" => "LDBL",
     "float" => "FLT",
+    "float128" => "FLT128",
   );
 
 %beautify =

math/libm-test-exp10.inc

@@ -39,8 +39,10 @@ exp10_test (void)
 static void
 pow10_test (void)
 {
+#if !TEST_FLOATN
   /* pow10 uses the same test data as exp10.  */
   ALL_RM_TEST (pow10, 0, exp10_test_data, RUN_TEST_LOOP_f_f, END);
+#endif
 }
 
 static void

math/libm-test-isfinite.inc

@@ -47,7 +47,9 @@ static void
 finite_test (void)
 {
   /* finite uses the same test data as isfinite.  */
+#if !TEST_FLOATN
   ALL_RM_TEST (finite, 1, isfinite_test_data, RUN_TEST_LOOP_f_b, END);
+#endif
 }
 
 static void

math/libm-test-lgamma.inc

@@ -45,8 +45,10 @@ lgamma_test (void)
 static void
 gamma_test (void)
 {
+#if !TEST_FLOATN
   /* gamma uses the same test data as lgamma.  */
   ALL_RM_TEST (gamma, 0, lgamma_test_data, RUN_TEST_LOOP_f_f1, END, signgam);
+#endif
 }
 
 static void

math/libm-test-nexttoward.inc

@@ -20,6 +20,7 @@
 
 static const struct test_fj_f_data nexttoward_test_data[] =
   {
+#if !TEST_FLOATN
     TEST_fj_f (nexttoward, 0, 0, 0, NO_INEXACT_EXCEPTION|ERRNO_UNCHANGED),
     TEST_fj_f (nexttoward, minus_zero, 0, 0, NO_INEXACT_EXCEPTION|ERRNO_UNCHANGED),
     TEST_fj_f (nexttoward, 0, minus_zero, minus_zero, NO_INEXACT_EXCEPTION|ERRNO_UNCHANGED),
@@ -80,6 +81,7 @@ static const struct test_fj_f_data nexttoward_test_data[] =
     TEST_fj_f (nexttoward, min_subnorm_value, minus_zero, 0, INEXACT_EXCEPTION|UNDERFLOW_EXCEPTION|ERRNO_ERANGE),
     TEST_fj_f (nexttoward, -min_subnorm_value, 0, minus_zero, INEXACT_EXCEPTION|UNDERFLOW_EXCEPTION|ERRNO_ERANGE),
     TEST_fj_f (nexttoward, -min_subnorm_value, minus_zero, minus_zero, INEXACT_EXCEPTION|UNDERFLOW_EXCEPTION|ERRNO_ERANGE),
+#endif
 
 #if TEST_COND_binary32
     TEST_fj_f (nexttoward, 1.0, 1.1L, 0x1.000002p0, NO_INEXACT_EXCEPTION|ERRNO_UNCHANGED),
@@ -153,7 +155,9 @@ static const struct test_fj_f_data nexttoward_test_data[] =
 static void
 nexttoward_test (void)
 {
+#if !TEST_FLOATN
   ALL_RM_TEST (nexttoward, 1, nexttoward_test_data, RUN_TEST_LOOP_fj_f, END);
+#endif
 }
 
 static void

math/libm-test-remainder.inc

@@ -184,8 +184,10 @@ remainder_test (void)
 static void
 drem_test (void)
 {
+#if !TEST_FLOATN
   /* drem uses the same test data as remainder.  */
   ALL_RM_TEST (drem, 1, remainder_test_data, RUN_TEST_LOOP_ff_f, END);
+#endif
 }
 
 static void

math/libm-test-scalb.inc

@@ -20,6 +20,7 @@
 
 static const struct test_ff_f_data scalb_test_data[] =
   {
+#if !TEST_FLOATN
     /* Results in this case are unspecified by POSIX, so, for an
        otherwise fully-determined function, spurious "inexact"
        exceptions are OK.  */
@@ -144,12 +145,15 @@ static const struct test_ff_f_data scalb_test_data[] =
 
     TEST_ff_f (scalb, 0.8L, 4, 12.8L, NO_INEXACT_EXCEPTION|ERRNO_UNCHANGED),
     TEST_ff_f (scalb, -0.854375L, 5, -27.34L, NO_INEXACT_EXCEPTION|ERRNO_UNCHANGED),
+#endif
   };
 
 static void
 scalb_test (void)
 {
+#if !TEST_FLOATN
   ALL_RM_TEST (scalb, 1, scalb_test_data, RUN_TEST_LOOP_ff_f, END);
+#endif
 }
 
 static void

math/libm-test-significand.inc

@@ -20,6 +20,7 @@
 
 static const struct test_f_f_data significand_test_data[] =
   {
+#if !TEST_FLOATN
     /* significand returns the mantissa of the exponential
        representation.  Tests for infinities, zero and NaNs reflect
        the implementation (including possibility of "inexact" and
@@ -45,12 +46,15 @@ static const struct test_f_f_data significand_test_data[] =
     TEST_f_f (significand, -6.0, -1.5, NO_INEXACT_EXCEPTION|ERRNO_UNCHANGED),
     TEST_f_f (significand, 8.0, 1.0, NO_INEXACT_EXCEPTION|ERRNO_UNCHANGED),
     TEST_f_f (significand, -8.0, -1.0, NO_INEXACT_EXCEPTION|ERRNO_UNCHANGED),
+#endif
   };
 
 static void
 significand_test (void)
 {
+#if !TEST_FLOATN
   ALL_RM_TEST (significand, 1, significand_test_data, RUN_TEST_LOOP_f_f, END);
+#endif
 }
 
 static void

math/libm-test-support.c

@@ -732,8 +732,8 @@ check_float (const char *test_name, FLOAT computed, FLOAT expected,
 
 
 void
-check_complex (const char *test_name, __complex__ FLOAT computed,
+check_complex (const char *test_name, CFLOAT computed,
-	       __complex__ FLOAT expected,
+	       CFLOAT expected,
 	       int exception)
 {
   FLOAT part_comp, part_exp;

math/libm-test-support.h

@@ -143,7 +143,7 @@ void init_max_error (const char *, int);
 void print_max_error (const char *);
 void print_complex_max_error (const char *);
 void check_float (const char *, FLOAT, FLOAT, int);
-void check_complex (const char *, __complex__ FLOAT, __complex__ FLOAT, int);
+void check_complex (const char *, CFLOAT, CFLOAT, int);
 void check_int (const char *, int, int, int);
 void check_long (const char *, long int, long int, int);
 void check_bool (const char *, int, int, int);

math/test-double.h

@@ -18,6 +18,7 @@
 
 #define FUNC(function) function
 #define FLOAT double
+#define CFLOAT __complex__ double
 #define BUILD_COMPLEX(real, imag) (CMPLX ((real), (imag)))
 #define PREFIX DBL
 #define LIT(x) (x)
@@ -25,3 +26,4 @@
 #define LITM(x) x
 #define FTOSTR strfromd
 #define snan_value_MACRO SNAN
+#define TEST_FLOATN 0

math/test-float.h

@@ -18,6 +18,7 @@
 
 #define FUNC(function) function ## f
 #define FLOAT float
+#define CFLOAT __complex__ float
 #define BUILD_COMPLEX(real, imag) (CMPLXF ((real), (imag)))
 #define PREFIX FLT
 #define TYPE_STR "float"
@@ -26,3 +27,4 @@
 #define LITM(x) x
 #define FTOSTR strfromf
 #define snan_value_MACRO SNANF
+#define TEST_FLOATN 0

math/test-float128.h

@@ -0,0 +1,35 @@
+/* Common definitions for libm tests for _Float128.
+
+   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
+   <http://www.gnu.org/licenses/>.  */
+
+#include "test-math-floatn.h"
+
+/* Fixup builtins and constants for older compilers.  */
+#include <bits/floatn.h>
+#include <float.h>
+
+#define FUNC(function) function ## f128
+#define FLOAT _Float128
+#define CFLOAT __CFLOAT128
+#define BUILD_COMPLEX(real, imag) (CMPLXF128 ((real), (imag)))
+#define PREFIX FLT128
+#define TYPE_STR "float128"
+#define LIT(x) __f128 (x)
+#define LITM(x) x ## f128
+#define FTOSTR strfromf128
+#define snan_value_MACRO SNANF128

math/test-ldouble.h

@@ -18,6 +18,7 @@
 
 #define FUNC(function) function##l
 #define FLOAT long double
+#define CFLOAT __complex__ long double
 #define BUILD_COMPLEX(real, imag) (CMPLXL ((real), (imag)))
 #define PREFIX LDBL
 #define TYPE_STR "ldouble"
@@ -25,3 +26,4 @@
 #define LITM(x) x ## l
 #define FTOSTR strfroml
 #define snan_value_MACRO SNANL
+#define TEST_FLOATN 0

math/test-math-floatn.h

@@ -0,0 +1,20 @@
+/* Common definitions for libm tests for TS 18661-3 derived types.
+   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
+   <http://www.gnu.org/licenses/>.  */
+
+#define TEST_FLOATN 1
+#define __STDC_WANT_IEC_60559_TYPES_EXT__