Make common fdim implementation generic.

The only difference is the usage of math_narrow_eval when
building s_fdiml.c.  This should be harmless for long double,
but I did observe some code generation changes on m68k, but
lack the resources to test it.

Likewise, to more easily support overriding symbol generation,
the aliasing macros are always conditionally defined on their
absence to reduce boilerplate.

I also ran builds for i486, ppc64, sparcv9, aarch64,
s390x and observed no changes to s_fdim* objects.

ChangeLog

@@ -1,3 +1,37 @@
+2016-09-01  Paul E. Murphy  <murphyp@linux.vnet.ibm.com>
+
+	* math/Makefile (gen-libm-calls): Add s_fdim.
+	(libm-calls): Move to above.
+
+	* math/s_fdim.c: Renamed and refactored into to ...
+	* math/s_fdim_template.c: New file.
+	* math/s_fdiml.c: Removed.
+	* sysdeps/ieee754/ldbl-opt/s_fdim.c: Removed.
+	* sysdeps/ieee754/ldbl-opt/s_fdiml.c: Removed.
+
+	* sysdeps/generic/math-type-macros-double.h:
+	(declare_mgen_alias): Don't declare if already declared.
+	* sysdeps/ieee754/ldbl-opt/math-type-macros-ldouble.h:
+	Likewise.
+
+	* sysdeps/ieee754/ldbl-opt/math-type-macros-double.h:
+	(declare_mgen_alias): Don't declare if already declared.
+	(M_LIBM_NEED_COMPAT): Likewise.
+	(declare_mgen_libm_compat): Likewise.
+	(LDOUBLE_fdiml_libm_version): New macro.
+
+	* sysdeps/sparc/sparc32/sparcv9/fpu/multiarch/s_fdim.c:
+	Include math/s_fdim.c now that ldbl-opt version is not needed.
+	(declare_mgen_alias): New macro to disable aliasing.
+
+	* sysdeps/sparc/sparc32/sparcv9/fpu/multiarch/s_fdimf-vis3.c:
+	(declare_mgen_alias): Likewise.
+	* sysdeps/sparc/sparc32/sparcv9/fpu/multiarch/s_fdimf.c:
+	(declare_mgen_alias): Likewise.
+	* sysdeps/sparc/sparc32/sparcv9/fpu/multiarch/s_fdim-vis3.c:
+	(declare_mgen_alias): Likewise.
+	(M_LIBM_NEED_COMPAT): Likewise.
+
 2016-09-01  Florian Weimer  <fweimer@redhat.com>
 
 	[BZ #20525]

math/Makefile

@@ -49,7 +49,7 @@ gen-libm-calls = cargF conjF cimagF crealF cabsF s_cacosF		  \
 	         s_cacoshF s_ccosF s_ccoshF s_casinF s_csinF s_casinhF	  \
 		 k_casinhF s_csinhF k_casinhF s_csinhF s_catanhF s_catanF \
 		 s_ctanF s_ctanhF s_cexpF s_clogF s_cprojF s_csqrtF	  \
-		 s_cpowF s_clog10F
+		 s_cpowF s_clog10F s_fdimF
 
 libm-calls =								  \
 	e_acosF e_acoshF e_asinF e_atan2F e_atanhF e_coshF e_expF e_fmodF \
@@ -65,7 +65,7 @@ libm-calls =								  \
 	w_tgammaF w_hypotF w_j0F w_j1F w_jnF w_lgammaF w_lgammaF_r	  \
 	w_logF w_log10F w_powF w_remainderF w_scalbF w_sinhF w_sqrtF	  \
 	w_ilogbF							  \
-	s_fpclassifyF s_fmaxF s_fminF s_fdimF s_nanF s_truncF		  \
+	s_fpclassifyF s_fmaxF s_fminF s_nanF s_truncF		  	  \
 	s_remquoF e_log2F e_exp2F s_roundF s_nearbyintF s_sincosF	  \
 	s_fmaF s_lrintF s_llrintF s_lroundF s_llroundF e_exp10F w_log2F	  \
 	s_issignalingF $(calls:s_%=m_%) x2y2m1F				  \

math/s_fdim.c

@@ -1,42 +0,0 @@
-/* Return positive difference between arguments.
-   Copyright (C) 1997-2016 Free Software Foundation, Inc.
-   This file is part of the GNU C Library.
-   Contributed by Ulrich Drepper <drepper@cygnus.com>, 1997.
-
-   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 <errno.h>
-#include <math.h>
-#include <math_private.h>
-
-double
-__fdim (double x, double y)
-{
-  if (islessequal (x, y))
-    return 0.0;
-
-  double r = math_narrow_eval (x - y);
-  if (isinf (r) && !isinf (x) && !isinf (y))
-    __set_errno (ERANGE);
-
-  return r;
-}
-#ifndef __fdim
-weak_alias (__fdim, fdim)
-# ifdef NO_LONG_DOUBLE
-strong_alias (__fdim, __fdiml)
-weak_alias (__fdim, fdiml)
-# endif
-#endif

math/s_fdim_template.c

@@ -21,18 +21,20 @@
 #include <math.h>
 #include <math_private.h>
 
-float
-__fdimf (float x, float y)
+FLOAT
+M_DECL_FUNC (__fdim) (FLOAT x, FLOAT y)
 {
   if (islessequal (x, y))
-    return 0.0f;
+    return 0;
 
-  float r = math_narrow_eval (x - y);
+  FLOAT r = math_narrow_eval (x - y);
   if (isinf (r) && !isinf (x) && !isinf (y))
     __set_errno (ERANGE);
 
   return r;
 }
-#ifndef __fdimf
-weak_alias (__fdimf, fdimf)
+declare_mgen_alias (__fdim, fdim);
+
+#if M_LIBM_NEED_COMPAT (fdim)
+declare_mgen_libm_compat (__fdim, fdim)
 #endif

math/s_fdiml.c

@@ -1,35 +0,0 @@
-/* Return positive difference between arguments.
-   Copyright (C) 1997-2016 Free Software Foundation, Inc.
-   This file is part of the GNU C Library.
-   Contributed by Ulrich Drepper <drepper@cygnus.com>, 1997.
-
-   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 <errno.h>
-#include <math.h>
-
-long double
-__fdiml (long double x, long double y)
-{
-  if (islessequal (x, y))
-    return 0.0f;
-
-  long double r = x - y;
-  if (isinf (r) && !isinf (x) && !isinf (y))
-    __set_errno (ERANGE);
-
-  return r;
-}
-weak_alias (__fdiml, fdiml)

sysdeps/generic/math-type-macros-double.h

@@ -29,12 +29,12 @@
 /* Machines without a distinct long double type
    alias long double functions to their double
    equivalent.  */
-#if defined NO_LONG_DOUBLE
+#if defined NO_LONG_DOUBLE && !defined declare_mgen_alias
 # define declare_mgen_alias(from, to)	    \
    weak_alias (from, to)		    \
    strong_alias (from, from ## l)	    \
    weak_alias (from, to ## l)
-#else
+#elif !defined declare_mgen_alias
 # define declare_mgen_alias(from, to)	    \
    weak_alias (M_SUF (from), M_SUF (to))
 #endif

sysdeps/ieee754/ldbl-opt/math-type-macros-double.h

@@ -44,15 +44,20 @@
 #define LDOUBLE_cpowl_libm_version GLIBC_2_1
 #define LDOUBLE_clog10l_libm_version GLIBC_2_1
 #define LDOUBLE___clog10l_libm_version GLIBC_2_1
+#define LDOUBLE_fdiml_libm_version GLIBC_2_1
 
 /* Define compat symbols for long double on platforms
    where it was not always a distinct type.  */
+#if !defined M_LIBM_NEED_COMPAT
 # define M_LIBM_NEED_COMPAT(f) \
   LONG_DOUBLE_COMPAT (libm, LDOUBLE_ ## f ## l_libm_version)
+#endif
 
+#if !defined declare_mgen_libm_compat
 # define declare_mgen_libm_compat(from, to)	      \
   compat_symbol (libm, from, to ## l,		      \
 		 LDOUBLE_ ## to ## l_libm_version);
+#endif
 
 #include_next <math-type-macros-double.h>
 #endif

sysdeps/ieee754/ldbl-opt/math-type-macros-ldouble.h

@@ -22,8 +22,10 @@
 
 /* Use properly versioned symbols for long double on platforms where
    it was not always a distinct type.  */
+#if !defined declare_mgen_alias
 # define declare_mgen_alias(from, to) \
   long_double_symbol (libm, from ## l, to ## l);
+#endif
 
 #include_next <math-type-macros-ldouble.h>
 #endif

sysdeps/ieee754/ldbl-opt/s_fdim.c

@@ -1,5 +0,0 @@
-#include <math_ldbl_opt.h>
-#include <math/s_fdim.c>
-#if LONG_DOUBLE_COMPAT(libm, GLIBC_2_1)
-compat_symbol (libm, __fdim, fdiml, GLIBC_2_1);
-#endif

sysdeps/ieee754/ldbl-opt/s_fdiml.c

@@ -1,5 +0,0 @@
-#include <math_ldbl_opt.h>
-#undef weak_alias
-#define weak_alias(n,a)
-#include <math/s_fdiml.c>
-long_double_symbol (libm, __fdiml, fdiml);

sysdeps/sparc/sparc32/sparcv9/fpu/multiarch/s_fdim-vis3.c

@@ -19,5 +19,7 @@
 #include <math.h>
 
 #define __fdim __fdim_vis3
+#define declare_mgen_alias(t, f)
+#define M_LIBM_NEED_COMPAT(f) 0
 
 #include <math/s_fdim.c>

sysdeps/sparc/sparc32/sparcv9/fpu/multiarch/s_fdim.c

@@ -27,6 +27,7 @@ sparc_libm_ifunc(__fdim, hwcap & HWCAP_SPARC_VIS3 ? __fdim_vis3 : __fdim_generic
 weak_alias (__fdim, fdim)
 
 # define __fdim __fdim_generic
+# define declare_mgen_alias(t, f)
 #endif
 
-#include <ldbl-opt/s_fdim.c>
+#include <math/s_fdim.c>

sysdeps/sparc/sparc32/sparcv9/fpu/multiarch/s_fdimf-vis3.c

@@ -19,5 +19,6 @@
 #include <math.h>
 
 #define __fdimf __fdimf_vis3
+#define declare_mgen_alias(t, f)
 
 #include <math/s_fdimf.c>

sysdeps/sparc/sparc32/sparcv9/fpu/multiarch/s_fdimf.c

@@ -27,6 +27,8 @@ sparc_libm_ifunc(__fdimf, hwcap & HWCAP_SPARC_VIS3 ? __fdimf_vis3 : __fdimf_gene
 weak_alias (__fdimf, fdimf)
 
 # define __fdimf __fdimf_generic
+# define declare_mgen_alias(t, f)
+
 #endif
 
 #include <math/s_fdimf.c>