Add fmaximum, fminimum functions

C2X adds new <math.h> functions for floating-point maximum and
minimum, corresponding to the new operations that were added in IEEE
754-2019 because of concerns about the old operations not being
associative in the presence of signaling NaNs.  fmaximum and fminimum
handle NaNs like most <math.h> functions (any NaN argument means the
result is a quiet NaN).  fmaximum_num and fminimum_num handle both
quiet and signaling NaNs the way fmax and fmin handle quiet NaNs (if
one argument is a number and the other is a NaN, return the number),
but still raise "invalid" for a signaling NaN argument, making them
exceptions to the normal rule that a function with a floating-point
result raising "invalid" also returns a quiet NaN.  fmaximum_mag,
fminimum_mag, fmaximum_mag_num and fminimum_mag_num are corresponding
functions returning the argument with greatest or least absolute
value.  All these functions also treat +0 as greater than -0.  There
are also corresponding <tgmath.h> type-generic macros.

Add these functions to glibc.  The implementations use type-generic
templates based on those for fmax, fmin, fmaxmag and fminmag, and test
inputs are based on those for those functions with appropriate
adjustments to the expected results.  The RISC-V maintainers might
wish to add optimized versions of fmaximum_num and fminimum_num (for
float and double), since RISC-V (F extension version 2.2 and later)
provides instructions corresponding to those functions - though it
might be at least as useful to add architecture-independent built-in
functions to GCC and teach the RISC-V back end to expand those
functions inline, which is what you generally want for functions that
can be implemented with a single instruction.

Tested for x86_64 and x86, and with build-many-glibcs.py.

10 files changed, 227 insertions, 1 deletions

diff --git a/sysdeps/ieee754/ldbl-128ibm-compat/Versions b/sysdeps/ieee754/ldbl-128ibm-compat/Versions
index 03a66134c4..d4ffd3762b 100644
--- a/sysdeps/ieee754/ldbl-128ibm-compat/Versions
+++ b/sysdeps/ieee754/ldbl-128ibm-compat/Versions
@@ -128,6 +128,14 @@ libm {
     __f64fmaieee128;
     __f32sqrtieee128;
     __f64sqrtieee128;
+    __fmaximumieee128;
+    __fmaximum_numieee128;
+    __fmaximum_magieee128;
+    __fmaximum_mag_numieee128;
+    __fminimumieee128;
+    __fminimum_numieee128;
+    __fminimum_magieee128;
+    __fminimum_mag_numieee128;
   }
 }
 libc {
diff --git a/sysdeps/ieee754/ldbl-opt/Makefile b/sysdeps/ieee754/ldbl-opt/Makefile
index 6b21680033..1d01846476 100644
--- a/sysdeps/ieee754/ldbl-opt/Makefile
+++ b/sysdeps/ieee754/ldbl-opt/Makefile
@@ -45,7 +45,9 @@ libnldbl-calls = asprintf dprintf fprintf fscanf fwprintf fwscanf iovfscanf \
 		 nextup nextdown totalorder totalordermag getpayload \
 		 canonicalize setpayload setpayloadsig llogb fmaxmag fminmag \
 		 roundeven fromfp ufromfp fromfpx ufromfpx fadd dadd \
-		 fdiv ddiv ffma dfma fmul dmul fsqrt dsqrt fsub dsub
+		 fdiv ddiv ffma dfma fmul dmul fsqrt dsqrt fsub dsub \
+		 fmaximum fmaximum_mag fmaximum_num fmaximum_mag_num \
+		 fminimum fminimum_mag fminimum_num fminimum_mag_num
 libnldbl-routines = $(libnldbl-calls:%=nldbl-%)
 libnldbl-inhibit-o = $(object-suffixes)
 libnldbl-static-only-routines = $(libnldbl-routines)
@@ -108,8 +110,16 @@ CFLAGS-nldbl-finite.c = -fno-builtin-finitel
 CFLAGS-nldbl-floor.c = -fno-builtin-floorl
 CFLAGS-nldbl-fma.c = -fno-builtin-fmal
 CFLAGS-nldbl-fmax.c = -fno-builtin-fmaxl
+CFLAGS-nldbl-fmaximum.c += -fno-builtin-fmaximuml
+CFLAGS-nldbl-fmaximum_mag.c += -fno-builtin-fmaximum_magl
+CFLAGS-nldbl-fmaximum_mag_num.c += -fno-builtin-fmaximum_mag_numl
+CFLAGS-nldbl-fmaximum_num.c += -fno-builtin-fmaximum_numl
 CFLAGS-nldbl-fmaxmag.c = -fno-builtin-fmaxmagl
 CFLAGS-nldbl-fmin.c = -fno-builtin-fminl
+CFLAGS-nldbl-fminimum.c += -fno-builtin-fminimuml
+CFLAGS-nldbl-fminimum_mag.c += -fno-builtin-fminimum_magl
+CFLAGS-nldbl-fminimum_mag_num.c += -fno-builtin-fminimum_mag_numl
+CFLAGS-nldbl-fminimum_num.c += -fno-builtin-fminimum_numl
 CFLAGS-nldbl-fminmag.c = -fno-builtin-fminmagl
 CFLAGS-nldbl-fmod.c = -fno-builtin-fmodl
 CFLAGS-nldbl-fmul.c = -fno-builtin-fmull
diff --git a/sysdeps/ieee754/ldbl-opt/nldbl-fmaximum.c b/sysdeps/ieee754/ldbl-opt/nldbl-fmaximum.c
new file mode 100644
index 0000000000..f26de00184
--- /dev/null
+++ b/sysdeps/ieee754/ldbl-opt/nldbl-fmaximum.c
@@ -0,0 +1,26 @@
+/* Compatibility routine for IEEE double as long double for fmaximum.
+   Copyright (C) 2021 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
+   <https://www.gnu.org/licenses/>.  */
+
+#include "nldbl-compat.h"
+
+double
+attribute_hidden
+fmaximuml (double x, double y)
+{
+  return fmaximum (x, y);
+}
diff --git a/sysdeps/ieee754/ldbl-opt/nldbl-fmaximum_mag.c b/sysdeps/ieee754/ldbl-opt/nldbl-fmaximum_mag.c
new file mode 100644
index 0000000000..5c6f9e6fbb
--- /dev/null
+++ b/sysdeps/ieee754/ldbl-opt/nldbl-fmaximum_mag.c
@@ -0,0 +1,26 @@
+/* Compatibility routine for IEEE double as long double for fmaximum_mag.
+   Copyright (C) 2021 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
+   <https://www.gnu.org/licenses/>.  */
+
+#include "nldbl-compat.h"
+
+double
+attribute_hidden
+fmaximum_magl (double x, double y)
+{
+  return fmaximum_mag (x, y);
+}
diff --git a/sysdeps/ieee754/ldbl-opt/nldbl-fmaximum_mag_num.c b/sysdeps/ieee754/ldbl-opt/nldbl-fmaximum_mag_num.c
new file mode 100644
index 0000000000..66c84e3ec0
--- /dev/null
+++ b/sysdeps/ieee754/ldbl-opt/nldbl-fmaximum_mag_num.c
@@ -0,0 +1,26 @@
+/* Compatibility routine for IEEE double as long double for fmaximum_mag_num.
+   Copyright (C) 2021 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
+   <https://www.gnu.org/licenses/>.  */
+
+#include "nldbl-compat.h"
+
+double
+attribute_hidden
+fmaximum_mag_numl (double x, double y)
+{
+  return fmaximum_mag_num (x, y);
+}
diff --git a/sysdeps/ieee754/ldbl-opt/nldbl-fmaximum_num.c b/sysdeps/ieee754/ldbl-opt/nldbl-fmaximum_num.c
new file mode 100644
index 0000000000..72fba4a477
--- /dev/null
+++ b/sysdeps/ieee754/ldbl-opt/nldbl-fmaximum_num.c
@@ -0,0 +1,26 @@
+/* Compatibility routine for IEEE double as long double for fmaximum_num.
+   Copyright (C) 2021 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
+   <https://www.gnu.org/licenses/>.  */
+
+#include "nldbl-compat.h"
+
+double
+attribute_hidden
+fmaximum_numl (double x, double y)
+{
+  return fmaximum_num (x, y);
+}
diff --git a/sysdeps/ieee754/ldbl-opt/nldbl-fminimum.c b/sysdeps/ieee754/ldbl-opt/nldbl-fminimum.c
new file mode 100644
index 0000000000..9247a814fe
--- /dev/null
+++ b/sysdeps/ieee754/ldbl-opt/nldbl-fminimum.c
@@ -0,0 +1,26 @@
+/* Compatibility routine for IEEE double as long double for fminimum.
+   Copyright (C) 2021 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
+   <https://www.gnu.org/licenses/>.  */
+
+#include "nldbl-compat.h"
+
+double
+attribute_hidden
+fminimuml (double x, double y)
+{
+  return fminimum (x, y);
+}
diff --git a/sysdeps/ieee754/ldbl-opt/nldbl-fminimum_mag.c b/sysdeps/ieee754/ldbl-opt/nldbl-fminimum_mag.c
new file mode 100644
index 0000000000..175f6d7d92
--- /dev/null
+++ b/sysdeps/ieee754/ldbl-opt/nldbl-fminimum_mag.c
@@ -0,0 +1,26 @@
+/* Compatibility routine for IEEE double as long double for fminimum_mag.
+   Copyright (C) 2021 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
+   <https://www.gnu.org/licenses/>.  */
+
+#include "nldbl-compat.h"
+
+double
+attribute_hidden
+fminimum_magl (double x, double y)
+{
+  return fminimum_mag (x, y);
+}
diff --git a/sysdeps/ieee754/ldbl-opt/nldbl-fminimum_mag_num.c b/sysdeps/ieee754/ldbl-opt/nldbl-fminimum_mag_num.c
new file mode 100644
index 0000000000..b37b4a5e68
--- /dev/null
+++ b/sysdeps/ieee754/ldbl-opt/nldbl-fminimum_mag_num.c
@@ -0,0 +1,26 @@
+/* Compatibility routine for IEEE double as long double for fminimum_mag_num.
+   Copyright (C) 2021 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
+   <https://www.gnu.org/licenses/>.  */
+
+#include "nldbl-compat.h"
+
+double
+attribute_hidden
+fminimum_mag_numl (double x, double y)
+{
+  return fminimum_mag_num (x, y);
+}
diff --git a/sysdeps/ieee754/ldbl-opt/nldbl-fminimum_num.c b/sysdeps/ieee754/ldbl-opt/nldbl-fminimum_num.c
new file mode 100644
index 0000000000..2cbd5a8188
--- /dev/null
+++ b/sysdeps/ieee754/ldbl-opt/nldbl-fminimum_num.c
@@ -0,0 +1,26 @@
+/* Compatibility routine for IEEE double as long double for fminimum_num.
+   Copyright (C) 2021 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
+   <https://www.gnu.org/licenses/>.  */
+
+#include "nldbl-compat.h"
+
+double
+attribute_hidden
+fminimum_numl (double x, double y)
+{
+  return fminimum_num (x, y);
+}