math: Use log1pf from CORE-MATH

The CORE-MATH implementation is correctly rounded (for any rounding mode)
and shows slight better performance to the generic log1pf.

The code was adapted to glibc style and to use the definition of
math_config.h (to handle errno, overflow, and underflow).

Benchtest on x64_64 (Ryzen 9 5900X, gcc 14.2.1), aarch64 (M1,
gcc 13.2.1), and powerpc (POWER10, gcc 13.2.1):

Latency                      master        patched   improvement
x86_64                      71.8142        38.9668        45.74%
x86_64v2                    71.9094        39.1321        45.58%
x86_64v3                    60.1000        32.4016        46.09%
i686                        147.105        104.258        29.13%
aarch64                     26.4439        14.0050        47.04%
power10                     19.4874         9.4146        51.69%
powerpc                     17.6145        8.00736        54.54%

reciprocal-throughput        master        patched   improvement
x86_64                      19.7604        12.7254        35.60%
x86_64v2                    19.0039        11.9455        37.14%
x86_64v3                    16.8559        11.9317        29.21%
i686                        82.3426        73.9718        10.17%
aarch64                     14.4665         7.9614        44.97%
power10                     11.9974         8.4117        29.89%
powerpc                     7.15222         6.0914        14.83%

Signed-off-by: Alexei Sibidanov <sibid@uvic.ca>
Signed-off-by: Paul Zimmermann <Paul.Zimmermann@inria.fr>
Signed-off-by: Adhemerval Zanella <adhemerval.zanella@linaro.org>
Reviewed-by: DJ Delorie <dj@redhat.com>

29 files changed, 189 insertions, 340 deletions

diff --git a/sysdeps/aarch64/libm-test-ulps b/sysdeps/aarch64/libm-test-ulps
index cb1c18c..5700c71 100644
--- a/sysdeps/aarch64/libm-test-ulps
+++ b/sysdeps/aarch64/libm-test-ulps
@@ -1366,7 +1366,6 @@ ldouble: 4
 
 Function: "log1p":
 double: 1
-float: 1
 ldouble: 3
 
 Function: "log1p_advsimd":
@@ -1375,7 +1374,6 @@ float: 1
 
 Function: "log1p_downward":
 double: 1
-float: 2
 ldouble: 3
 
 Function: "log1p_sve":
@@ -1384,12 +1382,10 @@ float: 1
 
 Function: "log1p_towardzero":
 double: 2
-float: 2
 ldouble: 3
 
 Function: "log1p_upward":
 double: 2
-float: 2
 ldouble: 2
 
 Function: "log2":
@@ -1459,22 +1455,18 @@ ldouble: 1
 
 Function: "logp1":
 double: 1
-float: 1
 ldouble: 3
 
 Function: "logp1_downward":
 double: 1
-float: 2
 ldouble: 3
 
 Function: "logp1_towardzero":
 double: 2
-float: 2
 ldouble: 3
 
 Function: "logp1_upward":
 double: 2
-float: 2
 ldouble: 2
 
 Function: "pow":
diff --git a/sysdeps/alpha/fpu/libm-test-ulps b/sysdeps/alpha/fpu/libm-test-ulps
index 09185dd..212c52c 100644
--- a/sysdeps/alpha/fpu/libm-test-ulps
+++ b/sysdeps/alpha/fpu/libm-test-ulps
@@ -1202,22 +1202,18 @@ ldouble: 3
 
 Function: "log1p":
 double: 1
-float: 1
 ldouble: 3
 
 Function: "log1p_downward":
 double: 2
-float: 2
 ldouble: 3
 
 Function: "log1p_towardzero":
 double: 2
-float: 2
 ldouble: 3
 
 Function: "log1p_upward":
 double: 2
-float: 2
 ldouble: 2
 
 Function: "log2":
@@ -1267,22 +1263,18 @@ ldouble: 1
 
 Function: "logp1":
 double: 1
-float: 1
 ldouble: 3
 
 Function: "logp1_downward":
 double: 2
-float: 2
 ldouble: 3
 
 Function: "logp1_towardzero":
 double: 2
-float: 2
 ldouble: 3
 
 Function: "logp1_upward":
 double: 2
-float: 2
 ldouble: 2
 
 Function: "pow":
diff --git a/sysdeps/arc/fpu/libm-test-ulps b/sysdeps/arc/fpu/libm-test-ulps
index d2ec67b..f73aae2 100644
--- a/sysdeps/arc/fpu/libm-test-ulps
+++ b/sysdeps/arc/fpu/libm-test-ulps
@@ -981,19 +981,15 @@ float: 3
 
 Function: "log1p":
 double: 1
-float: 1
 
 Function: "log1p_downward":
 double: 1
-float: 2
 
 Function: "log1p_towardzero":
 double: 2
-float: 2
 
 Function: "log1p_upward":
 double: 2
-float: 2
 
 Function: "log2":
 double: 2
@@ -1029,19 +1025,15 @@ float: 1
 
 Function: "logp1":
 double: 1
-float: 1
 
 Function: "logp1_downward":
 double: 1
-float: 2
 
 Function: "logp1_towardzero":
 double: 2
-float: 2
 
 Function: "logp1_upward":
 double: 2
-float: 2
 
 Function: "pow":
 double: 1
diff --git a/sysdeps/arc/nofpu/libm-test-ulps b/sysdeps/arc/nofpu/libm-test-ulps
index fe74fb8..f6d1ef0 100644
--- a/sysdeps/arc/nofpu/libm-test-ulps
+++ b/sysdeps/arc/nofpu/libm-test-ulps
@@ -238,7 +238,6 @@ float: 2
 
 Function: "log1p":
 double: 1
-float: 1
 
 Function: "log2":
 double: 2
diff --git a/sysdeps/arm/libm-test-ulps b/sysdeps/arm/libm-test-ulps
index b6b1941..502680f 100644
--- a/sysdeps/arm/libm-test-ulps
+++ b/sysdeps/arm/libm-test-ulps
@@ -974,19 +974,15 @@ float: 3
 
 Function: "log1p":
 double: 1
-float: 1
 
 Function: "log1p_downward":
 double: 2
-float: 2
 
 Function: "log1p_towardzero":
 double: 2
-float: 2
 
 Function: "log1p_upward":
 double: 2
-float: 2
 
 Function: "log2":
 double: 2
@@ -1028,19 +1024,15 @@ float: 2
 
 Function: "logp1":
 double: 1
-float: 1
 
 Function: "logp1_downward":
 double: 2
-float: 2
 
 Function: "logp1_towardzero":
 double: 2
-float: 2
 
 Function: "logp1_upward":
 double: 2
-float: 2
 
 Function: "pow":
 double: 1
diff --git a/sysdeps/csky/fpu/libm-test-ulps b/sysdeps/csky/fpu/libm-test-ulps
index 3f17778..a7b2bec 100644
--- a/sysdeps/csky/fpu/libm-test-ulps
+++ b/sysdeps/csky/fpu/libm-test-ulps
@@ -919,19 +919,15 @@ float: 2
 
 Function: "log1p":
 double: 1
-float: 1
 
 Function: "log1p_downward":
 double: 2
-float: 2
 
 Function: "log1p_towardzero":
 double: 2
-float: 2
 
 Function: "log1p_upward":
 double: 2
-float: 2
 
 Function: "log2":
 double: 2
@@ -948,19 +944,15 @@ double: 3
 
 Function: "logp1":
 double: 1
-float: 1
 
 Function: "logp1_downward":
 double: 2
-float: 2
 
 Function: "logp1_towardzero":
 double: 2
-float: 2
 
 Function: "logp1_upward":
 double: 2
-float: 2
 
 Function: "pow":
 double: 1
diff --git a/sysdeps/csky/nofpu/libm-test-ulps b/sysdeps/csky/nofpu/libm-test-ulps
index fdef935..4e4451a 100644
--- a/sysdeps/csky/nofpu/libm-test-ulps
+++ b/sysdeps/csky/nofpu/libm-test-ulps
@@ -920,19 +920,15 @@ float: 2
 
 Function: "log1p":
 double: 1
-float: 1
 
 Function: "log1p_downward":
 double: 2
-float: 2
 
 Function: "log1p_towardzero":
 double: 2
-float: 2
 
 Function: "log1p_upward":
 double: 2
-float: 2
 
 Function: "log2":
 double: 2
@@ -962,19 +958,15 @@ float: 2
 
 Function: "logp1":
 double: 1
-float: 1
 
 Function: "logp1_downward":
 double: 2
-float: 2
 
 Function: "logp1_towardzero":
 double: 2
-float: 2
 
 Function: "logp1_upward":
 double: 2
-float: 2
 
 Function: "pow":
 double: 1
diff --git a/sysdeps/hppa/fpu/libm-test-ulps b/sysdeps/hppa/fpu/libm-test-ulps
index 9f89528..b56f512 100644
--- a/sysdeps/hppa/fpu/libm-test-ulps
+++ b/sysdeps/hppa/fpu/libm-test-ulps
@@ -997,19 +997,15 @@ float: 3
 
 Function: "log1p":
 double: 1
-float: 1
 
 Function: "log1p_downward":
 double: 2
-float: 2
 
 Function: "log1p_towardzero":
 double: 2
-float: 2
 
 Function: "log1p_upward":
 double: 2
-float: 2
 
 Function: "log2":
 double: 2
@@ -1051,19 +1047,15 @@ float: 2
 
 Function: "logp1":
 double: 1
-float: 1
 
 Function: "logp1_downward":
 double: 2
-float: 2
 
 Function: "logp1_towardzero":
 double: 2
-float: 2
 
 Function: "logp1_upward":
 double: 2
-float: 2
 
 Function: "pow":
 double: 1
diff --git a/sysdeps/i386/fpu/libm-test-ulps b/sysdeps/i386/fpu/libm-test-ulps
index d0bd633..63ef356 100644
--- a/sysdeps/i386/fpu/libm-test-ulps
+++ b/sysdeps/i386/fpu/libm-test-ulps
@@ -1443,19 +1443,16 @@ ldouble: 2
 
 Function: "log1p_downward":
 double: 1
-float: 1
 float128: 3
 ldouble: 4
 
 Function: "log1p_towardzero":
 double: 1
-float: 1
 float128: 3
 ldouble: 4
 
 Function: "log1p_upward":
 double: 1
-float: 1
 float128: 2
 ldouble: 3
 
@@ -1524,19 +1521,16 @@ ldouble: 2
 
 Function: "logp1_downward":
 double: 1
-float: 1
 float128: 3
 ldouble: 4
 
 Function: "logp1_towardzero":
 double: 1
-float: 1
 float128: 3
 ldouble: 4
 
 Function: "logp1_upward":
 double: 1
-float: 1
 float128: 2
 ldouble: 3
 
diff --git a/sysdeps/i386/fpu/s_log1pf.S b/sysdeps/i386/fpu/s_log1pf.S
deleted file mode 100644
index 82a3f44..0000000
--- a/sysdeps/i386/fpu/s_log1pf.S
+++ /dev/null
@@ -1,66 +0,0 @@
-/*
- * Public domain.
- */
-
-#include <machine/asm.h>
-#include <i386-math-asm.h>
-
-RCSID("$NetBSD: s_log1pf.S,v 1.4 1995/05/09 00:13:05 jtc Exp $")
-
-	.section .rodata
-
-	.align ALIGNARG(4)
-	/* The fyl2xp1 can only be used for values in
-		-1 + sqrt(2) / 2 <= x <= 1 - sqrt(2) / 2
-	   0.29 is a safe value.
-	*/
-limit:	.float 0.29
-one:	.float 1.0
-
-DEFINE_FLT_MIN
-
-#ifdef PIC
-# define MO(op) op##@GOTOFF(%edx)
-#else
-# define MO(op) op
-#endif
-
-/*
- * Use the fyl2xp1 function when the argument is in the range -0.29 to 0.29,
- * otherwise fyl2x with the needed extra computation.
- */
-	.text
-ENTRY(__log1pf)
-	fldln2
-
-	flds	4(%esp)
-
-#ifdef	PIC
-	LOAD_PIC_REG (dx)
-#endif
-
-	fxam
-	fnstsw
-	fld	%st
-	sahf
-	jc	3f		// in case x is NaN or ħInf
-4:	fabs
-	fcomps	MO(limit)
-	fnstsw
-	sahf
-	jc	2f
-
-	fadds	MO(one)
-	fyl2x
-	ret
-
-2:	fyl2xp1
-	FLT_CHECK_FORCE_UFLOW_NONNAN
-	ret
-
-3:	jp	4b		// in case x is ħInf
-	fstp	%st(1)
-	fstp	%st(1)
-	ret
-
-END (__log1pf)
diff --git a/sysdeps/i386/i686/fpu/multiarch/libm-test-ulps b/sysdeps/i386/i686/fpu/multiarch/libm-test-ulps
index a97722b..f6393db 100644
--- a/sysdeps/i386/i686/fpu/multiarch/libm-test-ulps
+++ b/sysdeps/i386/i686/fpu/multiarch/libm-test-ulps
@@ -1446,19 +1446,16 @@ ldouble: 2
 
 Function: "log1p_downward":
 double: 1
-float: 1
 float128: 3
 ldouble: 4
 
 Function: "log1p_towardzero":
 double: 1
-float: 1
 float128: 3
 ldouble: 4
 
 Function: "log1p_upward":
 double: 1
-float: 1
 float128: 2
 ldouble: 3
 
@@ -1527,19 +1524,16 @@ ldouble: 2
 
 Function: "logp1_downward":
 double: 1
-float: 1
 float128: 3
 ldouble: 4
 
 Function: "logp1_towardzero":
 double: 1
-float: 1
 float128: 3
 ldouble: 4
 
 Function: "logp1_upward":
 double: 1
-float: 1
 float128: 2
 ldouble: 3
 
diff --git a/sysdeps/ieee754/flt-32/s_log1pf.c b/sysdeps/ieee754/flt-32/s_log1pf.c
index 555f0f8..d1686e7 100644
--- a/sysdeps/ieee754/flt-32/s_log1pf.c
+++ b/sysdeps/ieee754/flt-32/s_log1pf.c
@@ -1,116 +1,181 @@
-/* s_log1pf.c -- float version of s_log1p.c.
- */
+/* Correctly-rounded biased argument natural logarithm function for binary32
+   value.
 
-/*
- * ====================================================
- * Copyright (C) 1993 by Sun Microsystems, Inc. All rights reserved.
- *
- * Developed at SunPro, a Sun Microsystems, Inc. business.
- * Permission to use, copy, modify, and distribute this
- * software is freely granted, provided that this notice
- * is preserved.
- * ====================================================
- */
+Copyright (c) 2023, 2024  Alexei Sibidanov.
 
-#include <float.h>
-#include <math.h>
-#include <math-barriers.h>
-#include <math_private.h>
-#include <math-underflow.h>
-#include <libc-diag.h>
+This file is part of the CORE-MATH project
+project (file src/binary32/log1p/log1pf.c revision bc385c2).
 
-static const float
-ln2_hi =   6.9313812256e-01,	/* 0x3f317180 */
-ln2_lo =   9.0580006145e-06,	/* 0x3717f7d1 */
-two25 =    3.355443200e+07,	/* 0x4c000000 */
-Lp1 = 6.6666668653e-01,	/* 3F2AAAAB */
-Lp2 = 4.0000000596e-01,	/* 3ECCCCCD */
-Lp3 = 2.8571429849e-01, /* 3E924925 */
-Lp4 = 2.2222198546e-01, /* 3E638E29 */
-Lp5 = 1.8183572590e-01, /* 3E3A3325 */
-Lp6 = 1.5313838422e-01, /* 3E1CD04F */
-Lp7 = 1.4798198640e-01; /* 3E178897 */
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
 
-static const float zero = 0.0;
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
 
-float
-__log1pf(float x)
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.
+*/
+
+#include <math.h>
+#include <stdint.h>
+#include <errno.h>
+#include <libm-alias-float.h>
+#include "math_config.h"
+
+static __attribute__ ((noinline)) float
+as_special (float x)
 {
-	float hfsq,f,c,s,z,R,u;
-	int32_t k,hx,hu,ax;
+  uint32_t t = asuint (x);
+  if (t == 0xbf800000u)
+    return __math_divzerof (1);
+  if (t == 0x7f800000u)
+    return x; /* +inf */
+  uint32_t ax = t << 1;
+  if (ax > 0xff000000u)
+    return x + x; /* nan */
+  return __math_invalidf (0.0f);
+}
 
-	GET_FLOAT_WORD(hx,x);
-	ax = hx&0x7fffffff;
+float
+__log1pf (float x)
+{
+  static const double x0[] =
+    {
+      0x1.f81f82p-1,  0x1.e9131acp-1, 0x1.dae6077p-1, 0x1.cd85689p-1,
+      0x1.c0e0704p-1, 0x1.b4e81b5p-1, 0x1.a98ef6p-1,  0x1.9ec8e95p-1,
+      0x1.948b0fdp-1, 0x1.8acb90fp-1, 0x1.8181818p-1, 0x1.78a4c81p-1,
+      0x1.702e05cp-1, 0x1.6816817p-1, 0x1.605816p-1,  0x1.58ed231p-1,
+      0x1.51d07ebp-1, 0x1.4afd6ap-1,  0x1.446f865p-1, 0x1.3e22cbdp-1,
+      0x1.3813814p-1, 0x1.323e34ap-1, 0x1.2c9fb4ep-1, 0x1.27350b9p-1,
+      0x1.21fb781p-1, 0x1.1cf06aep-1, 0x1.1811812p-1, 0x1.135c811p-1,
+      0x1.0ecf56cp-1, 0x1.0a6810ap-1, 0x1.0624dd3p-1, 0x1.0204081p-1
+    };
+  static const double lixb[] =
+    {
+      0x1.fc0a8909b4218p-7, 0x1.77458f51aac89p-5, 0x1.341d793afb997p-4,
+      0x1.a926d3a5ebd2ap-4, 0x1.0d77e7a8a823dp-3, 0x1.44d2b6c557102p-3,
+      0x1.7ab89040accecp-3, 0x1.af3c94ecab3d6p-3, 0x1.e27076d54e6c9p-3,
+      0x1.0a324e3888ad5p-2, 0x1.22941fc0c7357p-2, 0x1.3a64c56ae3fdbp-2,
+      0x1.51aad874af21fp-2, 0x1.686c81d300eap-2,  0x1.7eaf83c7fa9b5p-2,
+      0x1.947941aa610ecp-2, 0x1.a9cec9a3f023bp-2, 0x1.beb4d9ea4156ep-2,
+      0x1.d32fe7f35e5c7p-2, 0x1.e7442617b817ap-2, 0x1.faf588dd5ed1p-2,
+      0x1.0723e5c635c39p-1, 0x1.109f39d53c99p-1,  0x1.19ee6b38a4668p-1,
+      0x1.23130d7f93c3bp-1, 0x1.2c0e9ec9b0b85p-1, 0x1.34e289cb35eccp-1,
+      0x1.3d9026ad3d3f3p-1, 0x1.4618bc1eadbbbp-1, 0x1.4e7d8127dd8a9p-1,
+      0x1.56bf9d5967092p-1, 0x1.5ee02a926936ep-1
+    };
+  static const double lix[] =
+    {
+      0x1.fc0a890fc03e4p-7, 0x1.77458f532dcfcp-5, 0x1.341d793bbd1d1p-4,
+      0x1.a926d3a6ad563p-4, 0x1.0d77e7a908e59p-3, 0x1.44d2b6c5b7d1ep-3,
+      0x1.7ab890410d909p-3, 0x1.af3c94ed0bff3p-3, 0x1.e27076d5af2e6p-3,
+      0x1.0a324e38b90e3p-2, 0x1.22941fc0f7966p-2, 0x1.3a64c56b145eap-2,
+      0x1.51aad874df82dp-2, 0x1.686c81d3314afp-2, 0x1.7eaf83c82afc3p-2,
+      0x1.947941aa916fbp-2, 0x1.a9cec9a42084ap-2, 0x1.beb4d9ea71b7cp-2,
+      0x1.d32fe7f38ebd5p-2, 0x1.e7442617e8788p-2, 0x1.faf588dd8f31fp-2,
+      0x1.0723e5c64df4p-1,  0x1.109f39d554c97p-1, 0x1.19ee6b38bc96fp-1,
+      0x1.23130d7fabf43p-1, 0x1.2c0e9ec9c8e8cp-1, 0x1.34e289cb4e1d3p-1,
+      0x1.3d9026ad556fbp-1, 0x1.4618bc1ec5ec2p-1, 0x1.4e7d8127f5bb1p-1,
+      0x1.56bf9d597f399p-1, 0x1.5ee02a9281675p-1
+    };
+  static const double b[] =
+    {
+      0x1p+0,
+      -0x1p-1,
+      0x1.5555555556f6bp-2,
+      -0x1.00000000029b9p-2,
+      0x1.9999988d176e4p-3,
+      -0x1.55555418889a7p-3,
+      0x1.24adeca50e2bcp-3,
+      -0x1.001ba33bf57cfp-3
+    };
 
-	k = 1;
-	if (hx < 0x3ed413d7) {			/* x < 0.41422  */
-	    if(ax>=0x3f800000) {		/* x <= -1.0 */
-		if(x==(float)-1.0) return -two25/zero; /* log1p(-1)=-inf */
-		else return (x-x)/(x-x);	/* log1p(x<-1)=NaN */
-	    }
-	    if(ax<0x31000000) {			/* |x| < 2**-29 */
-		math_force_eval(two25+x);	/* raise inexact */
-		if (ax<0x24800000)		/* |x| < 2**-54 */
-		  {
-		    math_check_force_underflow (x);
-		    return x;
-		  }
-		else
-		    return x - x*x*(float)0.5;
-	    }
-	    if(hx>0||hx<=((int32_t)0xbe95f61f)) {
-		k=0;f=x;hu=1;}	/* -0.2929<x<0.41422 */
+  double z = x;
+  uint32_t ux = asuint (x);
+  uint32_t ax = ux & (~0u >> 1);
+  if (__glibc_likely (ax < 0x3c880000))
+    {
+      if (__glibc_unlikely (ax < 0x33000000))
+	{
+	  if (!ax)
+	    return x;
+	  return fmaf (x, -x, x);
 	}
-	if (hx >= 0x7f800000) return x+x;
-	if(k!=0) {
-	    if(hx<0x5a000000) {
-		u  = (float)1.0+x;
-		GET_FLOAT_WORD(hu,u);
-		k  = (hu>>23)-127;
-		/* correction term */
-		c  = (k>0)? (float)1.0-(u-x):x-(u-(float)1.0);
-		c /= u;
-	    } else {
-		u  = x;
-		GET_FLOAT_WORD(hu,u);
-		k  = (hu>>23)-127;
-		c  = 0;
-	    }
-	    hu &= 0x007fffff;
-	    if(hu<0x3504f7) {
-		SET_FLOAT_WORD(u,hu|0x3f800000);/* normalize u */
-	    } else {
-		k += 1;
-		SET_FLOAT_WORD(u,hu|0x3f000000);	/* normalize u/2 */
-		hu = (0x00800000-hu)>>2;
+      double z2 = z * z, z4 = z2 * z2;
+      double f = z2
+		 * ((b[1] + z * b[2]) + z2 * (b[3] + z * b[4])
+		    + z4 * ((b[5] + z * b[6]) + z2 * b[7]));
+      double r = z + f;
+      if (__glibc_unlikely ((asuint64 (r) & 0xfffffffll) == 0))
+	r += 0x1p14 * (f + (z - r));
+      return r;
+    }
+  else
+    {
+      if (__glibc_unlikely (ux >= 0xbf800000u || ax >= 0x7f800000))
+	return as_special (x);
+      uint64_t tp = asuint64 (z + 1);
+      int e = tp >> 52;
+      uint64_t m52 = tp & (~(uint64_t) 0 >> 12);
+      unsigned int j = (tp >> (52 - 5)) & 31;
+      e -= 0x3ff;
+      double xd = asdouble (m52 | ((uint64_t) 0x3ff << 52));
+      z = xd * x0[j] - 1;
+      static const double c[] =
+	{
+	  -0x1.3902c33434e7fp-43, 0x1.ffffffe1cbed5p-1, -0x1.ffffff7d1b014p-2,
+	  0x1.5564e0ed3613ap-2, -0x1.0012232a00d4ap-2
+	};
+      const double ln2 = 0x1.62e42fefa39efp-1;
+      double z2 = z * z,
+	     r = (ln2 * e + lixb[j])
+		 + z * ((c[1] + z * c[2]) + z2 * (c[3] + z * c[4]));
+      float ub = r;
+      float lb = r + 2.2e-11;
+      if (__glibc_unlikely (ub != lb))
+	{
+	  double z4 = z2 * z2,
+		 f = z
+		     * ((b[0] + z * b[1]) + z2 * (b[2] + z * b[3])
+			+ z4 * ((b[4] + z * b[5]) + z2 * (b[6] + z * b[7])));
+	  const double ln2l = 0x1.7f7d1cf79abcap-20, ln2h = 0x1.62e4p-1;
+	  double Lh = ln2h * e;
+	  double Ll = ln2l * e;
+	  double rl = f + Ll + lix[j];
+	  double tr = rl + Lh;
+	  if (__glibc_unlikely ((asuint64 (tr) & 0xfffffffll) == 0))
+	    {
+	      if (x == -0x1.247ab0p-6)
+		return -0x1.271f0ep-6f - 0x1p-31f;
+	      if (x == -0x1.3a415ep-5)
+		return -0x1.407112p-5f + 0x1p-30f;
+	      if (x == 0x1.fb035ap-2)
+		return 0x1.9bddc2p-2f + 0x1p-27f;
+	      tr += 64 * (rl + (Lh - tr));
 	    }
-	    f = u-(float)1.0;
-	}
-	hfsq=(float)0.5*f*f;
-	if(hu==0) {	/* |f| < 2**-20 */
-	    if(f==zero) {
-		if(k==0) return zero;
-		else {c += k*ln2_lo; return k*ln2_hi+c;}
+	  else if (rl + (Lh - tr) == 0.0)
+	    {
+	      if (x == 0x1.b7fd86p-4)
+		return 0x1.a1ece2p-4f + 0x1p-29f;
+	      if (x == -0x1.3a415ep-5)
+		return -0x1.407112p-5f + 0x1p-30f;
+	      if (x == 0x1.43c7e2p-6)
+		return 0x1.409f80p-6f + 0x1p-31f;
 	    }
-	    R = hfsq*(1.0f-0.66666666666666666f*f);
-	    if(k==0) return f-R; else
-		     return k*ln2_hi-((R-(k*ln2_lo+c))-f);
+	  ub = tr;
 	}
-	s = f/((float)2.0+f);
-	z = s*s;
-	R = z*(Lp1+z*(Lp2+z*(Lp3+z*(Lp4+z*(Lp5+z*(Lp6+z*Lp7))))));
-	if (k == 0)
-	  return f - (hfsq - s * (hfsq + R));
-	else
-	  {
-	    /* With GCC 7 when compiling with -Os the compiler warns
-	       that c might be used uninitialized.  This can't be true
-	       because k must be 0 for c to be uninitialized and we
-	       handled that computation earlier without using c.  */
-	    DIAG_PUSH_NEEDS_COMMENT;
-	    DIAG_IGNORE_Os_NEEDS_COMMENT (7, "-Wmaybe-uninitialized");
-	    return k * ln2_hi - ((hfsq - (s * (hfsq + R)
-					  + (k * ln2_lo + c))) - f);
-	    DIAG_POP_NEEDS_COMMENT;
-	  }
+      return ub;
+    }
 }
+libm_alias_float (__log1p, log1p)
+strong_alias (__log1pf, __logp1f)
+libm_alias_float (__logp1, logp1)
diff --git a/sysdeps/ieee754/flt-32/w_log1pf.c b/sysdeps/ieee754/flt-32/w_log1pf.c
new file mode 100644
index 0000000..1cc8931
--- /dev/null
+++ b/sysdeps/ieee754/flt-32/w_log1pf.c
@@ -0,0 +1 @@
+/* Not needed.  */
diff --git a/sysdeps/loongarch/lp64/libm-test-ulps b/sysdeps/loongarch/lp64/libm-test-ulps
index b59d682..846ddb4 100644
--- a/sysdeps/loongarch/lp64/libm-test-ulps
+++ b/sysdeps/loongarch/lp64/libm-test-ulps
@@ -1215,22 +1215,18 @@ ldouble: 4
 
 Function: "log1p":
 double: 1
-float: 1
 ldouble: 3
 
 Function: "log1p_downward":
 double: 2
-float: 2
 ldouble: 3
 
 Function: "log1p_towardzero":
 double: 2
-float: 2
 ldouble: 3
 
 Function: "log1p_upward":
 double: 2
-float: 2
 ldouble: 2
 
 Function: "log2":
@@ -1278,22 +1274,18 @@ ldouble: 2
 
 Function: "logp1":
 double: 1
-float: 1
 ldouble: 3
 
 Function: "logp1_downward":
 double: 2
-float: 2
 ldouble: 3
 
 Function: "logp1_towardzero":
 double: 2
-float: 2
 ldouble: 3
 
 Function: "logp1_upward":
 double: 2
-float: 2
 ldouble: 2
 
 Function: "pow":
diff --git a/sysdeps/m68k/coldfire/fpu/libm-test-ulps b/sysdeps/m68k/coldfire/fpu/libm-test-ulps
index ae05498..8130d49 100644
--- a/sysdeps/m68k/coldfire/fpu/libm-test-ulps
+++ b/sysdeps/m68k/coldfire/fpu/libm-test-ulps
@@ -131,12 +131,6 @@ Function: "log10":
 double: 1
 float: 2
 
-Function: "log1p":
-float: 1
-
-Function: "logp1":
-float: 1
-
 Function: "sincos":
 double: 1
 float: 1
diff --git a/sysdeps/m68k/m680x0/fpu/w_log1pf.c b/sysdeps/m68k/m680x0/fpu/w_log1pf.c
new file mode 100644
index 0000000..5f363cc
--- /dev/null
+++ b/sysdeps/m68k/m680x0/fpu/w_log1pf.c
@@ -0,0 +1,20 @@
+/* Wrapper for __log1p that handles setting errno.
+   Copyright (C) 2015-2024 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 <math-type-macros-float.h>
+#include <w_log1p_template.c>
diff --git a/sysdeps/microblaze/libm-test-ulps b/sysdeps/microblaze/libm-test-ulps
index a84494b..328e315 100644
--- a/sysdeps/microblaze/libm-test-ulps
+++ b/sysdeps/microblaze/libm-test-ulps
@@ -224,7 +224,6 @@ float: 2
 
 Function: "log1p":
 double: 1
-float: 1
 
 Function: "log2":
 double: 2
@@ -232,7 +231,6 @@ float: 1
 
 Function: "logp1":
 double: 1
-float: 1
 
 Function: "pow":
 float: 1
diff --git a/sysdeps/mips/mips32/libm-test-ulps b/sysdeps/mips/mips32/libm-test-ulps
index f1d0543..2ba46a0 100644
--- a/sysdeps/mips/mips32/libm-test-ulps
+++ b/sysdeps/mips/mips32/libm-test-ulps
@@ -971,19 +971,15 @@ float: 3
 
 Function: "log1p":
 double: 1
-float: 1
 
 Function: "log1p_downward":
 double: 2
-float: 2
 
 Function: "log1p_towardzero":
 double: 2
-float: 2
 
 Function: "log1p_upward":
 double: 2
-float: 2
 
 Function: "log2":
 double: 2
@@ -1029,19 +1025,15 @@ float: 2
 
 Function: "logp1":
 double: 1
-float: 1
 
 Function: "logp1_downward":
 double: 2
-float: 2
 
 Function: "logp1_towardzero":
 double: 2
-float: 2
 
 Function: "logp1_upward":
 double: 2
-float: 2
 
 Function: "pow":
 double: 1
diff --git a/sysdeps/mips/mips64/libm-test-ulps b/sysdeps/mips/mips64/libm-test-ulps
index fecd998..f73f95c 100644
--- a/sysdeps/mips/mips64/libm-test-ulps
+++ b/sysdeps/mips/mips64/libm-test-ulps
@@ -1220,22 +1220,18 @@ ldouble: 4
 
 Function: "log1p":
 double: 1
-float: 1
 ldouble: 3
 
 Function: "log1p_downward":
 double: 2
-float: 2
 ldouble: 3
 
 Function: "log1p_towardzero":
 double: 2
-float: 2
 ldouble: 3
 
 Function: "log1p_upward":
 double: 2
-float: 2
 ldouble: 2
 
 Function: "log2":
@@ -1289,22 +1285,18 @@ ldouble: 1
 
 Function: "logp1":
 double: 1
-float: 1
 ldouble: 3
 
 Function: "logp1_downward":
 double: 2
-float: 2
 ldouble: 3
 
 Function: "logp1_towardzero":
 double: 2
-float: 2
 ldouble: 3
 
 Function: "logp1_upward":
 double: 2
-float: 2
 ldouble: 2
 
 Function: "pow":
diff --git a/sysdeps/nios2/libm-test-ulps b/sysdeps/nios2/libm-test-ulps
index 33f714e..5240767 100644
--- a/sysdeps/nios2/libm-test-ulps
+++ b/sysdeps/nios2/libm-test-ulps
@@ -230,7 +230,6 @@ float: 2
 
 Function: "log1p":
 double: 1
-float: 1
 
 Function: "log2":
 double: 2
@@ -238,7 +237,6 @@ float: 1
 
 Function: "logp1":
 double: 1
-float: 1
 
 Function: "pow":
 double: 1
diff --git a/sysdeps/or1k/nofpu/libm-test-ulps b/sysdeps/or1k/nofpu/libm-test-ulps
index 6862551..c7ae0f0 100644
--- a/sysdeps/or1k/nofpu/libm-test-ulps
+++ b/sysdeps/or1k/nofpu/libm-test-ulps
@@ -923,19 +923,15 @@ float: 2
 
 Function: "log1p":
 double: 1
-float: 1
 
 Function: "log1p_downward":
 double: 2
-float: 2
 
 Function: "log1p_towardzero":
 double: 2
-float: 2
 
 Function: "log1p_upward":
 double: 2
-float: 2
 
 Function: "log2":
 float: 1
@@ -951,19 +947,15 @@ double: 1
 
 Function: "logp1":
 double: 1
-float: 1
 
 Function: "logp1_downward":
 double: 2
-float: 2
 
 Function: "logp1_towardzero":
 double: 2
-float: 2
 
 Function: "logp1_upward":
 double: 2
-float: 2
 
 Function: "pow":
 double: 1
diff --git a/sysdeps/powerpc/fpu/libm-test-ulps b/sysdeps/powerpc/fpu/libm-test-ulps
index f775c56..48a3ea1 100644
--- a/sysdeps/powerpc/fpu/libm-test-ulps
+++ b/sysdeps/powerpc/fpu/libm-test-ulps
@@ -1521,25 +1521,21 @@ ldouble: 6
 
 Function: "log1p":
 double: 1
-float: 1
 float128: 3
 ldouble: 2
 
 Function: "log1p_downward":
 double: 1
-float: 2
 float128: 3
 ldouble: 2
 
 Function: "log1p_towardzero":
 double: 2
-float: 2
 float128: 3
 ldouble: 3
 
 Function: "log1p_upward":
 double: 2
-float: 2
 float128: 2
 ldouble: 3
 
@@ -1605,25 +1601,21 @@ ldouble: 1
 
 Function: "logp1":
 double: 1
-float: 1
 float128: 3
 ldouble: 2
 
 Function: "logp1_downward":
 double: 1
-float: 2
 float128: 3
 ldouble: 2
 
 Function: "logp1_towardzero":
 double: 2
-float: 2
 float128: 3
 ldouble: 3
 
 Function: "logp1_upward":
 double: 2
-float: 2
 float128: 2
 ldouble: 3
 
diff --git a/sysdeps/powerpc/nofpu/libm-test-ulps b/sysdeps/powerpc/nofpu/libm-test-ulps
index a732166..ec91048 100644
--- a/sysdeps/powerpc/nofpu/libm-test-ulps
+++ b/sysdeps/powerpc/nofpu/libm-test-ulps
@@ -1277,22 +1277,18 @@ ldouble: 6
 
 Function: "log1p":
 double: 1
-float: 1
 ldouble: 2
 
 Function: "log1p_downward":
 double: 2
-float: 2
 ldouble: 2
 
 Function: "log1p_towardzero":
 double: 2
-float: 2
 ldouble: 2
 
 Function: "log1p_upward":
 double: 2
-float: 2
 ldouble: 2
 
 Function: "log2":
@@ -1346,22 +1342,18 @@ ldouble: 1
 
 Function: "logp1":
 double: 1
-float: 1
 ldouble: 2
 
 Function: "logp1_downward":
 double: 2
-float: 2
 ldouble: 2
 
 Function: "logp1_towardzero":
 double: 2
-float: 2
 ldouble: 2
 
 Function: "logp1_upward":
 double: 2
-float: 2
 ldouble: 2
 
 Function: "mul_downward_ldouble":
diff --git a/sysdeps/riscv/nofpu/libm-test-ulps b/sysdeps/riscv/nofpu/libm-test-ulps
index 739d04b..0189fc5 100644
--- a/sysdeps/riscv/nofpu/libm-test-ulps
+++ b/sysdeps/riscv/nofpu/libm-test-ulps
@@ -1172,22 +1172,18 @@ ldouble: 3
 
 Function: "log1p":
 double: 1
-float: 1
 ldouble: 3
 
 Function: "log1p_downward":
 double: 1
-float: 2
 ldouble: 3
 
 Function: "log1p_towardzero":
 double: 2
-float: 2
 ldouble: 3
 
 Function: "log1p_upward":
 double: 2
-float: 2
 ldouble: 2
 
 Function: "log2":
@@ -1223,22 +1219,18 @@ ldouble: 1
 
 Function: "logp1":
 double: 1
-float: 1
 ldouble: 3
 
 Function: "logp1_downward":
 double: 1
-float: 2
 ldouble: 3
 
 Function: "logp1_towardzero":
 double: 2
-float: 2
 ldouble: 3
 
 Function: "logp1_upward":
 double: 2
-float: 2
 ldouble: 2
 
 Function: "pow":
diff --git a/sysdeps/riscv/rvd/libm-test-ulps b/sysdeps/riscv/rvd/libm-test-ulps
index f6bb6e0..a0b4376 100644
--- a/sysdeps/riscv/rvd/libm-test-ulps
+++ b/sysdeps/riscv/rvd/libm-test-ulps
@@ -1215,22 +1215,18 @@ ldouble: 4
 
 Function: "log1p":
 double: 1
-float: 1
 ldouble: 3
 
 Function: "log1p_downward":
 double: 1
-float: 2
 ldouble: 3
 
 Function: "log1p_towardzero":
 double: 2
-float: 2
 ldouble: 3
 
 Function: "log1p_upward":
 double: 2
-float: 2
 ldouble: 2
 
 Function: "log2":
@@ -1278,22 +1274,18 @@ ldouble: 2
 
 Function: "logp1":
 double: 1
-float: 1
 ldouble: 3
 
 Function: "logp1_downward":
 double: 1
-float: 2
 ldouble: 3
 
 Function: "logp1_towardzero":
 double: 2
-float: 2
 ldouble: 3
 
 Function: "logp1_upward":
 double: 2
-float: 2
 ldouble: 2
 
 Function: "pow":
diff --git a/sysdeps/s390/fpu/libm-test-ulps b/sysdeps/s390/fpu/libm-test-ulps
index b444d44..658752c 100644
--- a/sysdeps/s390/fpu/libm-test-ulps
+++ b/sysdeps/s390/fpu/libm-test-ulps
@@ -1216,22 +1216,18 @@ ldouble: 4
 
 Function: "log1p":
 double: 1
-float: 1
 ldouble: 3
 
 Function: "log1p_downward":
 double: 1
-float: 2
 ldouble: 3
 
 Function: "log1p_towardzero":
 double: 2
-float: 2
 ldouble: 3
 
 Function: "log1p_upward":
 double: 2
-float: 2
 ldouble: 2
 
 Function: "log2":
@@ -1276,22 +1272,18 @@ ldouble: 1
 
 Function: "logp1":
 double: 1
-float: 1
 ldouble: 3
 
 Function: "logp1_downward":
 double: 1
-float: 2
 ldouble: 3
 
 Function: "logp1_towardzero":
 double: 2
-float: 2
 ldouble: 3
 
 Function: "logp1_upward":
 double: 2
-float: 2
 ldouble: 2
 
 Function: "pow":
diff --git a/sysdeps/sh/libm-test-ulps b/sysdeps/sh/libm-test-ulps
index fc4556e..8562796 100644
--- a/sysdeps/sh/libm-test-ulps
+++ b/sysdeps/sh/libm-test-ulps
@@ -460,11 +460,9 @@ float: 1
 
 Function: "log1p":
 double: 1
-float: 1
 
 Function: "log1p_towardzero":
 double: 2
-float: 2
 
 Function: "log2":
 double: 2
@@ -475,11 +473,9 @@ double: 2
 
 Function: "logp1":
 double: 1
-float: 1
 
 Function: "logp1_towardzero":
 double: 2
-float: 2
 
 Function: "pow":
 double: 1
diff --git a/sysdeps/sparc/fpu/libm-test-ulps b/sysdeps/sparc/fpu/libm-test-ulps
index c7f8531..4c4d8ff 100644
--- a/sysdeps/sparc/fpu/libm-test-ulps
+++ b/sysdeps/sparc/fpu/libm-test-ulps
@@ -1220,22 +1220,18 @@ ldouble: 4
 
 Function: "log1p":
 double: 1
-float: 1
 ldouble: 3
 
 Function: "log1p_downward":
 double: 2
-float: 2
 ldouble: 3
 
 Function: "log1p_towardzero":
 double: 2
-float: 2
 ldouble: 3
 
 Function: "log1p_upward":
 double: 2
-float: 2
 ldouble: 2
 
 Function: "log2":
@@ -1289,22 +1285,18 @@ ldouble: 1
 
 Function: "logp1":
 double: 1
-float: 1
 ldouble: 3
 
 Function: "logp1_downward":
 double: 2
-float: 2
 ldouble: 3
 
 Function: "logp1_towardzero":
 double: 2
-float: 2
 ldouble: 3
 
 Function: "logp1_upward":
 double: 2
-float: 2
 ldouble: 2
 
 Function: "pow":
diff --git a/sysdeps/x86_64/fpu/libm-test-ulps b/sysdeps/x86_64/fpu/libm-test-ulps
index de19bf2..ef9ef51 100644
--- a/sysdeps/x86_64/fpu/libm-test-ulps
+++ b/sysdeps/x86_64/fpu/libm-test-ulps
@@ -1822,25 +1822,21 @@ ldouble: 6
 
 Function: "log1p":
 double: 1
-float: 1
 float128: 3
 ldouble: 2
 
 Function: "log1p_downward":
 double: 2
-float: 2
 float128: 3
 ldouble: 4
 
 Function: "log1p_towardzero":
 double: 2
-float: 2
 float128: 3
 ldouble: 4
 
 Function: "log1p_upward":
 double: 2
-float: 2
 float128: 2
 ldouble: 3
 
@@ -1966,25 +1962,21 @@ float: 3
 
 Function: "logp1":
 double: 1
-float: 1
 float128: 3
 ldouble: 2
 
 Function: "logp1_downward":
 double: 2
-float: 2
 float128: 3
 ldouble: 4
 
 Function: "logp1_towardzero":
 double: 2
-float: 2
 float128: 3
 ldouble: 4
 
 Function: "logp1_upward":
 double: 2
-float: 2
 float128: 2
 ldouble: 3