Fix log10 (1) in round-downward mode (bug 16977).

As with various other issues of this kind, bug 16977 is log10 (1)
wrongly returning -0 rather than +0 in round-downward mode because of
an implementation effectively in terms of log1p (x - 1).  This patch
fixes the issue in the same way used for log.

Tested x86_64 and x86 and ulps updated accordingly.  Also tested for
mips64 to confirm a fix was needed for ldbl-128 and to validate that
fix (also applied to ldbl-128ibm since that version of logl is
essentially the same as the ldbl-128 one).

	[BZ #16977]
	* sysdeps/i386/fpu/e_log10.S (__ieee754_log10): Take absolute
	value when x - 1 is zero.
	* sysdeps/i386/fpu/e_log10f.S (__ieee754_log10f): Likewise.
	* sysdeps/i386/fpu/e_log10l.S (__ieee754_log10l): Likewise.
	* sysdeps/ieee754/ldbl-128/e_log10l.c (__ieee754_log10l): Return
	0.0L for an argument of 1.0L.
	* sysdeps/ieee754/ldbl-128ibm/e_log10l.c (__ieee754_log10l):
	Likewise.
	* sysdeps/x86_64/fpu/e_log10l.S (__ieee754_log10l): Take absolute
	value when x - 1 is zero.
	* math/libm-test.inc (log10_test): Use ALL_RM_TEST.
	* sysdeps/i386/fpu/libm-test-ulps: Update.
	* sysdeps/x86_64/fpu/libm-test-ulps: Likewise.

8 files changed, 82 insertions, 4 deletions

diff --git a/sysdeps/i386/fpu/e_log10.S b/sysdeps/i386/fpu/e_log10.S
index ce6a81abb6..17277084ca 100644
--- a/sysdeps/i386/fpu/e_log10.S
+++ b/sysdeps/i386/fpu/e_log10.S
@@ -46,7 +46,13 @@ ENTRY(__ieee754_log10)
 	fnstsw			// x-1 : x : log10(2)
 	andb	$0x45, %ah
 	jz	2f
-	fstp	%st(1)		// x-1 : log10(2)
+	fxam
+	fnstsw
+	andb	$0x45, %ah
+	cmpb	$0x40, %ah
+	jne	5f
+	fabs			// log10(1) is +0 in all rounding modes.
+5:	fstp	%st(1)		// x-1 : log10(2)
 	fyl2xp1			// log10(x)
 	ret
 
diff --git a/sysdeps/i386/fpu/e_log10f.S b/sysdeps/i386/fpu/e_log10f.S
index 8c20723555..72a3b88251 100644
--- a/sysdeps/i386/fpu/e_log10f.S
+++ b/sysdeps/i386/fpu/e_log10f.S
@@ -47,7 +47,13 @@ ENTRY(__ieee754_log10f)
 	fnstsw			// x-1 : x : log10(2)
 	andb	$0x45, %ah
 	jz	2f
-	fstp	%st(1)		// x-1 : log10(2)
+	fxam
+	fnstsw
+	andb	$0x45, %ah
+	cmpb	$0x40, %ah
+	jne	5f
+	fabs			// log10(1) is +0 in all rounding modes.
+5:	fstp	%st(1)		// x-1 : log10(2)
 	fyl2xp1			// log10(x)
 	ret
 
diff --git a/sysdeps/i386/fpu/e_log10l.S b/sysdeps/i386/fpu/e_log10l.S
index cde987b137..45b9c6d21d 100644
--- a/sysdeps/i386/fpu/e_log10l.S
+++ b/sysdeps/i386/fpu/e_log10l.S
@@ -48,7 +48,13 @@ ENTRY(__ieee754_log10l)
 	fnstsw			// x-1 : x : log10(2)
 	andb	$0x45, %ah
 	jz	2f
-	fstp	%st(1)		// x-1 : log10(2)
+	fxam
+	fnstsw
+	andb	$0x45, %ah
+	cmpb	$0x40, %ah
+	jne	5f
+	fabs			// log10(1) is +0 in all rounding modes.
+5:	fstp	%st(1)		// x-1 : log10(2)
 	fyl2xp1			// log10(x)
 	ret
 
diff --git a/sysdeps/i386/fpu/libm-test-ulps b/sysdeps/i386/fpu/libm-test-ulps
index 946cad489b..1e89284455 100644
--- a/sysdeps/i386/fpu/libm-test-ulps
+++ b/sysdeps/i386/fpu/libm-test-ulps
@@ -1536,6 +1536,30 @@ Function: "log10":
 ildouble: 1
 ldouble: 1
 
+Function: "log10_downward":
+double: 1
+float: 1
+idouble: 1
+ifloat: 1
+ildouble: 1
+ldouble: 1
+
+Function: "log10_towardzero":
+double: 1
+float: 1
+idouble: 1
+ifloat: 1
+ildouble: 1
+ldouble: 1
+
+Function: "log10_upward":
+double: 1
+float: 1
+idouble: 1
+ifloat: 1
+ildouble: 1
+ldouble: 1
+
 Function: "log1p":
 ildouble: 1
 ldouble: 1
diff --git a/sysdeps/ieee754/ldbl-128/e_log10l.c b/sysdeps/ieee754/ldbl-128/e_log10l.c
index b403f815fb..618255f2fa 100644
--- a/sysdeps/ieee754/ldbl-128/e_log10l.c
+++ b/sysdeps/ieee754/ldbl-128/e_log10l.c
@@ -193,6 +193,9 @@ __ieee754_log10l (long double x)
   if (hx >= 0x7fff000000000000LL)
     return (x + x);
 
+  if (x == 1.0L)
+    return 0.0L;
+
 /* separate mantissa from exponent */
 
 /* Note, frexp is used so that denormal numbers
diff --git a/sysdeps/ieee754/ldbl-128ibm/e_log10l.c b/sysdeps/ieee754/ldbl-128ibm/e_log10l.c
index 1a6a4a0fa3..7477791b77 100644
--- a/sysdeps/ieee754/ldbl-128ibm/e_log10l.c
+++ b/sysdeps/ieee754/ldbl-128ibm/e_log10l.c
@@ -195,6 +195,9 @@ __ieee754_log10l (long double x)
   if (hx >= 0x7ff0000000000000LL)
     return (x + x);
 
+  if (x == 1.0L)
+    return 0.0L;
+
 /* separate mantissa from exponent */
 
 /* Note, frexp is used so that denormal numbers
diff --git a/sysdeps/x86_64/fpu/e_log10l.S b/sysdeps/x86_64/fpu/e_log10l.S
index 6c07024c19..2607ad199b 100644
--- a/sysdeps/x86_64/fpu/e_log10l.S
+++ b/sysdeps/x86_64/fpu/e_log10l.S
@@ -46,7 +46,13 @@ ENTRY(__ieee754_log10l)
 	fnstsw			// x-1 : x : log10(2)
 	andb	$0x45, %ah
 	jz	2f
-	fstp	%st(1)		// x-1 : log10(2)
+	fxam
+	fnstsw
+	andb	$0x45, %ah
+	cmpb	$0x40, %ah
+	jne	5f
+	fabs			// log10(1) is +0 in all rounding modes.
+5:	fstp	%st(1)		// x-1 : log10(2)
 	fyl2xp1			// log10(x)
 	ret
 
diff --git a/sysdeps/x86_64/fpu/libm-test-ulps b/sysdeps/x86_64/fpu/libm-test-ulps
index d47287696c..bb549d2b0d 100644
--- a/sysdeps/x86_64/fpu/libm-test-ulps
+++ b/sysdeps/x86_64/fpu/libm-test-ulps
@@ -1611,6 +1611,30 @@ ifloat: 2
 ildouble: 1
 ldouble: 1
 
+Function: "log10_downward":
+double: 1
+float: 1
+idouble: 1
+ifloat: 1
+ildouble: 1
+ldouble: 1
+
+Function: "log10_towardzero":
+double: 1
+float: 1
+idouble: 1
+ifloat: 1
+ildouble: 1
+ldouble: 1
+
+Function: "log10_upward":
+double: 1
+float: 1
+idouble: 1
+ifloat: 1
+ildouble: 1
+ldouble: 1
+
 Function: "log1p":
 float: 1
 ifloat: 1