PowerPC: Fix nearbyint/nearbyintf result for FE_DOWNWARD

This patch fixes the powerpc32 optimized nearbyint/nearbyintf bogus
results for FE_DOWNWARD rounding mode.  This is due wrong instructions
sequence used in the rounding calculation (two subtractions instead of
adition and a subtraction).

Fixes BZ#16815.

5 files changed, 213 insertions, 11 deletions

diff --git a/ChangeLog b/ChangeLog
index a0bb465..faaea3c 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,3 +1,12 @@
+2014-04-06  Adhemerval Zanella  <azanella@linux.vnet.ibm.com>
+
+	[BZ #16815]
+	* sysdeps/powerpc/powerpc32/fpu/s_nearbyint.S (__nearbyint): Fix
+	result for FE_DOWNWARD rounding mode.
+	* sysdeps/powerpc/powerpc32/fpu/s_nearbyintf.S (__nearbyintf):
+	Likewise.
+	* sysdeps/powerpc/fpu/libm-test-ulps: Update.
+
 2014-04-04  Chris Metcalf  <cmetcalf@tilera.com>
 
 	* sysdeps/tile/dl-runtime.c (_dl_unmap): Fix cut-and-paste bug
diff --git a/NEWS b/NEWS
index c7114da..db8cc10 100644
--- a/NEWS
+++ b/NEWS
@@ -14,7 +14,7 @@ Version 2.20
   16611, 16613, 16623, 16632, 16634, 16639, 16642, 16648, 16649, 16670,
   16674, 16677, 16680, 16683, 16689, 16695, 16701, 16706, 16707, 16712,
   16713, 16714, 16731, 16739, 16743, 16758, 16759, 16760, 16770, 16786,
-  16789, 16799, 16800.
+  16789, 16799, 16800, 16815.
 
 * Running the testsuite no longer terminates as soon as a test fails.
   Instead, a file tests.sum (xtests.sum from "make xcheck") is generated,
diff --git a/sysdeps/powerpc/fpu/libm-test-ulps b/sysdeps/powerpc/fpu/libm-test-ulps
index 27f228c..3319233 100644
--- a/sysdeps/powerpc/fpu/libm-test-ulps
+++ b/sysdeps/powerpc/fpu/libm-test-ulps
@@ -452,6 +452,54 @@ ifloat: 1
 ildouble: 1
 ldouble: 1
 
+Function: Real part of "catan_downward":
+double: 1
+float: 1
+idouble: 1
+ifloat: 1
+ildouble: 5
+ldouble: 5
+
+Function: Imaginary part of "catan_downward":
+double: 2
+float: 2
+idouble: 2
+ifloat: 2
+ildouble: 6
+ldouble: 6
+
+Function: Real part of "catan_towardzero":
+double: 1
+float: 1
+idouble: 1
+ifloat: 1
+ildouble: 6
+ldouble: 6
+
+Function: Imaginary part of "catan_towardzero":
+double: 2
+float: 1
+idouble: 2
+ifloat: 1
+ildouble: 2
+ldouble: 2
+
+Function: Real part of "catan_upward":
+double: 1
+float: 1
+idouble: 1
+ifloat: 1
+ildouble: 3
+ldouble: 3
+
+Function: Imaginary part of "catan_upward":
+double: 3
+float: 3
+idouble: 3
+ifloat: 3
+ildouble: 5
+ldouble: 5
+
 Function: Real part of "catanh":
 double: 4
 float: 1
@@ -466,6 +514,54 @@ ifloat: 1
 ildouble: 1
 ldouble: 1
 
+Function: Real part of "catanh_downward":
+double: 2
+float: 2
+idouble: 2
+ifloat: 2
+ildouble: 3
+ldouble: 3
+
+Function: Imaginary part of "catanh_downward":
+double: 1
+float: 2
+idouble: 1
+ifloat: 2
+ildouble: 5
+ldouble: 5
+
+Function: Real part of "catanh_towardzero":
+double: 2
+float: 1
+idouble: 2
+ifloat: 1
+ildouble: 2
+ldouble: 2
+
+Function: Imaginary part of "catanh_towardzero":
+double: 1
+float: 2
+idouble: 1
+ifloat: 2
+ildouble: 6
+ldouble: 6
+
+Function: Real part of "catanh_upward":
+double: 4
+float: 3
+idouble: 4
+ifloat: 3
+ildouble: 8
+ldouble: 8
+
+Function: Imaginary part of "catanh_upward":
+double: 1
+float: 1
+idouble: 1
+ifloat: 1
+ildouble: 3
+ldouble: 3
+
 Function: "cbrt":
 double: 1
 float: 1
@@ -672,6 +768,102 @@ ifloat: 1
 ildouble: 2
 ldouble: 2
 
+Function: Real part of "clog10_downward":
+double: 3
+float: 3
+idouble: 3
+ifloat: 3
+ildouble: 7
+ldouble: 7
+
+Function: Imaginary part of "clog10_downward":
+double: 3
+float: 2
+idouble: 3
+ifloat: 2
+ildouble: 7
+ldouble: 7
+
+Function: Real part of "clog10_towardzero":
+double: 3
+float: 2
+idouble: 3
+ifloat: 2
+ildouble: 8
+ldouble: 8
+
+Function: Imaginary part of "clog10_towardzero":
+double: 3
+float: 2
+idouble: 3
+ifloat: 2
+ildouble: 7
+ldouble: 7
+
+Function: Real part of "clog10_upward":
+double: 4
+float: 3
+idouble: 4
+ifloat: 3
+ildouble: 8
+ldouble: 8
+
+Function: Imaginary part of "clog10_upward":
+double: 2
+float: 2
+idouble: 2
+ifloat: 2
+ildouble: 6
+ldouble: 6
+
+Function: Real part of "clog_downward":
+double: 2
+float: 2
+idouble: 2
+ifloat: 2
+ildouble: 3
+ldouble: 3
+
+Function: Imaginary part of "clog_downward":
+double: 1
+float: 2
+idouble: 1
+ifloat: 2
+ildouble: 3
+ldouble: 3
+
+Function: Real part of "clog_towardzero":
+double: 2
+float: 2
+idouble: 2
+ifloat: 2
+ildouble: 3
+ldouble: 3
+
+Function: Imaginary part of "clog_towardzero":
+double: 1
+float: 2
+idouble: 1
+ifloat: 2
+ildouble: 4
+ldouble: 4
+
+Function: Real part of "clog_upward":
+double: 2
+float: 1
+idouble: 2
+ifloat: 1
+ildouble: 6
+ldouble: 6
+
+Function: Imaginary part of "clog_upward":
+double: 2
+float: 1
+idouble: 2
+ifloat: 1
+ildouble: 3
+ldouble: 3
+
 Function: "cos":
 double: 2
 float: 1
diff --git a/sysdeps/powerpc/powerpc32/fpu/s_nearbyint.S b/sysdeps/powerpc/powerpc32/fpu/s_nearbyint.S
index 2734738..05ab40e 100644
--- a/sysdeps/powerpc/powerpc32/fpu/s_nearbyint.S
+++ b/sysdeps/powerpc/powerpc32/fpu/s_nearbyint.S
@@ -53,17 +53,17 @@ ENTRY (__nearbyint)
 	fcmpu	cr7,fp1,fp12	/* if (x > 0.0 */
 	ble	cr7,L(lessthanzero)
 	mtfsb0	4*cr7+lt	/* Disable FE_INEXACT exception */
-	fadd	fp0,fp1,fp13	/* x += TWO52 */
-	fsub	fp1,fp0,fp13	/* x -= TWO52 */
+	fadd	fp1,fp1,fp13	/* x += TWO52 */
+	fsub	fp1,fp1,fp13	/* x -= TWO52 */
 	fabs	fp1,fp1		/* if (x == 0.0 */
 	mtfsb0	4*cr1+eq	/* Clear any FE_INEXACT exception */
 	blr
 L(lessthanzero):
 	bgelr	cr7
 	mtfsb0	4*cr7+lt	/* Disable FE_INEXACT exception */
-	fsub	fp0,fp13,fp1	/* x -= TWO52 */
-	fsub	fp0,fp0,fp13	/* x += TWO52 */
-	fneg	fp1,fp0		/* if (x == 0.0) */
+	fsub	fp1,fp1,fp13	/* x -= TWO52 */
+	fadd	fp1,fp1,fp13	/* x += TWO52 */
+	fnabs	fp1,fp1		/* if (x == 0.0) */
 	mtfsb0	4*cr1+eq	/* Clear any FE_INEXACT exception */
 	blr
 END (__nearbyint)
diff --git a/sysdeps/powerpc/powerpc32/fpu/s_nearbyintf.S b/sysdeps/powerpc/powerpc32/fpu/s_nearbyintf.S
index 11bdc77..7449a5f 100644
--- a/sysdeps/powerpc/powerpc32/fpu/s_nearbyintf.S
+++ b/sysdeps/powerpc/powerpc32/fpu/s_nearbyintf.S
@@ -52,16 +52,17 @@ ENTRY (__nearbyintf)
 	fcmpu	cr7,fp1,fp12		/* if (x > 0.0 */
 	ble	cr7,L(lessthanzero)
 	mtfsb0	4*cr7+lt		/* Disable FE_INEXACT exception */
-	fadds	fp0,fp1,fp13		/* x += TWO23 */
-	fsubs	fp1,fp0,fp13		/* x -= TWO23 */
+	fadds	fp1,fp1,fp13		/* x += TWO23 */
+	fsubs	fp1,fp1,fp13		/* x -= TWO23 */
+	fabs	fp1,fp1			/* if (x == 0.0) */
 	mtfsb0	4*cr1+eq		/* Clear any FE_INEXACT exception */
 	blr
 L(lessthanzero):
 	bgelr	cr7
 	mtfsb0	4*cr7+lt		/* Disable FE_INEXACT exception */
-	fsubs	fp0,fp13,fp1		/* x -= TWO23 */
-	fsubs	fp0,fp0,fp13		/* x += TWO23 */
-	fneg	fp1,fp0			/* if (x == 0.0) */
+	fsubs	fp1,fp1,fp13		/* x -= TWO23 */
+	fadds	fp1,fp1,fp13		/* x += TWO23 */
+	fnabs	fp1,fp1			/* if (x == 0.0) */
 	mtfsb0	4*cr1+eq		/* Clear any FE_INEXACT exception */
 	blr
 END (__nearbyintf)