i686: Update libm-test-ulps

Update i686 libm-test-ulps to fix

FAIL: math/test-float64x-cospi
FAIL: math/test-float64x-sinpi
FAIL: math/test-ldouble-cospi
FAIL: math/test-ldouble-sinpi

when building glibc with GCC 7.4.

Signed-off-by: H.J. Lu <hjl.tools@gmail.com>

diff --git a/sysdeps/i386/i686/fpu/multiarch/libm-test-ulps b/sysdeps/i386/i686/fpu/multiarch/libm-test-ulps
index 2bf55cd04681d2fafa830ecb23f5f29393dd7f80..326dfeb95725a1d597640a397baf34731cba92a7 100644 (file)
--- a/sysdeps/i386/i686/fpu/multiarch/libm-test-ulps
+++ b/sysdeps/i386/i686/fpu/multiarch/libm-test-ulps
@@ -802,13 +802,13 @@ Function: "cospi_downward":
 double: 1
 float: 1
 float128: 2
-ldouble: 1
+ldouble: 2
 
 Function: "cospi_towardzero":
 double: 1
 float: 1
 float128: 2
-ldouble: 1
+ldouble: 2
 
 Function: "cospi_upward":
 double: 1
@@ -1652,13 +1652,13 @@ Function: "sinpi_downward":
 double: 2
 float: 2
 float128: 2
-ldouble: 2
+ldouble: 3
 
 Function: "sinpi_towardzero":
 double: 2
 float: 1
 float128: 1
-ldouble: 1
+ldouble: 2
 
 Function: "sinpi_upward":
 double: 2