r5846 - in glibc-package/branches/eglibc-2.18/debian: . patches patches/i386
Author: adconrad
Date: 2013-12-01 19:49:11 +0000 (Sun, 01 Dec 2013)
New Revision: 5846
Added:
glibc-package/branches/eglibc-2.18/debian/patches/i386/local-i586-ulps.diff
Modified:
glibc-package/branches/eglibc-2.18/debian/changelog
glibc-package/branches/eglibc-2.18/debian/patches/series
Log:
debian/patches/i386/local-i586-ulps.diff: Regen ulps for i586, not i686.
Modified: glibc-package/branches/eglibc-2.18/debian/changelog
===================================================================
--- glibc-package/branches/eglibc-2.18/debian/changelog 2013-12-01 16:58:05 UTC (rev 5845)
+++ glibc-package/branches/eglibc-2.18/debian/changelog 2013-12-01 19:49:11 UTC (rev 5846)
@@ -85,6 +85,7 @@
matches the default that gcc-4.8 has had on i386 for the last two years.
* debian/rules: Set configure_build the same as configure_target above, so
we don't think we're cross-building and skip half the testsuite on i386.
+ * debian/patches/i386/local-i586-ulps.diff: Regen ulps for i586, not i686.
[ Samuel Thibault ]
* patches/hurd-i386/libpthread_hurd_cond_timedwait.diff: Rebased.
Added: glibc-package/branches/eglibc-2.18/debian/patches/i386/local-i586-ulps.diff
===================================================================
--- glibc-package/branches/eglibc-2.18/debian/patches/i386/local-i586-ulps.diff (rev 0)
+++ glibc-package/branches/eglibc-2.18/debian/patches/i386/local-i586-ulps.diff 2013-12-01 19:49:11 UTC (rev 5846)
@@ -0,0 +1,158 @@
+Description: Regen ulps for i586 instead of i686
+ This patch represents slightly looser libm-test-ulps results for
+ i386 representing the fact that i486 and i586 code generation
+ produces slightly sloppier precision than the i686 target that
+ the upstream test results were targetted at.
+Author: Adam Conrad <adconrad@debian.org>
+Reviewed-by: Carlos O'Donell <carlos@redhat.com>
+
+--- a/sysdeps/i386/fpu/libm-test-ulps
++++ b/sysdeps/i386/fpu/libm-test-ulps
+@@ -4132,6 +4132,9 @@ ifloat: 1
+ Test "Imaginary part of: ccos (-0.75 - 89.5 i)":
+ float: 1
+ ifloat: 1
++Test "Real part of: ccos (-2 - 3 i)":
++float: 1
++ifloat: 1
+ Test "Imaginary part of: ccos (-2 - 3 i)":
+ float: 1
+ ifloat: 1
+@@ -4758,6 +4761,9 @@ idouble: 1
+ ifloat: 1
+
+ # cos
++Test "cos (0x1p+50)":
++float: 1
++ifloat: 1
+ Test "cos (M_PI_6l * 2.0)":
+ double: 1
+ idouble: 1
+@@ -4769,6 +4775,8 @@ ifloat: 1
+
+ # cos_downward
+ Test "cos_downward (1)":
++float: 1
++ifloat: 1
+ ildouble: 1
+ ldouble: 1
+ Test "cos_downward (10)":
+@@ -4800,6 +4808,11 @@ Test "cos_downward (9)":
+ float: 1
+ ifloat: 1
+
++# cos_tonearest
++Test "cos_tonearest (2)":
++float: 1
++ifloat: 1
++
+ # cos_towardzero
+ Test "cos_towardzero (1)":
+ ildouble: 1
+@@ -4995,6 +5008,9 @@ ifloat: 1
+ Test "Imaginary part of: csin (-0.75 - 89.5 i)":
+ float: 1
+ ifloat: 1
++Test "Imaginary part of: csin (-2 - 3 i)":
++float: 1
++ifloat: 1
+ Test "Real part of: csin (0.75 + 1.25 i)":
+ float: 1
+ ifloat: 1
+@@ -5226,8 +5242,8 @@ ifloat: 1
+ ildouble: 1
+ ldouble: 1
+ Test "Imaginary part of: ctan_towardzero (0x1.921fb6p+0 + 0x1p-149 i)":
+-float: 1
+-ifloat: 1
++float: 3
++ifloat: 3
+ ildouble: 4
+ ldouble: 4
+
+@@ -5270,6 +5286,8 @@ Test "Imaginary part of: ctanh (0 + 0x3.
+ float: 1
+ ifloat: 1
+ Test "Imaginary part of: ctanh (0 + pi/4 i)":
++float: 1
++ifloat: 1
+ ildouble: 1
+ ldouble: 1
+ Test "Imaginary part of: ctanh (0.75 + 1.25 i)":
+@@ -5341,8 +5359,8 @@ idouble: 1
+ ildouble: 1
+ ldouble: 1
+ Test "Real part of: ctanh_towardzero (0x1p-149 + 0x1.921fb6p+0 i)":
+-float: 1
+-ifloat: 1
++float: 3
++ifloat: 3
+ ildouble: 4
+ ldouble: 4
+ Test "Imaginary part of: ctanh_towardzero (0x1p-149 + 0x1.921fb6p+0 i)":
+@@ -5796,6 +5814,11 @@ ifloat: 1
+ ildouble: 1
+ ldouble: 1
+
++# sin_tonearest
++Test "sin_tonearest (1)":
++float: 1
++ifloat: 1
++
+ # sin_towardzero
+ Test "sin_towardzero (1)":
+ ildouble: 1
+@@ -5864,6 +5887,9 @@ float: 1
+ ifloat: 1
+
+ # sincos
++Test "sincos (0x1p+50) extra output 2":
++float: 1
++ifloat: 1
+ Test "sincos (M_PI_6l*2.0) extra output 1":
+ double: 1
+ float: 1
+@@ -7084,6 +7110,10 @@ ifloat: 1
+ ildouble: 1
+ ldouble: 1
+
++Function: "cos_tonearest":
++float: 1
++ifloat: 1
++
+ Function: "cos_towardzero":
+ float: 1
+ ifloat: 1
+@@ -7218,8 +7248,8 @@ ildouble: 1
+ ldouble: 1
+
+ Function: Imaginary part of "ctan_towardzero":
+-float: 1
+-ifloat: 1
++float: 3
++ifloat: 3
+ ildouble: 4
+ ldouble: 4
+
+@@ -7274,8 +7304,8 @@ float: 1
+ ifloat: 1
+
+ Function: Real part of "ctanh_towardzero":
+-float: 1
+-ifloat: 1
++float: 3
++ifloat: 3
+ ildouble: 4
+ ldouble: 4
+
+@@ -7427,6 +7457,10 @@ ifloat: 1
+ ildouble: 1
+ ldouble: 1
+
++Function: "sin_tonearest":
++float: 1
++ifloat: 1
++
+ Function: "sin_towardzero":
+ float: 1
+ ifloat: 1
Modified: glibc-package/branches/eglibc-2.18/debian/patches/series
===================================================================
--- glibc-package/branches/eglibc-2.18/debian/patches/series 2013-12-01 16:58:05 UTC (rev 5845)
+++ glibc-package/branches/eglibc-2.18/debian/patches/series 2013-12-01 19:49:11 UTC (rev 5846)
@@ -147,6 +147,7 @@
i386/local-biarch.diff
i386/local-cmov.diff
i386/local-cpuid-level2.diff
+i386/local-i586-ulps.diff
i386/submitted-i686-timing.diff
i386/unsubmitted-quiet-ldconfig.diff
Reply to: