Patchwork powerpc: Shrink ULPs for float128

login
register
mail settings
Submitter Gabriel F T Gomes
Date Aug. 11, 2017, 2:01 p.m.
Message ID <20170811140124.30153-1-gftg@linux.vnet.ibm.com>
Download mbox | patch
Permalink /patch/22080/
State Committed
Delegated to: Tulio Magno Quites Machado Filho
Headers show

Comments

Gabriel F T Gomes - Aug. 11, 2017, 2:01 p.m.
During the development of float128 on powerpc64le, the ulps have been
generated since early versions of the patches.  On those versions, the
functions gamma and pow10 were mistakenly thought to be part of the API.
After review, the functions were removed, however the ulps for them were
carried in the ulps file.

This patch removes such entries from the ulps file, as well as it
shrinks the ulps for the cpow and lgamma functions.

Tested for powerpc64le.

	* sysdeps/powerpc/fpu/libm-test-ulps: Update.
---
 sysdeps/powerpc/fpu/libm-test-ulps | 36 ++++++++++--------------------------
 1 file changed, 10 insertions(+), 26 deletions(-)
Tulio Magno Quites Machado Filho - Aug. 16, 2017, 7:47 p.m.
"Gabriel F. T. Gomes" <gftg@linux.vnet.ibm.com> writes:

> During the development of float128 on powerpc64le, the ulps have been
> generated since early versions of the patches.  On those versions, the
> functions gamma and pow10 were mistakenly thought to be part of the API.
> After review, the functions were removed, however the ulps for them were
> carried in the ulps file.
>
> This patch removes such entries from the ulps file, as well as it
> shrinks the ulps for the cpow and lgamma functions.

LGTM.
Gabriel F T Gomes - Aug. 22, 2017, 6:31 p.m.
On Wed, 16 Aug 2017 16:47:49 -0300
"Tulio Magno Quites Machado Filho" <tuliom@linux.vnet.ibm.com> wrote:

>"Gabriel F. T. Gomes" <gftg@linux.vnet.ibm.com> writes:
>
>> During the development of float128 on powerpc64le, the ulps have been
>> generated since early versions of the patches.  On those versions, the
>> functions gamma and pow10 were mistakenly thought to be part of the API.
>> After review, the functions were removed, however the ulps for them were
>> carried in the ulps file.
>>
>> This patch removes such entries from the ulps file, as well as it
>> shrinks the ulps for the cpow and lgamma functions.  
>
>LGTM.

Thanks, pushed as f2f048a533bd.

Patch

diff --git a/sysdeps/powerpc/fpu/libm-test-ulps b/sysdeps/powerpc/fpu/libm-test-ulps
index 7fb67446a5..d4d03311e3 100644
--- a/sysdeps/powerpc/fpu/libm-test-ulps
+++ b/sysdeps/powerpc/fpu/libm-test-ulps
@@ -1357,9 +1357,9 @@  ldouble: 4
 
 Function: Imaginary part of "cpow":
 float: 2
-float128: 9
+float128: 1
 ifloat: 2
-ifloat128: 9
+ifloat128: 1
 ildouble: 2
 ldouble: 2
 
@@ -2080,40 +2080,32 @@  ldouble: 1
 Function: "gamma":
 double: 3
 float: 4
-float128: 9
 idouble: 3
 ifloat: 4
-ifloat128: 9
 ildouble: 3
 ldouble: 3
 
 Function: "gamma_downward":
 double: 4
 float: 4
-float128: 9
 idouble: 4
 ifloat: 4
-ifloat128: 9
 ildouble: 15
 ldouble: 15
 
 Function: "gamma_towardzero":
 double: 4
 float: 3
-float128: 9
 idouble: 4
 ifloat: 3
-ifloat128: 9
 ildouble: 16
 ldouble: 16
 
 Function: "gamma_upward":
 double: 4
 float: 5
-float128: 9
 idouble: 4
 ifloat: 5
-ifloat128: 9
 ildouble: 11
 ldouble: 11
 
@@ -2272,40 +2264,40 @@  ldouble: 5
 Function: "lgamma":
 double: 3
 float: 4
-float128: 9
+float128: 5
 idouble: 3
 ifloat: 4
-ifloat128: 9
+ifloat128: 5
 ildouble: 3
 ldouble: 3
 
 Function: "lgamma_downward":
 double: 4
 float: 4
-float128: 9
+float128: 8
 idouble: 4
 ifloat: 4
-ifloat128: 9
+ifloat128: 8
 ildouble: 15
 ldouble: 15
 
 Function: "lgamma_towardzero":
 double: 4
 float: 3
-float128: 9
+float128: 5
 idouble: 4
 ifloat: 3
-ifloat128: 9
+ifloat128: 5
 ildouble: 16
 ldouble: 16
 
 Function: "lgamma_upward":
 double: 4
 float: 5
-float128: 9
+float128: 8
 idouble: 4
 ifloat: 5
-ifloat128: 9
+ifloat128: 8
 ildouble: 11
 ldouble: 11
 
@@ -2481,39 +2473,31 @@  ldouble: 1
 
 Function: "pow10":
 double: 2
-float128: 2
 idouble: 2
-ifloat128: 2
 ildouble: 1
 ldouble: 1
 
 Function: "pow10_downward":
 double: 2
 float: 1
-float128: 3
 idouble: 2
 ifloat: 1
-ifloat128: 3
 ildouble: 9
 ldouble: 9
 
 Function: "pow10_towardzero":
 double: 2
 float: 1
-float128: 3
 idouble: 2
 ifloat: 1
-ifloat128: 3
 ildouble: 9
 ldouble: 9
 
 Function: "pow10_upward":
 double: 2
 float: 1
-float128: 3
 idouble: 2
 ifloat: 1
-ifloat128: 3
 ildouble: 4
 ldouble: 4