Re: [PATCH] New generic sinf

[Rajalakshmi Srinivasaraghavan <raji at linux dot vnet dot ibm dot com>]

On 10/27/2017 10:08 PM, Joseph Myers wrote:
> On Fri, 27 Oct 2017, Rajalakshmi Srinivasaraghavan wrote:
>
> >   #if defined(LIBM_SCCS) && !defined(lint)
> >   static char rcsid[] = "$NetBSD: s_sinf.c,v 1.4 1995/05/10 20:48:16 jtc Exp $";
> If you're replacing the whole implementation, it definitely doesn't make
> sense to keep this.

Ack.
> How does this patch affect the performance on the testcase in bug 5997?
> It's not a*good*  benchmark, but if the patch sufficiently improves
> performance it could be considered to fix [BZ #5997].


On s390 for the testcase mentioned in BZ #5997 for sinf:
Time without patch:
real	0m4.931s
user	0m4.924s
sys	0m0.002s

Time with patch:

real	0m0.223s
user	0m0.220s
sys	0m0.002s


--
Thanks
Rajalakshmi S