This is the mail archive of the
libc-ports@sources.redhat.com
mailing list for the libc-ports project.
Re: [PATCH] Optimize MIPS memcpy (mips glibc test results)
On Fri, 2012-09-21 at 20:48 +0000, Joseph S. Myers wrote:
> On Fri, 21 Sep 2012, Steve Ellcey wrote:
>
> > > > glibc/math/test-float.out
> > > > glibc/math/test-double.out
> > > > glibc/math/test-ifloat.out
> > > > glibc/math/test-idouble.out
> > >
> > > Expected for soft-float configurations.
> >
> > But I thought I was doing a hard float build. It looks like the
> > failures are with cos, sincos, clog10, clog, ctan*. Maybe not having
> > those instructions in hardware counts as soft-float?
>
> Those have had tests added since I updated libm-test-ulps for 2.16, so if
> it's just small ulps values for new tests then it's also fine and will go
> away when the ulps are updated again.
Yes, it is mostly 1 ulp differences with a couple of 2 ulp diffs.
Steve Ellcey
sellcey@mips.com