This is the mail archive of the
libc-alpha@sourceware.org
mailing list for the glibc project.
Re: Optimized with SSE2 sinf and cof for x86_32
- From: "Joseph S. Myers" <joseph at codesourcery dot com>
- To: Dmitrieva Liubov <liubov dot dmitrieva at gmail dot com>
- Cc: <libc-alpha at sourceware dot org>, Ulrich Drepper <drepper at gmail dot com>
- Date: Fri, 22 Jun 2012 14:55:33 +0000
- Subject: Re: Optimized with SSE2 sinf and cof for x86_32
- References: <CAHjhQ91iSZzUSaDs_MRO3AzzS-KjH7nBp=4ze_NoSTyomsbcHg@mail.gmail.com>
On Fri, 22 Jun 2012, Dmitrieva Liubov wrote:
> Our test system observed more than 1e4 ulp errors for |x|>1e4 for
> current GLIBC. New asm versions, provided here, are maximum 0.500121
> ulp for sinf, 0.500573 ulp for cosf.
Are you sure about *current* there - that is, using the flt-32 C
implementations? If so, please report those as bugs, with testcases
included; those implementations are used on other systems and would likely
have similar errors there as well. Since Andreas and I eliminated use of
the x87 trig instructions a few months ago, accurate range reduction
should mean such errors do not occur, but maybe there are bugs in the
range reduction.
Does your implementation address any of the open sinf or cosf bugs?
--
Joseph S. Myers
joseph@codesourcery.com