PATCH: optimized libm single precision routines: erfcf, erff, expf for x86_64.

[Dmitrieva Liubov]

Exponent code was fixed according to H.J comments.
Thanks.

--
Liubov Dmitrieva
Intel Corporation

2012/2/16 H.J. Lu <hjl.tools@gmail.com>:
> On Wed, Feb 15, 2012 at 4:02 AM, Dmitrieva Liubov
> <liubov.dmitrieva@gmail.com> wrote:
>> Hello,
>>
>> Let's start with exponent function at the moment. We significantly
>> improved a design of source code and would like suggest to release it.
>> A patch is attached.
>>
>> ChangeLog:
>>
>> 2012-02-15  Liubov Dmitrieva  <liubov.dmitrieva@gmail.com>
>>
>>       * sysdeps/x86_64/fpu/e_expf.S: New file.
>>
>
> My comments:
>
> 1.  Do you need 2
>
> +       .align  ALIGNARG(5)
> +       .align  ALIGNARG(5)
>
> The table is an array of double.  Will 8byte alignment also work?
>
> 2. All local labels should have L().
>
>
> H.J.
>
> --
> H.J.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: exp.patch
Type: application/octet-stream
Size: 12130 bytes
Desc: not available
URL: <http://sourceware.org/pipermail/libc-alpha/attachments/20120216/230791ff/attachment.obj>