Re: [AVX] Document changes

["H.J. Lu" <hjl dot tools at gmail dot com>]

On Wed, Aug 20, 2008 at 6:05 PM, Ye, Joey <joey.ye@intel.com> wrote:
> HJ wrote:
>
>> Please sort them by names.
>
>> Please mention that -mavx will generate AVX instructions, instead of
> SSE.
> See updated patch. OK for AVX branch?
>
> ChangeLog:
> 2008-08-21      Joey Ye <joey.ye@intel.com>
>
>        * doc/extend.texi: Document AVX built-in function.
>
>        * doc/invoke.texi: Document -mavx.
>
> Index: extend.texi
> ===================================================================
> --- extend.texi (revision 3736)
> +++ extend.texi (working copy)
> @@ -8271,6 +8271,141 @@
>  Generates the @code{popcntq} machine instruction.
>  @end table
>
> +The following built-in functions are available when @option{-mavx} is
> +used. All of them generate the machine instruction that is part of the
> +name.
> +
> +@smallexample
> +v4df __builtin_ia32_unpckhpd256 (v4df,v4df)
> +v8sf __builtin_ia32_unpckhps256 (v8sf,v8sf)
> +v4df __builtin_ia32_unpcklpd256 (v4df,v4df)
> +v8sf __builtin_ia32_unpcklps256 (v8sf,v8sf)
> +v2df __builtin_ia32_vpermil2pd (v2df,v2df,v2di,int)
> +v4df __builtin_ia32_vpermil2pd256 (v4df,v4df,v4di,int)
> +v4sf __builtin_ia32_vpermil2ps (v4sf,v4sf,v4si,int)
> +v8sf __builtin_ia32_vpermil2ps256 (v8sf,v8sf,v8si,int)
> +v4df __builtin_ia32_addpd256 (v4df,v4df)

Please sort the builtin list with "sort -k2".  OK with this change.  Or
you can check it in as is and I will sort it later.

Thanks.


-- 
H.J.