FAT GMP 5 binaries

Karl Hasselstrom kha@treskal.com
Wed, 14 May 2003 17:57:27 +0200

Content-Type: text/plain; charset=iso-8859-1
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

On 2003-05-14 16:49:47 +0200, Torbjorn Granlund wrote:

> Karl Hasselstrom <kha@treskal.com> writes:
>   On 2003-05-14 14:53:44 +0200, Torbjorn Granlund wrote:
>   > We need to decide what routines to put in __gmp_cpuvec. All GMP
>   > rouines is one choice. All mpn routines or a subset of mpn
>   > rouines are other choices.
>   The routines that have been hand-coded in assembly are obvious
>   candidates for this. For routines written in C, would the speedup
>   be worth the effort?
> Important routines written in C could call mpn_addmul_1_CPU,
> directly, instead of going via __gmp_cpuvec.

GMP is pretty big as it is, so duplicating _all_ functions doen't
strike me as a particularly good idea. On the other hand, just
selecting a predefined subset, such as all mpn functions, is sure to
miss a few functions that would benefit, and include some functions
that don't really need it.

So I'll chicken out and recommend that including/excluding a function
=66rom __gmp_cpuvec be made really easy, so it'll be easy to change.
Maybe it would be a good idea to have the possiblity for _every_
function to be duplicated, controlled by a single #define per
function. (Some of that functionality will be needed anyway, since we
presumably still want to be able to compile a library optimized for a
single computer, preprocessing the whole __gmp_cpuvec machinery into

Karl Hasselstr=F6m, kha@treskal.com

Content-Type: application/pgp-signature
Content-Disposition: inline

Version: GnuPG v1.2.2 (GNU/Linux)