mpn_mul is embarrassingly slow

[Niels Möller]

paul zimmermann <Paul.Zimmermann at inria.fr> writes:

>        Niels,
>
>> Such an assembly routine would need access to the threshold between
>> basecase and generic, which in the case of fat builds isn't a compile
>> time constant. 
>
> but you could determine at compile time a lower bound for fat builds, no?

If it's too inconvenient to get the runtime value, one could use the
compile-time minimum over those cpu flavors which will be using the
mul_basecase.asm file in question.

This can surely be solved, I just wanted to point out that it can get a
bit messy.

Regards,
/Niels

-- 
Niels Möller. PGP-encrypted email is preferred. Keyid 368C6677.
Internet email is subject to wholesale government surveillance.