Pentium and Repentium

[Kevin Ryde]

Torbjorn Granlund <tg at swox.com> writes:
>
> I'd expect we could get things to run well on this processor with
> some software pipelining.  Our current code relies on out-of-order
> execution, but Prescott's out-of-order capability seems
> insufficient for its over 30 steps deep pipeline.

There's some spare registers in most funcs, but not a lot.

> mpn_lshift      1.87            2.12

The aim there is 1.5 c/l, being 3 mmx insns for each 2 limbs.  I don't
know why it doesn't run at that pace now, in fact.  Probably
scheduling.