Re: Auto-Adaptive scheduler - Final chapter ( the numbers ) ...

[Davide Libenzi]

On Sat, 29 Jan 2000, Jamie Lokier wrote:
> Davide Libenzi wrote:
> > I can agree with You about cache optimization, but the fact is that the patch,
> > for RQ < limit introduced only a memory load ( 4 bytes ) and memory write ( 4
> > bytes ).
> 
> You forget about the icache footprint of the code itself.

No Jamie, the code leaves a footprint only when it is executed.
I You exclude the one or two "cmps" ( tipically less than 10 bytes of code,
adresses included ), the only thing at which We can shoot is at the block of
code of the cluster scheduler that I leaved into the fast path to better
optimize longer RQ versus shorter ones.

Cheers,
Davide.

-- 
All this stuff is IMVHO