[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[f-cpu] manual update
- To: <f-cpu@seul.org>
- Subject: [f-cpu] manual update
- From: "Nicolas Boulay" <nicolas.boulay@ifrance.com>
- Date: Thu, 6 Jun 2002 12:50:05 GMT
- Delivered-To: archiver@seul.org
- Delivered-To: f-cpu-outgoing@seul.org
- Delivered-To: f-cpu@seul.org
- Delivery-Date: Thu, 06 Jun 2002 08:50:14 -0400
- Reply-To: f-cpu@seul.org
- Send-By: 140.94.82.18 with Mozilla/4.0 (compatible; MSIE 5.0; Windows NT; DigExt; FR 15/06/2000)
- Sender: owner-f-cpu@seul.org
p195 Chapter 3
- We should thought away the reference to the xbar. If in our head xbar
== buses, it will not be the case of newcomers.
- We can't impose that more than 2 register can't point on the same
cache line. What happen with :
foo (int toto, int tata)
{
int a,b,c;
int * pa, *pb, *pc;
pa = &a;
pb=&b;
pc = &c;
}
a,b,c are allocate in the stack, so there are contigü in memory, so
there are on the same cache line.
We can't leave such enormous limitation to the fcpu.
- Memory stream are to differentiate data flow by not checking coherency
between them and then we could buffer write and note checking the adress
of an immediat following load. We absolutely never mind about the memory
technonoly used behind the memory controler (SDRAM bank in the manual
case). L2 cache are there to help the management.
So we should change all of this in the manual. Comment ?
nicO
______________________________________________________________________________
ifrance.com, l'email gratuit le plus complet de l'Internet !
vos emails depuis un navigateur, en POP3, sur Minitel, sur le WAP...
http://www.ifrance.com/_reloc/email.emailif
*************************************************************
To unsubscribe, send an e-mail to majordomo@seul.org with
unsubscribe f-cpu in the body. http://f-cpu.seul.org/