[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[f-cpu] 64 K processes



ok ok I misunderstood :)

Well suppose that 64 K processES are running, do you compute how much it takes
as physical memory for all 64 K processes if we use a hierachical paging for
example ?

Ok take an example from ia32 :

paging is 2-level :

    pde = pdbr[virtaddr(31..22)].pde;
    pte = pde[virtaddr(21..12)].pte;
    physaddr = pte.pfn + virtaddr(11..0);

So for 64 K processes we need 64 K pdbrs, i.e, 64 K x 4 KB = 256 MB (ouch ! our
pdbrs will take 256 MB).

Now if each process has, say,  a 16 MB virtual space, we also need 16 MB / 4 MB
for pdes, i.e, 4 pdes and 16 MB / 4 KB for ptes, i.e, 4 K ptes. I'm speaking
about the worst case where all page frames are not shared !!!

So for 64 K processes we get at least 256 MB + (64 K x 4 x 4 KB) = 512 MB +
1024 MB =  1,5 MB if we suppose all ptes have invalid page frame ! quite
impressive isn't it ?

well my computer can handle only a max of 1,5 MB so using 64 K processes with
an average of 16 MB virtal space is impossible on my computer.

----- Original Message -----
From: Nicolas Boulay <nicolas.boulay@ifrance.com>
To: <f-cpu@seul.org>
Sent: Tuesday, March 05, 2002 3:32 PM
Subject: Rep:Re: Rep:Re: Rep:Re: [f-cpu] virtually or physically-addressed
cache ?


I wrote 64 K process, not 64 KB. It's just because i'm lazy to write
65535 as Marco suggest it.
Sorry, Christophe.

nicO

-----Message d'origine-----
De: "Christophe" <christophe.avoinne@laposte.net>
A: <f-cpu@seul.org>
Date: 05/03/02
Objet: Re: Rep:Re: Rep:Re: [f-cpu] virtually or physically-addressed
cache ?

64 KB process is not enough but threads can surely use less than 64 KB.

I prefer to speak about teams and threads. I consider a team as a
collection of
threads sharing the same address space and resources. The minimal
process
(without pthread) is quite like a team in fact with one thread. If you
take a
big application a 64 KB space would be largely insufficient even if
using
several threads because code can exceed 64 KB !

----- Original Message -----
From: Nicolas Boulay <nicolas.boulay@ifrance.com>
To: <f-cpu@seul.org>
Sent: Tuesday, March 05, 2002 9:17 AM
Subject: Rep:Re: Rep:Re: [f-cpu] virtually or physically-addressed cache
?


Nowadays isn't possible to create long term computer which manage only 4
Gb (it's the current physical limit of the pc)! 64 K process it's far
enough but it should address 1 To at least ! (nowdays big machine
address 64 Go)

nicO

-----Message d'origine-----
De: "Marco Al" <marco@simplex.nl>
A: <f-cpu@seul.org>
Date: 05/03/02
Objet: Re: Rep:Re: [f-cpu] virtually or physically-addressed cache ?

From: "Michael Riepe" <michael@stud.uni-hannover.de>

> The main drawback is that you'll have to know the number of processes
and
> the maximum total memory size in advance. You can do that in an
embedded
> system (works pretty well there), but not in a general-purpose OS.

There is a difference between what you can do and what you can
efficiently
do. In such a scheme if you have say a 48 bit global address space and
you
are using 32 bit addresses with byte addressing you can efficiently run
65K
processes with 2 GB address spaces each, that is not to say you cant run
more you will just have to make changes in the global address space and
either flush the cache or walk it to invalidate affected cachelines each
time a process not on the list is "swapped" in. Would even a hard limit
of
65K 4GB address spaces really present any issue for a general purpose OS
though? :)

Marco

*************************************************************
To unsubscribe, send an e-mail to majordomo@seul.org with
unsubscribe f-cpu       in the body. http://f-cpu.seul.org/


________________________________________________________________________
______
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/


*************************************************************
To unsubscribe, send an e-mail to majordomo@seul.org with
unsubscribe f-cpu       in the body. http://f-cpu.seul.org/


______________________________________________________________________________
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/

*************************************************************
To unsubscribe, send an e-mail to majordomo@seul.org with
unsubscribe f-cpu       in the body. http://f-cpu.seul.org/