Subject: Re: ffs fragmentation
To: Manuel Bouyer <>
From: Mike Cheponis <mac@Wireless.Com>
List: tech-kern
Date: 05/15/1999 15:01:33
> Manuel Bouyer, LIP6, Universite Paris VI. muses...

>On May 14, Mike Cheponis wrote
>> The advantages of DynaSwap are:
>> (1) You don't need to have a separate swap partition
>> (2) You don't need to know the size of any of the disk used for "swap" - 
>>     the OS dyanmically does it for you
>> (3) Large (up to the size of unallocated disk space) arrays enable huge
>>     datasets.
> And, for most applications I can think, this is a "feature" I really don't
> want. In fact, I can't see the interest of this. When you set up a system,
> you still need to know how much swap and filesystem sizes you need, to
> get a disk large enouth. Once you have these values you can set up your
> partitions properly.

Think of applications that use algorithm similar to chess-playing programs
go deep and wide in their searches, and it's unknown at the outset how much
dynamic memory will be required.

It seems to me silly and antiquated to have a "fixed" swap.

The suggestion to open a file as swap is better, but still does not do
what I want, which is something like this:

char big[9876543210]; /* That's 9.876.543.210 bytes */
void main(){
  big[ ] = foo;

or, even better:

void main(){
  char big[9876543210]; /* That's 9.876.543.210 bytes */
  big[ ] = foo;

In the second case, it's stack space that's used.


IMHO, an OS should provide the proper support to handle the
above-mentioned cases.