From: Michele Andreoli (m.andreoli@tin.it)
Date: Thu Nov 16 2000 - 21:03:36 CET
On Thu, Nov 16, 2000 at 07:19:44PM +0100, Dumas Patrice nicely wrote:
> > can't run the command that unpack in the way you explained :-(
>
> zut ! :-(.
> But there is something I don't understand. Does it happens even when the root
> file tree isn't in ram, but on a floppy ?
Oh, no: in this case surely no.
>That't quite strange... What does
> ash fill the memory with ?
The problem is intrisec to concept of shell script: any new command in
the script is called as a separated child process (the fork() stuff).
Very different if you write a C program: you have now only a single
process, if you dot use the fork() in your program.
>
> > If you plan to do that with scripts, you have no hope. CAN'T FORK
> > will be the result. The only solution is to write the /linuxrc in C.
>
> It wouldn't be suficient, as after the boot a lot of shell scripts would be
> called.
No if /linuxrc is really able to clone the system immediately. But to
write a command like that is very difficult. You must call compressors etc
as external programs: this requires to use system() in C, which call
itself the shell interpreter, and so on. Otherwise, you have to
reimplement decompression in /linuxrc!
>
> Does someone know a way to restrict the memory linux sees, so that I can test
> with my box with plenty of RAM what it happens in boxes with low memory ?
>
This is a good idea: if you set ramdisk_size= (in your boot loader) to
a value close to your physical ram, you can waste the RAM with a command
like:
dd if=/dev/zero of=/dev/ram1 bs=1k count=20000 (20000 kB)!
Put that in the top of /linuxrc and experiment!
Michele
-- In summing up, I wish I had some kind of affirmative message to leave you with, I don't. Would you take two negative messages? - Woody Allen --------------------------------------------------------------------- To unsubscribe, e-mail: mulinux-unsubscribe@sunsite.auc.dk For additional commands, e-mail: mulinux-help@sunsite.auc.dk
This archive was generated by hypermail 2.1.6 : Sat Feb 08 2003 - 15:27:16 CET