Re: reserved system calls?

Dominik Kubla (dominik.kubla@uni-mainz.de)
Thu, 11 Nov 1999 22:32:33 +0100


On Thu, Nov 11, 1999 at 01:02:41PM -0500, sergey@memco.com wrote:
...
>
> Actualy system_call_table[] ( for Intel platform ) contains 256 elements
> and one quater of them is still empty. There is no rule ( for now ) how to
> use it. Actualy I think Linux should have special mechanism for register
> and unregister dynamic system calls ( like Solaris OS ).
>

Actually there is one: new system calls are _ALWAYS_ appended at the end,
unless they had a fixed value reserved earlier in the development cycle.

Everything else kills backward compatibility for static binaries (failing
because a legacy syscall is no longer provided is preferred over executing
a random _WRONG_ call!).

BTW. Don't be so focused on the number 256: Linux/Alpha for example has over
300 due to support for OSF syscalls (IMHO a design bug: should have used
personalities for this, but maybe i am mistake and this is the only way to
do this...)

Yours
Dominik Kubla

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/