Re: Solution: Modules under 2.1.0

Matthias Urlichs (smurf@smurf.noris.de)
Thu, 26 Sep 1996 13:50:03 +0100

In linux.dev.kernel, article <Pine.LNX.3.91.960926082135.418A-100000@HarryH.wg41.rus.uni-stuttgart.de>,
"Hoyer, Harald" <root@HarryH.wg41.rus.uni-stuttgart.de> writes:
> After 3 hours of lookin around what is wrong with insmod ... here is my
> small solution. Insmod is backward-compatible. I changed the
> parameterlist for sys_create_module because an unsigned long > MAX_INT is
> treated as en error for libc or whoever. SO it's a bad idea to take the

But that's a generic problem which affects some of other kernel functions
too (mmap(), brk(), ...). Adding another parameter to create_module() might
work for create_module(), but not for the others.

Besides, you now vane effectively created a new system call. For backwards
compatibility, that system call needs a new number, IMHO.

There's another patch which counts only negative numbers from -1 to -4095
as errors; I think that's preferable, and we can live without being able to
access the last page if somebody ever runs Linux with 1 GByte of main
memory.

-- 
Harry's bar has a new cocktail.  It's called MRS punch.  They make it with
milk, rum and sugar and it's wonderful.  The milk is for vitality and the 
sugar is for pep.  They put in the rum so that people will know what to do
with all that pep and vitality.

-- 
Matthias Urlichs         \  noris network GmbH  /  Xlink-POP Nürnberg 
Schleiermacherstraße 12   \   Linux+Internet   /   EMail: urlichs@noris.de
90491 Nürnberg (Germany)   \    Consulting+Programming+Networking+etc'ing
   PGP: 1024/4F578875   1B 89 E2 1C 43 EA 80 44  15 D2 29 CF C6 C7 E0 DE
       Click <A HREF="http://info.noris.de/~smurf/finger">here</A>.    42