Re: insmod question

Rogier Wolff (R.E.Wolff@BitWizard.nl)
Mon, 4 May 1998 13:31:53 +0200 (MET DST)


Rik van Riel wrote:
>
> On Sun, 3 May 1998, Alex Buell wrote:
> > On Sun, 3 May 1998, Albert D. Cahalan wrote:
> >
> > > It would be very good to fix this. Many other unix systems put
> > > the whole symbol table in the kernel. We should do that too.
> > > With compression, I get a 46kB map file.
> >
> > Perhaps this should be placed into a loadable module?
>
> No,

I'd say "no" for a completely different reason: If you put it in a
module you get the same "version trouble" that you get when you use
"system.map": You can get an old version of the module that seems to
belong to the current kernel.

The kernel exports symbols that you can use in modules through ksyms.
To prevent the API from becoming riddled with thousands of calls (Like
the well-known MS operating systems), we purposely keep the number
of exported functions down as much as possible.

The "Hack" of using system.map is a great hack to get stuff to work as
soon as possible, but shouldn't be used for production.

If you need a symbol, convince Linus that you really need it and have
it exported, or use a workaround.

Roger.

-- 
If it's there and you can see it, it's REAL      |___R.E.Wolff@BitWizard.nl  |
If it's there and you can't see it, it's TRANSPARENT |  Tel: +31-15-2137555  |
If it's not there and you can see it, it's VIRTUAL   |__FAX:_+31-15-2138217  |
If it's not there and you can't see it, it's GONE! -- Roy Wilks, 1983  |_____|

- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu