Re: [PATCH] Module loader preparation

[Christoph Hellwig]

On Thu, Oct 17, 2002 at 01:31:54PM +1000, Rusty Russell wrote:
> This patch fixes a couple of places using the old "I can just call my
> function init_module() and it will be called at module init" and a
> couple of modules without module_init() declarations.
> 
> These uses are obsolete with the in-kernel module loader, because the
> module_init() is where we put the module name in the ".modulename"
> section (we could have a "no_init_func()" thing, but it's fairly rare
> and hardly intuitive).

I don't think requiring a init func is a good idea.  Please fix
your module loader to generate a stub if no module_init() is
present instead.

init_module() sounds like a good idea to me, though.