Re: [PATCH] modules: add default loader hook implementations

[Rusty Russell <rusty@rustcorp.com.au>]

On Mon, 27 Jun 2011 13:05:44 +0200, Arnd Bergmann <arnd@arndb.de> wrote:
> On Monday 27 June 2011, Geert Uytterhoeven wrote:
> > >> I don't care much either way, you would get my Ack for both solutions.
> > >> The __weak approach would definitely make a simpler patch, and the
> > >> patch you sent adds extra complexity because of the
> > >> asm_generic_moduleloader_hooks macro you used to avoid having to
> > >> change all other architectures.
> > >
> > > I think you misread me.  If all else is equal, I dislike weak functions.
> > > But AFAICT the two standard mechanisms are #ifdef HAVE_ARCH and __weak.
> > > Inventing a third one is not going to be a win.
> > 
> > It's not inventing a new one, the third one is already in use.
> 
> True. In fact, we are (slowly) migrating away from HAVE_ARCH_* elsewhere.
> In include/asm-generic/*.h, the common method is now to #define the exact
> symbol if an architecture wants to override the generic version.
> 
> Weak symbols are fairly obscure in comparison, but they are actively
> used by a few architectures (mips, sh) and some core code in kernel/
> and mm/.

I stand corrected.  It seems to be done one way or another pretty much
on a whim.

To avoid more bikeshedding, smallest patch wins.  That's __weak, right?

Thanks,
Rusty.