Re: [PATCH] Allow linker to eliminate unused functions in lib/*

[Dave Chinner]

On Mon, Jan 11, 2010 at 03:24:12PM +0100, Andi Kleen wrote:
> On Mon, Jan 11, 2010 at 10:17:08AM +1100, Dave Chinner wrote:
> > On Sat, Jan 09, 2010 at 05:51:56PM +0100, Andi Kleen wrote:
> > > > Sneaky.
> > > > 
> > > > It's a bit sad to reduce the code cleanliness and maintainability in
> > > > this way, but 1k is 1k.
> > > 
> > > Yes. I had actually hoped for more too. I think it might bring 
> > > more with -ffunction-sections/--gc-sections, but I haven't tried that.
> > > 
> > > An alternative might be also to give some more of those explicit Kconfig
> > > symbols -- then kernels with module loading enabled would benefit more
> > > too -- but I haven't investigated how many changes that would need.
> > > 
> > > But even with that it's a reasonable improvement and putting the exports
> > > separately is not too bad.
> > > 
> > > > The patch assumes a pre-existing lib-syms.c.  Confused.
> > > 
> > > Sorry, some day I'll learn how to operate quilt properly.
> > > 
> > > Updated patch with all files included appended.
> > 
> > Andi, how do you want to handle the overlap between this and the
> > list_sort() patch I sent? I'm going to need a list_sort() available
> > in the XFS tree well before the 2.6.34 merge window....
> 
> Shouldn't be difficult to merge, it's just a few lines. Someone will
> do that.
> 
> It wou;ld be better if you put it into a separate file like I suggested
> earlier though, then the list_sort() could be actually eliminated when
> not needed.

OK, I'll respin it to do that, then push it onwards.

Cheers,

Dave.
-- 
Dave Chinner
david@xxxxxxxxxxxxx
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/