Re: [RFC] [PATCH] Fix b43 driver build for arm

[Michael Buesch <mb@bu3sch.de>]

On Monday 18 February 2008 23:50:30 Harvey Harrison wrote:
> On Mon, 2008-02-18 at 23:43 +0100, Michael Buesch wrote:
> > On Monday 18 February 2008 23:34:10 Russell King wrote:
> > > 
> > > Well, don't expect this driver to work until you fix your broken
> > > assumptions about alignment requirements.
> > 
> > Mr King, I'm not an idiot!
> > 
> > Can you _please_ explain what makes ARM so special here?
> > Why can't we have an array of this structure on ARM?
> > 
> > struct ssb_device_id {
> >        __u16   vendor;
> >        __u16   coreid;
> >        __u8    revision;
> > };
> > 
> > I will not apply any patches that I don't understand.
> > Why doesn't the compiler handle this? What's special? Can you please explain?
> > 
> 
> I believe this is a good place to start (although I could be totally
> off-base)
> 
> http://lkml.org/lkml/2007/11/22/120

I know very well what unaligned access is. As I said the code
works on MIPS, which can't do unaligned accesses.

The _real_ question is, why doesn't align the compiler the stuff properly
on ARM? It does the right thing on x86_32/64, powerpc and MIPS. Why doesn't
it do the right thing on ARM and we have to manually align stuff?

See section "Code that doesn't cause unaligned access"

-- 
Greetings Michael.