Re: [PATCH] pci: introduce users of ioremap_pcibar()

[Arjan van de Ven <arjan@infradead.org>]

On Wed, 1 Oct 2008 09:07:17 -0400
Kyle McMartin <kyle@mcmartin.ca> wrote:
> > 
> > OK, not all architectures use the generic code, but I've been
> > through and they all do more or less the above (mn10300 and frv
> > just return the address, but their readl() and inl() are identical)
> > 
> 
> I don't recall anyone ever promising that the iomap interfaces would
> be usable with legacy accessors. I'd certainly prefer it if we
> didn't, as well, as it makes for more explicitly written drivers...
> 
> Just because you can use them, doesn't mean you should.
> 

ok so now we're full circle.
I started with a real ioremap, was told "no must use iomap" despite the
same argument you make, and now we're back to square one.

What I want is an interface that can replace ioremap() for the common
"I want the bar uncached" case. Nothing more nothing less.... sigh.


-- 
Arjan van de Ven 	Intel Open Source Technology Centre
For development, discussion and tips for power savings, 
visit http://www.lesswatts.org