[PATCH 06/19] ARM: mach-at91: remove arch specific special handling for ioremap

[Nicolas Pitre]

On Mon, 19 Sep 2011, Russell King - ARM Linux wrote:

> On Mon, Sep 19, 2011 at 03:35:47PM +0200, Jean-Christophe PLAGNIOL-VILLARD wrote:
> > On 03:07 Fri 16 Sep     , Nicolas Pitre wrote:
> > > From: Nicolas Pitre <nicolas.pitre at linaro.org>
> > > 
> > > A generic version should replace this later.
> > > 
> > > Signed-off-by: Nicolas Pitre <nicolas.pitre at linaro.org>
> > Untill the replace will availlable
> > 
> > do not remove it
> 
> This patch series is all about adding that replacement.
> 
> But yes, as a general point of principle, the replacement should be
> in place before the old stuff is pulled out.  I don't see any reason
> why that can't be the case, because __arm_ioremap() can be rigged up
> to use the static mappings first, and then the individual platform
> interception of ioremap() can be removed afterwards.

Some of them were using IOREMAP_END assuming it was below the static 
mappings, so it was easier to simply remove those first before the 
IOREMAP_END was redefined.  And then it looks strange to remove some of 
them before, then some others after.  Hence I opted for removing them 
all first.  And as I said, this doesn't create any functional 
regression to have this optimization removed temporarily.


Nicolas