Fwd: [PATCH 5/9] magician: enable flash VPP GPIO and build in MTD, physmap-flash and JFFS2

[David Woodhouse]

On Fri, 2008-03-21 at 17:33 +0100, pHilipp Zabel wrote:
> Hi, I'd like to forward you a question about this patch that I'd like
> to get in via Russell King's queue.
> 
> The way the MTD subsystem / physmap driver handle set_vpp, can the
> magician_set_vpp function below race with itself? Or would it be
> better to use a counting approach like omap_set_vpp in the omap-nor
> driver?

I believe that when used for a single chip (or set of chips with a
single map driver), all access is serialised. It's only if you use the
same set_vpp() function for multiple different maps that you need to
play counting tricks.

-- 
dwmw2