Re: [RFC][PATCH] Expanding the size of "start" and "end" field in "struct resource"

[Vivek Goyal <vgoyal@in.ibm.com>]

On Wed, Mar 15, 2006 at 01:13:56PM -0700, Eric W. Biederman wrote:
> Kumar Gala <galak@kernel.crashing.org> writes:
> 
> > On Mar 15, 2006, at 1:57 PM, Arjan van de Ven wrote:
> >
> >>
> >>> One of the possible solutions to this problem is that expand the size
> >>> of "start" and "end" to "unsigned long long". But whole of the PCI  and
> >>> driver code has been written assuming start and end to be unsigned  long
> >>> and compiler starts throwing warnings.
> >>
> >>
> >> please use dma_addr_t then instead of unsigned long long
> >>
> >> this is the right size on all platforms afaik (could a ppc64 person
> >> verify this?> ;)
> >
> > Actually we really just want "start" and "end" to be u64 on all  platforms.
> > Linus was ok with this change but no one has gone through  and fixed everything
> > that would be required for it.
> 
> Since it is faster to ask :)
> 
> How is it that other pieces of code have problems?
> Warnings or something nasty?

Few problems which I have noticed so far.

- Many printk() warnings. Wherever start and end are being printed,
  the format specifier being used is %lx. Needs to be changed to %Lx.

- Some folks save a pointer of type (unsigned long *) to start and end field
  and then try to operate on it. This pointer type shall have to be changed
  to something like u64*.

	unsigned long *port, *end, *tport, *tend;
	port = &dev->res.port_resource[idx].start;

- Some folks cast "start" to a pointer and then use it. Compiler gives warning. 

	addr_reg = (void __iomem *) addr->start;

-vivek