Re: Does io_remap_page_range() take 5 or 6 args?

[Richard B. Johnson]

On Wed, 18 Aug 2004, William Lee Irwin III wrote:

> On Wed, 18 Aug 2004 13:53:38 -0700 William Lee Irwin III wrote:
> >> Once it's decided how many it really takes, I'll fix up sparc32 as-needed.
>
> On Wed, Aug 18, 2004 at 01:56:38PM -0700, David S. Miller wrote:
> > (sorry for the emply reply previously)
> > It needs 6 unless we start passing in a 64-bit value to
> > io_remap_page_range()
> > for the 'offset' parameter.
> > Physical I/O addresses are 36-bits or so on sparc32 systems, which is
> > why we need to pass in "offset" and "space".
>

Maybe a pointer?

> We should pass 64-bit values to remap_page_range() also, then. Or
> perhaps passing pfn's to both suffices, as it all has to be page
> aligned anyway.
>
>
> -- wli

With a pointer to some kind of descriptor (structure), the arch-dependent
pointer size is handled by the compiler/linker and the arch-dependent
content is handled with the arch-dependent header defining the
structure. You get to keep both pieces and they don't break when
porting to a new box.

Cheers,
Dick Johnson
Penguin : Linux version 2.4.26 on an i686 machine (5570.56 BogoMips).
            Note 96.31% of all statistics are fiction.


-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/