Re: [PATCH] mm/memblock: replace u64 with phys_addr_t where appropriate

[Mike Rapoport]

On Wed, Jul 04, 2018 at 03:05:00PM +0200, Michal Hocko wrote:
> On Tue 03-07-18 20:05:06, Mike Rapoport wrote:
> > Most functions in memblock already use phys_addr_t to represent a physical
> > address with __memblock_free_late() being an exception.
> > 
> > This patch replaces u64 with phys_addr_t in __memblock_free_late() and
> > switches several format strings from %llx to %pa to avoid casting from
> > phys_addr_t to u64.
> > 
> > CC: Michal Hocko <mhocko@xxxxxxxxxx>
> > CC: Matthew Wilcox <willy@xxxxxxxxxxxxx>
> > Signed-off-by: Mike Rapoport <rppt@xxxxxxxxxxxxxxxxxx>
> > ---
> >  mm/memblock.c | 46 +++++++++++++++++++++++-----------------------
> >  1 file changed, 23 insertions(+), 23 deletions(-)
> > 
> > diff --git a/mm/memblock.c b/mm/memblock.c
> > index 03d48d8..20ad8e9 100644
> > --- a/mm/memblock.c
> > +++ b/mm/memblock.c
> > @@ -330,7 +330,7 @@ static int __init_memblock memblock_double_array(struct memblock_type *type,
> >  {
> >  	struct memblock_region *new_array, *old_array;
> >  	phys_addr_t old_alloc_size, new_alloc_size;
> > -	phys_addr_t old_size, new_size, addr;
> > +	phys_addr_t old_size, new_size, addr, new_end;
> >  	int use_slab = slab_is_available();
> >  	int *in_slab;
> >  
> > @@ -391,9 +391,9 @@ static int __init_memblock memblock_double_array(struct memblock_type *type,
> >  		return -1;
> >  	}
> >  
> > -	memblock_dbg("memblock: %s is doubled to %ld at [%#010llx-%#010llx]",
> > -			type->name, type->max * 2, (u64)addr,
> > -			(u64)addr + new_size - 1);
> > +	new_end = addr + new_size - 1;
> > +	memblock_dbg("memblock: %s is doubled to %ld at [%pa-%pa]",
> > +			type->name, type->max * 2, &addr, &new_end);
> 
> I didn't get to check this carefully but this surely looks suspicious. I
> am pretty sure you wanted to print the value here rather than address of
> the local variable, right?

It's the semantics of %pa:

Physical address types phys_addr_t
----------------------------------

::

	%pa[p]	0x01234567 or 0x0123456789abcdef

For printing a phys_addr_t type (and its derivatives, such as
resource_size_t) which can vary based on build options, regardless of the
width of the CPU data path.

Passed by reference.


> -- 
> Michal Hocko
> SUSE Labs
> 

-- 
Sincerely yours,
Mike.