Re: [ckrm-tech] [PATCH 2/6] CKRM: Core framework support

[Dave Hansen <haveblue@us.ibm.com>]

On Thu, 2005-06-30 at 11:08 -0700, Chandra Seetharaman wrote:
> On Wed, 2005-06-29 at 10:22 +0900, YAMAMOTO Takashi wrote:
> > > > > +	if (pud_none(*pud))
> > > > > +		return 0;
> > > > > +	BUG_ON(pud_bad(*pud));
> > > > > +	pmd = pmd_offset(pud, address);
> > > > > +	pgd_end = (address + PGDIR_SIZE) & PGDIR_MASK;
> > > > 
> > > > why didn't you introduce class_migrate_pud?
> > > 
> > > Because there is no list to iterate through. 
> > 
> > i don't understand what you mean.
> > why you don't iterate pud, while you iterate pgdir and pmd?
> 
> what i meant was that the pmd are an array and there is no array
> w.r.t puds. correct me if i am wrong.

You are wrong :)

PUDs are the new level added to the pagetables for the architectures
with larger address spaces.  They are equivalent in functionality to all
other levels, and must be treated as such in any generic code.  

For 2 or 3-level pagetables, they do effectively collapse down to
nothing, but you still have to be concerned with it in generic code.

-- Dave

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"aart@kvack.org"> aart@kvack.org </a>