Re: [PATCH 3/5] x86: pgtable: Drop pXd_none() checks from pXd_free_pYd_table()

[Kani, Toshi]

On Mon, 2018-09-17 at 12:33 +0100, Will Deacon wrote:
> On Fri, Sep 14, 2018 at 08:37:48PM +0000, Kani, Toshi wrote:
> > On Wed, 2018-09-12 at 11:26 +0100, Will Deacon wrote:
> > > Now that the core code checks this for us, we don't need to do it in the
> > > backend.
> > > 
> > > Cc: Chintan Pandya <cpandya@xxxxxxxxxxxxxx>
> > > Cc: Toshi Kani <toshi.kani@xxxxxxx>
> > > Cc: Thomas Gleixner <tglx@xxxxxxxxxxxxx>
> > > Cc: Michal Hocko <mhocko@xxxxxxxx>
> > > Cc: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
> > > Signed-off-by: Will Deacon <will.deacon@xxxxxxx>
> > > ---
> > >  arch/x86/mm/pgtable.c | 6 ------
> > >  1 file changed, 6 deletions(-)
> > > 
> > > diff --git a/arch/x86/mm/pgtable.c b/arch/x86/mm/pgtable.c
> > > index ae394552fb94..b4919c44a194 100644
> > > --- a/arch/x86/mm/pgtable.c
> > > +++ b/arch/x86/mm/pgtable.c
> > > @@ -796,9 +796,6 @@ int pud_free_pmd_page(pud_t *pud, unsigned long addr)
> > >  	pte_t *pte;
> > >  	int i;
> > >  
> > > -	if (pud_none(*pud))
> > > -		return 1;
> > > -
> > 
> > Do we need to remove this safe guard?  I feel list this is same as
> > kfree() accepting NULL.
> 
> I think two big differences with kfree() are (1) that this function has
> exactly one caller in the tree and (2) it's implemented per-arch. Therefore
> we're in a good position to give it some simple semantics and implement
> those. Of course, if the x86 people would like to keep the redundant check,
> that's up to them, but I think it makes the function more confusing and
> tempts people into calling it for present entries.

With patch 1/5 change to have pXd_present() check, I agree that we can
remove this pXd_none() check to avoid any confusion.

Reviewed-by: Toshi Kani <toshi.kani@xxxxxxx>

Thanks,
-Toshi