Re: [RFC PATCH 4/6] mm, compaction: skip buddy pages by their order in the migrate scanner

From: David Rientjes
Date: Mon Jun 09 2014 - 18:25:23 EST


On Mon, 9 Jun 2014, Vlastimil Babka wrote:

> > > > Sorry, I meant ACCESS_ONCE(page_private(page)) in the migration scanner
> > >
> > > Hm but that's breaking the abstraction of page_order(). I don't know if
> > > it's
> > > worse to create a new variant of page_order() or to do this. BTW, seems
> > > like
> > > next_active_pageblock() in memory-hotplug.c should use this variant too.
> > >
> >
> > The compiler seems free to disregard the access of a volatile object above
> > because the return value of the inline function is unsigned long. What's
> > the difference between unsigned long order = page_order_unsafe(page) and
> > unsigned long order = (unsigned long)ACCESS_ONCE(page_private(page)) and
>
> I think there's none functionally, but one is abstraction layer violation and
> the other imply the context of usage as you say (but is that so uncommon?).
>
> > the compiler being able to reaccess page_private() because the result is
> > no longer volatile qualified?
>
> You think it will reaccess? That would defeat all current ACCESS_ONCE usages,
> no?
>

I think the compiler is allowed to turn this into

if (ACCESS_ONCE(page_private(page)) > 0 &&
ACCESS_ONCE(page_private(page)) < MAX_ORDER)
low_pfn += (1UL << ACCESS_ONCE(page_private(page))) - 1;

since the inline function has a return value of unsigned long but gcc may
not do this. I think

/*
* Big fat comment describing why we're using ACCESS_ONCE(), that
* we're ok to race, and that this is meaningful only because of
* the previous PageBuddy() check.
*/
unsigned long pageblock_order = ACCESS_ONCE(page_private(page));

is better.
--
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/