Re: [PATCH 10/22] HWPOISON: check and isolate corrupted free pagesv2

[KAMEZAWA Hiroyuki]

On Mon, 15 Jun 2009 18:16:20 +0800
Wu Fengguang <fengguang.wu@xxxxxxxxx> wrote:

> On Mon, Jun 15, 2009 at 05:41:12PM +0800, KAMEZAWA Hiroyuki wrote:
> > On Mon, 15 Jun 2009 10:45:30 +0800
> > Wu Fengguang <fengguang.wu@xxxxxxxxx> wrote:
> > 
> > > From: Wu Fengguang <fengguang.wu@xxxxxxxxx>
> > > 
> > > If memory corruption hits the free buddy pages, we can safely ignore them.
> > > No one will access them until page allocation time, then prep_new_page()
> > > will automatically check and isolate PG_hwpoison page for us (for 0-order
> > > allocation).
> > > 
> > > This patch expands prep_new_page() to check every component page in a high
> > > order page allocation, in order to completely stop PG_hwpoison pages from
> > > being recirculated.
> > > 
> > > Note that the common case -- only allocating a single page, doesn't
> > > do any more work than before. Allocating > order 0 does a bit more work,
> > > but that's relatively uncommon.
> > > 
> > > This simple implementation may drop some innocent neighbor pages, hopefully
> > > it is not a big problem because the event should be rare enough.
> > > 
> > > This patch adds some runtime costs to high order page users.
> > > 
> > > [AK: Improved description]
> > > 
> > > v2: Andi Kleen:
> > > Port to -mm code
> > > Move check into separate function.
> > > Don't dump stack in bad_pages for hwpoisoned pages.
> > > 
> > > Signed-off-by: Wu Fengguang <fengguang.wu@xxxxxxxxx>
> > > Signed-off-by: Andi Kleen <ak@xxxxxxxxxxxxxxx>
> > > 
> > > ---
> > >  mm/page_alloc.c |   20 +++++++++++++++++++-
> > >  1 file changed, 19 insertions(+), 1 deletion(-)
> > > 
> > > --- sound-2.6.orig/mm/page_alloc.c
> > > +++ sound-2.6/mm/page_alloc.c
> > > @@ -233,6 +233,12 @@ static void bad_page(struct page *page)
> > >  	static unsigned long nr_shown;
> > >  	static unsigned long nr_unshown;
> > >  
> > > +	/* Don't complain about poisoned pages */
> > > +	if (PageHWPoison(page)) {
> > > +		__ClearPageBuddy(page);
> > > +		return;
> > > +	}
> > 
> > Hmm ? why __ClearPageBuddy() is necessary ?
> 
> Because this page is considered to be "allocated" out of the buddy
> system, even though we fail the allocation here.
> 
> The page is now owned by no one, especially not owned by the buddy
> allocator.
> 
I just wonder "why __ClearPageBuddy() is necessary."

When bad_page() is called,  a page is removed from buddy allocator and no
PG_buddy flag at all....I'm sorry if you added bad_page() caller in buddy allocator.

Buddy Allocator I call here is just 2 functions.
 - __free_one_page()
 - expand()


Bye,
-Kame



> Thanks,
> Fengguang
> 

--
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/