2015-02-25 00:16:10

by David Rientjes

Subject: [patch] mm, hugetlb: close race when setting PageTail for gigantic pages

Now that gigantic pages are dynamically allocatable, care must be taken
to ensure that p->first_page is valid before setting PageTail.

If this isn't done, then it is possible to race and have compound_head()
return NULL.

Signed-off-by: David Rientjes <[email protected]>
---
mm/hugetlb.c | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/mm/hugetlb.c b/mm/hugetlb.c
--- a/mm/hugetlb.c
+++ b/mm/hugetlb.c
@@ -917,7 +917,6 @@ static void prep_compound_gigantic_page(struct page *page, unsigned long order)
__SetPageHead(page);
__ClearPageReserved(page);
for (i = 1; i < nr_pages; i++, p = mem_map_next(p, page, i)) {
- __SetPageTail(p);
/*
* For gigantic hugepages allocated through bootmem at
* boot, it's safer to be consistent with the not-gigantic
@@ -933,6 +932,9 @@ static void prep_compound_gigantic_page(struct page *page, unsigned long order)
__ClearPageReserved(p);
set_page_count(p, 0);
p->first_page = page;
+ /* Make sure p->first_page is always valid for PageTail() */
+ smp_wmb();
+ __SetPageTail(p);
}
}

2015-02-25 02:20:35

by Davidlohr Bueso

Subject: Re: [patch] mm, hugetlb: close race when setting PageTail for gigantic pages

On Tue, 2015-02-24 at 16:16 -0800, David Rientjes wrote:
> Now that gigantic pages are dynamically allocatable, care must be taken
> to ensure that p->first_page is valid before setting PageTail.
>
> If this isn't done, then it is possible to race and have compound_head()
> return NULL.
>
> Signed-off-by: David Rientjes <[email protected]>

Acked-by: Davidlohr Bueso <[email protected]>

2015-02-25 02:33:39

by Hillf Danton

Subject: Re: [patch] mm, hugetlb: close race when setting PageTail for gigantic pages

> Now that gigantic pages are dynamically allocatable, care must be taken
> to ensure that p->first_page is valid before setting PageTail.
>
> If this isn't done, then it is possible to race and have compound_head()
> return NULL.
>
> Signed-off-by: David Rientjes <[email protected]>
> ---
Acked-by: Hillf Danton <[email protected]>

> mm/hugetlb.c | 4 +++-
> 1 file changed, 3 insertions(+), 1 deletion(-)
>
> diff --git a/mm/hugetlb.c b/mm/hugetlb.c
> --- a/mm/hugetlb.c
> +++ b/mm/hugetlb.c
> @@ -917,7 +917,6 @@ static void prep_compound_gigantic_page(struct page *page, unsigned long order)
> __SetPageHead(page);
> __ClearPageReserved(page);
> for (i = 1; i < nr_pages; i++, p = mem_map_next(p, page, i)) {
> - __SetPageTail(p);
> /*
> * For gigantic hugepages allocated through bootmem at
> * boot, it's safer to be consistent with the not-gigantic
> @@ -933,6 +932,9 @@ static void prep_compound_gigantic_page(struct page *page, unsigned long order)
> __ClearPageReserved(p);
> set_page_count(p, 0);
> p->first_page = page;
> + /* Make sure p->first_page is always valid for PageTail() */
> + smp_wmb();
> + __SetPageTail(p);
> }
> }
>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to [email protected]
> More majordomo info at http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at http://www.tux.org/lkml/