Re: [PATCH 1/3] HWPOISON, hugetlbfs: fix warning on freeing hwpoisoned hugepage
From: Naoya Horiguchi
Date: Fri Dec 07 2012 - 10:51:56 EST
On Fri, Dec 07, 2012 at 01:24:45PM +0530, Aneesh Kumar K.V wrote:
> Naoya Horiguchi <n-horiguchi@xxxxxxxxxxxxx> writes:
>
> > On Fri, Dec 07, 2012 at 11:06:41AM +0530, Aneesh Kumar K.V wrote:
> > ...
> >> > From: Naoya Horiguchi <n-horiguchi@xxxxxxxxxxxxx>
> >> > Date: Thu, 6 Dec 2012 20:54:30 -0500
> >> > Subject: [PATCH v2] HWPOISON, hugetlbfs: fix warning on freeing hwpoisoned
> >> > hugepage
> >> >
> >> > This patch fixes the warning from __list_del_entry() which is triggered
> >> > when a process tries to do free_huge_page() for a hwpoisoned hugepage.
> >>
> >>
> >> Can you get a dump stack for that. I am confused because the page was
> >> already in freelist, and we deleted it from the list and set the
> >> refcount to 1. So how are we reaching free_huge_page() again ?
> >
> > free_huge_page() can be called for hwpoisoned hugepage from unpoison_memory().
> > This function gets refcount once and clears PageHWPoison, and then puts
> > refcount twice to return the hugepage back to free pool.
> > The second put_page() finally reaches free_huge_page().
> >
>
> Can we add this also to the commit message ?. With that you can add
>
> Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@xxxxxxxxxxxxxxxxxx>
OK, I added it.
Thanks for the review!
Naoya
---
From: Naoya Horiguchi <n-horiguchi@xxxxxxxxxxxxx>
Date: Thu, 6 Dec 2012 20:54:30 -0500
Subject: [PATCH v4] HWPOISON, hugetlbfs: fix warning on freeing hwpoisoned
hugepage
This patch fixes the warning from __list_del_entry() which is triggered
when a process tries to do free_huge_page() for a hwpoisoned hugepage.
ChangeLog v4:
- Add comment about when the warning is triggered
ChangeLog v3:
- Add comment
ChangeLog v2:
- simply use list_del_init instead of introducing new hugepage list
Signed-off-by: Naoya Horiguchi <n-horiguchi@xxxxxxxxxxxxx>
Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@xxxxxxxxxxxxxxxxxx>
---
mm/hugetlb.c | 8 +++++++-
1 file changed, 7 insertions(+), 1 deletion(-)
diff --git a/mm/hugetlb.c b/mm/hugetlb.c
index 59a0059..2511bcb 100644
--- a/mm/hugetlb.c
+++ b/mm/hugetlb.c
@@ -3170,7 +3170,13 @@ int dequeue_hwpoisoned_huge_page(struct page *hpage)
spin_lock(&hugetlb_lock);
if (is_hugepage_on_freelist(hpage)) {
- list_del(&hpage->lru);
+ /*
+ * Hwpoisoned hugepage isn't linked to activelist or freelist,
+ * but dangling hpage->lru can trigger list-debug warnings
+ * (this happens when we call unpoison_memory() on it),
+ * so let it point to itself with list_del_init().
+ */
+ list_del_init(&hpage->lru);
set_page_refcounted(hpage);
h->free_huge_pages--;
h->free_huge_pages_node[nid]--;
--
1.7.11.7
--
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/