On Wed 20-07-16 18:29:02, Naoya Horiguchi wrote:
[...]
> >From 7da52ca6920dcd84e3da2df619bd5242f9c3ccec Mon Sep 17 00:00:00 2001
> From: Naoya Horiguchi <n-horigu...@ah.jp.nec.com>
> Date: Wed, 20 Jul 2016 18:21:33 +0900
> Subject: [PATCH v2] mm: hwpoison: remove incorrect comment
>
> dequeue_hwpoisoned_huge_page() can be called without page lock hold,
> so let's remove incorrect comment.
>
> The reason why the page lock is not really needed is that
> dequeue_hwpoisoned_huge_page() checks page_huge_active() inside hugetlb_lock,
> which allows us to avoid trying to dequeue a hugepage that are just allocated
> but not linked to active list yet, even without taking page lock.
Thank you for the clarification!
> Reported-by: Zhan Chen <zha...@andrew.cmu.edu>
> Signed-off-by: Naoya Horiguchi <n-horigu...@ah.jp.nec.com>
Acked-by: Michal Hocko <mho...@suse.com>
> ---
> mm/hugetlb.c | 1 -
> mm/memory-failure.c | 2 --
> 2 files changed, 3 deletions(-)
>
> diff --git a/mm/hugetlb.c b/mm/hugetlb.c
> index c1f3c0be150a..26f735cc7478 100644
> --- a/mm/hugetlb.c
> +++ b/mm/hugetlb.c
> @@ -4401,7 +4401,6 @@ follow_huge_pud(struct mm_struct *mm, unsigned long
> address,
>
> /*
> * This function is called from memory failure code.
> - * Assume the caller holds page lock of the head page.
> */
> int dequeue_hwpoisoned_huge_page(struct page *hpage)
> {
> diff --git a/mm/memory-failure.c b/mm/memory-failure.c
> index 2fcca6b0e005..7532c3a8a39c 100644
> --- a/mm/memory-failure.c
> +++ b/mm/memory-failure.c
> @@ -741,8 +741,6 @@ static int me_huge_page(struct page *p, unsigned long pfn)
> * page->lru because it can be used in other hugepage operations,
> * such as __unmap_hugepage_range() and gather_surplus_pages().
> * So instead we use page_mapping() and PageAnon().
> - * We assume that this function is called with page lock held,
> - * so there is no race between isolation and mapping/unmapping.
> */
> if (!(page_mapping(hpage) || PageAnon(hpage))) {
> res = dequeue_hwpoisoned_huge_page(hpage);
> --
> 2.7.4
>
> --
> To unsubscribe, send a message with 'unsubscribe linux-mm' in
> the body to majord...@kvack.org. For more info on Linux MM,
> see: http://www.linux-mm.org/ .
> Don't email: <a href=mailto:"d...@kvack.org";> em...@kvack.org </a>
--
Michal Hocko
SUSE Labs
