* Re: [PATCH] mm/hugetlb.c: add NULL check of return value of huge_pte_offset
[not found] <533efd68.435fe00a.6936.ffffa5e7SMTPIN_ADDED_BROKEN@mx.google.com>
@ 2014-04-04 22:03 ` Andrew Morton
0 siblings, 0 replies; only message in thread
From: Andrew Morton @ 2014-04-04 22:03 UTC (permalink / raw)
To: Naoya Horiguchi
Cc: linux-kernel, mgorman, andi, sasha.levin, kirill.shutemov,
aneesh.kumar, linux-mm
On Fri, 04 Apr 2014 14:43:33 -0400 Naoya Horiguchi <n-horiguchi@ah.jp.nec.com> wrote:
> huge_pte_offset() could return NULL, so we need NULL check to avoid
> potential NULL pointer dereferences.
>
> --- a/mm/hugetlb.c
> +++ b/mm/hugetlb.c
> @@ -2662,7 +2662,8 @@ static int hugetlb_cow(struct mm_struct *mm, struct vm_area_struct *vma,
> BUG_ON(huge_pte_none(pte));
> spin_lock(ptl);
> ptep = huge_pte_offset(mm, address & huge_page_mask(h));
> - if (likely(pte_same(huge_ptep_get(ptep), pte)))
> + if (likely(ptep &&
> + pte_same(huge_ptep_get(ptep), pte)))
> goto retry_avoidcopy;
> /*
> * race occurs while re-acquiring page table
> @@ -2706,7 +2707,7 @@ static int hugetlb_cow(struct mm_struct *mm, struct vm_area_struct *vma,
> */
> spin_lock(ptl);
> ptep = huge_pte_offset(mm, address & huge_page_mask(h));
> - if (likely(pte_same(huge_ptep_get(ptep), pte))) {
> + if (likely(ptep && pte_same(huge_ptep_get(ptep), pte))) {
> ClearPagePrivate(new_page);
>
> /* Break COW */
Has anyone been hitting oopses here or was this from code inspection?
^ permalink raw reply [flat|nested] only message in thread