On Mon, Jun 11, 2012 at 05:17:29AM -0400, kosaki.motohiro@xxxxxxxxx wrote:
> From: KOSAKI Motohiro <kosaki.motohiro@xxxxxxxxxxxxxx>
>
> commit cc9a6c8776 (cpuset: mm: reduce large amounts of memory barrier related
> damage v3) introduced a memory corruption.
>
Ouch. No biscuits for Mel.
> shmem_alloc_page() passes pseudo vma and it has one significant unique
> combination, vma->vm_ops=NULL and (vma->policy->flags & MPOL_F_SHARED).
>
> Now, get_vma_policy() does NOT increase a policy ref when vma->vm_ops=NULL
> and mpol_cond_put() DOES decrease a policy ref when a policy has MPOL_F_SHARED.
> Therefore, when alloc_pages_vma() goes 'goto retry_cpuset' path, a policy
> refcount will be decreased too much and therefore it will make a memory corruption.
>
Yes, this is true. Hitting the bug requires that the cpuset is being
updated during the allocation so it's not a common but it is real. I'm
surprised I did not hit this while I was running the cpuset stress test
that originally introduced [get|put]_mems_allowed().
> This patch fixes it.
>
> Cc: Dave Jones <davej@xxxxxxxxxx>,
> Cc: Mel Gorman <mgorman@xxxxxxx>
> Cc: Christoph Lameter <cl@xxxxxxxxx>,
> Cc: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
> Cc: <stable@xxxxxxxxxxxxxxx>
> Cc: Miao Xie <miaox@xxxxxxxxxxxxxx>
> Cc: Peter Zijlstra <a.p.zijlstra@xxxxxxxxx>
> Signed-off-by: KOSAKI Motohiro <kosaki.motohiro@xxxxxxxxxxxxxx>
> Acked-by: Andi Kleen <ak@xxxxxxxxxxxxxxx>
> ---
> mm/mempolicy.c | 13 ++++++++++++-
> mm/shmem.c | 9 +++++----
> 2 files changed, 17 insertions(+), 5 deletions(-)
>
> diff --git a/mm/mempolicy.c b/mm/mempolicy.c
> index 7fb7d51..0da0969 100644
> --- a/mm/mempolicy.c
> +++ b/mm/mempolicy.c
> @@ -1544,18 +1544,29 @@ struct mempolicy *get_vma_policy(struct task_struct *task,
> struct vm_area_struct *vma, unsigned long addr)
> {
> struct mempolicy *pol = task->mempolicy;
> + int got_ref;
>
> if (vma) {
> if (vma->vm_ops && vma->vm_ops->get_policy) {
> struct mempolicy *vpol = vma->vm_ops->get_policy(vma,
> addr);
> - if (vpol)
> + if (vpol) {
> pol = vpol;
> + got_ref = 1;
> + }
> } else if (vma->vm_policy)
> pol = vma->vm_policy;
> }
> if (!pol)
> pol = &default_policy;
> +
> + /*
> + * shmem_alloc_page() passes MPOL_F_SHARED policy with vma->vm_ops=NULL.
> + * Thus, we need to take additional ref for avoiding refcount imbalance.
> + */
> + if (!got_ref && mpol_needs_cond_ref(pol))
> + mpol_get(pol);
> +
> return pol;
> }
>
> diff --git a/mm/shmem.c b/mm/shmem.c
> index d576b84..eb5f1eb 100644
> --- a/mm/shmem.c
> +++ b/mm/shmem.c
> @@ -919,6 +919,7 @@ static struct page *shmem_alloc_page(gfp_t gfp,
> struct shmem_inode_info *info, pgoff_t index)
> {
> struct vm_area_struct pvma;
> + struct page *page;
>
> /* Create a pseudo vma that just contains the policy */
> pvma.vm_start = 0;
> @@ -926,10 +927,10 @@ static struct page *shmem_alloc_page(gfp_t gfp,
> pvma.vm_ops = NULL;
> pvma.vm_policy = mpol_shared_policy_lookup(&info->policy, index);
>
> - /*
> - * alloc_page_vma() will drop the shared policy reference
> - */
> - return alloc_page_vma(gfp, &pvma, 0);
> + page = alloc_page_vma(gfp, &pvma, 0);
> +
> + mpol_put(pvma.vm_policy);
> + return page;
> }
Why does dequeue_huge_page_vma() not need to be changed as well? It's
currently using mpol_cond_put() but if there is a goto retry_cpuset then
will it have not take an additional reference count and leak?
Would it be more straight forward to put the mpol_cond_put() and __mpol_put()
calls after the "goto retry_cpuset" checks instead?
> #else /* !CONFIG_NUMA */
> #ifdef CONFIG_TMPFS
> --
> 1.7.1
>
--
Mel Gorman
SUSE Labs
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@xxxxxxxxx. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>