On Tue, 6 Aug 2013 12:06:56 +0800 Jianguo Wu <wujian...@huawei.com> wrote:

> If node == NUMA_NO_NODE, pol is NULL, we should return NULL instead of
> do "if (!pol->mode)" check.
> 
> Signed-off-by: Jianguo Wu <wujian...@huawei.com>
> ---
>  mm/mempolicy.c | 2 ++
>  1 file changed, 2 insertions(+)
> 
> diff --git a/mm/mempolicy.c b/mm/mempolicy.c
> index 4baf12e..e0e3398 100644
> --- a/mm/mempolicy.c
> +++ b/mm/mempolicy.c
> @@ -129,6 +129,8 @@ static struct mempolicy *get_task_policy(struct 
> task_struct *p)
>               node = numa_node_id();
>               if (node != NUMA_NO_NODE)
>                       pol = &preferred_node_policy[node];
> +             else
> +                     return NULL;
>  
>               /* preferred_node_policy is not initialised early in boot */
>               if (!pol->mode)

Well yes, it'll dereference a null pointer

This is neater, I think:

--- 
a/mm/mempolicy.c~mm-mempolicy-return-null-if-node-is-numa_no_node-in-get_task_policy
+++ a/mm/mempolicy.c
@@ -123,16 +123,19 @@ static struct mempolicy preferred_node_p
 static struct mempolicy *get_task_policy(struct task_struct *p)
 {
        struct mempolicy *pol = p->mempolicy;
-       int node;
 
        if (!pol) {
-               node = numa_node_id();
-               if (node != NUMA_NO_NODE)
-                       pol = &preferred_node_policy[node];
+               int node = numa_node_id();
 
-               /* preferred_node_policy is not initialised early in boot */
-               if (!pol->mode)
-                       pol = NULL;
+               if (node != NUMA_NO_NODE) {
+                       pol = &preferred_node_policy[node];
+                       /*
+                        * preferred_node_policy is not initialised early in
+                        * boot
+                        */
+                       if (!pol->mode)
+                               pol = NULL;
+               }
        }
 
        return pol;
_

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

Reply via email to