Against 2.6.35-rc1+
My patch to "Factor out duplicate put/frees in mpol_shared_policy_init()
to a common return path"; and Dan Carpenter's fix thereto both left a
dangling reference to the incoming tmpfs superblock mempolicy structure.
A similar leak was introduced earlier when the nodemask was moved offstack
to the scratch area despite the note in the comment block regarding the
incoming ref.
Move the remaining 'put of the incoming "mpol" to the common exit path to
drop the reference.
Signed-off-by: Lee Schermerhorn <[email protected]>
mm/mempolicy.c | 9 +++++----
1 file changed, 5 insertions(+), 4 deletions(-)
Index: linux-2.6.35-rc2-mmotm-100611-1640/mm/mempolicy.c
===================================================================
--- linux-2.6.35-rc2-mmotm-100611-1640.orig/mm/mempolicy.c
+++ linux-2.6.35-rc2-mmotm-100611-1640/mm/mempolicy.c
@@ -2138,7 +2138,7 @@ void mpol_shared_policy_init(struct shar
NODEMASK_SCRATCH(scratch);
if (!scratch)
- return;
+ goto put_mpol;
/* contextualize the tmpfs mount point mempolicy */
new = mpol_new(mpol->mode, mpol->flags, &mpol->w.user_nodemask);
if (IS_ERR(new))
@@ -2147,19 +2147,20 @@ void mpol_shared_policy_init(struct shar
task_lock(current);
ret = mpol_set_nodemask(new, &mpol->w.user_nodemask, scratch);
task_unlock(current);
- mpol_put(mpol); /* drop our ref on sb mpol */
if (ret)
- goto put_free;
+ goto put_new;
/* Create pseudo-vma that contains just the policy */
memset(&pvma, 0, sizeof(struct vm_area_struct));
pvma.vm_end = TASK_SIZE; /* policy covers entire file */
mpol_set_shared_policy(sp, &pvma, new); /* adds ref */
-put_free:
+put_new:
mpol_put(new); /* drop initial ref */
free_scratch:
NODEMASK_SCRATCH_FREE(scratch);
+put_mpol:
+ mpol_put(mpol); /* drop our incoming ref on sb mpol */
}
}
On Tue, Jun 15, 2010 at 01:47:17PM -0400, Lee Schermerhorn wrote:
> Against 2.6.35-rc1+
>
> My patch to "Factor out duplicate put/frees in mpol_shared_policy_init()
> to a common return path"; and Dan Carpenter's fix thereto both left a
> dangling reference to the incoming tmpfs superblock mempolicy structure.
> A similar leak was introduced earlier when the nodemask was moved offstack
> to the scratch area despite the note in the comment block regarding the
> incoming ref.
>
Yup. You're right.
Acked-by: Dan Carpenter <[email protected]>
regards,
dan carpenter