mempolicy: fix a race in shared_policy_replace()

shared_policy_replace() use of sp_alloc() is unsafe.  1) sp_node cannot be
dereferenced if sp->lock is not held and 2) another thread can modify
sp_node between spin_unlock for allocating a new sp node and next
spin_lock.  The bug was introduced before 2.6.12-rc2.

Kosaki's original patch for this problem was to allocate an sp node and
policy within shared_policy_replace and initialise it when the lock is
reacquired.  I was not keen on this approach because it partially
duplicates sp_alloc().  As the paths were sp->lock is taken are not that
performance critical this patch converts sp->lock to sp->mutex so it can
sleep when calling sp_alloc().

[kosaki.motohiro@jp.fujitsu.com: Original patch]
Signed-off-by: Mel Gorman <mgorman@suse.de>
Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Reviewed-by: Christoph Lameter <cl@linux.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

diff --git a/include/linux/mempolicy.h b/include/linux/mempolicy.h
index ba7a0ff19d39997f84f483343bfecc5a15886db2..cec56932560823ac6a34ce5c5c21bea24316a9ea 100644 (file)
--- a/include/linux/mempolicy.h
+++ b/include/linux/mempolicy.h
@@ -188,7 +188,7 @@ struct sp_node {
 
 struct shared_policy {
        struct rb_root root;
-       spinlock_t lock;
+       struct mutex mutex;
 };
 
 void mpol_shared_policy_init(struct shared_policy *sp, struct mempolicy *mpol);

diff --git a/mm/mempolicy.c b/mm/mempolicy.c
index f0728ae74672262939e58e920be13f60b893e18f..b2f12ecc1b346902a5a756e0d74f3abbe15f9763 100644 (file)
--- a/mm/mempolicy.c
+++ b/mm/mempolicy.c
@@ -2083,7 +2083,7 @@ bool __mpol_equal(struct mempolicy *a, struct mempolicy *b)
  */
 
 /* lookup first element intersecting start-end */
-/* Caller holds sp->lock */
+/* Caller holds sp->mutex */
 static struct sp_node *
 sp_lookup(struct shared_policy *sp, unsigned long start, unsigned long end)
 {
@@ -2147,13 +2147,13 @@ mpol_shared_policy_lookup(struct shared_policy *sp, unsigned long idx)
 
        if (!sp->root.rb_node)
                return NULL;
-       spin_lock(&sp->lock);
+       mutex_lock(&sp->mutex);
        sn = sp_lookup(sp, idx, idx+1);
        if (sn) {
                mpol_get(sn->policy);
                pol = sn->policy;
        }
-       spin_unlock(&sp->lock);
+       mutex_unlock(&sp->mutex);
        return pol;
 }
 
@@ -2193,10 +2193,10 @@ static struct sp_node *sp_alloc(unsigned long start, unsigned long end,
 static int shared_policy_replace(struct shared_policy *sp, unsigned long start,
                                 unsigned long end, struct sp_node *new)
 {
-       struct sp_node *n, *new2 = NULL;
+       struct sp_node *n;
+       int ret = 0;
 
-restart:
-       spin_lock(&sp->lock);
+       mutex_lock(&sp->mutex);
        n = sp_lookup(sp, start, end);
        /* Take care of old policies in the same range. */
        while (n && n->start < end) {
@@ -2209,16 +2209,14 @@ restart:
                } else {
                        /* Old policy spanning whole new range. */
                        if (n->end > end) {
+                               struct sp_node *new2;
+                               new2 = sp_alloc(end, n->end, n->policy);
                                if (!new2) {
-                                       spin_unlock(&sp->lock);
-                                       new2 = sp_alloc(end, n->end, n->policy);
-                                       if (!new2)
-                                               return -ENOMEM;
-                                       goto restart;
+                                       ret = -ENOMEM;
+                                       goto out;
                                }
                                n->end = start;
                                sp_insert(sp, new2);
-                               new2 = NULL;
                                break;
                        } else
                                n->end = start;
@@ -2229,12 +2227,9 @@ restart:
        }
        if (new)
                sp_insert(sp, new);
-       spin_unlock(&sp->lock);
-       if (new2) {
-               mpol_put(new2->policy);
-               kmem_cache_free(sn_cache, new2);
-       }
-       return 0;
+out:
+       mutex_unlock(&sp->mutex);
+       return ret;
 }
 
 /**
@@ -2252,7 +2247,7 @@ void mpol_shared_policy_init(struct shared_policy *sp, struct mempolicy *mpol)
        int ret;
 
        sp->root = RB_ROOT;             /* empty tree == default mempolicy */
-       spin_lock_init(&sp->lock);
+       mutex_init(&sp->mutex);
 
        if (mpol) {
                struct vm_area_struct pvma;
@@ -2318,7 +2313,7 @@ void mpol_free_shared_policy(struct shared_policy *p)
 
        if (!p->root.rb_node)
                return;
-       spin_lock(&p->lock);
+       mutex_lock(&p->mutex);
        next = rb_first(&p->root);
        while (next) {
                n = rb_entry(next, struct sp_node, nd);
@@ -2327,7 +2322,7 @@ void mpol_free_shared_policy(struct shared_policy *p)
                mpol_put(n->policy);
                kmem_cache_free(sn_cache, n);
        }
-       spin_unlock(&p->lock);
+       mutex_unlock(&p->mutex);
 }
 
 /* assumes fs == KERNEL_DS */