lists.openwall.net   lists  /  announce  owl-users  owl-dev  john-users  john-dev  passwdqc-users  yescrypt  popa3d-users  /  oss-security  kernel-hardening  musl  sabotage  tlsify  passwords  /  crypt-dev  xvendor  /  Bugtraq  Full-Disclosure  linux-kernel  linux-netdev  linux-ext4  linux-hardening  linux-cve-announce  PHC 
Open Source and information security mailing list archives
 
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-Id: <20190827072743.798323282@linuxfoundation.org>
Date:   Tue, 27 Aug 2019 09:51:14 +0200
From:   Greg Kroah-Hartman <gregkh@...uxfoundation.org>
To:     linux-kernel@...r.kernel.org
Cc:     Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
        stable@...r.kernel.org, Henry Burns <henryburns@...gle.com>,
        Vitaly Wool <vitalywool@...il.com>,
        Shakeel Butt <shakeelb@...gle.com>,
        Jonathan Adams <jwadams@...gle.com>,
        Henry Burns <henrywolfeburns@...il.com>,
        Andrew Morton <akpm@...ux-foundation.org>,
        Linus Torvalds <torvalds@...ux-foundation.org>
Subject: [PATCH 5.2 146/162] mm/z3fold.c: fix race between migration and destruction

From: Henry Burns <henryburns@...gle.com>

commit d776aaa9895eb6eb770908e899cb7f5bd5025b3c upstream.

In z3fold_destroy_pool() we call destroy_workqueue(&pool->compact_wq).
However, we have no guarantee that migration isn't happening in the
background at that time.

Migration directly calls queue_work_on(pool->compact_wq), if destruction
wins that race we are using a destroyed workqueue.

Link: http://lkml.kernel.org/r/20190809213828.202833-1-henryburns@google.com
Signed-off-by: Henry Burns <henryburns@...gle.com>
Cc: Vitaly Wool <vitalywool@...il.com>
Cc: Shakeel Butt <shakeelb@...gle.com>
Cc: Jonathan Adams <jwadams@...gle.com>
Cc: Henry Burns <henrywolfeburns@...il.com>
Cc: <stable@...r.kernel.org>
Signed-off-by: Andrew Morton <akpm@...ux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@...ux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@...uxfoundation.org>

---
 mm/z3fold.c |   89 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 89 insertions(+)

--- a/mm/z3fold.c
+++ b/mm/z3fold.c
@@ -41,6 +41,7 @@
 #include <linux/workqueue.h>
 #include <linux/slab.h>
 #include <linux/spinlock.h>
+#include <linux/wait.h>
 #include <linux/zpool.h>
 
 /*
@@ -144,6 +145,8 @@ struct z3fold_header {
  * @release_wq:	workqueue for safe page release
  * @work:	work_struct for safe page release
  * @inode:	inode for z3fold pseudo filesystem
+ * @destroying: bool to stop migration once we start destruction
+ * @isolated: int to count the number of pages currently in isolation
  *
  * This structure is allocated at pool creation time and maintains metadata
  * pertaining to a particular z3fold pool.
@@ -162,8 +165,11 @@ struct z3fold_pool {
 	const struct zpool_ops *zpool_ops;
 	struct workqueue_struct *compact_wq;
 	struct workqueue_struct *release_wq;
+	struct wait_queue_head isolate_wait;
 	struct work_struct work;
 	struct inode *inode;
+	bool destroying;
+	int isolated;
 };
 
 /*
@@ -771,6 +777,7 @@ static struct z3fold_pool *z3fold_create
 		goto out_c;
 	spin_lock_init(&pool->lock);
 	spin_lock_init(&pool->stale_lock);
+	init_waitqueue_head(&pool->isolate_wait);
 	pool->unbuddied = __alloc_percpu(sizeof(struct list_head)*NCHUNKS, 2);
 	if (!pool->unbuddied)
 		goto out_pool;
@@ -810,6 +817,15 @@ out:
 	return NULL;
 }
 
+static bool pool_isolated_are_drained(struct z3fold_pool *pool)
+{
+	bool ret;
+
+	spin_lock(&pool->lock);
+	ret = pool->isolated == 0;
+	spin_unlock(&pool->lock);
+	return ret;
+}
 /**
  * z3fold_destroy_pool() - destroys an existing z3fold pool
  * @pool:	the z3fold pool to be destroyed
@@ -819,6 +835,22 @@ out:
 static void z3fold_destroy_pool(struct z3fold_pool *pool)
 {
 	kmem_cache_destroy(pool->c_handle);
+	/*
+	 * We set pool-> destroying under lock to ensure that
+	 * z3fold_page_isolate() sees any changes to destroying. This way we
+	 * avoid the need for any memory barriers.
+	 */
+
+	spin_lock(&pool->lock);
+	pool->destroying = true;
+	spin_unlock(&pool->lock);
+
+	/*
+	 * We need to ensure that no pages are being migrated while we destroy
+	 * these workqueues, as migration can queue work on either of the
+	 * workqueues.
+	 */
+	wait_event(pool->isolate_wait, !pool_isolated_are_drained(pool));
 
 	/*
 	 * We need to destroy pool->compact_wq before pool->release_wq,
@@ -1309,6 +1341,28 @@ static u64 z3fold_get_pool_size(struct z
 	return atomic64_read(&pool->pages_nr);
 }
 
+/*
+ * z3fold_dec_isolated() expects to be called while pool->lock is held.
+ */
+static void z3fold_dec_isolated(struct z3fold_pool *pool)
+{
+	assert_spin_locked(&pool->lock);
+	VM_BUG_ON(pool->isolated <= 0);
+	pool->isolated--;
+
+	/*
+	 * If we have no more isolated pages, we have to see if
+	 * z3fold_destroy_pool() is waiting for a signal.
+	 */
+	if (pool->isolated == 0 && waitqueue_active(&pool->isolate_wait))
+		wake_up_all(&pool->isolate_wait);
+}
+
+static void z3fold_inc_isolated(struct z3fold_pool *pool)
+{
+	pool->isolated++;
+}
+
 static bool z3fold_page_isolate(struct page *page, isolate_mode_t mode)
 {
 	struct z3fold_header *zhdr;
@@ -1335,6 +1389,33 @@ static bool z3fold_page_isolate(struct p
 		spin_lock(&pool->lock);
 		if (!list_empty(&page->lru))
 			list_del(&page->lru);
+		/*
+		 * We need to check for destruction while holding pool->lock, as
+		 * otherwise destruction could see 0 isolated pages, and
+		 * proceed.
+		 */
+		if (unlikely(pool->destroying)) {
+			spin_unlock(&pool->lock);
+			/*
+			 * If this page isn't stale, somebody else holds a
+			 * reference to it. Let't drop our refcount so that they
+			 * can call the release logic.
+			 */
+			if (unlikely(kref_put(&zhdr->refcount,
+					      release_z3fold_page_locked))) {
+				/*
+				 * If we get here we have kref problems, so we
+				 * should freak out.
+				 */
+				WARN(1, "Z3fold is experiencing kref problems\n");
+				return false;
+			}
+			z3fold_page_unlock(zhdr);
+			return false;
+		}
+
+
+		z3fold_inc_isolated(pool);
 		spin_unlock(&pool->lock);
 		z3fold_page_unlock(zhdr);
 		return true;
@@ -1408,6 +1489,10 @@ static int z3fold_page_migrate(struct ad
 
 	queue_work_on(new_zhdr->cpu, pool->compact_wq, &new_zhdr->work);
 
+	spin_lock(&pool->lock);
+	z3fold_dec_isolated(pool);
+	spin_unlock(&pool->lock);
+
 	page_mapcount_reset(page);
 	unlock_page(page);
 	put_page(page);
@@ -1428,10 +1513,14 @@ static void z3fold_page_putback(struct p
 	INIT_LIST_HEAD(&page->lru);
 	if (kref_put(&zhdr->refcount, release_z3fold_page_locked)) {
 		atomic64_dec(&pool->pages_nr);
+		spin_lock(&pool->lock);
+		z3fold_dec_isolated(pool);
+		spin_unlock(&pool->lock);
 		return;
 	}
 	spin_lock(&pool->lock);
 	list_add(&page->lru, &pool->lru);
+	z3fold_dec_isolated(pool);
 	spin_unlock(&pool->lock);
 	z3fold_page_unlock(zhdr);
 }


Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ