[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-Id: <1476275641-4697-10-git-send-email-lizf@kernel.org>
Date: Wed, 12 Oct 2016 20:32:06 +0800
From: lizf@...nel.org
To: stable@...r.kernel.org
Cc: linux-kernel@...r.kernel.org, Eric Dumazet <edumazet@...gle.com>,
"David S. Miller" <davem@...emloft.net>,
Zefan Li <lizefan@...wei.com>
Subject: [PATCH 3.4 010/125] net: fix a race in dst_release()
From: Eric Dumazet <edumazet@...gle.com>
3.4.113-rc1 review patch. If anyone has any objections, please let me know.
------------------
commit d69bbf88c8d0b367cf3e3a052f6daadf630ee566 upstream.
Only cpu seeing dst refcount going to 0 can safely
dereference dst->flags.
Otherwise an other cpu might already have freed the dst.
Fixes: 27b75c95f10d ("net: avoid RCU for NOCACHE dst")
Reported-by: Greg Thelen <gthelen@...gle.com>
Signed-off-by: Eric Dumazet <edumazet@...gle.com>
Signed-off-by: David S. Miller <davem@...emloft.net>
[lizf: Backported to 3.4: adjust context]
Signed-off-by: Zefan Li <lizefan@...wei.com>
---
net/core/dst.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/net/core/dst.c b/net/core/dst.c
index 43d94ce..54ba1eb 100644
--- a/net/core/dst.c
+++ b/net/core/dst.c
@@ -272,7 +272,7 @@ void dst_release(struct dst_entry *dst)
newrefcnt = atomic_dec_return(&dst->__refcnt);
WARN_ON(newrefcnt < 0);
- if (unlikely(dst->flags & DST_NOCACHE) && !newrefcnt) {
+ if (!newrefcnt && unlikely(dst->flags & DST_NOCACHE)) {
dst = dst_destroy(dst);
if (dst)
__dst_free(dst);
--
1.9.1
Powered by blists - more mailing lists