From: Ying Xue Date: Tue, 10 Dec 2013 06:54:47 +0000 (-0800) Subject: tipc: protect handler_enabled variable with qitem_lock spin lock X-Git-Tag: v5.15~18775^2~25^2 X-Git-Url: http://review.tizen.org/git/?a=commitdiff_plain;h=00ede977098be3296d42d05a4265ec5ec4a28419;p=platform%2Fkernel%2Flinux-starfive.git tipc: protect handler_enabled variable with qitem_lock spin lock 'handler_enabled' is a global flag indicating whether the TIPC signal handling service is enabled or not. The lack of lock protection for this flag incurs a risk for contention, so that a tipc_k_signal() call might queue a signal handler to a destroyed signal queue, with unpredictable results. To correct this, we let the already existing 'qitem_lock' protect the flag, as it already does with the queue itself. This way, we ensure that the flag always is consistent across all cores. Signed-off-by: Ying Xue Reviewed-by: Paul Gortmaker Signed-off-by: Jon Maloy Signed-off-by: David S. Miller --- diff --git a/net/tipc/handler.c b/net/tipc/handler.c index b36f0fcd9bdf..e4bc8a296744 100644 --- a/net/tipc/handler.c +++ b/net/tipc/handler.c @@ -56,12 +56,13 @@ unsigned int tipc_k_signal(Handler routine, unsigned long argument) { struct queue_item *item; + spin_lock_bh(&qitem_lock); if (!handler_enabled) { pr_err("Signal request ignored by handler\n"); + spin_unlock_bh(&qitem_lock); return -ENOPROTOOPT; } - spin_lock_bh(&qitem_lock); item = kmem_cache_alloc(tipc_queue_item_cache, GFP_ATOMIC); if (!item) { pr_err("Signal queue out of memory\n"); @@ -112,10 +113,14 @@ void tipc_handler_stop(void) struct list_head *l, *n; struct queue_item *item; - if (!handler_enabled) + spin_lock_bh(&qitem_lock); + if (!handler_enabled) { + spin_unlock_bh(&qitem_lock); return; - + } handler_enabled = 0; + spin_unlock_bh(&qitem_lock); + tasklet_kill(&tipc_tasklet); spin_lock_bh(&qitem_lock);