From dea4096bc41a9642039840ced91e585d04883a16 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Thu, 12 May 2011 15:03:32 +0200 Subject: [PATCH] mac80211: remove pointless mesh path timer RCU code The code here to RCU-dereference a pointer that's on the stack is totally pointless, RCU isn't magic (like say Java's weak references are), so the code can't work like whoever wrote it thought it might. Remove it so readers don't get confused. Note that it seems that a bug is there anyway: I don't see any code that cancels the timer when a mesh path struct is destroyed. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville --- net/mac80211/mesh_hwmp.c | 17 +++-------------- 1 file changed, 3 insertions(+), 14 deletions(-) diff --git a/net/mac80211/mesh_hwmp.c b/net/mac80211/mesh_hwmp.c index 849fecd0820e..2aec7c4f357b 100644 --- a/net/mac80211/mesh_hwmp.c +++ b/net/mac80211/mesh_hwmp.c @@ -966,20 +966,11 @@ int mesh_nexthop_lookup(struct sk_buff *skb, void mesh_path_timer(unsigned long data) { - struct ieee80211_sub_if_data *sdata; - struct mesh_path *mpath; - - rcu_read_lock(); - mpath = (struct mesh_path *) data; - mpath = rcu_dereference(mpath); - if (!mpath) - goto endmpathtimer; - sdata = mpath->sdata; + struct mesh_path *mpath = (void *) data; + struct ieee80211_sub_if_data *sdata = mpath->sdata; - if (sdata->local->quiescing) { - rcu_read_unlock(); + if (sdata->local->quiescing) return; - } spin_lock_bh(&mpath->state_lock); if (mpath->flags & MESH_PATH_RESOLVED || @@ -996,8 +987,6 @@ void mesh_path_timer(unsigned long data) } spin_unlock_bh(&mpath->state_lock); -endmpathtimer: - rcu_read_unlock(); } void -- GitLab