net_sched: remove tcf_block_put_deferred()

In commit 7aa0045dad ("net_sched: introduce a workqueue for RCU callbacks of tc filter")
I defer tcf_chain_flush() to a workqueue, this causes a use-after-free
because qdisc is already destroyed after we queue this work.

The tcf_block_put_deferred() is no longer necessary after we get RTNL
for each tc filter destroy work, no others could jump in at this point.
Same for tcf_chain_hold(), we are fully serialized now.

This also reduces one indirection therefore makes the code more
readable. Note this brings back a rcu_barrier(), however comparing
to the code prior to commit 7aa0045dad we still reduced one
rcu_barrier(). For net-next, we can consider to refcnt tcf block to
avoid it.

Fixes: 7aa0045dad ("net_sched: introduce a workqueue for RCU callbacks of tc filter")
Cc: Daniel Borkmann <daniel@iogearbox.net>
Cc: Jiri Pirko <jiri@resnulli.us>
Cc: John Fastabend <john.fastabend@gmail.com>
Cc: Jamal Hadi Salim <jhs@mojatatu.com>
Cc: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com>
Cc: Eric Dumazet <edumazet@google.com>
Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This commit is contained in:
Cong Wang 2017-10-30 11:10:09 -07:00 committed by David S. Miller
parent f9e56baf03
commit 822e86d997

View File

@ -280,8 +280,8 @@ static void tcf_block_put_final(struct work_struct *work)
struct tcf_block *block = container_of(work, struct tcf_block, work);
struct tcf_chain *chain, *tmp;
/* At this point, all the chains should have refcnt == 1. */
rtnl_lock();
/* Only chain 0 should be still here. */
list_for_each_entry_safe(chain, tmp, &block->chain_list, list)
tcf_chain_put(chain);
rtnl_unlock();
@ -289,23 +289,17 @@ static void tcf_block_put_final(struct work_struct *work)
}
/* XXX: Standalone actions are not allowed to jump to any chain, and bound
* actions should be all removed after flushing. However, filters are destroyed
* in RCU callbacks, we have to hold the chains first, otherwise we would
* always race with RCU callbacks on this list without proper locking.
* actions should be all removed after flushing. However, filters are now
* destroyed in tc filter workqueue with RTNL lock, they can not race here.
*/
static void tcf_block_put_deferred(struct work_struct *work)
void tcf_block_put(struct tcf_block *block)
{
struct tcf_block *block = container_of(work, struct tcf_block, work);
struct tcf_chain *chain;
struct tcf_chain *chain, *tmp;
rtnl_lock();
/* Hold a refcnt for all chains, except 0, in case they are gone. */
list_for_each_entry(chain, &block->chain_list, list)
if (chain->index)
tcf_chain_hold(chain);
if (!block)
return;
/* No race on the list, because no chain could be destroyed. */
list_for_each_entry(chain, &block->chain_list, list)
list_for_each_entry_safe(chain, tmp, &block->chain_list, list)
tcf_chain_flush(chain);
INIT_WORK(&block->work, tcf_block_put_final);
@ -314,21 +308,6 @@ static void tcf_block_put_deferred(struct work_struct *work)
*/
rcu_barrier();
tcf_queue_work(&block->work);
rtnl_unlock();
}
void tcf_block_put(struct tcf_block *block)
{
if (!block)
return;
INIT_WORK(&block->work, tcf_block_put_deferred);
/* Wait for existing RCU callbacks to cool down, make sure their works
* have been queued before this. We can not flush pending works here
* because we are holding the RTNL lock.
*/
rcu_barrier();
tcf_queue_work(&block->work);
}
EXPORT_SYMBOL(tcf_block_put);