[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Tue, 10 Oct 2017 08:56:50 -0700
From: "Paul E. McKenney" <paulmck@...ux.vnet.ibm.com>
To: Peter Zijlstra <peterz@...radead.org>
Cc: linux-kernel@...r.kernel.org, mingo@...nel.org,
torvalds@...ux-foundation.org, mark.rutland@....com,
dhowells@...hat.com, linux-arch@...r.kernel.org,
will.deacon@....com, Pablo Neira Ayuso <pablo@...filter.org>,
Jozsef Kadlecsik <kadlec@...ckhole.kfki.hu>,
Florian Westphal <fw@...len.de>,
"David S. Miller" <davem@...emloft.net>,
netfilter-devel@...r.kernel.org, coreteam@...filter.org,
netdev@...r.kernel.org
Subject: Re: [PATCH RFC tip/core/rcu 14/15] netfilter: Remove now-redundant
smp_read_barrier_depends()
On Tue, Oct 10, 2017 at 10:43:34AM +0200, Peter Zijlstra wrote:
> On Mon, Oct 09, 2017 at 05:22:48PM -0700, Paul E. McKenney wrote:
> > READ_ONCE() now implies smp_read_barrier_depends(), which means that
> > the instances in arpt_do_table(), ipt_do_table(), and ip6t_do_table()
> > are now redundant. This commit removes them and adjusts the comments.
>
> Similar to the previous patch, the lack of READ_ONCE() in the original
> code is a pre-existing bug. It would allow the compiler to tear the load
> and observe a composite of two difference pointer values, or reload the
> private pointer and result in table_base and jumpstacl being part of
> different objects.
>
> It would be good to point out this actually fixes a bug in the code.
Assuming that these changes actually fixed something, agreed. ;-)
Thanx, Paul
Powered by blists - more mailing lists