From mboxrd@z Thu Jan 1 00:00:00 1970 From: Richard Guy Briggs Subject: Re: [Patch net-next] audit: remove useless synchronize_net() Date: Wed, 30 Nov 2016 04:16:43 -0500 Message-ID: <20161130091643.GA32562@madcap2.tricolour.ca> References: <1480439696-21818-1-git-send-email-xiyou.wangcong@gmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Content-Disposition: inline In-Reply-To: <1480439696-21818-1-git-send-email-xiyou.wangcong@gmail.com> Sender: netdev-owner@vger.kernel.org To: Cong Wang Cc: netdev@vger.kernel.org, linux-audit@redhat.com, pmoore@redhat.com List-Id: linux-audit@redhat.com On 2016-11-29 09:14, Cong Wang wrote: > netlink kernel socket is protected by refcount, not RCU. > Its rcv path is neither protected by RCU. So the synchronize_net() > is just pointless. If I understand correctly, xfrm_user_net_exit() usage of RCU_INIT_POINTER() and synchronize_net() is similarly pointless? Also net/phonet/socket.c? I probably modelled things based on the former... > Cc: Richard Guy Briggs > Signed-off-by: Cong Wang > --- > kernel/audit.c | 3 +-- > 1 file changed, 1 insertion(+), 2 deletions(-) > > diff --git a/kernel/audit.c b/kernel/audit.c > index 92c463d..67b9fbd8 100644 > --- a/kernel/audit.c > +++ b/kernel/audit.c > @@ -1172,9 +1172,8 @@ static void __net_exit audit_net_exit(struct net *net) > audit_sock = NULL; > } > > - RCU_INIT_POINTER(aunet->nlsk, NULL); > - synchronize_net(); > netlink_kernel_release(sock); > + aunet->nlsk = NULL; > } > > static struct pernet_operations audit_net_ops __net_initdata = { > -- > 2.1.0 > - RGB