From mboxrd@z Thu Jan 1 00:00:00 1970 From: Eric Dumazet Subject: Re: Suspicious RCU usage in bridge with Linux v4.0-9362-g1fc149933fd4 Date: Mon, 11 May 2015 07:53:06 -0700 Message-ID: <1431355986.566.18.camel@edumazet-glaptop2.roam.corp.google.com> References: <20150504133943.GA17043@x131e> <20150504132714.55dca5b0@urahara> <20150511131516.GC10375@x131e> Mime-Version: 1.0 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: 7bit Cc: netdev@vger.kernel.org To: Dominick Grift Return-path: Received: from mail-ig0-f171.google.com ([209.85.213.171]:33124 "EHLO mail-ig0-f171.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751184AbbEKOxI (ORCPT ); Mon, 11 May 2015 10:53:08 -0400 Received: by igbpi8 with SMTP id pi8so73537995igb.0 for ; Mon, 11 May 2015 07:53:08 -0700 (PDT) In-Reply-To: <20150511131516.GC10375@x131e> Sender: netdev-owner@vger.kernel.org List-ID: On Mon, 2015-05-11 at 15:15 +0200, Dominick Grift wrote: > On Mon, May 04, 2015 at 02:35:27PM -0700, Cong Wang wrote: > > > > > It _looks like_ we only retrieve vlan info to fill netlink > > messages in timer context, so it doesn't seem we need to > > hold br->lock here. > > > > But I never look into br vlan code of course. > > This is just a friendly reminder that this issue still exits in: 4.1.0-0.rc2.git3.1 > > Ignore me if you are aware of the above > Hi Dominick Have you tried this patch I sent last monday ? I will submit formally when I get a test result. Thanks diff --git a/net/bridge/br_stp_timer.c b/net/bridge/br_stp_timer.c index 4fcaa67750fda845ad0a180332c4cd96a9524086..7caf7fae2d5b8aa369b924e1c87a47c343fb8954 100644 --- a/net/bridge/br_stp_timer.c +++ b/net/bridge/br_stp_timer.c @@ -97,7 +97,9 @@ static void br_forward_delay_timer_expired(unsigned long arg) netif_carrier_on(br->dev); } br_log_state(p); + rcu_read_lock(); br_ifinfo_notify(RTM_NEWLINK, p); + rcu_read_unlock(); spin_unlock(&br->lock); }