Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Date:   Tue, 16 Oct 2018 23:50:38 -0400
From:   Dave Jones <davej@codemonkey.org.uk>
To:     Cong Wang <xiyou.wangcong@gmail.com>, Meelis Roos <mroos@linux.ee>,
        LKML <linux-kernel@vger.kernel.org>,
        Linux Kernel Network Developers <netdev@vger.kernel.org>
Subject: Re: netconsole warning in 4.19.0-rc7
Message-ID: <20181017035038.czaqown24rnjn2pw@codemonkey.org.uk>
Mail-Followup-To: Dave Jones <davej@codemonkey.org.uk>,
        Cong Wang <xiyou.wangcong@gmail.com>, Meelis Roos <mroos@linux.ee>,
        LKML <linux-kernel@vger.kernel.org>,
        Linux Kernel Network Developers <netdev@vger.kernel.org>
References: <e0611057-1b5b-9990-b9fa-060a9b5eba40@linux.ee>
 <CAM_iQpU0+LVwgMea32KQWf8z6Bp1VTmWOGnWNwNSmV0PZzf_ew@mail.gmail.com>
 <20181017034047.nrhstc3b4wzevnri@codemonkey.org.uk>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20181017034047.nrhstc3b4wzevnri@codemonkey.org.uk>
User-Agent: NeoMutt/20170113 (1.7.2)
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Tue, Oct 16, 2018 at 11:40:47PM -0400, Dave Jones wrote:

 >  > This is exactly what I mentioned in my review here:
 >  > https://marc.info/?l=linux-netdev&m=153816136624679&w=2
 >  > 
 >  > "But irq is disabled here, so not sure if rcu_read_lock_bh()
 >  > could cause trouble... "
 > 
 > Not sure why this didn't show up for me when I was developing that
 > patch, but I can now reproduce this.  The patch below fixes it for
 > me, but I'm not sure if there are still any side-effects.
 > There's also a missed unlock in the error path.

I took another look at that error path. Turns out this is all we need I
think..


diff --git a/net/core/netpoll.c b/net/core/netpoll.c
index de1d1ba92f2d..f9322d5db899 100644
--- a/net/core/netpoll.c
+++ b/net/core/netpoll.c
@@ -318,6 +318,7 @@ void netpoll_send_skb_on_dev(struct netpoll *np, struct sk_buff *skb,
 	npinfo = rcu_dereference_bh(np->dev->npinfo);
 	if (!npinfo || !netif_running(dev) || !netif_device_present(dev)) {
 		dev_kfree_skb_irq(skb);
+		rcu_read_unlock_bh();
 		return;
 	}