All of lore.kernel.org
 help / color / mirror / Atom feed
From: Jukka Rissanen <jukka.rissanen@linux.intel.com>
To: Michael Scott <michael.scott@linaro.org>,
	Marcel Holtmann <marcel@holtmann.org>,
	Gustavo Padovan <gustavo@padovan.org>,
	Johan Hedberg <johan.hedberg@gmail.com>
Cc: "David S . Miller" <davem@davemloft.net>,
	linux-bluetooth@vger.kernel.org, linux-wpan@vger.kernel.org,
	netdev@vger.kernel.org, linux-kernel@vger.kernel.org
Subject: Re: [PATCH] bluetooth: 6lowpan: fix delay work init in add_peer_chan()
Date: Fri, 31 Mar 2017 12:37:07 +0300	[thread overview]
Message-ID: <1490953027.4892.14.camel@linux.intel.com> (raw)
In-Reply-To: <20170329061054.4300-1-michael.scott@linaro.org>

Hi Michael,

On Tue, 2017-03-28 at 23:10 -0700, Michael Scott wrote:
> When adding 6lowpan devices very rapidly we sometimes see a crash:
> [23122.306615] CPU: 2 PID: 0 Comm: swapper/2 Not tainted 4.9.0-43-
> arm64 #1 Debian 4.9.9.linaro.43-1
> [23122.315400] Hardware name: HiKey Development Board (DT)
> [23122.320623] task: ffff800075443080 task.stack: ffff800075484000
> [23122.326551] PC is at expire_timers+0x70/0x150
> [23122.330907] LR is at run_timer_softirq+0xa0/0x1a0
> [23122.335616] pc : [<ffff000008142dd8>] lr : [<ffff000008142f58>]
> pstate: 600001c5
> 
> This was due to add_peer_chan() unconditionally initializing the
> lowpan_btle_dev->notify_peers delayed work structure, even if the
> lowpan_btle_dev passed into add_peer_chan() had previously been
> initialized.
> 
> Normally, this would go unnoticed as the delayed work timer is set
> for
> 100 msec, however when calling add_peer_chan() faster than 100 msec
> it
> clears out a previously queued delay work causing the crash above.
> 
> To fix this, let add_peer_chan() know when a new lowpan_btle_dev is
> passed
> in so that it only performs the delay work initialization when
> needed.
> 
> Signed-off-by: Michael Scott <michael.scott@linaro.org>
> ---
>  net/bluetooth/6lowpan.c | 10 +++++++---
>  1 file changed, 7 insertions(+), 3 deletions(-)
> 
> diff --git a/net/bluetooth/6lowpan.c b/net/bluetooth/6lowpan.c
> index e27be3ca0a0c..c282482edc2c 100644
> --- a/net/bluetooth/6lowpan.c
> +++ b/net/bluetooth/6lowpan.c
> @@ -754,7 +754,8 @@ static void set_ip_addr_bits(u8 addr_type, u8
> *addr)
>  }
>  
>  static struct l2cap_chan *add_peer_chan(struct l2cap_chan *chan,
> -					struct lowpan_btle_dev *dev)
> +					struct lowpan_btle_dev *dev,
> +					bool new_netdev)
>  {
>  	struct lowpan_peer *peer;
>  
> @@ -785,7 +786,8 @@ static struct l2cap_chan *add_peer_chan(struct
> l2cap_chan *chan,
>  	spin_unlock(&devices_lock);
>  
>  	/* Notifying peers about us needs to be done without locks
> held */
> -	INIT_DELAYED_WORK(&dev->notify_peers, do_notify_peers);
> +	if (new_netdev)
> +		INIT_DELAYED_WORK(&dev->notify_peers,
> do_notify_peers);
>  	schedule_delayed_work(&dev->notify_peers,
> msecs_to_jiffies(100));
>  
>  	return peer->chan;
> @@ -842,6 +844,7 @@ static int setup_netdev(struct l2cap_chan *chan,
> struct lowpan_btle_dev **dev)
>  static inline void chan_ready_cb(struct l2cap_chan *chan)
>  {
>  	struct lowpan_btle_dev *dev;
> +	bool new_netdev = false;
>  
>  	dev = lookup_dev(chan->conn);
>  
> @@ -852,12 +855,13 @@ static inline void chan_ready_cb(struct
> l2cap_chan *chan)
>  			l2cap_chan_del(chan, -ENOENT);
>  			return;
>  		}
> +		new_netdev = true;
>  	}
>  
>  	if (!try_module_get(THIS_MODULE))
>  		return;
>  
> -	add_peer_chan(chan, dev);
> +	add_peer_chan(chan, dev, new_netdev);
>  	ifup(dev->netdev);
>  }
>  

Good catch!

Acked-by: Jukka Rissanen <jukka.rissanen@linux.intel.com>


Cheers,
Jukka

WARNING: multiple messages have this Message-ID (diff)
From: Jukka Rissanen <jukka.rissanen-VuQAYsv1563Yd54FQh9/CA@public.gmane.org>
To: Michael Scott
	<michael.scott-QSEj5FYQhm4dnm+yROfE0A@public.gmane.org>,
	Marcel Holtmann <marcel-kz+m5ild9QBg9hUCZPvPmw@public.gmane.org>,
	Gustavo Padovan <gustavo-THi1TnShQwVAfugRpC6u6w@public.gmane.org>,
	Johan Hedberg
	<johan.hedberg-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
Cc: "David S . Miller"
	<davem-fT/PcQaiUtIeIZ0/mPfg9Q@public.gmane.org>,
	linux-bluetooth-u79uwXL29TY76Z2rM5mHXA@public.gmane.org,
	linux-wpan-u79uwXL29TY76Z2rM5mHXA@public.gmane.org,
	netdev-u79uwXL29TY76Z2rM5mHXA@public.gmane.org,
	linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
Subject: Re: [PATCH] bluetooth: 6lowpan: fix delay work init in add_peer_chan()
Date: Fri, 31 Mar 2017 12:37:07 +0300	[thread overview]
Message-ID: <1490953027.4892.14.camel@linux.intel.com> (raw)
In-Reply-To: <20170329061054.4300-1-michael.scott-QSEj5FYQhm4dnm+yROfE0A@public.gmane.org>

Hi Michael,

On Tue, 2017-03-28 at 23:10 -0700, Michael Scott wrote:
> When adding 6lowpan devices very rapidly we sometimes see a crash:
> [23122.306615] CPU: 2 PID: 0 Comm: swapper/2 Not tainted 4.9.0-43-
> arm64 #1 Debian 4.9.9.linaro.43-1
> [23122.315400] Hardware name: HiKey Development Board (DT)
> [23122.320623] task: ffff800075443080 task.stack: ffff800075484000
> [23122.326551] PC is at expire_timers+0x70/0x150
> [23122.330907] LR is at run_timer_softirq+0xa0/0x1a0
> [23122.335616] pc : [<ffff000008142dd8>] lr : [<ffff000008142f58>]
> pstate: 600001c5
> 
> This was due to add_peer_chan() unconditionally initializing the
> lowpan_btle_dev->notify_peers delayed work structure, even if the
> lowpan_btle_dev passed into add_peer_chan() had previously been
> initialized.
> 
> Normally, this would go unnoticed as the delayed work timer is set
> for
> 100 msec, however when calling add_peer_chan() faster than 100 msec
> it
> clears out a previously queued delay work causing the crash above.
> 
> To fix this, let add_peer_chan() know when a new lowpan_btle_dev is
> passed
> in so that it only performs the delay work initialization when
> needed.
> 
> Signed-off-by: Michael Scott <michael.scott-QSEj5FYQhm4dnm+yROfE0A@public.gmane.org>
> ---
>  net/bluetooth/6lowpan.c | 10 +++++++---
>  1 file changed, 7 insertions(+), 3 deletions(-)
> 
> diff --git a/net/bluetooth/6lowpan.c b/net/bluetooth/6lowpan.c
> index e27be3ca0a0c..c282482edc2c 100644
> --- a/net/bluetooth/6lowpan.c
> +++ b/net/bluetooth/6lowpan.c
> @@ -754,7 +754,8 @@ static void set_ip_addr_bits(u8 addr_type, u8
> *addr)
>  }
>  
>  static struct l2cap_chan *add_peer_chan(struct l2cap_chan *chan,
> -					struct lowpan_btle_dev *dev)
> +					struct lowpan_btle_dev *dev,
> +					bool new_netdev)
>  {
>  	struct lowpan_peer *peer;
>  
> @@ -785,7 +786,8 @@ static struct l2cap_chan *add_peer_chan(struct
> l2cap_chan *chan,
>  	spin_unlock(&devices_lock);
>  
>  	/* Notifying peers about us needs to be done without locks
> held */
> -	INIT_DELAYED_WORK(&dev->notify_peers, do_notify_peers);
> +	if (new_netdev)
> +		INIT_DELAYED_WORK(&dev->notify_peers,
> do_notify_peers);
>  	schedule_delayed_work(&dev->notify_peers,
> msecs_to_jiffies(100));
>  
>  	return peer->chan;
> @@ -842,6 +844,7 @@ static int setup_netdev(struct l2cap_chan *chan,
> struct lowpan_btle_dev **dev)
>  static inline void chan_ready_cb(struct l2cap_chan *chan)
>  {
>  	struct lowpan_btle_dev *dev;
> +	bool new_netdev = false;
>  
>  	dev = lookup_dev(chan->conn);
>  
> @@ -852,12 +855,13 @@ static inline void chan_ready_cb(struct
> l2cap_chan *chan)
>  			l2cap_chan_del(chan, -ENOENT);
>  			return;
>  		}
> +		new_netdev = true;
>  	}
>  
>  	if (!try_module_get(THIS_MODULE))
>  		return;
>  
> -	add_peer_chan(chan, dev);
> +	add_peer_chan(chan, dev, new_netdev);
>  	ifup(dev->netdev);
>  }
>  

Good catch!

Acked-by: Jukka Rissanen <jukka.rissanen-VuQAYsv1563Yd54FQh9/CA@public.gmane.org>


Cheers,
Jukka

  reply	other threads:[~2017-03-31  9:37 UTC|newest]

Thread overview: 5+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2017-03-29  6:10 [PATCH] bluetooth: 6lowpan: fix delay work init in add_peer_chan() Michael Scott
2017-03-31  9:37 ` Jukka Rissanen [this message]
2017-03-31  9:37   ` Jukka Rissanen
2017-03-31  9:37   ` Jukka Rissanen
2017-03-31 10:10 ` Marcel Holtmann

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=1490953027.4892.14.camel@linux.intel.com \
    --to=jukka.rissanen@linux.intel.com \
    --cc=davem@davemloft.net \
    --cc=gustavo@padovan.org \
    --cc=johan.hedberg@gmail.com \
    --cc=linux-bluetooth@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-wpan@vger.kernel.org \
    --cc=marcel@holtmann.org \
    --cc=michael.scott@linaro.org \
    --cc=netdev@vger.kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.