netdev - Re: [1/4] bonding: don't call slave_xxx

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:	Sat, 20 Jul 2013 12:38:01 +0200
From:	Veaceslav Falico <vfalico@...hat.com>
To:	dingtianhong <dingtianhong@...wei.com>
Cc:	Jay Vosburgh <fubar@...ibm.com>,
	Andy Gospodarek <andy@...yhouse.net>,
	"David S. Miller" <davem@...emloft.net>,
	Netdev <netdev@...r.kernel.org>
Subject: Re: [1/4] bonding: don't call slave_xxx_netpoll under spinlocks

On Sat, Jul 20, 2013 at 03:23:47PM +0800, dingtianhong wrote:
>the slave_xxx_netpoll may sleep, so it should't be called under spinlocks.

I don't really see how it may sleep, it was specifically changed to not
sleep actually. However, see below...

>
>the slave point of the bonding will not be changed outside rtnl lock,
>so rtnl lock is enough here.

Yep, as far as I see there's really no need to take the lock, both the
slave list and the netpoll part are always protected by rtnl lock, unless
I'm missing something, and indeed .ndo_netpoll_setup() is always called
under rtnl.

BTW, bond_netpoll_cleanup() has the same problem - maybe you could check if
we can remove the bond->lock from there also and update the patch?

>
>Signed-off-by: Ding Tianhong <dingtianhong@...wei.com>
>
>---
>drivers/net/bonding/bond_main.c | 15 +++------------
> 1 file changed, 3 insertions(+), 12 deletions(-)
>
>diff --git a/drivers/net/bonding/bond_main.c b/drivers/net/bonding/bond_main.c
>index 07f257d4..5eb75ef 100644
>--- a/drivers/net/bonding/bond_main.c
>+++ b/drivers/net/bonding/bond_main.c
>@@ -1249,8 +1249,9 @@ static void bond_poll_controller(struct net_device *bond_dev)
> {
> }
>
>-static void __bond_netpoll_cleanup(struct bonding *bond)
>+static void bond_netpoll_cleanup(struct net_device *bond_dev)
> {
>+	struct bonding *bond = netdev_priv(bond_dev);
> 	struct slave *slave;
> 	int i;
>
>@@ -1258,14 +1259,6 @@ static void __bond_netpoll_cleanup(struct bonding *bond)
> 		if (IS_UP(slave->dev))
> 			slave_disable_netpoll(slave);
> }
>-static void bond_netpoll_cleanup(struct net_device *bond_dev)
>-{
>-	struct bonding *bond = netdev_priv(bond_dev);
>-
>-	read_lock(&bond->lock);
>-	__bond_netpoll_cleanup(bond);
>-	read_unlock(&bond->lock);
>-}
>
> static int bond_netpoll_setup(struct net_device *dev, struct netpoll_info *ni, gfp_t gfp)
> {
>@@ -1273,15 +1266,13 @@ static int bond_netpoll_setup(struct net_device *dev, struct netpoll_info *ni, g
> 	struct slave *slave;
> 	int i, err = 0;
>
>-	read_lock(&bond->lock);
> 	bond_for_each_slave(bond, slave, i) {
> 		err = slave_enable_netpoll(slave);
> 		if (err) {
>-			__bond_netpoll_cleanup(bond);
>+			bond_netpoll_cleanup(dev);
> 			break;
> 		}
> 	}
>-	read_unlock(&bond->lock);
> 	return err;
> }
>
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html