bonding: restructure locking of bond_ab_arp_probe()

Currently we're calling it from under RCU context, however we're using some
functions that require rtnl to be held.

Fix this by restructuring the locking - don't call it under any locks,
aquire rcu_read_lock() if we're sending _only_ (i.e. we have the active
slave present), and use rtnl locking otherwise - if we need to modify
(in)active flags of a slave.

CC: Jay Vosburgh <fubar@us.ibm.com>
CC: Andy Gospodarek <andy@greyhouse.net>
Signed-off-by: Veaceslav Falico <vfalico@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This commit is contained in:
Veaceslav Falico 2014-01-27 14:37:32 +01:00 committed by David S. Miller
parent 98b90f2665
commit f2ebd477f1

View File

@ -2599,17 +2599,18 @@ do_failover:
/*
* Send ARP probes for active-backup mode ARP monitor.
*
* Called with rcu_read_lock hold.
*/
static void bond_ab_arp_probe(struct bonding *bond)
static bool bond_ab_arp_probe(struct bonding *bond)
{
struct slave *slave, *before = NULL, *new_slave = NULL,
*curr_arp_slave = rcu_dereference(bond->current_arp_slave),
*curr_active_slave = rcu_dereference(bond->curr_active_slave);
*curr_arp_slave, *curr_active_slave;
struct list_head *iter;
bool found = false;
rcu_read_lock();
curr_arp_slave = rcu_dereference(bond->current_arp_slave);
curr_active_slave = rcu_dereference(bond->curr_active_slave);
if (curr_arp_slave && curr_active_slave)
pr_info("PROBE: c_arp %s && cas %s BAD\n",
curr_arp_slave->dev->name,
@ -2617,23 +2618,32 @@ static void bond_ab_arp_probe(struct bonding *bond)
if (curr_active_slave) {
bond_arp_send_all(bond, curr_active_slave);
return;
rcu_read_unlock();
return true;
}
rcu_read_unlock();
/* if we don't have a curr_active_slave, search for the next available
* backup slave from the current_arp_slave and make it the candidate
* for becoming the curr_active_slave
*/
if (!rtnl_trylock())
return false;
/* curr_arp_slave might have gone away */
curr_arp_slave = ACCESS_ONCE(bond->current_arp_slave);
if (!curr_arp_slave) {
curr_arp_slave = bond_first_slave_rcu(bond);
if (!curr_arp_slave)
return;
curr_arp_slave = bond_first_slave(bond);
if (!curr_arp_slave) {
rtnl_unlock();
return true;
}
}
bond_set_slave_inactive_flags(curr_arp_slave);
bond_for_each_slave_rcu(bond, slave, iter) {
bond_for_each_slave(bond, slave, iter) {
if (!found && !before && IS_UP(slave->dev))
before = slave;
@ -2663,21 +2673,26 @@ static void bond_ab_arp_probe(struct bonding *bond)
if (!new_slave && before)
new_slave = before;
if (!new_slave)
return;
if (!new_slave) {
rtnl_unlock();
return true;
}
new_slave->link = BOND_LINK_BACK;
bond_set_slave_active_flags(new_slave);
bond_arp_send_all(bond, new_slave);
new_slave->jiffies = jiffies;
rcu_assign_pointer(bond->current_arp_slave, new_slave);
rtnl_unlock();
return true;
}
static void bond_activebackup_arp_mon(struct work_struct *work)
{
struct bonding *bond = container_of(work, struct bonding,
arp_work.work);
bool should_notify_peers = false;
bool should_notify_peers = false, should_commit = false;
int delta_in_ticks;
delta_in_ticks = msecs_to_jiffies(bond->params.arp_interval);
@ -2686,12 +2701,11 @@ static void bond_activebackup_arp_mon(struct work_struct *work)
goto re_arm;
rcu_read_lock();
should_notify_peers = bond_should_notify_peers(bond);
should_commit = bond_ab_arp_inspect(bond);
rcu_read_unlock();
if (bond_ab_arp_inspect(bond)) {
rcu_read_unlock();
if (should_commit) {
/* Race avoidance with bond_close flush of workqueue */
if (!rtnl_trylock()) {
delta_in_ticks = 1;
@ -2700,13 +2714,14 @@ static void bond_activebackup_arp_mon(struct work_struct *work)
}
bond_ab_arp_commit(bond);
rtnl_unlock();
rcu_read_lock();
}
bond_ab_arp_probe(bond);
rcu_read_unlock();
if (!bond_ab_arp_probe(bond)) {
/* rtnl locking failed, re-arm */
delta_in_ticks = 1;
should_notify_peers = false;
}
re_arm:
if (bond->params.arp_interval)