[PATCH net-next v3 04/11] net: Export xmit_recursion
From: Pravin B Shelar <hidden>
Date: 2014-07-16 07:32:12
Subsystem:
networking drivers, networking [general], the rest · Maintainers:
Andrew Lunn, "David S. Miller", Eric Dumazet, Jakub Kicinski, Paolo Abeni, Linus Torvalds
Next patch adds re-circulation action to OVS. This is recursive action needs to be limited to safe number recursions to avoid overflowing kernel stack space. This patch export xmit_recursion so that OVS can increment and check the value before executing recirc action. Signed-off-by: Pravin B Shelar <redacted> --- include/linux/netdevice.h | 3 +++ net/core/dev.c | 10 +++++----- 2 files changed, 8 insertions(+), 5 deletions(-)
diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index 3a320db..ca7a1ab 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h@@ -1941,6 +1941,9 @@ static inline struct net_device *first_net_device_rcu(struct net *net) return lh == &net->dev_base_head ? NULL : net_device_entry(lh); } +#define NET_RECURSION_LIMIT 10 +DECLARE_PER_CPU(int, net_xmit_recursion); + int netdev_boot_setup_check(struct net_device *dev); unsigned long netdev_boot_base(const char *prefix, int unit); struct net_device *dev_getbyhwaddr_rcu(struct net *net, unsigned short type,
diff --git a/net/core/dev.c b/net/core/dev.c
index 6e2a2cd..7f36a4f 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c@@ -2804,8 +2804,8 @@ static void skb_update_prio(struct sk_buff *skb) #define skb_update_prio(skb) #endif -static DEFINE_PER_CPU(int, xmit_recursion); -#define RECURSION_LIMIT 10 +DEFINE_PER_CPU(int, net_xmit_recursion); +EXPORT_SYMBOL_GPL(net_xmit_recursion); /** * dev_loopback_xmit - loop back @skb
@@ -2895,15 +2895,15 @@ static int __dev_queue_xmit(struct sk_buff *skb, void *accel_priv) if (txq->xmit_lock_owner != cpu) { - if (__this_cpu_read(xmit_recursion) > RECURSION_LIMIT) + if (__this_cpu_read(net_xmit_recursion) > NET_RECURSION_LIMIT) goto recursion_alert; HARD_TX_LOCK(dev, txq, cpu); if (!netif_xmit_stopped(txq)) { - __this_cpu_inc(xmit_recursion); + __this_cpu_inc(net_xmit_recursion); rc = dev_hard_start_xmit(skb, dev, txq); - __this_cpu_dec(xmit_recursion); + __this_cpu_dec(net_xmit_recursion); if (dev_xmit_complete(rc)) { HARD_TX_UNLOCK(dev, txq); goto out;
--
1.9.3