From ac3d9dd034e565df2c034ab2ca71f0a9f69153c1 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Fri, 21 Sep 2018 15:27:38 -0700 Subject: netpoll: make ndo_poll_controller() optional As diagnosed by Song Liu, ndo_poll_controller() can be very dangerous on loaded hosts, since the cpu calling ndo_poll_controller() might steal all NAPI contexts (for all RX/TX queues of the NIC). This capture can last for unlimited amount of time, since one cpu is generally not able to drain all the queues under load. It seems that all networking drivers that do use NAPI for their TX completions, should not provide a ndo_poll_controller(). NAPI drivers have netpoll support already handled in core networking stack, since netpoll_poll_dev() uses poll_napi(dev) to iterate through registered NAPI contexts for a device. This patch allows netpoll_poll_dev() to process NAPI contexts even for drivers not providing ndo_poll_controller(), allowing for following patches in NAPI drivers. Also we export netpoll_poll_dev() so that it can be called by bonding/team drivers in following patches. Reported-by: Song Liu Signed-off-by: Eric Dumazet Tested-by: Song Liu Signed-off-by: David S. Miller --- include/linux/netpoll.h | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) (limited to 'include/linux/netpoll.h') diff --git a/include/linux/netpoll.h b/include/linux/netpoll.h index 67662d01130a..3ef82d3a78db 100644 --- a/include/linux/netpoll.h +++ b/include/linux/netpoll.h @@ -49,8 +49,9 @@ struct netpoll_info { }; #ifdef CONFIG_NETPOLL -extern void netpoll_poll_disable(struct net_device *dev); -extern void netpoll_poll_enable(struct net_device *dev); +void netpoll_poll_dev(struct net_device *dev); +void netpoll_poll_disable(struct net_device *dev); +void netpoll_poll_enable(struct net_device *dev); #else static inline void netpoll_poll_disable(struct net_device *dev) { return; } static inline void netpoll_poll_enable(struct net_device *dev) { return; } -- cgit v1.2.3 From c9fbd71f73094311b31ee703a918e9e0df502cef Mon Sep 17 00:00:00 2001 From: Debabrata Banerjee Date: Thu, 18 Oct 2018 11:18:26 -0400 Subject: netpoll: allow cleanup to be synchronous This fixes a problem introduced by: commit 2cde6acd49da ("netpoll: Fix __netpoll_rcu_free so that it can hold the rtnl lock") When using netconsole on a bond, __netpoll_cleanup can asynchronously recurse multiple times, each __netpoll_free_async call can result in more __netpoll_free_async's. This means there is now a race between cleanup_work queues on multiple netpoll_info's on multiple devices and the configuration of a new netpoll. For example if a netconsole is set to enable 0, reconfigured, and enable 1 immediately, this netconsole will likely not work. Given the reason for __netpoll_free_async is it can be called when rtnl is not locked, if it is locked, we should be able to execute synchronously. It appears to be locked everywhere it's called from. Generalize the design pattern from the teaming driver for current callers of __netpoll_free_async. CC: Neil Horman CC: "David S. Miller" Signed-off-by: Debabrata Banerjee Signed-off-by: David S. Miller --- include/linux/netpoll.h | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-) (limited to 'include/linux/netpoll.h') diff --git a/include/linux/netpoll.h b/include/linux/netpoll.h index 3ef82d3a78db..676f1ff161a9 100644 --- a/include/linux/netpoll.h +++ b/include/linux/netpoll.h @@ -31,8 +31,6 @@ struct netpoll { bool ipv6; u16 local_port, remote_port; u8 remote_mac[ETH_ALEN]; - - struct work_struct cleanup_work; }; struct netpoll_info { @@ -63,7 +61,7 @@ int netpoll_parse_options(struct netpoll *np, char *opt); int __netpoll_setup(struct netpoll *np, struct net_device *ndev); int netpoll_setup(struct netpoll *np); void __netpoll_cleanup(struct netpoll *np); -void __netpoll_free_async(struct netpoll *np); +void __netpoll_free(struct netpoll *np); void netpoll_cleanup(struct netpoll *np); void netpoll_send_skb_on_dev(struct netpoll *np, struct sk_buff *skb, struct net_device *dev); -- cgit v1.2.3