net-device: move lstats in net_device_read_txrx

dev->lstats is notably used from loopback ndo_start_xmit()
and other virtual drivers.

Per cpu stats updates are dirtying per-cpu data,
but the pointer itself is read-only.

Fixes: 43a71cd ("net-device: reorganize net_device fast path variables")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Coco Li <lixiaoyan@google.com>
Cc: Simon Horman <horms@kernel.org>
Reviewed-by: Simon Horman <horms@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

Author: edumazet

Documentation/networking/net_cachelines/net_device.rst

@@ -136,8 +136,8 @@ struct_netpoll_info*                npinfo                  -
 possible_net_t                      nd_net                  -                   read_mostly         (dev_net)napi_busy_loop,tcp_v(4/6)_rcv,ip(v6)_rcv,ip(6)_input,ip(6)_input_finish
 void*                               ml_priv                                                         
 enum_netdev_ml_priv_type            ml_priv_type                                                    
-struct_pcpu_lstats__percpu*         lstats                                                          
-struct_pcpu_sw_netstats__percpu*    tstats                                                          
+struct_pcpu_lstats__percpu*         lstats                  read_mostly                             dev_lstats_add()
+struct_pcpu_sw_netstats__percpu*    tstats                  read_mostly                             dev_sw_netstats_tx_add()
 struct_pcpu_dstats__percpu*         dstats                                                          
 struct_garp_port*                   garp_port                                                       
 struct_mrp_port*                    mrp_port                                                        

include/linux/netdevice.h

@@ -2141,6 +2141,11 @@ struct net_device {
 
 	/* TXRX read-mostly hotpath */
 	__cacheline_group_begin(net_device_read_txrx);
+	union {
+		struct pcpu_lstats __percpu		*lstats;
+		struct pcpu_sw_netstats __percpu	*tstats;
+		struct pcpu_dstats __percpu		*dstats;
+	};
 	unsigned int		flags;
 	unsigned short		hard_header_len;
 	netdev_features_t	features;
@@ -2395,11 +2400,6 @@ struct net_device {
 	enum netdev_ml_priv_type	ml_priv_type;
 
 	enum netdev_stat_type		pcpu_stat_type:8;
-	union {
-		struct pcpu_lstats __percpu		*lstats;
-		struct pcpu_sw_netstats __percpu	*tstats;
-		struct pcpu_dstats __percpu		*dstats;
-	};
 
 #if IS_ENABLED(CONFIG_GARP)
 	struct garp_port __rcu	*garp_port;

net/core/dev.c

@@ -11652,11 +11652,12 @@ static void __init net_dev_struct_check(void)
 	CACHELINE_ASSERT_GROUP_SIZE(struct net_device, net_device_read_tx, 160);
 
 	/* TXRX read-mostly hotpath */
+	CACHELINE_ASSERT_GROUP_MEMBER(struct net_device, net_device_read_txrx, lstats);
 	CACHELINE_ASSERT_GROUP_MEMBER(struct net_device, net_device_read_txrx, flags);
 	CACHELINE_ASSERT_GROUP_MEMBER(struct net_device, net_device_read_txrx, hard_header_len);
 	CACHELINE_ASSERT_GROUP_MEMBER(struct net_device, net_device_read_txrx, features);
 	CACHELINE_ASSERT_GROUP_MEMBER(struct net_device, net_device_read_txrx, ip6_ptr);
-	CACHELINE_ASSERT_GROUP_SIZE(struct net_device, net_device_read_txrx, 30);
+	CACHELINE_ASSERT_GROUP_SIZE(struct net_device, net_device_read_txrx, 38);
 
 	/* RX read-mostly hotpath */
 	CACHELINE_ASSERT_GROUP_MEMBER(struct net_device, net_device_read_rx, ptype_specific);