summaryrefslogtreecommitdiff
path: root/net/core/gro_cells.c
blob: ff8e5b64bf6b76451a69e3eae132b593c60ee204 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
// SPDX-License-Identifier: GPL-2.0
#include <linux/skbuff.h>
#include <linux/slab.h>
#include <linux/netdevice.h>
#include <net/gro_cells.h>
#include <net/hotdata.h>

struct gro_cell {
	struct sk_buff_head	napi_skbs;
	struct napi_struct	napi;
};

int gro_cells_receive(struct gro_cells *gcells, struct sk_buff *skb)
{
	struct net_device *dev = skb->dev;
	struct gro_cell *cell;
	int res;

	rcu_read_lock();
	if (unlikely(!(dev->flags & IFF_UP)))
		goto drop;

	if (!gcells->cells || skb_cloned(skb) || netif_elide_gro(dev)) {
		res = netif_rx(skb);
		goto unlock;
	}

	cell = this_cpu_ptr(gcells->cells);

	if (skb_queue_len(&cell->napi_skbs) > READ_ONCE(net_hotdata.max_backlog)) {
drop:
		dev_core_stats_rx_dropped_inc(dev);
		kfree_skb(skb);
		res = NET_RX_DROP;
		goto unlock;
	}

	__skb_queue_tail(&cell->napi_skbs, skb);
	if (skb_queue_len(&cell->napi_skbs) == 1)
		napi_schedule(&cell->napi);

	res = NET_RX_SUCCESS;

unlock:
	rcu_read_unlock();
	return res;
}
EXPORT_SYMBOL(gro_cells_receive);

/* called under BH context */
static int gro_cell_poll(struct napi_struct *napi, int budget)
{
	struct gro_cell *cell = container_of(napi, struct gro_cell, napi);
	struct sk_buff *skb;
	int work_done = 0;

	while (work_done < budget) {
		skb = __skb_dequeue(&cell->napi_skbs);
		if (!skb)
			break;
		napi_gro_receive(napi, skb);
		work_done++;
	}

	if (work_done < budget)
		napi_complete_done(napi, work_done);
	return work_done;
}

int gro_cells_init(struct gro_cells *gcells, struct net_device *dev)
{
	int i;

	gcells->cells = alloc_percpu(struct gro_cell);
	if (!gcells->cells)
		return -ENOMEM;

	for_each_possible_cpu(i) {
		struct gro_cell *cell = per_cpu_ptr(gcells->cells, i);

		__skb_queue_head_init(&cell->napi_skbs);

		set_bit(NAPI_STATE_NO_BUSY_POLL, &cell->napi.state);

		netif_napi_add(dev, &cell->napi, gro_cell_poll);
		napi_enable(&cell->napi);
	}
	return 0;
}
EXPORT_SYMBOL(gro_cells_init);

struct percpu_free_defer {
	struct rcu_head rcu;
	void __percpu	*ptr;
};

static void percpu_free_defer_callback(struct rcu_head *head)
{
	struct percpu_free_defer *defer;

	defer = container_of(head, struct percpu_free_defer, rcu);
	free_percpu(defer->ptr);
	kfree(defer);
}

void gro_cells_destroy(struct gro_cells *gcells)
{
	struct percpu_free_defer *defer;
	int i;

	if (!gcells->cells)
		return;
	for_each_possible_cpu(i) {
		struct gro_cell *cell = per_cpu_ptr(gcells->cells, i);

		napi_disable(&cell->napi);
		__netif_napi_del(&cell->napi);
		__skb_queue_purge(&cell->napi_skbs);
	}
	/* We need to observe an rcu grace period before freeing ->cells,
	 * because netpoll could access dev->napi_list under rcu protection.
	 * Try hard using call_rcu() instead of synchronize_rcu(),
	 * because we might be called from cleanup_net(), and we
	 * definitely do not want to block this critical task.
	 */
	defer = kmalloc(sizeof(*defer), GFP_KERNEL | __GFP_NOWARN);
	if (likely(defer)) {
		defer->ptr = gcells->cells;
		call_rcu(&defer->rcu, percpu_free_defer_callback);
	} else {
		/* We do not hold RTNL at this point, synchronize_net()
		 * would not be able to expedite this sync.
		 */
		synchronize_rcu_expedited();
		free_percpu(gcells->cells);
	}
	gcells->cells = NULL;
}
EXPORT_SYMBOL(gro_cells_destroy);