[linux.git] / net / bpf / test_run.c

/* Copyright (c) 2017 Facebook
 *
 * This program is free software; you can redistribute it and/or
 * modify it under the terms of version 2 of the GNU General Public
 * License as published by the Free Software Foundation.
 */
#include <linux/bpf.h>
#include <linux/slab.h>
#include <linux/vmalloc.h>
#include <linux/etherdevice.h>
#include <linux/filter.h>
#include <linux/sched/signal.h>
#include <net/sock.h>
#include <net/tcp.h>

static int bpf_test_run(struct bpf_prog *prog, void *ctx, u32 repeat,
			u32 *retval, u32 *time)
{
	struct bpf_cgroup_storage *storage[MAX_BPF_CGROUP_STORAGE_TYPE] = { 0 };
	enum bpf_cgroup_storage_type stype;
	u64 time_start, time_spent = 0;
	int ret = 0;
	u32 i;

	for_each_cgroup_storage_type(stype) {
		storage[stype] = bpf_cgroup_storage_alloc(prog, stype);
		if (IS_ERR(storage[stype])) {
			storage[stype] = NULL;
			for_each_cgroup_storage_type(stype)
				bpf_cgroup_storage_free(storage[stype]);
			return -ENOMEM;
		}
	}

	if (!repeat)
		repeat = 1;

	rcu_read_lock();
	preempt_disable();
	time_start = ktime_get_ns();
	for (i = 0; i < repeat; i++) {
		bpf_cgroup_storage_set(storage);
		*retval = BPF_PROG_RUN(prog, ctx);

		if (signal_pending(current)) {
			ret = -EINTR;
			break;
		}

		if (need_resched()) {
			time_spent += ktime_get_ns() - time_start;
			preempt_enable();
			rcu_read_unlock();

			cond_resched();

			rcu_read_lock();
			preempt_disable();
			time_start = ktime_get_ns();
		}
	}
	time_spent += ktime_get_ns() - time_start;
	preempt_enable();
	rcu_read_unlock();

	do_div(time_spent, repeat);
	*time = time_spent > U32_MAX ? U32_MAX : (u32)time_spent;

	for_each_cgroup_storage_type(stype)
		bpf_cgroup_storage_free(storage[stype]);

	return ret;
}

static int bpf_test_finish(const union bpf_attr *kattr,
			   union bpf_attr __user *uattr, const void *data,
			   u32 size, u32 retval, u32 duration)
{
	void __user *data_out = u64_to_user_ptr(kattr->test.data_out);
	int err = -EFAULT;
	u32 copy_size = size;

	/* Clamp copy if the user has provided a size hint, but copy the full
	 * buffer if not to retain old behaviour.
	 */
	if (kattr->test.data_size_out &&
	    copy_size > kattr->test.data_size_out) {
		copy_size = kattr->test.data_size_out;
		err = -ENOSPC;
	}

	if (data_out && copy_to_user(data_out, data, copy_size))
		goto out;
	if (copy_to_user(&uattr->test.data_size_out, &size, sizeof(size)))
		goto out;
	if (copy_to_user(&uattr->test.retval, &retval, sizeof(retval)))
		goto out;
	if (copy_to_user(&uattr->test.duration, &duration, sizeof(duration)))
		goto out;
	if (err != -ENOSPC)
		err = 0;
out:
	return err;
}

static void *bpf_test_init(const union bpf_attr *kattr, u32 size,
			   u32 headroom, u32 tailroom)
{
	void __user *data_in = u64_to_user_ptr(kattr->test.data_in);
	void *data;

	if (size < ETH_HLEN || size > PAGE_SIZE - headroom - tailroom)
		return ERR_PTR(-EINVAL);

	data = kzalloc(size + headroom + tailroom, GFP_USER);
	if (!data)
		return ERR_PTR(-ENOMEM);

	if (copy_from_user(data + headroom, data_in, size)) {
		kfree(data);
		return ERR_PTR(-EFAULT);
	}
	return data;
}

int bpf_prog_test_run_skb(struct bpf_prog *prog, const union bpf_attr *kattr,
			  union bpf_attr __user *uattr)
{
	bool is_l2 = false, is_direct_pkt_access = false;
	u32 size = kattr->test.data_size_in;
	u32 repeat = kattr->test.repeat;
	u32 retval, duration;
	int hh_len = ETH_HLEN;
	struct sk_buff *skb;
	struct sock *sk;
	void *data;
	int ret;

	data = bpf_test_init(kattr, size, NET_SKB_PAD + NET_IP_ALIGN,
			     SKB_DATA_ALIGN(sizeof(struct skb_shared_info)));
	if (IS_ERR(data))
		return PTR_ERR(data);

	switch (prog->type) {
	case BPF_PROG_TYPE_SCHED_CLS:
	case BPF_PROG_TYPE_SCHED_ACT:
		is_l2 = true;
		/* fall through */
	case BPF_PROG_TYPE_LWT_IN:
	case BPF_PROG_TYPE_LWT_OUT:
	case BPF_PROG_TYPE_LWT_XMIT:
		is_direct_pkt_access = true;
		break;
	default:
		break;
	}

	sk = kzalloc(sizeof(struct sock), GFP_USER);
	if (!sk) {
		kfree(data);
		return -ENOMEM;
	}
	sock_net_set(sk, current->nsproxy->net_ns);
	sock_init_data(NULL, sk);

	skb = build_skb(data, 0);
	if (!skb) {
		kfree(data);
		kfree(sk);
		return -ENOMEM;
	}
	skb->sk = sk;

	skb_reserve(skb, NET_SKB_PAD + NET_IP_ALIGN);
	__skb_put(skb, size);
	skb->protocol = eth_type_trans(skb, current->nsproxy->net_ns->loopback_dev);
	skb_reset_network_header(skb);

	if (is_l2)
		__skb_push(skb, hh_len);
	if (is_direct_pkt_access)
		bpf_compute_data_pointers(skb);
	ret = bpf_test_run(prog, skb, repeat, &retval, &duration);
	if (ret) {
		kfree_skb(skb);
		kfree(sk);
		return ret;
	}
	if (!is_l2) {
		if (skb_headroom(skb) < hh_len) {
			int nhead = HH_DATA_ALIGN(hh_len - skb_headroom(skb));

			if (pskb_expand_head(skb, nhead, 0, GFP_USER)) {
				kfree_skb(skb);
				kfree(sk);
				return -ENOMEM;
			}
		}
		memset(__skb_push(skb, hh_len), 0, hh_len);
	}

	size = skb->len;
	/* bpf program can never convert linear skb to non-linear */
	if (WARN_ON_ONCE(skb_is_nonlinear(skb)))
		size = skb_headlen(skb);
	ret = bpf_test_finish(kattr, uattr, skb->data, size, retval, duration);
	kfree_skb(skb);
	kfree(sk);
	return ret;
}

int bpf_prog_test_run_xdp(struct bpf_prog *prog, const union bpf_attr *kattr,
			  union bpf_attr __user *uattr)
{
	u32 size = kattr->test.data_size_in;
	u32 repeat = kattr->test.repeat;
	struct netdev_rx_queue *rxqueue;
	struct xdp_buff xdp = {};
	u32 retval, duration;
	void *data;
	int ret;

	data = bpf_test_init(kattr, size, XDP_PACKET_HEADROOM + NET_IP_ALIGN, 0);
	if (IS_ERR(data))
		return PTR_ERR(data);

	xdp.data_hard_start = data;
	xdp.data = data + XDP_PACKET_HEADROOM + NET_IP_ALIGN;
	xdp.data_meta = xdp.data;
	xdp.data_end = xdp.data + size;

	rxqueue = __netif_get_rx_queue(current->nsproxy->net_ns->loopback_dev, 0);
	xdp.rxq = &rxqueue->xdp_rxq;

	ret = bpf_test_run(prog, &xdp, repeat, &retval, &duration);
	if (ret)
		goto out;
	if (xdp.data != data + XDP_PACKET_HEADROOM + NET_IP_ALIGN ||
	    xdp.data_end != xdp.data + size)
		size = xdp.data_end - xdp.data;
	ret = bpf_test_finish(kattr, uattr, xdp.data, size, retval, duration);
out:
	kfree(data);
	return ret;
}
Commit	Line	Data
1cf1cae9 AS	1	/* Copyright (c) 2017 Facebook
	2	*
	3	* This program is free software; you can redistribute it and/or
	4	* modify it under the terms of version 2 of the GNU General Public
	5	* License as published by the Free Software Foundation.
	6	*/
	7	#include <linux/bpf.h>
	8	#include <linux/slab.h>
	9	#include <linux/vmalloc.h>
	10	#include <linux/etherdevice.h>
	11	#include <linux/filter.h>
	12	#include <linux/sched/signal.h>
2cb494a3 SL	13	#include <net/sock.h>
2cb494a3 SL	14	#include <net/tcp.h>
1cf1cae9	15
df1a2cb7 SF	16	static int bpf_test_run(struct bpf_prog prog, void ctx, u32 repeat,
df1a2cb7 SF	17	u32 retval, u32 time)
1cf1cae9	18	{
8bad74f9 RG	19	struct bpf_cgroup_storage *storage[MAX_BPF_CGROUP_STORAGE_TYPE] = { 0 };
8bad74f9 RG	20	enum bpf_cgroup_storage_type stype;
1cf1cae9	21	u64 time_start, time_spent = 0;
df1a2cb7	22	int ret = 0;
dcb40590	23	u32 i;
1cf1cae9	24
8bad74f9 RG	25	for_each_cgroup_storage_type(stype) {
	26	storage[stype] = bpf_cgroup_storage_alloc(prog, stype);
	27	if (IS_ERR(storage[stype])) {
	28	storage[stype] = NULL;
	29	for_each_cgroup_storage_type(stype)
	30	bpf_cgroup_storage_free(storage[stype]);
	31	return -ENOMEM;
	32	}
	33	}
f42ee093	34
1cf1cae9 AS	35	if (!repeat)
1cf1cae9 AS	36	repeat = 1;
df1a2cb7 SF	37
	38	rcu_read_lock();
	39	preempt_disable();
1cf1cae9 AS	40	time_start = ktime_get_ns();
1cf1cae9 AS	41	for (i = 0; i < repeat; i++) {
df1a2cb7 SF	42	bpf_cgroup_storage_set(storage);
	43	*retval = BPF_PROG_RUN(prog, ctx);
	44
	45	if (signal_pending(current)) {
	46	ret = -EINTR;
	47	break;
	48	}
	49
1cf1cae9	50	if (need_resched()) {
1cf1cae9	51	time_spent += ktime_get_ns() - time_start;
df1a2cb7 SF	52	preempt_enable();
	53	rcu_read_unlock();
	54
1cf1cae9	55	cond_resched();
df1a2cb7 SF	56
	57	rcu_read_lock();
	58	preempt_disable();
1cf1cae9 AS	59	time_start = ktime_get_ns();
	60	}
	61	}
	62	time_spent += ktime_get_ns() - time_start;
df1a2cb7 SF	63	preempt_enable();
	64	rcu_read_unlock();
	65
1cf1cae9 AS	66	do_div(time_spent, repeat);
	67	*time = time_spent > U32_MAX ? U32_MAX : (u32)time_spent;
	68
8bad74f9 RG	69	for_each_cgroup_storage_type(stype)
8bad74f9 RG	70	bpf_cgroup_storage_free(storage[stype]);
f42ee093	71
df1a2cb7	72	return ret;
1cf1cae9 AS	73	}
1cf1cae9 AS	74
78e52272 DM	75	static int bpf_test_finish(const union bpf_attr *kattr,
78e52272 DM	76	union bpf_attr __user uattr, const void data,
1cf1cae9 AS	77	u32 size, u32 retval, u32 duration)
1cf1cae9 AS	78	{
78e52272	79	void __user *data_out = u64_to_user_ptr(kattr->test.data_out);
1cf1cae9	80	int err = -EFAULT;
b5a36b1e	81	u32 copy_size = size;
1cf1cae9	82
b5a36b1e LB	83	/* Clamp copy if the user has provided a size hint, but copy the full
	84	* buffer if not to retain old behaviour.
	85	*/
	86	if (kattr->test.data_size_out &&
	87	copy_size > kattr->test.data_size_out) {
	88	copy_size = kattr->test.data_size_out;
	89	err = -ENOSPC;
	90	}
	91
	92	if (data_out && copy_to_user(data_out, data, copy_size))
1cf1cae9 AS	93	goto out;
	94	if (copy_to_user(&uattr->test.data_size_out, &size, sizeof(size)))
	95	goto out;
	96	if (copy_to_user(&uattr->test.retval, &retval, sizeof(retval)))
	97	goto out;
	98	if (copy_to_user(&uattr->test.duration, &duration, sizeof(duration)))
	99	goto out;
b5a36b1e LB	100	if (err != -ENOSPC)
b5a36b1e LB	101	err = 0;
1cf1cae9 AS	102	out:
	103	return err;
	104	}
	105
	106	static void bpf_test_init(const union bpf_attr kattr, u32 size,
	107	u32 headroom, u32 tailroom)
	108	{
	109	void __user *data_in = u64_to_user_ptr(kattr->test.data_in);
	110	void *data;
	111
	112	if (size < ETH_HLEN \|\| size > PAGE_SIZE - headroom - tailroom)
	113	return ERR_PTR(-EINVAL);
	114
	115	data = kzalloc(size + headroom + tailroom, GFP_USER);
	116	if (!data)
	117	return ERR_PTR(-ENOMEM);
	118
	119	if (copy_from_user(data + headroom, data_in, size)) {
	120	kfree(data);
	121	return ERR_PTR(-EFAULT);
	122	}
	123	return data;
	124	}
	125
	126	int bpf_prog_test_run_skb(struct bpf_prog prog, const union bpf_attr kattr,
	127	union bpf_attr __user *uattr)
	128	{
	129	bool is_l2 = false, is_direct_pkt_access = false;
	130	u32 size = kattr->test.data_size_in;
	131	u32 repeat = kattr->test.repeat;
	132	u32 retval, duration;
6e6fddc7	133	int hh_len = ETH_HLEN;
1cf1cae9	134	struct sk_buff *skb;
2cb494a3	135	struct sock *sk;
1cf1cae9 AS	136	void *data;
	137	int ret;
	138
586f8525	139	data = bpf_test_init(kattr, size, NET_SKB_PAD + NET_IP_ALIGN,
1cf1cae9 AS	140	SKB_DATA_ALIGN(sizeof(struct skb_shared_info)));
	141	if (IS_ERR(data))
	142	return PTR_ERR(data);
	143
	144	switch (prog->type) {
	145	case BPF_PROG_TYPE_SCHED_CLS:
	146	case BPF_PROG_TYPE_SCHED_ACT:
	147	is_l2 = true;
	148	/* fall through */
	149	case BPF_PROG_TYPE_LWT_IN:
	150	case BPF_PROG_TYPE_LWT_OUT:
	151	case BPF_PROG_TYPE_LWT_XMIT:
	152	is_direct_pkt_access = true;
	153	break;
	154	default:
	155	break;
	156	}
	157
2cb494a3 SL	158	sk = kzalloc(sizeof(struct sock), GFP_USER);
	159	if (!sk) {
	160	kfree(data);
	161	return -ENOMEM;
	162	}
	163	sock_net_set(sk, current->nsproxy->net_ns);
	164	sock_init_data(NULL, sk);
	165
1cf1cae9 AS	166	skb = build_skb(data, 0);
	167	if (!skb) {
	168	kfree(data);
2cb494a3	169	kfree(sk);
1cf1cae9 AS	170	return -ENOMEM;
1cf1cae9 AS	171	}
2cb494a3	172	skb->sk = sk;
1cf1cae9	173
586f8525	174	skb_reserve(skb, NET_SKB_PAD + NET_IP_ALIGN);
1cf1cae9 AS	175	__skb_put(skb, size);
	176	skb->protocol = eth_type_trans(skb, current->nsproxy->net_ns->loopback_dev);
	177	skb_reset_network_header(skb);
	178
	179	if (is_l2)
6e6fddc7	180	__skb_push(skb, hh_len);
1cf1cae9	181	if (is_direct_pkt_access)
6aaae2b6	182	bpf_compute_data_pointers(skb);
dcb40590 RG	183	ret = bpf_test_run(prog, skb, repeat, &retval, &duration);
	184	if (ret) {
	185	kfree_skb(skb);
	186	kfree(sk);
	187	return ret;
	188	}
6e6fddc7 DB	189	if (!is_l2) {
	190	if (skb_headroom(skb) < hh_len) {
	191	int nhead = HH_DATA_ALIGN(hh_len - skb_headroom(skb));
	192
	193	if (pskb_expand_head(skb, nhead, 0, GFP_USER)) {
	194	kfree_skb(skb);
2cb494a3	195	kfree(sk);
6e6fddc7 DB	196	return -ENOMEM;
	197	}
	198	}
	199	memset(__skb_push(skb, hh_len), 0, hh_len);
	200	}
	201
1cf1cae9 AS	202	size = skb->len;
	203	/* bpf program can never convert linear skb to non-linear */
	204	if (WARN_ON_ONCE(skb_is_nonlinear(skb)))
	205	size = skb_headlen(skb);
78e52272	206	ret = bpf_test_finish(kattr, uattr, skb->data, size, retval, duration);
1cf1cae9	207	kfree_skb(skb);
2cb494a3	208	kfree(sk);
1cf1cae9 AS	209	return ret;
	210	}
	211
	212	int bpf_prog_test_run_xdp(struct bpf_prog prog, const union bpf_attr kattr,
	213	union bpf_attr __user *uattr)
	214	{
	215	u32 size = kattr->test.data_size_in;
	216	u32 repeat = kattr->test.repeat;
65073a67	217	struct netdev_rx_queue *rxqueue;
1cf1cae9 AS	218	struct xdp_buff xdp = {};
	219	u32 retval, duration;
	220	void *data;
	221	int ret;
	222
586f8525	223	data = bpf_test_init(kattr, size, XDP_PACKET_HEADROOM + NET_IP_ALIGN, 0);
1cf1cae9 AS	224	if (IS_ERR(data))
	225	return PTR_ERR(data);
	226
	227	xdp.data_hard_start = data;
586f8525	228	xdp.data = data + XDP_PACKET_HEADROOM + NET_IP_ALIGN;
de8f3a83	229	xdp.data_meta = xdp.data;
1cf1cae9 AS	230	xdp.data_end = xdp.data + size;
1cf1cae9 AS	231
65073a67 DB	232	rxqueue = __netif_get_rx_queue(current->nsproxy->net_ns->loopback_dev, 0);
	233	xdp.rxq = &rxqueue->xdp_rxq;
	234
dcb40590 RG	235	ret = bpf_test_run(prog, &xdp, repeat, &retval, &duration);
	236	if (ret)
	237	goto out;
587b80cc NS	238	if (xdp.data != data + XDP_PACKET_HEADROOM + NET_IP_ALIGN \|\|
587b80cc NS	239	xdp.data_end != xdp.data + size)
1cf1cae9	240	size = xdp.data_end - xdp.data;
78e52272	241	ret = bpf_test_finish(kattr, uattr, xdp.data, size, retval, duration);
dcb40590	242	out:
1cf1cae9 AS	243	kfree(data);
	244	return ret;
	245	}