[linux-2.6-block.git] / net / bpf / test_run.c

// SPDX-License-Identifier: GPL-2.0-only
/* Copyright (c) 2017 Facebook
 */
#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/bpf_sk_storage.h>
#include <net/sock.h>
#include <net/tcp.h>
#include <linux/error-injection.h>

#define CREATE_TRACE_POINTS
#include <trace/events/bpf_test_run.h>

static int bpf_test_run(struct bpf_prog *prog, void *ctx, u32 repeat,
			u32 *retval, u32 *time, bool xdp)
{
	struct bpf_cgroup_storage *storage[MAX_BPF_CGROUP_STORAGE_TYPE] = { NULL };
	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();
	migrate_disable();
	time_start = ktime_get_ns();
	for (i = 0; i < repeat; i++) {
		bpf_cgroup_storage_set(storage);

		if (xdp)
			*retval = bpf_prog_run_xdp(prog, ctx);
		else
			*retval = BPF_PROG_RUN(prog, ctx);

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

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

			cond_resched();

			rcu_read_lock();
			migrate_disable();
			time_start = ktime_get_ns();
		}
	}
	time_spent += ktime_get_ns() - time_start;
	migrate_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:
	trace_bpf_test_finish(&err);
	return err;
}

/* Integer types of various sizes and pointer combinations cover variety of
 * architecture dependent calling conventions. 7+ can be supported in the
 * future.
 */
__diag_push();
__diag_ignore(GCC, 8, "-Wmissing-prototypes",
	      "Global functions as their definitions will be in vmlinux BTF");
int noinline bpf_fentry_test1(int a)
{
	return a + 1;
}

int noinline bpf_fentry_test2(int a, u64 b)
{
	return a + b;
}

int noinline bpf_fentry_test3(char a, int b, u64 c)
{
	return a + b + c;
}

int noinline bpf_fentry_test4(void *a, char b, int c, u64 d)
{
	return (long)a + b + c + d;
}

int noinline bpf_fentry_test5(u64 a, void *b, short c, int d, u64 e)
{
	return a + (long)b + c + d + e;
}

int noinline bpf_fentry_test6(u64 a, void *b, short c, int d, void *e, u64 f)
{
	return a + (long)b + c + d + (long)e + f;
}

int noinline bpf_modify_return_test(int a, int *b)
{
	*b += 1;
	return a + *b;
}
__diag_pop();

ALLOW_ERROR_INJECTION(bpf_modify_return_test, ERRNO);

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_tracing(struct bpf_prog *prog,
			      const union bpf_attr *kattr,
			      union bpf_attr __user *uattr)
{
	u16 side_effect = 0, ret = 0;
	int b = 2, err = -EFAULT;
	u32 retval = 0;

	switch (prog->expected_attach_type) {
	case BPF_TRACE_FENTRY:
	case BPF_TRACE_FEXIT:
		if (bpf_fentry_test1(1) != 2 ||
		    bpf_fentry_test2(2, 3) != 5 ||
		    bpf_fentry_test3(4, 5, 6) != 15 ||
		    bpf_fentry_test4((void *)7, 8, 9, 10) != 34 ||
		    bpf_fentry_test5(11, (void *)12, 13, 14, 15) != 65 ||
		    bpf_fentry_test6(16, (void *)17, 18, 19, (void *)20, 21) != 111)
			goto out;
		break;
	case BPF_MODIFY_RETURN:
		ret = bpf_modify_return_test(1, &b);
		if (b != 2)
			side_effect = 1;
		break;
	default:
		goto out;
	}

	retval = ((u32)side_effect << 16) | ret;
	if (copy_to_user(&uattr->test.retval, &retval, sizeof(retval)))
		goto out;

	err = 0;
out:
	trace_bpf_test_finish(&err);
	return err;
}

static void *bpf_ctx_init(const union bpf_attr *kattr, u32 max_size)
{
	void __user *data_in = u64_to_user_ptr(kattr->test.ctx_in);
	void __user *data_out = u64_to_user_ptr(kattr->test.ctx_out);
	u32 size = kattr->test.ctx_size_in;
	void *data;
	int err;

	if (!data_in && !data_out)
		return NULL;

	data = kzalloc(max_size, GFP_USER);
	if (!data)
		return ERR_PTR(-ENOMEM);

	if (data_in) {
		err = bpf_check_uarg_tail_zero(data_in, max_size, size);
		if (err) {
			kfree(data);
			return ERR_PTR(err);
		}

		size = min_t(u32, max_size, size);
		if (copy_from_user(data, data_in, size)) {
			kfree(data);
			return ERR_PTR(-EFAULT);
		}
	}
	return data;
}

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

	if (!data || !data_out)
		return 0;

	if (copy_size > kattr->test.ctx_size_out) {
		copy_size = kattr->test.ctx_size_out;
		err = -ENOSPC;
	}

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

/**
 * range_is_zero - test whether buffer is initialized
 * @buf: buffer to check
 * @from: check from this position
 * @to: check up until (excluding) this position
 *
 * This function returns true if the there is a non-zero byte
 * in the buf in the range [from,to).
 */
static inline bool range_is_zero(void *buf, size_t from, size_t to)
{
	return !memchr_inv((u8 *)buf + from, 0, to - from);
}

static int convert___skb_to_skb(struct sk_buff *skb, struct __sk_buff *__skb)
{
	struct qdisc_skb_cb *cb = (struct qdisc_skb_cb *)skb->cb;

	if (!__skb)
		return 0;

	/* make sure the fields we don't use are zeroed */
	if (!range_is_zero(__skb, 0, offsetof(struct __sk_buff, mark)))
		return -EINVAL;

	/* mark is allowed */

	if (!range_is_zero(__skb, offsetofend(struct __sk_buff, mark),
			   offsetof(struct __sk_buff, priority)))
		return -EINVAL;

	/* priority is allowed */

	if (!range_is_zero(__skb, offsetofend(struct __sk_buff, priority),
			   offsetof(struct __sk_buff, cb)))
		return -EINVAL;

	/* cb is allowed */

	if (!range_is_zero(__skb, offsetofend(struct __sk_buff, cb),
			   offsetof(struct __sk_buff, tstamp)))
		return -EINVAL;

	/* tstamp is allowed */
	/* wire_len is allowed */
	/* gso_segs is allowed */

	if (!range_is_zero(__skb, offsetofend(struct __sk_buff, gso_segs),
			   offsetof(struct __sk_buff, gso_size)))
		return -EINVAL;

	/* gso_size is allowed */

	if (!range_is_zero(__skb, offsetofend(struct __sk_buff, gso_size),
			   sizeof(struct __sk_buff)))
		return -EINVAL;

	skb->mark = __skb->mark;
	skb->priority = __skb->priority;
	skb->tstamp = __skb->tstamp;
	memcpy(&cb->data, __skb->cb, QDISC_CB_PRIV_LEN);

	if (__skb->wire_len == 0) {
		cb->pkt_len = skb->len;
	} else {
		if (__skb->wire_len < skb->len ||
		    __skb->wire_len > GSO_MAX_SIZE)
			return -EINVAL;
		cb->pkt_len = __skb->wire_len;
	}

	if (__skb->gso_segs > GSO_MAX_SEGS)
		return -EINVAL;
	skb_shinfo(skb)->gso_segs = __skb->gso_segs;
	skb_shinfo(skb)->gso_size = __skb->gso_size;

	return 0;
}

static void convert_skb_to___skb(struct sk_buff *skb, struct __sk_buff *__skb)
{
	struct qdisc_skb_cb *cb = (struct qdisc_skb_cb *)skb->cb;

	if (!__skb)
		return;

	__skb->mark = skb->mark;
	__skb->priority = skb->priority;
	__skb->tstamp = skb->tstamp;
	memcpy(__skb->cb, &cb->data, QDISC_CB_PRIV_LEN);
	__skb->wire_len = cb->pkt_len;
	__skb->gso_segs = skb_shinfo(skb)->gso_segs;
}

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;
	struct __sk_buff *ctx = NULL;
	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);

	ctx = bpf_ctx_init(kattr, sizeof(struct __sk_buff));
	if (IS_ERR(ctx)) {
		kfree(data);
		return PTR_ERR(ctx);
	}

	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);
		kfree(ctx);
		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(ctx);
		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 = convert___skb_to_skb(skb, ctx);
	if (ret)
		goto out;
	ret = bpf_test_run(prog, skb, repeat, &retval, &duration, false);
	if (ret)
		goto out;
	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)) {
				ret = -ENOMEM;
				goto out;
			}
		}
		memset(__skb_push(skb, hh_len), 0, hh_len);
	}
	convert_skb_to___skb(skb, ctx);

	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);
	if (!ret)
		ret = bpf_ctx_finish(kattr, uattr, ctx,
				     sizeof(struct __sk_buff));
out:
	kfree_skb(skb);
	bpf_sk_storage_free(sk);
	kfree(sk);
	kfree(ctx);
	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;

	if (kattr->test.ctx_in || kattr->test.ctx_out)
		return -EINVAL;

	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;
	bpf_prog_change_xdp(NULL, prog);
	ret = bpf_test_run(prog, &xdp, repeat, &retval, &duration, true);
	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:
	bpf_prog_change_xdp(prog, NULL);
	kfree(data);
	return ret;
}

static int verify_user_bpf_flow_keys(struct bpf_flow_keys *ctx)
{
	/* make sure the fields we don't use are zeroed */
	if (!range_is_zero(ctx, 0, offsetof(struct bpf_flow_keys, flags)))
		return -EINVAL;

	/* flags is allowed */

	if (!range_is_zero(ctx, offsetofend(struct bpf_flow_keys, flags),
			   sizeof(struct bpf_flow_keys)))
		return -EINVAL;

	return 0;
}

int bpf_prog_test_run_flow_dissector(struct bpf_prog *prog,
				     const union bpf_attr *kattr,
				     union bpf_attr __user *uattr)
{
	u32 size = kattr->test.data_size_in;
	struct bpf_flow_dissector ctx = {};
	u32 repeat = kattr->test.repeat;
	struct bpf_flow_keys *user_ctx;
	struct bpf_flow_keys flow_keys;
	u64 time_start, time_spent = 0;
	const struct ethhdr *eth;
	unsigned int flags = 0;
	u32 retval, duration;
	void *data;
	int ret;
	u32 i;

	if (prog->type != BPF_PROG_TYPE_FLOW_DISSECTOR)
		return -EINVAL;

	if (size < ETH_HLEN)
		return -EINVAL;

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

	eth = (struct ethhdr *)data;

	if (!repeat)
		repeat = 1;

	user_ctx = bpf_ctx_init(kattr, sizeof(struct bpf_flow_keys));
	if (IS_ERR(user_ctx)) {
		kfree(data);
		return PTR_ERR(user_ctx);
	}
	if (user_ctx) {
		ret = verify_user_bpf_flow_keys(user_ctx);
		if (ret)
			goto out;
		flags = user_ctx->flags;
	}

	ctx.flow_keys = &flow_keys;
	ctx.data = data;
	ctx.data_end = (__u8 *)data + size;

	rcu_read_lock();
	preempt_disable();
	time_start = ktime_get_ns();
	for (i = 0; i < repeat; i++) {
		retval = bpf_flow_dissect(prog, &ctx, eth->h_proto, ETH_HLEN,
					  size, flags);

		if (signal_pending(current)) {
			preempt_enable();
			rcu_read_unlock();

			ret = -EINTR;
			goto out;
		}

		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);
	duration = time_spent > U32_MAX ? U32_MAX : (u32)time_spent;

	ret = bpf_test_finish(kattr, uattr, &flow_keys, sizeof(flow_keys),
			      retval, duration);
	if (!ret)
		ret = bpf_ctx_finish(kattr, uattr, user_ctx,
				     sizeof(struct bpf_flow_keys));

out:
	kfree(user_ctx);
	kfree(data);
	return ret;
}
Commit	Line	Data
25763b3c	1	// SPDX-License-Identifier: GPL-2.0-only
1cf1cae9	2	/* Copyright (c) 2017 Facebook
1cf1cae9 AS	3	*/
	4	#include <linux/bpf.h>
	5	#include <linux/slab.h>
	6	#include <linux/vmalloc.h>
	7	#include <linux/etherdevice.h>
	8	#include <linux/filter.h>
	9	#include <linux/sched/signal.h>
6ac99e8f	10	#include <net/bpf_sk_storage.h>
2cb494a3 SL	11	#include <net/sock.h>
2cb494a3 SL	12	#include <net/tcp.h>
3d08b6f2	13	#include <linux/error-injection.h>
1cf1cae9	14
e950e843 MM	15	#define CREATE_TRACE_POINTS
	16	#include <trace/events/bpf_test_run.h>
	17
df1a2cb7	18	static int bpf_test_run(struct bpf_prog prog, void ctx, u32 repeat,
f23c4b39	19	u32 retval, u32 time, bool xdp)
1cf1cae9	20	{
71b91a50	21	struct bpf_cgroup_storage *storage[MAX_BPF_CGROUP_STORAGE_TYPE] = { NULL };
8bad74f9	22	enum bpf_cgroup_storage_type stype;
1cf1cae9	23	u64 time_start, time_spent = 0;
df1a2cb7	24	int ret = 0;
dcb40590	25	u32 i;
1cf1cae9	26
8bad74f9 RG	27	for_each_cgroup_storage_type(stype) {
	28	storage[stype] = bpf_cgroup_storage_alloc(prog, stype);
	29	if (IS_ERR(storage[stype])) {
	30	storage[stype] = NULL;
	31	for_each_cgroup_storage_type(stype)
	32	bpf_cgroup_storage_free(storage[stype]);
	33	return -ENOMEM;
	34	}
	35	}
f42ee093	36
1cf1cae9 AS	37	if (!repeat)
1cf1cae9 AS	38	repeat = 1;
df1a2cb7 SF	39
df1a2cb7 SF	40	rcu_read_lock();
6eac7795	41	migrate_disable();
1cf1cae9 AS	42	time_start = ktime_get_ns();
1cf1cae9 AS	43	for (i = 0; i < repeat; i++) {
df1a2cb7	44	bpf_cgroup_storage_set(storage);
f23c4b39 BT	45
	46	if (xdp)
	47	*retval = bpf_prog_run_xdp(prog, ctx);
	48	else
	49	*retval = BPF_PROG_RUN(prog, ctx);
df1a2cb7 SF	50
	51	if (signal_pending(current)) {
	52	ret = -EINTR;
	53	break;
	54	}
	55
1cf1cae9	56	if (need_resched()) {
1cf1cae9	57	time_spent += ktime_get_ns() - time_start;
6eac7795	58	migrate_enable();
df1a2cb7 SF	59	rcu_read_unlock();
df1a2cb7 SF	60
1cf1cae9	61	cond_resched();
df1a2cb7 SF	62
df1a2cb7 SF	63	rcu_read_lock();
6eac7795	64	migrate_disable();
1cf1cae9 AS	65	time_start = ktime_get_ns();
	66	}
	67	}
	68	time_spent += ktime_get_ns() - time_start;
6eac7795	69	migrate_enable();
df1a2cb7 SF	70	rcu_read_unlock();
df1a2cb7 SF	71
1cf1cae9 AS	72	do_div(time_spent, repeat);
	73	*time = time_spent > U32_MAX ? U32_MAX : (u32)time_spent;
	74
8bad74f9 RG	75	for_each_cgroup_storage_type(stype)
8bad74f9 RG	76	bpf_cgroup_storage_free(storage[stype]);
f42ee093	77
df1a2cb7	78	return ret;
1cf1cae9 AS	79	}
1cf1cae9 AS	80
78e52272 DM	81	static int bpf_test_finish(const union bpf_attr *kattr,
78e52272 DM	82	union bpf_attr __user uattr, const void data,
1cf1cae9 AS	83	u32 size, u32 retval, u32 duration)
1cf1cae9 AS	84	{
78e52272	85	void __user *data_out = u64_to_user_ptr(kattr->test.data_out);
1cf1cae9	86	int err = -EFAULT;
b5a36b1e	87	u32 copy_size = size;
1cf1cae9	88
b5a36b1e LB	89	/* Clamp copy if the user has provided a size hint, but copy the full
	90	* buffer if not to retain old behaviour.
	91	*/
	92	if (kattr->test.data_size_out &&
	93	copy_size > kattr->test.data_size_out) {
	94	copy_size = kattr->test.data_size_out;
	95	err = -ENOSPC;
	96	}
	97
	98	if (data_out && copy_to_user(data_out, data, copy_size))
1cf1cae9 AS	99	goto out;
	100	if (copy_to_user(&uattr->test.data_size_out, &size, sizeof(size)))
	101	goto out;
	102	if (copy_to_user(&uattr->test.retval, &retval, sizeof(retval)))
	103	goto out;
	104	if (copy_to_user(&uattr->test.duration, &duration, sizeof(duration)))
	105	goto out;
b5a36b1e LB	106	if (err != -ENOSPC)
b5a36b1e LB	107	err = 0;
1cf1cae9	108	out:
e950e843	109	trace_bpf_test_finish(&err);
1cf1cae9 AS	110	return err;
	111	}
	112
faeb2dce AS	113	/* Integer types of various sizes and pointer combinations cover variety of
	114	* architecture dependent calling conventions. 7+ can be supported in the
	115	* future.
	116	*/
e9ff9d52 JPM	117	__diag_push();
	118	__diag_ignore(GCC, 8, "-Wmissing-prototypes",
	119	"Global functions as their definitions will be in vmlinux BTF");
faeb2dce AS	120	int noinline bpf_fentry_test1(int a)
	121	{
	122	return a + 1;
	123	}
	124
	125	int noinline bpf_fentry_test2(int a, u64 b)
	126	{
	127	return a + b;
	128	}
	129
	130	int noinline bpf_fentry_test3(char a, int b, u64 c)
	131	{
	132	return a + b + c;
	133	}
	134
	135	int noinline bpf_fentry_test4(void *a, char b, int c, u64 d)
	136	{
	137	return (long)a + b + c + d;
	138	}
	139
	140	int noinline bpf_fentry_test5(u64 a, void *b, short c, int d, u64 e)
	141	{
	142	return a + (long)b + c + d + e;
	143	}
	144
	145	int noinline bpf_fentry_test6(u64 a, void b, short c, int d, void e, u64 f)
	146	{
	147	return a + (long)b + c + d + (long)e + f;
	148	}
	149
3d08b6f2 KS	150	int noinline bpf_modify_return_test(int a, int *b)
	151	{
	152	*b += 1;
	153	return a + *b;
	154	}
e9ff9d52	155	__diag_pop();
3d08b6f2 KS	156
	157	ALLOW_ERROR_INJECTION(bpf_modify_return_test, ERRNO);
	158
1cf1cae9 AS	159	static void bpf_test_init(const union bpf_attr kattr, u32 size,
	160	u32 headroom, u32 tailroom)
	161	{
	162	void __user *data_in = u64_to_user_ptr(kattr->test.data_in);
	163	void *data;
	164
	165	if (size < ETH_HLEN \|\| size > PAGE_SIZE - headroom - tailroom)
	166	return ERR_PTR(-EINVAL);
	167
	168	data = kzalloc(size + headroom + tailroom, GFP_USER);
	169	if (!data)
	170	return ERR_PTR(-ENOMEM);
	171
	172	if (copy_from_user(data + headroom, data_in, size)) {
	173	kfree(data);
	174	return ERR_PTR(-EFAULT);
	175	}
da00d2f1	176
1cf1cae9 AS	177	return data;
	178	}
	179
da00d2f1 KS	180	int bpf_prog_test_run_tracing(struct bpf_prog *prog,
	181	const union bpf_attr *kattr,
	182	union bpf_attr __user *uattr)
	183	{
3d08b6f2 KS	184	u16 side_effect = 0, ret = 0;
	185	int b = 2, err = -EFAULT;
	186	u32 retval = 0;
da00d2f1 KS	187
	188	switch (prog->expected_attach_type) {
	189	case BPF_TRACE_FENTRY:
	190	case BPF_TRACE_FEXIT:
	191	if (bpf_fentry_test1(1) != 2 \|\|
	192	bpf_fentry_test2(2, 3) != 5 \|\|
	193	bpf_fentry_test3(4, 5, 6) != 15 \|\|
	194	bpf_fentry_test4((void *)7, 8, 9, 10) != 34 \|\|
	195	bpf_fentry_test5(11, (void *)12, 13, 14, 15) != 65 \|\|
	196	bpf_fentry_test6(16, (void )17, 18, 19, (void )20, 21) != 111)
	197	goto out;
	198	break;
3d08b6f2 KS	199	case BPF_MODIFY_RETURN:
	200	ret = bpf_modify_return_test(1, &b);
	201	if (b != 2)
	202	side_effect = 1;
	203	break;
da00d2f1 KS	204	default:
	205	goto out;
	206	}
	207
3d08b6f2 KS	208	retval = ((u32)side_effect << 16) \| ret;
	209	if (copy_to_user(&uattr->test.retval, &retval, sizeof(retval)))
	210	goto out;
	211
da00d2f1 KS	212	err = 0;
	213	out:
	214	trace_bpf_test_finish(&err);
	215	return err;
	216	}
	217
b0b9395d SF	218	static void bpf_ctx_init(const union bpf_attr kattr, u32 max_size)
	219	{
	220	void __user *data_in = u64_to_user_ptr(kattr->test.ctx_in);
	221	void __user *data_out = u64_to_user_ptr(kattr->test.ctx_out);
	222	u32 size = kattr->test.ctx_size_in;
	223	void *data;
	224	int err;
	225
	226	if (!data_in && !data_out)
	227	return NULL;
	228
	229	data = kzalloc(max_size, GFP_USER);
	230	if (!data)
	231	return ERR_PTR(-ENOMEM);
	232
	233	if (data_in) {
	234	err = bpf_check_uarg_tail_zero(data_in, max_size, size);
	235	if (err) {
	236	kfree(data);
	237	return ERR_PTR(err);
	238	}
	239
	240	size = min_t(u32, max_size, size);
	241	if (copy_from_user(data, data_in, size)) {
	242	kfree(data);
	243	return ERR_PTR(-EFAULT);
	244	}
	245	}
	246	return data;
	247	}
	248
	249	static int bpf_ctx_finish(const union bpf_attr *kattr,
	250	union bpf_attr __user uattr, const void data,
	251	u32 size)
	252	{
	253	void __user *data_out = u64_to_user_ptr(kattr->test.ctx_out);
	254	int err = -EFAULT;
	255	u32 copy_size = size;
	256
	257	if (!data \|\| !data_out)
	258	return 0;
	259
	260	if (copy_size > kattr->test.ctx_size_out) {
	261	copy_size = kattr->test.ctx_size_out;
	262	err = -ENOSPC;
	263	}
	264
	265	if (copy_to_user(data_out, data, copy_size))
	266	goto out;
	267	if (copy_to_user(&uattr->test.ctx_size_out, &size, sizeof(size)))
	268	goto out;
	269	if (err != -ENOSPC)
	270	err = 0;
	271	out:
	272	return err;
	273	}
	274
	275	/**
	276	* range_is_zero - test whether buffer is initialized
	277	* @buf: buffer to check
	278	* @from: check from this position
	279	* @to: check up until (excluding) this position
	280	*
	281	* This function returns true if the there is a non-zero byte
282	* in the buf in the range [from,to).
283	*/
284	static inline bool range_is_zero(void *buf, size_t from, size_t to)
285	{
286	return !memchr_inv((u8 *)buf + from, 0, to - from);
287	}
288
289	static int convert___skb_to_skb(struct sk_buff skb, struct __sk_buff __skb)
290	{
291	struct qdisc_skb_cb cb = (struct qdisc_skb_cb )skb->cb;
292
293	if (!__skb)
294	return 0;
295
296	/* make sure the fields we don't use are zeroed */
6de6c1f8 NS	297	if (!range_is_zero(__skb, 0, offsetof(struct __sk_buff, mark)))
	298	return -EINVAL;
	299
	300	/* mark is allowed */
	301
	302	if (!range_is_zero(__skb, offsetofend(struct __sk_buff, mark),
	303	offsetof(struct __sk_buff, priority)))
b0b9395d SF	304	return -EINVAL;
	305
	306	/* priority is allowed */
	307
b590cb5f	308	if (!range_is_zero(__skb, offsetofend(struct __sk_buff, priority),
b0b9395d SF	309	offsetof(struct __sk_buff, cb)))
	310	return -EINVAL;
	311
	312	/* cb is allowed */
	313
b590cb5f	314	if (!range_is_zero(__skb, offsetofend(struct __sk_buff, cb),
ba940948 SF	315	offsetof(struct __sk_buff, tstamp)))
	316	return -EINVAL;
	317
	318	/* tstamp is allowed */
850a88cc SF	319	/* wire_len is allowed */
850a88cc SF	320	/* gso_segs is allowed */
ba940948	321
850a88cc	322	if (!range_is_zero(__skb, offsetofend(struct __sk_buff, gso_segs),
cf62089b WB	323	offsetof(struct __sk_buff, gso_size)))
	324	return -EINVAL;
	325
	326	/* gso_size is allowed */
	327
	328	if (!range_is_zero(__skb, offsetofend(struct __sk_buff, gso_size),
b0b9395d SF	329	sizeof(struct __sk_buff)))
	330	return -EINVAL;
	331
6de6c1f8	332	skb->mark = __skb->mark;
b0b9395d	333	skb->priority = __skb->priority;
ba940948	334	skb->tstamp = __skb->tstamp;
b0b9395d SF	335	memcpy(&cb->data, __skb->cb, QDISC_CB_PRIV_LEN);
b0b9395d SF	336
850a88cc SF	337	if (__skb->wire_len == 0) {
	338	cb->pkt_len = skb->len;
	339	} else {
	340	if (__skb->wire_len < skb->len \|\|
	341	__skb->wire_len > GSO_MAX_SIZE)
	342	return -EINVAL;
	343	cb->pkt_len = __skb->wire_len;
	344	}
	345
	346	if (__skb->gso_segs > GSO_MAX_SEGS)
	347	return -EINVAL;
	348	skb_shinfo(skb)->gso_segs = __skb->gso_segs;
cf62089b	349	skb_shinfo(skb)->gso_size = __skb->gso_size;
850a88cc	350
b0b9395d SF	351	return 0;
	352	}
	353
	354	static void convert_skb_to___skb(struct sk_buff skb, struct __sk_buff __skb)
	355	{
	356	struct qdisc_skb_cb cb = (struct qdisc_skb_cb )skb->cb;
	357
	358	if (!__skb)
	359	return;
	360
6de6c1f8	361	__skb->mark = skb->mark;
b0b9395d	362	__skb->priority = skb->priority;
ba940948	363	__skb->tstamp = skb->tstamp;
b0b9395d	364	memcpy(__skb->cb, &cb->data, QDISC_CB_PRIV_LEN);
850a88cc SF	365	__skb->wire_len = cb->pkt_len;
850a88cc SF	366	__skb->gso_segs = skb_shinfo(skb)->gso_segs;
b0b9395d SF	367	}
b0b9395d SF	368
1cf1cae9 AS	369	int bpf_prog_test_run_skb(struct bpf_prog prog, const union bpf_attr kattr,
	370	union bpf_attr __user *uattr)
	371	{
	372	bool is_l2 = false, is_direct_pkt_access = false;
	373	u32 size = kattr->test.data_size_in;
	374	u32 repeat = kattr->test.repeat;
b0b9395d	375	struct __sk_buff *ctx = NULL;
1cf1cae9	376	u32 retval, duration;
6e6fddc7	377	int hh_len = ETH_HLEN;
1cf1cae9	378	struct sk_buff *skb;
2cb494a3	379	struct sock *sk;
1cf1cae9 AS	380	void *data;
	381	int ret;
	382
586f8525	383	data = bpf_test_init(kattr, size, NET_SKB_PAD + NET_IP_ALIGN,
1cf1cae9 AS	384	SKB_DATA_ALIGN(sizeof(struct skb_shared_info)));
	385	if (IS_ERR(data))
	386	return PTR_ERR(data);
	387
b0b9395d SF	388	ctx = bpf_ctx_init(kattr, sizeof(struct __sk_buff));
	389	if (IS_ERR(ctx)) {
	390	kfree(data);
	391	return PTR_ERR(ctx);
	392	}
	393
1cf1cae9 AS	394	switch (prog->type) {
	395	case BPF_PROG_TYPE_SCHED_CLS:
	396	case BPF_PROG_TYPE_SCHED_ACT:
	397	is_l2 = true;
	398	/* fall through */
	399	case BPF_PROG_TYPE_LWT_IN:
	400	case BPF_PROG_TYPE_LWT_OUT:
	401	case BPF_PROG_TYPE_LWT_XMIT:
	402	is_direct_pkt_access = true;
	403	break;
	404	default:
	405	break;
	406	}
	407
2cb494a3 SL	408	sk = kzalloc(sizeof(struct sock), GFP_USER);
	409	if (!sk) {
	410	kfree(data);
b0b9395d	411	kfree(ctx);
2cb494a3 SL	412	return -ENOMEM;
	413	}
	414	sock_net_set(sk, current->nsproxy->net_ns);
	415	sock_init_data(NULL, sk);
	416
1cf1cae9 AS	417	skb = build_skb(data, 0);
	418	if (!skb) {
	419	kfree(data);
b0b9395d	420	kfree(ctx);
2cb494a3	421	kfree(sk);
1cf1cae9 AS	422	return -ENOMEM;
1cf1cae9 AS	423	}
2cb494a3	424	skb->sk = sk;
1cf1cae9	425
586f8525	426	skb_reserve(skb, NET_SKB_PAD + NET_IP_ALIGN);
1cf1cae9 AS	427	__skb_put(skb, size);
	428	skb->protocol = eth_type_trans(skb, current->nsproxy->net_ns->loopback_dev);
	429	skb_reset_network_header(skb);
	430
	431	if (is_l2)
6e6fddc7	432	__skb_push(skb, hh_len);
1cf1cae9	433	if (is_direct_pkt_access)
6aaae2b6	434	bpf_compute_data_pointers(skb);
b0b9395d SF	435	ret = convert___skb_to_skb(skb, ctx);
	436	if (ret)
	437	goto out;
f23c4b39	438	ret = bpf_test_run(prog, skb, repeat, &retval, &duration, false);
b0b9395d SF	439	if (ret)
b0b9395d SF	440	goto out;
6e6fddc7 DB	441	if (!is_l2) {
	442	if (skb_headroom(skb) < hh_len) {
	443	int nhead = HH_DATA_ALIGN(hh_len - skb_headroom(skb));
	444
	445	if (pskb_expand_head(skb, nhead, 0, GFP_USER)) {
b0b9395d SF	446	ret = -ENOMEM;
b0b9395d SF	447	goto out;
6e6fddc7 DB	448	}
	449	}
	450	memset(__skb_push(skb, hh_len), 0, hh_len);
	451	}
b0b9395d	452	convert_skb_to___skb(skb, ctx);
6e6fddc7	453
1cf1cae9 AS	454	size = skb->len;
	455	/* bpf program can never convert linear skb to non-linear */
	456	if (WARN_ON_ONCE(skb_is_nonlinear(skb)))
	457	size = skb_headlen(skb);
78e52272	458	ret = bpf_test_finish(kattr, uattr, skb->data, size, retval, duration);
b0b9395d SF	459	if (!ret)
	460	ret = bpf_ctx_finish(kattr, uattr, ctx,
	461	sizeof(struct __sk_buff));
	462	out:
1cf1cae9	463	kfree_skb(skb);
6ac99e8f	464	bpf_sk_storage_free(sk);
2cb494a3	465	kfree(sk);
b0b9395d	466	kfree(ctx);
1cf1cae9 AS	467	return ret;
	468	}
	469
	470	int bpf_prog_test_run_xdp(struct bpf_prog prog, const union bpf_attr kattr,
	471	union bpf_attr __user *uattr)
	472	{
	473	u32 size = kattr->test.data_size_in;
	474	u32 repeat = kattr->test.repeat;
65073a67	475	struct netdev_rx_queue *rxqueue;
1cf1cae9 AS	476	struct xdp_buff xdp = {};
	477	u32 retval, duration;
	478	void *data;
	479	int ret;
	480
947e8b59 SF	481	if (kattr->test.ctx_in \|\| kattr->test.ctx_out)
	482	return -EINVAL;
	483
586f8525	484	data = bpf_test_init(kattr, size, XDP_PACKET_HEADROOM + NET_IP_ALIGN, 0);
1cf1cae9 AS	485	if (IS_ERR(data))
	486	return PTR_ERR(data);
	487
	488	xdp.data_hard_start = data;
586f8525	489	xdp.data = data + XDP_PACKET_HEADROOM + NET_IP_ALIGN;
de8f3a83	490	xdp.data_meta = xdp.data;
1cf1cae9 AS	491	xdp.data_end = xdp.data + size;
1cf1cae9 AS	492
65073a67 DB	493	rxqueue = __netif_get_rx_queue(current->nsproxy->net_ns->loopback_dev, 0);
65073a67 DB	494	xdp.rxq = &rxqueue->xdp_rxq;
f23c4b39 BT	495	bpf_prog_change_xdp(NULL, prog);
f23c4b39 BT	496	ret = bpf_test_run(prog, &xdp, repeat, &retval, &duration, true);
dcb40590 RG	497	if (ret)
dcb40590 RG	498	goto out;
587b80cc NS	499	if (xdp.data != data + XDP_PACKET_HEADROOM + NET_IP_ALIGN \|\|
587b80cc NS	500	xdp.data_end != xdp.data + size)
1cf1cae9	501	size = xdp.data_end - xdp.data;
78e52272	502	ret = bpf_test_finish(kattr, uattr, xdp.data, size, retval, duration);
dcb40590	503	out:
f23c4b39	504	bpf_prog_change_xdp(prog, NULL);
1cf1cae9 AS	505	kfree(data);
	506	return ret;
	507	}
b7a1848e	508
b2ca4e1c SF	509	static int verify_user_bpf_flow_keys(struct bpf_flow_keys *ctx)
	510	{
	511	/* make sure the fields we don't use are zeroed */
	512	if (!range_is_zero(ctx, 0, offsetof(struct bpf_flow_keys, flags)))
	513	return -EINVAL;
	514
	515	/* flags is allowed */
	516
b590cb5f	517	if (!range_is_zero(ctx, offsetofend(struct bpf_flow_keys, flags),
b2ca4e1c SF	518	sizeof(struct bpf_flow_keys)))
	519	return -EINVAL;
	520
	521	return 0;
	522	}
	523
b7a1848e SF	524	int bpf_prog_test_run_flow_dissector(struct bpf_prog *prog,
	525	const union bpf_attr *kattr,
	526	union bpf_attr __user *uattr)
	527	{
	528	u32 size = kattr->test.data_size_in;
7b8a1304	529	struct bpf_flow_dissector ctx = {};
b7a1848e	530	u32 repeat = kattr->test.repeat;
b2ca4e1c	531	struct bpf_flow_keys *user_ctx;
b7a1848e SF	532	struct bpf_flow_keys flow_keys;
b7a1848e SF	533	u64 time_start, time_spent = 0;
7b8a1304	534	const struct ethhdr *eth;
b2ca4e1c	535	unsigned int flags = 0;
b7a1848e	536	u32 retval, duration;
b7a1848e SF	537	void *data;
	538	int ret;
	539	u32 i;
	540
	541	if (prog->type != BPF_PROG_TYPE_FLOW_DISSECTOR)
	542	return -EINVAL;
	543
7b8a1304 SF	544	if (size < ETH_HLEN)
	545	return -EINVAL;
	546
	547	data = bpf_test_init(kattr, size, 0, 0);
b7a1848e SF	548	if (IS_ERR(data))
	549	return PTR_ERR(data);
	550
7b8a1304	551	eth = (struct ethhdr *)data;
b7a1848e	552
b7a1848e SF	553	if (!repeat)
	554	repeat = 1;
	555
b2ca4e1c SF	556	user_ctx = bpf_ctx_init(kattr, sizeof(struct bpf_flow_keys));
	557	if (IS_ERR(user_ctx)) {
	558	kfree(data);
	559	return PTR_ERR(user_ctx);
	560	}
	561	if (user_ctx) {
	562	ret = verify_user_bpf_flow_keys(user_ctx);
	563	if (ret)
	564	goto out;
	565	flags = user_ctx->flags;
	566	}
	567
7b8a1304 SF	568	ctx.flow_keys = &flow_keys;
	569	ctx.data = data;
	570	ctx.data_end = (__u8 *)data + size;
	571
a439184d SF	572	rcu_read_lock();
a439184d SF	573	preempt_disable();
b7a1848e SF	574	time_start = ktime_get_ns();
b7a1848e SF	575	for (i = 0; i < repeat; i++) {
7b8a1304	576	retval = bpf_flow_dissect(prog, &ctx, eth->h_proto, ETH_HLEN,
b2ca4e1c	577	size, flags);
7b8a1304	578
a439184d SF	579	if (signal_pending(current)) {
	580	preempt_enable();
	581	rcu_read_unlock();
	582
	583	ret = -EINTR;
	584	goto out;
	585	}
b7a1848e SF	586
b7a1848e SF	587	if (need_resched()) {
b7a1848e	588	time_spent += ktime_get_ns() - time_start;
a439184d SF	589	preempt_enable();
	590	rcu_read_unlock();
	591
b7a1848e	592	cond_resched();
a439184d SF	593
	594	rcu_read_lock();
	595	preempt_disable();
b7a1848e SF	596	time_start = ktime_get_ns();
	597	}
	598	}
	599	time_spent += ktime_get_ns() - time_start;
a439184d SF	600	preempt_enable();
	601	rcu_read_unlock();
	602
b7a1848e SF	603	do_div(time_spent, repeat);
	604	duration = time_spent > U32_MAX ? U32_MAX : (u32)time_spent;
	605
	606	ret = bpf_test_finish(kattr, uattr, &flow_keys, sizeof(flow_keys),
	607	retval, duration);
b2ca4e1c SF	608	if (!ret)
	609	ret = bpf_ctx_finish(kattr, uattr, user_ctx,
	610	sizeof(struct bpf_flow_keys));
b7a1848e	611
a439184d	612	out:
b2ca4e1c	613	kfree(user_ctx);
7b8a1304	614	kfree(data);
b7a1848e SF	615	return ret;
b7a1848e SF	616	}