[thirdparty/linux.git] / include / linux / skmsg.h

/* SPDX-License-Identifier: GPL-2.0 */
/* Copyright (c) 2017 - 2018 Covalent IO, Inc. http://covalent.io */

#ifndef _LINUX_SKMSG_H
#define _LINUX_SKMSG_H

#include <linux/bpf.h>
#include <linux/filter.h>
#include <linux/scatterlist.h>
#include <linux/skbuff.h>

#include <net/sock.h>
#include <net/tcp.h>
#include <net/strparser.h>

#define MAX_MSG_FRAGS			MAX_SKB_FRAGS
#define NR_MSG_FRAG_IDS			(MAX_MSG_FRAGS + 1)

enum __sk_action {
	__SK_DROP = 0,
	__SK_PASS,
	__SK_REDIRECT,
	__SK_NONE,
};

struct sk_msg_sg {
	u32				start;
	u32				curr;
	u32				end;
	u32				size;
	u32				copybreak;
	DECLARE_BITMAP(copy, MAX_MSG_FRAGS + 2);
	/* The extra two elements:
	 * 1) used for chaining the front and sections when the list becomes
	 *    partitioned (e.g. end < start). The crypto APIs require the
	 *    chaining;
	 * 2) to chain tailer SG entries after the message.
	 */
	struct scatterlist		data[MAX_MSG_FRAGS + 2];
};

/* UAPI in filter.c depends on struct sk_msg_sg being first element. */
struct sk_msg {
	struct sk_msg_sg		sg;
	void				*data;
	void				*data_end;
	u32				apply_bytes;
	u32				cork_bytes;
	u32				flags;
	struct sk_buff			*skb;
	struct sock			*sk_redir;
	struct sock			*sk;
	struct list_head		list;
};

struct sk_psock_progs {
	struct bpf_prog			*msg_parser;
	struct bpf_prog			*stream_parser;
	struct bpf_prog			*stream_verdict;
	struct bpf_prog			*skb_verdict;
	struct bpf_link			*msg_parser_link;
	struct bpf_link			*stream_parser_link;
	struct bpf_link			*stream_verdict_link;
	struct bpf_link			*skb_verdict_link;
};

enum sk_psock_state_bits {
	SK_PSOCK_TX_ENABLED,
	SK_PSOCK_RX_STRP_ENABLED,
};

struct sk_psock_link {
	struct list_head		list;
	struct bpf_map			*map;
	void				*link_raw;
};

struct sk_psock_work_state {
	u32				len;
	u32				off;
};

struct sk_psock {
	struct sock			*sk;
	struct sock			*sk_redir;
	u32				apply_bytes;
	u32				cork_bytes;
	u32				eval;
	bool				redir_ingress; /* undefined if sk_redir is null */
	struct sk_msg			*cork;
	struct sk_psock_progs		progs;
#if IS_ENABLED(CONFIG_BPF_STREAM_PARSER)
	struct strparser		strp;
	u32				copied_seq;
	u32				ingress_bytes;
#endif
	struct sk_buff_head		ingress_skb;
	struct list_head		ingress_msg;
	spinlock_t			ingress_lock;
	unsigned long			state;
	struct list_head		link;
	spinlock_t			link_lock;
	refcount_t			refcnt;
	void (*saved_unhash)(struct sock *sk);
	void (*saved_destroy)(struct sock *sk);
	void (*saved_close)(struct sock *sk, long timeout);
	void (*saved_write_space)(struct sock *sk);
	void (*saved_data_ready)(struct sock *sk);
	/* psock_update_sk_prot may be called with restore=false many times
	 * so the handler must be safe for this case. It will be called
	 * exactly once with restore=true when the psock is being destroyed
	 * and psock refcnt is zero, but before an RCU grace period.
	 */
	int  (*psock_update_sk_prot)(struct sock *sk, struct sk_psock *psock,
				     bool restore);
	struct proto			*sk_proto;
	struct mutex			work_mutex;
	struct sk_psock_work_state	work_state;
	struct delayed_work		work;
	struct sock			*sk_pair;
	struct rcu_work			rwork;
};

int sk_msg_alloc(struct sock *sk, struct sk_msg *msg, int len,
		 int elem_first_coalesce);
int sk_msg_clone(struct sock *sk, struct sk_msg *dst, struct sk_msg *src,
		 u32 off, u32 len);
void sk_msg_trim(struct sock *sk, struct sk_msg *msg, int len);
int sk_msg_free(struct sock *sk, struct sk_msg *msg);
int sk_msg_free_nocharge(struct sock *sk, struct sk_msg *msg);
void sk_msg_free_partial(struct sock *sk, struct sk_msg *msg, u32 bytes);
void sk_msg_free_partial_nocharge(struct sock *sk, struct sk_msg *msg,
				  u32 bytes);

void sk_msg_return(struct sock *sk, struct sk_msg *msg, int bytes);
void sk_msg_return_zero(struct sock *sk, struct sk_msg *msg, int bytes);

int sk_msg_zerocopy_from_iter(struct sock *sk, struct iov_iter *from,
			      struct sk_msg *msg, u32 bytes);
int sk_msg_memcopy_from_iter(struct sock *sk, struct iov_iter *from,
			     struct sk_msg *msg, u32 bytes);
int sk_msg_recvmsg(struct sock *sk, struct sk_psock *psock, struct msghdr *msg,
		   int len, int flags);
bool sk_msg_is_readable(struct sock *sk);

static inline void sk_msg_check_to_free(struct sk_msg *msg, u32 i, u32 bytes)
{
	WARN_ON(i == msg->sg.end && bytes);
}

static inline void sk_msg_apply_bytes(struct sk_psock *psock, u32 bytes)
{
	if (psock->apply_bytes) {
		if (psock->apply_bytes < bytes)
			psock->apply_bytes = 0;
		else
			psock->apply_bytes -= bytes;
	}
}

static inline u32 sk_msg_iter_dist(u32 start, u32 end)
{
	return end >= start ? end - start : end + (NR_MSG_FRAG_IDS - start);
}

#define sk_msg_iter_var_prev(var)			\
	do {						\
		if (var == 0)				\
			var = NR_MSG_FRAG_IDS - 1;	\
		else					\
			var--;				\
	} while (0)

#define sk_msg_iter_var_next(var)			\
	do {						\
		var++;					\
		if (var == NR_MSG_FRAG_IDS)		\
			var = 0;			\
	} while (0)

#define sk_msg_iter_prev(msg, which)			\
	sk_msg_iter_var_prev(msg->sg.which)

#define sk_msg_iter_next(msg, which)			\
	sk_msg_iter_var_next(msg->sg.which)

static inline void sk_msg_init(struct sk_msg *msg)
{
	BUILD_BUG_ON(ARRAY_SIZE(msg->sg.data) - 1 != NR_MSG_FRAG_IDS);
	memset(msg, 0, sizeof(*msg));
	sg_init_marker(msg->sg.data, NR_MSG_FRAG_IDS);
}

static inline void sk_msg_xfer(struct sk_msg *dst, struct sk_msg *src,
			       int which, u32 size)
{
	dst->sg.data[which] = src->sg.data[which];
	dst->sg.data[which].length  = size;
	dst->sg.size		   += size;
	src->sg.size		   -= size;
	src->sg.data[which].length -= size;
	src->sg.data[which].offset += size;
}

static inline void sk_msg_xfer_full(struct sk_msg *dst, struct sk_msg *src)
{
	memcpy(dst, src, sizeof(*src));
	sk_msg_init(src);
}

static inline bool sk_msg_full(const struct sk_msg *msg)
{
	return sk_msg_iter_dist(msg->sg.start, msg->sg.end) == MAX_MSG_FRAGS;
}

static inline u32 sk_msg_elem_used(const struct sk_msg *msg)
{
	return sk_msg_iter_dist(msg->sg.start, msg->sg.end);
}

static inline struct scatterlist *sk_msg_elem(struct sk_msg *msg, int which)
{
	return &msg->sg.data[which];
}

static inline struct scatterlist sk_msg_elem_cpy(struct sk_msg *msg, int which)
{
	return msg->sg.data[which];
}

static inline struct page *sk_msg_page(struct sk_msg *msg, int which)
{
	return sg_page(sk_msg_elem(msg, which));
}

static inline bool sk_msg_to_ingress(const struct sk_msg *msg)
{
	return msg->flags & BPF_F_INGRESS;
}

static inline void sk_msg_compute_data_pointers(struct sk_msg *msg)
{
	struct scatterlist *sge = sk_msg_elem(msg, msg->sg.start);

	if (test_bit(msg->sg.start, msg->sg.copy)) {
		msg->data = NULL;
		msg->data_end = NULL;
	} else {
		msg->data = sg_virt(sge);
		msg->data_end = msg->data + sge->length;
	}
}

static inline void sk_msg_page_add(struct sk_msg *msg, struct page *page,
				   u32 len, u32 offset)
{
	struct scatterlist *sge;

	get_page(page);
	sge = sk_msg_elem(msg, msg->sg.end);
	sg_set_page(sge, page, len, offset);
	sg_unmark_end(sge);

	__set_bit(msg->sg.end, msg->sg.copy);
	msg->sg.size += len;
	sk_msg_iter_next(msg, end);
}

static inline void sk_msg_sg_copy(struct sk_msg *msg, u32 i, bool copy_state)
{
	do {
		if (copy_state)
			__set_bit(i, msg->sg.copy);
		else
			__clear_bit(i, msg->sg.copy);
		sk_msg_iter_var_next(i);
		if (i == msg->sg.end)
			break;
	} while (1);
}

static inline void sk_msg_sg_copy_set(struct sk_msg *msg, u32 start)
{
	sk_msg_sg_copy(msg, start, true);
}

static inline void sk_msg_sg_copy_clear(struct sk_msg *msg, u32 start)
{
	sk_msg_sg_copy(msg, start, false);
}

static inline struct sk_psock *sk_psock(const struct sock *sk)
{
	return __rcu_dereference_sk_user_data_with_flags(sk,
							 SK_USER_DATA_PSOCK);
}

static inline void sk_psock_set_state(struct sk_psock *psock,
				      enum sk_psock_state_bits bit)
{
	set_bit(bit, &psock->state);
}

static inline void sk_psock_clear_state(struct sk_psock *psock,
					enum sk_psock_state_bits bit)
{
	clear_bit(bit, &psock->state);
}

static inline bool sk_psock_test_state(const struct sk_psock *psock,
				       enum sk_psock_state_bits bit)
{
	return test_bit(bit, &psock->state);
}

static inline void sock_drop(struct sock *sk, struct sk_buff *skb)
{
	sk_drops_add(sk, skb);
	kfree_skb(skb);
}

static inline bool sk_psock_queue_msg(struct sk_psock *psock,
				      struct sk_msg *msg)
{
	bool ret;

	spin_lock_bh(&psock->ingress_lock);
	if (sk_psock_test_state(psock, SK_PSOCK_TX_ENABLED)) {
		list_add_tail(&msg->list, &psock->ingress_msg);
		ret = true;
	} else {
		sk_msg_free(psock->sk, msg);
		kfree(msg);
		ret = false;
	}
	spin_unlock_bh(&psock->ingress_lock);
	return ret;
}

static inline struct sk_msg *sk_psock_dequeue_msg(struct sk_psock *psock)
{
	struct sk_msg *msg;

	spin_lock_bh(&psock->ingress_lock);
	msg = list_first_entry_or_null(&psock->ingress_msg, struct sk_msg, list);
	if (msg)
		list_del(&msg->list);
	spin_unlock_bh(&psock->ingress_lock);
	return msg;
}

static inline struct sk_msg *sk_psock_peek_msg(struct sk_psock *psock)
{
	struct sk_msg *msg;

	spin_lock_bh(&psock->ingress_lock);
	msg = list_first_entry_or_null(&psock->ingress_msg, struct sk_msg, list);
	spin_unlock_bh(&psock->ingress_lock);
	return msg;
}

static inline struct sk_msg *sk_psock_next_msg(struct sk_psock *psock,
					       struct sk_msg *msg)
{
	struct sk_msg *ret;

	spin_lock_bh(&psock->ingress_lock);
	if (list_is_last(&msg->list, &psock->ingress_msg))
		ret = NULL;
	else
		ret = list_next_entry(msg, list);
	spin_unlock_bh(&psock->ingress_lock);
	return ret;
}

static inline bool sk_psock_queue_empty(const struct sk_psock *psock)
{
	return psock ? list_empty(&psock->ingress_msg) : true;
}

static inline void kfree_sk_msg(struct sk_msg *msg)
{
	if (msg->skb)
		consume_skb(msg->skb);
	kfree(msg);
}

static inline void sk_psock_report_error(struct sk_psock *psock, int err)
{
	struct sock *sk = psock->sk;

	sk->sk_err = err;
	sk_error_report(sk);
}

struct sk_psock *sk_psock_init(struct sock *sk, int node);
void sk_psock_stop(struct sk_psock *psock);

#if IS_ENABLED(CONFIG_BPF_STREAM_PARSER)
int sk_psock_init_strp(struct sock *sk, struct sk_psock *psock);
void sk_psock_start_strp(struct sock *sk, struct sk_psock *psock);
void sk_psock_stop_strp(struct sock *sk, struct sk_psock *psock);
#else
static inline int sk_psock_init_strp(struct sock *sk, struct sk_psock *psock)
{
	return -EOPNOTSUPP;
}

static inline void sk_psock_start_strp(struct sock *sk, struct sk_psock *psock)
{
}

static inline void sk_psock_stop_strp(struct sock *sk, struct sk_psock *psock)
{
}
#endif

void sk_psock_start_verdict(struct sock *sk, struct sk_psock *psock);
void sk_psock_stop_verdict(struct sock *sk, struct sk_psock *psock);

int sk_psock_msg_verdict(struct sock *sk, struct sk_psock *psock,
			 struct sk_msg *msg);

/*
 * This specialized allocator has to be a macro for its allocations to be
 * accounted separately (to have a separate alloc_tag). The typecast is
 * intentional to enforce typesafety.
 */
#define sk_psock_init_link()	\
		((struct sk_psock_link *)kzalloc(sizeof(struct sk_psock_link),	\
						 GFP_ATOMIC | __GFP_NOWARN))

static inline void sk_psock_free_link(struct sk_psock_link *link)
{
	kfree(link);
}

struct sk_psock_link *sk_psock_link_pop(struct sk_psock *psock);

static inline void sk_psock_cork_free(struct sk_psock *psock)
{
	if (psock->cork) {
		sk_msg_free(psock->sk, psock->cork);
		kfree(psock->cork);
		psock->cork = NULL;
	}
}

static inline void sk_psock_restore_proto(struct sock *sk,
					  struct sk_psock *psock)
{
	if (psock->psock_update_sk_prot)
		psock->psock_update_sk_prot(sk, psock, true);
}

static inline struct sk_psock *sk_psock_get(struct sock *sk)
{
	struct sk_psock *psock;

	rcu_read_lock();
	psock = sk_psock(sk);
	if (psock && !refcount_inc_not_zero(&psock->refcnt))
		psock = NULL;
	rcu_read_unlock();
	return psock;
}

void sk_psock_drop(struct sock *sk, struct sk_psock *psock);

static inline void sk_psock_put(struct sock *sk, struct sk_psock *psock)
{
	if (refcount_dec_and_test(&psock->refcnt))
		sk_psock_drop(sk, psock);
}

static inline void sk_psock_data_ready(struct sock *sk, struct sk_psock *psock)
{
	read_lock_bh(&sk->sk_callback_lock);
	if (psock->saved_data_ready)
		psock->saved_data_ready(sk);
	else
		sk->sk_data_ready(sk);
	read_unlock_bh(&sk->sk_callback_lock);
}

static inline void psock_set_prog(struct bpf_prog **pprog,
				  struct bpf_prog *prog)
{
	prog = xchg(pprog, prog);
	if (prog)
		bpf_prog_put(prog);
}

static inline int psock_replace_prog(struct bpf_prog **pprog,
				     struct bpf_prog *prog,
				     struct bpf_prog *old)
{
	if (cmpxchg(pprog, old, prog) != old)
		return -ENOENT;

	if (old)
		bpf_prog_put(old);

	return 0;
}

static inline void psock_progs_drop(struct sk_psock_progs *progs)
{
	psock_set_prog(&progs->msg_parser, NULL);
	psock_set_prog(&progs->stream_parser, NULL);
	psock_set_prog(&progs->stream_verdict, NULL);
	psock_set_prog(&progs->skb_verdict, NULL);
}

int sk_psock_tls_strp_read(struct sk_psock *psock, struct sk_buff *skb);

static inline bool sk_psock_strp_enabled(struct sk_psock *psock)
{
	if (!psock)
		return false;
	return !!psock->saved_data_ready;
}

#if IS_ENABLED(CONFIG_NET_SOCK_MSG)

#define BPF_F_STRPARSER	(1UL << 1)

/* We only have two bits so far. */
#define BPF_F_PTR_MASK ~(BPF_F_INGRESS | BPF_F_STRPARSER)

static inline bool skb_bpf_strparser(const struct sk_buff *skb)
{
	unsigned long sk_redir = skb->_sk_redir;

	return sk_redir & BPF_F_STRPARSER;
}

static inline void skb_bpf_set_strparser(struct sk_buff *skb)
{
	skb->_sk_redir |= BPF_F_STRPARSER;
}

static inline bool skb_bpf_ingress(const struct sk_buff *skb)
{
	unsigned long sk_redir = skb->_sk_redir;

	return sk_redir & BPF_F_INGRESS;
}

static inline void skb_bpf_set_ingress(struct sk_buff *skb)
{
	skb->_sk_redir |= BPF_F_INGRESS;
}

static inline void skb_bpf_set_redir(struct sk_buff *skb, struct sock *sk_redir,
				     bool ingress)
{
	skb->_sk_redir = (unsigned long)sk_redir;
	if (ingress)
		skb->_sk_redir |= BPF_F_INGRESS;
}

static inline struct sock *skb_bpf_redirect_fetch(const struct sk_buff *skb)
{
	unsigned long sk_redir = skb->_sk_redir;

	return (struct sock *)(sk_redir & BPF_F_PTR_MASK);
}

static inline void skb_bpf_redirect_clear(struct sk_buff *skb)
{
	skb->_sk_redir = 0;
}
#endif /* CONFIG_NET_SOCK_MSG */
#endif /* _LINUX_SKMSG_H */
Commit	Line	Data
604326b4 DB	1	/* SPDX-License-Identifier: GPL-2.0 */
	2	/* Copyright (c) 2017 - 2018 Covalent IO, Inc. http://covalent.io */
	3
	4	#ifndef _LINUX_SKMSG_H
	5	#define _LINUX_SKMSG_H
	6
	7	#include <linux/bpf.h>
	8	#include <linux/filter.h>
	9	#include <linux/scatterlist.h>
	10	#include <linux/skbuff.h>
	11
	12	#include <net/sock.h>
	13	#include <net/tcp.h>
	14	#include <net/strparser.h>
	15
	16	#define MAX_MSG_FRAGS MAX_SKB_FRAGS
031097d9	17	#define NR_MSG_FRAG_IDS (MAX_MSG_FRAGS + 1)
604326b4 DB	18
	19	enum __sk_action {
	20	__SK_DROP = 0,
	21	__SK_PASS,
	22	__SK_REDIRECT,
	23	__SK_NONE,
	24	};
	25
	26	struct sk_msg_sg {
	27	u32 start;
	28	u32 curr;
	29	u32 end;
	30	u32 size;
	31	u32 copybreak;
5a8fb33e	32	DECLARE_BITMAP(copy, MAX_MSG_FRAGS + 2);
031097d9 JK	33	/* The extra two elements:
	34	* 1) used for chaining the front and sections when the list becomes
	35	* partitioned (e.g. end < start). The crypto APIs require the
	36	* chaining;
	37	* 2) to chain tailer SG entries after the message.
d3b18ad3	38	*/
031097d9	39	struct scatterlist data[MAX_MSG_FRAGS + 2];
604326b4 DB	40	};
604326b4 DB	41
7a69c0f2	42	/* UAPI in filter.c depends on struct sk_msg_sg being first element. */
604326b4 DB	43	struct sk_msg {
	44	struct sk_msg_sg sg;
	45	void *data;
	46	void *data_end;
	47	u32 apply_bytes;
	48	u32 cork_bytes;
	49	u32 flags;
	50	struct sk_buff *skb;
	51	struct sock *sk_redir;
	52	struct sock *sk;
	53	struct list_head list;
	54	};
	55
	56	struct sk_psock_progs {
	57	struct bpf_prog *msg_parser;
ae8b8332 CW	58	struct bpf_prog *stream_parser;
ae8b8332 CW	59	struct bpf_prog *stream_verdict;
a7ba4558	60	struct bpf_prog *skb_verdict;
699c23f0 YS	61	struct bpf_link *msg_parser_link;
	62	struct bpf_link *stream_parser_link;
	63	struct bpf_link *stream_verdict_link;
	64	struct bpf_link *skb_verdict_link;
604326b4 DB	65	};
	66
	67	enum sk_psock_state_bits {
	68	SK_PSOCK_TX_ENABLED,
809e4dc7	69	SK_PSOCK_RX_STRP_ENABLED,
604326b4 DB	70	};
	71
	72	struct sk_psock_link {
	73	struct list_head list;
	74	struct bpf_map *map;
	75	void *link_raw;
	76	};
	77
604326b4	78	struct sk_psock_work_state {
604326b4 DB	79	u32 len;
	80	u32 off;
	81	};
	82
	83	struct sk_psock {
	84	struct sock *sk;
	85	struct sock *sk_redir;
	86	u32 apply_bytes;
	87	u32 cork_bytes;
	88	u32 eval;
a351d608	89	bool redir_ingress; /* undefined if sk_redir is null */
604326b4 DB	90	struct sk_msg *cork;
604326b4 DB	91	struct sk_psock_progs progs;
5a685cd9 CW	92	#if IS_ENABLED(CONFIG_BPF_STREAM_PARSER)
5a685cd9 CW	93	struct strparser strp;
36b62df5 JC	94	u32 copied_seq;
36b62df5 JC	95	u32 ingress_bytes;
5a685cd9	96	#endif
604326b4 DB	97	struct sk_buff_head ingress_skb;
604326b4 DB	98	struct list_head ingress_msg;
b01fd6e8	99	spinlock_t ingress_lock;
604326b4 DB	100	unsigned long state;
	101	struct list_head link;
	102	spinlock_t link_lock;
	103	refcount_t refcnt;
	104	void (saved_unhash)(struct sock sk);
d8616ee2	105	void (saved_destroy)(struct sock sk);
604326b4 DB	106	void (saved_close)(struct sock sk, long timeout);
604326b4 DB	107	void (saved_write_space)(struct sock sk);
5a685cd9	108	void (saved_data_ready)(struct sock sk);
7865dfb1 JF	109	/* psock_update_sk_prot may be called with restore=false many times
	110	* so the handler must be safe for this case. It will be called
	111	* exactly once with restore=true when the psock is being destroyed
	112	* and psock refcnt is zero, but before an RCU grace period.
	113	*/
51e0158a CW	114	int (psock_update_sk_prot)(struct sock sk, struct sk_psock *psock,
51e0158a CW	115	bool restore);
604326b4	116	struct proto *sk_proto;
799aa7f9	117	struct mutex work_mutex;
604326b4	118	struct sk_psock_work_state work_state;
29173d07	119	struct delayed_work work;
8866730a	120	struct sock *sk_pair;
7786dfc4	121	struct rcu_work rwork;
604326b4 DB	122	};
	123
	124	int sk_msg_alloc(struct sock sk, struct sk_msg msg, int len,
	125	int elem_first_coalesce);
d829e9c4 DB	126	int sk_msg_clone(struct sock sk, struct sk_msg dst, struct sk_msg *src,
d829e9c4 DB	127	u32 off, u32 len);
604326b4 DB	128	void sk_msg_trim(struct sock sk, struct sk_msg msg, int len);
	129	int sk_msg_free(struct sock sk, struct sk_msg msg);
	130	int sk_msg_free_nocharge(struct sock sk, struct sk_msg msg);
	131	void sk_msg_free_partial(struct sock sk, struct sk_msg msg, u32 bytes);
	132	void sk_msg_free_partial_nocharge(struct sock sk, struct sk_msg msg,
	133	u32 bytes);
	134
	135	void sk_msg_return(struct sock sk, struct sk_msg msg, int bytes);
d3b18ad3	136	void sk_msg_return_zero(struct sock sk, struct sk_msg msg, int bytes);
604326b4 DB	137
	138	int sk_msg_zerocopy_from_iter(struct sock sk, struct iov_iter from,
	139	struct sk_msg *msg, u32 bytes);
	140	int sk_msg_memcopy_from_iter(struct sock sk, struct iov_iter from,
	141	struct sk_msg *msg, u32 bytes);
2bc793e3 CW	142	int sk_msg_recvmsg(struct sock sk, struct sk_psock psock, struct msghdr *msg,
2bc793e3 CW	143	int len, int flags);
fb4e0a5e	144	bool sk_msg_is_readable(struct sock *sk);
604326b4 DB	145
	146	static inline void sk_msg_check_to_free(struct sk_msg *msg, u32 i, u32 bytes)
	147	{
	148	WARN_ON(i == msg->sg.end && bytes);
	149	}
	150
	151	static inline void sk_msg_apply_bytes(struct sk_psock *psock, u32 bytes)
	152	{
	153	if (psock->apply_bytes) {
	154	if (psock->apply_bytes < bytes)
	155	psock->apply_bytes = 0;
	156	else
	157	psock->apply_bytes -= bytes;
	158	}
	159	}
	160
683916f6 JK	161	static inline u32 sk_msg_iter_dist(u32 start, u32 end)
683916f6 JK	162	{
031097d9	163	return end >= start ? end - start : end + (NR_MSG_FRAG_IDS - start);
683916f6 JK	164	}
683916f6 JK	165
604326b4 DB	166	#define sk_msg_iter_var_prev(var) \
	167	do { \
	168	if (var == 0) \
031097d9	169	var = NR_MSG_FRAG_IDS - 1; \
604326b4 DB	170	else \
	171	var--; \
	172	} while (0)
	173
	174	#define sk_msg_iter_var_next(var) \
	175	do { \
	176	var++; \
031097d9	177	if (var == NR_MSG_FRAG_IDS) \
604326b4 DB	178	var = 0; \
	179	} while (0)
	180
	181	#define sk_msg_iter_prev(msg, which) \
	182	sk_msg_iter_var_prev(msg->sg.which)
	183
	184	#define sk_msg_iter_next(msg, which) \
	185	sk_msg_iter_var_next(msg->sg.which)
	186
604326b4 DB	187	static inline void sk_msg_init(struct sk_msg *msg)
604326b4 DB	188	{
031097d9	189	BUILD_BUG_ON(ARRAY_SIZE(msg->sg.data) - 1 != NR_MSG_FRAG_IDS);
604326b4	190	memset(msg, 0, sizeof(*msg));
031097d9	191	sg_init_marker(msg->sg.data, NR_MSG_FRAG_IDS);
604326b4 DB	192	}
	193
	194	static inline void sk_msg_xfer(struct sk_msg dst, struct sk_msg src,
	195	int which, u32 size)
	196	{
	197	dst->sg.data[which] = src->sg.data[which];
	198	dst->sg.data[which].length = size;
3f4c3127	199	dst->sg.size += size;
81aabbb9	200	src->sg.size -= size;
604326b4 DB	201	src->sg.data[which].length -= size;
	202	src->sg.data[which].offset += size;
	203	}
	204
d3b18ad3 JF	205	static inline void sk_msg_xfer_full(struct sk_msg dst, struct sk_msg src)
	206	{
	207	memcpy(dst, src, sizeof(*src));
	208	sk_msg_init(src);
	209	}
	210
8734a162 JF	211	static inline bool sk_msg_full(const struct sk_msg *msg)
8734a162 JF	212	{
031097d9	213	return sk_msg_iter_dist(msg->sg.start, msg->sg.end) == MAX_MSG_FRAGS;
8734a162 JF	214	}
8734a162 JF	215
604326b4 DB	216	static inline u32 sk_msg_elem_used(const struct sk_msg *msg)
604326b4 DB	217	{
683916f6	218	return sk_msg_iter_dist(msg->sg.start, msg->sg.end);
604326b4 DB	219	}
604326b4 DB	220
604326b4 DB	221	static inline struct scatterlist sk_msg_elem(struct sk_msg msg, int which)
	222	{
	223	return &msg->sg.data[which];
	224	}
	225
6fff607e JF	226	static inline struct scatterlist sk_msg_elem_cpy(struct sk_msg *msg, int which)
	227	{
	228	return msg->sg.data[which];
	229	}
	230
604326b4 DB	231	static inline struct page sk_msg_page(struct sk_msg msg, int which)
	232	{
	233	return sg_page(sk_msg_elem(msg, which));
	234	}
	235
	236	static inline bool sk_msg_to_ingress(const struct sk_msg *msg)
	237	{
	238	return msg->flags & BPF_F_INGRESS;
	239	}
	240
	241	static inline void sk_msg_compute_data_pointers(struct sk_msg *msg)
	242	{
	243	struct scatterlist *sge = sk_msg_elem(msg, msg->sg.start);
	244
5a8fb33e	245	if (test_bit(msg->sg.start, msg->sg.copy)) {
604326b4 DB	246	msg->data = NULL;
	247	msg->data_end = NULL;
	248	} else {
	249	msg->data = sg_virt(sge);
	250	msg->data_end = msg->data + sge->length;
	251	}
	252	}
	253
	254	static inline void sk_msg_page_add(struct sk_msg msg, struct page page,
	255	u32 len, u32 offset)
	256	{
	257	struct scatterlist *sge;
	258
	259	get_page(page);
	260	sge = sk_msg_elem(msg, msg->sg.end);
	261	sg_set_page(sge, page, len, offset);
	262	sg_unmark_end(sge);
	263
5a8fb33e	264	__set_bit(msg->sg.end, msg->sg.copy);
604326b4 DB	265	msg->sg.size += len;
	266	sk_msg_iter_next(msg, end);
	267	}
	268
d3b18ad3 JF	269	static inline void sk_msg_sg_copy(struct sk_msg *msg, u32 i, bool copy_state)
	270	{
	271	do {
163ab96b	272	if (copy_state)
5a8fb33e	273	__set_bit(i, msg->sg.copy);
163ab96b	274	else
5a8fb33e	275	__clear_bit(i, msg->sg.copy);
d3b18ad3 JF	276	sk_msg_iter_var_next(i);
	277	if (i == msg->sg.end)
	278	break;
	279	} while (1);
	280	}
	281
	282	static inline void sk_msg_sg_copy_set(struct sk_msg *msg, u32 start)
	283	{
	284	sk_msg_sg_copy(msg, start, true);
	285	}
	286
	287	static inline void sk_msg_sg_copy_clear(struct sk_msg *msg, u32 start)
	288	{
	289	sk_msg_sg_copy(msg, start, false);
	290	}
	291
604326b4 DB	292	static inline struct sk_psock sk_psock(const struct sock sk)
604326b4 DB	293	{
2a013372 HJ	294	return __rcu_dereference_sk_user_data_with_flags(sk,
2a013372 HJ	295	SK_USER_DATA_PSOCK);
604326b4 DB	296	}
604326b4 DB	297
9635720b JF	298	static inline void sk_psock_set_state(struct sk_psock *psock,
	299	enum sk_psock_state_bits bit)
	300	{
	301	set_bit(bit, &psock->state);
	302	}
	303
	304	static inline void sk_psock_clear_state(struct sk_psock *psock,
	305	enum sk_psock_state_bits bit)
	306	{
	307	clear_bit(bit, &psock->state);
	308	}
	309
	310	static inline bool sk_psock_test_state(const struct sk_psock *psock,
	311	enum sk_psock_state_bits bit)
	312	{
	313	return test_bit(bit, &psock->state);
	314	}
	315
	316	static inline void sock_drop(struct sock sk, struct sk_buff skb)
	317	{
	318	sk_drops_add(sk, skb);
	319	kfree_skb(skb);
	320	}
	321
d888b7af	322	static inline bool sk_psock_queue_msg(struct sk_psock *psock,
604326b4 DB	323	struct sk_msg *msg)
604326b4 DB	324	{
d888b7af ZZ	325	bool ret;
d888b7af ZZ	326
b01fd6e8	327	spin_lock_bh(&psock->ingress_lock);
d888b7af	328	if (sk_psock_test_state(psock, SK_PSOCK_TX_ENABLED)) {
9635720b	329	list_add_tail(&msg->list, &psock->ingress_msg);
d888b7af ZZ	330	ret = true;
d888b7af ZZ	331	} else {
938d3480 WY	332	sk_msg_free(psock->sk, msg);
938d3480 WY	333	kfree(msg);
d888b7af	334	ret = false;
938d3480	335	}
b01fd6e8	336	spin_unlock_bh(&psock->ingress_lock);
d888b7af	337	return ret;
b01fd6e8 CW	338	}
	339
	340	static inline struct sk_msg sk_psock_dequeue_msg(struct sk_psock psock)
	341	{
	342	struct sk_msg *msg;
	343
	344	spin_lock_bh(&psock->ingress_lock);
	345	msg = list_first_entry_or_null(&psock->ingress_msg, struct sk_msg, list);
	346	if (msg)
	347	list_del(&msg->list);
	348	spin_unlock_bh(&psock->ingress_lock);
	349	return msg;
	350	}
	351
	352	static inline struct sk_msg sk_psock_peek_msg(struct sk_psock psock)
	353	{
	354	struct sk_msg *msg;
	355
	356	spin_lock_bh(&psock->ingress_lock);
	357	msg = list_first_entry_or_null(&psock->ingress_msg, struct sk_msg, list);
	358	spin_unlock_bh(&psock->ingress_lock);
	359	return msg;
	360	}
	361
	362	static inline struct sk_msg sk_psock_next_msg(struct sk_psock psock,
	363	struct sk_msg *msg)
	364	{
	365	struct sk_msg *ret;
	366
	367	spin_lock_bh(&psock->ingress_lock);
	368	if (list_is_last(&msg->list, &psock->ingress_msg))
	369	ret = NULL;
	370	else
	371	ret = list_next_entry(msg, list);
	372	spin_unlock_bh(&psock->ingress_lock);
	373	return ret;
604326b4 DB	374	}
604326b4 DB	375
d3b18ad3 JF	376	static inline bool sk_psock_queue_empty(const struct sk_psock *psock)
	377	{
	378	return psock ? list_empty(&psock->ingress_msg) : true;
	379	}
	380
b01fd6e8 CW	381	static inline void kfree_sk_msg(struct sk_msg *msg)
	382	{
	383	if (msg->skb)
	384	consume_skb(msg->skb);
	385	kfree(msg);
	386	}
	387
604326b4 DB	388	static inline void sk_psock_report_error(struct sk_psock *psock, int err)
	389	{
	390	struct sock *sk = psock->sk;
	391
	392	sk->sk_err = err;
e3ae2365	393	sk_error_report(sk);
604326b4 DB	394	}
	395
	396	struct sk_psock sk_psock_init(struct sock sk, int node);
8bbabb3f	397	void sk_psock_stop(struct sk_psock *psock);
604326b4	398
88759609	399	#if IS_ENABLED(CONFIG_BPF_STREAM_PARSER)
604326b4 DB	400	int sk_psock_init_strp(struct sock sk, struct sk_psock psock);
	401	void sk_psock_start_strp(struct sock sk, struct sk_psock psock);
	402	void sk_psock_stop_strp(struct sock sk, struct sk_psock psock);
88759609 CW	403	#else
	404	static inline int sk_psock_init_strp(struct sock sk, struct sk_psock psock)
	405	{
	406	return -EOPNOTSUPP;
	407	}
	408
	409	static inline void sk_psock_start_strp(struct sock sk, struct sk_psock psock)
	410	{
	411	}
	412
	413	static inline void sk_psock_stop_strp(struct sock sk, struct sk_psock psock)
	414	{
	415	}
	416	#endif
	417
ef565928 JF	418	void sk_psock_start_verdict(struct sock sk, struct sk_psock psock);
ef565928 JF	419	void sk_psock_stop_verdict(struct sock sk, struct sk_psock psock);
604326b4 DB	420
	421	int sk_psock_msg_verdict(struct sock sk, struct sk_psock psock,
	422	struct sk_msg *msg);
	423
3b0ba54d SB	424	/*
	425	* This specialized allocator has to be a macro for its allocations to be
	426	* accounted separately (to have a separate alloc_tag). The typecast is
	427	* intentional to enforce typesafety.
	428	*/
2c321f3f SB	429	#define sk_psock_init_link() \
	430	((struct sk_psock_link *)kzalloc(sizeof(struct sk_psock_link), \
	431	GFP_ATOMIC \| __GFP_NOWARN))
604326b4 DB	432
	433	static inline void sk_psock_free_link(struct sk_psock_link *link)
	434	{
	435	kfree(link);
	436	}
	437
	438	struct sk_psock_link sk_psock_link_pop(struct sk_psock psock);
604326b4	439
604326b4 DB	440	static inline void sk_psock_cork_free(struct sk_psock *psock)
	441	{
	442	if (psock->cork) {
	443	sk_msg_free(psock->sk, psock->cork);
	444	kfree(psock->cork);
	445	psock->cork = NULL;
	446	}
	447	}
	448
604326b4 DB	449	static inline void sk_psock_restore_proto(struct sock *sk,
	450	struct sk_psock *psock)
	451	{
8a59f9d1	452	if (psock->psock_update_sk_prot)
51e0158a	453	psock->psock_update_sk_prot(sk, psock, true);
604326b4 DB	454	}
604326b4 DB	455
604326b4 DB	456	static inline struct sk_psock sk_psock_get(struct sock sk)
	457	{
	458	struct sk_psock *psock;
	459
	460	rcu_read_lock();
	461	psock = sk_psock(sk);
	462	if (psock && !refcount_inc_not_zero(&psock->refcnt))
	463	psock = NULL;
	464	rcu_read_unlock();
	465	return psock;
	466	}
	467
604326b4 DB	468	void sk_psock_drop(struct sock sk, struct sk_psock psock);
	469
	470	static inline void sk_psock_put(struct sock sk, struct sk_psock psock)
	471	{
	472	if (refcount_dec_and_test(&psock->refcnt))
	473	sk_psock_drop(sk, psock);
	474	}
	475
552de910 JF	476	static inline void sk_psock_data_ready(struct sock sk, struct sk_psock psock)
552de910 JF	477	{
6648e613	478	read_lock_bh(&sk->sk_callback_lock);
5a685cd9 CW	479	if (psock->saved_data_ready)
5a685cd9 CW	480	psock->saved_data_ready(sk);
552de910 JF	481	else
552de910 JF	482	sk->sk_data_ready(sk);
6648e613	483	read_unlock_bh(&sk->sk_callback_lock);
552de910 JF	484	}
552de910 JF	485
604326b4 DB	486	static inline void psock_set_prog(struct bpf_prog **pprog,
	487	struct bpf_prog *prog)
	488	{
	489	prog = xchg(pprog, prog);
	490	if (prog)
	491	bpf_prog_put(prog);
	492	}
	493
bb0de313 LB	494	static inline int psock_replace_prog(struct bpf_prog **pprog,
	495	struct bpf_prog *prog,
	496	struct bpf_prog *old)
	497	{
	498	if (cmpxchg(pprog, old, prog) != old)
	499	return -ENOENT;
	500
	501	if (old)
	502	bpf_prog_put(old);
	503
	504	return 0;
	505	}
	506
604326b4 DB	507	static inline void psock_progs_drop(struct sk_psock_progs *progs)
	508	{
	509	psock_set_prog(&progs->msg_parser, NULL);
ae8b8332 CW	510	psock_set_prog(&progs->stream_parser, NULL);
ae8b8332 CW	511	psock_set_prog(&progs->stream_verdict, NULL);
a7ba4558	512	psock_set_prog(&progs->skb_verdict, NULL);
604326b4 DB	513	}
604326b4 DB	514
e91de6af JF	515	int sk_psock_tls_strp_read(struct sk_psock psock, struct sk_buff skb);
	516
	517	static inline bool sk_psock_strp_enabled(struct sk_psock *psock)
	518	{
	519	if (!psock)
	520	return false;
5a685cd9	521	return !!psock->saved_data_ready;
e91de6af	522	}
e3526bb9 CW	523
	524	#if IS_ENABLED(CONFIG_NET_SOCK_MSG)
	525
7303524e LJ	526	#define BPF_F_STRPARSER (1UL << 1)
	527
	528	/* We only have two bits so far. */
	529	#define BPF_F_PTR_MASK ~(BPF_F_INGRESS \| BPF_F_STRPARSER)
	530
	531	static inline bool skb_bpf_strparser(const struct sk_buff *skb)
	532	{
	533	unsigned long sk_redir = skb->_sk_redir;
	534
	535	return sk_redir & BPF_F_STRPARSER;
	536	}
	537
	538	static inline void skb_bpf_set_strparser(struct sk_buff *skb)
	539	{
	540	skb->_sk_redir \|= BPF_F_STRPARSER;
	541	}
e3526bb9 CW	542
	543	static inline bool skb_bpf_ingress(const struct sk_buff *skb)
	544	{
	545	unsigned long sk_redir = skb->_sk_redir;
	546
	547	return sk_redir & BPF_F_INGRESS;
	548	}
	549
	550	static inline void skb_bpf_set_ingress(struct sk_buff *skb)
	551	{
	552	skb->_sk_redir \|= BPF_F_INGRESS;
	553	}
	554
	555	static inline void skb_bpf_set_redir(struct sk_buff skb, struct sock sk_redir,
	556	bool ingress)
	557	{
	558	skb->_sk_redir = (unsigned long)sk_redir;
	559	if (ingress)
	560	skb->_sk_redir \|= BPF_F_INGRESS;
	561	}
	562
	563	static inline struct sock skb_bpf_redirect_fetch(const struct sk_buff skb)
	564	{
	565	unsigned long sk_redir = skb->_sk_redir;
	566
	567	return (struct sock *)(sk_redir & BPF_F_PTR_MASK);
	568	}
	569
	570	static inline void skb_bpf_redirect_clear(struct sk_buff *skb)
	571	{
	572	skb->_sk_redir = 0;
	573	}
	574	#endif /* CONFIG_NET_SOCK_MSG */
604326b4	575	#endif /* _LINUX_SKMSG_H */