Merge branch 'sockmap: introduce BPF_SK_SKB_VERDICT and support UDP'

Cong Wang says:

====================

From: Cong Wang <cong.wang@bytedance.com>

We have thousands of services connected to a daemon on every host
via AF_UNIX dgram sockets, after they are moved into VM, we have to
add a proxy to forward these communications from VM to host, because
rewriting thousands of them is not practical. This proxy uses an
AF_UNIX socket connected to services and a UDP socket to connect to
the host. It is inefficient because data is copied between kernel
space and user space twice, and we can not use splice() which only
supports TCP. Therefore, we want to use sockmap to do the splicing
without going to user-space at all (after the initial setup).

Currently sockmap only fully supports TCP, UDP is partially supported
as it is only allowed to add into sockmap. This patchset, as the second
part of the original large patchset, extends sockmap with:
1) cross-protocol support with BPF_SK_SKB_VERDICT; 2) full UDP support.

On the high level, ->read_sock() is required for each protocol to support
sockmap redirection, and in order to do sock proto update, a new ops
->psock_update_sk_prot() is introduced, which is also required. And the
BPF ->recvmsg() is also needed to replace the original ->recvmsg() to
retrieve skmsg. To make life easier, we have to get rid of lock_sock()
in sk_psock_handle_skb(), otherwise we would have to implement
->sendmsg_locked() on top of ->sendmsg(), which is ugly.

Please see each patch for more details.

To see the big picture, the original patchset is available here:
https://github.com/congwang/linux/tree/sockmap
this patchset is also available:
https://github.com/congwang/linux/tree/sockmap2
---
v8: get rid of 'offset' in udp_read_sock()
    add checks for skb_verdict/stream_verdict conflict
    add two cleanup patches for sock_map_link()
    add a new test case

v7: use work_mutex to protect psock->work
    return err in udp_read_sock()
    add patch 6/13
    clean up test case

v6: get rid of sk_psock_zap_ingress()
    add rcu work patch

v5: use INDIRECT_CALL_2() for function pointers
    use ingress_lock to fix a race condition found by Jacub
    rename two helper functions

v4: get rid of lock_sock() in sk_psock_handle_skb()
    get rid of udp_sendmsg_locked()
    remove an empty line
    update cover letter

v3: export tcp/udp_update_proto()
    rename sk->sk_prot->psock_update_sk_prot()
    improve changelogs

v2: separate from the original large patchset
    rebase to the latest bpf-next
    split UDP test case
    move inet_csk_has_ulp() check to tcp_bpf.c
    clean up udp_read_sock()
====================

Signed-off-by: Alexei Starovoitov <ast@kernel.org>

26 files changed, 677 insertions, 237 deletions

diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h
index c8def85fcc22..dbf820a50a39 100644
--- a/include/linux/skbuff.h
+++ b/include/linux/skbuff.h
@@ -3626,6 +3626,7 @@ int skb_splice_bits(struct sk_buff *skb, struct sock *sk, unsigned int offset,
 		    unsigned int flags);
 int skb_send_sock_locked(struct sock *sk, struct sk_buff *skb, int offset,
 			 int len);
+int skb_send_sock(struct sock *sk, struct sk_buff *skb, int offset, int len);
 void skb_copy_and_csum_dev(const struct sk_buff *skb, u8 *to);
 unsigned int skb_zerocopy_headlen(const struct sk_buff *from);
 int skb_zerocopy(struct sk_buff *to, struct sk_buff *from,
diff --git a/include/linux/skmsg.h b/include/linux/skmsg.h
index 6c09d94be2e9..f78e90a04a69 100644
--- a/include/linux/skmsg.h
+++ b/include/linux/skmsg.h
@@ -58,6 +58,7 @@ struct sk_psock_progs {
 	struct bpf_prog			*msg_parser;
 	struct bpf_prog			*stream_parser;
 	struct bpf_prog			*stream_verdict;
+	struct bpf_prog			*skb_verdict;
 };
 
 enum sk_psock_state_bits {
@@ -89,6 +90,7 @@ struct sk_psock {
 #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;
@@ -97,13 +99,12 @@ struct sk_psock {
 	void (*saved_close)(struct sock *sk, long timeout);
 	void (*saved_write_space)(struct sock *sk);
 	void (*saved_data_ready)(struct sock *sk);
+	int  (*psock_update_sk_prot)(struct sock *sk, bool restore);
 	struct proto			*sk_proto;
+	struct mutex			work_mutex;
 	struct sk_psock_work_state	work_state;
 	struct work_struct		work;
-	union {
-		struct rcu_head		rcu;
-		struct work_struct	gc;
-	};
+	struct rcu_work			rwork;
 };
 
 int sk_msg_alloc(struct sock *sk, struct sk_msg *msg, int len,
@@ -124,6 +125,10 @@ 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_wait_data(struct sock *sk, struct sk_psock *psock, int flags,
+		     long timeo, int *err);
+int sk_msg_recvmsg(struct sock *sk, struct sk_psock *psock, struct msghdr *msg,
+		   int len, int flags);
 
 static inline void sk_msg_check_to_free(struct sk_msg *msg, u32 i, u32 bytes)
 {
@@ -284,7 +289,45 @@ static inline struct sk_psock *sk_psock(const struct sock *sk)
 static inline void sk_psock_queue_msg(struct sk_psock *psock,
 				      struct sk_msg *msg)
 {
+	spin_lock_bh(&psock->ingress_lock);
 	list_add_tail(&msg->list, &psock->ingress_msg);
+	spin_unlock_bh(&psock->ingress_lock);
+}
+
+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)
@@ -292,6 +335,13 @@ 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;
@@ -301,6 +351,7 @@ static inline void sk_psock_report_error(struct sk_psock *psock, int err)
 }
 
 struct sk_psock *sk_psock_init(struct sock *sk, int node);
+void sk_psock_stop(struct sk_psock *psock, bool wait);
 
 #if IS_ENABLED(CONFIG_BPF_STREAM_PARSER)
 int sk_psock_init_strp(struct sock *sk, struct sk_psock *psock);
@@ -349,25 +400,12 @@ static inline void sk_psock_cork_free(struct sk_psock *psock)
 	}
 }
 
-static inline void sk_psock_update_proto(struct sock *sk,
-					 struct sk_psock *psock,
-					 struct proto *ops)
-{
-	/* Pairs with lockless read in sk_clone_lock() */
-	WRITE_ONCE(sk->sk_prot, ops);
-}
-
 static inline void sk_psock_restore_proto(struct sock *sk,
 					  struct sk_psock *psock)
 {
 	sk->sk_prot->unhash = psock->saved_unhash;
-	if (inet_csk_has_ulp(sk)) {
-		tcp_update_ulp(sk, psock->sk_proto, psock->saved_write_space);
-	} else {
-		sk->sk_write_space = psock->saved_write_space;
-		/* Pairs with lockless read in sk_clone_lock() */
-		WRITE_ONCE(sk->sk_prot, psock->sk_proto);
-	}
+	if (psock->psock_update_sk_prot)
+		psock->psock_update_sk_prot(sk, true);
 }
 
 static inline void sk_psock_set_state(struct sk_psock *psock,
@@ -442,6 +480,7 @@ 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);
diff --git a/include/net/sock.h b/include/net/sock.h
index 0b6266fd6bf6..8b4155e756c2 100644
--- a/include/net/sock.h
+++ b/include/net/sock.h
@@ -1184,6 +1184,9 @@ struct proto {
 	void			(*unhash)(struct sock *sk);
 	void			(*rehash)(struct sock *sk);
 	int			(*get_port)(struct sock *sk, unsigned short snum);
+#ifdef CONFIG_BPF_SYSCALL
+	int			(*psock_update_sk_prot)(struct sock *sk, bool restore);
+#endif
 
 	/* Keeping track of sockets in use */
 #ifdef CONFIG_PROC_FS
diff --git a/include/net/tcp.h b/include/net/tcp.h
index 075de26f449d..31b1696c62ba 100644
--- a/include/net/tcp.h
+++ b/include/net/tcp.h
@@ -2203,13 +2203,12 @@ struct sk_psock;
 
 #ifdef CONFIG_BPF_SYSCALL
 struct proto *tcp_bpf_get_proto(struct sock *sk, struct sk_psock *psock);
+int tcp_bpf_update_proto(struct sock *sk, bool restore);
 void tcp_bpf_clone(const struct sock *sk, struct sock *newsk);
 #endif /* CONFIG_BPF_SYSCALL */
 
 int tcp_bpf_sendmsg_redir(struct sock *sk, struct sk_msg *msg, u32 bytes,
 			  int flags);
-int __tcp_bpf_recvmsg(struct sock *sk, struct sk_psock *psock,
-		      struct msghdr *msg, int len, int flags);
 #endif /* CONFIG_NET_SOCK_MSG */
 
 #if !defined(CONFIG_BPF_SYSCALL) || !defined(CONFIG_NET_SOCK_MSG)
diff --git a/include/net/udp.h b/include/net/udp.h
index d4d064c59232..347b62a753c3 100644
--- a/include/net/udp.h
+++ b/include/net/udp.h
@@ -329,6 +329,8 @@ struct sock *__udp6_lib_lookup(struct net *net,
 			       struct sk_buff *skb);
 struct sock *udp6_lib_lookup_skb(const struct sk_buff *skb,
 				 __be16 sport, __be16 dport);
+int udp_read_sock(struct sock *sk, read_descriptor_t *desc,
+		  sk_read_actor_t recv_actor);
 
 /* UDP uses skb->dev_scratch to cache as much information as possible and avoid
  * possibly multiple cache miss on dequeue()
@@ -518,6 +520,7 @@ static inline struct sk_buff *udp_rcv_segment(struct sock *sk,
 #ifdef CONFIG_BPF_SYSCALL
 struct sk_psock;
 struct proto *udp_bpf_get_proto(struct sock *sk, struct sk_psock *psock);
+int udp_bpf_update_proto(struct sock *sk, bool restore);
 #endif
 
 #endif	/* _UDP_H */
diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h
index 598716742593..49371eba98ba 100644
--- a/include/uapi/linux/bpf.h
+++ b/include/uapi/linux/bpf.h
@@ -957,6 +957,7 @@ enum bpf_attach_type {
 	BPF_XDP_CPUMAP,
 	BPF_SK_LOOKUP,
 	BPF_XDP,
+	BPF_SK_SKB_VERDICT,
 	__MAX_BPF_ATTACH_TYPE
 };
 
diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c
index 9603de81811a..6428634da57e 100644
--- a/kernel/bpf/syscall.c
+++ b/kernel/bpf/syscall.c
@@ -2948,6 +2948,7 @@ attach_type_to_prog_type(enum bpf_attach_type attach_type)
 		return BPF_PROG_TYPE_SK_MSG;
 	case BPF_SK_SKB_STREAM_PARSER:
 	case BPF_SK_SKB_STREAM_VERDICT:
+	case BPF_SK_SKB_VERDICT:
 		return BPF_PROG_TYPE_SK_SKB;
 	case BPF_LIRC_MODE2:
 		return BPF_PROG_TYPE_LIRC_MODE2;
diff --git a/net/core/skbuff.c b/net/core/skbuff.c
index e8320b5d651a..3ad9e8425ab2 100644
--- a/net/core/skbuff.c
+++ b/net/core/skbuff.c
@@ -2500,9 +2500,32 @@ int skb_splice_bits(struct sk_buff *skb, struct sock *sk, unsigned int offset,
 }
 EXPORT_SYMBOL_GPL(skb_splice_bits);
 
-/* Send skb data on a socket. Socket must be locked. */
-int skb_send_sock_locked(struct sock *sk, struct sk_buff *skb, int offset,
-			 int len)
+static int sendmsg_unlocked(struct sock *sk, struct msghdr *msg,
+			    struct kvec *vec, size_t num, size_t size)
+{
+	struct socket *sock = sk->sk_socket;
+
+	if (!sock)
+		return -EINVAL;
+	return kernel_sendmsg(sock, msg, vec, num, size);
+}
+
+static int sendpage_unlocked(struct sock *sk, struct page *page, int offset,
+			     size_t size, int flags)
+{
+	struct socket *sock = sk->sk_socket;
+
+	if (!sock)
+		return -EINVAL;
+	return kernel_sendpage(sock, page, offset, size, flags);
+}
+
+typedef int (*sendmsg_func)(struct sock *sk, struct msghdr *msg,
+			    struct kvec *vec, size_t num, size_t size);
+typedef int (*sendpage_func)(struct sock *sk, struct page *page, int offset,
+			     size_t size, int flags);
+static int __skb_send_sock(struct sock *sk, struct sk_buff *skb, int offset,
+			   int len, sendmsg_func sendmsg, sendpage_func sendpage)
 {
 	unsigned int orig_len = len;
 	struct sk_buff *head = skb;
@@ -2522,7 +2545,8 @@ do_frag_list:
 		memset(&msg, 0, sizeof(msg));
 		msg.msg_flags = MSG_DONTWAIT;
 
-		ret = kernel_sendmsg_locked(sk, &msg, &kv, 1, slen);
+		ret = INDIRECT_CALL_2(sendmsg, kernel_sendmsg_locked,
+				      sendmsg_unlocked, sk, &msg, &kv, 1, slen);
 		if (ret <= 0)
 			goto error;
 
@@ -2553,9 +2577,11 @@ do_frag_list:
 		slen = min_t(size_t, len, skb_frag_size(frag) - offset);
 
 		while (slen) {
-			ret = kernel_sendpage_locked(sk, skb_frag_page(frag),
-						     skb_frag_off(frag) + offset,
-						     slen, MSG_DONTWAIT);
+			ret = INDIRECT_CALL_2(sendpage, kernel_sendpage_locked,
+					      sendpage_unlocked, sk,
+					      skb_frag_page(frag),
+					      skb_frag_off(frag) + offset,
+					      slen, MSG_DONTWAIT);
 			if (ret <= 0)
 				goto error;
 
@@ -2587,8 +2613,23 @@ out:
 error:
 	return orig_len == len ? ret : orig_len - len;
 }
+
+/* Send skb data on a socket. Socket must be locked. */
+int skb_send_sock_locked(struct sock *sk, struct sk_buff *skb, int offset,
+			 int len)
+{
+	return __skb_send_sock(sk, skb, offset, len, kernel_sendmsg_locked,
+			       kernel_sendpage_locked);
+}
 EXPORT_SYMBOL_GPL(skb_send_sock_locked);
 
+/* Send skb data on a socket. Socket must be unlocked. */
+int skb_send_sock(struct sock *sk, struct sk_buff *skb, int offset, int len)
+{
+	return __skb_send_sock(sk, skb, offset, len, sendmsg_unlocked,
+			       sendpage_unlocked);
+}
+
 /**
  *	skb_store_bits - store bits from kernel buffer to skb
  *	@skb: destination buffer
diff --git a/net/core/skmsg.c b/net/core/skmsg.c
index 07f54015238a..92a83c02562a 100644
--- a/net/core/skmsg.c
+++ b/net/core/skmsg.c
@@ -399,6 +399,104 @@ out:
 }
 EXPORT_SYMBOL_GPL(sk_msg_memcopy_from_iter);
 
+int sk_msg_wait_data(struct sock *sk, struct sk_psock *psock, int flags,
+		     long timeo, int *err)
+{
+	DEFINE_WAIT_FUNC(wait, woken_wake_function);
+	int ret = 0;
+
+	if (sk->sk_shutdown & RCV_SHUTDOWN)
+		return 1;
+
+	if (!timeo)
+		return ret;
+
+	add_wait_queue(sk_sleep(sk), &wait);
+	sk_set_bit(SOCKWQ_ASYNC_WAITDATA, sk);
+	ret = sk_wait_event(sk, &timeo,
+			    !list_empty(&psock->ingress_msg) ||
+			    !skb_queue_empty(&sk->sk_receive_queue), &wait);
+	sk_clear_bit(SOCKWQ_ASYNC_WAITDATA, sk);
+	remove_wait_queue(sk_sleep(sk), &wait);
+	return ret;
+}
+EXPORT_SYMBOL_GPL(sk_msg_wait_data);
+
+/* Receive sk_msg from psock->ingress_msg to @msg. */
+int sk_msg_recvmsg(struct sock *sk, struct sk_psock *psock, struct msghdr *msg,
+		   int len, int flags)
+{
+	struct iov_iter *iter = &msg->msg_iter;
+	int peek = flags & MSG_PEEK;
+	struct sk_msg *msg_rx;
+	int i, copied = 0;
+
+	msg_rx = sk_psock_peek_msg(psock);
+	while (copied != len) {
+		struct scatterlist *sge;
+
+		if (unlikely(!msg_rx))
+			break;
+
+		i = msg_rx->sg.start;
+		do {
+			struct page *page;
+			int copy;
+
+			sge = sk_msg_elem(msg_rx, i);
+			copy = sge->length;
+			page = sg_page(sge);
+			if (copied + copy > len)
+				copy = len - copied;
+			copy = copy_page_to_iter(page, sge->offset, copy, iter);
+			if (!copy)
+				return copied ? copied : -EFAULT;
+
+			copied += copy;
+			if (likely(!peek)) {
+				sge->offset += copy;
+				sge->length -= copy;
+				if (!msg_rx->skb)
+					sk_mem_uncharge(sk, copy);
+				msg_rx->sg.size -= copy;
+
+				if (!sge->length) {
+					sk_msg_iter_var_next(i);
+					if (!msg_rx->skb)
+						put_page(page);
+				}
+			} else {
+				/* Lets not optimize peek case if copy_page_to_iter
+				 * didn't copy the entire length lets just break.
+				 */
+				if (copy != sge->length)
+					return copied;
+				sk_msg_iter_var_next(i);
+			}
+
+			if (copied == len)
+				break;
+		} while (i != msg_rx->sg.end);
+
+		if (unlikely(peek)) {
+			msg_rx = sk_psock_next_msg(psock, msg_rx);
+			if (!msg_rx)
+				break;
+			continue;
+		}
+
+		msg_rx->sg.start = i;
+		if (!sge->length && msg_rx->sg.start == msg_rx->sg.end) {
+			msg_rx = sk_psock_dequeue_msg(psock);
+			kfree_sk_msg(msg_rx);
+		}
+		msg_rx = sk_psock_peek_msg(psock);
+	}
+
+	return copied;
+}
+EXPORT_SYMBOL_GPL(sk_msg_recvmsg);
+
 static struct sk_msg *sk_psock_create_ingress_msg(struct sock *sk,
 						  struct sk_buff *skb)
 {
@@ -410,7 +508,7 @@ static struct sk_msg *sk_psock_create_ingress_msg(struct sock *sk,
 	if (!sk_rmem_schedule(sk, skb, skb->truesize))
 		return NULL;
 
-	msg = kzalloc(sizeof(*msg), __GFP_NOWARN | GFP_ATOMIC);
+	msg = kzalloc(sizeof(*msg), __GFP_NOWARN | GFP_KERNEL);
 	if (unlikely(!msg))
 		return NULL;
 
@@ -497,7 +595,7 @@ static int sk_psock_handle_skb(struct sk_psock *psock, struct sk_buff *skb,
 	if (!ingress) {
 		if (!sock_writeable(psock->sk))
 			return -EAGAIN;
-		return skb_send_sock_locked(psock->sk, skb, off, len);
+		return skb_send_sock(psock->sk, skb, off, len);
 	}
 	return sk_psock_skb_ingress(psock, skb);
 }
@@ -511,8 +609,7 @@ static void sk_psock_backlog(struct work_struct *work)
 	u32 len, off;
 	int ret;
 
-	/* Lock sock to avoid losing sk_socket during loop. */
-	lock_sock(psock->sk);
+	mutex_lock(&psock->work_mutex);
 	if (state->skb) {
 		skb = state->skb;
 		len = state->len;
@@ -529,7 +626,7 @@ start:
 		skb_bpf_redirect_clear(skb);
 		do {
 			ret = -EIO;
-			if (likely(psock->sk->sk_socket))
+			if (!sock_flag(psock->sk, SOCK_DEAD))
 				ret = sk_psock_handle_skb(psock, skb, off,
 							  len, ingress);
 			if (ret <= 0) {
@@ -553,7 +650,7 @@ start:
 			kfree_skb(skb);
 	}
 end:
-	release_sock(psock->sk);
+	mutex_unlock(&psock->work_mutex);
 }
 
 struct sk_psock *sk_psock_init(struct sock *sk, int node)
@@ -563,11 +660,6 @@ struct sk_psock *sk_psock_init(struct sock *sk, int node)
 
 	write_lock_bh(&sk->sk_callback_lock);
 
-	if (inet_csk_has_ulp(sk)) {
-		psock = ERR_PTR(-EINVAL);
-		goto out;
-	}
-
 	if (sk->sk_user_data) {
 		psock = ERR_PTR(-EBUSY);
 		goto out;
@@ -591,7 +683,9 @@ struct sk_psock *sk_psock_init(struct sock *sk, int node)
 	spin_lock_init(&psock->link_lock);
 
 	INIT_WORK(&psock->work, sk_psock_backlog);
+	mutex_init(&psock->work_mutex);
 	INIT_LIST_HEAD(&psock->ingress_msg);
+	spin_lock_init(&psock->ingress_lock);
 	skb_queue_head_init(&psock->ingress_skb);
 
 	sk_psock_set_state(psock, SK_PSOCK_TX_ENABLED);
@@ -630,11 +724,11 @@ static void __sk_psock_purge_ingress_msg(struct sk_psock *psock)
 	}
 }
 
-static void sk_psock_zap_ingress(struct sk_psock *psock)
+static void __sk_psock_zap_ingress(struct sk_psock *psock)
 {
 	struct sk_buff *skb;
 
-	while ((skb = __skb_dequeue(&psock->ingress_skb)) != NULL) {
+	while ((skb = skb_dequeue(&psock->ingress_skb)) != NULL) {
 		skb_bpf_redirect_clear(skb);
 		kfree_skb(skb);
 	}
@@ -651,23 +745,35 @@ static void sk_psock_link_destroy(struct sk_psock *psock)
 	}
 }
 
+void sk_psock_stop(struct sk_psock *psock, bool wait)
+{
+	spin_lock_bh(&psock->ingress_lock);
+	sk_psock_clear_state(psock, SK_PSOCK_TX_ENABLED);
+	sk_psock_cork_free(psock);
+	__sk_psock_zap_ingress(psock);
+	spin_unlock_bh(&psock->ingress_lock);
+
+	if (wait)
+		cancel_work_sync(&psock->work);
+}
+
 static void sk_psock_done_strp(struct sk_psock *psock);
 
-static void sk_psock_destroy_deferred(struct work_struct *gc)
+static void sk_psock_destroy(struct work_struct *work)
 {
-	struct sk_psock *psock = container_of(gc, struct sk_psock, gc);
-
+	struct sk_psock *psock = container_of(to_rcu_work(work),
+					      struct sk_psock, rwork);
 	/* No sk_callback_lock since already detached. */
 
 	sk_psock_done_strp(psock);
 
 	cancel_work_sync(&psock->work);
+	mutex_destroy(&psock->work_mutex);
 
 	psock_progs_drop(&psock->progs);
 
 	sk_psock_link_destroy(psock);
 	sk_psock_cork_free(psock);
-	sk_psock_zap_ingress(psock);
 
 	if (psock->sk_redir)
 		sock_put(psock->sk_redir);
@@ -675,30 +781,21 @@ static void sk_psock_destroy_deferred(struct work_struct *gc)
 	kfree(psock);
 }
 
-static void sk_psock_destroy(struct rcu_head *rcu)
-{
-	struct sk_psock *psock = container_of(rcu, struct sk_psock, rcu);
-
-	INIT_WORK(&psock->gc, sk_psock_destroy_deferred);
-	schedule_work(&psock->gc);
-}
-
 void sk_psock_drop(struct sock *sk, struct sk_psock *psock)
 {
-	sk_psock_cork_free(psock);
-	sk_psock_zap_ingress(psock);
+	sk_psock_stop(psock, false);
 
 	write_lock_bh(&sk->sk_callback_lock);
 	sk_psock_restore_proto(sk, psock);
 	rcu_assign_sk_user_data(sk, NULL);
 	if (psock->progs.stream_parser)
 		sk_psock_stop_strp(sk, psock);
-	else if (psock->progs.stream_verdict)
+	else if (psock->progs.stream_verdict || psock->progs.skb_verdict)
 		sk_psock_stop_verdict(sk, psock);
 	write_unlock_bh(&sk->sk_callback_lock);
-	sk_psock_clear_state(psock, SK_PSOCK_TX_ENABLED);
 
-	call_rcu(&psock->rcu, sk_psock_destroy);
+	INIT_RCU_WORK(&psock->rwork, sk_psock_destroy);
+	queue_rcu_work(system_wq, &psock->rwork);
 }
 EXPORT_SYMBOL_GPL(sk_psock_drop);
 
@@ -767,14 +864,20 @@ static void sk_psock_skb_redirect(struct sk_buff *skb)
 	 * error that caused the pipe to break. We can't send a packet on
 	 * a socket that is in this state so we drop the skb.
 	 */
-	if (!psock_other || sock_flag(sk_other, SOCK_DEAD) ||
-	    !sk_psock_test_state(psock_other, SK_PSOCK_TX_ENABLED)) {
+	if (!psock_other || sock_flag(sk_other, SOCK_DEAD)) {
+		kfree_skb(skb);
+		return;
+	}
+	spin_lock_bh(&psock_other->ingress_lock);
+	if (!sk_psock_test_state(psock_other, SK_PSOCK_TX_ENABLED)) {
+		spin_unlock_bh(&psock_other->ingress_lock);
 		kfree_skb(skb);
 		return;
 	}
 
 	skb_queue_tail(&psock_other->ingress_skb, skb);
 	schedule_work(&psock_other->work);
+	spin_unlock_bh(&psock_other->ingress_lock);
 }
 
 static void sk_psock_tls_verdict_apply(struct sk_buff *skb, struct sock *sk, int verdict)
@@ -842,8 +945,12 @@ static void sk_psock_verdict_apply(struct sk_psock *psock,
 			err = sk_psock_skb_ingress_self(psock, skb);
 		}
 		if (err < 0) {
-			skb_queue_tail(&psock->ingress_skb, skb);
-			schedule_work(&psock->work);
+			spin_lock_bh(&psock->ingress_lock);
+			if (sk_psock_test_state(psock, SK_PSOCK_TX_ENABLED)) {
+				skb_queue_tail(&psock->ingress_skb, skb);
+				schedule_work(&psock->work);
+			}
+			spin_unlock_bh(&psock->ingress_lock);
 		}
 		break;
 	case __SK_REDIRECT:
@@ -1010,6 +1117,8 @@ static int sk_psock_verdict_recv(read_descriptor_t *desc, struct sk_buff *skb,
 	}
 	skb_set_owner_r(skb, sk);
 	prog = READ_ONCE(psock->progs.stream_verdict);
+	if (!prog)
+		prog = READ_ONCE(psock->progs.skb_verdict);
 	if (likely(prog)) {
 		skb_dst_drop(skb);
 		skb_bpf_redirect_clear(skb);
diff --git a/net/core/sock_map.c b/net/core/sock_map.c
index dd53a7771d7e..3d190d22b0d8 100644
--- a/net/core/sock_map.c
+++ b/net/core/sock_map.c
@@ -26,6 +26,7 @@ struct bpf_stab {
 
 static int sock_map_prog_update(struct bpf_map *map, struct bpf_prog *prog,
 				struct bpf_prog *old, u32 which);
+static struct sk_psock_progs *sock_map_progs(struct bpf_map *map);
 
 static struct bpf_map *sock_map_alloc(union bpf_attr *attr)
 {
@@ -155,6 +156,8 @@ static void sock_map_del_link(struct sock *sk,
 				strp_stop = true;
 			if (psock->saved_data_ready && stab->progs.stream_verdict)
 				verdict_stop = true;
+			if (psock->saved_data_ready && stab->progs.skb_verdict)
+				verdict_stop = true;
 			list_del(&link->list);
 			sk_psock_free_link(link);
 		}
@@ -182,26 +185,10 @@ static void sock_map_unref(struct sock *sk, void *link_raw)
 
 static int sock_map_init_proto(struct sock *sk, struct sk_psock *psock)
 {
-	struct proto *prot;
-
-	switch (sk->sk_type) {
-	case SOCK_STREAM:
-		prot = tcp_bpf_get_proto(sk, psock);
-		break;
-
-	case SOCK_DGRAM:
-		prot = udp_bpf_get_proto(sk, psock);
-		break;
-
-	default:
+	if (!sk->sk_prot->psock_update_sk_prot)
 		return -EINVAL;
-	}
-
-	if (IS_ERR(prot))
-		return PTR_ERR(prot);
-
-	sk_psock_update_proto(sk, psock, prot);
-	return 0;
+	psock->psock_update_sk_prot = sk->sk_prot->psock_update_sk_prot;
+	return sk->sk_prot->psock_update_sk_prot(sk, false);
 }
 
 static struct sk_psock *sock_map_psock_get_checked(struct sock *sk)
@@ -224,13 +211,25 @@ out:
 	return psock;
 }
 
-static int sock_map_link(struct bpf_map *map, struct sk_psock_progs *progs,
-			 struct sock *sk)
+static bool sock_map_redirect_allowed(const struct sock *sk);
+
+static int sock_map_link(struct bpf_map *map, struct sock *sk)
 {
-	struct bpf_prog *msg_parser, *stream_parser, *stream_verdict;
+	struct sk_psock_progs *progs = sock_map_progs(map);
+	struct bpf_prog *stream_verdict = NULL;
+	struct bpf_prog *stream_parser = NULL;
+	struct bpf_prog *skb_verdict = NULL;
+	struct bpf_prog *msg_parser = NULL;
 	struct sk_psock *psock;
 	int ret;
 
+	/* Only sockets we can redirect into/from in BPF need to hold
+	 * refs to parser/verdict progs and have their sk_data_ready
+	 * and sk_write_space callbacks overridden.
+	 */
+	if (!sock_map_redirect_allowed(sk))
+		goto no_progs;
+
 	stream_verdict = READ_ONCE(progs->stream_verdict);
 	if (stream_verdict) {
 		stream_verdict = bpf_prog_inc_not_zero(stream_verdict);
@@ -256,6 +255,16 @@ static int sock_map_link(struct bpf_map *map, struct sk_psock_progs *progs,
 		}
 	}
 
+	skb_verdict = READ_ONCE(progs->skb_verdict);
+	if (skb_verdict) {
+		skb_verdict = bpf_prog_inc_not_zero(skb_verdict);
+		if (IS_ERR(skb_verdict)) {
+			ret = PTR_ERR(skb_verdict);
+			goto out_put_msg_parser;
+		}
+	}
+
+no_progs:
 	psock = sock_map_psock_get_checked(sk);
 	if (IS_ERR(psock)) {
 		ret = PTR_ERR(psock);
@@ -265,6 +274,9 @@ static int sock_map_link(struct bpf_map *map, struct sk_psock_progs *progs,
 	if (psock) {
 		if ((msg_parser && READ_ONCE(psock->progs.msg_parser)) ||
 		    (stream_parser  && READ_ONCE(psock->progs.stream_parser)) ||
+		    (skb_verdict && READ_ONCE(psock->progs.skb_verdict)) ||
+		    (skb_verdict && READ_ONCE(psock->progs.stream_verdict)) ||
+		    (stream_verdict && READ_ONCE(psock->progs.skb_verdict)) ||
 		    (stream_verdict && READ_ONCE(psock->progs.stream_verdict))) {
 			sk_psock_put(sk, psock);
 			ret = -EBUSY;
@@ -296,6 +308,9 @@ static int sock_map_link(struct bpf_map *map, struct sk_psock_progs *progs,
 	} else if (!stream_parser && stream_verdict && !psock->saved_data_ready) {
 		psock_set_prog(&psock->progs.stream_verdict, stream_verdict);
 		sk_psock_start_verdict(sk,psock);
+	} else if (!stream_verdict && skb_verdict && !psock->saved_data_ready) {
+		psock_set_prog(&psock->progs.skb_verdict, skb_verdict);
+		sk_psock_start_verdict(sk, psock);
 	}
 	write_unlock_bh(&sk->sk_callback_lock);
 	return 0;
@@ -304,6 +319,9 @@ out_unlock_drop:
 out_drop:
 	sk_psock_put(sk, psock);
 out_progs:
+	if (skb_verdict)
+		bpf_prog_put(skb_verdict);
+out_put_msg_parser:
 	if (msg_parser)
 		bpf_prog_put(msg_parser);
 out_put_stream_parser:
@@ -315,27 +333,6 @@ out_put_stream_verdict:
 	return ret;
 }
 
-static int sock_map_link_no_progs(struct bpf_map *map, struct sock *sk)
-{
-	struct sk_psock *psock;
-	int ret;
-
-	psock = sock_map_psock_get_checked(sk);
-	if (IS_ERR(psock))
-		return PTR_ERR(psock);
-
-	if (!psock) {
-		psock = sk_psock_init(sk, map->numa_node);
-		if (IS_ERR(psock))
-			return PTR_ERR(psock);
-	}
-
-	ret = sock_map_init_proto(sk, psock);
-	if (ret < 0)
-		sk_psock_put(sk, psock);