From 92ef0fd55ac80dfc2e4654edfe5d1ddfa6e070fe Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Thu, 9 May 2024 09:20:08 -0600 Subject: [PATCH 1/4] net: change proto and proto_ops accept type Rather than pass in flags, error pointer, and whether this is a kernel invocation or not, add a struct proto_accept_arg struct as the argument. This then holds all of these arguments, and prepares accept for being able to pass back more information. No functional changes in this patch. Acked-by: Jakub Kicinski Signed-off-by: Jens Axboe --- crypto/af_alg.c | 11 ++++++----- crypto/algif_hash.c | 10 +++++----- drivers/xen/pvcalls-back.c | 6 +++++- fs/ocfs2/cluster/tcp.c | 5 ++++- include/crypto/if_alg.h | 3 ++- include/linux/net.h | 4 +++- include/net/inet_common.h | 4 ++-- include/net/inet_connection_sock.h | 2 +- include/net/sock.h | 12 +++++++++--- net/atm/svc.c | 8 ++++---- net/ax25/af_ax25.c | 6 +++--- net/bluetooth/iso.c | 4 ++-- net/bluetooth/l2cap_sock.c | 4 ++-- net/bluetooth/rfcomm/sock.c | 6 +++--- net/bluetooth/sco.c | 4 ++-- net/core/sock.c | 4 ++-- net/ipv4/af_inet.c | 10 +++++----- net/ipv4/inet_connection_sock.c | 6 +++--- net/iucv/af_iucv.c | 4 ++-- net/llc/af_llc.c | 7 +++---- net/mptcp/protocol.c | 11 +++++------ net/netrom/af_netrom.c | 6 +++--- net/nfc/llcp_sock.c | 4 ++-- net/phonet/pep.c | 12 ++++++------ net/phonet/socket.c | 7 +++---- net/rds/tcp_listen.c | 6 +++++- net/rose/af_rose.c | 6 +++--- net/sctp/socket.c | 8 ++++---- net/smc/af_smc.c | 6 +++--- net/socket.c | 13 ++++++++++--- net/tipc/socket.c | 13 +++++-------- net/unix/af_unix.c | 21 ++++++++++----------- net/vmw_vsock/af_vsock.c | 6 +++--- net/x25/af_x25.c | 4 ++-- 34 files changed, 132 insertions(+), 111 deletions(-) diff --git a/crypto/af_alg.c b/crypto/af_alg.c index 5bc6d0fa7498..18cfead0081d 100644 --- a/crypto/af_alg.c +++ b/crypto/af_alg.c @@ -407,7 +407,8 @@ unlock: return err; } -int af_alg_accept(struct sock *sk, struct socket *newsock, bool kern) +int af_alg_accept(struct sock *sk, struct socket *newsock, + struct proto_accept_arg *arg) { struct alg_sock *ask = alg_sk(sk); const struct af_alg_type *type; @@ -422,7 +423,7 @@ int af_alg_accept(struct sock *sk, struct socket *newsock, bool kern) if (!type) goto unlock; - sk2 = sk_alloc(sock_net(sk), PF_ALG, GFP_KERNEL, &alg_proto, kern); + sk2 = sk_alloc(sock_net(sk), PF_ALG, GFP_KERNEL, &alg_proto, arg->kern); err = -ENOMEM; if (!sk2) goto unlock; @@ -468,10 +469,10 @@ unlock: } EXPORT_SYMBOL_GPL(af_alg_accept); -static int alg_accept(struct socket *sock, struct socket *newsock, int flags, - bool kern) +static int alg_accept(struct socket *sock, struct socket *newsock, + struct proto_accept_arg *arg) { - return af_alg_accept(sock->sk, newsock, kern); + return af_alg_accept(sock->sk, newsock, arg); } static const struct proto_ops alg_proto_ops = { diff --git a/crypto/algif_hash.c b/crypto/algif_hash.c index e24c829d7a01..7c7394d46a23 100644 --- a/crypto/algif_hash.c +++ b/crypto/algif_hash.c @@ -223,8 +223,8 @@ unlock: return err ?: len; } -static int hash_accept(struct socket *sock, struct socket *newsock, int flags, - bool kern) +static int hash_accept(struct socket *sock, struct socket *newsock, + struct proto_accept_arg *arg) { struct sock *sk = sock->sk; struct alg_sock *ask = alg_sk(sk); @@ -252,7 +252,7 @@ static int hash_accept(struct socket *sock, struct socket *newsock, int flags, if (err) goto out_free_state; - err = af_alg_accept(ask->parent, newsock, kern); + err = af_alg_accept(ask->parent, newsock, arg); if (err) goto out_free_state; @@ -355,7 +355,7 @@ static int hash_recvmsg_nokey(struct socket *sock, struct msghdr *msg, } static int hash_accept_nokey(struct socket *sock, struct socket *newsock, - int flags, bool kern) + struct proto_accept_arg *arg) { int err; @@ -363,7 +363,7 @@ static int hash_accept_nokey(struct socket *sock, struct socket *newsock, if (err) return err; - return hash_accept(sock, newsock, flags, kern); + return hash_accept(sock, newsock, arg); } static struct proto_ops algif_hash_ops_nokey = { diff --git a/drivers/xen/pvcalls-back.c b/drivers/xen/pvcalls-back.c index d52593466a79..fd7ed65e0197 100644 --- a/drivers/xen/pvcalls-back.c +++ b/drivers/xen/pvcalls-back.c @@ -517,6 +517,10 @@ static void __pvcalls_back_accept(struct work_struct *work) { struct sockpass_mapping *mappass = container_of( work, struct sockpass_mapping, register_work); + struct proto_accept_arg arg = { + .flags = O_NONBLOCK, + .kern = true, + }; struct sock_mapping *map; struct pvcalls_ioworker *iow; struct pvcalls_fedata *fedata; @@ -548,7 +552,7 @@ static void __pvcalls_back_accept(struct work_struct *work) sock->type = mappass->sock->type; sock->ops = mappass->sock->ops; - ret = inet_accept(mappass->sock, sock, O_NONBLOCK, true); + ret = inet_accept(mappass->sock, sock, &arg); if (ret == -EAGAIN) { sock_release(sock); return; diff --git a/fs/ocfs2/cluster/tcp.c b/fs/ocfs2/cluster/tcp.c index 960080753d3b..2b8fa3e782fb 100644 --- a/fs/ocfs2/cluster/tcp.c +++ b/fs/ocfs2/cluster/tcp.c @@ -1784,6 +1784,9 @@ static int o2net_accept_one(struct socket *sock, int *more) struct o2nm_node *node = NULL; struct o2nm_node *local_node = NULL; struct o2net_sock_container *sc = NULL; + struct proto_accept_arg arg = { + .flags = O_NONBLOCK, + }; struct o2net_node *nn; unsigned int nofs_flag; @@ -1802,7 +1805,7 @@ static int o2net_accept_one(struct socket *sock, int *more) new_sock->type = sock->type; new_sock->ops = sock->ops; - ret = sock->ops->accept(sock, new_sock, O_NONBLOCK, false); + ret = sock->ops->accept(sock, new_sock, &arg); if (ret < 0) goto out; diff --git a/include/crypto/if_alg.h b/include/crypto/if_alg.h index 78ecaf5db04c..f7b3b93f3a49 100644 --- a/include/crypto/if_alg.h +++ b/include/crypto/if_alg.h @@ -166,7 +166,8 @@ int af_alg_unregister_type(const struct af_alg_type *type); int af_alg_release(struct socket *sock); void af_alg_release_parent(struct sock *sk); -int af_alg_accept(struct sock *sk, struct socket *newsock, bool kern); +int af_alg_accept(struct sock *sk, struct socket *newsock, + struct proto_accept_arg *arg); void af_alg_free_sg(struct af_alg_sgl *sgl); diff --git a/include/linux/net.h b/include/linux/net.h index 15df6d5f27a7..688320b79fcc 100644 --- a/include/linux/net.h +++ b/include/linux/net.h @@ -153,6 +153,7 @@ struct sockaddr; struct msghdr; struct module; struct sk_buff; +struct proto_accept_arg; typedef int (*sk_read_actor_t)(read_descriptor_t *, struct sk_buff *, unsigned int, size_t); typedef int (*skb_read_actor_t)(struct sock *, struct sk_buff *); @@ -171,7 +172,8 @@ struct proto_ops { int (*socketpair)(struct socket *sock1, struct socket *sock2); int (*accept) (struct socket *sock, - struct socket *newsock, int flags, bool kern); + struct socket *newsock, + struct proto_accept_arg *arg); int (*getname) (struct socket *sock, struct sockaddr *addr, int peer); diff --git a/include/net/inet_common.h b/include/net/inet_common.h index f50a644d87a9..c17a6585d0b0 100644 --- a/include/net/inet_common.h +++ b/include/net/inet_common.h @@ -29,8 +29,8 @@ int __inet_stream_connect(struct socket *sock, struct sockaddr *uaddr, int addr_len, int flags, int is_sendmsg); int inet_dgram_connect(struct socket *sock, struct sockaddr *uaddr, int addr_len, int flags); -int inet_accept(struct socket *sock, struct socket *newsock, int flags, - bool kern); +int inet_accept(struct socket *sock, struct socket *newsock, + struct proto_accept_arg *arg); void __inet_accept(struct socket *sock, struct socket *newsock, struct sock *newsk); int inet_send_prepare(struct sock *sk); diff --git a/include/net/inet_connection_sock.h b/include/net/inet_connection_sock.h index 20e7b0c0b3d1..7d6b1254c92d 100644 --- a/include/net/inet_connection_sock.h +++ b/include/net/inet_connection_sock.h @@ -250,7 +250,7 @@ inet_csk_rto_backoff(const struct inet_connection_sock *icsk, return (unsigned long)min_t(u64, when, max_when); } -struct sock *inet_csk_accept(struct sock *sk, int flags, int *err, bool kern); +struct sock *inet_csk_accept(struct sock *sk, struct proto_accept_arg *arg); int inet_csk_get_port(struct sock *sk, unsigned short snum); diff --git a/include/net/sock.h b/include/net/sock.h index 0450494a1766..217079b3e3e8 100644 --- a/include/net/sock.h +++ b/include/net/sock.h @@ -1194,6 +1194,12 @@ static inline void sk_prot_clear_nulls(struct sock *sk, int size) size - offsetof(struct sock, sk_node.pprev)); } +struct proto_accept_arg { + int flags; + int err; + bool kern; +}; + /* Networking protocol blocks we attach to sockets. * socket layer -> transport layer interface */ @@ -1208,8 +1214,8 @@ struct proto { int addr_len); int (*disconnect)(struct sock *sk, int flags); - struct sock * (*accept)(struct sock *sk, int flags, int *err, - bool kern); + struct sock * (*accept)(struct sock *sk, + struct proto_accept_arg *arg); int (*ioctl)(struct sock *sk, int cmd, int *karg); @@ -1804,7 +1810,7 @@ int sock_cmsg_send(struct sock *sk, struct msghdr *msg, int sock_no_bind(struct socket *, struct sockaddr *, int); int sock_no_connect(struct socket *, struct sockaddr *, int, int); int sock_no_socketpair(struct socket *, struct socket *); -int sock_no_accept(struct socket *, struct socket *, int, bool); +int sock_no_accept(struct socket *, struct socket *, struct proto_accept_arg *); int sock_no_getname(struct socket *, struct sockaddr *, int); int sock_no_ioctl(struct socket *, unsigned int, unsigned long); int sock_no_listen(struct socket *, int); diff --git a/net/atm/svc.c b/net/atm/svc.c index 36a814f1fbd1..f8137ae693b0 100644 --- a/net/atm/svc.c +++ b/net/atm/svc.c @@ -324,8 +324,8 @@ out: return error; } -static int svc_accept(struct socket *sock, struct socket *newsock, int flags, - bool kern) +static int svc_accept(struct socket *sock, struct socket *newsock, + struct proto_accept_arg *arg) { struct sock *sk = sock->sk; struct sk_buff *skb; @@ -336,7 +336,7 @@ static int svc_accept(struct socket *sock, struct socket *newsock, int flags, lock_sock(sk); - error = svc_create(sock_net(sk), newsock, 0, kern); + error = svc_create(sock_net(sk), newsock, 0, arg->kern); if (error) goto out; @@ -355,7 +355,7 @@ static int svc_accept(struct socket *sock, struct socket *newsock, int flags, error = -sk->sk_err; break; } - if (flags & O_NONBLOCK) { + if (arg->flags & O_NONBLOCK) { error = -EAGAIN; break; } diff --git a/net/ax25/af_ax25.c b/net/ax25/af_ax25.c index 9169efb2f43a..8077cf2ee448 100644 --- a/net/ax25/af_ax25.c +++ b/net/ax25/af_ax25.c @@ -1373,8 +1373,8 @@ out_release: return err; } -static int ax25_accept(struct socket *sock, struct socket *newsock, int flags, - bool kern) +static int ax25_accept(struct socket *sock, struct socket *newsock, + struct proto_accept_arg *arg) { struct sk_buff *skb; struct sock *newsk; @@ -1409,7 +1409,7 @@ static int ax25_accept(struct socket *sock, struct socket *newsock, int flags, if (skb) break; - if (flags & O_NONBLOCK) { + if (arg->flags & O_NONBLOCK) { err = -EWOULDBLOCK; break; } diff --git a/net/bluetooth/iso.c b/net/bluetooth/iso.c index ef0cc80b4c0c..2a075119d65d 100644 --- a/net/bluetooth/iso.c +++ b/net/bluetooth/iso.c @@ -1186,7 +1186,7 @@ done: } static int iso_sock_accept(struct socket *sock, struct socket *newsock, - int flags, bool kern) + struct proto_accept_arg *arg) { DEFINE_WAIT_FUNC(wait, woken_wake_function); struct sock *sk = sock->sk, *ch; @@ -1195,7 +1195,7 @@ static int iso_sock_accept(struct socket *sock, struct socket *newsock, lock_sock(sk); - timeo = sock_rcvtimeo(sk, flags & O_NONBLOCK); + timeo = sock_rcvtimeo(sk, arg->flags & O_NONBLOCK); BT_DBG("sk %p timeo %ld", sk, timeo); diff --git a/net/bluetooth/l2cap_sock.c b/net/bluetooth/l2cap_sock.c index 5cc83f906c12..125dddc77452 100644 --- a/net/bluetooth/l2cap_sock.c +++ b/net/bluetooth/l2cap_sock.c @@ -327,7 +327,7 @@ done: } static int l2cap_sock_accept(struct socket *sock, struct socket *newsock, - int flags, bool kern) + struct proto_accept_arg *arg) { DEFINE_WAIT_FUNC(wait, woken_wake_function); struct sock *sk = sock->sk, *nsk; @@ -336,7 +336,7 @@ static int l2cap_sock_accept(struct socket *sock, struct socket *newsock, lock_sock_nested(sk, L2CAP_NESTING_PARENT); - timeo = sock_rcvtimeo(sk, flags & O_NONBLOCK); + timeo = sock_rcvtimeo(sk, arg->flags & O_NONBLOCK); BT_DBG("sk %p timeo %ld", sk, timeo); diff --git a/net/bluetooth/rfcomm/sock.c b/net/bluetooth/rfcomm/sock.c index 29aa07e9db9d..37d63d768afb 100644 --- a/net/bluetooth/rfcomm/sock.c +++ b/net/bluetooth/rfcomm/sock.c @@ -468,8 +468,8 @@ done: return err; } -static int rfcomm_sock_accept(struct socket *sock, struct socket *newsock, int flags, - bool kern) +static int rfcomm_sock_accept(struct socket *sock, struct socket *newsock, + struct proto_accept_arg *arg) { DEFINE_WAIT_FUNC(wait, woken_wake_function); struct sock *sk = sock->sk, *nsk; @@ -483,7 +483,7 @@ static int rfcomm_sock_accept(struct socket *sock, struct socket *newsock, int f goto done; } - timeo = sock_rcvtimeo(sk, flags & O_NONBLOCK); + timeo = sock_rcvtimeo(sk, arg->flags & O_NONBLOCK); BT_DBG("sk %p timeo %ld", sk, timeo); diff --git a/net/bluetooth/sco.c b/net/bluetooth/sco.c index e0ad30862ee4..94c6f2b46279 100644 --- a/net/bluetooth/sco.c +++ b/net/bluetooth/sco.c @@ -647,7 +647,7 @@ done: } static int sco_sock_accept(struct socket *sock, struct socket *newsock, - int flags, bool kern) + struct proto_accept_arg *arg) { DEFINE_WAIT_FUNC(wait, woken_wake_function); struct sock *sk = sock->sk, *ch; @@ -656,7 +656,7 @@ static int sco_sock_accept(struct socket *sock, struct socket *newsock, lock_sock(sk); - timeo = sock_rcvtimeo(sk, flags & O_NONBLOCK); + timeo = sock_rcvtimeo(sk, arg->flags & O_NONBLOCK); BT_DBG("sk %p timeo %ld", sk, timeo); diff --git a/net/core/sock.c b/net/core/sock.c index 8d6e638b5426..8629f9aecf91 100644 --- a/net/core/sock.c +++ b/net/core/sock.c @@ -3241,8 +3241,8 @@ int sock_no_socketpair(struct socket *sock1, struct socket *sock2) } EXPORT_SYMBOL(sock_no_socketpair); -int sock_no_accept(struct socket *sock, struct socket *newsock, int flags, - bool kern) +int sock_no_accept(struct socket *sock, struct socket *newsock, + struct proto_accept_arg *arg) { return -EOPNOTSUPP; } diff --git a/net/ipv4/af_inet.c b/net/ipv4/af_inet.c index a7bad18bc8b5..de3449e16b89 100644 --- a/net/ipv4/af_inet.c +++ b/net/ipv4/af_inet.c @@ -771,16 +771,16 @@ void __inet_accept(struct socket *sock, struct socket *newsock, struct sock *new * Accept a pending connection. The TCP layer now gives BSD semantics. */ -int inet_accept(struct socket *sock, struct socket *newsock, int flags, - bool kern) +int inet_accept(struct socket *sock, struct socket *newsock, + struct proto_accept_arg *arg) { struct sock *sk1 = sock->sk, *sk2; - int err = -EINVAL; /* IPV6_ADDRFORM can change sk->sk_prot under us. */ - sk2 = READ_ONCE(sk1->sk_prot)->accept(sk1, flags, &err, kern); + arg->err = -EINVAL; + sk2 = READ_ONCE(sk1->sk_prot)->accept(sk1, arg); if (!sk2) - return err; + return arg->err; lock_sock(sk2); __inet_accept(sock, newsock, sk2); diff --git a/net/ipv4/inet_connection_sock.c b/net/ipv4/inet_connection_sock.c index 3b38610958ee..7734d189c66b 100644 --- a/net/ipv4/inet_connection_sock.c +++ b/net/ipv4/inet_connection_sock.c @@ -661,7 +661,7 @@ static int inet_csk_wait_for_connect(struct sock *sk, long timeo) /* * This will accept the next outstanding connection. */ -struct sock *inet_csk_accept(struct sock *sk, int flags, int *err, bool kern) +struct sock *inet_csk_accept(struct sock *sk, struct proto_accept_arg *arg) { struct inet_connection_sock *icsk = inet_csk(sk); struct request_sock_queue *queue = &icsk->icsk_accept_queue; @@ -680,7 +680,7 @@ struct sock *inet_csk_accept(struct sock *sk, int flags, int *err, bool kern) /* Find already established connection */ if (reqsk_queue_empty(queue)) { - long timeo = sock_rcvtimeo(sk, flags & O_NONBLOCK); + long timeo = sock_rcvtimeo(sk, arg->flags & O_NONBLOCK); /* If this is a non blocking socket don't sleep */ error = -EAGAIN; @@ -745,7 +745,7 @@ out: out_err: newsk = NULL; req = NULL; - *err = error; + arg->err = error; goto out; } EXPORT_SYMBOL(inet_csk_accept); diff --git a/net/iucv/af_iucv.c b/net/iucv/af_iucv.c index c951bb9cc2e0..c3b0b610b0aa 100644 --- a/net/iucv/af_iucv.c +++ b/net/iucv/af_iucv.c @@ -795,7 +795,7 @@ done: /* Accept a pending connection */ static int iucv_sock_accept(struct socket *sock, struct socket *newsock, - int flags, bool kern) + struct proto_accept_arg *arg) { DECLARE_WAITQUEUE(wait, current); struct sock *sk = sock->sk, *nsk; @@ -809,7 +809,7 @@ static int iucv_sock_accept(struct socket *sock, struct socket *newsock, goto done; } - timeo = sock_rcvtimeo(sk, flags & O_NONBLOCK); + timeo = sock_rcvtimeo(sk, arg->flags & O_NONBLOCK); /* Wait for an incoming connection */ add_wait_queue_exclusive(sk_sleep(sk), &wait); diff --git a/net/llc/af_llc.c b/net/llc/af_llc.c index fde1140d899e..4eb52add7103 100644 --- a/net/llc/af_llc.c +++ b/net/llc/af_llc.c @@ -688,14 +688,13 @@ static void llc_cmsg_rcv(struct msghdr *msg, struct sk_buff *skb) * llc_ui_accept - accept a new incoming connection. * @sock: Socket which connections arrive on. * @newsock: Socket to move incoming connection to. - * @flags: User specified operational flags. - * @kern: If the socket is kernel internal + * @arg: User specified arguments * * Accept a new incoming connection. * Returns 0 upon success, negative otherwise. */ -static int llc_ui_accept(struct socket *sock, struct socket *newsock, int flags, - bool kern) +static int llc_ui_accept(struct socket *sock, struct socket *newsock, + struct proto_accept_arg *arg) { struct sock *sk = sock->sk, *newsk; struct llc_sock *llc, *newllc; diff --git a/net/mptcp/protocol.c b/net/mptcp/protocol.c index bb8f96f2b86f..815ce439183c 100644 --- a/net/mptcp/protocol.c +++ b/net/mptcp/protocol.c @@ -3881,11 +3881,10 @@ unlock: } static int mptcp_stream_accept(struct socket *sock, struct socket *newsock, - int flags, bool kern) + struct proto_accept_arg *arg) { struct mptcp_sock *msk = mptcp_sk(sock->sk); struct sock *ssk, *newsk; - int err; pr_debug("msk=%p", msk); @@ -3897,9 +3896,9 @@ static int mptcp_stream_accept(struct socket *sock, struct socket *newsock, return -EINVAL; pr_debug("ssk=%p, listener=%p", ssk, mptcp_subflow_ctx(ssk)); - newsk = inet_csk_accept(ssk, flags, &err, kern); + newsk = inet_csk_accept(ssk, arg); if (!newsk) - return err; + return arg->err; pr_debug("newsk=%p, subflow is mptcp=%d", newsk, sk_is_mptcp(newsk)); if (sk_is_mptcp(newsk)) { @@ -3920,7 +3919,7 @@ static int mptcp_stream_accept(struct socket *sock, struct socket *newsock, newsk = new_mptcp_sock; MPTCP_INC_STATS(sock_net(ssk), MPTCP_MIB_MPCAPABLEPASSIVEACK); - newsk->sk_kern_sock = kern; + newsk->sk_kern_sock = arg->kern; lock_sock(newsk); __inet_accept(sock, newsock, newsk); @@ -3949,7 +3948,7 @@ static int mptcp_stream_accept(struct socket *sock, struct socket *newsock, } } else { tcpfallback: - newsk->sk_kern_sock = kern; + newsk->sk_kern_sock = arg->kern; lock_sock(newsk); __inet_accept(sock, newsock, newsk); /* we are being invoked after accepting a non-mp-capable diff --git a/net/netrom/af_netrom.c b/net/netrom/af_netrom.c index 104a80b75477..6ee148f0e6d0 100644 --- a/net/netrom/af_netrom.c +++ b/net/netrom/af_netrom.c @@ -772,8 +772,8 @@ out_release: return err; } -static int nr_accept(struct socket *sock, struct socket *newsock, int flags, - bool kern) +static int nr_accept(struct socket *sock, struct socket *newsock, + struct proto_accept_arg *arg) { struct sk_buff *skb; struct sock *newsk; @@ -805,7 +805,7 @@ static int nr_accept(struct socket *sock, struct socket *newsock, int flags, if (skb) break; - if (flags & O_NONBLOCK) { + if (arg->flags & O_NONBLOCK) { err = -EWOULDBLOCK; break; } diff --git a/net/nfc/llcp_sock.c b/net/nfc/llcp_sock.c index d5344563e525..57a2f97004e1 100644 --- a/net/nfc/llcp_sock.c +++ b/net/nfc/llcp_sock.c @@ -447,7 +447,7 @@ struct sock *nfc_llcp_accept_dequeue(struct sock *parent, } static int llcp_sock_accept(struct socket *sock, struct socket *newsock, - int flags, bool kern) + struct proto_accept_arg *arg) { DECLARE_WAITQUEUE(wait, current); struct sock *sk = sock->sk, *new_sk; @@ -463,7 +463,7 @@ static int llcp_sock_accept(struct socket *sock, struct socket *newsock, goto error; } - timeo = sock_rcvtimeo(sk, flags & O_NONBLOCK); + timeo = sock_rcvtimeo(sk, arg->flags & O_NONBLOCK); /* Wait for an incoming connection. */ add_wait_queue_exclusive(sk_sleep(sk), &wait); diff --git a/net/phonet/pep.c b/net/phonet/pep.c index 3dd5f52bc1b5..53a858478e22 100644 --- a/net/phonet/pep.c +++ b/net/phonet/pep.c @@ -759,8 +759,8 @@ static void pep_sock_close(struct sock *sk, long timeout) sock_put(sk); } -static struct sock *pep_sock_accept(struct sock *sk, int flags, int *errp, - bool kern) +static struct sock *pep_sock_accept(struct sock *sk, + struct proto_accept_arg *arg) { struct pep_sock *pn = pep_sk(sk), *newpn; struct sock *newsk = NULL; @@ -772,8 +772,8 @@ static struct sock *pep_sock_accept(struct sock *sk, int flags, int *errp, u8 pipe_handle, enabled, n_sb; u8 aligned = 0; - skb = skb_recv_datagram(sk, (flags & O_NONBLOCK) ? MSG_DONTWAIT : 0, - errp); + skb = skb_recv_datagram(sk, (arg->flags & O_NONBLOCK) ? MSG_DONTWAIT : 0, + &arg->err); if (!skb) return NULL; @@ -836,7 +836,7 @@ static struct sock *pep_sock_accept(struct sock *sk, int flags, int *errp, /* Create a new to-be-accepted sock */ newsk = sk_alloc(sock_net(sk), PF_PHONET, GFP_KERNEL, sk->sk_prot, - kern); + arg->kern); if (!newsk) { pep_reject_conn(sk, skb, PN_PIPE_ERR_OVERLOAD, GFP_KERNEL); err = -ENOBUFS; @@ -878,7 +878,7 @@ static struct sock *pep_sock_accept(struct sock *sk, int flags, int *errp, drop: release_sock(sk); kfree_skb(skb); - *errp = err; + arg->err = err; return newsk; } diff --git a/net/phonet/socket.c b/net/phonet/socket.c index 1018340d89a7..5ce0b3ee5def 100644 --- a/net/phonet/socket.c +++ b/net/phonet/socket.c @@ -292,18 +292,17 @@ out: } static int pn_socket_accept(struct socket *sock, struct socket *newsock, - int flags, bool kern) + struct proto_accept_arg *arg) { struct sock *sk = sock->sk; struct sock *newsk; - int err; if (unlikely(sk->sk_state != TCP_LISTEN)) return -EINVAL; - newsk = sk->sk_prot->accept(sk, flags, &err, kern); + newsk = sk->sk_prot->accept(sk, arg); if (!newsk) - return err; + return arg->err; lock_sock(newsk); sock_graft(newsk, newsock); diff --git a/net/rds/tcp_listen.c b/net/rds/tcp_listen.c index 05008ce5c421..d89bd8d0c354 100644 --- a/net/rds/tcp_listen.c +++ b/net/rds/tcp_listen.c @@ -105,6 +105,10 @@ int rds_tcp_accept_one(struct socket *sock) int conn_state; struct rds_conn_path *cp; struct in6_addr *my_addr, *peer_addr; + struct proto_accept_arg arg = { + .flags = O_NONBLOCK, + .kern = true, + }; #if !IS_ENABLED(CONFIG_IPV6) struct in6_addr saddr, daddr; #endif @@ -119,7 +123,7 @@ int rds_tcp_accept_one(struct socket *sock) if (ret) goto out; - ret = sock->ops->accept(sock, new_sock, O_NONBLOCK, true); + ret = sock->ops->accept(sock, new_sock, &arg); if (ret < 0) goto out; diff --git a/net/rose/af_rose.c b/net/rose/af_rose.c index ef81d019b20f..59050caab65c 100644 --- a/net/rose/af_rose.c +++ b/net/rose/af_rose.c @@ -919,8 +919,8 @@ out_release: return err; } -static int rose_accept(struct socket *sock, struct socket *newsock, int flags, - bool kern) +static int rose_accept(struct socket *sock, struct socket *newsock, + struct proto_accept_arg *arg) { struct sk_buff *skb; struct sock *newsk; @@ -953,7 +953,7 @@ static int rose_accept(struct socket *sock, struct socket *newsock, int flags, if (skb) break; - if (flags & O_NONBLOCK) { + if (arg->flags & O_NONBLOCK) { err = -EWOULDBLOCK; break; } diff --git a/net/sctp/socket.c b/net/sctp/socket.c index 64196b1dce1d..c009383369b2 100644 --- a/net/sctp/socket.c +++ b/net/sctp/socket.c @@ -4847,7 +4847,7 @@ static int sctp_disconnect(struct sock *sk, int flags) * descriptor will be returned from accept() to represent the newly * formed association. */ -static struct sock *sctp_accept(struct sock *sk, int flags, int *err, bool kern) +static struct sock *sctp_accept(struct sock *sk, struct proto_accept_arg *arg) { struct sctp_sock *sp; struct sctp_endpoint *ep; @@ -4871,7 +4871,7 @@ static struct sock *sctp_accept(struct sock *sk, int flags, int *err, bool kern) goto out; } - timeo = sock_rcvtimeo(sk, flags & O_NONBLOCK); + timeo = sock_rcvtimeo(sk, arg->flags & O_NONBLOCK); error = sctp_wait_for_accept(sk, timeo); if (error) @@ -4882,7 +4882,7 @@ static struct sock *sctp_accept(struct sock *sk, int flags, int *err, bool kern) */ asoc = list_entry(ep->asocs.next, struct sctp_association, asocs); - newsk = sp->pf->create_accept_sk(sk, asoc, kern); + newsk = sp->pf->create_accept_sk(sk, asoc, arg->kern); if (!newsk) { error = -ENOMEM; goto out; @@ -4899,7 +4899,7 @@ static struct sock *sctp_accept(struct sock *sk, int flags, int *err, bool kern) out: release_sock(sk); - *err = error; + arg->err = error; return newsk; } diff --git a/net/smc/af_smc.c b/net/smc/af_smc.c index 9389f0cfa374..e50a286fd0fb 100644 --- a/net/smc/af_smc.c +++ b/net/smc/af_smc.c @@ -2689,7 +2689,7 @@ out: } static int smc_accept(struct socket *sock, struct socket *new_sock, - int flags, bool kern) + struct proto_accept_arg *arg) { struct sock *sk = sock->sk, *nsk; DECLARE_WAITQUEUE(wait, current); @@ -2708,7 +2708,7 @@ static int smc_accept(struct socket *sock, struct socket *new_sock, } /* Wait for an incoming connection */ - timeo = sock_rcvtimeo(sk, flags & O_NONBLOCK); + timeo = sock_rcvtimeo(sk, arg->flags & O_NONBLOCK); add_wait_queue_exclusive(sk_sleep(sk), &wait); while (!(nsk = smc_accept_dequeue(sk, new_sock))) { set_current_state(TASK_INTERRUPTIBLE); @@ -2735,7 +2735,7 @@ static int smc_accept(struct socket *sock, struct socket *new_sock, if (rc) goto out; - if (lsmc->sockopt_defer_accept && !(flags & O_NONBLOCK)) { + if (lsmc->sockopt_defer_accept && !(arg->flags & O_NONBLOCK)) { /* wait till data arrives on the socket */ timeo = msecs_to_jiffies(lsmc->sockopt_defer_accept * MSEC_PER_SEC); diff --git a/net/socket.c b/net/socket.c index 01a71ae10c35..6ff5f21d9633 100644 --- a/net/socket.c +++ b/net/socket.c @@ -1898,6 +1898,9 @@ struct file *do_accept(struct file *file, unsigned file_flags, struct file *newfile; int err, len; struct sockaddr_storage address; + struct proto_accept_arg arg = { + .flags = file_flags, + }; const struct proto_ops *ops; sock = sock_from_file(file); @@ -1926,8 +1929,8 @@ struct file *do_accept(struct file *file, unsigned file_flags, if (err) goto out_fd; - err = ops->accept(sock, newsock, sock->file->f_flags | file_flags, - false); + arg.flags |= sock->file->f_flags; + err = ops->accept(sock, newsock, &arg); if (err < 0) goto out_fd; @@ -3580,6 +3583,10 @@ int kernel_accept(struct socket *sock, struct socket **newsock, int flags) { struct sock *sk = sock->sk; const struct proto_ops *ops = READ_ONCE(sock->ops); + struct proto_accept_arg arg = { + .flags = flags, + .kern = true, + }; int err; err = sock_create_lite(sk->sk_family, sk->sk_type, sk->sk_protocol, @@ -3587,7 +3594,7 @@ int kernel_accept(struct socket *sock, struct socket **newsock, int flags) if (err < 0) goto done; - err = ops->accept(sock, *newsock, flags, true); + err = ops->accept(sock, *newsock, &arg); if (err < 0) { sock_release(*newsock); *newsock = NULL; diff --git a/net/tipc/socket.c b/net/tipc/socket.c index 798397b6811e..2d58ecae4e21 100644 --- a/net/tipc/socket.c +++ b/net/tipc/socket.c @@ -146,8 +146,6 @@ static void tipc_data_ready(struct sock *sk); static void tipc_write_space(struct sock *sk); static void tipc_sock_destruct(struct sock *sk); static int tipc_release(struct socket *sock); -static int tipc_accept(struct socket *sock, struct socket *new_sock, int flags, - bool kern); static void tipc_sk_timeout(struct timer_list *t); static int tipc_sk_publish(struct tipc_sock *tsk, struct tipc_uaddr *ua); static int tipc_sk_withdraw(struct tipc_sock *tsk, struct tipc_uaddr *ua); @@ -2711,13 +2709,12 @@ static int tipc_wait_for_accept(struct socket *sock, long timeo) * tipc_accept - wait for connection request * @sock: listening socket * @new_sock: new socket that is to be connected - * @flags: file-related flags associated with socket - * @kern: caused by kernel or by userspace? + * @arg: arguments for accept * * Return: 0 on success, errno otherwise */ -static int tipc_accept(struct socket *sock, struct socket *new_sock, int flags, - bool kern) +static int tipc_accept(struct socket *sock, struct socket *new_sock, + struct proto_accept_arg *arg) { struct sock *new_sk, *sk = sock->sk; struct tipc_sock *new_tsock; @@ -2733,14 +2730,14 @@ static int tipc_accept(struct socket *sock, struct socket *new_sock, int flags, res = -EINVAL; goto exit; } - timeo = sock_rcvtimeo(sk, flags & O_NONBLOCK); + timeo = sock_rcvtimeo(sk, arg->flags & O_NONBLOCK); res = tipc_wait_for_accept(sock, timeo); if (res) goto exit; buf = skb_peek(&sk->sk_receive_queue); - res = tipc_sk_create(sock_net(sock->sk), new_sock, 0, kern); + res = tipc_sk_create(sock_net(sock->sk), new_sock, 0, arg->kern); if (res) goto exit; security_sk_clone(sock->sk, new_sock->sk); diff --git a/net/unix/af_unix.c b/net/unix/af_unix.c index dc1651541723..26e3b5f1ee46 100644 --- a/net/unix/af_unix.c +++ b/net/unix/af_unix.c @@ -755,7 +755,7 @@ static int unix_bind(struct socket *, struct sockaddr *, int); static int unix_stream_connect(struct socket *, struct sockaddr *, int addr_len, int flags); static int unix_socketpair(struct socket *, struct socket *); -static int unix_accept(struct socket *, struct socket *, int, bool); +static int unix_accept(struct socket *, struct socket *, struct proto_accept_arg *arg); static int unix_getname(struct socket *, struct sockaddr *, int); static __poll_t unix_poll(struct file *, struct socket *, poll_table *); static __poll_t unix_dgram_poll(struct file *, struct socket *, @@ -1689,19 +1689,18 @@ static void unix_sock_inherit_flags(const struct socket *old, set_bit(SOCK_PASSSEC, &new->flags); } -static int unix_accept(struct socket *sock, struct socket *newsock, int flags, - bool kern) +static int unix_accept(struct socket *sock, struct socket *newsock, + struct proto_accept_arg *arg) { struct sock *sk = sock->sk; struct sk_buff *skb; struct sock *tsk; - int err; - err = -EOPNOTSUPP; + arg->err = -EOPNOTSUPP; if (sock->type != SOCK_STREAM && sock->type != SOCK_SEQPACKET) goto out; - err = -EINVAL; + arg->err = -EINVAL; if (sk->sk_state != TCP_LISTEN) goto out; @@ -1709,12 +1708,12 @@ static int unix_accept(struct socket *sock, struct socket *newsock, int flags, * so that no locks are necessary. */ - skb = skb_recv_datagram(sk, (flags & O_NONBLOCK) ? MSG_DONTWAIT : 0, - &err); + skb = skb_recv_datagram(sk, (arg->flags & O_NONBLOCK) ? MSG_DONTWAIT : 0, + &arg->err); if (!skb) { /* This means receive shutdown. */ - if (err == 0) - err = -EINVAL; + if (arg->err == 0) + arg->err = -EINVAL; goto out; } @@ -1732,7 +1731,7 @@ static int unix_accept(struct socket *sock, struct socket *newsock, int flags, return 0; out: - return err; + return arg->err; } diff --git a/net/vmw_vsock/af_vsock.c b/net/vmw_vsock/af_vsock.c index 54ba7316f808..4b040285aa78 100644 --- a/net/vmw_vsock/af_vsock.c +++ b/net/vmw_vsock/af_vsock.c @@ -1500,8 +1500,8 @@ out: return err; } -static int vsock_accept(struct socket *sock, struct socket *newsock, int flags, - bool kern) +static int vsock_accept(struct socket *sock, struct socket *newsock, + struct proto_accept_arg *arg) { struct sock *listener; int err; @@ -1528,7 +1528,7 @@ static int vsock_accept(struct socket *sock, struct socket *newsock, int flags, /* Wait for children sockets to appear; these are the new sockets * created upon connection establishment. */ - timeout = sock_rcvtimeo(listener, flags & O_NONBLOCK); + timeout = sock_rcvtimeo(listener, arg->flags & O_NONBLOCK); prepare_to_wait(sk_sleep(listener), &wait, TASK_INTERRUPTIBLE); while ((connected = vsock_dequeue_accept(listener)) == NULL && diff --git a/net/x25/af_x25.c b/net/x25/af_x25.c index d18d51412cc0..8dda4178497c 100644 --- a/net/x25/af_x25.c +++ b/net/x25/af_x25.c @@ -871,8 +871,8 @@ static int x25_wait_for_data(struct sock *sk, long timeout) return rc; } -static int x25_accept(struct socket *sock, struct socket *newsock, int flags, - bool kern) +static int x25_accept(struct socket *sock, struct socket *newsock, + struct proto_accept_arg *arg) { struct sock *sk = sock->sk; struct sock *newsk; From 0645fbe760afcc5332c858d1cbf416bf77ef3c29 Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Thu, 9 May 2024 09:31:05 -0600 Subject: [PATCH 2/4] net: have do_accept() take a struct proto_accept_arg argument In preparation for passing in more information via this API, change do_accept() to take a proto_accept_arg struct pointer rather than just the file flags separately. No functional changes in this patch. Acked-by: Jakub Kicinski Signed-off-by: Jens Axboe --- include/linux/socket.h | 3 ++- io_uring/net.c | 6 ++++-- net/socket.c | 12 +++++------- 3 files changed, 11 insertions(+), 10 deletions(-) diff --git a/include/linux/socket.h b/include/linux/socket.h index 139c330ccf2c..89d16b90370b 100644 --- a/include/linux/socket.h +++ b/include/linux/socket.h @@ -16,6 +16,7 @@ struct cred; struct socket; struct sock; struct sk_buff; +struct proto_accept_arg; #define __sockaddr_check_size(size) \ BUILD_BUG_ON(((size) > sizeof(struct __kernel_sockaddr_storage))) @@ -433,7 +434,7 @@ extern int __sys_recvfrom(int fd, void __user *ubuf, size_t size, extern int __sys_sendto(int fd, void __user *buff, size_t len, unsigned int flags, struct sockaddr __user *addr, int addr_len); -extern struct file *do_accept(struct file *file, unsigned file_flags, +extern struct file *do_accept(struct file *file, struct proto_accept_arg *arg, struct sockaddr __user *upeer_sockaddr, int __user *upeer_addrlen, int flags); extern int __sys_accept4(int fd, struct sockaddr __user *upeer_sockaddr, diff --git a/io_uring/net.c b/io_uring/net.c index 070dea9a4eda..d4d1fc93635c 100644 --- a/io_uring/net.c +++ b/io_uring/net.c @@ -1528,8 +1528,10 @@ int io_accept(struct io_kiocb *req, unsigned int issue_flags) { struct io_accept *accept = io_kiocb_to_cmd(req, struct io_accept); bool force_nonblock = issue_flags & IO_URING_F_NONBLOCK; - unsigned int file_flags = force_nonblock ? O_NONBLOCK : 0; bool fixed = !!accept->file_slot; + struct proto_accept_arg arg = { + .flags = force_nonblock ? O_NONBLOCK : 0, + }; struct file *file; int ret, fd; @@ -1543,7 +1545,7 @@ retry: if (unlikely(fd < 0)) return fd; } - file = do_accept(req->file, file_flags, accept->addr, accept->addr_len, + file = do_accept(req->file, &arg, accept->addr, accept->addr_len, accept->flags); if (IS_ERR(file)) { if (!fixed) diff --git a/net/socket.c b/net/socket.c index 6ff5f21d9633..e416920e9399 100644 --- a/net/socket.c +++ b/net/socket.c @@ -1890,7 +1890,7 @@ SYSCALL_DEFINE2(listen, int, fd, int, backlog) return __sys_listen(fd, backlog); } -struct file *do_accept(struct file *file, unsigned file_flags, +struct file *do_accept(struct file *file, struct proto_accept_arg *arg, struct sockaddr __user *upeer_sockaddr, int __user *upeer_addrlen, int flags) { @@ -1898,9 +1898,6 @@ struct file *do_accept(struct file *file, unsigned file_flags, struct file *newfile; int err, len; struct sockaddr_storage address; - struct proto_accept_arg arg = { - .flags = file_flags, - }; const struct proto_ops *ops; sock = sock_from_file(file); @@ -1929,8 +1926,8 @@ struct file *do_accept(struct file *file, unsigned file_flags, if (err) goto out_fd; - arg.flags |= sock->file->f_flags; - err = ops->accept(sock, newsock, &arg); + arg->flags |= sock->file->f_flags; + err = ops->accept(sock, newsock, arg); if (err < 0) goto out_fd; @@ -1956,6 +1953,7 @@ out_fd: static int __sys_accept4_file(struct file *file, struct sockaddr __user *upeer_sockaddr, int __user *upeer_addrlen, int flags) { + struct proto_accept_arg arg = { }; struct file *newfile; int newfd; @@ -1969,7 +1967,7 @@ static int __sys_accept4_file(struct file *file, struct sockaddr __user *upeer_s if (unlikely(newfd < 0)) return newfd; - newfile = do_accept(file, 0, upeer_sockaddr, upeer_addrlen, + newfile = do_accept(file, &arg, upeer_sockaddr, upeer_addrlen, flags); if (IS_ERR(newfile)) { put_unused_fd(newfd); From 7951e36ac620a9ba1bae0ac0ddd62d2e82285725 Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Thu, 9 May 2024 09:35:01 -0600 Subject: [PATCH 3/4] net: pass back whether socket was empty post accept This adds an 'is_empty' argument to struct proto_accept_arg, which can be used to pass back information on whether or not the given socket has more connections to accept post the one just accepted. To utilize this information, the caller should initialize the 'is_empty' field to, eg, -1 and then check for 0/1 after the accept. If the field has been set, the caller knows whether there are more pending connections or not. If the field remains -1 after the accept call, the protocol doesn't support passing back this information. This patch wires it up for ipv4/6 TCP. Acked-by: Jakub Kicinski Signed-off-by: Jens Axboe --- include/net/sock.h | 1 + net/ipv4/inet_connection_sock.c | 1 + 2 files changed, 2 insertions(+) diff --git a/include/net/sock.h b/include/net/sock.h index 217079b3e3e8..5f4d0629348f 100644 --- a/include/net/sock.h +++ b/include/net/sock.h @@ -1197,6 +1197,7 @@ static inline void sk_prot_clear_nulls(struct sock *sk, int size) struct proto_accept_arg { int flags; int err; + int is_empty; bool kern; }; diff --git a/net/ipv4/inet_connection_sock.c b/net/ipv4/inet_connection_sock.c index 7734d189c66b..d81f74ce0f02 100644 --- a/net/ipv4/inet_connection_sock.c +++ b/net/ipv4/inet_connection_sock.c @@ -692,6 +692,7 @@ struct sock *inet_csk_accept(struct sock *sk, struct proto_accept_arg *arg) goto out_err; } req = reqsk_queue_remove(queue, sk); + arg->is_empty = reqsk_queue_empty(queue); newsk = req->sk; if (sk->sk_protocol == IPPROTO_TCP && From ac287da2e0ea5be2523222981efec86f0ca977cd Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Thu, 9 May 2024 09:41:10 -0600 Subject: [PATCH 4/4] io_uring/net: wire up IORING_CQE_F_SOCK_NONEMPTY for accept If the given protocol supports passing back whether or not we had more pending accept post this one, pass back this information to userspace. This is done by setting IORING_CQE_F_SOCK_NONEMPTY in the CQE flags, just like we do for recv/recvmsg if there's more data available post a receive operation. We can also use this information to be smarter about multishot retry, as we don't need to do a pointless retry if we know for a fact that there aren't any more connections to accept. Suggested-by: Norman Maurer Acked-by: Jakub Kicinski Signed-off-by: Jens Axboe --- io_uring/net.c | 20 ++++++++++++++++---- 1 file changed, 16 insertions(+), 4 deletions(-) diff --git a/io_uring/net.c b/io_uring/net.c index d4d1fc93635c..0a48596429d9 100644 --- a/io_uring/net.c +++ b/io_uring/net.c @@ -1533,6 +1533,7 @@ int io_accept(struct io_kiocb *req, unsigned int issue_flags) .flags = force_nonblock ? O_NONBLOCK : 0, }; struct file *file; + unsigned cflags; int ret, fd; if (!(req->flags & REQ_F_POLLED) && @@ -1545,6 +1546,8 @@ retry: if (unlikely(fd < 0)) return fd; } + arg.err = 0; + arg.is_empty = -1; file = do_accept(req->file, &arg, accept->addr, accept->addr_len, accept->flags); if (IS_ERR(file)) { @@ -1573,17 +1576,26 @@ retry: accept->file_slot); } + cflags = 0; + if (!arg.is_empty) + cflags |= IORING_CQE_F_SOCK_NONEMPTY; + if (!(req->flags & REQ_F_APOLL_MULTISHOT)) { - io_req_set_res(req, ret, 0); + io_req_set_res(req, ret, cflags); return IOU_OK; } if (ret < 0) return ret; - if (io_req_post_cqe(req, ret, IORING_CQE_F_MORE)) - goto retry; + if (io_req_post_cqe(req, ret, cflags | IORING_CQE_F_MORE)) { + if (cflags & IORING_CQE_F_SOCK_NONEMPTY || arg.is_empty == -1) + goto retry; + if (issue_flags & IO_URING_F_MULTISHOT) + return IOU_ISSUE_SKIP_COMPLETE; + return -EAGAIN; + } - io_req_set_res(req, ret, 0); + io_req_set_res(req, ret, cflags); return IOU_STOP_MULTISHOT; }