From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-16.8 required=3.0 tests=BAYES_00,DKIM_SIGNED, DKIM_VALID,HEADER_FROM_DIFFERENT_DOMAINS,INCLUDES_CR_TRAILER,INCLUDES_PATCH, MAILING_LIST_MULTI,SPF_HELO_NONE,SPF_PASS,URIBL_BLOCKED,USER_AGENT_GIT autolearn=unavailable autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id EF3D8C4320E for ; Thu, 5 Aug 2021 05:14:10 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by mail.kernel.org (Postfix) with ESMTP id D17DE610A8 for ; Thu, 5 Aug 2021 05:14:10 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S237234AbhHEFOW (ORCPT ); Thu, 5 Aug 2021 01:14:22 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:56244 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S237182AbhHEFOT (ORCPT ); Thu, 5 Aug 2021 01:14:19 -0400 Received: from mail-pl1-x633.google.com (mail-pl1-x633.google.com [IPv6:2607:f8b0:4864:20::633]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 0FBC5C0613C1 for ; Wed, 4 Aug 2021 22:14:06 -0700 (PDT) Received: by mail-pl1-x633.google.com with SMTP id q2so5660583plr.11 for ; Wed, 04 Aug 2021 22:14:06 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=bytedance-com.20150623.gappssmtp.com; s=20150623; h=from:to:cc:subject:date:message-id:in-reply-to:references :mime-version:content-transfer-encoding; bh=LNY3e4ptwysnI+KP7F89eycCkLBETHgwxYKM8oSuHDo=; b=BlCXomTK/b0a3A2Bs9K3ENvAm2i3zUkZ/gqLpzJefGploz+UhoOZCzXA/ljxnCTvV3 eEbFxu0auaCBn1OrLvH8kQ3MQmFmHtsnLKPQt8o0mYl8vWgjxxx5jiwfctEYet2B9q8u 2y1rpH9q4EUzlyYxKqjsh4AG+YDWE7hqLfqtfoJLhK7Tm4corDuBPzL6o26kN6fhDz+4 TqUYWPpJ38AeWOK8WgCYNi6A6Rg+dY+X8/iMM21oNXwhUu0HaIOis0Tn2RBewsDurWOv k4/KRpMtOx5uG44r6UVaFIxoQ9GXfYr2Z70cPEPOc1M4qDUgHHv5PwAffMLkZ2GauDb5 a9TQ== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:from:to:cc:subject:date:message-id:in-reply-to :references:mime-version:content-transfer-encoding; bh=LNY3e4ptwysnI+KP7F89eycCkLBETHgwxYKM8oSuHDo=; b=OtYgytlrIB1j1R32u79zvnt1bSTNqJq4qTuDj9vXjHQDV5XbeSoNRi74PDfpbViOch ROMLidjyzxMGPEmmjgypMeufEm05a1ctOFW9hbSTLz6Ro5HwTP4y7OjqaTuqi+fKHdVs EVMnlUHl/LSwTCHO09oGWiszEkPFk/pVybHseRlRqEXETP+K8502Qpa3KCF7VWegjPtZ rUDK7+oC7MuQhL2dKUgDohjGyrhhBcWR5YbSLZ2EFc+QrVf6L//nN7+z3eG7qX1fE7sE RTPGcsrZvAaR8SvvGvOSlwsO5vuiWv2g1AZKwrrCpYzvSK11zU8M6I90Jtvj1wBNOfES WQyw== X-Gm-Message-State: AOAM530u+EqYBoo+s6UivQx7oBUSHX7bodHDfw2XURGPJmnoPfWEhG6S 6bXBrK66tUNkCzgz+X+q8sj1nQ== X-Google-Smtp-Source: ABdhPJz2mCEPMwe3te2f2WDkRsK0Pa9drZeYDUcbK8AqRPqawfdhonc2caYBeYiicnpyTtuouo52UA== X-Received: by 2002:a63:d757:: with SMTP id w23mr27987pgi.434.1628140445484; Wed, 04 Aug 2021 22:14:05 -0700 (PDT) Received: from ip-10-124-121-13.byted.org (ec2-54-241-92-238.us-west-1.compute.amazonaws.com. [54.241.92.238]) by smtp.gmail.com with ESMTPSA id k4sm4201098pjs.55.2021.08.04.22.14.03 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Wed, 04 Aug 2021 22:14:05 -0700 (PDT) From: Jiang Wang To: netdev@vger.kernel.org Cc: cong.wang@bytedance.com, duanxiongchun@bytedance.com, xieyongji@bytedance.com, chaiwen.cc@bytedance.com, "David S. Miller" , Jakub Kicinski , John Fastabend , Daniel Borkmann , Jakub Sitnicki , Lorenz Bauer , Alexei Starovoitov , Andrii Nakryiko , Martin KaFai Lau , Song Liu , Yonghong Song , KP Singh , Shuah Khan , Al Viro , Christian Brauner , Johan Almbladh , linux-kernel@vger.kernel.org, bpf@vger.kernel.org, linux-kselftest@vger.kernel.org Subject: [PATCH bpf-next v4 2/5] af_unix: add unix_stream_proto for sockmap Date: Thu, 5 Aug 2021 05:13:34 +0000 Message-Id: <20210805051340.3798543-3-jiang.wang@bytedance.com> X-Mailer: git-send-email 2.20.1 In-Reply-To: <20210805051340.3798543-1-jiang.wang@bytedance.com> References: <20210805051340.3798543-1-jiang.wang@bytedance.com> MIME-Version: 1.0 Content-Transfer-Encoding: 8bit Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Previously, sockmap for AF_UNIX protocol only supports dgram type. This patch add unix stream type support, which is similar to unix_dgram_proto. To support sockmap, dgram and stream cannot share the same unix_proto anymore, because they have different implementations, such as unhash for stream type (which will remove closed or disconnected sockets from the map), so rename unix_proto to unix_dgram_proto and add a new unix_stream_proto. Also implement stream related sockmap functions. And add dgram key words to those dgram specific functions. Signed-off-by: Jiang Wang Reviewed-by: Cong Wang --- include/net/af_unix.h | 8 +++- net/unix/af_unix.c | 74 +++++++++++++++++++++++++++++----- net/unix/unix_bpf.c | 93 +++++++++++++++++++++++++++++++++---------- 3 files changed, 140 insertions(+), 35 deletions(-) diff --git a/include/net/af_unix.h b/include/net/af_unix.h index 435a2c3d5a6f..5d04fbf8aac8 100644 --- a/include/net/af_unix.h +++ b/include/net/af_unix.h @@ -84,6 +84,8 @@ long unix_outq_len(struct sock *sk); int __unix_dgram_recvmsg(struct sock *sk, struct msghdr *msg, size_t size, int flags); +int __unix_stream_recvmsg(struct sock *sk, struct msghdr *msg, size_t size, + int flags); #ifdef CONFIG_SYSCTL int unix_sysctl_register(struct net *net); void unix_sysctl_unregister(struct net *net); @@ -93,9 +95,11 @@ static inline void unix_sysctl_unregister(struct net *net) {} #endif #ifdef CONFIG_BPF_SYSCALL -extern struct proto unix_proto; +extern struct proto unix_dgram_proto; +extern struct proto unix_stream_proto; -int unix_bpf_update_proto(struct sock *sk, struct sk_psock *psock, bool restore); +int unix_dgram_bpf_update_proto(struct sock *sk, struct sk_psock *psock, bool restore); +int unix_stream_bpf_update_proto(struct sock *sk, struct sk_psock *psock, bool restore); void __init unix_bpf_build_proto(void); #else static inline void __init unix_bpf_build_proto(void) diff --git a/net/unix/af_unix.c b/net/unix/af_unix.c index c020ad0e8438..1eae92a7340b 100644 --- a/net/unix/af_unix.c +++ b/net/unix/af_unix.c @@ -791,17 +791,35 @@ static void unix_close(struct sock *sk, long timeout) */ } -struct proto unix_proto = { - .name = "UNIX", +static void unix_unhash(struct sock *sk) +{ + /* Nothing to do here, unix socket does not need a ->unhash(). + * This is merely for sockmap. + */ +} + +struct proto unix_dgram_proto = { + .name = "UNIX-DGRAM", + .owner = THIS_MODULE, + .obj_size = sizeof(struct unix_sock), + .close = unix_close, +#ifdef CONFIG_BPF_SYSCALL + .psock_update_sk_prot = unix_dgram_bpf_update_proto, +#endif +}; + +struct proto unix_stream_proto = { + .name = "UNIX-STREAM", .owner = THIS_MODULE, .obj_size = sizeof(struct unix_sock), .close = unix_close, + .unhash = unix_unhash, #ifdef CONFIG_BPF_SYSCALL - .psock_update_sk_prot = unix_bpf_update_proto, + .psock_update_sk_prot = unix_stream_bpf_update_proto, #endif }; -static struct sock *unix_create1(struct net *net, struct socket *sock, int kern) +static struct sock *unix_create1(struct net *net, struct socket *sock, int kern, int type) { struct sock *sk = NULL; struct unix_sock *u; @@ -810,7 +828,11 @@ static struct sock *unix_create1(struct net *net, struct socket *sock, int kern) if (atomic_long_read(&unix_nr_socks) > 2 * get_max_files()) goto out; - sk = sk_alloc(net, PF_UNIX, GFP_KERNEL, &unix_proto, kern); + if (type == SOCK_STREAM) + sk = sk_alloc(net, PF_UNIX, GFP_KERNEL, &unix_stream_proto, kern); + else /*dgram and seqpacket */ + sk = sk_alloc(net, PF_UNIX, GFP_KERNEL, &unix_dgram_proto, kern); + if (!sk) goto out; @@ -872,7 +894,7 @@ static int unix_create(struct net *net, struct socket *sock, int protocol, return -ESOCKTNOSUPPORT; } - return unix_create1(net, sock, kern) ? 0 : -ENOMEM; + return unix_create1(net, sock, kern, sock->type) ? 0 : -ENOMEM; } static int unix_release(struct socket *sock) @@ -1286,7 +1308,7 @@ static int unix_stream_connect(struct socket *sock, struct sockaddr *uaddr, err = -ENOMEM; /* create new sock for complete connection */ - newsk = unix_create1(sock_net(sk), NULL, 0); + newsk = unix_create1(sock_net(sk), NULL, 0, sock->type); if (newsk == NULL) goto out; @@ -2261,7 +2283,7 @@ static int unix_dgram_recvmsg(struct socket *sock, struct msghdr *msg, size_t si struct sock *sk = sock->sk; #ifdef CONFIG_BPF_SYSCALL - if (sk->sk_prot != &unix_proto) + if (READ_ONCE(sk->sk_prot) != &unix_dgram_proto) return sk->sk_prot->recvmsg(sk, msg, size, flags & MSG_DONTWAIT, flags & ~MSG_DONTWAIT, NULL); #endif @@ -2580,6 +2602,20 @@ static int unix_stream_read_actor(struct sk_buff *skb, return ret ?: chunk; } +int __unix_stream_recvmsg(struct sock *sk, struct msghdr *msg, + size_t size, int flags) +{ + struct unix_stream_read_state state = { + .recv_actor = unix_stream_read_actor, + .socket = sk->sk_socket, + .msg = msg, + .size = size, + .flags = flags + }; + + return unix_stream_read_generic(&state, true); +} + static int unix_stream_recvmsg(struct socket *sock, struct msghdr *msg, size_t size, int flags) { @@ -2591,6 +2627,12 @@ static int unix_stream_recvmsg(struct socket *sock, struct msghdr *msg, .flags = flags }; +#ifdef CONFIG_BPF_SYSCALL + struct sock *sk = sock->sk; + if (sk->sk_prot != &unix_stream_proto) + return sk->sk_prot->recvmsg(sk, msg, size, flags & MSG_DONTWAIT, + flags & ~MSG_DONTWAIT, NULL); +#endif return unix_stream_read_generic(&state, true); } @@ -2652,6 +2694,7 @@ static int unix_shutdown(struct socket *sock, int mode) int peer_mode = 0; + other->sk_prot->unhash(other); if (mode&RCV_SHUTDOWN) peer_mode |= SEND_SHUTDOWN; if (mode&SEND_SHUTDOWN) @@ -2660,8 +2703,10 @@ static int unix_shutdown(struct socket *sock, int mode) other->sk_shutdown |= peer_mode; unix_state_unlock(other); other->sk_state_change(other); - if (peer_mode == SHUTDOWN_MASK) + if (peer_mode == SHUTDOWN_MASK) { sk_wake_async(other, SOCK_WAKE_WAITD, POLL_HUP); + other->sk_state = TCP_CLOSE; + } else if (peer_mode & RCV_SHUTDOWN) sk_wake_async(other, SOCK_WAKE_WAITD, POLL_IN); } @@ -3040,7 +3085,13 @@ static int __init af_unix_init(void) BUILD_BUG_ON(sizeof(struct unix_skb_parms) > sizeof_field(struct sk_buff, cb)); - rc = proto_register(&unix_proto, 1); + rc = proto_register(&unix_dgram_proto, 1); + if (rc != 0) { + pr_crit("%s: Cannot create unix_sock SLAB cache!\n", __func__); + goto out; + } + + rc = proto_register(&unix_stream_proto, 1); if (rc != 0) { pr_crit("%s: Cannot create unix_sock SLAB cache!\n", __func__); goto out; @@ -3056,7 +3107,8 @@ static int __init af_unix_init(void) static void __exit af_unix_exit(void) { sock_unregister(PF_UNIX); - proto_unregister(&unix_proto); + proto_unregister(&unix_dgram_proto); + proto_unregister(&unix_stream_proto); unregister_pernet_subsys(&unix_net_ops); } diff --git a/net/unix/unix_bpf.c b/net/unix/unix_bpf.c index 20f53575b5c9..cbb0cdad63ba 100644 --- a/net/unix/unix_bpf.c +++ b/net/unix/unix_bpf.c @@ -38,9 +38,18 @@ static int unix_msg_wait_data(struct sock *sk, struct sk_psock *psock, return ret; } -static int unix_dgram_bpf_recvmsg(struct sock *sk, struct msghdr *msg, - size_t len, int nonblock, int flags, - int *addr_len) +static int __unix_recvmsg(struct sock *sk, struct msghdr *msg, + size_t len, int flags) +{ + if (sk->sk_type == SOCK_DGRAM) + return __unix_dgram_recvmsg(sk, msg, len, flags); + else + return __unix_stream_recvmsg(sk, msg, len, flags); +} + +static int unix_bpf_recvmsg(struct sock *sk, struct msghdr *msg, + size_t len, int nonblock, int flags, + int *addr_len) { struct unix_sock *u = unix_sk(sk); struct sk_psock *psock; @@ -48,14 +57,14 @@ static int unix_dgram_bpf_recvmsg(struct sock *sk, struct msghdr *msg, psock = sk_psock_get(sk); if (unlikely(!psock)) - return __unix_dgram_recvmsg(sk, msg, len, flags); + return __unix_recvmsg(sk, msg, len, flags); mutex_lock(&u->iolock); if (!skb_queue_empty(&sk->sk_receive_queue) && sk_psock_queue_empty(psock)) { mutex_unlock(&u->iolock); sk_psock_put(sk, psock); - return __unix_dgram_recvmsg(sk, msg, len, flags); + return __unix_recvmsg(sk, msg, len, flags); } msg_bytes_ready: @@ -71,7 +80,7 @@ static int unix_dgram_bpf_recvmsg(struct sock *sk, struct msghdr *msg, goto msg_bytes_ready; mutex_unlock(&u->iolock); sk_psock_put(sk, psock); - return __unix_dgram_recvmsg(sk, msg, len, flags); + return __unix_recvmsg(sk, msg, len, flags); } copied = -EAGAIN; } @@ -80,30 +89,55 @@ static int unix_dgram_bpf_recvmsg(struct sock *sk, struct msghdr *msg, return copied; } -static struct proto *unix_prot_saved __read_mostly; -static DEFINE_SPINLOCK(unix_prot_lock); -static struct proto unix_bpf_prot; +static struct proto *unix_dgram_prot_saved __read_mostly; +static DEFINE_SPINLOCK(unix_dgram_prot_lock); +static struct proto unix_dgram_bpf_prot; + +static struct proto *unix_stream_prot_saved __read_mostly; +static DEFINE_SPINLOCK(unix_stream_prot_lock); +static struct proto unix_stream_bpf_prot; -static void unix_bpf_rebuild_protos(struct proto *prot, const struct proto *base) +static void unix_dgram_bpf_rebuild_protos(struct proto *prot, const struct proto *base) { *prot = *base; prot->close = sock_map_close; - prot->recvmsg = unix_dgram_bpf_recvmsg; + prot->recvmsg = unix_bpf_recvmsg; +} + +static void unix_stream_bpf_rebuild_protos(struct proto *prot, + const struct proto *base) +{ + *prot = *base; + prot->close = sock_map_close; + prot->recvmsg = unix_bpf_recvmsg; + prot->unhash = sock_map_unhash; +} + +static void unix_dgram_bpf_check_needs_rebuild(struct proto *ops) +{ + if (unlikely(ops != smp_load_acquire(&unix_dgram_prot_saved))) { + spin_lock_bh(&unix_dgram_prot_lock); + if (likely(ops != unix_dgram_prot_saved)) { + unix_dgram_bpf_rebuild_protos(&unix_dgram_bpf_prot, ops); + smp_store_release(&unix_dgram_prot_saved, ops); + } + spin_unlock_bh(&unix_dgram_prot_lock); + } } -static void unix_bpf_check_needs_rebuild(struct proto *ops) +static void unix_stream_bpf_check_needs_rebuild(struct proto *ops) { - if (unlikely(ops != smp_load_acquire(&unix_prot_saved))) { - spin_lock_bh(&unix_prot_lock); - if (likely(ops != unix_prot_saved)) { - unix_bpf_rebuild_protos(&unix_bpf_prot, ops); - smp_store_release(&unix_prot_saved, ops); + if (unlikely(ops != smp_load_acquire(&unix_stream_prot_saved))) { + spin_lock_bh(&unix_stream_prot_lock); + if (likely(ops != unix_stream_prot_saved)) { + unix_stream_bpf_rebuild_protos(&unix_stream_bpf_prot, ops); + smp_store_release(&unix_stream_prot_saved, ops); } - spin_unlock_bh(&unix_prot_lock); + spin_unlock_bh(&unix_stream_prot_lock); } } -int unix_bpf_update_proto(struct sock *sk, struct sk_psock *psock, bool restore) +int unix_dgram_bpf_update_proto(struct sock *sk, struct sk_psock *psock, bool restore) { if (sk->sk_type != SOCK_DGRAM) return -EOPNOTSUPP; @@ -114,12 +148,27 @@ int unix_bpf_update_proto(struct sock *sk, struct sk_psock *psock, bool restore) return 0; } - unix_bpf_check_needs_rebuild(psock->sk_proto); - WRITE_ONCE(sk->sk_prot, &unix_bpf_prot); + unix_dgram_bpf_check_needs_rebuild(psock->sk_proto); + WRITE_ONCE(sk->sk_prot, &unix_dgram_bpf_prot); + return 0; +} + +int unix_stream_bpf_update_proto(struct sock *sk, struct sk_psock *psock, bool restore) +{ + if (restore) { + sk->sk_write_space = psock->saved_write_space; + WRITE_ONCE(sk->sk_prot, psock->sk_proto); + return 0; + } + + unix_stream_bpf_check_needs_rebuild(psock->sk_proto); + WRITE_ONCE(sk->sk_prot, &unix_stream_bpf_prot); return 0; } void __init unix_bpf_build_proto(void) { - unix_bpf_rebuild_protos(&unix_bpf_prot, &unix_proto); + unix_dgram_bpf_rebuild_protos(&unix_dgram_bpf_prot, &unix_dgram_proto); + unix_stream_bpf_rebuild_protos(&unix_stream_bpf_prot, &unix_stream_proto); + } -- 2.20.1