Merge tag 'net-next-5.15' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net-next

Pull networking updates from Jakub Kicinski:
 "Core:

   - Enable memcg accounting for various networking objects.

  BPF:

   - Introduce bpf timers.

   - Add perf link and opaque bpf_cookie which the program can read out
     again, to be used in libbpf-based USDT library.

   - Add bpf_task_pt_regs() helper to access user space pt_regs in
     kprobes, to help user space stack unwinding.

   - Add support for UNIX sockets for BPF sockmap.

   - Extend BPF iterator support for UNIX domain sockets.

   - Allow BPF TCP congestion control progs and bpf iterators to call
     bpf_setsockopt(), e.g. to switch to another congestion control
     algorithm.

  Protocols:

   - Support IOAM Pre-allocated Trace with IPv6.

   - Support Management Component Transport Protocol.

   - bridge: multicast: add vlan support.

   - netfilter: add hooks for the SRv6 lightweight tunnel driver.

   - tcp:
       - enable mid-stream window clamping (by user space or BPF)
       - allow data-less, empty-cookie SYN with TFO_SERVER_COOKIE_NOT_REQD
       - more accurate DSACK processing for RACK-TLP

   - mptcp:
       - add full mesh path manager option
       - add partial support for MP_FAIL
       - improve use of backup subflows
       - optimize option processing

   - af_unix: add OOB notification support.

   - ipv6: add IFLA_INET6_RA_MTU to expose MTU value advertised by the
     router.

   - mac80211: Target Wake Time support in AP mode.

   - can: j1939: extend UAPI to notify about RX status.

  Driver APIs:

   - Add page frag support in page pool API.

   - Many improvements to the DSA (distributed switch) APIs.

   - ethtool: extend IRQ coalesce uAPI with timer reset modes.

   - devlink: control which auxiliary devices are created.

   - Support CAN PHYs via the generic PHY subsystem.

   - Proper cross-chip support for tag_8021q.

   - Allow TX forwarding for the software bridge data path to be
     offloaded to capable devices.

  Drivers:

   - veth: more flexible channels number configuration.

   - openvswitch: introduce per-cpu upcall dispatch.

   - Add internet mix (IMIX) mode to pktgen.

   - Transparently handle XDP operations in the bonding driver.

   - Add LiteETH network driver.

   - Renesas (ravb):
       - support Gigabit Ethernet IP

   - NXP Ethernet switch (sja1105):
       - fast aging support
       - support for "H" switch topologies
       - traffic termination for ports under VLAN-aware bridge

   - Intel 1G Ethernet
       - support getcrosststamp() with PCIe PTM (Precision Time
         Measurement) for better time sync
       - support Credit-Based Shaper (CBS) offload, enabling HW traffic
         prioritization and bandwidth reservation

   - Broadcom Ethernet (bnxt)
       - support pulse-per-second output
       - support larger Rx rings

   - Mellanox Ethernet (mlx5)
       - support ethtool RSS contexts and MQPRIO channel mode
       - support LAG offload with bridging
       - support devlink rate limit API
       - support packet sampling on tunnels

   - Huawei Ethernet (hns3):
       - basic devlink support
       - add extended IRQ coalescing support
       - report extended link state

   - Netronome Ethernet (nfp):
       - add conntrack offload support

   - Broadcom WiFi (brcmfmac):
       - add WPA3 Personal with FT to supported cipher suites
       - support 43752 SDIO device

   - Intel WiFi (iwlwifi):
       - support scanning hidden 6GHz networks
       - support for a new hardware family (Bz)

   - Xen pv driver:
       - harden netfront against malicious backends

   - Qualcomm mobile
       - ipa: refactor power management and enable automatic suspend
       - mhi: move MBIM to WWAN subsystem interfaces

  Refactor:

   - Ambient BPF run context and cgroup storage cleanup.

   - Compat rework for ndo_ioctl.

  Old code removal:

   - prism54 remove the obsoleted driver, deprecated by the p54 driver.

   - wan: remove sbni/granch driver"

* tag 'net-next-5.15' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net-next: (1715 commits)
  net: Add depends on OF_NET for LiteX's LiteETH
  ipv6: seg6: remove duplicated include
  net: hns3: remove unnecessary spaces
  net: hns3: add some required spaces
  net: hns3: clean up a type mismatch warning
  net: hns3: refine function hns3_set_default_feature()
  ipv6: remove duplicated 'net/lwtunnel.h' include
  net: w5100: check return value after calling platform_get_resource()
  net/mlxbf_gige: Make use of devm_platform_ioremap_resourcexxx()
  net: mdio: mscc-miim: Make use of the helper function devm_platform_ioremap_resource()
  net: mdio-ipq4019: Make use of devm_platform_ioremap_resource()
  fou: remove sparse errors
  ipv4: fix endianness issue in inet_rtm_getroute_build_skb()
  octeontx2-af: Set proper errorcode for IPv4 checksum errors
  octeontx2-af: Fix static code analyzer reported issues
  octeontx2-af: Fix mailbox errors in nix_rss_flowkey_cfg
  octeontx2-af: Fix loop in free and unmap counter
  af_unix: fix potential NULL deref in unix_dgram_connect()
  dpaa2-eth: Replace strlcpy with strscpy
  octeontx2-af: Use NDC TX for transmit packet data
  ...

1 files changed, 174 insertions, 0 deletions

diff --git a/net/unix/unix_bpf.c b/net/unix/unix_bpf.c
new file mode 100644
index 000000000000..b927e2baae50
--- /dev/null
+++ b/net/unix/unix_bpf.c
@@ -0,0 +1,174 @@
+// SPDX-License-Identifier: GPL-2.0
+/* Copyright (c) 2021 Cong Wang <cong.wang@bytedance.com> */
+
+#include <linux/skmsg.h>
+#include <linux/bpf.h>
+#include <net/sock.h>
+#include <net/af_unix.h>
+
+#define unix_sk_has_data(__sk, __psock)					\
+		({	!skb_queue_empty(&__sk->sk_receive_queue) ||	\
+			!skb_queue_empty(&__psock->ingress_skb) ||	\
+			!list_empty(&__psock->ingress_msg);		\
+		})
+
+static int unix_msg_wait_data(struct sock *sk, struct sk_psock *psock,
+			      long timeo)
+{
+	DEFINE_WAIT_FUNC(wait, woken_wake_function);
+	struct unix_sock *u = unix_sk(sk);
+	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);
+	if (!unix_sk_has_data(sk, psock)) {
+		mutex_unlock(&u->iolock);
+		wait_woken(&wait, TASK_INTERRUPTIBLE, timeo);
+		mutex_lock(&u->iolock);
+		ret = unix_sk_has_data(sk, psock);
+	}
+	sk_clear_bit(SOCKWQ_ASYNC_WAITDATA, sk);
+	remove_wait_queue(sk_sleep(sk), &wait);
+	return ret;
+}
+
+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;
+	int copied;
+
+	psock = sk_psock_get(sk);
+	if (unlikely(!psock))
+		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_recvmsg(sk, msg, len, flags);
+	}
+
+msg_bytes_ready:
+	copied = sk_msg_recvmsg(sk, psock, msg, len, flags);
+	if (!copied) {
+		long timeo;
+		int data;
+
+		timeo = sock_rcvtimeo(sk, nonblock);
+		data = unix_msg_wait_data(sk, psock, timeo);
+		if (data) {
+			if (!sk_psock_queue_empty(psock))
+				goto msg_bytes_ready;
+			mutex_unlock(&u->iolock);
+			sk_psock_put(sk, psock);
+			return __unix_recvmsg(sk, msg, len, flags);
+		}
+		copied = -EAGAIN;
+	}
+	mutex_unlock(&u->iolock);
+	sk_psock_put(sk, psock);
+	return copied;
+}
+
+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_dgram_bpf_rebuild_protos(struct proto *prot, const struct proto *base)
+{
+	*prot        = *base;
+	prot->close  = sock_map_close;
+	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_stream_bpf_check_needs_rebuild(struct proto *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_stream_prot_lock);
+	}
+}
+
+int unix_dgram_bpf_update_proto(struct sock *sk, struct sk_psock *psock, bool restore)
+{
+	if (sk->sk_type != SOCK_DGRAM)
+		return -EOPNOTSUPP;
+
+	if (restore) {
+		sk->sk_write_space = psock->saved_write_space;
+		WRITE_ONCE(sk->sk_prot, psock->sk_proto);
+		return 0;
+	}
+
+	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_dgram_bpf_rebuild_protos(&unix_dgram_bpf_prot, &unix_dgram_proto);
+	unix_stream_bpf_rebuild_protos(&unix_stream_bpf_prot, &unix_stream_proto);
+
+}