Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net

Pull networking fixes from David Miller:

 1) The wireless rate info fix from Johannes Berg.

 2) When a RAW socket is in hdrincl mode, we need to make sure that the
    user provided at least a minimally sized ipv4/ipv6 header. Fix from
    Alexander Potapenko.

 3) We must emit IFLA_PHYS_PORT_NAME netlink attributes using
    nla_put_string() so that it is NULL terminated.

 4) Fix a bug in TCP fastopen handling, wherein child sockets
    erroneously inherit the fastopen_req from the parent, and later can
    end up derefencing freed memory or doing a double free. From Eric
    Dumazet.

 5) Don't clear out netdev stats at close time in tg3 driver, from
    YueHaibing.

 6) Fix refcount leak in xt_CT, from Gao Feng.

 7) In nft_set_bitmap() don't leak dummy elements, from Liping Zhang.

 8) Fix deadlock due to taking the expectation lock twice, also from
    Liping Zhang.

 9) Make xt_socket work again with ipv6, from Peter Tirsek.

10) Don't allow IPV6 to be used with IPVS if ipv6.disable=1, from Paolo
    Abeni.

11) Make the BPF loader more flexible wrt. changes to the bpf MAP entry
    layout. From Jesper Dangaard Brouer.

12) Fix ethtool reported device name in aquantia driver, from Pavel
    Belous.

13) Fix build failures due to the compile time size test not working in
    netfilter conntrack. From Geert Uytterhoeven.

* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net: (52 commits)
  cfg80211: make RATE_INFO_BW_20 the default
  ipv6: initialize route null entry in addrconf_init()
  qede: Fix possible misconfiguration of advertised autoneg value.
  qed: Fix overriding of supported autoneg value.
  qed*: Fix possible overflow for status block id field.
  rtnetlink: NUL-terminate IFLA_PHYS_PORT_NAME string
  netvsc: make sure napi enabled before vmbus_open
  aquantia: Fix driver name reported by ethtool
  ipv4, ipv6: ensure raw socket message is big enough to hold an IP header
  net/sched: remove redundant null check on head
  tcp: do not inherit fastopen_req from parent
  forcedeth: remove unnecessary carrier status check
  ibmvnic: Move queue restarting in ibmvnic_tx_complete
  ibmvnic: Record SKB RX queue during poll
  ibmvnic: Continue skb processing after skb completion error
  ibmvnic: Check for driver reset first in ibmvnic_xmit
  ibmvnic: Wait for any pending scrqs entries at driver close
  ibmvnic: Clean up tx pools when closing
  ibmvnic: Whitespace correction in release_rx_pools
  ibmvnic: Delete napi's when releasing driver resources
  ...

4 files changed, 278 insertions, 3 deletions

diff --git a/tools/testing/selftests/bpf/Makefile b/tools/testing/selftests/bpf/Makefile
index d8d94b9bd76c..91edd0566237 100644
--- a/tools/testing/selftests/bpf/Makefile
+++ b/tools/testing/selftests/bpf/Makefile
@@ -13,7 +13,7 @@ LDLIBS += -lcap -lelf
 
 TEST_GEN_PROGS = test_verifier test_tag test_maps test_lru_map test_lpm_map test_progs
 
-TEST_GEN_FILES = test_pkt_access.o test_xdp.o test_l4lb.o
+TEST_GEN_FILES = test_pkt_access.o test_xdp.o test_l4lb.o test_tcp_estats.o
 
 TEST_PROGS := test_kmod.sh
 
@@ -34,6 +34,6 @@ $(BPFOBJ): force
 CLANG ?= clang
 
 %.o: %.c
-	$(CLANG) -I../../../include/uapi -I../../../../samples/bpf/ \
-		-D__x86_64__ -Wno-compare-distinct-pointer-types \
+	$(CLANG) -I. -I../../../include/uapi -I../../../../samples/bpf/ \
+		-Wno-compare-distinct-pointer-types \
 		-O2 -target bpf -c $< -o $@
diff --git a/tools/testing/selftests/bpf/gnu/stubs.h b/tools/testing/selftests/bpf/gnu/stubs.h
new file mode 100644
index 000000000000..719225b16626
--- /dev/null
+++ b/tools/testing/selftests/bpf/gnu/stubs.h
@@ -0,0 +1 @@
+/* dummy .h to trick /usr/include/features.h to work with 'clang -target bpf' */
diff --git a/tools/testing/selftests/bpf/test_progs.c b/tools/testing/selftests/bpf/test_progs.c
index 4ed049a0b14b..b59f5ed4ae40 100644
--- a/tools/testing/selftests/bpf/test_progs.c
+++ b/tools/testing/selftests/bpf/test_progs.c
@@ -268,6 +268,21 @@ out:
 	bpf_object__close(obj);
 }
 
+static void test_tcp_estats(void)
+{
+	const char *file = "./test_tcp_estats.o";
+	int err, prog_fd;
+	struct bpf_object *obj;
+	__u32 duration = 0;
+
+	err = bpf_prog_load(file, BPF_PROG_TYPE_TRACEPOINT, &obj, &prog_fd);
+	CHECK(err, "", "err %d errno %d\n", err, errno);
+	if (err)
+		return;
+
+	bpf_object__close(obj);
+}
+
 int main(void)
 {
 	struct rlimit rinf = { RLIM_INFINITY, RLIM_INFINITY };
@@ -277,6 +292,7 @@ int main(void)
 	test_pkt_access();
 	test_xdp();
 	test_l4lb();
+	test_tcp_estats();
 
 	printf("Summary: %d PASSED, %d FAILED\n", pass_cnt, error_cnt);
 	return 0;
diff --git a/tools/testing/selftests/bpf/test_tcp_estats.c b/tools/testing/selftests/bpf/test_tcp_estats.c
new file mode 100644
index 000000000000..bee3bbecc0c4
--- /dev/null
+++ b/tools/testing/selftests/bpf/test_tcp_estats.c
@@ -0,0 +1,258 @@
+/* Copyright (c) 2017 Facebook
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of version 2 of the GNU General Public
+ * License as published by the Free Software Foundation.
+ */
+
+/* This program shows clang/llvm is able to generate code pattern
+ * like:
+ *   _tcp_send_active_reset:
+ *      0:       bf 16 00 00 00 00 00 00         r6 = r1
+ *    ......
+ *    335:       b7 01 00 00 0f 00 00 00         r1 = 15
+ *    336:       05 00 48 00 00 00 00 00         goto 72
+ *
+ *   LBB0_3:
+ *    337:       b7 01 00 00 01 00 00 00         r1 = 1
+ *    338:       63 1a d0 ff 00 00 00 00         *(u32 *)(r10 - 48) = r1
+ *    408:       b7 01 00 00 03 00 00 00         r1 = 3
+ *
+ *   LBB0_4:
+ *    409:       71 a2 fe ff 00 00 00 00         r2 = *(u8 *)(r10 - 2)
+ *    410:       bf a7 00 00 00 00 00 00         r7 = r10
+ *    411:       07 07 00 00 b8 ff ff ff         r7 += -72
+ *    412:       bf 73 00 00 00 00 00 00         r3 = r7
+ *    413:       0f 13 00 00 00 00 00 00         r3 += r1
+ *    414:       73 23 2d 00 00 00 00 00         *(u8 *)(r3 + 45) = r2
+ *
+ * From the above code snippet, the code generated by the compiler
+ * is reasonable. The "r1" is assigned to different values in basic
+ * blocks "_tcp_send_active_reset" and "LBB0_3", and used in "LBB0_4".
+ * The verifier should be able to handle such code patterns.
+ */
+#include <string.h>
+#include <linux/bpf.h>
+#include <linux/ipv6.h>
+#include <linux/version.h>
+#include <sys/socket.h>
+#include "bpf_helpers.h"
+
+#define _(P) ({typeof(P) val = 0; bpf_probe_read(&val, sizeof(val), &P); val;})
+#define TCP_ESTATS_MAGIC 0xBAADBEEF
+
+/* This test case needs "sock" and "pt_regs" data structure.
+ * Recursively, "sock" needs "sock_common" and "inet_sock".
+ * However, this is a unit test case only for
+ * verifier purpose without bpf program execution.
+ * We can safely mock much simpler data structures, basically
+ * only taking the necessary fields from kernel headers.
+ */
+typedef __u32 __bitwise __portpair;
+typedef __u64 __bitwise __addrpair;
+
+struct sock_common {
+	unsigned short		skc_family;
+	union {
+		__addrpair	skc_addrpair;
+		struct {
+			__be32	skc_daddr;
+			__be32	skc_rcv_saddr;
+		};
+	};
+	union {
+		__portpair	skc_portpair;
+		struct {
+			__be16	skc_dport;
+			__u16	skc_num;
+		};
+	};
+	struct in6_addr		skc_v6_daddr;
+	struct in6_addr		skc_v6_rcv_saddr;
+};
+
+struct sock {
+	struct sock_common	__sk_common;
+#define sk_family		__sk_common.skc_family
+#define sk_v6_daddr		__sk_common.skc_v6_daddr
+#define sk_v6_rcv_saddr		__sk_common.skc_v6_rcv_saddr
+};
+
+struct inet_sock {
+	struct sock		sk;
+#define inet_daddr		sk.__sk_common.skc_daddr
+#define inet_dport		sk.__sk_common.skc_dport
+	__be32			inet_saddr;
+	__be16			inet_sport;
+};
+
+struct pt_regs {
+	long di;
+};
+
+static inline struct inet_sock *inet_sk(const struct sock *sk)
+{
+	return (struct inet_sock *)sk;
+}
+
+/* Define various data structures for state recording.
+ * Some fields are not used due to test simplification.
+ */
+enum tcp_estats_addrtype {
+	TCP_ESTATS_ADDRTYPE_IPV4 = 1,
+	TCP_ESTATS_ADDRTYPE_IPV6 = 2
+};
+
+enum tcp_estats_event_type {
+	TCP_ESTATS_ESTABLISH,
+	TCP_ESTATS_PERIODIC,
+	TCP_ESTATS_TIMEOUT,
+	TCP_ESTATS_RETRANSMIT_TIMEOUT,
+	TCP_ESTATS_RETRANSMIT_OTHER,
+	TCP_ESTATS_SYN_RETRANSMIT,
+	TCP_ESTATS_SYNACK_RETRANSMIT,
+	TCP_ESTATS_TERM,
+	TCP_ESTATS_TX_RESET,
+	TCP_ESTATS_RX_RESET,
+	TCP_ESTATS_WRITE_TIMEOUT,
+	TCP_ESTATS_CONN_TIMEOUT,
+	TCP_ESTATS_ACK_LATENCY,
+	TCP_ESTATS_NEVENTS,
+};
+
+struct tcp_estats_event {
+	int pid;
+	int cpu;
+	unsigned long ts;
+	unsigned int magic;
+	enum tcp_estats_event_type event_type;
+};
+
+/* The below data structure is packed in order for
+ * llvm compiler to generate expected code.
+ */
+struct tcp_estats_conn_id {
+	unsigned int localaddressType;
+	struct {
+		unsigned char data[16];
+	} localaddress;
+	struct {
+		unsigned char data[16];
+	} remaddress;
+	unsigned short    localport;
+	unsigned short    remport;
+} __attribute__((__packed__));
+
+struct tcp_estats_basic_event {
+	struct tcp_estats_event event;
+	struct tcp_estats_conn_id conn_id;
+};
+
+struct bpf_map_def SEC("maps") ev_record_map = {
+	.type = BPF_MAP_TYPE_HASH,
+	.key_size = sizeof(__u32),
+	.value_size = sizeof(struct tcp_estats_basic_event),
+	.max_entries = 1024,
+};
+
+struct dummy_tracepoint_args {
+	unsigned long long pad;
+	struct sock *sock;
+};
+
+static __always_inline void tcp_estats_ev_init(struct tcp_estats_event *event,
+					       enum tcp_estats_event_type type)
+{
+	event->magic = TCP_ESTATS_MAGIC;
+	event->ts = bpf_ktime_get_ns();
+	event->event_type = type;
+}
+
+static __always_inline void unaligned_u32_set(unsigned char *to, __u8 *from)
+{
+	to[0] = _(from[0]);
+	to[1] = _(from[1]);
+	to[2] = _(from[2]);
+	to[3] = _(from[3]);
+}
+
+static __always_inline void conn_id_ipv4_init(struct tcp_estats_conn_id *conn_id,
+					      __be32 *saddr, __be32 *daddr)
+{
+	conn_id->localaddressType = TCP_ESTATS_ADDRTYPE_IPV4;
+
+	unaligned_u32_set(conn_id->localaddress.data, (__u8 *)saddr);
+	unaligned_u32_set(conn_id->remaddress.data, (__u8 *)daddr);
+}
+
+static __always_inline void conn_id_ipv6_init(struct tcp_estats_conn_id *conn_id,
+					      __be32 *saddr, __be32 *daddr)
+{
+	conn_id->localaddressType = TCP_ESTATS_ADDRTYPE_IPV6;
+
+	unaligned_u32_set(conn_id->localaddress.data, (__u8 *)saddr);
+	unaligned_u32_set(conn_id->localaddress.data + sizeof(__u32),
+			  (__u8 *)(saddr + 1));
+	unaligned_u32_set(conn_id->localaddress.data + sizeof(__u32) * 2,
+			  (__u8 *)(saddr + 2));
+	unaligned_u32_set(conn_id->localaddress.data + sizeof(__u32) * 3,
+			  (__u8 *)(saddr + 3));
+
+	unaligned_u32_set(conn_id->remaddress.data,
+			  (__u8 *)(daddr));
+	unaligned_u32_set(conn_id->remaddress.data + sizeof(__u32),
+			  (__u8 *)(daddr + 1));
+	unaligned_u32_set(conn_id->remaddress.data + sizeof(__u32) * 2,
+			  (__u8 *)(daddr + 2));
+	unaligned_u32_set(conn_id->remaddress.data + sizeof(__u32) * 3,
+			  (__u8 *)(daddr + 3));
+}
+
+static __always_inline void tcp_estats_conn_id_init(struct tcp_estats_conn_id *conn_id,
+						    struct sock *sk)
+{
+	conn_id->localport = _(inet_sk(sk)->inet_sport);
+	conn_id->remport = _(inet_sk(sk)->inet_dport);
+
+	if (_(sk->sk_family) == AF_INET6)
+		conn_id_ipv6_init(conn_id,
+				  sk->sk_v6_rcv_saddr.s6_addr32,
+				  sk->sk_v6_daddr.s6_addr32);
+	else
+		conn_id_ipv4_init(conn_id,
+				  &inet_sk(sk)->inet_saddr,
+				  &inet_sk(sk)->inet_daddr);
+}
+
+static __always_inline void tcp_estats_init(struct sock *sk,
+					    struct tcp_estats_event *event,
+					    struct tcp_estats_conn_id *conn_id,
+					    enum tcp_estats_event_type type)
+{
+	tcp_estats_ev_init(event, type);
+	tcp_estats_conn_id_init(conn_id, sk);
+}
+
+static __always_inline void send_basic_event(struct sock *sk,
+					     enum tcp_estats_event_type type)
+{
+	struct tcp_estats_basic_event ev;
+	__u32 key = bpf_get_prandom_u32();
+
+	memset(&ev, 0, sizeof(ev));
+	tcp_estats_init(sk, &ev.event, &ev.conn_id, type);
+	bpf_map_update_elem(&ev_record_map, &key, &ev, BPF_ANY);
+}
+
+SEC("dummy_tracepoint")
+int _dummy_tracepoint(struct dummy_tracepoint_args *arg)
+{
+	if (!arg->sock)
+		return 0;
+
+	send_basic_event(arg->sock, TCP_ESTATS_TX_RESET);
+	return 0;
+}
+
+char _license[] SEC("license") = "GPL";
+__u32 _version SEC("version") = 1; /* ignored by tracepoints, required by libbpf.a */