Merge git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net-next

Pull networking updates from David Miller: 1) Allow setting bluetooth L2CAP modes via socket option, from Luiz Augusto von Dentz. 2) Add GSO partial support to igc, from Sasha Neftin. 3) Several cleanups and improvements to r8169 from Heiner Kallweit. 4) Add IF_OPER_TESTING link state and use it when ethtool triggers a device self-test. From Andrew Lunn. 5) Start moving away from custom driver versions, use the globally defined kernel version instead, from Leon Romanovsky. 6) Support GRO vis gro_cells in DSA layer, from Alexander Lobakin. 7) Allow hard IRQ deferral during NAPI, from Eric Dumazet. 8) Add sriov and vf support to hinic, from Luo bin. 9) Support Media Redundancy Protocol (MRP) in the bridging code, from Horatiu Vultur. 10) Support netmap in the nft_nat code, from Pablo Neira Ayuso. 11) Allow UDPv6 encapsulation of ESP in the ipsec code, from Sabrina Dubroca. Also add ipv6 support for espintcp. 12) Lots of ReST conversions of the networking documentation, from Mauro Carvalho Chehab. 13) Support configuration of ethtool rxnfc flows in bcmgenet driver, from Doug Berger. 14) Allow to dump cgroup id and filter by it in inet_diag code, from Dmitry Yakunin. 15) Add infrastructure to export netlink attribute policies to userspace, from Johannes Berg. 16) Several optimizations to sch_fq scheduler, from Eric Dumazet. 17) Fallback to the default qdisc if qdisc init fails because otherwise a packet scheduler init failure will make a device inoperative. From Jesper Dangaard Brouer. 18) Several RISCV bpf jit optimizations, from Luke Nelson. 19) Correct the return type of the ->ndo_start_xmit() method in several drivers, it's netdev_tx_t but many drivers were using 'int'. From Yunjian Wang. 20) Add an ethtool interface for PHY master/slave config, from Oleksij Rempel. 21) Add BPF iterators, from Yonghang Song. 22) Add cable test infrastructure, including ethool interfaces, from Andrew Lunn. Marvell PHY driver is the first to support this facility. 23) Remove zero-length arrays all over, from Gustavo A. R. Silva. 24) Calculate and maintain an explicit frame size in XDP, from Jesper Dangaard Brouer. 25) Add CAP_BPF, from Alexei Starovoitov. 26) Support terse dumps in the packet scheduler, from Vlad Buslov. 27) Support XDP_TX bulking in dpaa2 driver, from Ioana Ciornei. 28) Add devm_register_netdev(), from Bartosz Golaszewski. 29) Minimize qdisc resets, from Cong Wang. 30) Get rid of kernel_getsockopt and kernel_setsockopt in order to eliminate set_fs/get_fs calls. From Christoph Hellwig. * git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net-next: (2517 commits) selftests: net: ip_defrag: ignore EPERM net_failover: fixed rollback in net_failover_open() Revert "tipc: Fix potential tipc_aead refcnt leak in tipc_crypto_rcv" Revert "tipc: Fix potential tipc_node refcnt leak in tipc_rcv" vmxnet3: allow rx flow hash ops only when rss is enabled hinic: add set_channels ethtool_ops support selftests/bpf: Add a default $(CXX) value tools/bpf: Don't use $(COMPILE.c) bpf, selftests: Use bpf_probe_read_kernel s390/bpf: Use bcr 0,%0 as tail call nop filler s390/bpf: Maintain 8-byte stack alignment selftests/bpf: Fix verifier test selftests/bpf: Fix sample_cnt shared between two threads bpf, selftests: Adapt cls_redirect to call csum_level helper bpf: Add csum_level helper for fixing up csum levels bpf: Fix up bpf_skb_adjust_room helper's skb csum setting sfc: add missing annotation for efx_ef10_try_update_nic_stats_vf() crypto/chtls: IPv6 support for inline TLS Crypto/chcr: Fixes a coccinile check error Crypto/chcr: Fixes compilations warnings ...

Linus Torvalds committed 6 years ago

Commit cb8e59c2 parents 2e63f6c 065fcfd2097 files · +151766 −46213Browse files

Showing the changes to tools/include/uapi (2 of 2097 files). Show all

Unified · Split

2 files

tools/include/uapi/linux/bpf.h +344 −46
tools/include/uapi/linux/if_link.h +1 −0

tools/include/uapi/linux/bpf.h+344−46

@@ -73,7 +73,7 @@ struct bpf_insn { · 72 unmodified lines
73	73	/* Key of an a BPF_MAP_TYPE_LPM_TRIE entry */
74	74	struct bpf_lpm_trie_key {
75	75	__u32 prefixlen; /* up to 32 for AF_INET, 128 for AF_INET6 */
76		− __u8 data[]; /* Arbitrary size */
	76	+ __u8 data[0]; /* Arbitrary size */
77	77	};
78	78
79	79	struct bpf_cgroup_storage_key {
@@ -113,6 +113,10 @@ enum bpf_cmd { · 33 unmodified lines
113	113	BPF_MAP_DELETE_BATCH,
114	114	BPF_LINK_CREATE,
115	115	BPF_LINK_UPDATE,
	116	+ BPF_LINK_GET_FD_BY_ID,
	117	+ BPF_LINK_GET_NEXT_ID,
	118	+ BPF_ENABLE_STATS,
	119	+ BPF_ITER_CREATE,
116	120	};
117	121
118	122	enum bpf_map_type {
@@ -143,6 +147,7 @@ enum bpf_map_type { · 24 unmodified lines
143	147	BPF_MAP_TYPE_SK_STORAGE,
144	148	BPF_MAP_TYPE_DEVMAP_HASH,
145	149	BPF_MAP_TYPE_STRUCT_OPS,
	150	+ BPF_MAP_TYPE_RINGBUF,
146	151	};
147	152
148	153	/* Note that tracing related programs such as
@@ -215,11 +220,28 @@ enum bpf_attach_type { · 66 unmodified lines
215	220	BPF_TRACE_FEXIT,
216	221	BPF_MODIFY_RETURN,
217	222	BPF_LSM_MAC,
	223	+ BPF_TRACE_ITER,
	224	+ BPF_CGROUP_INET4_GETPEERNAME,
	225	+ BPF_CGROUP_INET6_GETPEERNAME,
	226	+ BPF_CGROUP_INET4_GETSOCKNAME,
	227	+ BPF_CGROUP_INET6_GETSOCKNAME,
	228	+ BPF_XDP_DEVMAP,
218	229	__MAX_BPF_ATTACH_TYPE
219	230	};
220	231
221	232	#define MAX_BPF_ATTACH_TYPE __MAX_BPF_ATTACH_TYPE
222	233
	234	+enum bpf_link_type {
	235	+ BPF_LINK_TYPE_UNSPEC = 0,
	236	+ BPF_LINK_TYPE_RAW_TRACEPOINT = 1,
	237	+ BPF_LINK_TYPE_TRACING = 2,
	238	+ BPF_LINK_TYPE_CGROUP = 3,
	239	+ BPF_LINK_TYPE_ITER = 4,
	240	+ BPF_LINK_TYPE_NETNS = 5,
	241	+
	242	+ MAX_BPF_LINK_TYPE,
	243	+};
	244	+
223	245	/* cgroup-bpf attach flags used in BPF_PROG_ATTACH command
224	246	*
225	247	* NONE(default): No further bpf programs allowed in the subtree.
@@ -379,6 +401,12 @@ enum { · 153 unmodified lines
379	401	*/
380	402	#define BPF_F_QUERY_EFFECTIVE (1U << 0)
381	403
	404	+/* type for BPF_ENABLE_STATS */
	405	+enum bpf_stats_type {
	406	+ /* enabled run_time_ns and run_cnt */
	407	+ BPF_STATS_RUN_TIME = 0,
	408	+};
	409	+
382	410	enum bpf_stack_build_id_status {
383	411	/* user space need an empty entry to identify end of a trace */
384	412	BPF_STACK_BUILD_ID_EMPTY = 0,
@@ -523,6 +551,7 @@ union bpf_attr { · 138 unmodified lines
523	551	__u32 prog_id;
524	552	__u32 map_id;
525	553	__u32 btf_id;
	554	+ __u32 link_id;
526	555	};
527	556	__u32 next_id;
528	557	__u32 open_flags;
@@ -589,6 +618,15 @@ union bpf_attr { · 60 unmodified lines
589	618	__u32 old_prog_fd;
590	619	} link_update;
591	620
	621	+ struct { /* struct used by BPF_ENABLE_STATS command */
	622	+ __u32 type;
	623	+ } enable_stats;
	624	+
	625	+ struct { /* struct used by BPF_ITER_CREATE command */
	626	+ __u32 link_fd;
	627	+ __u32 flags;
	628	+ } iter_create;
	629	+
592	630	} __attribute__((aligned(8)));
593	631
594	632	/* The description below is an attempt at providing documentation to eBPF
@@ -644,14 +682,16 @@ union bpf_attr { · 49 unmodified lines
644	682	* For tracing programs, safely attempt to read size bytes from
645	683	* kernel space address unsafe_ptr and store the data in dst.
646	684	*
647		− * Generally, use bpf_probe_read_user() or bpf_probe_read_kernel()
648		− * instead.
	685	+ * Generally, use bpf_probe_read_user\ () or
	686	+ * bpf_probe_read_kernel\ () instead.
649	687	* Return
650	688	* 0 on success, or a negative error in case of failure.
651	689	*
652	690	* u64 bpf_ktime_get_ns(void)
653	691	* Description
654	692	* Return the time elapsed since system boot, in nanoseconds.
	693	+ * Does not include time the system was suspended.
	694	+ * See: clock_gettime\ (CLOCK_MONOTONIC)
655	695	* Return
656	696	* Current ktime.
657	697	*
@@ -1510,11 +1550,11 @@ union bpf_attr { · 852 unmodified lines
1510	1550	* int bpf_probe_read_str(void dst, u32 size, const void unsafe_ptr)
1511	1551	* Description
1512	1552	* Copy a NUL terminated string from an unsafe kernel address
1513		− * unsafe_ptr to dst. See bpf_probe_read_kernel_str() for
	1553	+ * unsafe_ptr to dst. See bpf_probe_read_kernel_str\ () for
1514	1554	* more details.
1515	1555	*
1516		− * Generally, use bpf_probe_read_user_str() or bpf_probe_read_kernel_str()
1517		− * instead.
	1556	+ * Generally, use bpf_probe_read_user_str\ () or
	1557	+ * bpf_probe_read_kernel_str\ () instead.
1518	1558	* Return
1519	1559	* On success, the strictly positive length of the string,
1520	1560	* including the trailing NUL character. On error, a negative
@@ -1542,7 +1582,7 @@ union bpf_attr { · 21 unmodified lines
1542	1582	*
1543	1583	* u64 bpf_get_socket_cookie(struct bpf_sock_ops *ctx)
1544	1584	* Description
1545		− * Equivalent to bpf_get_socket_cookie() helper that accepts
	1585	+ * Equivalent to bpf_get_socket_cookie\ () helper that accepts
1546	1586	* skb, but gets socket from struct bpf_sock_ops context.
1547	1587	* Return
1548	1588	* A 8-byte long non-decreasing number.
@@ -1562,7 +1602,7 @@ union bpf_attr { · 13 unmodified lines
1562	1602	* Return
1563	1603	* 0
1564	1604	*
1565		− * int bpf_setsockopt(struct bpf_sock_ops bpf_socket, int level, int optname, void optval, int optlen)
	1605	+ * int bpf_setsockopt(void bpf_socket, int level, int optname, void optval, int optlen)
1566	1606	* Description
1567	1607	* Emulate a call to setsockopt() on the socket associated to
1568	1608	* bpf_socket, which must be a full socket. The level at
@@ -1570,6 +1610,12 @@ union bpf_attr { · 1 unmodified line
1570	1610	* must be specified, see setsockopt(2) for more information.
1571	1611	* The option value of length optlen is pointed by optval.
1572	1612	*
	1613	+ * bpf_socket should be one of the following:
	1614	+ *
	1615	+ * * struct bpf_sock_ops for BPF_PROG_TYPE_SOCK_OPS.
	1616	+ * * struct bpf_sock_addr for BPF_CGROUP_INET4_CONNECT
	1617	+ * and BPF_CGROUP_INET6_CONNECT.
	1618	+ *
1573	1619	* This helper actually implements a subset of setsockopt().
1574	1620	* It supports the following level\ s:
1575	1621	*
@@ -1589,6 +1635,13 @@ union bpf_attr { · 13 unmodified lines
1589	1635	* Grow or shrink the room for data in the packet associated to
1590	1636	* skb by len_diff, and according to the selected mode.
1591	1637	*
	1638	+ * By default, the helper will reset any offloaded checksum
	1639	+ * indicator of the skb to CHECKSUM_NONE. This can be avoided
	1640	+ * by the following flag:
	1641	+ *
	1642	+ * * BPF_F_ADJ_ROOM_NO_CSUM_RESET: Do not reset offloaded
	1643	+ * checksum data of the skb to CHECKSUM_NONE.
	1644	+ *
1592	1645	* There are two supported modes at this time:
1593	1646	*
1594	1647	* * BPF_ADJ_ROOM_MAC: Adjust room at the mac layer
@@ -1634,12 +1687,12 @@ union bpf_attr { · 39 unmodified lines
1634	1687	*
1635	1688	* The lower two bits of flags are used as the return code if
1636	1689	* the map lookup fails. This is so that the return value can be
1637		− * one of the XDP program return codes up to XDP_TX, as chosen by
1638		− * the caller. Any higher bits in the flags argument must be
	1690	+ * one of the XDP program return codes up to XDP_TX, as chosen
	1691	+ * by the caller. Any higher bits in the flags argument must be
1639	1692	* unset.
1640	1693	*
1641		− * See also bpf_redirect(), which only supports redirecting to an
1642		− * ifindex, but doesn't require a map to do so.
	1694	+ * See also bpf_redirect\ (), which only supports redirecting
	1695	+ * to an ifindex, but doesn't require a map to do so.
1643	1696	* Return
1644	1697	* XDP_REDIRECT on success, or the value of the two lower bits
1645	1698	* of the flags argument on error.
@@ -1747,7 +1800,7 @@ union bpf_attr { · 101 unmodified lines
1747	1800	* the time running for event since last normalization. The
1748	1801	* enabled and running times are accumulated since the perf event
1749	1802	* open. To achieve scaling factor between two invocations of an
1750		− * eBPF program, users can can use CPU id as the key (which is
	1803	+ * eBPF program, users can use CPU id as the key (which is
1751	1804	* typical for perf array usage model) to remember the previous
1752	1805	* value and do the calculation inside the eBPF program.
1753	1806	* Return
@@ -1764,7 +1817,7 @@ union bpf_attr { · 10 unmodified lines
1764	1817	* Return
1765	1818	* 0 on success, or a negative error in case of failure.
1766	1819	*
1767		− * int bpf_getsockopt(struct bpf_sock_ops bpf_socket, int level, int optname, void optval, int optlen)
	1820	+ * int bpf_getsockopt(void bpf_socket, int level, int optname, void optval, int optlen)
1768	1821	* Description
1769	1822	* Emulate a call to getsockopt() on the socket associated to
1770	1823	* bpf_socket, which must be a full socket. The level at
@@ -1773,6 +1826,12 @@ union bpf_attr { · 2 unmodified lines
1773	1826	* The retrieved value is stored in the structure pointed by
1774	1827	* opval and of length optlen.
1775	1828	*
	1829	+ * bpf_socket should be one of the following:
	1830	+ *
	1831	+ * * struct bpf_sock_ops for BPF_PROG_TYPE_SOCK_OPS.
	1832	+ * * struct bpf_sock_addr for BPF_CGROUP_INET4_CONNECT
	1833	+ * and BPF_CGROUP_INET6_CONNECT.
	1834	+ *
1776	1835	* This helper actually implements a subset of getsockopt().
1777	1836	* It supports the following level\ s:
1778	1837	*
@@ -1790,7 +1849,7 @@ union bpf_attr { · 11 unmodified lines
1790	1849	* The first argument is the context regs on which the kprobe
1791	1850	* works.
1792	1851	*
1793		− * This helper works by setting setting the PC (program counter)
	1852	+ * This helper works by setting the PC (program counter)
1794	1853	* to an override function which is run in place of the original
1795	1854	* probed function. This means the probed function is not run at
1796	1855	* all. The replacement function just returns with the required
@@ -1959,18 +2018,19 @@ union bpf_attr { · 162 unmodified lines
1959	2018	*
1960	2019	* This helper works for IPv4 and IPv6, TCP and UDP sockets. The
1961	2020	* domain (addr\ ->sa_family) must be AF_INET (or
1962		− * AF_INET6). Looking for a free port to bind to can be
1963		− * expensive, therefore binding to port is not permitted by the
1964		− * helper: addr\ ->sin_port (or sin6_port, respectively)
1965		− * must be set to zero.
	2021	+ * AF_INET6). It's advised to pass zero port (sin_port
	2022	+ * or sin6_port) which triggers IP_BIND_ADDRESS_NO_PORT-like
	2023	+ * behavior and lets the kernel efficiently pick up an unused
	2024	+ * port as long as 4-tuple is unique. Passing non-zero port might
	2025	+ * lead to degraded performance.
1966	2026	* Return
1967	2027	* 0 on success, or a negative error in case of failure.
1968	2028	*
1969	2029	* int bpf_xdp_adjust_tail(struct xdp_buff *xdp_md, int delta)
1970	2030	* Description
1971	2031	* Adjust (move) xdp_md\ ->data_end by delta bytes. It is
1972		− * only possible to shrink the packet as of this writing,
1973		− * therefore delta must be a negative integer.
	2032	+ * possible to both shrink and grow the packet tail.
	2033	+ * Shrink done via delta being a negative integer.
1974	2034	*
1975	2035	* A call to this helper is susceptible to change the underlying
1976	2036	* packet buffer. Therefore, at load time, all checks on pointers
@@ -2256,7 +2316,7 @@ union bpf_attr { · 279 unmodified lines
2256	2316	* bpf_rc_keydown\ () again with the same values, or calling
2257	2317	* bpf_rc_repeat\ ().
2258	2318	*
2259		− * Some protocols include a toggle bit, in case the button was
	2319	+ * Some protocols include a toggle bit, in case the button was
2260	2320	* released and pressed again between consecutive scancodes.
2261	2321	*
2262	2322	* The ctx should point to the lirc sample as passed into
@@ -2602,7 +2662,6 @@ union bpf_attr { · 339 unmodified lines
2602	2662	*
2603	2663	* th points to the start of the TCP header, while th_len
2604	2664	* contains sizeof\ (struct tcphdr).
2605		− *
2606	2665	* Return
2607	2666	* 0 if iph and th are a valid SYN cookie ACK, or a negative
2608	2667	* error otherwise.
@@ -2785,7 +2844,6 @@ union bpf_attr { · 176 unmodified lines
2785	2844	*
2786	2845	* th points to the start of the TCP header, while th_len
2787	2846	* contains the length of the TCP header.
2788		− *
2789	2847	* Return
2790	2848	* On success, lower 32 bits hold the generated SYN cookie in
2791	2849	* followed by 16 bits which hold the MSS value for that cookie,
@@ -2868,7 +2926,7 @@ union bpf_attr { · 76 unmodified lines
2868	2926	* // size, after checking its boundaries.
2869	2927	* }
2870	2928	*
2871		− * In comparison, using bpf_probe_read_user() helper here
	2929	+ * In comparison, using bpf_probe_read_user\ () helper here
2872	2930	* instead to read the string would require to estimate the length
2873	2931	* at compile time, and would often result in copying more memory
2874	2932	* than necessary.
@@ -2886,14 +2944,14 @@ union bpf_attr { · 11 unmodified lines
2886	2944	* int bpf_probe_read_kernel_str(void dst, u32 size, const void unsafe_ptr)
2887	2945	* Description
2888	2946	* Copy a NUL terminated string from an unsafe kernel address unsafe_ptr
2889		− * to dst. Same semantics as with bpf_probe_read_user_str() apply.
	2947	+ * to dst. Same semantics as with bpf_probe_read_user_str\ () apply.
2890	2948	* Return
2891		− * On success, the strictly positive length of the string, including
	2949	+ * On success, the strictly positive length of the string, including
2892	2950	* the trailing NUL character. On error, a negative value.
2893	2951	*
2894	2952	* int bpf_tcp_send_ack(void *tp, u32 rcv_nxt)
2895	2953	* Description
2896		− * Send out a tcp-ack. tp is the in-kernel struct tcp_sock.
	2954	+ * Send out a tcp-ack. tp is the in-kernel struct tcp_sock.
2897	2955	* rcv_nxt is the ack_seq to be sent out.
2898	2956	* Return
2899	2957	* 0 on success, or a negative error in case of failure.
@@ -2921,19 +2979,19 @@ union bpf_attr { · 21 unmodified lines
2921	2979	* int bpf_read_branch_records(struct bpf_perf_event_data ctx, void buf, u32 size, u64 flags)
2922	2980	* Description
2923	2981	* For an eBPF program attached to a perf event, retrieve the
2924		− * branch records (struct perf_branch_entry) associated to ctx
2925		− * and store it in the buffer pointed by buf up to size
	2982	+ * branch records (struct perf_branch_entry) associated to ctx
	2983	+ * and store it in the buffer pointed by buf up to size
2926	2984	* size bytes.
2927	2985	* Return
2928	2986	* On success, number of bytes written to buf. On error, a
2929	2987	* negative value.
2930	2988	*
2931	2989	* The flags can be set to BPF_F_GET_BRANCH_RECORDS_SIZE to
2932		− * instead return the number of bytes required to store all the
	2990	+ * instead return the number of bytes required to store all the
2933	2991	* branch entries. If this flag is set, buf may be NULL.
2934	2992	*
2935	2993	* -EINVAL if arguments invalid or size not a multiple
2936		− * of sizeof(struct perf_branch_entry).
	2994	+ * of sizeof\ (struct perf_branch_entry\ ).
2937	2995	*
2938	2996	* -ENOENT if architecture does not support branch records.
2939	2997	*
@@ -2941,8 +2999,8 @@ union bpf_attr { · 1 unmodified line
2941	2999	* Description
2942	3000	* Returns 0 on success, values for pid and tgid as seen from the current
2943	3001	* namespace will be returned in nsdata.
2944		− *
2945		− * On failure, the returned value is one of the following:
	3002	+ * Return
	3003	+ * 0 on success, or one of the following in case of failure:
2946	3004	*
2947	3005	* -EINVAL if dev and inum supplied don't match dev_t and inode number
2948	3006	* with nsfs of current task, or if dev conversion to dev_t lost high bits.
@@ -2981,8 +3039,8 @@ union bpf_attr { · 32 unmodified lines
2981	3039	* a global identifier that can be assumed unique. If ctx is
2982	3040	* NULL, then the helper returns the cookie for the initial
2983	3041	* network namespace. The cookie itself is very similar to that
2984		− * of bpf_get_socket_cookie() helper, but for network namespaces
2985		− * instead of sockets.
	3042	+ * of bpf_get_socket_cookie\ () helper, but for network
	3043	+ * namespaces instead of sockets.
2986	3044	* Return
2987	3045	* A 8-byte long opaque number.
2988	3046	*
@@ -3017,14 +3075,183 @@ union bpf_attr { · 28 unmodified lines
3017	3075	*
3018	3076	* The flags argument must be zero.
3019	3077	* Return
3020		− * 0 on success, or a negative errno in case of failure.
3021		− *
3022		− * * -EINVAL Unsupported flags specified.
3023		− * * -ENOENT Socket is unavailable for assignment.
3024		− * * -ENETUNREACH Socket is unreachable (wrong netns).
3025		− * * -EOPNOTSUPP Unsupported operation, for example a
3026		− * call from outside of TC ingress.
3027		− * * -ESOCKTNOSUPPORT Socket type not supported (reuseport).
	3078	+ * 0 on success, or a negative error in case of failure:
	3079	+ *
	3080	+ * -EINVAL if specified flags are not supported.
	3081	+ *
	3082	+ * -ENOENT if the socket is unavailable for assignment.
	3083	+ *
	3084	+ * -ENETUNREACH if the socket is unreachable (wrong netns).
	3085	+ *
	3086	+ * -EOPNOTSUPP if the operation is not supported, for example
	3087	+ * a call from outside of TC ingress.
	3088	+ *
	3089	+ * -ESOCKTNOSUPPORT if the socket type is not supported
	3090	+ * (reuseport).
	3091	+ *
	3092	+ * u64 bpf_ktime_get_boot_ns(void)
	3093	+ * Description
	3094	+ * Return the time elapsed since system boot, in nanoseconds.
	3095	+ * Does include the time the system was suspended.
	3096	+ * See: clock_gettime\ (CLOCK_BOOTTIME)
	3097	+ * Return
	3098	+ * Current ktime.
	3099	+ *
	3100	+ * int bpf_seq_printf(struct seq_file m, const char fmt, u32 fmt_size, const void *data, u32 data_len)
	3101	+ * Description
	3102	+ * bpf_seq_printf\ () uses seq_file seq_printf\ () to print
	3103	+ * out the format string.
	3104	+ * The m represents the seq_file. The fmt and fmt_size are for
	3105	+ * the format string itself. The data and data_len are format string
	3106	+ * arguments. The data are a u64 array and corresponding format string
	3107	+ * values are stored in the array. For strings and pointers where pointees
	3108	+ * are accessed, only the pointer values are stored in the data array.
	3109	+ * The data_len is the size of data in bytes.
	3110	+ *
	3111	+ * Formats %s, %p{i,I}{4,6} requires to read kernel memory.
	3112	+ * Reading kernel memory may fail due to either invalid address or
	3113	+ * valid address but requiring a major memory fault. If reading kernel memory
	3114	+ * fails, the string for %s will be an empty string, and the ip
	3115	+ * address for %p{i,I}{4,6} will be 0. Not returning error to
	3116	+ * bpf program is consistent with what bpf_trace_printk\ () does for now.
	3117	+ * Return
	3118	+ * 0 on success, or a negative error in case of failure:
	3119	+ *
	3120	+ * -EBUSY if per-CPU memory copy buffer is busy, can try again
	3121	+ * by returning 1 from bpf program.
	3122	+ *
	3123	+ * -EINVAL if arguments are invalid, or if fmt is invalid/unsupported.
	3124	+ *
	3125	+ * -E2BIG if fmt contains too many format specifiers.
	3126	+ *
	3127	+ * -EOVERFLOW if an overflow happened: The same object will be tried again.
	3128	+ *
	3129	+ * int bpf_seq_write(struct seq_file m, const void data, u32 len)
	3130	+ * Description
	3131	+ * bpf_seq_write\ () uses seq_file seq_write\ () to write the data.
	3132	+ * The m represents the seq_file. The data and len represent the
	3133	+ * data to write in bytes.
	3134	+ * Return
	3135	+ * 0 on success, or a negative error in case of failure:
	3136	+ *
	3137	+ * -EOVERFLOW if an overflow happened: The same object will be tried again.
	3138	+ *
	3139	+ * u64 bpf_sk_cgroup_id(struct bpf_sock *sk)
	3140	+ * Description
	3141	+ * Return the cgroup v2 id of the socket sk.
	3142	+ *
	3143	+ * sk must be a non-NULL pointer to a full socket, e.g. one
	3144	+ * returned from bpf_sk_lookup_xxx\ (),
	3145	+ * bpf_sk_fullsock\ (), etc. The format of returned id is
	3146	+ * same as in bpf_skb_cgroup_id\ ().
	3147	+ *
	3148	+ * This helper is available only if the kernel was compiled with
	3149	+ * the CONFIG_SOCK_CGROUP_DATA configuration option.
	3150	+ * Return
	3151	+ * The id is returned or 0 in case the id could not be retrieved.
	3152	+ *
	3153	+ * u64 bpf_sk_ancestor_cgroup_id(struct bpf_sock *sk, int ancestor_level)
	3154	+ * Description
	3155	+ * Return id of cgroup v2 that is ancestor of cgroup associated
	3156	+ * with the sk at the ancestor_level. The root cgroup is at
	3157	+ * ancestor_level zero and each step down the hierarchy
	3158	+ * increments the level. If ancestor_level == level of cgroup
	3159	+ * associated with sk, then return value will be same as that
	3160	+ * of bpf_sk_cgroup_id\ ().
	3161	+ *
	3162	+ * The helper is useful to implement policies based on cgroups
	3163	+ * that are upper in hierarchy than immediate cgroup associated
	3164	+ * with sk.
	3165	+ *
	3166	+ * The format of returned id and helper limitations are same as in
	3167	+ * bpf_sk_cgroup_id\ ().
	3168	+ * Return
	3169	+ * The id is returned or 0 in case the id could not be retrieved.
	3170	+ *
	3171	+ * void bpf_ringbuf_output(void ringbuf, void *data, u64 size, u64 flags)
	3172	+ * Description
	3173	+ * Copy size bytes from data into a ring buffer ringbuf.
	3174	+ * If BPF_RB_NO_WAKEUP is specified in flags, no notification of
	3175	+ * new data availability is sent.
	3176	+ * IF BPF_RB_FORCE_WAKEUP is specified in flags, notification of
	3177	+ * new data availability is sent unconditionally.
	3178	+ * Return
	3179	+ * 0, on success;
	3180	+ * < 0, on error.
	3181	+ *
	3182	+ * void bpf_ringbuf_reserve(void ringbuf, u64 size, u64 flags)
	3183	+ * Description
	3184	+ * Reserve size bytes of payload in a ring buffer ringbuf.
	3185	+ * Return
	3186	+ * Valid pointer with size bytes of memory available; NULL,
	3187	+ * otherwise.
	3188	+ *
	3189	+ * void bpf_ringbuf_submit(void *data, u64 flags)
	3190	+ * Description
	3191	+ * Submit reserved ring buffer sample, pointed to by data.
	3192	+ * If BPF_RB_NO_WAKEUP is specified in flags, no notification of
	3193	+ * new data availability is sent.
	3194	+ * IF BPF_RB_FORCE_WAKEUP is specified in flags, notification of
	3195	+ * new data availability is sent unconditionally.
	3196	+ * Return
	3197	+ * Nothing. Always succeeds.
	3198	+ *
	3199	+ * void bpf_ringbuf_discard(void *data, u64 flags)
	3200	+ * Description
	3201	+ * Discard reserved ring buffer sample, pointed to by data.
	3202	+ * If BPF_RB_NO_WAKEUP is specified in flags, no notification of
	3203	+ * new data availability is sent.
	3204	+ * IF BPF_RB_FORCE_WAKEUP is specified in flags, notification of
	3205	+ * new data availability is sent unconditionally.
	3206	+ * Return
	3207	+ * Nothing. Always succeeds.
	3208	+ *
	3209	+ * u64 bpf_ringbuf_query(void *ringbuf, u64 flags)
	3210	+ * Description
	3211	+ * Query various characteristics of provided ring buffer. What
	3212	+ * exactly is queries is determined by flags:
	3213	+ * - BPF_RB_AVAIL_DATA - amount of data not yet consumed;
	3214	+ * - BPF_RB_RING_SIZE - the size of ring buffer;
	3215	+ * - BPF_RB_CONS_POS - consumer position (can wrap around);
	3216	+ * - BPF_RB_PROD_POS - producer(s) position (can wrap around);
	3217	+ * Data returned is just a momentary snapshots of actual values
	3218	+ * and could be inaccurate, so this facility should be used to
	3219	+ * power heuristics and for reporting, not to make 100% correct
	3220	+ * calculation.
	3221	+ * Return
	3222	+ * Requested value, or 0, if flags are not recognized.
	3223	+ *
	3224	+ * int bpf_csum_level(struct sk_buff *skb, u64 level)
	3225	+ * Description
	3226	+ * Change the skbs checksum level by one layer up or down, or
	3227	+ * reset it entirely to none in order to have the stack perform
	3228	+ * checksum validation. The level is applicable to the following
	3229	+ * protocols: TCP, UDP, GRE, SCTP, FCOE. For example, a decap of
	3230	+ * \| ETH \| IP \| UDP \| GUE \| IP \| TCP \| into \| ETH \| IP \| TCP \|
	3231	+ * through bpf_skb_adjust_room\ () helper with passing in
	3232	+ * BPF_F_ADJ_ROOM_NO_CSUM_RESET flag would require one call
	3233	+ * to bpf_csum_level\ () with BPF_CSUM_LEVEL_DEC since
	3234	+ * the UDP header is removed. Similarly, an encap of the latter
	3235	+ * into the former could be accompanied by a helper call to
	3236	+ * bpf_csum_level\ () with BPF_CSUM_LEVEL_INC if the
	3237	+ * skb is still intended to be processed in higher layers of the
	3238	+ * stack instead of just egressing at tc.
	3239	+ *
	3240	+ * There are three supported level settings at this time:
	3241	+ *
	3242	+ * * BPF_CSUM_LEVEL_INC: Increases skb->csum_level for skbs
	3243	+ * with CHECKSUM_UNNECESSARY.
	3244	+ * * BPF_CSUM_LEVEL_DEC: Decreases skb->csum_level for skbs
	3245	+ * with CHECKSUM_UNNECESSARY.
	3246	+ * * BPF_CSUM_LEVEL_RESET: Resets skb->csum_level to 0 and
	3247	+ * sets CHECKSUM_NONE to force checksum validation by the stack.
	3248	+ * * BPF_CSUM_LEVEL_QUERY: No-op, returns the current
	3249	+ * skb->csum_level.
	3250	+ * Return
	3251	+ * 0 on success, or a negative error in case of failure. In the
	3252	+ * case of BPF_CSUM_LEVEL_QUERY, the current skb->csum_level
	3253	+ * is returned or the error code -EACCES in case the skb is not
	3254	+ * subject to CHECKSUM_UNNECESSARY.
3028	3255	*/
3029	3256	#define __BPF_FUNC_MAPPER(FN) \
3030	3257	FN(unspec), \
@@ -3151,7 +3378,18 @@ union bpf_attr { · 120 unmodified lines
3151	3378	FN(xdp_output), \
3152	3379	FN(get_netns_cookie), \
3153	3380	FN(get_current_ancestor_cgroup_id), \
3154		− FN(sk_assign),
	3381	+ FN(sk_assign), \
	3382	+ FN(ktime_get_boot_ns), \
	3383	+ FN(seq_printf), \
	3384	+ FN(seq_write), \
	3385	+ FN(sk_cgroup_id), \
	3386	+ FN(sk_ancestor_cgroup_id), \
	3387	+ FN(ringbuf_output), \
	3388	+ FN(ringbuf_reserve), \
	3389	+ FN(ringbuf_submit), \
	3390	+ FN(ringbuf_discard), \
	3391	+ FN(ringbuf_query), \
	3392	+ FN(csum_level),
3155	3393
3156	3394	/* integer value in 'imm' field of BPF_CALL instruction selects which helper
3157	3395	* function eBPF program intends to call
@@ -3228,6 +3466,14 @@ enum { · 70 unmodified lines
3228	3466	BPF_F_CURRENT_NETNS = (-1L),
3229	3467	};
3230	3468
	3469	+/* BPF_FUNC_csum_level level values. */
	3470	+enum {
	3471	+ BPF_CSUM_LEVEL_QUERY,
	3472	+ BPF_CSUM_LEVEL_INC,
	3473	+ BPF_CSUM_LEVEL_DEC,
	3474	+ BPF_CSUM_LEVEL_RESET,
	3475	+};
	3476	+
3231	3477	/* BPF_FUNC_skb_adjust_room flags. */
3232	3478	enum {
3233	3479	BPF_F_ADJ_ROOM_FIXED_GSO = (1ULL << 0),
@@ -3235,6 +3481,7 @@ enum { · 1 unmodified line
3235	3481	BPF_F_ADJ_ROOM_ENCAP_L3_IPV6 = (1ULL << 2),
3236	3482	BPF_F_ADJ_ROOM_ENCAP_L4_GRE = (1ULL << 3),
3237	3483	BPF_F_ADJ_ROOM_ENCAP_L4_UDP = (1ULL << 4),
	3484	+ BPF_F_ADJ_ROOM_NO_CSUM_RESET = (1ULL << 5),
3238	3485	};
3239	3486
3240	3487	enum {
@@ -3261,6 +3508,29 @@ enum { · 20 unmodified lines
3261	3508	BPF_F_GET_BRANCH_RECORDS_SIZE = (1ULL << 0),
3262	3509	};
3263	3510
	3511	+/* BPF_FUNC_bpf_ringbuf_commit, BPF_FUNC_bpf_ringbuf_discard, and
	3512	+ * BPF_FUNC_bpf_ringbuf_output flags.
	3513	+ */
	3514	+enum {
	3515	+ BPF_RB_NO_WAKEUP = (1ULL << 0),
	3516	+ BPF_RB_FORCE_WAKEUP = (1ULL << 1),
	3517	+};
	3518	+
	3519	+/* BPF_FUNC_bpf_ringbuf_query flags */
	3520	+enum {
	3521	+ BPF_RB_AVAIL_DATA = 0,
	3522	+ BPF_RB_RING_SIZE = 1,
	3523	+ BPF_RB_CONS_POS = 2,
	3524	+ BPF_RB_PROD_POS = 3,
	3525	+};
	3526	+
	3527	+/* BPF ring buffer constants */
	3528	+enum {
	3529	+ BPF_RINGBUF_BUSY_BIT = (1U << 31),
	3530	+ BPF_RINGBUF_DISCARD_BIT = (1U << 30),
	3531	+ BPF_RINGBUF_HDR_SZ = 8,
	3532	+};
	3533	+
3264	3534	/* Mode for BPF_FUNC_skb_adjust_room helper. */
3265	3535	enum bpf_adj_room_mode {
3266	3536	BPF_ADJ_ROOM_NET,
@@ -3393,6 +3663,7 @@ struct bpf_sock { · 126 unmodified lines
3393	3663	__u32 dst_ip4;
3394	3664	__u32 dst_ip6[4];
3395	3665	__u32 state;
	3666	+ __s32 rx_queue_mapping;
3396	3667	};
3397	3668
3398	3669	struct bpf_tcp_sock {
@@ -3486,6 +3757,8 @@ struct xdp_md { · 87 unmodified lines
3486	3757	/* Below access go through struct xdp_rxq_info */
3487	3758	__u32 ingress_ifindex; /* rxq->dev->ifindex */
3488	3759	__u32 rx_queue_index; /* rxq->queue_index */
	3760	+
	3761	+ __u32 egress_ifindex; /* txq->dev->ifindex */
3489	3762	};
3490	3763
3491	3764	enum sk_action {
@@ -3508,6 +3781,8 @@ struct sk_msg_md { · 16 unmodified lines
3508	3781	__u32 remote_port; /* Stored in network byte order */
3509	3782	__u32 local_port; /* stored in host byte order */
3510	3783	__u32 size; /* Total size of sk_msg */
	3784	+
	3785	+ __bpf_md_ptr(struct bpf_sock , sk); / current socket */
3511	3786	};
3512	3787
3513	3788	struct sk_reuseport_md {
@@ -3598,6 +3873,29 @@ struct bpf_btf_info { · 84 unmodified lines
3598	3873	__u32 id;
3599	3874	} __attribute__((aligned(8)));
3600	3875
	3876	+struct bpf_link_info {
	3877	+ __u32 type;
	3878	+ __u32 id;
	3879	+ __u32 prog_id;
	3880	+ union {
	3881	+ struct {
	3882	+ __aligned_u64 tp_name; /* in/out: tp_name buffer ptr */
	3883	+ __u32 tp_name_len; /* in/out: tp_name buffer len */
	3884	+ } raw_tracepoint;
	3885	+ struct {
	3886	+ __u32 attach_type;
	3887	+ } tracing;
	3888	+ struct {
	3889	+ __u64 cgroup_id;
	3890	+ __u32 attach_type;
	3891	+ } cgroup;
	3892	+ struct {
	3893	+ __u32 netns_ino;
	3894	+ __u32 attach_type;
	3895	+ } netns;
	3896	+ };
	3897	+} __attribute__((aligned(8)));
	3898	+
3601	3899	/* User bpf_sock_addr struct to access socket fields and sockaddr struct passed
3602	3900	* by user and intended to be used by socket (e.g. to bind to, depends on
3603	3901	* attach attach type).
@@ -3610,7 +3908,7 @@ struct bpf_sock_addr { · 6 unmodified lines
3610	3908	__u32 user_ip6[4]; /* Allows 1,2,4,8-byte read and 4,8-byte write.
3611	3909	* Stored in network byte order.
3612	3910	*/
3613		− __u32 user_port; /* Allows 4-byte read and write.
	3911	+ __u32 user_port; /* Allows 1,2,4-byte read and 4-byte write.
3614	3912	* Stored in network byte order
3615	3913	*/
3616	3914	__u32 family; /* Allows 4-byte read, but no write */

tools/include/uapi/linux/if_link.h+1−0

@@ -343,6 +343,7 @@ enum { · 342 unmodified lines
343	343	IFLA_BRPORT_NEIGH_SUPPRESS,
344	344	IFLA_BRPORT_ISOLATED,
345	345	IFLA_BRPORT_BACKUP_PORT,
	346	+ IFLA_BRPORT_MRP_RING_OPEN,
346	347	__IFLA_BRPORT_MAX
347	348	};
348	349	#define IFLA_BRPORT_MAX (__IFLA_BRPORT_MAX - 1)