kernel/linux.git/include/trace/events/sock.h, branch v6.6.131

inet: preserve const qualifier in inet_sk()

2023-03-17T08:56:37+00:00

We can change inet_sk() to propagate const qualifier of its argument. This should avoid some potential errors caused by accidental (const -> not_const) promotion. Other helpers like tcp_sk(), udp_sk(), raw_sk() will be handled in separate patch series. v2: use container_of_const() as advised by Jakub and Linus Signed-off-by: Eric Dumazet Link: https://lore.kernel.org/netdev/20230315142841.3a2ac99a@kernel.org/ Link: https://lore.kernel.org/netdev/CAHk-=wiOf12nrYEF2vJMcucKjWPN-Ns_SW9fA7LwST_2Dzp7rw@mail.gmail.com/ Signed-off-by: David S. Miller

net/sock: Introduce trace_sk_data_ready()

2023-01-23T11:26:50+00:00

As suggested by Cong, introduce a tracepoint for all ->sk_data_ready() callback implementations. For example: <...> iperf-609 [002] ..... 70.660425: sk_data_ready: family=2 protocol=6 func=sock_def_readable iperf-609 [002] ..... 70.660436: sk_data_ready: family=2 protocol=6 func=sock_def_readable <...> Suggested-by: Cong Wang Signed-off-by: Peilin Ye Signed-off-by: David S. Miller

sock: add tracepoint for send recv length

2023-01-13T10:25:10+00:00

Add 2 tracepoints to monitor the tcp/udp traffic of per process and per cgroup. Regarding monitoring the tcp/udp traffic of each process, there are two existing solutions, the first one is https://www.atoptool.nl/netatop.php. The second is via kprobe/kretprobe. Netatop solution is implemented by registering the hook function at the hook point provided by the netfilter framework. These hook functions may be in the soft interrupt context and cannot directly obtain the pid. Some data structures are added to bind packets and processes. For example, struct taskinfobucket, struct taskinfo ... Every time the process sends and receives packets it needs multiple hashmaps,resulting in low performance and it has the problem fo inaccurate tcp/udp traffic statistics(for example: multiple threads share sockets). We can obtain the information with kretprobe, but as we know, kprobe gets the result by trappig in an exception, which loses performance compared to tracepoint. We compared the performance of tracepoints with the above two methods, and the results are as follows: ab -n 1000000 -c 1000 -r http://127.0.0.1/index.html without trace: Time per request: 39.660 [ms] (mean) Time per request: 0.040 [ms] (mean, across all concurrent requests) netatop: Time per request: 50.717 [ms] (mean) Time per request: 0.051 [ms] (mean, across all concurrent requests) kr: Time per request: 43.168 [ms] (mean) Time per request: 0.043 [ms] (mean, across all concurrent requests) tracepoint: Time per request: 41.004 [ms] (mean) Time per request: 0.041 [ms] (mean, across all concurrent requests It can be seen that tracepoint has better performance. Signed-off-by: Yunhui Cui Signed-off-by: Xiongchun Duan Reviewed-by: Steven Rostedt (Google) Signed-off-by: David S. Miller

net: sock: tracing: Fix sock_exceed_buf_limit not to dereference stale pointer

2022-07-08T11:06:17+00:00

The trace event sock_exceed_buf_limit saves the prot->sysctl_mem pointer and then dereferences it in the TP_printk() portion. This is unsafe as the TP_printk() portion is executed at the time the buffer is read. That is, it can be seconds, minutes, days, months, even years later. If the proto is freed, then this dereference will can also lead to a kernel crash. Instead, save the sysctl_mem array into the ring buffer and have the TP_printk() reference that instead. This is the proper and safe way to read pointers in trace events. Link: https://lore.kernel.org/all/20220706052130.16368-12-kuniyu@amazon.com/ Cc: stable@vger.kernel.org Fixes: 3847ce32aea9f ("core: add tracepoints for queueing skb to rcvbuf") Signed-off-by: Steven Rostedt (Google) Acked-by: Kuniyuki Iwashima Signed-off-by: David S. Miller

net: sock: add trace for socket errors

2021-06-29T18:28:21+00:00

This patch will add tracers to trace inet socket errors only. A user space monitor application can track connection errors indepedent from socket lifetime and do additional handling. For example a cluster manager can fence a node if errors occurs in a specific heuristic. Signed-off-by: Alexander Aring Signed-off-by: David S. Miller

tcp: Define IPPROTO_MPTCP

2020-01-10T02:41:41+00:00

To open a MPTCP socket with socket(AF_INET, SOCK_STREAM, IPPROTO_MPTCP), IPPROTO_MPTCP needs a value that differs from IPPROTO_TCP. The existing IPPROTO numbers mostly map directly to IANA-specified protocol numbers. MPTCP does not have a protocol number allocated because MPTCP packets use the TCP protocol number. Use private number not used OTA. Reviewed-by: Eric Dumazet Signed-off-by: Mat Martineau Signed-off-by: David S. Miller

sock: Make sk_protocol a 16-bit value

2020-01-10T02:41:41+00:00

Match the 16-bit width of skbuff->protocol. Fills an 8-bit hole so sizeof(struct sock) does not change. Also take care of BPF field access for sk_type/sk_protocol. Both of them are now outside the bitfield, so we can use load instructions without further shifting/masking. v5 -> v6: - update eBPF accessors, too (Intel's kbuild test robot) v2 -> v3: - keep 'sk_type' 2 bytes aligned (Eric) v1 -> v2: - preserve sk_pacing_shift as bit field (Eric) Cc: Alexei Starovoitov Cc: Daniel Borkmann Cc: bpf@vger.kernel.org Co-developed-by: Paolo Abeni Signed-off-by: Paolo Abeni Co-developed-by: Matthieu Baerts Signed-off-by: Matthieu Baerts Signed-off-by: Mat Martineau Signed-off-by: David S. Miller

tcp: annotate sk->sk_wmem_queued lockless reads

2019-10-13T17:13:08+00:00

For the sake of tcp_poll(), there are few places where we fetch sk->sk_wmem_queued while this field can change from IRQ or other cpu. We need to add READ_ONCE() annotations, and also make sure write sides use corresponding WRITE_ONCE() to avoid store-tearing. sk_wmem_queued_add() helper is added so that we can in the future convert to ADD_ONCE() or equivalent if/when available. Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller

tcp: annotate sk->sk_rcvbuf lockless reads

2019-10-13T17:13:08+00:00

For the sake of tcp_poll(), there are few places where we fetch sk->sk_rcvbuf while this field can change from IRQ or other cpu. We need to add READ_ONCE() annotations, and also make sure write sides use corresponding WRITE_ONCE() to avoid store-tearing. Note that other transports probably need similar fixes. Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller

net: expose sk wmem in sock_exceed_buf_limit tracepoint

2018-07-02T13:40:56+00:00

Currently trace_sock_exceed_buf_limit() only show rmem info, but wmem limit may also be hit. So expose wmem info in this tracepoint as well. Regarding memcg, I think it is better to introduce a new tracepoint(if that is needed), i.e. trace_memcg_limit_hit other than show memcg info in trace_sock_exceed_buf_limit. Signed-off-by: Yafang Shao Signed-off-by: David S. Miller