display "last time" actions info
Subsystems (3)
Geliang Tang
reviewerv4 cover display "last time" actions info
From: Geliang Tang <[email protected]>
v4:
- address Mat's comments for v3 (thanks).
- merge patch 1 and patch 2 into one.
v3:
- address Matt's comments for patch 3 (thanks)
- add "dport", "sleep", and mptcp_lib_expect_all_features.
v2:
- address Mat's comments for v1 (thanks)
- set msk->last_data_sent only if err > 0
- set last_data_recv in __mptcp_move_skbs_from_subflow
- add three reserved bytes after mptcpi_subflows_total
- move selftests from mptcp_join.sh to diag.sh and check that the
timestamps move forward.
Geliang Tang (2):
mptcp: add last time fields in mptcp_info
selftests: mptcp: test last time mptcp_info
include/uapi/linux/mptcp.h | 4 ++ net/mptcp/options.c | 1 + net/mptcp/protocol.c | 7 ++++ net/mptcp/protocol.h | 3 ++ net/mptcp/sockopt.c | 4 ++ tools/testing/selftests/net/mptcp/diag.sh | 48 +++++++++++++++++++++++ 6 files changed, 67 insertions(+)
Signature · 1 line
2.40.1
Geliang Tang
reviewerPATCH v4 1/2 mptcp: add last time fields in mptcp_info
From: Geliang Tang <[email protected]>
This patch adds "last time" fields last_data_sent, last_data_recv and last_ack_recv in struct mptcp_sock to record the last time data_sent, data_recv and ack_recv happened. They all are initialized as tcp_jiffies32 in __mptcp_init_sock(), but updated as tcp_jiffies32 too when data is sent in __subflow_push_pending(), data is received in __mptcp_move_skbs_from_subflow(), and ack is received in ack_update_msk().
Similar to tcpi_last_data_sent, tcpi_last_data_recv and tcpi_last_ack_recv exposed with TCP, this patch exposes the last time "an action happened" for MPTCP in mptcp_info, named mptcpi_last_data_sent, mptcpi_last_data_recv and mptcpi_last_ack_recv, calculated in mptcp_diag_fill_info() as the time deltas between now and the newly added last time fields in mptcp_sock.
Also add three reserved bytes in struct mptcp_info.
--- include/uapi/linux/mptcp.h | 4 ++++ net/mptcp/options.c | 1 + net/mptcp/protocol.c | 7 +++++++ net/mptcp/protocol.h | 3 +++ net/mptcp/sockopt.c | 4 ++++ 5 files changed, 19 insertions(+)
include/uapi/linux/mptcp.h +4 -0
@@ -58,6 +58,10 @@ struct mptcp_info {__u64 mptcpi_bytes_received;__u64 mptcpi_bytes_acked;__u8 mptcpi_subflows_total; +__u8 reserved[3 ]; +__u32 mptcpi_last_data_sent; +__u32 mptcpi_last_data_recv; +__u32 mptcpi_last_ack_recv; };/* MPTCP Reset reason codes, rfc8684 */
net/mptcp/options.c +1 -0
@@ -1069,6 +1069,7 @@ static void ack_update_msk(struct mptcp_sock *msk,__mptcp_snd_una_update (msk, new_snd_una);__mptcp_data_acked (sk); } + msk->last_ack_recv = tcp_jiffies32;mptcp_data_unlock (sk); trace_ack_update_msk(mp_opt->data_ack,
net/mptcp/protocol.c +7 -0
@@ -706,6 +706,8 @@ static bool __mptcp_move_skbs_from_subflow(struct mptcp_sock *msk, } }while (more_data_avail); +if (moved >0 ) + msk->last_data_recv = tcp_jiffies32; *bytes += moved;return done; }
@@ -1556,6 +1558,8 @@ static int __subflow_push_pending(struct sock *sk, struct sock *ssk, err = copied; out: +if (err >0 ) + msk->last_data_sent = tcp_jiffies32;return err; }
@@ -2793,6 +2797,9 @@ static void __mptcp_init_sock(struct sock *sk)WRITE_ONCE (msk->allow_infinite_fallback, true); msk->recovery = false; msk->subflow_id =1 ; + msk->last_data_sent = tcp_jiffies32; + msk->last_data_recv = tcp_jiffies32; + msk->last_ack_recv = tcp_jiffies32;mptcp_pm_data_init (msk);
net/mptcp/protocol.h +3 -0
@@ -282,6 +282,9 @@ struct mptcp_sock {u64 bytes_acked;u64 snd_una;u64 wnd_end; +u32 last_data_sent; +u32 last_data_recv; +u32 last_ack_recv;unsigned long timer_ival;u32 token;int rmem_released;
net/mptcp/sockopt.c +4 -0
@@ -896,6 +896,7 @@ static int mptcp_getsockopt_first_sf_only(struct mptcp_sock *msk, int level, intvoid mptcp_diag_fill_info (struct mptcp_sock *msk,struct mptcp_info *info) {struct sock *sk = (struct sock *)msk; +u32 now = tcp_jiffies32;u32 flags =0 ;bool slow;
@@ -930,6 +931,7 @@ void mptcp_diag_fill_info(struct mptcp_sock *msk, struct mptcp_info *info) info->mptcpi_snd_una = msk->snd_una; info->mptcpi_rcv_nxt = msk->ack_seq; info->mptcpi_bytes_acked = msk->bytes_acked; + info->mptcpi_last_ack_recv =jiffies_to_msecs (now - msk->last_ack_recv);mptcp_data_unlock (sk); slow =lock_sock_fast (sk);
@@ -942,6 +944,8 @@ void mptcp_diag_fill_info(struct mptcp_sock *msk, struct mptcp_info *info) info->mptcpi_bytes_retrans = msk->bytes_retrans; info->mptcpi_subflows_total = info->mptcpi_subflows +__mptcp_has_initial_subflow (msk); + info->mptcpi_last_data_sent =jiffies_to_msecs (now - msk->last_data_sent); + info->mptcpi_last_data_recv =jiffies_to_msecs (now - msk->last_data_recv);unlock_sock_fast (sk, slow); }EXPORT_SYMBOL_GPL (mptcp_diag_fill_info);
Signature · 1 line
2.40.1
Geliang Tang
reviewerPATCH v4 2/2 selftests: mptcp: test last time mptcp_info
In reply to Geliang TangFrom: Geliang Tang <[email protected]>
This patch adds a new helper chk_msk_info() to show the counters in mptcp_info of the given info, and check that the timestamps move forward. Use it to show newly added last_data_sent, last_data_recv and last_ack_recv in mptcp_info in chk_last_time_info().
--- tools/testing/selftests/net/mptcp/diag.sh | 48 +++++++++++++++++++++++ 1 file changed, 48 insertions(+)
tools/testing/selftests/net/mptcp/diag.sh +48 -0
@@ -200,6 +200,53 @@ chk_msk_cestab() "${expected}" "${msg}" "" } +msk_info_get_value() +{ + local port="${1}" + local info="${2}" + + ss -N ${ns} -inHM dport ${port} | \ + mptcp_lib_get_info_value "${info}" "${info}" +} + +chk_msk_info() +{ + local port="${1}" + local info="${2}" + local cnt="${3}" + local now=$(msk_info_get_value ${port} "${info}") + local delta_ms=250 + + mptcp_lib_print_title "....chk ${info}" + if { [ -z "${cnt}" ] || [ -z "${now}" ]; } && + ! mptcp_lib_expect_all_features; then + mptcp_lib_pr_skip "Feature probably not supported" + mptcp_lib_result_skip "${info}" + elif [ "$((cnt1 + ${delta_ms}))" -lt "${now}" ]; then + mptcp_lib_pr_ok + mptcp_lib_result_pass "${info}" + else + mptcp_lib_pr_fail "value of ${info} changed by $((now - cnt))ms," \ + "expected at least ${delta_ms}ms" + mptcp_lib_result_fail "${info}" + ret=${KSFT_FAIL} + fi +} + +chk_last_time_info() +{ + local port="${1}" + local cnt1=$(msk_info_get_value ${port} "last_data_sent") + local cnt2=$(msk_info_get_value ${port} "last_data_recv") + local cnt3=$(msk_info_get_value ${port} "last_ack_recv") + + sleep 0.5 + + chk_msk_info "${port}" "last_data_sent" "${cnt1}" + chk_msk_info "${port}" "last_data_recv" "${cnt2}" + chk_msk_info "${port}" "last_ack_recv" "${cnt3}" +} + wait_connected() { local listener_ns="${1}"
@@ -233,6 +280,7 @@ echo "b" | \ 127.0.0.1 >/dev/null & wait_connected $ns 10000 chk_msk_nr 2 "after MPC handshake " +chk_last_time_info 10000 chk_msk_remote_key_nr 2 "....chk remote_key" chk_msk_fallback_nr 0 "....chk no fallback" chk_msk_inuse 2
Signature · 1 line
2.40.1
Matthieu Baerts
maintainerRe: [PATCH mptcp-next v4 2/2] selftests: mptcp: test last time mptcp_info
Hi Geliang,
I did a review of this patch before, but I didn't send it because I was looking at the CI in parallel (see below). Later, I noticed your squash-to patch (and the report from the CI). I think you fixed all the issues I reported here below, except my suggestion about not using 'cnt1', 'cnt2', etc. but something more explicit. Do you mind looking at that, please?
Also, do you mind sending a v5 instead? The CI will not pick-up the squash-to patch.
Please note that the CI will now [1] use your IPRoute2 patch [2], so the diag.sh test should no longer fail:
[1]
https://github.com/multipath-tcp/mptcp-upstream-virtme-docker/commit/fbf8f1a
[2] https://github.com/multipath-tcp/iproute2/commit/591f0654c
Quoted 33 lines · Geliang Tang
From: Geliang Tang <[email protected]>
This patch adds a new helper
chk_msk_info()to show the counters inmptcp_infoof the given info, and check that the timestamps move forward. Use it to show newly addedlast_data_sent,last_data_recvandlast_ack_recvinmptcp_infoinchk_last_time_info().Signed-off-by Geliang Tang <[email protected]>--- tools/testing/selftests/net/mptcp/diag.sh | 48 +++++++++++++++++++++++ 1 file changed, 48 insertions(+)tools/testing/selftests/net/mptcp/diag.sh +13
-0@@ -200,6 +200,53 @@ chk_msk_cestab() "${expected}" "${msg}" "" } +msk_info_get_value() +{ + local port="${1}" + local info="${2}" + + ss -N ${ns} -inHM dport ${port} | \ + mptcp_lib_get_info_value "${info}" "${info}" +} + +chk_msk_info() +{ + local port="${1}" + local info="${2}"
tools/testing/selftests/net/mptcp/diag.sh chk_msk_cestab() · line 225
Quoted 10 lines
tools/testing/selftests/net/mptcp/diag.sh +10
-0+ local cnt="${3}" + local now=$(msk_info_get_value ${port} "${info}") + local delta_ms=250 + + mptcp_lib_print_title "....chk ${info}" + if { [ -z "${cnt}" ] || [ -z "${now}" ]; } && + ! mptcp_lib_expect_all_features; then + mptcp_lib_pr_skip "Feature probably not supported" + mptcp_lib_result_skip "${info}" + elif [ "$((cnt1 + ${delta_ms}))" -lt "${now}" ]; then
It should be 'cnt', not 'cnt1', and without ${} around delta_ms when used inside '$((...))'. shellcheck should complain about that.
Quoted 6 lines
tools/testing/selftests/net/mptcp/diag.sh +6
-0+ mptcp_lib_pr_ok + mptcp_lib_result_pass "${info}" + else + mptcp_lib_pr_fail "value of ${info} changed by $((now - cnt))ms," \ + "expected at least ${delta_ms}ms" + mptcp_lib_result_fail "${info}"
tools/testing/selftests/net/mptcp/diag.sh chk_msk_cestab() · line 241
Quoted 10 lines
tools/testing/selftests/net/mptcp/diag.sh +10
-0+ ret=${KSFT_FAIL} + fi +} + +chk_last_time_info() +{ + local port="${1}" + local cnt1=$(msk_info_get_value ${port} "last_data_sent") + local cnt2=$(msk_info_get_value ${port} "last_data_recv") + local cnt3=$(msk_info_get_value ${port} "last_ack_recv")
I would recommend not using '<generic name>X', but something more readable:
data_sent=$(...) data_recv=$(...) ack_recv=$(...)
Also, I guess shellcheck will complain here: double quotes around "${port}", and declaring and assigning separately, no?
If you are not using them, there are shellcheck plugins for many text editors [1], or you can use a git hook [1].
[1] https://github.com/koalaman/shellcheck#user-content-in-your-editor
[2] https://github.com/koalaman/shellcheck-precommit
Quoted 12 lines
tools/testing/selftests/net/mptcp/diag.sh +8
-0+ + sleep 0.5 + + chk_msk_info "${port}" "last_data_sent" "${cnt1}" + chk_msk_info "${port}" "last_data_recv" "${cnt2}" + chk_msk_info "${port}" "last_ack_recv" "${cnt3}" +} + wait_connected() { local listener_ns="${1}"
tools/testing/selftests/net/mptcp/diag.sh echo "b" | \ · line 286
Quoted 7 lines
tools/testing/selftests/net/mptcp/diag.sh +1
-0@@ -233,6 +280,7 @@ echo "b" | \ 127.0.0.1 >/dev/null & wait_connected $ns 10000 chk_msk_nr 2 "after MPC handshake " +chk_last_time_info 10000 chk_msk_remote_key_nr 2 "....chk remote_key" chk_msk_fallback_nr 0 "....chk no fallback" chk_msk_inuse 2
Cheers,
Matt
Signature · 1 line
Sponsored by the NGI0 Core fund.
Hi Geliang,
Thank you for your modifications, that's great!
Our CI did some validations and here is its report:
- KVM Validation: normal: Unstable: 1 failed test(s): selftest_diag 🔴
- KVM Validation: debug: Unstable: 1 failed test(s): selftest_diag 🔴
- KVM Validation: btf (only bpftest_all): Success! ✅
- Task: https://github.com/multipath-tcp/mptcp_net-next/actions/runs/8520704616
Initiator: Patchew Applier
Commits: https://github.com/multipath-tcp/mptcp_net-next/commits/6b5cb3bcdad5
Patchwork: https://patchwork.kernel.org/project/mptcp/list/?series=840534
If there are some issues, you can reproduce them using the same environment as the one used by the CI thanks to a docker image, e.g.:
$ cd [kernel source code]
$ docker run -v "${PWD}:${PWD}:rw" -w "${PWD}" --privileged --rm -it \
--pull always mptcp/mptcp-upstream-virtme-docker:latest \
auto-normalFor more details:
https://github.com/multipath-tcp/mptcp-upstream-virtme-docker
Please note that despite all the efforts that have been already done to have a stable tests suite when executed on a public CI like here, it is possible some reported issues are not due to your modifications. Still, do not hesitate to help us improve that ;-)
Cheers,
MPTCP GH Action bot
Bot operated by Matthieu Baerts (NGI0 Core)