Skip to contentSkip to conversation

display "last time" actions info

PATCH Superseded Geliang Tang 3 participants 6 messages Mar 29, 2024

Subsystems (3)

v2 of 5 Superseded Latest is v5 →

lore
Share & tools
lore

Geliang Tang

author reviewer
#0 lore raw

v2 cover display "last time" actions info

From: Geliang Tang <[email protected]>

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 (3):
mptcp: add last time fields in mptcp_sock
mptcp: add last time fields in mptcp_info
selftests: mptcp: add last time actions tests

 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                       |  5 +++++
 tools/testing/selftests/net/mptcp/diag.sh | 24 +++++++++++++++++++++++
 6 files changed, 44 insertions(+)
Signature · 1 line
2.40.1
author reviewer
#1 lore raw

PATCH v2 1/3 mptcp: add last time fields in mptcp_sock

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().

---
 net/mptcp/options.c  | 1 +
 net/mptcp/protocol.c | 7 +++++++
 net/mptcp/protocol.h | 3 +++
 3 files changed, 11 insertions(+)
net/mptcp/options.c +1 -0
index 5926955625cf..c0832df3b0a3 100644
@@ -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
index 556b3b95c537..43318aa5f991 100644
@@ -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
index 5a4538205fd6..3a3fed3642dd 100644
@@ -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;
Signature · 1 line
2.40.1
author reviewer
#2 lore raw

PATCH v2 2/3 mptcp: add last time fields in mptcp_info

In reply to Geliang Tang

From: Geliang Tang <[email protected]>

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/sockopt.c        | 5 +++++
 2 files changed, 9 insertions(+)
include/uapi/linux/mptcp.h +4 -0
index 74cfe496891e..67d015df8893 100644
@@ -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/sockopt.c +5 -0
index dcd1c76d2a3b..1e74851614e8 100644
@@ -898,6 +898,7 @@ void mptcp_diag_fill_info(struct mptcp_sock *msk, struct mptcp_info *info)
 	struct sock *sk = (struct sock *)msk;
 	u32 flags = 0;
 	bool slow;
+	u32 now;
 
 	memset(info, 0, sizeof(*info));
 
@@ -942,6 +943,10 @@ 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);
+	now = tcp_jiffies32;
+	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);
+	info->mptcpi_last_ack_recv = jiffies_to_msecs(now - msk->last_ack_recv);
 	unlock_sock_fast(sk, slow);
 }
 EXPORT_SYMBOL_GPL(mptcp_diag_fill_info);
Signature · 1 line
2.40.1
author reviewer
#3 lore raw

PATCH v2 3/3 selftests: mptcp: add last time actions tests

In reply to Geliang Tang

From: Geliang Tang <[email protected]>

This patch adds a new helper chk_msk_info() to show the counters in mptcp_info of the given infos, 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 diag.sh.

---
 tools/testing/selftests/net/mptcp/diag.sh | 24 +++++++++++++++++++++++
 1 file changed, 24 insertions(+)
tools/testing/selftests/net/mptcp/diag.sh +24 -0
index bc97ab33a00e..6e865f95f85e 100755
@@ -200,6 +200,27 @@ chk_msk_cestab()
 		 "${expected}" "${msg}" ""
 }
 
+chk_msk_info()
+{
+	local info
+
+	for info in "${@}"; do
+		local cnt1 cnt2 msg
+
+		cnt1=$(ss -N ${ns} -inmHM | mptcp_lib_get_info_value "$info" "$info")
+		cnt2=$(ss -N ${ns} -inmHM | mptcp_lib_get_info_value "$info" "$info")
+		msg="....chk ${info:0:15}=$cnt1:$cnt2"
+		mptcp_lib_print_title "${msg}"
+		if [ "${cnt1}" -lt "${cnt2}" ]; then
+			mptcp_lib_pr_ok
+			mptcp_lib_result_pass "${msg}"
+		else
+			mptcp_lib_pr_skip
+			mptcp_lib_result_skip "${msg}"
+		fi
+	done
+}
+
 wait_connected()
 {
 	local listener_ns="${1}"
@@ -237,6 +258,7 @@ chk_msk_remote_key_nr 2 "....chk remote_key"
 chk_msk_fallback_nr 0 "....chk no fallback"
 chk_msk_inuse 2
 chk_msk_cestab 2
+chk_msk_info last_data_sent last_data_recv last_ack_recv
 flush_pids
 
 chk_msk_inuse 0 "2->0"
@@ -257,6 +279,7 @@ wait_connected $ns 10001
 chk_msk_fallback_nr 1 "check fallback"
 chk_msk_inuse 1
 chk_msk_cestab 1
+chk_msk_info last_data_sent last_data_recv last_ack_recv
 flush_pids
 
 chk_msk_inuse 0 "1->0"
@@ -283,6 +306,7 @@ done
 wait_msk_nr $((NR_CLIENTS*2)) "many msk socket present"
 chk_msk_inuse $((NR_CLIENTS*2)) "many"
 chk_msk_cestab $((NR_CLIENTS*2)) "many"
+chk_msk_info last_data_sent last_data_recv last_ack_recv
 flush_pids
 
 chk_msk_inuse 0 "many->0"
Signature · 1 line
2.40.1
maintainer
#4 lore raw

Re: [PATCH mptcp-next v2 3/3] selftests: mptcp: add last time actions tests

Hi Geliang,

Quoted 18 lines · Geliang Tang

From: Geliang Tang <[email protected]>

This patch adds a new helper chk_msk_info() to show the counters in mptcp_info of the given infos, 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 diag.sh.

---
 tools/testing/selftests/net/mptcp/diag.sh | 24 +++++++++++++++++++++++
 1 file changed, 24 insertions(+)
tools/testing/selftests/net/mptcp/diag.sh +0 -0
index bc97ab33a00e..6e865f95f85e 100755
@@ -200,6 +200,27 @@ chk_msk_cestab()
 		 "${expected}" "${msg}" ""

tools/testing/selftests/net/mptcp/diag.sh chk_msk_cestab() · line 210

Quoted 10 lines
tools/testing/selftests/net/mptcp/diag.sh +8 -0
 }
 
+chk_msk_info()
+{
+	local info
+
+	for info in "${@}"; do
+		local cnt1 cnt2 msg
+
+		cnt1=$(ss -N ${ns} -inmHM | mptcp_lib_get_info_value "$info" "$info")

Mmh, when there are many connections in parallel or a fallback to TCP, what are you going to look at? Would it not be better to restrict this test only when there is one connection ("after MPC handshake"), using a counter for the client side and one for the server side, using ss '[sd]port ${port}' filter?

Also, do you need 'ss -m' here?

tools/testing/selftests/net/mptcp/diag.sh chk_msk_cestab() · line 211

Quoted 1 line
tools/testing/selftests/net/mptcp/diag.sh +1 -0
+		cnt2=$(ss -N ${ns} -inmHM | mptcp_lib_get_info_value "$info" "$info")

Should you maybe wait half a second between the two call of 'ss', just to be sure data have been exchanged on very slow environments? By doing that, you can additionally check that the timestamps are different by at least 250ms, no?

tools/testing/selftests/net/mptcp/diag.sh chk_msk_cestab() · line 212

Quoted 1 line
tools/testing/selftests/net/mptcp/diag.sh +1 -0
+		msg="....chk ${info:0:15}=$cnt1:$cnt2"

${msg} is used in mptcp_lib_result_XXX(), which mean it is used as title for the subtest in the TAP format. This title should always have the same name in case of error or success for the same test. Here, the same test will always have a different name, that's not OK.

(It is also important to have different names for different subtests, so if the helper is used multiple time, you cannot just show '${info:0:15}' here)

tools/testing/selftests/net/mptcp/diag.sh chk_msk_cestab() · line 219

Quoted 7 lines
tools/testing/selftests/net/mptcp/diag.sh +7 -0
+		mptcp_lib_print_title "${msg}"
+		if [ "${cnt1}" -lt "${cnt2}" ]; then
+			mptcp_lib_pr_ok
+			mptcp_lib_result_pass "${msg}"
+		else
+			mptcp_lib_pr_skip
+			mptcp_lib_result_skip "${msg}"

(the fail part is missing, but I see you added it in a squash-to patch later, I will comment there).

Cheers,
Matt

Signature · 1 line
Sponsored by the NGI0 Core fund.
#5 lore raw
In reply to Geliang Tang

Hi Geliang,

Thank you for your modifications, that's great!

Our CI did some validations and here is its report:

- KVM Validation: normal: Success! ✅
- KVM Validation: debug: Success! ✅
- KVM Validation: btf (only bpftest_all): Success! ✅
- Task: https://github.com/multipath-tcp/mptcp_net-next/actions/runs/8479431546

Initiator: Patchew Applier
Commits: https://github.com/multipath-tcp/mptcp_net-next/commits/bf2927563d79
Patchwork: https://patchwork.kernel.org/project/mptcp/list/?series=839714

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-normal

For 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)