trace: add tracepoint for tcp_sendmsg_locked()
Subsystems (6)
v3 cover trace: add tracepoint for tcp_sendmsg_locked()
Meta has been using BPF programs to monitor tcp_sendmsg() for years, indicating significant interest in observing this important functionality. Adding a proper tracepoint provides a stable API for all users who need visibility into TCP message transmission.
David Ahern is using a similar functionality with a custom patch[1]. So, this means we have more than a single use case for this request, and it might be a good idea to have such feature upstream.
---
Changes in v3:
- Fixed identation (Kuniyuki Iwashi)
- Removed duplicated entry in the commit message (Kuniyuki Iwashi)
- Added tcp_sendmsg_locked in bpf_raw_tp_null_args[] (Kuniyuki Iwashi)
- Link to v2: https://lore.kernel.org/r/[email protected]
Changes in v2:
- Change to a full tracepoint inside tcp_sendmsg_locked(), heavily inspired in David's patch
- Link to v1: https://lore.kernel.org/r/[email protected]
---
Breno Leitao (2):
net: pass const to msg_data_left()
trace: tcp: Add tracepoint for tcp_sendmsg_locked()
include/linux/socket.h | 2 +- include/trace/events/tcp.h | 24 ++++++++++++++++++++++++ kernel/bpf/btf.c | 1 + net/ipv4/tcp.c | 2 ++ 4 files changed, 28 insertions(+), 1 deletion(-)
---
base-commit: 1a9239bb4253f9076b5b4b2a1a4e8d7defd77a95
change-id: 20250224-tcpsendmsg-4f0a236751d7
Best regards,
Signature · 1 line
Breno Leitao <[email protected]>
PATCH v3 1/2 net: pass const to msg_data_left()
The msg_data_left() function doesn't modify the struct msghdr parameter, so mark it as const. This allows the function to be used with const references, improving type safety and making the API more flexible.
--- include/linux/socket.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
include/linux/socket.h +1 -1
@@ -168,7 +168,7 @@ static inline struct cmsghdr * cmsg_nxthdr (struct msghdr *__msg, struct cmsghdrreturn __cmsg_nxthdr (__msg->msg_control, __msg->msg_controllen, __cmsg); } -static inline size_t msg_data_left (struct msghdr *msg) +static inline size_t msg_data_left (const struct msghdr *msg) {return iov_iter_count (&msg->msg_iter); }
Signature · 1 line
2.47.1
PATCH v3 2/2 trace: tcp: Add tracepoint for tcp_sendmsg_locked()
In reply to Breno LeitaoAdd a tracepoint to monitor TCP send operations, enabling detailed visibility into TCP message transmission.
Create a new tracepoint within the tcp_sendmsg_locked function, capturing traditional fields along with size_goal, which indicates the optimal data size for a single TCP segment. Additionally, a reference to the struct sock sk is passed, allowing direct access for BPF programs. The implementation is largely based on David's patch[1] and suggestions.
--- include/trace/events/tcp.h | 24 ++++++++++++++++++++++++ kernel/bpf/btf.c | 1 + net/ipv4/tcp.c | 2 ++ 3 files changed, 27 insertions(+)
include/trace/events/tcp.h +24 -0
@@ -259,6 +259,30 @@ TRACE_EVENT(tcp_retransmit_synack, __entry->saddr_v6, __entry->daddr_v6) ); +TRACE_EVENT (tcp_sendmsg_locked, + TP_PROTO(const struct sock *sk, const struct msghdr *msg, + const struct sk_buff *skb, int size_goal), + +TP_ARGS (sk, msg, skb, size_goal), + +TP_STRUCT__entry ( + __field(const void *, skb_addr) + __field(int, skb_len) + __field(int, msg_left) +__field (int, size_goal) + ), + +TP_fast_assign ( + __entry->skb_addr = skb; + __entry->skb_len = skb ? skb->len : 0; + __entry->msg_left = msg_data_left(msg); + __entry->size_goal = size_goal; + ), + +TP_printk ("skb_addr %p skb_len %d msg_left %d size_goal %d" , + __entry->skb_addr, __entry->skb_len, __entry->msg_left, + __entry->size_goal)); + DECLARE_TRACE(tcp_cwnd_reduction_tp, TP_PROTO(const struct sock *sk, int newly_acked_sacked, int newly_lost, int flag),
kernel/bpf/btf.c +1 -0
@@ -6518,6 +6518,7 @@ static const struct bpf_raw_tp_null_args raw_tp_null_args[] = { { "xprt_put_cong", 0x10 },/* tcp */ { "tcp_send_reset", 0x11 }, + { "tcp_sendmsg_locked", 0x100 },/* tegra_apb_dma */ { "tegra_dma_tx_status", 0x100 },/* timer_migration */
net/ipv4/tcp.c +2 -0
@@ -1160,6 +1160,8 @@ int tcp_sendmsg_locked(struct sock *sk, struct msghdr *msg, size_t size)if (skb) copy = size_goal - skb->len; +trace_tcp_sendmsg_locked (sk, msg, skb, size_goal); +if (copy <=0 || !tcp_skb_can_collapse_to (skb)) {bool first_skb;
Signature · 1 line
2.47.1
Re: [PATCH net-next v3 2/2] trace: tcp: Add tracepoint for tcp_sendmsg_locked()
Quoted 9 lines · Breno Leitao
Add a tracepoint to monitor TCP send operations, enabling detailed visibility into TCP message transmission.
Create a new tracepoint within the
tcp_sendmsg_lockedfunction, capturing traditional fields along withsize_goal, which indicates the optimal data size for a single TCP segment. Additionally, a reference to the struct sock sk is passed, allowing direct access for BPF programs. The implementation is largely based on David's patch[1] and suggestions.
Quoted 8 lines
Signed-off-by Breno Leitao <[email protected]>--- include/trace/events/tcp.h | 24 ++++++++++++++++++++++++ kernel/bpf/btf.c | 1 + net/ipv4/tcp.c | 2 ++ 3 files changed, 27 insertions(+)
Re: [PATCH net-next v3 2/2] trace: tcp: Add tracepoint for tcp_sendmsg_locked()
In reply to Breno LeitaoFrom: Breno Leitao <[email protected]>
Date: Tue, 08 Apr 2025 11:32:02 -0700
Quoted 11 lines
Add a tracepoint to monitor TCP send operations, enabling detailed visibility into TCP message transmission.
Create a new tracepoint within the
tcp_sendmsg_lockedfunction, capturing traditional fields along withsize_goal, which indicates the optimal data size for a single TCP segment. Additionally, a reference to the struct sock sk is passed, allowing direct access for BPF programs. The implementation is largely based on David's patch[1] and suggestions.Signed-off-by Breno Leitao <[email protected]>
Re: [PATCH net-next v3 2/2] trace: tcp: Add tracepoint for tcp_sendmsg_locked()
In reply to Breno LeitaoQuoted 4 lines · Breno Leitao
Add a tracepoint to monitor TCP send operations, enabling detailed visibility into TCP message transmission.
Quoted 8 lines
Create a new tracepoint within the
tcp_sendmsg_lockedfunction, capturing traditional fields along withsize_goal, which indicates the optimal data size for a single TCP segment. Additionally, a reference to the struct sock sk is passed, allowing direct access for BPF programs. The implementation is largely based on David's patch[1] and suggestions.Signed-off-by Breno Leitao <[email protected]>
This series was applied to netdev/net-next.git (main) by Jakub Kicinski <[email protected]>:
Hello:
This series was applied to netdev/net-next.git (main) by Jakub Kicinski <[email protected]>:
Quoted 10 lines · On Tue, 08 Apr 2025 11:32:00 -0700 you wrote:
Meta has been using BPF programs to monitor
tcp_sendmsg()for years, indicating significant interest in observing this important functionality. Adding a proper tracepoint provides a stable API for all users who need visibility into TCP message transmission.David Ahern is using a similar functionality with a custom patch[1]. So, this means we have more than a single use case for this request, and it might be a good idea to have such feature upstream.
[...]
Here is the summary with links:
- [net-next,v3,1/2] net: pass const to msg_data_left()
https://git.kernel.org/netdev/net-next/c/b1e904999542 - [net-next,v3,2/2] trace: tcp: Add tracepoint for tcp_sendmsg_locked() https://git.kernel.org/netdev/net-next/c/0f08335ade71
You are awesome, thank you!
Signature · 2 lines
Deet-doot-dot, I am a bot. https://korg.docs.kernel.org/patchwork/pwbot.html