On 1/20/25 5:28 PM, Jason Xing wrote:
Introduce the callback to correlate tcp_sendmsg timestamp with other
three points (SND/SW/ACK). We can let bpf trace the beginning of
tcp_sendmsg_locked() and fetch the socket addr, so that in
tcp_tx_timestamp() we can correlate the tskey with the socket addr.
It is accurate since they are under the protect of socket lock.
More details can be found in the selftest.
Signed-off-by: Jason Xing <kerneljasonxing@xxxxxxxxx>
---
include/uapi/linux/bpf.h | 3 +++
net/ipv4/tcp.c | 1 +
tools/include/uapi/linux/bpf.h | 3 +++
3 files changed, 7 insertions(+)
diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h
index 3b9bfc88345c..55c74fa18163 100644
--- a/include/uapi/linux/bpf.h
+++ b/include/uapi/linux/bpf.h
@@ -7042,6 +7042,9 @@ enum {
* feature is on. It indicates the
* recorded timestamp.
*/
+ BPF_SOCK_OPS_TS_TCP_SND_CB, /* Called when every tcp_sendmsg
+ * syscall is triggered
+ */
I recall we agreed in v5 to adjust the "TCP_" naming part because it will be
used in UDP also. Like completely remove the "TCP_" from the name?
};
/* List of TCP states. There is a build check in net/ipv4/tcp.c to detect
diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c
index 0a41006b10d1..49e489c346ea 100644
--- a/net/ipv4/tcp.c
+++ b/net/ipv4/tcp.c
@@ -500,6 +500,7 @@ static void tcp_tx_timestamp(struct sock *sk, struct sockcm_cookie *sockc)
tcb->txstamp_ack_bpf = 1;
shinfo->tx_flags |= SKBTX_BPF;
shinfo->tskey = TCP_SKB_CB(skb)->seq + skb->len - 1;
+ bpf_skops_tx_timestamping(sk, skb, BPF_SOCK_OPS_TS_TCP_SND_CB);
}
}
diff --git a/tools/include/uapi/linux/bpf.h b/tools/include/uapi/linux/bpf.h
index b463aa9c27da..38fc04a7ac20 100644
--- a/tools/include/uapi/linux/bpf.h
+++ b/tools/include/uapi/linux/bpf.h
@@ -7035,6 +7035,9 @@ enum {
* feature is on. It indicates the
* recorded timestamp.
*/
+ BPF_SOCK_OPS_TS_TCP_SND_CB, /* Called when every tcp_sendmsg
+ * syscall is triggered
+ */
};
/* List of TCP states. There is a build check in net/ipv4/tcp.c to detect
