Re: [PATCH bpf-next v9 11/12] bpf: support selective sampling for bpf timestamping

[Martin KaFai Lau <martin.lau@linux.dev>]

On 2/8/25 2:32 AM, Jason Xing wrote:
> Use __bpf_kfunc feature to allow bpf prog dynamically and selectively

s/Use/Add/

Remove "dynamically". A kfunc can only be called dynamically at runtime.

Like:

"Add the bpf_sock_ops_enable_tx_tstamp kfunc to allow BPF programs to 
selectively enable TX timestamping on a skb during tcp_sendmsg..."

> to sample/track the skb. For example, the bpf prog will limit tracking
> X numbers of packets and then will stop there instead of tracing
> all the sendmsgs of matched flow all along.
 > > Signed-off-by: Jason Xing <kerneljasonxing@gmail.com>
> ---
>   kernel/bpf/btf.c  |  1 +
>   net/core/filter.c | 27 ++++++++++++++++++++++++++-
>   2 files changed, 27 insertions(+), 1 deletion(-)
> 
> diff --git a/kernel/bpf/btf.c b/kernel/bpf/btf.c
> index 8396ce1d0fba..a65e2eeffb88 100644
> --- a/kernel/bpf/btf.c
> +++ b/kernel/bpf/btf.c
> @@ -8535,6 +8535,7 @@ static int bpf_prog_type_to_kfunc_hook(enum bpf_prog_type prog_type)
>   	case BPF_PROG_TYPE_CGROUP_SOCK_ADDR:
>   	case BPF_PROG_TYPE_CGROUP_SOCKOPT:
>   	case BPF_PROG_TYPE_CGROUP_SYSCTL:
> +	case BPF_PROG_TYPE_SOCK_OPS:
>   		return BTF_KFUNC_HOOK_CGROUP;
>   	case BPF_PROG_TYPE_SCHED_ACT:
>   		return BTF_KFUNC_HOOK_SCHED_ACT;
> diff --git a/net/core/filter.c b/net/core/filter.c
> index 7f56d0bbeb00..db20a947e757 100644
> --- a/net/core/filter.c
> +++ b/net/core/filter.c
> @@ -12102,6 +12102,21 @@ __bpf_kfunc int bpf_sk_assign_tcp_reqsk(struct __sk_buff *s, struct sock *sk,
>   #endif
>   }
>   
> +__bpf_kfunc int bpf_sock_ops_enable_tx_tstamp(struct bpf_sock_ops_kern *skops)

I am ok to always enable txstamp_ack here. Please still add a second "u64 flags" 
argument such that future disable/enable is still possible.

> +{
> +	struct sk_buff *skb;
> +
> +	if (skops->op != BPF_SOCK_OPS_TS_SND_CB)
 > +		return -EOPNOTSUPP;> +
> +	skb = skops->skb;
> +	TCP_SKB_CB(skb)->txstamp_ack = 2;

Willem (thanks!) has already mentioned there is a bug.

This also brought up that a test is missing: the bpf timestamping and user 
space's SO_TIMESTAMPING can work without interfering others. The current test 
only has SK_BPF_CB_TX_TIMESTAMPING on. A test is needed when both 
SK_BPF_CB_TX_TIMESTAMPING and the user space's SO_TIMESTAMPING are on. The 
expectation is both of them will work together.

> +	skb_shinfo(skb)->tx_flags |= SKBTX_BPF;
> +	skb_shinfo(skb)->tskey = TCP_SKB_CB(skb)->seq + skb->len - 1;
> +
> +	return 0;
> +}
> +
>   __bpf_kfunc_end_defs();
>   
>   int bpf_dynptr_from_skb_rdonly(struct __sk_buff *skb, u64 flags,
> @@ -12135,6 +12150,10 @@ BTF_KFUNCS_START(bpf_kfunc_check_set_tcp_reqsk)
>   BTF_ID_FLAGS(func, bpf_sk_assign_tcp_reqsk, KF_TRUSTED_ARGS)
>   BTF_KFUNCS_END(bpf_kfunc_check_set_tcp_reqsk)
>   
> +BTF_KFUNCS_START(bpf_kfunc_check_set_sock_ops)
> +BTF_ID_FLAGS(func, bpf_sock_ops_enable_tx_tstamp, KF_TRUSTED_ARGS)
> +BTF_KFUNCS_END(bpf_kfunc_check_set_sock_ops)
> +
>   static const struct btf_kfunc_id_set bpf_kfunc_set_skb = {
>   	.owner = THIS_MODULE,
>   	.set = &bpf_kfunc_check_set_skb,
> @@ -12155,6 +12174,11 @@ static const struct btf_kfunc_id_set bpf_kfunc_set_tcp_reqsk = {
>   	.set = &bpf_kfunc_check_set_tcp_reqsk,
>   };
>   
> +static const struct btf_kfunc_id_set bpf_kfunc_set_sock_ops = {
> +	.owner = THIS_MODULE,
> +	.set = &bpf_kfunc_check_set_sock_ops,
> +};
> +
>   static int __init bpf_kfunc_init(void)
>   {
>   	int ret;
> @@ -12173,7 +12197,8 @@ static int __init bpf_kfunc_init(void)
>   	ret = ret ?: register_btf_kfunc_id_set(BPF_PROG_TYPE_XDP, &bpf_kfunc_set_xdp);
>   	ret = ret ?: register_btf_kfunc_id_set(BPF_PROG_TYPE_CGROUP_SOCK_ADDR,
>   					       &bpf_kfunc_set_sock_addr);
> -	return ret ?: register_btf_kfunc_id_set(BPF_PROG_TYPE_SCHED_CLS, &bpf_kfunc_set_tcp_reqsk);
> +	ret = ret ?: register_btf_kfunc_id_set(BPF_PROG_TYPE_SCHED_CLS, &bpf_kfunc_set_tcp_reqsk);
> +	return ret ?: register_btf_kfunc_id_set(BPF_PROG_TYPE_SOCK_OPS, &bpf_kfunc_set_sock_ops);
>   }
>   late_initcall(bpf_kfunc_init);
>