On Tue, Aug 27, 2024 at 6:19 PM Rong Tao <rtoax@xxxxxxxxxxx> wrote:
>
> From: Rong Tao <rongtao@xxxxxxxx>
>
> In commit ba8de796baf4 ("net: introduce sk_skb_reason_drop function")
> kfree_skb_reason() becomes an inline function and cannot be traced.
> We can use the stable tracepoint kfree_skb to get 'ip'.
>
> Link: https://github.com/torvalds/linux/commit/ba8de796baf4bdc03530774fb284fe3c97875566
> Signed-off-by: Rong Tao <rongtao@xxxxxxxx>
> ---
> samples/bpf/tracex2.bpf.c | 13 ++++---------
> 1 file changed, 4 insertions(+), 9 deletions(-)
>
Maybe just remove this "sample". samples/bpf is abandonware by now,
and we should slowly but surely convert whatever makes sense into BPF
selftests under tools/testing/selftests/bpf and just get rid of the
rest.
> diff --git a/samples/bpf/tracex2.bpf.c b/samples/bpf/tracex2.bpf.c
> index 0a5c75b367be..dc3d91b65a6f 100644
> --- a/samples/bpf/tracex2.bpf.c
> +++ b/samples/bpf/tracex2.bpf.c
> @@ -17,20 +17,15 @@ struct {
> __uint(max_entries, 1024);
> } my_map SEC(".maps");
>
> -/* kprobe is NOT a stable ABI. If kernel internals change this bpf+kprobe
> - * example will no longer be meaningful
> - */
> -SEC("kprobe/kfree_skb_reason")
> -int bpf_prog2(struct pt_regs *ctx)
> +SEC("tracepoint/skb/kfree_skb")
> +int bpf_prog1(struct trace_event_raw_kfree_skb *ctx)
> {
> long loc = 0;
> long init_val = 1;
> long *value;
>
> - /* read ip of kfree_skb_reason caller.
> - * non-portable version of __builtin_return_address(0)
> - */
> - BPF_KPROBE_READ_RET_IP(loc, ctx);
> + /* read ip */
> + loc = (long)ctx->location;
>
> value = bpf_map_lookup_elem(&my_map, &loc);
> if (value)
> --
> 2.46.0
>
>
