Hi,
On 9/22/2021 5:42 AM, Andrii Nakryiko wrote:
> On Fri, Sep 17, 2021 at 6:56 PM Hou Tao <houtao1@xxxxxxxxxx> wrote:
>> Program on writable tracepoint is BPF_PROG_TYPE_RAW_TRACEPOINT_WRITABLE,
>> but its attachment is the same as BPF_PROG_TYPE_RAW_TRACEPOINT.
>>
>> Signed-off-by: Hou Tao <houtao1@xxxxxxxxxx>
>> ---
>> tools/lib/bpf/libbpf.c | 4 ++++
>> 1 file changed, 4 insertions(+)
>>
>> diff --git a/tools/lib/bpf/libbpf.c b/tools/lib/bpf/libbpf.c
>> index da65a1666a5e..981fcdd95bdc 100644
>> --- a/tools/lib/bpf/libbpf.c
>> +++ b/tools/lib/bpf/libbpf.c
>> @@ -7976,6 +7976,10 @@ static const struct bpf_sec_def section_defs[] = {
>> .attach_fn = attach_raw_tp),
>> SEC_DEF("raw_tp/", RAW_TRACEPOINT,
>> .attach_fn = attach_raw_tp),
>> + SEC_DEF("raw_tracepoint_writable/", RAW_TRACEPOINT_WRITABLE,
>> + .attach_fn = attach_raw_tp),
>> + SEC_DEF("raw_tp_writable/", RAW_TRACEPOINT_WRITABLE,
>> + .attach_fn = attach_raw_tp),
> _writable is a bit mouthful, maybe we should do the same we did for
> "sleepable", just add ".w" suffix? So it will be "raw_tp.w/..."? Or
> does anyone feel it's too subtle?
raw_tp.w is fine to me. Will update it in v3.
>
>> SEC_DEF("tp_btf/", TRACING,
>> .expected_attach_type = BPF_TRACE_RAW_TP,
>> .is_attach_btf = true,
>> --
>> 2.29.2
>>
> .
