On Wed, Aug 14, 2024 at 4:58 PM JP Kobryn <inwardvessel@xxxxxxxxx> wrote: > > Account for normal tracepoint programs by associating them with the kfunc > tracing hook. This allows kfuncs to be called within tracepoint programs. > > Signed-off-by: JP Kobryn <inwardvessel@xxxxxxxxx> > --- > kernel/bpf/btf.c | 1 + > 1 file changed, 1 insertion(+) > > diff --git a/kernel/bpf/btf.c b/kernel/bpf/btf.c > index 520f49f422fe..8b844d6fd041 100644 > --- a/kernel/bpf/btf.c > +++ b/kernel/bpf/btf.c > @@ -8303,6 +8303,7 @@ static int bpf_prog_type_to_kfunc_hook(enum bpf_prog_type prog_type) > return BTF_KFUNC_HOOK_TC; > case BPF_PROG_TYPE_STRUCT_OPS: > return BTF_KFUNC_HOOK_STRUCT_OPS; > + case BPF_PROG_TYPE_TRACEPOINT: > case BPF_PROG_TYPE_TRACING: > case BPF_PROG_TYPE_LSM: > return BTF_KFUNC_HOOK_TRACING; > -- > 2.46.0 > I'm not 100% sure it's ok to map TRACEPOINT prog type to HOOK_TRACING. But assuming it is, shouldn't we then also do the same for KPROBE and PERF_EVENT programs? Either way, please consider Eduard's suggestion about changing patch order and switching to TEST_LOADER approach. pw-bot: cr
