On Wed, Aug 28, 2019 at 02:03:04PM -0700, Julia Kartseva wrote:
> Similar to __MAX_BPF_ATTACH_TYPE identifying the number of elements in
> bpf_attach_type enum, add tailing enum values __MAX_BPF_PROG_TYPE
> and __MAX_BPF_MAP_TYPE to simplify e.g. iteration over enums values in
> the case when new values are added.
>
> Signed-off-by: Julia Kartseva <h...@fb.com>
> ---
> include/uapi/linux/bpf.h | 6 ++++++
> 1 file changed, 6 insertions(+)
>
> diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h
> index 5d2fb183ee2d..9b681bb82211 100644
> --- a/include/uapi/linux/bpf.h
> +++ b/include/uapi/linux/bpf.h
> @@ -136,8 +136,11 @@ enum bpf_map_type {
> BPF_MAP_TYPE_STACK,
> BPF_MAP_TYPE_SK_STORAGE,
> BPF_MAP_TYPE_DEVMAP_HASH,
> + __MAX_BPF_MAP_TYPE
> };
>
> +#define MAX_BPF_MAP_TYPE __MAX_BPF_MAP_TYPE
> +
> /* Note that tracing related programs such as
> * BPF_PROG_TYPE_{KPROBE,TRACEPOINT,PERF_EVENT,RAW_TRACEPOINT}
> * are not subject to a stable API since kernel internal data
> @@ -173,8 +176,11 @@ enum bpf_prog_type {
> BPF_PROG_TYPE_CGROUP_SYSCTL,
> BPF_PROG_TYPE_RAW_TRACEPOINT_WRITABLE,
> BPF_PROG_TYPE_CGROUP_SOCKOPT,
> + __MAX_BPF_PROG_TYPE
> };
>
> +#define MAX_BPF_PROG_TYPE __MAX_BPF_PROG_TYPE
> +
This came up before and my position is still the same.
I'm against this type of band-aid in uapi.
'bpftool feature probe' can easily discover all supported
prog and map types already.
