Re: [PATCH v2 bpf-next 1/2] libbpf: add helpers for mmapping maps

[Andrii Nakryiko]

On Wed, Jan 3, 2024 at 10:54 AM Barret Rhoden <brho@xxxxxxxxxx> wrote:
>
> bpf_map__mmap_size() was internal to bpftool.  Use that to make wrappers
> for mmap and munmap.
>
> Signed-off-by: Barret Rhoden <brho@xxxxxxxxxx>
> ---
>  tools/bpf/bpftool/gen.c  | 16 +++-------------
>  tools/lib/bpf/libbpf.c   | 18 ++++++++++++++++++
>  tools/lib/bpf/libbpf.h   |  6 ++++++
>  tools/lib/bpf/libbpf.map |  4 ++++
>  4 files changed, 31 insertions(+), 13 deletions(-)
>

There is very little added value provided by these APIs, while adding
API obligations. bpf_map__mmap() assumes READ/WRITE access, why? What
if the user needs only read-only? And so on.

Please drop this patch, we don't need to expose these functions as
stable APIs. Definitely not bpf_map__mmap/munmap. bpf_map__mmap_size()
might be useful to hide various per-map type details, but we'll need
to discuss all this separately.

> diff --git a/tools/bpf/bpftool/gen.c b/tools/bpf/bpftool/gen.c
> index ee3ce2b8000d..a328e960c141 100644
> --- a/tools/bpf/bpftool/gen.c
> +++ b/tools/bpf/bpftool/gen.c
> @@ -453,16 +453,6 @@ static void print_hex(const char *data, int data_sz)
>         }
>  }
>
> -static size_t bpf_map_mmap_sz(const struct bpf_map *map)
> -{
> -       long page_sz = sysconf(_SC_PAGE_SIZE);
> -       size_t map_sz;
> -
> -       map_sz = (size_t)roundup(bpf_map__value_size(map), 8) * bpf_map__max_entries(map);

this is specifically ARRAY map's rules, it might differ for other map
types (e.g., RINGBUF has different logic)

> -       map_sz = roundup(map_sz, page_sz);
> -       return map_sz;
> -}
> -

[...]