Re: [RFC PATCH 2/6] bpf: add BPF_MAP_DUMP command to access more than one entry per call

[Brian Vazquez <brianvv.kernel@xxxxxxxxx>]

On Mon, Jun 24, 2019 at 3:46 PM Jakub Kicinski
<jakub.kicinski@xxxxxxxxxxxxx> wrote:
>
> On Fri, 21 Jun 2019 16:16:46 -0700, Brian Vazquez wrote:
> > @@ -385,6 +386,14 @@ union bpf_attr {
> >               __u64           flags;
> >       };
> >
> > +     struct { /* struct used by BPF_MAP_DUMP command */
> > +             __u32           map_fd;
>
> There is a hole here, perhaps flags don't have to be 64 bit?
The command implementation is wrapping BPF_MAP_*_ELEM commands, I
would expect this one to handle the same flags which are 64 bit.
Note that there's a hole in the anonymous structure used by the other
commands too:
        struct { /* anonymous struct used by BPF_MAP_*_ELEM commands */
                __u32           map_fd;
                __aligned_u64   key;
                union {
                        __aligned_u64 value;
                        __aligned_u64 next_key;
                };
                __u64           flags;
        };