Re: [PATCH bpf-next 10/11] samples: bpf: fix -Wsequence-point

[Song Liu <song@kernel.org>]

On Thu, Apr 14, 2022 at 3:47 PM Alexander Lobakin <alobakin@pm.me> wrote:
>
> In some libc implementations, CPU_SET() may utilize its first
> argument several times. When combined with a post-increment, it
> leads to:
>
> samples/bpf/test_lru_dist.c:233:36: warning: operation on 'next_to_try' may be undefined [-Wsequence-point]
>   233 |                 CPU_SET(next_to_try++, &cpuset);
>       |                                    ^
>
> Split the sentence into two standalone operations to fix this.
>
> Fixes: 5db58faf989f ("bpf: Add tests for the LRU bpf_htab")
> Signed-off-by: Alexander Lobakin <alobakin@pm.me>

Acked-by: Song Liu <songliubraving@fb.com>

> ---
>  samples/bpf/test_lru_dist.c | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
>
> diff --git a/samples/bpf/test_lru_dist.c b/samples/bpf/test_lru_dist.c
> index be98ccb4952f..191643ec501e 100644
> --- a/samples/bpf/test_lru_dist.c
> +++ b/samples/bpf/test_lru_dist.c
> @@ -229,7 +229,8 @@ static int sched_next_online(int pid, int next_to_try)
>
>         while (next_to_try < nr_cpus) {
>                 CPU_ZERO(&cpuset);
> -               CPU_SET(next_to_try++, &cpuset);
> +               CPU_SET(next_to_try, &cpuset);
> +               next_to_try++;
>                 if (!sched_setaffinity(pid, sizeof(cpuset), &cpuset))
>                         break;
>         }
> --
> 2.35.2
>
>