Re: [PATCH v2 2/2] bpf: cgroup_iter: support cgroup1 using cgroup fd

[Martin KaFai Lau <martin.lau-fxUVXftIFDnyG1zEObXtfA@public.gmane.org>]

On 10/10/22 5:33 PM, Yosry Ahmed wrote:
> Use cgroup_v1v2_get_from_fd() in cgroup_iter to support attaching to
> both cgroup v1 and v2 using fds.
> 
> Signed-off-by: Yosry Ahmed <yosryahmed-hpIqsD4AKlfQT0dZR+AlfA@public.gmane.org>
> ---
>   kernel/bpf/cgroup_iter.c | 2 +-
>   1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/kernel/bpf/cgroup_iter.c b/kernel/bpf/cgroup_iter.c
> index 0d200a993489..9fcf09f2ef00 100644
> --- a/kernel/bpf/cgroup_iter.c
> +++ b/kernel/bpf/cgroup_iter.c
> @@ -196,7 +196,7 @@ static int bpf_iter_attach_cgroup(struct bpf_prog *prog,
>   		return -EINVAL;
>   
>   	if (fd)
> -		cgrp = cgroup_get_from_fd(fd);
> +		cgrp = cgroup_v1v2_get_from_fd(fd);

Acked-by: Martin KaFai Lau <martin.lau-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org>

Tejun, patch 1 should depend on a recent revert that is not in the bpf tree yet. 
  Do you want to take this set to the cgroup tree?

>   	else if (id)
>   		cgrp = cgroup_get_from_id(id);
>   	else /* walk the entire hierarchy by default. */