Re: [RFC PATCH 2/4] memcg_ext: add cgroup-attached bpf_memcg_ops

[sashiko-bot@kernel.org]

Thank you for your contribution! Sashiko AI review found 2 potential issue(s) to consider:
- [Medium] bpf: spurious boot warning when CONFIG_CGROUP_BPF is disabled
- [Low] bpf: missing newline in pr_warn for bpf_memcg_ops registration
--

commit f28f15056753b0bfc8cefa4ccf7793b6fb45150d
Author: Shakeel Butt <shakeel.butt@linux.dev>
Subject: memcg_ext: add cgroup-attached bpf_memcg_ops

This patch adds an empty bpf_memcg_ops type for attaching memory policies
to cgroups. It also moves the cgroup BPF attach types outside
CONFIG_CGROUP_BPF and updates the struct_ops registration stub.

> diff --git a/mm/bpf_memcontrol.c b/mm/bpf_memcontrol.c
> index d8f579c285600..fd6dff150f017 100644
> --- a/mm/bpf_memcontrol.c
> +++ b/mm/bpf_memcontrol.c
[ ... ]
> @@ -248,8 +348,14 @@ static int __init bpf_memcontrol_init(void)
>  
>  	err = register_btf_kfunc_id_set(BPF_PROG_TYPE_SYSCALL,
>  					&bpf_memcontrol_reclaim_kfunc_set);
> -	if (err)
> +	if (err) {
>  		pr_warn("error registering bpf reclaim kfuncs: %d\n", err);
> +		return err;
> +	}
> +
> +	err = bpf_memcg_ops_register();
> +	if (err)
> +		pr_warn("error while registering bpf_memcg_ops: %d", err);

[Severity: Medium]
Will this print a spurious warning on every boot when CONFIG_CGROUP_BPF is
disabled?

If a kernel is built with CONFIG_MEMCG=y, CONFIG_BPF_SYSCALL=y, and
CONFIG_BPF_JIT=y, but CONFIG_CGROUP_BPF=n, bpf_memcontrol_init() calls
bpf_memcg_ops_register().

This eventually calls register_bpf_struct_ops(), which relies on
cgroup_bpf_is_struct_ops_atype() in btf_add_struct_ops(). Since the
!CONFIG_CGROUP_BPF stub for cgroup_bpf_is_struct_ops_atype() unconditionally
returns false, this returns -EINVAL.

Does this mean the late initcall will return an error and unconditionally
pollute the kernel log with this warning on supported configurations?

[Severity: Low]
Does this pr_warn() need a terminating '\n'?

Without it, subsequent printk messages might be incorrectly appended to the
same line in the kernel log.

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260921192559.2619635-1-shakeel.butt@linux.dev?part=2