Re: [PATCHv7 bpf-next 13/29] bpf: Add support for tracing multi link

[sashiko-bot@kernel.org]

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Low] The patch references multiple undefined symbols (`struct bpf_tracing_multi_link`, `bpf_trampoline_multi_attach`, `bpf_trampoline_multi_detach`, and `is_tracing_multi`), which results in compilation failures.
--

commit 4e3377b79034e4eafa802e1a2382a4ca893a956f
Author: Jiri Olsa <jolsa@kernel.org>

bpf: Add support for tracing multi link

This commit adds a new link type to allow attaching a BPF program to
multiple function BTF IDs. It introduces the tracing_multi configuration
fields in bpf_attr::link_create and implements the attachment logic.

> diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c
> index 625a4366fe6d7..6f72ede0c8f04 100644
> --- a/kernel/bpf/syscall.c
> +++ b/kernel/bpf/syscall.c
> @@ -5842,6 +5842,8 @@ static int link_create(union bpf_attr *attr, bpfptr_t uattr)
>  			ret = bpf_iter_link_attach(attr, uattr, prog);
>  		else if (prog->expected_attach_type == BPF_LSM_CGROUP)
>  			ret = cgroup_bpf_link_attach(attr, prog);
> +		else if (is_tracing_multi(prog->expected_attach_type))

[Severity: Low]
Is is_tracing_multi() defined anywhere in this series? It appears this might
cause an implicit declaration error during the build, as the function or macro
seems to be missing.

> +			ret = bpf_tracing_multi_attach(prog, attr);
>  		else
>  			ret = bpf_tracing_prog_attach(prog,

[ ... ]

> diff --git a/kernel/trace/bpf_trace.c b/kernel/trace/bpf_trace.c
> index d853f97bd154f..9e3cb547651e5 100644
> --- a/kernel/trace/bpf_trace.c
> +++ b/kernel/trace/bpf_trace.c
> @@ -3641,3 +3642,132 @@ __bpf_kfunc int bpf_copy_from_user_task_str_dynptr(const struct bpf_dynptr *dptr
>  }
>  
>  __bpf_kfunc_end_defs();
> +
> +#if defined(CONFIG_DYNAMIC_FTRACE_WITH_DIRECT_CALLS) && \
> +    defined(CONFIG_HAVE_SINGLE_FTRACE_DIRECT_OPS)
> +
> +static void bpf_tracing_multi_link_release(struct bpf_link *link)
> +{
> +	struct bpf_tracing_multi_link *tr_link =
> +		container_of(link, struct bpf_tracing_multi_link, link);
> +
> +	WARN_ON_ONCE(bpf_trampoline_multi_detach(link->prog, tr_link));

[Severity: Low]
Where is bpf_trampoline_multi_detach() declared? It looks like the declaration
and definition for this function are missing, which will break the build.

> +}

[ ... ]

> +int bpf_tracing_multi_attach(struct bpf_prog *prog, const union bpf_attr *attr)
> +{
> +	struct bpf_tracing_multi_link *link = NULL;
> +	struct bpf_link_primer link_primer;
> +	u32 cnt, *ids = NULL;

[ ... ]

> +	err = check_dup_ids(ids, cnt);
> +	if (err)
> +		goto error;
> +
> +	link = kvzalloc_flex(*link, nodes, cnt);

[Severity: Low]
Is struct bpf_tracing_multi_link defined anywhere? Evaluating the size of an
incomplete type in this allocation macro will lead to a compiler error.

> +	if (!link) {
> +		err = -ENOMEM;
> +		goto error;
> +	}

[ ... ]

> +	link->nodes_cnt = cnt;
> +
> +	err = bpf_trampoline_multi_attach(prog, ids, link);

[Severity: Low]
Does bpf_trampoline_multi_attach() exist yet? Like the detach function, this
seems to be undeclared and undefined, leading to a build failure.

> +	kvfree(ids);

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260603110554.29590-14-jolsa@kernel.org?part=1