Re: [PATCH bpf-next v2 1/2] bpf: Add show_fdinfo for uprobe_multi

[Tao Chen]

在 2025/6/16 15:40, Jiri Olsa 写道:
> On Sun, Jun 15, 2025 at 11:05:13PM +0800, Tao Chen wrote:
> > Show uprobe_multi link info with fdinfo, the info as follows:
> >
> > link_type:	uprobe_multi
> > link_id:	9
> > prog_tag:	e729f789e34a8eca
> > prog_id:	58
> > type:	uprobe_multi
> > uprobe_cnt:	3
> > pid:	0
> > path:	/home/dylane/bpf/tools/testing/selftests/bpf/test_progs
> > offset:	0xa69ed7
> > ref_ctr_offset:	0x0
> > cookie:	3
> > offset:	0xa69ee2
> > ref_ctr_offset:	0x0
> > cookie:	1
> > offset:	0xa69eed
> > ref_ctr_offset:	0x0
> > cookie:	2
>
> hi,
> does this need to be 'tag: value' format ? bpftool uses:
>
>          offset             ref_ctr_offset     cookies
>          0xe558             0x0                0x0
>          0x2574e            0x0                0x0
>          0x6c393            0x0                0x0
>
> which might be more readable, or at least extra line after each uprobe?
> also using spaces instead of tabs  to align the values might help
>
> same for kprobe_multi, otherwise looks lgtm

That's a great suggestion, it does look more concise. I will change it 
in v3, thanks!

> thanks,
> jirka
>
>
> > Signed-off-by: Tao Chen <chen.dylane@xxxxxxxxx>
> > ---
> >   kernel/trace/bpf_trace.c | 48 ++++++++++++++++++++++++++++++++++++++++
> >   1 file changed, 48 insertions(+)
> >
> > Change list:
> >    v1 -> v2:
> >      - replace 'func_cnt' with 'uprobe_cnt'.(Andrii)
> >      - print func name is more readable and security for kprobe_multi.(Alexei)
> >    v1:
> >    https://lore.kernel.org/bpf/20250612115556.295103-1-chen.dylane@xxxxxxxxx
> >
> > diff --git a/kernel/trace/bpf_trace.c b/kernel/trace/bpf_trace.c
> > index 24b94870b50..9a8ca8a8e2b 100644
> > --- a/kernel/trace/bpf_trace.c
> > +++ b/kernel/trace/bpf_trace.c
> > @@ -3157,10 +3157,58 @@ static int bpf_uprobe_multi_link_fill_link_info(const struct bpf_link *link,
> >   	return err;
> >   }
> >   
> > +#ifdef CONFIG_PROC_FS
> > +static void bpf_uprobe_multi_show_fdinfo(const struct bpf_link *link,
> > +					 struct seq_file *seq)
> > +{
> > +	struct bpf_uprobe_multi_link *umulti_link;
> > +	char *p, *buf;
> > +
> > +	umulti_link = container_of(link, struct bpf_uprobe_multi_link, link);
> > +
> > +	buf = kmalloc(PATH_MAX, GFP_KERNEL);
> > +	if (!buf)
> > +		return;
> > +
> > +	p = d_path(&umulti_link->path, buf, PATH_MAX);
> > +	if (IS_ERR(p)) {
> > +		kfree(buf);
> > +		return;
> > +	}
> > +
> > +	seq_printf(seq,
> > +		   "type:\t%s\n"
> > +		   "uprobe_cnt:\t%u\n"
> > +		   "pid:\t%u\n"
> > +		   "path:\t%s\n",
> > +		   umulti_link->flags == BPF_F_UPROBE_MULTI_RETURN ?
> > +					 "uretprobe_multi" : "uprobe_multi",
> > +		   umulti_link->cnt,
> > +		   umulti_link->task ? task_pid_nr_ns(umulti_link->task,
> > +			   task_active_pid_ns(current)) : 0,
> > +		   p);
> > +
> > +	for (int i = 0; i < umulti_link->cnt; i++) {
> > +		seq_printf(seq,
> > +			   "offset:\t%#llx\n"
> > +			   "ref_ctr_offset:\t%#lx\n"
> > +			   "cookie:\t%llu\n",
> > +			   umulti_link->uprobes[i].offset,
> > +			   umulti_link->uprobes[i].ref_ctr_offset,
> > +			   umulti_link->uprobes[i].cookie);
> > +	}
> > +
> > +	kfree(buf);
> > +}
> > +#endif
> > +
> >   static const struct bpf_link_ops bpf_uprobe_multi_link_lops = {
> >   	.release = bpf_uprobe_multi_link_release,
> >   	.dealloc_deferred = bpf_uprobe_multi_link_dealloc,
> >   	.fill_link_info = bpf_uprobe_multi_link_fill_link_info,
> > +#ifdef CONFIG_PROC_FS
> > +	.show_fdinfo = bpf_uprobe_multi_show_fdinfo,
> > +#endif
> >   };
> >   
> >   static int uprobe_prog_run(struct bpf_uprobe *uprobe,
> > --
> > 2.48.1


--
Best Regards
Tao Chen