Re: [PATCH v3 2/8] fprobe: Use fprobe_regs in fprobe entry handler

[Google]

On Thu, 17 Aug 2023 10:57:26 +0200
Florent Revest <revest@xxxxxxxxxxxx> wrote:

> On Sat, Aug 12, 2023 at 7:37 AM Masami Hiramatsu (Google)
> <mhiramat@xxxxxxxxxx> wrote:
> > --- a/kernel/trace/bpf_trace.c
> > +++ b/kernel/trace/bpf_trace.c
> > @@ -2467,7 +2467,7 @@ static int __init bpf_event_init(void)
> >  fs_initcall(bpf_event_init);
> >  #endif /* CONFIG_MODULES */
> >
> > -#ifdef CONFIG_FPROBE
> > +#ifdef CONFIG_DYNAMIC_FTRACE_WITH_REGS
> 
> Shouldn't this be #if defined(CONFIG_FPROBE) &&
> defined(CONFIG_DYNAMIC_FTRACE_WITH_REGS) ?

Oops, that's right!

> 
> I believe one could build a kernel with FTRACE_WITH_REGS and without
> FPROBE and then this code would have undefined references to fprobe
> functions, wouldn't it ?

Yeah, ftrace with regs doesn't mean fprobe is enabled.

> 
> And then patch 7 should be "Enable kprobe_multi feature even if
> FTRACE_WITH_REGS is disabled"

OK.

Thank you!


-- 
Masami Hiramatsu (Google) <mhiramat@xxxxxxxxxx>