Re: [PATCH 4/4] kprobes: Remove core dependency on modules

[Google]

On Wed, 27 Mar 2024 19:46:50 +0200
"Jarkko Sakkinen" <jarkko@xxxxxxxxxx> wrote:

> On Wed Mar 27, 2024 at 2:01 AM EET, Masami Hiramatsu (Google) wrote:
> > On Tue, 26 Mar 2024 17:38:18 +0000
> > Mark Rutland <mark.rutland@xxxxxxx> wrote:
> >
> > > On Tue, Mar 26, 2024 at 07:13:51PM +0200, Jarkko Sakkinen wrote:
> > > > On Tue Mar 26, 2024 at 6:36 PM EET, Mark Rutland wrote:
> > > 
> > > > > +#ifdef CONFIG_MODULES
> > > > >  	/* Check if 'p' is probing a module. */
> > > > >  	*probed_mod = __module_text_address((unsigned long) p->addr);
> > > > >  	if (*probed_mod) {
> > > > > @@ -1605,6 +1606,8 @@ static int check_kprobe_address_safe(struct kprobe *p,
> > > > >  			ret = -ENOENT;
> > > > >  		}
> > > > >  	}
> > > > > +#endif
> > > > 
> > > > This can be scoped a bit more (see v7 of my patch set).
> > > 
> > > > > +#ifdef CONFIG_MODULES
> > > > >  static nokprobe_inline bool trace_kprobe_module_exist(struct trace_kprobe *tk)
> > > > >  {
> > > > >  	char *p;
> > > > > @@ -129,6 +130,9 @@ static nokprobe_inline bool trace_kprobe_module_exist(struct trace_kprobe *tk)
> > > > >  
> > > > >  	return ret;
> > > > >  }
> > > > > +#else
> > > > > +#define trace_kprobe_module_exist(tk) false /* aka a module never exists */
> > > > > +#endif /* CONFIG_MODULES */
> > > > >  
> > > > >  static bool trace_kprobe_is_busy(struct dyn_event *ev)
> > > > >  {
> > > > > @@ -670,6 +674,7 @@ static int register_trace_kprobe(struct trace_kprobe *tk)
> > > > >  	return ret;
> > > > >  }
> > > > >  
> > > > > +#ifdef CONFIG_MODULES
> > > > >  /* Module notifier call back, checking event on the module */
> > > > >  static int trace_kprobe_module_callback(struct notifier_block *nb,
> > > > >  				       unsigned long val, void *data)
> > > > > @@ -699,6 +704,9 @@ static int trace_kprobe_module_callback(struct notifier_block *nb,
> > > > >  
> > > > >  	return NOTIFY_DONE;
> > > > >  }
> > > > > +#else
> > > > > +#define trace_kprobe_module_callback (NULL)
> > > > > +#endif /* CONFIG_MODULES */
> > > > 
> > > > The last two CONFIG_MODULES sections could be combined. This was also in
> > > > v7.
> > > 
> > > > Other than lgtm.
> > > 
> > > Great! I've folded your v7 changes in, and pushed that out to:
> > > 
> > >   https://git.kernel.org/pub/scm/linux/kernel/git/mark/linux.git/log/?h=kprobes/without-modules
> > > 
> > > I'll hold off sending that out to the list until other folk have had a chance
> > > to comment.
> >
> > Yeah, the updated one looks good to me too.
> >
> > Thanks!
> 
> As for RISC-V:
> 
> Tested-by: Jarkko Sakkinen <jarkko@xxxxxxxxxx> # arch/riscv

Thank you for testing!

> 
> I'm fine with adding to all patches because it would be hard
> to place tested-by to any specific patch (e.g. if this was a
> syscall I would give tested-by just for that patch).

Except for the 1st patch because that is for arm64, right? :)

> 
> Just adding disclaimer because depending on subsystem people
> are more or less strict with this tag :-)
> 
> BR, Jarkko

Thanks,

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