Re: [PATCH 16/21] tracing/uprobe: Add per-probe delete from event

[Masami Hiramatsu]

On Mon, 17 Jun 2019 22:16:46 -0400
Steven Rostedt <rostedt@xxxxxxxxxxx> wrote:

> On Sat,  1 Jun 2019 00:19:17 +0900
> Masami Hiramatsu <mhiramat@xxxxxxxxxx> wrote:
> 
> > +static bool trace_uprobe_match_command_head(struct trace_uprobe *tu,
> > +					    int argc, const char **argv)
> > +{
> > +	char buf[MAX_ARGSTR_LEN + 1];
> > +	int len;
> > +
> > +	if (!argc)
> > +		return true;
> > +
> > +	len = strlen(tu->filename);
> > +	if (argv[0][len] != ':' || strncmp(tu->filename, argv[0], len))
> 
> Hmm, isn't it possible that 'len' can be greater than whatever argv[0] is?
> 
> The argv[0][len] looks very dangerous to me.

Ah, right! it can lead an unexpected memory access! 

> 
> Perhaps that should be changed to:
> 
> 	if (!(!strncmp(tu->filename, argv[0], len) && argv[0][len] == ':'))
> 
> That way, the test of argv[0][len] will only happen if argv[0] is of length len.

OK, I'll take it! Thank you!

> 
> -- Steve
> 
> 
> > +		return false;
> > +
> > +	if (tu->ref_ctr_offset == 0)
> > +		snprintf(buf, sizeof(buf), "0x%0*lx",
> > +				(int)(sizeof(void *) * 2), tu->offset);
> > +	else
> > +		snprintf(buf, sizeof(buf), "0x%0*lx(0x%lx)",
> > +				(int)(sizeof(void *) * 2), tu->offset,
> > +				tu->ref_ctr_offset);
> > +	if (strcmp(buf, &argv[0][len + 1]))
> > +		return false;
> > +
> > +	argc--; argv++;
> > +
> > +	return trace_probe_match_command_args(&tu->tp, argc, argv);
> > +}
> > +


-- 
Masami Hiramatsu <mhiramat@xxxxxxxxxx>