Re: [PATCH v5 5/5] tracing: Adopt __free() and guard() for trace_fprobe.c

[Google]

On Wed, 8 Jan 2025 10:07:36 -0500
Steven Rostedt <rostedt@xxxxxxxxxxx> wrote:

> On Wed,  8 Jan 2025 11:11:30 +0900
> "Masami Hiramatsu (Google)" <mhiramat@xxxxxxxxxx> wrote:
> 
> > @@ -1234,24 +1223,26 @@ static int __trace_fprobe_create(int argc, const char *argv[])
> >  			trace_probe_log_err(0, BAD_PROBE_ADDR);
> >  		else if (ret != -ENOMEM && ret != -EEXIST)
> >  			trace_probe_log_err(0, FAIL_REG_PROBE);
> > -		goto error;
> > -	}
> > +		ret = -EINVAL;
> > +	} else
> > +		/* 'tf' is successfully registered. To avoid freeing, assign NULL. */
> > +		tf = NULL;
> >  
> > -out:
> > -	if (tp_mod)
> > -		module_put(tp_mod);
> > +	return ret;
> > +}
> > +
> 
> Hmm, the above could probably be simplified as:
> 
> 	ret = register_trace_fprobe(tf);
> 	if (ret) {
> 		trace_probe_log_set_index(1);
> 		if (ret == -EILSEQ)
> 			trace_probe_log_err(0, BAD_INSN_BNDRY);
> 		else if (ret == -ENOENT)
> 			trace_probe_log_err(0, BAD_PROBE_ADDR);
> 		else if (ret != -ENOMEM && ret != -EEXIST)
> 			trace_probe_log_err(0, FAIL_REG_PROBE);
> 		return -EINVAL;
> 	}
> 
> 	/* 'tf' is successfully registered. To avoid freeing, assign NULL. */
> 	tf = NULL;
> 
> 	return 0;

Indeed. Thanks!

>  }
> 
> -- Steve


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