Re: [PATCH] kobject_set_name_vargs memory leak

[Kay Sievers <kay.sievers@vrfy.org>]

On Sat, 2009-06-27 at 12:39 +0300, Sergey Senozhatsky wrote:

> > >> Or something along those lines, I can't remember the exact reasoning
> > >> this early in the morning.
> > >> 
> > >> Kay, do you remember?

Hmm, yes, I think there was only something to work around during the
transition from the static name array, which is gone now. At least I
can't see anything we need to care about with the current code.

> Sorry, correct one.
> 
> diff --git a/lib/kobject.c b/lib/kobject.c
> index b512b74..3ab224b 100644
> --- a/lib/kobject.c
> +++ b/lib/kobject.c
> @@ -215,7 +215,6 @@ static int kobject_add_internal(struct kobject *kobj)
>  int kobject_set_name_vargs(struct kobject *kobj, const char *fmt,
>  				  va_list vargs)
>  {
> -	const char *old_name = kobj->name;

I guess, that would leak an allocated name, when it is set several times
in a row? Something like this?

Thanks,
Kay

diff --git a/lib/kobject.c b/lib/kobject.c
index b512b74..780e89f 100644
--- a/lib/kobject.c
+++ b/lib/kobject.c
@@ -215,21 +215,22 @@ static int kobject_add_internal(struct kobject *kobj)
 int kobject_set_name_vargs(struct kobject *kobj, const char *fmt,
 				  va_list vargs)
 {
-	const char *old_name = kobj->name;
+	const char *name = kobj->name;
 	char *s;
 
 	if (kobj->name && !fmt)
 		return 0;
 
-	kobj->name = kvasprintf(GFP_KERNEL, fmt, vargs);
-	if (!kobj->name)
+	name = kvasprintf(GFP_KERNEL, fmt, vargs);
+	if (!name)
 		return -ENOMEM;
 
 	/* ewww... some of these buggers have '/' in the name ... */
-	while ((s = strchr(kobj->name, '/')))
+	while ((s = strchr(name, '/')))
 		s[0] = '!';
 
-	kfree(old_name);
+	kfree(kobj->name);
+	kobj->name = name;
 	return 0;
 }