Linux bluetooth development
 help / color / mirror / Atom feed
* [PATCH] Fix invalid read and possible memory leaks
@ 2011-05-27 11:26 Dmitriy Paliy
  2011-05-27 11:56 ` Luiz Augusto von Dentz
  0 siblings, 1 reply; 4+ messages in thread
From: Dmitriy Paliy @ 2011-05-27 11:26 UTC (permalink / raw)
  To: linux-bluetooth; +Cc: Dmitriy Paliy

Fixed incorrect update of transport->owners GSlist in media_transport_free.
Removal of list entries within 'for' loop leads to invalid read of memory
(l = l->next) and memory leaks.
---
 audio/transport.c |   15 +++++++++++----
 1 files changed, 11 insertions(+), 4 deletions(-)

diff --git a/audio/transport.c b/audio/transport.c
index aa3308d..4047d80 100644
--- a/audio/transport.c
+++ b/audio/transport.c
@@ -805,14 +805,21 @@ static GDBusSignalTable transport_signals[] = {
 	{ }
 };
 
+static void media_transport_remove_helper(gpointer data, gpointer user_data)
+{
+	struct media_transport *transport = user_data;
+	struct media_owner *owner = data;
+
+	media_transport_remove(transport, owner);
+}
+
+
 static void media_transport_free(void *data)
 {
 	struct media_transport *transport = data;
-	GSList *l;
-
-	for (l = transport->owners; l; l = l->next)
-		media_transport_remove(transport, l->data);
 
+	g_slist_foreach(transport->owners, media_transport_remove_helper,
+								transport);
 	g_slist_free(transport->owners);
 
 	if (transport->session)
-- 
1.7.4.1


^ permalink raw reply related	[flat|nested] 4+ messages in thread

* Re: [PATCH] Fix invalid read and possible memory leaks
  2011-05-27 11:26 [PATCH] Fix invalid read and possible memory leaks Dmitriy Paliy
@ 2011-05-27 11:56 ` Luiz Augusto von Dentz
  2011-05-27 12:03   ` Dmitriy Paliy
  0 siblings, 1 reply; 4+ messages in thread
From: Luiz Augusto von Dentz @ 2011-05-27 11:56 UTC (permalink / raw)
  To: Dmitriy Paliy; +Cc: linux-bluetooth

Hi Dmitry,

On Fri, May 27, 2011 at 2:26 PM, Dmitriy Paliy <dmitriy.paliy@nokia.com> wrote:
> Fixed incorrect update of transport->owners GSlist in media_transport_free.
> Removal of list entries within 'for' loop leads to invalid read of memory
> (l = l->next) and memory leaks.
> ---
>  audio/transport.c |   15 +++++++++++----
>  1 files changed, 11 insertions(+), 4 deletions(-)
>
> diff --git a/audio/transport.c b/audio/transport.c
> index aa3308d..4047d80 100644
> --- a/audio/transport.c
> +++ b/audio/transport.c
> @@ -805,14 +805,21 @@ static GDBusSignalTable transport_signals[] = {
>        { }
>  };
>
> +static void media_transport_remove_helper(gpointer data, gpointer user_data)
> +{
> +       struct media_transport *transport = user_data;
> +       struct media_owner *owner = data;
> +
> +       media_transport_remove(transport, owner);
> +}
> +
> +
>  static void media_transport_free(void *data)
>  {
>        struct media_transport *transport = data;
> -       GSList *l;
> -
> -       for (l = transport->owners; l; l = l->next)
> -               media_transport_remove(transport, l->data);
>
> +       g_slist_foreach(transport->owners, media_transport_remove_helper,
> +                                                               transport);
>        g_slist_free(transport->owners);
>
>        if (transport->session)
> --
> 1.7.4.1

Usually we should just use while (list) and do the list = list->next
before item removal.

-- 
Luiz Augusto von Dentz
Computer Engineer

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: [PATCH] Fix invalid read and possible memory leaks
  2011-05-27 11:56 ` Luiz Augusto von Dentz
@ 2011-05-27 12:03   ` Dmitriy Paliy
  2011-05-27 13:03     ` Luiz Augusto von Dentz
  0 siblings, 1 reply; 4+ messages in thread
From: Dmitriy Paliy @ 2011-05-27 12:03 UTC (permalink / raw)
  To: ext Luiz Augusto von Dentz; +Cc: linux-bluetooth

Hi Luiz,

On Fri, 2011-05-27 at 14:56 +0300, ext Luiz Augusto von Dentz wrote:
> Hi Dmitry,
> 
> On Fri, May 27, 2011 at 2:26 PM, Dmitriy Paliy <dmitriy.paliy@nokia.com> wrote:
> > Fixed incorrect update of transport->owners GSlist in media_transport_free.
> > Removal of list entries within 'for' loop leads to invalid read of memory
> > (l = l->next) and memory leaks.
> > ---
> >  audio/transport.c |   15 +++++++++++----
> >  1 files changed, 11 insertions(+), 4 deletions(-)
> >
> > diff --git a/audio/transport.c b/audio/transport.c
> > index aa3308d..4047d80 100644
> > --- a/audio/transport.c
> > +++ b/audio/transport.c
> > @@ -805,14 +805,21 @@ static GDBusSignalTable transport_signals[] = {
> >        { }
> >  };
> >
> > +static void media_transport_remove_helper(gpointer data, gpointer user_data)
> > +{
> > +       struct media_transport *transport = user_data;
> > +       struct media_owner *owner = data;
> > +
> > +       media_transport_remove(transport, owner);
> > +}
> > +
> > +
> >  static void media_transport_free(void *data)
> >  {
> >        struct media_transport *transport = data;
> > -       GSList *l;
> > -
> > -       for (l = transport->owners; l; l = l->next)
> > -               media_transport_remove(transport, l->data);
> >
> > +       g_slist_foreach(transport->owners, media_transport_remove_helper,
> > +                                                               transport);
> >        g_slist_free(transport->owners);
> >
> >        if (transport->session)
> > --
> > 1.7.4.1
> 
> Usually we should just use while (list) and do the list = list->next
> before item removal.

This is how it is done in the g_slist_foreach function. Therefore, may
be it is better to use ready function without exposing implementation
details of this loop.

BR,
Dmitriy



^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: [PATCH] Fix invalid read and possible memory leaks
  2011-05-27 12:03   ` Dmitriy Paliy
@ 2011-05-27 13:03     ` Luiz Augusto von Dentz
  0 siblings, 0 replies; 4+ messages in thread
From: Luiz Augusto von Dentz @ 2011-05-27 13:03 UTC (permalink / raw)
  To: Dmitriy Paliy; +Cc: linux-bluetooth

Hi Dmitriy,

On Fri, May 27, 2011 at 3:03 PM, Dmitriy Paliy <dmitriy.paliy@nokia.com> wrote:
> Hi Luiz,
>
> On Fri, 2011-05-27 at 14:56 +0300, ext Luiz Augusto von Dentz wrote:
>> Hi Dmitry,
>>
>> On Fri, May 27, 2011 at 2:26 PM, Dmitriy Paliy <dmitriy.paliy@nokia.com> wrote:
>> > Fixed incorrect update of transport->owners GSlist in media_transport_free.
>> > Removal of list entries within 'for' loop leads to invalid read of memory
>> > (l = l->next) and memory leaks.
>> > ---
>> >  audio/transport.c |   15 +++++++++++----
>> >  1 files changed, 11 insertions(+), 4 deletions(-)
>> >
>> > diff --git a/audio/transport.c b/audio/transport.c
>> > index aa3308d..4047d80 100644
>> > --- a/audio/transport.c
>> > +++ b/audio/transport.c
>> > @@ -805,14 +805,21 @@ static GDBusSignalTable transport_signals[] = {
>> >        { }
>> >  };
>> >
>> > +static void media_transport_remove_helper(gpointer data, gpointer user_data)
>> > +{
>> > +       struct media_transport *transport = user_data;
>> > +       struct media_owner *owner = data;
>> > +
>> > +       media_transport_remove(transport, owner);
>> > +}
>> > +
>> > +
>> >  static void media_transport_free(void *data)
>> >  {
>> >        struct media_transport *transport = data;
>> > -       GSList *l;
>> > -
>> > -       for (l = transport->owners; l; l = l->next)
>> > -               media_transport_remove(transport, l->data);
>> >
>> > +       g_slist_foreach(transport->owners, media_transport_remove_helper,
>> > +                                                               transport);
>> >        g_slist_free(transport->owners);
>> >
>> >        if (transport->session)
>> > --
>> > 1.7.4.1
>>
>> Usually we should just use while (list) and do the list = list->next
>> before item removal.
>
> This is how it is done in the g_slist_foreach function. Therefore, may
> be it is better to use ready function without exposing implementation
> details of this loop.

Well normally it is more code to do it via g_slist_foreach than on
while loop, it is probably more expensive too since it has to call a
callback for each element and obviously the g_slist_foreach has to
take care of prefetch next element.

-- 
Luiz Augusto von Dentz
Computer Engineer

^ permalink raw reply	[flat|nested] 4+ messages in thread

end of thread, other threads:[~2011-05-27 13:03 UTC | newest]

Thread overview: 4+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2011-05-27 11:26 [PATCH] Fix invalid read and possible memory leaks Dmitriy Paliy
2011-05-27 11:56 ` Luiz Augusto von Dentz
2011-05-27 12:03   ` Dmitriy Paliy
2011-05-27 13:03     ` Luiz Augusto von Dentz

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox