Hi,
On Tue, Dec 05, 2017 at 08:41:05AM +0000, Frediano Ziglio wrote:
> This object was not freed.
AND, you are fixing the leak in the object itself, no?
One patch should be fine but adding that in commit log should do no
harm.
Still,
Acked-by: Victor Toso <victortoso@xxxxxxxxxx>
> Signed-off-by: Frediano Ziglio <fziglio@xxxxxxxxxx>
> ---
> server/reds.c | 9 ++++++++-
> 1 file changed, 8 insertions(+), 1 deletion(-)
>
> diff --git a/server/reds.c b/server/reds.c
> index 4e652253..40c82ccc 100644
> --- a/server/reds.c
> +++ b/server/reds.c
> @@ -3754,6 +3754,7 @@ SPICE_GNUC_VISIBLE void spice_server_destroy(SpiceServer *reds)
> reds_core_watch_remove(reds, reds->secure_listen_watch);
> close(reds->secure_listen_socket);
> }
> + g_clear_object(&reds->agent_dev);
> spice_buffer_free(&reds->client_monitors_config);
> red_record_unref(reds->record);
> reds_cleanup(reds);
> @@ -4535,8 +4536,14 @@ red_char_device_vdi_port_finalize(GObject *object)
> {
> RedCharDeviceVDIPort *dev = RED_CHAR_DEVICE_VDIPORT(object);
>
> + /* make sure we have no other references to RedVDIReadBuf buffers */
> + red_char_device_reset(RED_CHAR_DEVICE(dev));
> + if (dev->priv->current_read_buf) {
> + red_pipe_item_unref(&dev->priv->current_read_buf->base);
> + dev->priv->current_read_buf = NULL;
> + }
> + g_list_free_full(dev->priv->read_bufs, g_free);
> g_free(dev->priv->mig_data);
> - /* FIXME: need to free the RedVDIReadBuf allocated previously */
>
> G_OBJECT_CLASS(red_char_device_vdi_port_parent_class)->finalize(object);
> }
> --
> 2.14.3
>
> _______________________________________________
> Spice-devel mailing list
> Spice-devel@xxxxxxxxxxxxxxxxxxxxx
> https://lists.freedesktop.org/mailman/listinfo/spice-devel
Attachment:
signature.asc
Description: PGP signature
_______________________________________________ Spice-devel mailing list Spice-devel@xxxxxxxxxxxxxxxxxxxxx https://lists.freedesktop.org/mailman/listinfo/spice-devel