Hi Peter,
On Fri, Aug 12, 2011 at 3:56 PM, Peter Hurley <peter@xxxxxxxxxxxxxxxxxx> wrote:
> AVDTP_CONNECT_TIMEOUT controls the delay between HSP/HFP
> connection establishment and AVDTP signal channel establishment.
> The original value of 1 sec. is too short to avoid racing for AVDTP
> connection establishment (eg., if the device continues to configure
> the HFP service level connection with add'l AT cmds).
>
> Some devices have broken AVDTP implementations that just cannot
> handle the race conditions that arise if both devices are attempting
> stream establishment at the same time. However, these conditions arise
> only when the remote device is the ACL initiator (and in practice, the
> RFCOMM initiator as well). This fix bumps out the timeout value only
> when the headset has initiated the link.
> ---
> audio/device.c | 9 ++++++++-
> audio/headset.c | 16 ++++++++++++++++
> audio/headset.h | 3 +++
> audio/manager.c | 1 +
> 4 files changed, 28 insertions(+), 1 deletions(-)
>
> diff --git a/audio/device.c b/audio/device.c
> index 18e873e..0bd1443 100644
> --- a/audio/device.c
> +++ b/audio/device.c
> @@ -61,6 +61,7 @@
>
> #define CONTROL_CONNECT_TIMEOUT 2
> #define AVDTP_CONNECT_TIMEOUT 1
> +#define AVDTP_CONNECT_TIMEOUT_BOOST 1
> #define HEADSET_CONNECT_TIMEOUT 1
>
> typedef enum {
> @@ -305,6 +306,7 @@ static gboolean avdtp_connect_timeout(gpointer user_data)
> static gboolean device_set_avdtp_timer(struct audio_device *dev)
> {
> struct dev_priv *priv = dev->priv;
> + guint timeout = AVDTP_CONNECT_TIMEOUT;
>
> if (!dev->sink)
> return FALSE;
> @@ -312,7 +314,12 @@ static gboolean device_set_avdtp_timer(struct audio_device *dev)
> if (priv->avdtp_timer)
> return FALSE;
>
> - priv->avdtp_timer = g_timeout_add_seconds(AVDTP_CONNECT_TIMEOUT,
> + /* if the headset is the HSP/HFP RFCOMM initiator, give the headset
> + time to initiate AVDTP signalling (and avoid further racing) */
> + if (dev->headset && get_rfcomm_initiator(dev))
> + timeout += AVDTP_CONNECT_TIMEOUT_BOOST;
> +
> + priv->avdtp_timer = g_timeout_add_seconds(timeout,
> avdtp_connect_timeout,
> dev);
>
> diff --git a/audio/headset.c b/audio/headset.c
> index 8e63afc..66a28f2 100644
> --- a/audio/headset.c
> +++ b/audio/headset.c
> @@ -167,6 +167,7 @@ struct headset {
>
> gboolean hfp_active;
> gboolean search_hfp;
> + gboolean rfcomm_initiator;
>
> headset_state_t state;
> struct pending_connect *pending;
> @@ -1633,6 +1634,7 @@ static int rfcomm_connect(struct audio_device *dev, headset_stream_cb_t cb,
> }
>
> hs->hfp_active = hs->hfp_handle != 0 ? TRUE : FALSE;
> + hs->rfcomm_initiator = FALSE;
>
> headset_set_state(dev, HEADSET_STATE_CONNECTING);
>
> @@ -2442,6 +2444,20 @@ void set_hfp_active(struct audio_device *dev, gboolean active)
> hs->hfp_active = active;
> }
>
> +gboolean get_rfcomm_initiator(struct audio_device *dev)
> +{
> + struct headset *hs = dev->headset;
> +
> + return hs->rfcomm_initiator;
> +}
> +
> +void set_rfcomm_initiator(struct audio_device *dev, gboolean initiator)
> +{
> + struct headset *hs = dev->headset;
> +
> + hs->rfcomm_initiator = initiator;
> +}
> +
I would suggest you to follow the name convention e.g.
headset_get_rfcomm_initiator...
> GIOChannel *headset_get_rfcomm(struct audio_device *dev)
> {
> struct headset *hs = dev->headset;
> diff --git a/audio/headset.h b/audio/headset.h
> index 7ce88c8..f275e03 100644
> --- a/audio/headset.h
> +++ b/audio/headset.h
> @@ -82,6 +82,9 @@ gboolean headset_cancel_stream(struct audio_device *dev, unsigned int id);
> gboolean get_hfp_active(struct audio_device *dev);
> void set_hfp_active(struct audio_device *dev, gboolean active);
>
> +gboolean get_rfcomm_initiator(struct audio_device *dev);
> +void set_rfcomm_initiator(struct audio_device *dev, gboolean initiator);
> +
Here too.
--
Luiz Augusto von Dentz
--
To unsubscribe from this list: send the line "unsubscribe linux-bluetooth" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
