Return-Path: From: Mikel Astiz To: linux-bluetooth@vger.kernel.org Cc: Mikel Astiz Subject: [PATCH 2/4] audio: gateway_unlock together with in_use = FALSE Date: Thu, 15 Dec 2011 11:01:05 +0100 Message-Id: <1323943267-14576-3-git-send-email-mikel.astiz.oss@gmail.com> In-Reply-To: <1323943267-14576-1-git-send-email-mikel.astiz.oss@gmail.com> References: <1323943267-14576-1-git-send-email-mikel.astiz.oss@gmail.com> Sender: linux-bluetooth-owner@vger.kernel.org List-ID: From: Mikel Astiz Calling gateway_unlock seems safer in gateway_suspend_complete, in combination with the update of transport->in_use. This avoids the transitional state of having the gateway unlocked but the transport still in use. This approach is also more consistent with the way headset and a2dp handle it. --- audio/transport.c | 3 ++- 1 files changed, 2 insertions(+), 1 deletions(-) diff --git a/audio/transport.c b/audio/transport.c index 6029528..fc7026e 100644 --- a/audio/transport.c +++ b/audio/transport.c @@ -510,6 +510,7 @@ static gboolean gateway_suspend_complete(gpointer user_data) { struct media_owner *owner = user_data; struct media_transport *transport = owner->transport; + struct audio_device *device = transport->device; /* Release always succeeds */ if (owner->pending) { @@ -518,6 +519,7 @@ static gboolean gateway_suspend_complete(gpointer user_data) media_owner_remove(owner); } + gateway_unlock(device, GATEWAY_LOCK_READ | GATEWAY_LOCK_WRITE); transport->in_use = FALSE; media_transport_remove(transport, owner); return FALSE; @@ -535,7 +537,6 @@ static guint suspend_gateway(struct media_transport *transport, return 0; } - gateway_unlock(device, GATEWAY_LOCK_READ | GATEWAY_LOCK_WRITE); g_idle_add(gateway_suspend_complete, owner); return id++; } -- 1.7.6.4