Return-Path: MIME-Version: 1.0 In-Reply-To: <20120830233451.GA972@x220.sheraton.com> References: <1346364960-7699-1-git-send-email-jprvita@openbossa.org> <1346364960-7699-2-git-send-email-jprvita@openbossa.org> <20120830233451.GA972@x220.sheraton.com> Date: Mon, 3 Sep 2012 15:16:38 -0300 Message-ID: Subject: Re: [PATCH BlueZ v4 01/15] core: Fix missing g_io_channel_ref From: Joao Paulo Rechi Vita To: linux-bluetooth@vger.kernel.org Content-Type: text/plain; charset=UTF-8 Sender: linux-bluetooth-owner@vger.kernel.org List-ID: On Thu, Aug 30, 2012 at 8:34 PM, Johan Hedberg wrote: > Hi João Paulo, > > On Thu, Aug 30, 2012, João Paulo Rechi Vita wrote: >> --- >> src/device.c | 2 +- >> 1 file changed, 1 insertion(+), 1 deletion(-) >> >> diff --git a/src/device.c b/src/device.c >> index 3b44d9b..2c40ec2 100644 >> --- a/src/device.c >> +++ b/src/device.c >> @@ -2027,7 +2027,7 @@ static gboolean att_connect(gpointer user_data) >> return FALSE; >> } >> >> - device->att_io = io; >> + device->att_io = g_io_channel_ref(io); >> >> return FALSE; >> } > > This doesn't look right to me. bt_io_connect returns a reference for the > caller and you shouldn't need to re-increment the ref count again unless > you store a pointer in multiple places (which you don't). If this patch > fixes some behavior the real bug must be somewhere else. > Yes, this became wrong due to spliting this code out of "core: Mutually exclude concurrent connections". I'll fix that and re-send the series one more time. -- João Paulo Rechi Vita Openbossa Labs - INdT