Return-Path: <cyrus@holtmann.org>
Content-Type: text/plain; charset=US-ASCII
Mime-Version: 1.0 (Mac OS X Mail 6.6 \(1510\))
Subject: Re: [PATCH 4/7] Bluetooth: Remove hci_cs_le_create_conn event handler
From: Marcel Holtmann <marcel@holtmann.org>
In-Reply-To: <CACJA=fXGisEzc86N1YKL1Y84wVs5SMRaedEGuMhWJg9A-FV+xQ@mail.gmail.com>
Date: Thu, 3 Oct 2013 16:13:03 +0200
Cc: "linux-bluetooth@vger.kernel.org" <linux-bluetooth@vger.kernel.org>
Message-Id: <B392847A-6838-4999-A1D2-571B08F7115A@holtmann.org>
References: <1380668636-30654-1-git-send-email-andre.guedes@openbossa.org> <1380668636-30654-5-git-send-email-andre.guedes@openbossa.org> <3E999DA3-EE5E-41B3-8903-16C4A29FEB95@holtmann.org> <CACJA=fXGisEzc86N1YKL1Y84wVs5SMRaedEGuMhWJg9A-FV+xQ@mail.gmail.com>
To: Andre Guedes <andre.guedes@openbossa.org>
Sender: linux-bluetooth-owner@vger.kernel.org
List-ID: <linux-bluetooth.vger.kernel.org>

Hi Andre,

>>> This patch removes the hci_cs_le_create_conn event handler since this
>>> handling is now done in create_le_connection_complete() callback in
>>> hci_conn.c.
>>> 
>>> Signed-off-by: Andre Guedes <andre.guedes@openbossa.org>
>>> ---
>>> net/bluetooth/hci_event.c | 31 -------------------------------
>>> 1 file changed, 31 deletions(-)
>>> 
>>> diff --git a/net/bluetooth/hci_event.c b/net/bluetooth/hci_event.c
>>> index d171c04b..1d1ffa6 100644
>>> --- a/net/bluetooth/hci_event.c
>>> +++ b/net/bluetooth/hci_event.c
>>> @@ -1465,33 +1465,6 @@ static void hci_cs_disconnect(struct hci_dev *hdev, u8 status)
>>>      hci_dev_unlock(hdev);
>>> }
>>> 
>>> -static void hci_cs_le_create_conn(struct hci_dev *hdev, __u8 status)
>>> -{
>>> -     struct hci_conn *conn;
>>> -
>>> -     BT_DBG("%s status 0x%2.2x", hdev->name, status);
>>> -
>>> -     if (status) {
>>> -             hci_dev_lock(hdev);
>>> -
>>> -             conn = hci_conn_hash_lookup_state(hdev, LE_LINK, BT_CONNECT);
>>> -             if (!conn) {
>>> -                     hci_dev_unlock(hdev);
>>> -                     return;
>>> -             }
>>> -
>>> -             BT_DBG("%s bdaddr %pMR conn %p", hdev->name, &conn->dst, conn);
>>> -
>>> -             conn->state = BT_CLOSED;
>>> -             mgmt_connect_failed(hdev, &conn->dst, conn->type,
>>> -                                 conn->dst_type, status);
>>> -             hci_proto_connect_cfm(conn, status);
>>> -             hci_conn_del(conn);
>>> -
>>> -             hci_dev_unlock(hdev);
>>> -     }
>>> -}
>> 
>> this is dangerous since it actually breaks bisection. The code is never complete. So while this might turn into a larger patch, you might need to do it all 3 patches at once. With a length commit message explaining exactly what happens and why this is correct.
> 
> I failed to see how this breaks bisection since the handling is
> already done in initiate_le_connection_complete(). However, as
> commented in patch 2/7, I'll squash this into patch 2/7 as you
> suggested.

once you actually run the code it will break. Which means that bisecting is not possible anymore since now you are chasing the bug of duplicated connection handling.

Regards

Marcel