[v2,1/3] Bluetooth: hci_conn: Fix hci_connect_le_sync

Commit Message

Luiz Augusto von Dentz May 19, 2022, 8:22 p.m. UTC

From: Luiz Augusto von Dentz <[e-mail]>

The handling of connection failures shall be handled by the request
completion callback as already done by hci_cs_le_create_conn, also make
sure to use hci_conn_failed instead of hci_le_conn_failed as the later
don't actually call hci_conn_del to cleanup.

Fixes: 8e8b92ee60de5 ("Bluetooth: hci_sync: Add hci_le_create_conn_sync")
Signed-off-by: Luiz Augusto von Dentz <[e-mail]>
---
 net/bluetooth/hci_conn.c  | 5 +++--
 net/bluetooth/hci_event.c | 8 +++++---
 2 files changed, 8 insertions(+), 5 deletions(-)

Comments

[e-mail] May 19, 2022, 9:07 p.m. UTC | #1

This is automated email and please do not reply to this email!

Dear submitter,

Thank you for submitting the patches to the linux bluetooth mailing list.
This is a CI test results with your patch series:
PW Link:https://patchwork.kernel.org/project/bluetooth/list/?series=643316

---Test result---

Test Summary:
CheckPatch                    PASS      2.73 seconds
GitLint                       PASS      1.14 seconds
SubjectPrefix                 PASS      0.72 seconds
BuildKernel                   PASS      42.02 seconds
BuildKernel32                 PASS      37.35 seconds
Incremental Build with patchesPASS      91.80 seconds
TestRunner: Setup             PASS      624.66 seconds
TestRunner: l2cap-tester      PASS      20.89 seconds
TestRunner: bnep-tester       PASS      7.58 seconds
TestRunner: mgmt-tester       PASS      128.15 seconds
TestRunner: rfcomm-tester     PASS      11.71 seconds
TestRunner: sco-tester        PASS      11.35 seconds
TestRunner: smp-tester        PASS      11.41 seconds
TestRunner: userchan-tester   PASS      7.90 seconds



---
Regards,
Linux Bluetooth

Marcel Holtmann May 20, 2022, 9:18 a.m. UTC | #2

Hi Luiz,

> The handling of connection failures shall be handled by the request
> completion callback as already done by hci_cs_le_create_conn, also make
> sure to use hci_conn_failed instead of hci_le_conn_failed as the later
> don't actually call hci_conn_del to cleanup.
> 
> Fixes: 8e8b92ee60de5 ("Bluetooth: hci_sync: Add hci_le_create_conn_sync")
> Signed-off-by: Luiz Augusto von Dentz <[e-mail]>
> ---
> net/bluetooth/hci_conn.c  | 5 +++--
> net/bluetooth/hci_event.c | 8 +++++---
> 2 files changed, 8 insertions(+), 5 deletions(-)

you are making a bit mess with yet another v2 of this. I rather have you mark this as [PATCH -stable] or something so that we get this one into that tree and then later the rest into -next.

Regards

Marcel

Luiz Augusto von Dentz May 20, 2022, 6:39 p.m. UTC | #3

Hi Marcel,

On Fri, May 20, 2022 at 2:18 AM Marcel Holtmann <[e-mail]> wrote:
>
> Hi Luiz,
>
> > The handling of connection failures shall be handled by the request
> > completion callback as already done by hci_cs_le_create_conn, also make
> > sure to use hci_conn_failed instead of hci_le_conn_failed as the later
> > don't actually call hci_conn_del to cleanup.
> >
> > Fixes: 8e8b92ee60de5 ("Bluetooth: hci_sync: Add hci_le_create_conn_sync")
> > Signed-off-by: Luiz Augusto von Dentz <[e-mail]>
> > ---
> > net/bluetooth/hci_conn.c  | 5 +++--
> > net/bluetooth/hci_event.c | 8 +++++---
> > 2 files changed, 8 insertions(+), 5 deletions(-)
>
> you are making a bit mess with yet another v2 of this. I rather have you mark this as [PATCH -stable] or something so that we get this one into that tree and then later the rest into -next.
>
> Regards
>
> Marcel

Just sent one with -stable, is that something we should adapt going
forward? Perhaps we should document this somewhere.

Patch

diff --git a/net/bluetooth/hci_conn.c b/net/bluetooth/hci_conn.c
index 882a7df13005..ac06c9724c7f 100644
--- a/net/bluetooth/hci_conn.c
+++ b/net/bluetooth/hci_conn.c
@@ -943,10 +943,11 @@  static void create_le_conn_complete(struct hci_dev *hdev, void *data, int err)
 
 	bt_dev_err(hdev, "request failed to create LE connection: err %d", err);
 
-	if (!conn)
+	/* Check if connection is still pending */
+	if (conn != hci_lookup_le_connect(hdev))
 		goto done;
 
-	hci_le_conn_failed(conn, err);
+	hci_conn_failed(conn, err);
 
 done:
 	hci_dev_unlock(hdev);
diff --git a/net/bluetooth/hci_event.c b/net/bluetooth/hci_event.c
index 0270e597c285..af17dfb20e01 100644
--- a/net/bluetooth/hci_event.c
+++ b/net/bluetooth/hci_event.c
@@ -5632,10 +5632,12 @@  static void le_conn_complete_evt(struct hci_dev *hdev, u8 status,
 		status = HCI_ERROR_INVALID_PARAMETERS;
 	}
 
-	if (status) {
-		hci_conn_failed(conn, status);
+	/* All connection failure handling is taken care of by the
+	 * hci_conn_failed function which is triggered by the HCI
+	 * request completion callbacks used for connecting.
+	 */
+	if (status)
 		goto unlock;
-	}
 
 	if (conn->dst_type == ADDR_LE_DEV_PUBLIC)
 		addr_type = BDADDR_LE_PUBLIC;