From: David Herrmann <dh.herrmann@gmail.com>
To: linux-bluetooth@vger.kernel.org
Cc: Gustavo Padovan <gustavo@padovan.org>,
David Herrmann <dh.herrmann@gmail.com>
Subject: [PATCH v4 09/16] Bluetooth: l2cap: add l2cap_user sub-modules
Date: Sat, 6 Apr 2013 20:28:45 +0200 [thread overview]
Message-ID: <1365272932-571-10-git-send-email-dh.herrmann@gmail.com> (raw)
In-Reply-To: <1365272932-571-1-git-send-email-dh.herrmann@gmail.com>
Several sub-modules like HIDP, rfcomm, ... need to track l2cap
connections. The l2cap_conn->hcon->dev object is used as parent for sysfs
devices so the sub-modules need to be notified when the hci_conn object is
removed from sysfs.
As submodules normally use the l2cap layer, the l2cap_user objects are
registered there instead of on the underlying hci_conn object. This avoids
any direct dependency on the HCI layer and lets the l2cap core handle any
specifics.
This patch introduces l2cap_user objects which contain a "probe" and
"remove" callback. You can register them on any l2cap_conn object and if
it is active, the "probe" callback will get called. Otherwise, an error is
returned.
The l2cap_conn object will call your "remove" callback directly before it
is removed from user-space. This allows you to remove your submodules
_before_ the parent l2cap_conn and hci_conn object is removed.
At any time you can asynchronously unregister your l2cap_user object if
your submodule vanishes before the l2cap_conn object does.
There is no way around l2cap_user. If we want wire-protocols in the
kernel, we always want the hci_conn object as parent in the sysfs tree. We
cannot use a channel here since we might need multiple channels for a
single protocol.
But the problem is, we _must_ get notified when an l2cap_conn object is
removed. We cannot use reference-counting for object-removal! This is not
how it works. If a hardware is removed, we should immediately remove the
object from sysfs. Any other behavior would be inconsistent with the rest
of the system. Also note that device_del() might sleep, but it doesn't
wait for user-space or block very long. It only _unlinks_ the object from
sysfs and the whole device-tree. Everything else is handled by ref-counts!
This is exactly what the other sub-modules must do: unlink their devices
when the "remove" l2cap_user callback is called. They should not do any
cleanup or synchronous shutdowns.
Signed-off-by: David Herrmann <dh.herrmann@gmail.com>
Acked-by: Marcel Holtmann <marcel@holtmann.org>
---
include/net/bluetooth/l2cap.h | 10 +++++
net/bluetooth/l2cap_core.c | 86 +++++++++++++++++++++++++++++++++++++++++++
2 files changed, 96 insertions(+)
diff --git a/include/net/bluetooth/l2cap.h b/include/net/bluetooth/l2cap.h
index 7b4cc5b..fb94cf1 100644
--- a/include/net/bluetooth/l2cap.h
+++ b/include/net/bluetooth/l2cap.h
@@ -584,6 +584,13 @@ struct l2cap_conn {
struct list_head chan_l;
struct mutex chan_lock;
struct kref ref;
+ struct list_head users;
+};
+
+struct l2cap_user {
+ struct list_head list;
+ int (*probe) (struct l2cap_conn *conn, struct l2cap_user *user);
+ void (*remove) (struct l2cap_conn *conn, struct l2cap_user *user);
};
#define L2CAP_INFO_CL_MTU_REQ_SENT 0x01
@@ -817,4 +824,7 @@ void __l2cap_physical_cfm(struct l2cap_chan *chan, int result);
void l2cap_conn_get(struct l2cap_conn *conn);
void l2cap_conn_put(struct l2cap_conn *conn);
+int l2cap_register_user(struct l2cap_conn *conn, struct l2cap_user *user);
+void l2cap_unregister_user(struct l2cap_conn *conn, struct l2cap_user *user);
+
#endif /* __L2CAP_H */
diff --git a/net/bluetooth/l2cap_core.c b/net/bluetooth/l2cap_core.c
index 5f273a3..e7c06d1 100644
--- a/net/bluetooth/l2cap_core.c
+++ b/net/bluetooth/l2cap_core.c
@@ -1446,6 +1446,89 @@ static void l2cap_info_timeout(struct work_struct *work)
l2cap_conn_start(conn);
}
+/*
+ * l2cap_user
+ * External modules can register l2cap_user objects on l2cap_conn. The ->probe
+ * callback is called during registration. The ->remove callback is called
+ * during unregistration.
+ * An l2cap_user object can either be explicitly unregistered or when the
+ * underlying l2cap_conn object is deleted. This guarantees that l2cap->hcon,
+ * l2cap->hchan, .. are valid as long as the remove callback hasn't been called.
+ * External modules must own a reference to the l2cap_conn object if they intend
+ * to call l2cap_unregister_user(). The l2cap_conn object might get destroyed at
+ * any time if they don't.
+ */
+
+int l2cap_register_user(struct l2cap_conn *conn, struct l2cap_user *user)
+{
+ struct hci_dev *hdev = conn->hcon->hdev;
+ int ret;
+
+ /* We need to check whether l2cap_conn is registered. If it is not, we
+ * must not register the l2cap_user. l2cap_conn_del() is unregisters
+ * l2cap_conn objects, but doesn't provide its own locking. Instead, it
+ * relies on the parent hci_conn object to be locked. This itself relies
+ * on the hci_dev object to be locked. So we must lock the hci device
+ * here, too. */
+
+ hci_dev_lock(hdev);
+
+ if (user->list.next || user->list.prev) {
+ ret = -EINVAL;
+ goto out_unlock;
+ }
+
+ /* conn->hchan is NULL after l2cap_conn_del() was called */
+ if (!conn->hchan) {
+ ret = -ENODEV;
+ goto out_unlock;
+ }
+
+ ret = user->probe(conn, user);
+ if (ret)
+ goto out_unlock;
+
+ list_add(&user->list, &conn->users);
+ ret = 0;
+
+out_unlock:
+ hci_dev_unlock(hdev);
+ return ret;
+}
+EXPORT_SYMBOL(l2cap_register_user);
+
+void l2cap_unregister_user(struct l2cap_conn *conn, struct l2cap_user *user)
+{
+ struct hci_dev *hdev = conn->hcon->hdev;
+
+ hci_dev_lock(hdev);
+
+ if (!user->list.next || !user->list.prev)
+ goto out_unlock;
+
+ list_del(&user->list);
+ user->list.next = NULL;
+ user->list.prev = NULL;
+ user->remove(conn, user);
+
+out_unlock:
+ hci_dev_unlock(hdev);
+}
+EXPORT_SYMBOL(l2cap_unregister_user);
+
+static void l2cap_unregister_all_users(struct l2cap_conn *conn)
+{
+ struct l2cap_user *user;
+
+ while (!list_empty(&conn->users)) {
+ user = list_first_entry(&conn->users, struct l2cap_user, list);
+ list_del(&user->list);
+ user->list.next = NULL;
+ user->list.prev = NULL;
+ user->remove(conn, user);
+ }
+}
+
static void l2cap_conn_del(struct hci_conn *hcon, int err)
{
struct l2cap_conn *conn = hcon->l2cap_data;
@@ -1458,6 +1541,8 @@ static void l2cap_conn_del(struct hci_conn *hcon, int err)
kfree_skb(conn->rx_skb);
+ l2cap_unregister_all_users(conn);
+
mutex_lock(&conn->chan_lock);
/* Kill channels */
@@ -1550,6 +1635,7 @@ static struct l2cap_conn *l2cap_conn_add(struct hci_conn *hcon, u8 status)
mutex_init(&conn->chan_lock);
INIT_LIST_HEAD(&conn->chan_l);
+ INIT_LIST_HEAD(&conn->users);
if (hcon->type == LE_LINK)
INIT_DELAYED_WORK(&conn->security_timer, security_timeout);
--
1.8.2
next prev parent reply other threads:[~2013-04-06 18:28 UTC|newest]
Thread overview: 44+ messages / expand[flat|nested] mbox.gz Atom feed top
2013-04-05 12:57 [PATCH v3 00/18] Rework HIDP Session Management David Herrmann
2013-04-05 12:57 ` [PATCH v3 01/18] Bluetooth: hidp: remove redundant error message David Herrmann
2013-04-06 2:41 ` Gustavo Padovan
2013-04-05 12:57 ` [PATCH v3 02/18] Bluetooth: hidp: verify l2cap sockets David Herrmann
2013-04-06 2:44 ` Gustavo Padovan
2013-04-05 12:57 ` [PATCH v3 03/18] Bluetooth: rename hci_conn_put to hci_conn_drop David Herrmann
2013-04-06 2:48 ` Gustavo Padovan
2013-04-06 18:31 ` David Herrmann
2013-04-05 12:57 ` [PATCH v3 04/18] Bluetooth: remove unneeded hci_conn_hold/put_device() David Herrmann
2013-04-05 12:57 ` [PATCH v3 05/18] Bluetooth: introduce hci_conn ref-counting David Herrmann
2013-04-05 12:57 ` [PATCH v3 06/18] Bluetooth: hidp: remove unused session->state field David Herrmann
2013-04-05 12:57 ` [PATCH v3 07/18] Bluetooth: hidp: test "terminate" before sleeping David Herrmann
2013-04-05 12:57 ` [PATCH v3 08/18] Bluetooth: allow constant arguments for bacmp()/bacpy() David Herrmann
2013-04-05 12:57 ` [PATCH v3 09/18] Bluetooth: hidp: move hidp_schedule() to core.c David Herrmann
2013-04-05 12:57 ` [PATCH v3 10/18] Bluetooth: l2cap: introduce l2cap_conn ref-counting David Herrmann
2013-04-05 12:57 ` [PATCH v3 11/18] Bluetooth: l2cap: add l2cap_user sub-modules David Herrmann
2013-04-05 12:57 ` [PATCH v3 12/18] Bluetooth: hidp: add new session-management helpers David Herrmann
2013-04-05 12:57 ` [PATCH v3 13/18] Bluetooth: hidp: remove old session-management David Herrmann
2013-04-05 12:57 ` [PATCH v3 14/18] Bluetooth: hidp: handle kernel_sendmsg() errors correctly David Herrmann
2013-04-05 12:57 ` [PATCH v3 15/18] Bluetooth: hidp: merge hidp_process_{ctrl,intr}_transmit() David Herrmann
2013-04-05 12:57 ` [PATCH v3 16/18] Bluetooth: hidp: merge 'send' functions into hidp_send_message() David Herrmann
2013-04-05 12:57 ` [PATCH v3 17/18] Bluetooth: hidp: don't send boot-protocol messages as HID-reports David Herrmann
2013-04-05 12:57 ` [PATCH v3 18/18] Bluetooth: hidp: fix sending output reports on intr channel David Herrmann
2013-04-18 2:49 ` Gustavo Padovan
2013-04-05 19:01 ` [PATCH v3 00/18] Rework HIDP Session Management Marcel Holtmann
2013-04-06 18:28 ` [PATCH v4 00/16] " David Herrmann
2013-04-06 18:28 ` [PATCH v4 01/16] Bluetooth: rename hci_conn_put to hci_conn_drop David Herrmann
2013-04-11 19:45 ` Gustavo Padovan
2013-04-06 18:28 ` [PATCH v4 02/16] Bluetooth: remove unneeded hci_conn_hold/put_device() David Herrmann
2013-04-17 5:39 ` Gustavo Padovan
2013-04-06 18:28 ` [PATCH v4 03/16] Bluetooth: introduce hci_conn ref-counting David Herrmann
2013-04-06 18:28 ` [PATCH v4 04/16] Bluetooth: hidp: remove unused session->state field David Herrmann
2013-04-06 18:28 ` [PATCH v4 05/16] Bluetooth: hidp: test "terminate" before sleeping David Herrmann
2013-04-06 18:28 ` [PATCH v4 06/16] Bluetooth: allow constant arguments for bacmp()/bacpy() David Herrmann
2013-04-06 18:28 ` [PATCH v4 07/16] Bluetooth: hidp: move hidp_schedule() to core.c David Herrmann
2013-04-06 18:28 ` [PATCH v4 08/16] Bluetooth: l2cap: introduce l2cap_conn ref-counting David Herrmann
2013-04-06 18:28 ` David Herrmann [this message]
2013-04-06 18:28 ` [PATCH v4 10/16] Bluetooth: hidp: add new session-management helpers David Herrmann
2013-04-06 18:28 ` [PATCH v4 11/16] Bluetooth: hidp: remove old session-management David Herrmann
2013-04-06 18:28 ` [PATCH v4 12/16] Bluetooth: hidp: handle kernel_sendmsg() errors correctly David Herrmann
2013-04-06 18:28 ` [PATCH v4 13/16] Bluetooth: hidp: merge hidp_process_{ctrl,intr}_transmit() David Herrmann
2013-04-06 18:28 ` [PATCH v4 14/16] Bluetooth: hidp: merge 'send' functions into hidp_send_message() David Herrmann
2013-04-06 18:28 ` [PATCH v4 15/16] Bluetooth: hidp: don't send boot-protocol messages as HID-reports David Herrmann
2013-04-06 18:28 ` [PATCH v4 16/16] Bluetooth: hidp: fix sending output reports on intr channel David Herrmann
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1365272932-571-10-git-send-email-dh.herrmann@gmail.com \
--to=dh.herrmann@gmail.com \
--cc=gustavo@padovan.org \
--cc=linux-bluetooth@vger.kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.