- 14 Oct, 2013 15 commits
-
-
Marcel Holtmann authored
The scan window parameter for connection establishment and passive scanning needs to be smaller or equal than the scan interval. Instead of waiting for a controller to reject these values later on, just reject them right away. Signed-off-by:
Marcel Holtmann <marcel@holtmann.org> Signed-off-by:
Johan Hedberg <johan.hedberg@intel.com>
-
Johan Hedberg authored
If a socket was bound to an address type other than BR/EDR (such as LE) we should reject trying to connect it to a BR/EDR address. The same applies for binding to BR/EDR and trying to connect to non-BR/EDR. Signed-off-by:
-
Johan Hedberg authored
We need to verify that the bdaddr type passed to connect() and bind() is within the set of valid values. If it is not we need to cleanly fail with EINVAL. Signed-off-by:
-
Johan Hedberg authored
This patch converts Set Discoverable to use an asynchronous request along with its own completion callback. This is necessary for splitting raw HCI socket use cases from mgmt, as well as for enabling the hooking up of Advertising parameters together with the HCI_DISCOVERABLE flag (coming in later patches). Signed-off-by:
-
Johan Hedberg authored
Now that the connectable setting is also applicable for the LE side it's possible that the HCI_CONNECTABLE flag is already set when changing the BR/EDR setting from false to true while the controller is powered. In this situation we need to update the BR/EDR scan mode to reflect the setting. Additionally, since HCI_CONNECTABLE also applies to LE we must not clear the HCI_CONNECTABLE flag when disabling bredr. Signed-off-by:
-
Johan Hedberg authored
The set_bredr_scan() function will soon be needed by the set_bredr() function, so move it to a new location to avoid having to add a forward declaration. Signed-off-by:
-
Johan Hedberg authored
This patch updates the Set Connectable Management command to also update the LE advertising type to either connectable or non-connectable advertising. An extra helper function is needed for getting the right advertising type since we can not only rely on the HCI_CONNECTABLE flag but must also check for a pending Set Connectable command (in which case the flag does not yet have its final value). Signed-off-by:
-
Johan Hedberg authored
We need to ensure that the advertising data is up-to-date whenever advertising is enabled, but when disabling advertising we do not need to worry about it (since it will eventually get fixed as soon as advertising is enabled again). This patch fixes this in the command complete callback for set_adv_enable. Signed-off-by:
-
Johan Hedberg authored
These functions will soon be used by set_connectable() so move them to a location in mgmt.c that doesn't require forward declarations. Signed-off-by:
-
Johan Hedberg authored
If the HCI commands related to the Set Connectable command fail we will get a non-zero status in the request completion callback. In such a case we must respond with the appropriate command status message to user space. Signed-off-by:
-
Johan Hedberg authored
This patch moves the responsibility of setting/clearing the HCI_CONNECTABLE flag to the request completion callback of the Set Connectable command. This will allow us to cleanly add support for LE Advertising hooks in later patches. Signed-off-by:
-
Johan Hedberg authored
This patch moves all the decisions of which HCI commands to send (or not to send) to the code between hci_req_init() and hci_req_run() this allows us to further extend the request with further commands but still keep the same logic of handling whether to return a direct mgmt response in the case that no HCI commands were sent. Signed-off-by:
-
Marcel Holtmann authored
Clearing the BT_SK_SUSPEND socket flag from the L2CAP core is causing a dependency on the socket. So intead of doing that, use a channel callback into the socket handling to resume. Signed-off-by:
-
Marcel Holtmann authored
The L2CAP core should not look into the socket flags to figure out the setting of defer setup. So introduce a L2CAP channel flag that mirrors the socket flag. Since the defer setup option is only set in one place this becomes a really easy thing to do. Signed-off-by:
-
Marcel Holtmann authored
The exposed socket information do not contain source or destination addresses. So adjust the header accordingly. Signed-off-by:
-
- 13 Oct, 2013 25 commits
-
-
Marcel Holtmann authored
There have been a lot of changes in the core Bluetooth handling lately. So it is a good idea to increase the module version. The module version is not used anywhere, but it makes debugging a little bit simpler if versions can be distinguished. Signed-off-by:
-
Marcel Holtmann authored
The L2CAP connectionless channels use SOCK_DGRAM and recvmsg() and need to receive the remote BD_ADDR and PSM information via msg_name from the recvmsg() system call. So in case the L2CAP socket is for connectionless channels, provide a msg_name callback that can update the data. Also store the remote BD_ADDR and PSM in the skb so it can be extracted later on. Signed-off-by:
-
Marcel Holtmann authored
This allows to add a per socket msg_name callback that can be used for updating the msg_name information for recvmsg() system calls. This feature is used by another patch to support address information on L2CAP connectionless channels. Signed-off-by:
-
Marcel Holtmann authored
There are few places where it makes sense to use l2cap_pi(sk) directly instead of assigning it to temporary structure. Signed-off-by:
-
Marcel Holtmann authored
Every socket protocol now stores its own address information. So just remove the generic src and dst fields since they are no longer needed. Signed-off-by:
-
Marcel Holtmann authored
The address information of RFCOMM sockets should be stored in its own socket structure. Trying to generalize them is not helpful since different transports have different address types. Signed-off-by:
-
Marcel Holtmann authored
The address information of SCO sockets should be stored in its own socket structure. Trying to generalize them is not helpful since different transports have different address types. Signed-off-by:
-
Marcel Holtmann authored
Instead of storing a pointer to the addresses for the HCI device and HCI connection, use them directly. With the recent changes to address tracking of HCI connections, this becomes simple. Signed-off-by:
-
Marcel Holtmann authored
The L2CAP socket structure does not contain the address information anymore. They need to be accessed through the L2CAP channel. Signed-off-by:
-
Marcel Holtmann authored
The L2CAP socket structure does not contain the address information anymore. They need to be accessed through the L2CAP channel. Signed-off-by:
-
Marcel Holtmann authored
The L2CAP socket structure does not contain the address information anymore. They need to be accessed through the L2CAP channel. Signed-off-by:
-
Marcel Holtmann authored
The L2CAP socket structure does not contain the address information anymore. They need to be accessed through the L2CAP channel. Signed-off-by:
-
Marcel Holtmann authored
The L2CAP sockets can use BR/EDR public, LE public and LE random addresses for various combinations of source and destination devices. So make sure that getsockname(), getpeername() and accept() return the correct address type. For this the address type of the source and destination is stored with the L2CAP channel information. The stored address type is not the one specific for the HCI protocol. It is the address type used for the L2CAP sockets and the management interface. The underlying HCI connections store the HCI address type. If needed, it gets converted to the socket address type. Signed-off-by:
-
Marcel Holtmann authored
With the effort of abstracting the L2CAP socket from the underlying L2CAP channel it is important to store the source and destination address information directly in the L2CAP channel structure. Direct access to the HCI connection address information is not possible since they might not be avaiable at L2CAP channel creation time. The address information will be updated when the underlying BR/EDR or LE connection status changes. Signed-off-by:
-
Marcel Holtmann authored
When having LE connections, the source address is not always the public address of the controller. So update the socket address based on the actual used source address of the HCI connection. This also remove the pointless source address pointer and adds a proper lock around the socket structure. Signed-off-by:
-
Marcel Holtmann authored
The SMP source code has a few coding style violations. Fix them up all at once. No actual code has changed. Signed-off-by:
-
Marcel Holtmann authored
The smp_c1() so far always assumed public addresses as input for its operation. However it should provide actually the source address type of the actual connection. Finally the source address type is tracked in hci_conn->src_type and so use that one as input. Signed-off-by:
-
Marcel Holtmann authored
The source address is now stored in hci_conn->src and so use that one for L2CAP functions. Signed-off-by:
-
Marcel Holtmann authored
The source address is now stored in hci_conn->src and so use that one for SMP functions. Signed-off-by:
-
Marcel Holtmann authored
The incoming LE connections do not have a proper source address and address type set. The connection needs to be set with the same values as used for advertising parameters. Signed-off-by:
-
Marcel Holtmann authored
The source addressed was based on the public address of the HCI device, but with LE connections this not always the case. For example single mode LE-only controllers would use a static random address. And this address is configured by userspace. To not complicate the lookup of what kind of address is in use, store the correct source address for each HCI connection. Signed-off-by:
-
Marcel Holtmann authored
When establishing LE connections, it is possible to use a public address (if available) or a random address. The type of address is only known when creating connections, so make sure it is stored in hci_conn structure. Signed-off-by:
-
Marcel Holtmann authored
The bdaddr_to_le() function tries to convert the internal address type to one that matches the HCI address type for LE. It does not handle any address types not used by LE and in the end just make the code a lot harder to read. So instead of just hiding behind a magic function, just convert the internal address type where it needs to be converted. And it turns out that these are only two cases anyway. One when creating new LE connections and the other when loading the long term keys. In both cases this makes it more clear on what it going on. Signed-off-by:
-
Marcel Holtmann authored
The l2cap_conn->src and l2cap_conn->dst pointers are no longer in use and so just remove them. Signed-off-by:
-
Marcel Holtmann authored
The l2cap_conn->src and l2cap_conn->dst addresses are just a pointers to hci_conn structure. Use hci_conn->hdev->bdaddr and hci_conn->dst directly instead. Signed-off-by:
-
