summaryrefslogtreecommitdiff
path: root/net/bluetooth/hci_request.c
AgeCommit message (Collapse)Author
2017-06-16networking: make skb_put & friends return void pointersJohannes Berg
It seems like a historic accident that these return unsigned char *, and in many places that means casts are required, more often than not. Make these functions (skb_put, __skb_put and pskb_put) return void * and remove all the casts across the tree, adding a (u8 *) cast only where the unsigned char pointer was used directly, all done with the following spatch: @@ expression SKB, LEN; typedef u8; identifier fn = { skb_put, __skb_put }; @@ - *(fn(SKB, LEN)) + *(u8 *)fn(SKB, LEN) @@ expression E, SKB, LEN; identifier fn = { skb_put, __skb_put }; type T; @@ - E = ((T *)(fn(SKB, LEN))) + E = fn(SKB, LEN) which actually doesn't cover pskb_put since there are only three users overall. A handful of stragglers were converted manually, notably a macro in drivers/isdn/i4l/isdn_bsdcomp.c and, oddly enough, one of the many instances in net/bluetooth/hci_sock.c. In the former file, I also had to fix one whitespace problem spatch introduced. Signed-off-by: Johannes Berg <johannes.berg@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-06-16networking: introduce and use skb_put_data()Johannes Berg
A common pattern with skb_put() is to just want to memcpy() some data into the new space, introduce skb_put_data() for this. An spatch similar to the one for skb_put_zero() converts many of the places using it: @@ identifier p, p2; expression len, skb, data; type t, t2; @@ ( -p = skb_put(skb, len); +p = skb_put_data(skb, data, len); | -p = (t)skb_put(skb, len); +p = skb_put_data(skb, data, len); ) ( p2 = (t2)p; -memcpy(p2, data, len); | -memcpy(p, data, len); ) @@ type t, t2; identifier p, p2; expression skb, data; @@ t *p; ... ( -p = skb_put(skb, sizeof(t)); +p = skb_put_data(skb, data, sizeof(t)); | -p = (t *)skb_put(skb, sizeof(t)); +p = skb_put_data(skb, data, sizeof(t)); ) ( p2 = (t2)p; -memcpy(p2, data, sizeof(*p)); | -memcpy(p, data, sizeof(*p)); ) @@ expression skb, len, data; @@ -memcpy(skb_put(skb, len), data, len); +skb_put_data(skb, data, len); (again, manually post-processed to retain some comments) Reviewed-by: Stephen Hemminger <stephen@networkplumber.org> Signed-off-by: Johannes Berg <johannes.berg@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-03-02sched/headers: Prepare to move signal wakeup & sigpending methods from ↵Ingo Molnar
<linux/sched.h> into <linux/sched/signal.h> Fix up affected files that include this signal functionality via sched.h. Acked-by: Linus Torvalds <torvalds@linux-foundation.org> Cc: Mike Galbraith <efault@gmx.de> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: linux-kernel@vger.kernel.org Signed-off-by: Ingo Molnar <mingo@kernel.org>
2016-10-19Bluetooth: Fix append max 11 bytes of name to scan rsp dataMichał Narajowski
Append maximum of 10 + 1 bytes of name to scan response data. Complete name is appended only if exists and is <= 10 characters. Else append short name if exists or shorten complete name if not. This makes sure name is consistent across multiple advertising instances. Signed-off-by: Michał Narajowski <michal.narajowski@codecoup.pl> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2016-10-06Bluetooth: Refactor append name and appearanceMichał Narajowski
Use eir_append_data to remove code duplication. Signed-off-by: Michał Narajowski <michal.narajowski@codecoup.pl> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2016-10-06Bluetooth: Add appearance to default scan rsp dataMichał Narajowski
Add appearance value to beginning of scan rsp data for default advertising instance if the value is not 0. Signed-off-by: Michał Narajowski <michal.narajowski@codecoup.pl> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2016-10-06Bluetooth: Fix local name in scan rspMichał Narajowski
Use complete name if it fits. If not and there is short name check if it fits. If not then use shortened name as prefix of complete name. Signed-off-by: Michał Narajowski <michal.narajowski@codecoup.pl> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2016-09-19Bluetooth: Add support for appearance in scan rspMichał Narajowski
This patch enables prepending appearance value to scan response data. It also adds support for setting appearance value through mgmt command. If currently advertised instance has apperance flag set it is expired immediately. Signed-off-by: Michał Narajowski <michal.narajowski@codecoup.pl> Signed-off-by: Szymon Janc <szymon.janc@codecoup.pl> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2016-09-19Bluetooth: Add support for local name in scan rspMichał Narajowski
This patch enables appending local name to scan response data. If currently advertised instance has name flag set it is expired immediately. Signed-off-by: Michał Narajowski <michal.narajowski@codecoup.pl> Signed-off-by: Szymon Janc <szymon.janc@codecoup.pl> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2016-09-19Bluetooth: Fix reason code used for rejecting SCO connectionsFrédéric Dalleau
A comment in the code states that SCO connection should be rejected with the proper error value between 0xd-0xf. The code uses HCI_ERROR_REMOTE_LOW_RESOURCES which is 0x14. This led to following error: < HCI Command: Reject Synchronous Co.. (0x01|0x002a) plen 7 Address: 34:51:C9:EF:02:CA (Apple, Inc.) Reason: Remote Device Terminated due to Low Resources (0x14) > HCI Event: Command Status (0x0f) plen 4 Reject Synchronous Connection Request (0x01|0x002a) ncmd 1 Status: Invalid HCI Command Parameters (0x12) Instead make use of HCI_ERROR_REJ_LIMITED_RESOURCES which is 0xd. Signed-off-by: Frédéric Dalleau <frederic.dalleau@collabora.co.uk> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2016-09-19Bluetooth: mgmt: Fix sending redundant event for Advertising InstanceJohan Hedberg
When an Advertising Instance is removed, the Advertising Removed event shouldn't be sent to the same socket that issued the Remove Advertising command (it gets a command complete event instead). The mgmt_advertising_removed() function already has a parameter for skipping a specific socket, but there was no code to propagate the right value to this parameter. This patch fixes the issue by making sure the intermediate hci_req_clear_adv_instance() function gets the socket pointer. Signed-off-by: Johan Hedberg <johan.hedberg@intel.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2016-08-24Bluetooth: Fix memory leak at end of hci requestsFrederic Dalleau
In hci_req_sync_complete the event skb is referenced in hdev->req_skb. It is used (via hci_req_run_skb) from either __hci_cmd_sync_ev which will pass the skb to the caller, or __hci_req_sync which leaks. unreferenced object 0xffff880005339a00 (size 256): comm "kworker/u3:1", pid 1011, jiffies 4294671976 (age 107.389s) backtrace: [<ffffffff818d89d9>] kmemleak_alloc+0x49/0xa0 [<ffffffff8116bba8>] kmem_cache_alloc+0x128/0x180 [<ffffffff8167c1df>] skb_clone+0x4f/0xa0 [<ffffffff817aa351>] hci_event_packet+0xc1/0x3290 [<ffffffff8179a57b>] hci_rx_work+0x18b/0x360 [<ffffffff810692ea>] process_one_work+0x14a/0x440 [<ffffffff81069623>] worker_thread+0x43/0x4d0 [<ffffffff8106ead4>] kthread+0xc4/0xe0 [<ffffffff818dd38f>] ret_from_fork+0x1f/0x40 [<ffffffffffffffff>] 0xffffffffffffffff Signed-off-by: Frédéric Dalleau <frederic.dalleau@collabora.co.uk> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2016-04-08Bluetooth: Fix setting NO_BREDR advertising flagJohan Hedberg
If we're dealing with a single-mode controller or BR/EDR is disable for a dual-mode one, the NO_BREDR flag needs to be unconditionally present in the advertising data. This patch moves it out from behind an extra condition to be always set in the create_instance_adv_data() function if BR/EDR is disabled. Signed-off-by: Johan Hedberg <johan.hedberg@intel.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2016-03-11Bluetooth: Fix setting correct flags in ADJohan Hedberg
A recent change added MGMT_ADV_FLAG_DISCOV to the flags returned by get_adv_instance_flags(), however failed to take into account limited discoverable mode. This patch fixes the issue by setting the correct discoverability flag in the AD data. Signed-off-by: Johan Hedberg <johan.hedberg@intel.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2016-03-10Bluetooth: Add support for limited privacy modeJohan Hedberg
Introduce a limited privacy mode indicated by value 0x02 to the mgmt Set Privacy command. With value 0x02 the kernel will use privacy mode with a resolvable private address. In case the controller is bondable and discoverable the identity address will be used. Signed-off-by: Johan Hedberg <johan.hedberg@intel.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2016-03-10Bluetooth: Fix adding discoverable to adv instance flagsJohan Hedberg
When lookup up the advertising instance flags for the default advertising instance (0) the discoverable flag should be filled in based on the HCI_DISCOVERABLE flag. Signed-off-by: Johan Hedberg <johan.hedberg@intel.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2016-01-29Bluetooth: Fix incorrect removing of IRKsJohan Hedberg
The commit cad20c278085d893ebd616cd20c0747a8e9d53c7 was supposed to fix handling of devices first using public addresses and then switching to RPAs after pairing. Unfortunately it missed a couple of key places in the code. 1. When evaluating which devices should be removed from the existing white list we also need to consider whether we have an IRK for them or not, i.e. a call to hci_find_irk_by_addr() is needed. 2. In smp_notify_keys() we should not be requiring the knowledge of the RPA, but should simply keep the IRK around if the other conditions require it. Signed-off-by: Johan Hedberg <johan.hedberg@intel.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org> Cc: stable@vger.kernel.org # 4.4+
2016-01-05Bluetooth: Add support for Start Limited Discovery commandJohan Hedberg
This patch implements the mgmt Start Limited Discovery command. Most of existing Start Discovery code is reused since the only difference is the presence of a 'limited' flag as part of the discovery state. Signed-off-by: Johan Hedberg <johan.hedberg@intel.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2015-12-10Bluetooth: Clean up current advertising instance trackingJohan Hedberg
We can simplify a lot of code by making sure hdev->cur_adv_instance is always up-to-date. This allows e.g. the removal of the get_current_adv_instance() helper function and the special HCI_ADV_CURRENT value. This patch also makes selecting instance 0x00 explicit in the various calls where advertising instances aren't enabled, e.g. when HCI_ADVERTISING is set or we've just finished enabling LE. Signed-off-by: Johan Hedberg <johan.hedberg@intel.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2015-12-10Bluetooth: Clean up advertising initialization in powered_update_hci()Johan Hedberg
The logic in powered_update_hci() to initialize the advertising data & state is a bit more complicated than it needs to be. It was previously not doing anything if HCI_LE_ENABLED wasn't set, but this was not obvious by quickly looking at the code. Now the conditions for the various actions are more explicit. Another simplification is due to the fact that __hci_req_schedule_adv_instance() takes care of setting hdev->cur_adv_instance so there's no need to set it before calling the function. Signed-off-by: Johan Hedberg <johan.hedberg@intel.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2015-12-10Bluetooth: Remove redundant check for req.cmd_qJohan Hedberg
The hci_req_run() function already checks for empty cmd_q and bails out if necessary. Also, req.cmd_q should really be treated as private data of the request and not accessed directly. Signed-off-by: Johan Hedberg <johan.hedberg@intel.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2015-12-10Bluetooth: Fix updating wrong instance's scan_rsp dataJohan Hedberg
The __hci_req_update_scan_rsp_data gets the instance to be updated which should get passed to update_inst_scan_rsp_data() instead of always enabling the current instance. Signed-off-by: Johan Hedberg <johan.hedberg@intel.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2015-12-10Bluetooth: Remove unnecessary HCI_ADVERTISING_INSTANCE flagJohan Hedberg
This flag just tells us whether hdev->adv_instances is empty or not. We can equally well use the list_empty() function to get this information. Signed-off-by: Johan Hedberg <johan.hedberg@intel.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2015-12-10Bluetooth: Perform HCI update for power on synchronouslyJohan Hedberg
The request to update HCI during power on is always coming either from hdev->req_workqueue or through an ioctl, so it's safe to use hci_req_sync for it. This way we also eliminate potential races with incoming mgmt commands or other actions while powering on. Part of this refactoring is the splitting of mgmt_powered() into mgmt_power_on() and __mgmt_power_off() functions. The main reason is the different requirements as far as hdev locking is concerned, as highlighted with the __ prefix of the power off API. Since the power on in the case of clearing the AUTO_OFF flag cannot be done synchronously in the set_powered mgmt handler, the hci_power_on work callback is extended to cover this (which also simplifies the set_powered helper a lot). Signed-off-by: Johan Hedberg <johan.hedberg@intel.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2015-12-10Bluetooth: Move fast connectable code to hci_request.cJohan Hedberg
We'll soon need this both in hci_request.c and mgmt.c so move it to hci_request.c as a generic helper. Signed-off-by: Johan Hedberg <johan.hedberg@intel.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2015-12-10Bluetooth: Move EIR update to hci_request.cJohan Hedberg
We'll soon need to update the EIR both from hci_request.c and mgmt.c so move update_eir() as a more generic request helper to hci_request.c. Signed-off-by: Johan Hedberg <johan.hedberg@intel.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2015-12-10Bluetooth: HCI name update to hci_request.cJohan Hedberg
We'll soon need this both from hci_request.c and mgmt.c so move it as a request helper function to hci_request.c. Signed-off-by: Johan Hedberg <johan.hedberg@intel.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2015-12-10Bluetooth: Move discoverable timeout behind hdev->req_workqueueJohan Hedberg
Since the other discoverable changes are behind req_workqueue now it only makes sense to move the discoverable timeout there as well. Signed-off-by: Johan Hedberg <johan.hedberg@intel.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2015-12-10Bluetooth: Move discoverable changes to hdev->req_workqueueJohan Hedberg
The discoverable mode is intrinsically linked with the connectable mode e.g. through sharing the same HCI command (Write Scan Enable) for BR/EDR. It makes therefore sense to move it to hci_request.c and run the changes through the same hdev->req_workqueue. Signed-off-by: Johan Hedberg <johan.hedberg@intel.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2015-12-10Bluetooth: Perform Class of Device changes through hdev->req_workqueueJohan Hedberg
The Class of Device needs to be changed e.g. for limited discoverable mode. In preparation of moving the discoverable mode to hci_request.c and hdev->req_workqueue, move the Class of Device helpers there first. Signed-off-by: Johan Hedberg <johan.hedberg@intel.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2015-12-10Bluetooth: Move connectable changes to hdev->req_workqueueJohan Hedberg
This way the connectable changes are synchronized against each other, which helps avoid potential races. The connectable mode is also linked together with LE advertising which makes is more convenient to have it behind the same workqueue. Signed-off-by: Johan Hedberg <johan.hedberg@intel.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2015-12-10Bluetooth: Move advertising instance management to hci_request.cJohan Hedberg
This paves the way for eventually performing advertising changes through the hdev->req_workqueue. Some new APIs need to be exposed from mgmt.c to hci_request.c and vice-versa, but many of them will go away once hdev->req_workqueue gets used. Signed-off-by: Johan Hedberg <johan.hedberg@intel.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2015-12-10Bluetooth: Move __hci_update_background_scan up in hci_request.cJohan Hedberg
This way we avoid the need to do a forward declaration in later patches. Signed-off-by: Johan Hedberg <johan.hedberg@intel.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2015-12-10Bluetooth: Run page scan updates through hdev->req_workqueueJohan Hedberg
Since Add/Remove Device perform the page scan updates independently from the HCI command completion we've introduced a potential race when multiple mgmt commands are queued. Doing the page scan updates through the req_workqueue ensures that the state changes are performed in a race-free manner. At the same time, to make the request helper more widely usable, extend it to also cover Inquiry Scan changes since those are behind the same HCI command. This is also reflected in the new name of the API as well as the work struct name. Signed-off-by: Johan Hedberg <johan.hedberg@intel.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2015-11-23Bluetooth: Fix returning proper HCI status from __hci_req_syncJohan Hedberg
There were a couple of code paths missed by the previous patch that added a HCI status return parameter to __hci_req_sync. This patch adds the missing assignments for them. Signed-off-by: Johan Hedberg <johan.hedberg@intel.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2015-11-19Bluetooth: Simplify request cleanup codeJohan Hedberg
The hci_req_sync_cancel() is just as much related to the request cleanup as hci_request_cancel_all() is. Just move the former into the latter and do the cleanup from a single place in hci_dev_do_close(). The important thing is to avoid deadlocks by holding the req_sync lock: previously hci_request_cancel_all was done right after releasing the lock and with this patch it's right before taking it. Signed-off-by: Johan Hedberg <johan.hedberg@intel.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2015-11-19Bluetooth: Remove unnecessary le_scan_restart_work_complete() functionJohan Hedberg
The only user of this, le_scan_restart_work(), is so short and simple that it makes sense to just merge the code there. Signed-off-by: Johan Hedberg <johan.hedberg@intel.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2015-11-19Bluetooth: Simplify le_scan_disable_work()Johan Hedberg
Merge le_scan_disable_work_complete into the main le_scan_disable_work function and take advantage of the updated bredr_inquiry() to run the Inquiry through hci_req_sync(). Signed-off-by: Johan Hedberg <johan.hedberg@intel.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2015-11-19Bluetooth: Pass inquiry length to bredr_inquiry()Johan Hedberg
Passing the needed inquiry length to bredr_inquiry() makes it possible to also use this helper for interleaved discovery where the controller doesn't support simultaneous Inquiry & LE scan. Signed-off-by: Johan Hedberg <johan.hedberg@intel.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2015-11-19Bluetooth: Move Stop Discovery to req_workqueueJohan Hedberg
Since discovery also deals with LE scanning it makes sense to move it behind the same req_workqueue as other LE scanning changes. This also simplifies the logic since we do many of the actions in a synchronous manner. Part of this refactoring is moving hci_req_stop_discovery() to hci_request.c. At the same time the function receives support for properly handling the STOPPING state since that's the state we'll be in when stopping through the req_workqueue. Signed-off-by: Johan Hedberg <johan.hedberg@intel.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2015-11-19Bluetooth: Move Start Discovery to req_workqueueJohan Hedberg
Since discovery also deals with LE scanning it makes sense to move it behind the same req_workqueue as other LE scanning changes. This also simplifies the logic since we do many of the actions in a synchronous manner. Signed-off-by: Johan Hedberg <johan.hedberg@intel.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2015-11-19Bluetooth: Add error return value to hci_req_sync callbackJohan Hedberg
In some circumstances it may be useful to abort the request through checks done in the request callback. To make the feature possible this patch changes the return value of the request callback from void to int and aborts the request if a non-zero value is returned. Signed-off-by: Johan Hedberg <johan.hedberg@intel.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2015-11-19Bluetooth: Move LE scan disable/restart behind req_workqueueJohan Hedberg
To avoid any risks of races, place also these LE scan modification work callbacks behind the same work queue as the other LE scan changes. Signed-off-by: Johan Hedberg <johan.hedberg@intel.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2015-11-19Bluetooth: Make __hci_update_background_scan private to hci_request.cJohan Hedberg
There are no more external users so this API can be made private. Signed-off-by: Johan Hedberg <johan.hedberg@intel.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2015-11-19Bluetooth: Use req_workqueue for explicit connect requestsJohan Hedberg
Since explicit connect requests are also a sub-category of passive scan updates, run them through the same workqueue as the other passive scan changes. Signed-off-by: Johan Hedberg <johan.hedberg@intel.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2015-11-19Bluetooth: Add HCI status return parameter to hci_req_sync()Johan Hedberg
In some cases it may be important to get the exact HCI status rather than the converted HCI-to-errno value. Add an optional return parameter to the hci_req_sync() API to allow for this. Since there are no good HCI translation candidates for cancelation and timeout, use the "unknown" status code for those cases. Signed-off-by: Johan Hedberg <johan.hedberg@intel.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2015-11-19Bluetooth: Run all background scan updates through req_workqueueJohan Hedberg
Instead of firing off a simple async request queue all background scan updates through req_workqueue and use hci_req_sync() there to ensure that no two updates overlap with each other. Signed-off-by: Johan Hedberg <johan.hedberg@intel.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2015-11-19Bluetooth: Add stubs for synchronous HCI request functionalityJohan Hedberg
Prepare hci_request.c to have code for doing synchronous HCI requests, such as LE scanning or advertising changes. The necessary work callbacks will be set up in hci_request_setup() and cleaned up in hci_request_cancel_all(). The former is used when an HCI device get registered, and the latter each time it is powered off (or unregistered). Signed-off-by: Johan Hedberg <johan.hedberg@intel.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2015-11-19Bluetooth: Add 'sync' specifier to synchronous request APIsJohan Hedberg
To make it clear which HCI request APIs target specifically synchronous requests, add 'sync' to the API names. Signed-off-by: Johan Hedberg <johan.hedberg@intel.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2015-11-19Bluetooth: Move synchronous request handling into hci_request.cJohan Hedberg
hci_request.c is a more natural place for the synchronous request handling. Furthermore, we will soon need access to some of the previously private-to-hci_core.c functions from hci_request.c. Signed-off-by: Johan Hedberg <johan.hedberg@intel.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>