About Kernel Documentation Linux Kernel Contact Linux Resources Linux Blog

Documentation / usb / WUSB-Design-overview.txt


Based on kernel version 4.16.1. Page generated on 2018-04-09 11:53 EST.

1	
2	Linux UWB + Wireless USB + WiNET
3	
4	   (C) 2005-2006 Intel Corporation
5	   Inaky Perez-Gonzalez <inaky.perez-gonzalez@intel.com>
6	
7	   This program is free software; you can redistribute it and/or
8	   modify it under the terms of the GNU General Public License version
9	   2 as published by the Free Software Foundation.
10	
11	   This program is distributed in the hope that it will be useful,
12	   but WITHOUT ANY WARRANTY; without even the implied warranty of
13	   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
14	   GNU General Public License for more details.
15	
16	   You should have received a copy of the GNU General Public License
17	   along with this program; if not, write to the Free Software
18	   Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
19	   02110-1301, USA.
20	
21	
22	Please visit http://bughost.org/thewiki/Design-overview.txt-1.8 for
23	updated content.
24	
25	    * Design-overview.txt-1.8
26	
27	This code implements a Ultra Wide Band stack for Linux, as well as
28	drivers for the USB based UWB radio controllers defined in the
29	Wireless USB 1.0 specification (including Wireless USB host controller
30	and an Intel WiNET controller).
31	
32	   1. Introduction
33	         1. HWA: Host Wire adapters, your Wireless USB dongle
34	
35	         2. DWA: Device Wired Adaptor, a Wireless USB hub for wired
36	            devices
37	         3. WHCI: Wireless Host Controller Interface, the PCI WUSB host
38	            adapter
39	   2. The UWB stack
40	         1. Devices and hosts: the basic structure
41	
42	         2. Host Controller life cycle
43	
44	         3. On the air: beacons and enumerating the radio neighborhood
45	
46	         4. Device lists
47	         5. Bandwidth allocation
48	
49	   3. Wireless USB Host Controller drivers
50	
51	   4. Glossary
52	
53	
54	    Introduction
55	
56	UWB is a wide-band communication protocol that is to serve also as the
57	low-level protocol for others (much like TCP sits on IP). Currently
58	these others are Wireless USB and TCP/IP, but seems Bluetooth and
59	Firewire/1394 are coming along.
60	
61	UWB uses a band from roughly 3 to 10 GHz, transmitting at a max of
62	~-41dB (or 0.074 uW/MHz--geography specific data is still being
63	negotiated w/ regulators, so watch for changes). That band is divided in
64	a bunch of ~1.5 GHz wide channels (or band groups) composed of three
65	subbands/subchannels (528 MHz each). Each channel is independent of each
66	other, so you could consider them different "busses". Initially this
67	driver considers them all a single one.
68	
69	Radio time is divided in 65536 us long /superframes/, each one divided
70	in 256 256us long /MASs/ (Media Allocation Slots), which are the basic
71	time/media allocation units for transferring data. At the beginning of
72	each superframe there is a Beacon Period (BP), where every device
73	transmit its beacon on a single MAS. The length of the BP depends on how
74	many devices are present and the length of their beacons.
75	
76	Devices have a MAC (fixed, 48 bit address) and a device (changeable, 16
77	bit address) and send periodic beacons to advertise themselves and pass
78	info on what they are and do. They advertise their capabilities and a
79	bunch of other stuff.
80	
81	The different logical parts of this driver are:
82	
83	    *
84	
85	      *UWB*: the Ultra-Wide-Band stack -- manages the radio and
86	      associated spectrum to allow for devices sharing it. Allows to
87	      control bandwidth assignment, beaconing, scanning, etc
88	
89	    *
90	
91	      *WUSB*: the layer that sits on top of UWB to provide Wireless USB.
92	      The Wireless USB spec defines means to control a UWB radio and to
93	      do the actual WUSB.
94	
95	
96	      HWA: Host Wire adapters, your Wireless USB dongle
97	
98	WUSB also defines a device called a Host Wire Adaptor (HWA), which in
99	mere terms is a USB dongle that enables your PC to have UWB and Wireless
100	USB. The Wireless USB Host Controller in a HWA looks to the host like a
101	[Wireless] USB controller connected via USB (!)
102	
103	The HWA itself is broken in two or three main interfaces:
104	
105	    *
106	
107	      *RC*: Radio control -- this implements an interface to the
108	      Ultra-Wide-Band radio controller. The driver for this implements a
109	      USB-based UWB Radio Controller to the UWB stack.
110	
111	    *
112	
113	      *HC*: the wireless USB host controller. It looks like a USB host
114	      whose root port is the radio and the WUSB devices connect to it.
115	      To the system it looks like a separate USB host. The driver (will)
116	      implement a USB host controller (similar to UHCI, OHCI or EHCI)
117	      for which the root hub is the radio...To reiterate: it is a USB
118	      controller that is connected via USB instead of PCI.
119	
120	    *
121	
122	      *WINET*: some HW provide a WiNET interface (IP over UWB). This
123	      package provides a driver for it (it looks like a network
124	      interface, winetX). The driver detects when there is a link up for
125	      their type and kick into gear.
126	
127	
128	      DWA: Device Wired Adaptor, a Wireless USB hub for wired devices
129	
130	These are the complement to HWAs. They are a USB host for connecting
131	wired devices, but it is connected to your PC connected via Wireless
132	USB. To the system it looks like yet another USB host. To the untrained
133	eye, it looks like a hub that connects upstream wirelessly.
134	
135	We still offer no support for this; however, it should share a lot of
136	code with the HWA-RC driver; there is a bunch of factorization work that
137	has been done to support that in upcoming releases.
138	
139	
140	      WHCI: Wireless Host Controller Interface, the PCI WUSB host adapter
141	
142	This is your usual PCI device that implements WHCI. Similar in concept
143	to EHCI, it allows your wireless USB devices (including DWAs) to connect
144	to your host via a PCI interface. As in the case of the HWA, it has a
145	Radio Control interface and the WUSB Host Controller interface per se.
146	
147	There is still no driver support for this, but will be in upcoming
148	releases.
149	
150	
151	    The UWB stack
152	
153	The main mission of the UWB stack is to keep a tally of which devices
154	are in radio proximity to allow drivers to connect to them. As well, it
155	provides an API for controlling the local radio controllers (RCs from
156	now on), such as to start/stop beaconing, scan, allocate bandwidth, etc.
157	
158	
159	      Devices and hosts: the basic structure
160	
161	The main building block here is the UWB device (struct uwb_dev). For
162	each device that pops up in radio presence (ie: the UWB host receives a
163	beacon from it) you get a struct uwb_dev that will show up in
164	/sys/bus/uwb/devices.
165	
166	For each RC that is detected, a new struct uwb_rc and struct uwb_dev are
167	created. An entry is also created in /sys/class/uwb_rc for each RC.
168	
169	Each RC driver is implemented by a separate driver that plugs into the
170	interface that the UWB stack provides through a struct uwb_rc_ops. The
171	spec creators have been nice enough to make the message format the same
172	for HWA and WHCI RCs, so the driver is really a very thin transport that
173	moves the requests from the UWB API to the device [/uwb_rc_ops->cmd()/]
174	and sends the replies and notifications back to the API
175	[/uwb_rc_neh_grok()/]. Notifications are handled to the UWB daemon, that
176	is chartered, among other things, to keep the tab of how the UWB radio
177	neighborhood looks, creating and destroying devices as they show up or
178	disappear.
179	
180	Command execution is very simple: a command block is sent and a event
181	block or reply is expected back. For sending/receiving command/events, a
182	handle called /neh/ (Notification/Event Handle) is opened with
183	/uwb_rc_neh_open()/.
184	
185	The HWA-RC (USB dongle) driver (drivers/uwb/hwa-rc.c) does this job for
186	the USB connected HWA. Eventually, drivers/whci-rc.c will do the same
187	for the PCI connected WHCI controller.
188	
189	
190	      Host Controller life cycle
191	
192	So let's say we connect a dongle to the system: it is detected and
193	firmware uploaded if needed [for Intel's i1480
194	/drivers/uwb/ptc/usb.c:ptc_usb_probe()/] and then it is reenumerated.
195	Now we have a real HWA device connected and
196	/drivers/uwb/hwa-rc.c:hwarc_probe()/ picks it up, that will set up the
197	Wire-Adaptor environment and then suck it into the UWB stack's vision of
198	the world [/drivers/uwb/lc-rc.c:uwb_rc_add()/].
199	
200	    *
201	
202	      [*] The stack should put a new RC to scan for devices
203	      [/uwb_rc_scan()/] so it finds what's available around and tries to
204	      connect to them, but this is policy stuff and should be driven
205	      from user space. As of now, the operator is expected to do it
206	      manually; see the release notes for documentation on the procedure.
207	
208	When a dongle is disconnected, /drivers/uwb/hwa-rc.c:hwarc_disconnect()/
209	takes time of tearing everything down safely (or not...).
210	
211	
212	      On the air: beacons and enumerating the radio neighborhood
213	
214	So assuming we have devices and we have agreed for a channel to connect
215	on (let's say 9), we put the new RC to beacon:
216	
217	    *
218	
219	            $ echo 9 0 > /sys/class/uwb_rc/uwb0/beacon
220	
221	Now it is visible. If there were other devices in the same radio channel
222	and beacon group (that's what the zero is for), the dongle's radio
223	control interface will send beacon notifications on its
224	notification/event endpoint (NEEP). The beacon notifications are part of
225	the event stream that is funneled into the API with
226	/drivers/uwb/neh.c:uwb_rc_neh_grok()/ and delivered to the UWBD, the UWB
227	daemon through a notification list.
228	
229	UWBD wakes up and scans the event list; finds a beacon and adds it to
230	the BEACON CACHE (/uwb_beca/). If he receives a number of beacons from
231	the same device, he considers it to be 'onair' and creates a new device
232	[/drivers/uwb/lc-dev.c:uwbd_dev_onair()/]. Similarly, when no beacons
233	are received in some time, the device is considered gone and wiped out
234	[uwbd calls periodically /uwb/beacon.c:uwb_beca_purge()/ that will purge
235	the beacon cache of dead devices].
236	
237	
238	      Device lists
239	
240	All UWB devices are kept in the list of the struct bus_type uwb_bus_type.
241	
242	
243	      Bandwidth allocation
244	
245	The UWB stack maintains a local copy of DRP availability through
246	processing of incoming *DRP Availability Change* notifications. This
247	local copy is currently used to present the current bandwidth
248	availability to the user through the sysfs file
249	/sys/class/uwb_rc/uwbx/bw_avail. In the future the bandwidth
250	availability information will be used by the bandwidth reservation
251	routines.
252	
253	The bandwidth reservation routines are in progress and are thus not
254	present in the current release. When completed they will enable a user
255	to initiate DRP reservation requests through interaction with sysfs. DRP
256	reservation requests from remote UWB devices will also be handled. The
257	bandwidth management done by the UWB stack will include callbacks to the
258	higher layers will enable the higher layers to use the reservations upon
259	completion. [Note: The bandwidth reservation work is in progress and
260	subject to change.]
261	
262	
263	    Wireless USB Host Controller drivers
264	
265	*WARNING* This section needs a lot of work!
266	
267	As explained above, there are three different types of HCs in the WUSB
268	world: HWA-HC, DWA-HC and WHCI-HC.
269	
270	HWA-HC and DWA-HC share that they are Wire-Adapters (USB or WUSB
271	connected controllers), and their transfer management system is almost
272	identical. So is their notification delivery system.
273	
274	HWA-HC and WHCI-HC share that they are both WUSB host controllers, so
275	they have to deal with WUSB device life cycle and maintenance, wireless
276	root-hub
277	
278	HWA exposes a Host Controller interface (HWA-HC 0xe0/02/02). This has
279	three endpoints (Notifications, Data Transfer In and Data Transfer
280	Out--known as NEP, DTI and DTO in the code).
281	
282	We reserve UWB bandwidth for our Wireless USB Cluster, create a Cluster
283	ID and tell the HC to use all that. Then we start it. This means the HC
284	starts sending MMCs.
285	
286	    *
287	
288	      The MMCs are blocks of data defined somewhere in the WUSB1.0 spec
289	      that define a stream in the UWB channel time allocated for sending
290	      WUSB IEs (host to device commands/notifications) and Device
291	      Notifications (device initiated to host). Each host defines a
292	      unique Wireless USB cluster through MMCs. Devices can connect to a
293	      single cluster at the time. The IEs are Information Elements, and
294	      among them are the bandwidth allocations that tell each device
295	      when can they transmit or receive.
296	
297	Now it all depends on external stimuli.
298	
299	*New device connection*
300	
301	A new device pops up, it scans the radio looking for MMCs that give out
302	the existence of Wireless USB channels. Once one (or more) are found,
303	selects which one to connect to. Sends a /DN_Connect/ (device
304	notification connect) during the DNTS (Device Notification Time
305	Slot--announced in the MMCs
306	
307	HC picks the /DN_Connect/ out (nep module sends to notif.c for delivery
308	into /devconnect/). This process starts the authentication process for
309	the device. First we allocate a /fake port/ and assign an
310	unauthenticated address (128 to 255--what we really do is
311	0x80 | fake_port_idx). We fiddle with the fake port status and /hub_wq/
312	sees a new connection, so he moves on to enable the fake port with a reset.
313	
314	So now we are in the reset path -- we know we have a non-yet enumerated
315	device with an unauthorized address; we ask user space to authenticate
316	(FIXME: not yet done, similar to bluetooth pairing), then we do the key
317	exchange (FIXME: not yet done) and issue a /set address 0/ to bring the
318	device to the default state. Device is authenticated.
319	
320	From here, the USB stack takes control through the usb_hcd ops. hub_wq
321	has seen the port status changes, as we have been toggling them. It will
322	start enumerating and doing transfers through usb_hcd->urb_enqueue() to
323	read descriptors and move our data.
324	
325	*Device life cycle and keep alives*
326	
327	Every time there is a successful transfer to/from a device, we update a
328	per-device activity timestamp. If not, every now and then we check and
329	if the activity timestamp gets old, we ping the device by sending it a
330	Keep Alive IE; it responds with a /DN_Alive/ pong during the DNTS (this
331	arrives to us as a notification through
332	devconnect.c:wusb_handle_dn_alive(). If a device times out, we
333	disconnect it from the system (cleaning up internal information and
334	toggling the bits in the fake hub port, which kicks hub_wq into removing
335	the rest of the stuff).
336	
337	This is done through devconnect:__wusb_check_devs(), which will scan the
338	device list looking for whom needs refreshing.
339	
340	If the device wants to disconnect, it will either die (ugly) or send a
341	/DN_Disconnect/ that will prompt a disconnection from the system.
342	
343	*Sending and receiving data*
344	
345	Data is sent and received through /Remote Pipes/ (rpipes). An rpipe is
346	/aimed/ at an endpoint in a WUSB device. This is the same for HWAs and
347	DWAs.
348	
349	Each HC has a number of rpipes and buffers that can be assigned to them;
350	when doing a data transfer (xfer), first the rpipe has to be aimed and
351	prepared (buffers assigned), then we can start queueing requests for
352	data in or out.
353	
354	Data buffers have to be segmented out before sending--so we send first a
355	header (segment request) and then if there is any data, a data buffer
356	immediately after to the DTI interface (yep, even the request). If our
357	buffer is bigger than the max segment size, then we just do multiple
358	requests.
359	
360	[This sucks, because doing USB scatter gatter in Linux is resource
361	intensive, if any...not that the current approach is not. It just has to
362	be cleaned up a lot :)].
363	
364	If reading, we don't send data buffers, just the segment headers saying
365	we want to read segments.
366	
367	When the xfer is executed, we receive a notification that says data is
368	ready in the DTI endpoint (handled through
369	xfer.c:wa_handle_notif_xfer()). In there we read from the DTI endpoint a
370	descriptor that gives us the status of the transfer, its identification
371	(given when we issued it) and the segment number. If it was a data read,
372	we issue another URB to read into the destination buffer the chunk of
373	data coming out of the remote endpoint. Done, wait for the next guy. The
374	callbacks for the URBs issued from here are the ones that will declare
375	the xfer complete at some point and call its callback.
376	
377	Seems simple, but the implementation is not trivial.
378	
379	    *
380	
381	      *WARNING* Old!!
382	
383	The main xfer descriptor, wa_xfer (equivalent to a URB) contains an
384	array of segments, tallys on segments and buffers and callback
385	information. Buried in there is a lot of URBs for executing the segments
386	and buffer transfers.
387	
388	For OUT xfers, there is an array of segments, one URB for each, another
389	one of buffer URB. When submitting, we submit URBs for segment request
390	1, buffer 1, segment 2, buffer 2...etc. Then we wait on the DTI for xfer
391	result data; when all the segments are complete, we call the callback to
392	finalize the transfer.
393	
394	For IN xfers, we only issue URBs for the segments we want to read and
395	then wait for the xfer result data.
396	
397	*URB mapping into xfers*
398	
399	This is done by hwahc_op_urb_[en|de]queue(). In enqueue() we aim an
400	rpipe to the endpoint where we have to transmit, create a transfer
401	context (wa_xfer) and submit it. When the xfer is done, our callback is
402	called and we assign the status bits and release the xfer resources.
403	
404	In dequeue() we are basically cancelling/aborting the transfer. We issue
405	a xfer abort request to the HC, cancel all the URBs we had submitted
406	and not yet done and when all that is done, the xfer callback will be
407	called--this will call the URB callback.
408	
409	
410	    Glossary
411	
412	*DWA* -- Device Wire Adapter
413	
414	USB host, wired for downstream devices, upstream connects wirelessly
415	with Wireless USB.
416	
417	*EVENT* -- Response to a command on the NEEP
418	
419	*HWA* -- Host Wire Adapter / USB dongle for UWB and Wireless USB
420	
421	*NEH* -- Notification/Event Handle
422	
423	Handle/file descriptor for receiving notifications or events. The WA
424	code requires you to get one of this to listen for notifications or
425	events on the NEEP.
426	
427	*NEEP* -- Notification/Event EndPoint
428	
429	Stuff related to the management of the first endpoint of a HWA USB
430	dongle that is used to deliver an stream of events and notifications to
431	the host.
432	
433	*NOTIFICATION* -- Message coming in the NEEP as response to something.
434	
435	*RC* -- Radio Control
436	
437	Design-overview.txt-1.8 (last edited 2006-11-04 12:22:24 by
438	InakyPerezGonzalez)
Hide Line Numbers


About Kernel Documentation Linux Kernel Contact Linux Resources Linux Blog