Commit | Line | Data |
---|---|---|
1da177e4 LT |
1 | * Introduction |
2 | ||
3 | The name "usbmon" in lowercase refers to a facility in kernel which is | |
4 | used to collect traces of I/O on the USB bus. This function is analogous | |
5 | to a packet socket used by network monitoring tools such as tcpdump(1) | |
6 | or Ethereal. Similarly, it is expected that a tool such as usbdump or | |
7 | USBMon (with uppercase letters) is used to examine raw traces produced | |
8 | by usbmon. | |
9 | ||
10 | The usbmon reports requests made by peripheral-specific drivers to Host | |
11 | Controller Drivers (HCD). So, if HCD is buggy, the traces reported by | |
12 | usbmon may not correspond to bus transactions precisely. This is the same | |
13 | situation as with tcpdump. | |
14 | ||
15 | * How to use usbmon to collect raw text traces | |
16 | ||
17 | Unlike the packet socket, usbmon has an interface which provides traces | |
18 | in a text format. This is used for two purposes. First, it serves as a | |
f1c9e30b | 19 | common trace exchange format for tools while more sophisticated formats |
1da177e4 LT |
20 | are finalized. Second, humans can read it in case tools are not available. |
21 | ||
22 | To collect a raw text trace, execute following steps. | |
23 | ||
24 | 1. Prepare | |
25 | ||
26 | Mount debugfs (it has to be enabled in your kernel configuration), and | |
27 | load the usbmon module (if built as module). The second step is skipped | |
28 | if usbmon is built into the kernel. | |
29 | ||
30 | # mount -t debugfs none_debugs /sys/kernel/debug | |
31 | # modprobe usbmon | |
d9ac2cfc | 32 | # |
1da177e4 LT |
33 | |
34 | Verify that bus sockets are present. | |
35 | ||
d9ac2cfc | 36 | # ls /sys/kernel/debug/usbmon |
f1c9e30b | 37 | 1s 1t 1u 2s 2t 2u 3s 3t 3u 4s 4t 4u |
d9ac2cfc | 38 | # |
1da177e4 LT |
39 | |
40 | 2. Find which bus connects to the desired device | |
41 | ||
42 | Run "cat /proc/bus/usb/devices", and find the T-line which corresponds to | |
43 | the device. Usually you do it by looking for the vendor string. If you have | |
44 | many similar devices, unplug one and compare two /proc/bus/usb/devices outputs. | |
45 | The T-line will have a bus number. Example: | |
46 | ||
47 | T: Bus=03 Lev=01 Prnt=01 Port=00 Cnt=01 Dev#= 2 Spd=12 MxCh= 0 | |
48 | D: Ver= 1.10 Cls=00(>ifc ) Sub=00 Prot=00 MxPS= 8 #Cfgs= 1 | |
49 | P: Vendor=0557 ProdID=2004 Rev= 1.00 | |
50 | S: Manufacturer=ATEN | |
51 | S: Product=UC100KM V2.00 | |
52 | ||
53 | Bus=03 means it's bus 3. | |
54 | ||
55 | 3. Start 'cat' | |
56 | ||
f1c9e30b | 57 | # cat /sys/kernel/debug/usbmon/3u > /tmp/1.mon.out |
1da177e4 LT |
58 | |
59 | This process will be reading until killed. Naturally, the output can be | |
60 | redirected to a desirable location. This is preferred, because it is going | |
61 | to be quite long. | |
62 | ||
63 | 4. Perform the desired operation on the USB bus | |
64 | ||
65 | This is where you do something that creates the traffic: plug in a flash key, | |
66 | copy files, control a webcam, etc. | |
67 | ||
68 | 5. Kill cat | |
69 | ||
70 | Usually it's done with a keyboard interrupt (Control-C). | |
71 | ||
72 | At this point the output file (/tmp/1.mon.out in this example) can be saved, | |
73 | sent by e-mail, or inspected with a text editor. In the last case make sure | |
74 | that the file size is not excessive for your favourite editor. | |
75 | ||
76 | * Raw text data format | |
77 | ||
f1c9e30b PZ |
78 | Two formats are supported currently: the original, or '1t' format, and |
79 | the '1u' format. The '1t' format is deprecated in kernel 2.6.21. The '1u' | |
80 | format adds a few fields, such as ISO frame descriptors, interval, etc. | |
81 | It produces slightly longer lines, but otherwise is a perfect superset | |
82 | of '1t' format. | |
83 | ||
84 | If it is desired to recognize one from the other in a program, look at the | |
85 | "address" word (see below), where '1u' format adds a bus number. If 2 colons | |
86 | are present, it's the '1t' format, otherwise '1u'. | |
87 | ||
88 | Any text format data consists of a stream of events, such as URB submission, | |
1da177e4 | 89 | URB callback, submission error. Every event is a text line, which consists |
6f23ee1f | 90 | of whitespace separated words. The number or position of words may depend |
1da177e4 LT |
91 | on the event type, but there is a set of words, common for all types. |
92 | ||
93 | Here is the list of words, from left to right: | |
f1c9e30b | 94 | |
1da177e4 LT |
95 | - URB Tag. This is used to identify URBs is normally a kernel mode address |
96 | of the URB structure in hexadecimal. | |
f1c9e30b | 97 | |
1da177e4 LT |
98 | - Timestamp in microseconds, a decimal number. The timestamp's resolution |
99 | depends on available clock, and so it can be much worse than a microsecond | |
100 | (if the implementation uses jiffies, for example). | |
f1c9e30b | 101 | |
1da177e4 LT |
102 | - Event Type. This type refers to the format of the event, not URB type. |
103 | Available types are: S - submission, C - callback, E - submission error. | |
f1c9e30b PZ |
104 | |
105 | - "Address" word (formerly a "pipe"). It consists of four fields, separated by | |
106 | colons: URB type and direction, Bus number, Device address, Endpoint number. | |
1da177e4 LT |
107 | Type and direction are encoded with two bytes in the following manner: |
108 | Ci Co Control input and output | |
109 | Zi Zo Isochronous input and output | |
110 | Ii Io Interrupt input and output | |
111 | Bi Bo Bulk input and output | |
f1c9e30b PZ |
112 | Bus number, Device address, and Endpoint are decimal numbers, but they may |
113 | have leading zeros, for the sake of human readers. | |
114 | ||
115 | - URB Status word. This is either a letter, or several numbers separated | |
116 | by colons: URB status, interval, start frame, and error count. Unlike the | |
117 | "address" word, all fields save the status are optional. Interval is printed | |
118 | only for interrupt and isochronous URBs. Start frame is printed only for | |
119 | isochronous URBs. Error count is printed only for isochronous callback | |
120 | events. | |
121 | ||
122 | The status field is a decimal number, sometimes negative, which represents | |
123 | a "status" field of the URB. This field makes no sense for submissions, but | |
124 | is present anyway to help scripts with parsing. When an error occurs, the | |
125 | field contains the error code. | |
126 | ||
127 | In case of a submission of a Control packet, this field contains a Setup Tag | |
128 | instead of an group of numbers. It is easy to tell whether the Setup Tag is | |
129 | present because it is never a number. Thus if scripts find a set of numbers | |
130 | in this word, they proceed to read Data Length (except for isochronous URBs). | |
131 | If they find something else, like a letter, they read the setup packet before | |
132 | reading the Data Length or isochronous descriptors. | |
133 | ||
ae0d6cce PZ |
134 | - Setup packet, if present, consists of 5 words: one of each for bmRequestType, |
135 | bRequest, wValue, wIndex, wLength, as specified by the USB Specification 2.0. | |
136 | These words are safe to decode if Setup Tag was 's'. Otherwise, the setup | |
137 | packet was present, but not captured, and the fields contain filler. | |
f1c9e30b PZ |
138 | |
139 | - Number of isochronous frame descriptors and descriptors themselves. | |
140 | If an Isochronous transfer event has a set of descriptors, a total number | |
141 | of them in an URB is printed first, then a word per descriptor, up to a | |
142 | total of 5. The word consists of 3 colon-separated decimal numbers for | |
143 | status, offset, and length respectively. For submissions, initial length | |
144 | is reported. For callbacks, actual length is reported. | |
145 | ||
d9ac2cfc PZ |
146 | - Data Length. For submissions, this is the requested length. For callbacks, |
147 | this is the actual length. | |
f1c9e30b | 148 | |
1da177e4 | 149 | - Data tag. The usbmon may not always capture data, even if length is nonzero. |
d9ac2cfc | 150 | The data words are present only if this tag is '='. |
f1c9e30b | 151 | |
1da177e4 LT |
152 | - Data words follow, in big endian hexadecimal format. Notice that they are |
153 | not machine words, but really just a byte stream split into words to make | |
154 | it easier to read. Thus, the last word may contain from one to four bytes. | |
155 | The length of collected data is limited and can be less than the data length | |
156 | report in Data Length word. | |
157 | ||
158 | Here is an example of code to read the data stream in a well known programming | |
159 | language: | |
160 | ||
161 | class ParsedLine { | |
162 | int data_len; /* Available length of data */ | |
163 | byte data[]; | |
164 | ||
165 | void parseData(StringTokenizer st) { | |
166 | int availwords = st.countTokens(); | |
167 | data = new byte[availwords * 4]; | |
168 | data_len = 0; | |
169 | while (st.hasMoreTokens()) { | |
170 | String data_str = st.nextToken(); | |
171 | int len = data_str.length() / 2; | |
172 | int i; | |
ae0d6cce | 173 | int b; // byte is signed, apparently?! XXX |
1da177e4 | 174 | for (i = 0; i < len; i++) { |
ae0d6cce PZ |
175 | // data[data_len] = Byte.parseByte( |
176 | // data_str.substring(i*2, i*2 + 2), | |
177 | // 16); | |
178 | b = Integer.parseInt( | |
179 | data_str.substring(i*2, i*2 + 2), | |
180 | 16); | |
181 | if (b >= 128) | |
182 | b *= -1; | |
183 | data[data_len] = (byte) b; | |
1da177e4 LT |
184 | data_len++; |
185 | } | |
186 | } | |
187 | } | |
188 | } | |
189 | ||
1da177e4 LT |
190 | Examples: |
191 | ||
ae0d6cce | 192 | An input control transfer to get a port status. |
1da177e4 | 193 | |
f1c9e30b PZ |
194 | d5ea89a0 3575914555 S Ci:1:001:0 s a3 00 0000 0003 0004 4 < |
195 | d5ea89a0 3575914560 C Ci:1:001:0 0 4 = 01050000 | |
1da177e4 LT |
196 | |
197 | An output bulk transfer to send a SCSI command 0x5E in a 31-byte Bulk wrapper | |
198 | to a storage device at address 5: | |
199 | ||
f1c9e30b PZ |
200 | dd65f0e8 4128379752 S Bo:1:005:2 -115 31 = 55534243 5e000000 00000000 00000600 00000000 00000000 00000000 000000 |
201 | dd65f0e8 4128379808 C Bo:1:005:2 0 31 > | |
1da177e4 LT |
202 | |
203 | * Raw binary format and API | |
204 | ||
6f23ee1f PZ |
205 | The overall architecture of the API is about the same as the one above, |
206 | only the events are delivered in binary format. Each event is sent in | |
207 | the following structure (its name is made up, so that we can refer to it): | |
208 | ||
209 | struct usbmon_packet { | |
210 | u64 id; /* 0: URB ID - from submission to callback */ | |
211 | unsigned char type; /* 8: Same as text; extensible. */ | |
212 | unsigned char xfer_type; /* ISO (0), Intr, Control, Bulk (3) */ | |
213 | unsigned char epnum; /* Endpoint number and transfer direction */ | |
214 | unsigned char devnum; /* Device address */ | |
215 | u16 busnum; /* 12: Bus number */ | |
216 | char flag_setup; /* 14: Same as text */ | |
217 | char flag_data; /* 15: Same as text; Binary zero is OK. */ | |
218 | s64 ts_sec; /* 16: gettimeofday */ | |
219 | s32 ts_usec; /* 24: gettimeofday */ | |
220 | int status; /* 28: */ | |
221 | unsigned int length; /* 32: Length of data (submitted or actual) */ | |
222 | unsigned int len_cap; /* 36: Delivered length */ | |
223 | unsigned char setup[8]; /* 40: Only for Control 'S' */ | |
224 | }; /* 48 bytes total */ | |
225 | ||
226 | These events can be received from a character device by reading with read(2), | |
227 | with an ioctl(2), or by accessing the buffer with mmap. | |
228 | ||
229 | The character device is usually called /dev/usbmonN, where N is the USB bus | |
230 | number. Number zero (/dev/usbmon0) is special and means "all buses". | |
231 | However, this feature is not implemented yet. Note that specific naming | |
232 | policy is set by your Linux distribution. | |
233 | ||
234 | If you create /dev/usbmon0 by hand, make sure that it is owned by root | |
235 | and has mode 0600. Otherwise, unpriviledged users will be able to snoop | |
236 | keyboard traffic. | |
237 | ||
238 | The following ioctl calls are available, with MON_IOC_MAGIC 0x92: | |
239 | ||
240 | MON_IOCQ_URB_LEN, defined as _IO(MON_IOC_MAGIC, 1) | |
241 | ||
242 | This call returns the length of data in the next event. Note that majority of | |
243 | events contain no data, so if this call returns zero, it does not mean that | |
244 | no events are available. | |
245 | ||
246 | MON_IOCG_STATS, defined as _IOR(MON_IOC_MAGIC, 3, struct mon_bin_stats) | |
247 | ||
248 | The argument is a pointer to the following structure: | |
249 | ||
250 | struct mon_bin_stats { | |
251 | u32 queued; | |
252 | u32 dropped; | |
253 | }; | |
254 | ||
255 | The member "queued" refers to the number of events currently queued in the | |
256 | buffer (and not to the number of events processed since the last reset). | |
257 | ||
258 | The member "dropped" is the number of events lost since the last call | |
259 | to MON_IOCG_STATS. | |
260 | ||
261 | MON_IOCT_RING_SIZE, defined as _IO(MON_IOC_MAGIC, 4) | |
262 | ||
263 | This call sets the buffer size. The argument is the size in bytes. | |
264 | The size may be rounded down to the next chunk (or page). If the requested | |
265 | size is out of [unspecified] bounds for this kernel, the call fails with | |
266 | -EINVAL. | |
267 | ||
268 | MON_IOCQ_RING_SIZE, defined as _IO(MON_IOC_MAGIC, 5) | |
269 | ||
270 | This call returns the current size of the buffer in bytes. | |
271 | ||
272 | MON_IOCX_GET, defined as _IOW(MON_IOC_MAGIC, 6, struct mon_get_arg) | |
273 | ||
274 | This call waits for events to arrive if none were in the kernel buffer, | |
275 | then returns the first event. Its argument is a pointer to the following | |
276 | structure: | |
277 | ||
278 | struct mon_get_arg { | |
279 | struct usbmon_packet *hdr; | |
280 | void *data; | |
281 | size_t alloc; /* Length of data (can be zero) */ | |
282 | }; | |
283 | ||
284 | Before the call, hdr, data, and alloc should be filled. Upon return, the area | |
285 | pointed by hdr contains the next event structure, and the data buffer contains | |
286 | the data, if any. The event is removed from the kernel buffer. | |
287 | ||
288 | MON_IOCX_MFETCH, defined as _IOWR(MON_IOC_MAGIC, 7, struct mon_mfetch_arg) | |
289 | ||
290 | This ioctl is primarily used when the application accesses the buffer | |
291 | with mmap(2). Its argument is a pointer to the following structure: | |
292 | ||
293 | struct mon_mfetch_arg { | |
294 | uint32_t *offvec; /* Vector of events fetched */ | |
295 | uint32_t nfetch; /* Number of events to fetch (out: fetched) */ | |
296 | uint32_t nflush; /* Number of events to flush */ | |
297 | }; | |
298 | ||
299 | The ioctl operates in 3 stages. | |
300 | ||
301 | First, it removes and discards up to nflush events from the kernel buffer. | |
302 | The actual number of events discarded is returned in nflush. | |
303 | ||
304 | Second, it waits for an event to be present in the buffer, unless the pseudo- | |
305 | device is open with O_NONBLOCK. | |
306 | ||
307 | Third, it extracts up to nfetch offsets into the mmap buffer, and stores | |
308 | them into the offvec. The actual number of event offsets is stored into | |
309 | the nfetch. | |
310 | ||
311 | MON_IOCH_MFLUSH, defined as _IO(MON_IOC_MAGIC, 8) | |
312 | ||
313 | This call removes a number of events from the kernel buffer. Its argument | |
314 | is the number of events to remove. If the buffer contains fewer events | |
315 | than requested, all events present are removed, and no error is reported. | |
316 | This works when no events are available too. | |
317 | ||
318 | FIONBIO | |
319 | ||
320 | The ioctl FIONBIO may be implemented in the future, if there's a need. | |
321 | ||
322 | In addition to ioctl(2) and read(2), the special file of binary API can | |
323 | be polled with select(2) and poll(2). But lseek(2) does not work. | |
324 | ||
325 | * Memory-mapped access of the kernel buffer for the binary API | |
326 | ||
327 | The basic idea is simple: | |
328 | ||
329 | To prepare, map the buffer by getting the current size, then using mmap(2). | |
330 | Then, execute a loop similar to the one written in pseudo-code below: | |
331 | ||
332 | struct mon_mfetch_arg fetch; | |
333 | struct usbmon_packet *hdr; | |
334 | int nflush = 0; | |
335 | for (;;) { | |
336 | fetch.offvec = vec; // Has N 32-bit words | |
337 | fetch.nfetch = N; // Or less than N | |
338 | fetch.nflush = nflush; | |
339 | ioctl(fd, MON_IOCX_MFETCH, &fetch); // Process errors, too | |
340 | nflush = fetch.nfetch; // This many packets to flush when done | |
341 | for (i = 0; i < nflush; i++) { | |
342 | hdr = (struct ubsmon_packet *) &mmap_area[vec[i]]; | |
343 | if (hdr->type == '@') // Filler packet | |
344 | continue; | |
345 | caddr_t data = &mmap_area[vec[i]] + 64; | |
346 | process_packet(hdr, data); | |
347 | } | |
348 | } | |
349 | ||
350 | Thus, the main idea is to execute only one ioctl per N events. | |
351 | ||
352 | Although the buffer is circular, the returned headers and data do not cross | |
353 | the end of the buffer, so the above pseudo-code does not need any gathering. |