blob: 54504c5aa047d32992bfe2dc1c2fe12b4e00d67f
1 | If variable is of Type, use printk format specifier: |
2 | --------------------------------------------------------- |
3 | int %d or %x |
4 | unsigned int %u or %x |
5 | long %ld or %lx |
6 | unsigned long %lu or %lx |
7 | long long %lld or %llx |
8 | unsigned long long %llu or %llx |
9 | size_t %zu or %zx |
10 | ssize_t %zd or %zx |
11 | s32 %d or %x |
12 | u32 %u or %x |
13 | s64 %lld or %llx |
14 | u64 %llu or %llx |
15 | |
16 | If <type> is dependent on a config option for its size (e.g., sector_t, |
17 | blkcnt_t) or is architecture-dependent for its size (e.g., tcflag_t), use a |
18 | format specifier of its largest possible type and explicitly cast to it. |
19 | Example: |
20 | |
21 | printk("test: sector number/total blocks: %llu/%llu\n", |
22 | (unsigned long long)sector, (unsigned long long)blockcount); |
23 | |
24 | Reminder: sizeof() result is of type size_t. |
25 | |
26 | The kernel's printf does not support %n. For obvious reasons, floating |
27 | point formats (%e, %f, %g, %a) are also not recognized. Use of any |
28 | unsupported specifier or length qualifier results in a WARN and early |
29 | return from vsnprintf. |
30 | |
31 | Raw pointer value SHOULD be printed with %p. The kernel supports |
32 | the following extended format specifiers for pointer types: |
33 | |
34 | Pointer Types: |
35 | |
36 | Pointers printed without a specifier extension (i.e unadorned %p) are |
37 | hashed to give a unique identifier without leaking kernel addresses to user |
38 | space. On 64 bit machines the first 32 bits are zeroed. If you _really_ |
39 | want the address see %px below. |
40 | |
41 | %p abcdef12 or 00000000abcdef12 |
42 | |
43 | Symbols/Function Pointers: |
44 | |
45 | %pF versatile_init+0x0/0x110 |
46 | %pf versatile_init |
47 | %pS versatile_init+0x0/0x110 |
48 | %pSR versatile_init+0x9/0x110 |
49 | (with __builtin_extract_return_addr() translation) |
50 | %ps versatile_init |
51 | %pB prev_fn_of_versatile_init+0x88/0x88 |
52 | |
53 | For printing symbols and function pointers. The 'S' and 's' specifiers |
54 | result in the symbol name with ('S') or without ('s') offsets. Where |
55 | this is used on a kernel without KALLSYMS - the symbol address is |
56 | printed instead. |
57 | |
58 | The 'B' specifier results in the symbol name with offsets and should be |
59 | used when printing stack backtraces. The specifier takes into |
60 | consideration the effect of compiler optimisations which may occur |
61 | when tail-call's are used and marked with the noreturn GCC attribute. |
62 | |
63 | On ia64, ppc64 and parisc64 architectures function pointers are |
64 | actually function descriptors which must first be resolved. The 'F' and |
65 | 'f' specifiers perform this resolution and then provide the same |
66 | functionality as the 'S' and 's' specifiers. |
67 | |
68 | Kernel Pointers: |
69 | |
70 | %pK 01234567 or 0123456789abcdef |
71 | |
72 | For printing kernel pointers which should be hidden from unprivileged |
73 | users. The behaviour of %pK depends on the kptr_restrict sysctl - see |
74 | Documentation/sysctl/kernel.txt for more details. |
75 | |
76 | Unmodified Addresses: |
77 | |
78 | %px 01234567 or 0123456789abcdef |
79 | |
80 | For printing pointers when you _really_ want to print the address. Please |
81 | consider whether or not you are leaking sensitive information about the |
82 | Kernel layout in memory before printing pointers with %px. %px is |
83 | functionally equivalent to %lx. %px is preferred to %lx because it is more |
84 | uniquely grep'able. If, in the future, we need to modify the way the Kernel |
85 | handles printing pointers it will be nice to be able to find the call |
86 | sites. |
87 | |
88 | Struct Resources: |
89 | |
90 | %pr [mem 0x60000000-0x6fffffff flags 0x2200] or |
91 | [mem 0x0000000060000000-0x000000006fffffff flags 0x2200] |
92 | %pR [mem 0x60000000-0x6fffffff pref] or |
93 | [mem 0x0000000060000000-0x000000006fffffff pref] |
94 | |
95 | For printing struct resources. The 'R' and 'r' specifiers result in a |
96 | printed resource with ('R') or without ('r') a decoded flags member. |
97 | Passed by reference. |
98 | |
99 | Physical addresses types phys_addr_t: |
100 | |
101 | %pa[p] 0x01234567 or 0x0123456789abcdef |
102 | |
103 | For printing a phys_addr_t type (and its derivatives, such as |
104 | resource_size_t) which can vary based on build options, regardless of |
105 | the width of the CPU data path. Passed by reference. |
106 | |
107 | DMA addresses types dma_addr_t: |
108 | |
109 | %pad 0x01234567 or 0x0123456789abcdef |
110 | |
111 | For printing a dma_addr_t type which can vary based on build options, |
112 | regardless of the width of the CPU data path. Passed by reference. |
113 | |
114 | Raw buffer as an escaped string: |
115 | |
116 | %*pE[achnops] |
117 | |
118 | For printing raw buffer as an escaped string. For the following buffer |
119 | |
120 | 1b 62 20 5c 43 07 22 90 0d 5d |
121 | |
122 | few examples show how the conversion would be done (the result string |
123 | without surrounding quotes): |
124 | |
125 | %*pE "\eb \C\a"\220\r]" |
126 | %*pEhp "\x1bb \C\x07"\x90\x0d]" |
127 | %*pEa "\e\142\040\\\103\a\042\220\r\135" |
128 | |
129 | The conversion rules are applied according to an optional combination |
130 | of flags (see string_escape_mem() kernel documentation for the |
131 | details): |
132 | a - ESCAPE_ANY |
133 | c - ESCAPE_SPECIAL |
134 | h - ESCAPE_HEX |
135 | n - ESCAPE_NULL |
136 | o - ESCAPE_OCTAL |
137 | p - ESCAPE_NP |
138 | s - ESCAPE_SPACE |
139 | By default ESCAPE_ANY_NP is used. |
140 | |
141 | ESCAPE_ANY_NP is the sane choice for many cases, in particularly for |
142 | printing SSIDs. |
143 | |
144 | If field width is omitted the 1 byte only will be escaped. |
145 | |
146 | Raw buffer as a hex string: |
147 | |
148 | %*ph 00 01 02 ... 3f |
149 | %*phC 00:01:02: ... :3f |
150 | %*phD 00-01-02- ... -3f |
151 | %*phN 000102 ... 3f |
152 | |
153 | For printing a small buffers (up to 64 bytes long) as a hex string with |
154 | certain separator. For the larger buffers consider to use |
155 | print_hex_dump(). |
156 | |
157 | MAC/FDDI addresses: |
158 | |
159 | %pM 00:01:02:03:04:05 |
160 | %pMR 05:04:03:02:01:00 |
161 | %pMF 00-01-02-03-04-05 |
162 | %pm 000102030405 |
163 | %pmR 050403020100 |
164 | |
165 | For printing 6-byte MAC/FDDI addresses in hex notation. The 'M' and 'm' |
166 | specifiers result in a printed address with ('M') or without ('m') byte |
167 | separators. The default byte separator is the colon (':'). |
168 | |
169 | Where FDDI addresses are concerned the 'F' specifier can be used after |
170 | the 'M' specifier to use dash ('-') separators instead of the default |
171 | separator. |
172 | |
173 | For Bluetooth addresses the 'R' specifier shall be used after the 'M' |
174 | specifier to use reversed byte order suitable for visual interpretation |
175 | of Bluetooth addresses which are in the little endian order. |
176 | |
177 | Passed by reference. |
178 | |
179 | IPv4 addresses: |
180 | |
181 | %pI4 1.2.3.4 |
182 | %pi4 001.002.003.004 |
183 | %p[Ii]4[hnbl] |
184 | |
185 | For printing IPv4 dot-separated decimal addresses. The 'I4' and 'i4' |
186 | specifiers result in a printed address with ('i4') or without ('I4') |
187 | leading zeros. |
188 | |
189 | The additional 'h', 'n', 'b', and 'l' specifiers are used to specify |
190 | host, network, big or little endian order addresses respectively. Where |
191 | no specifier is provided the default network/big endian order is used. |
192 | |
193 | Passed by reference. |
194 | |
195 | IPv6 addresses: |
196 | |
197 | %pI6 0001:0002:0003:0004:0005:0006:0007:0008 |
198 | %pi6 00010002000300040005000600070008 |
199 | %pI6c 1:2:3:4:5:6:7:8 |
200 | |
201 | For printing IPv6 network-order 16-bit hex addresses. The 'I6' and 'i6' |
202 | specifiers result in a printed address with ('I6') or without ('i6') |
203 | colon-separators. Leading zeros are always used. |
204 | |
205 | The additional 'c' specifier can be used with the 'I' specifier to |
206 | print a compressed IPv6 address as described by |
207 | http://tools.ietf.org/html/rfc5952 |
208 | |
209 | Passed by reference. |
210 | |
211 | IPv4/IPv6 addresses (generic, with port, flowinfo, scope): |
212 | |
213 | %pIS 1.2.3.4 or 0001:0002:0003:0004:0005:0006:0007:0008 |
214 | %piS 001.002.003.004 or 00010002000300040005000600070008 |
215 | %pISc 1.2.3.4 or 1:2:3:4:5:6:7:8 |
216 | %pISpc 1.2.3.4:12345 or [1:2:3:4:5:6:7:8]:12345 |
217 | %p[Ii]S[pfschnbl] |
218 | |
219 | For printing an IP address without the need to distinguish whether it's |
220 | of type AF_INET or AF_INET6, a pointer to a valid 'struct sockaddr', |
221 | specified through 'IS' or 'iS', can be passed to this format specifier. |
222 | |
223 | The additional 'p', 'f', and 's' specifiers are used to specify port |
224 | (IPv4, IPv6), flowinfo (IPv6) and scope (IPv6). Ports have a ':' prefix, |
225 | flowinfo a '/' and scope a '%', each followed by the actual value. |
226 | |
227 | In case of an IPv6 address the compressed IPv6 address as described by |
228 | http://tools.ietf.org/html/rfc5952 is being used if the additional |
229 | specifier 'c' is given. The IPv6 address is surrounded by '[', ']' in |
230 | case of additional specifiers 'p', 'f' or 's' as suggested by |
231 | https://tools.ietf.org/html/draft-ietf-6man-text-addr-representation-07 |
232 | |
233 | In case of IPv4 addresses, the additional 'h', 'n', 'b', and 'l' |
234 | specifiers can be used as well and are ignored in case of an IPv6 |
235 | address. |
236 | |
237 | Passed by reference. |
238 | |
239 | Further examples: |
240 | |
241 | %pISfc 1.2.3.4 or [1:2:3:4:5:6:7:8]/123456789 |
242 | %pISsc 1.2.3.4 or [1:2:3:4:5:6:7:8]%1234567890 |
243 | %pISpfc 1.2.3.4:12345 or [1:2:3:4:5:6:7:8]:12345/123456789 |
244 | |
245 | UUID/GUID addresses: |
246 | |
247 | %pUb 00010203-0405-0607-0809-0a0b0c0d0e0f |
248 | %pUB 00010203-0405-0607-0809-0A0B0C0D0E0F |
249 | %pUl 03020100-0504-0706-0809-0a0b0c0e0e0f |
250 | %pUL 03020100-0504-0706-0809-0A0B0C0E0E0F |
251 | |
252 | For printing 16-byte UUID/GUIDs addresses. The additional 'l', 'L', |
253 | 'b' and 'B' specifiers are used to specify a little endian order in |
254 | lower ('l') or upper case ('L') hex characters - and big endian order |
255 | in lower ('b') or upper case ('B') hex characters. |
256 | |
257 | Where no additional specifiers are used the default big endian |
258 | order with lower case hex characters will be printed. |
259 | |
260 | Passed by reference. |
261 | |
262 | dentry names: |
263 | |
264 | %pd{,2,3,4} |
265 | %pD{,2,3,4} |
266 | |
267 | For printing dentry name; if we race with d_move(), the name might be |
268 | a mix of old and new ones, but it won't oops. %pd dentry is a safer |
269 | equivalent of %s dentry->d_name.name we used to use, %pd<n> prints |
270 | n last components. %pD does the same thing for struct file. |
271 | |
272 | Passed by reference. |
273 | |
274 | block_device names: |
275 | |
276 | %pg sda, sda1 or loop0p1 |
277 | |
278 | For printing name of block_device pointers. |
279 | |
280 | struct va_format: |
281 | |
282 | %pV |
283 | |
284 | For printing struct va_format structures. These contain a format string |
285 | and va_list as follows: |
286 | |
287 | struct va_format { |
288 | const char *fmt; |
289 | va_list *va; |
290 | }; |
291 | |
292 | Implements a "recursive vsnprintf". |
293 | |
294 | Do not use this feature without some mechanism to verify the |
295 | correctness of the format string and va_list arguments. |
296 | |
297 | Passed by reference. |
298 | |
299 | struct clk: |
300 | |
301 | %pC pll1 |
302 | %pCn pll1 |
303 | |
304 | For printing struct clk structures. '%pC' and '%pCn' print the name |
305 | (Common Clock Framework) or address (legacy clock framework) of the |
306 | structure. |
307 | |
308 | Passed by reference. |
309 | |
310 | bitmap and its derivatives such as cpumask and nodemask: |
311 | |
312 | %*pb 0779 |
313 | %*pbl 0,3-6,8-10 |
314 | |
315 | For printing bitmap and its derivatives such as cpumask and nodemask, |
316 | %*pb output the bitmap with field width as the number of bits and %*pbl |
317 | output the bitmap as range list with field width as the number of bits. |
318 | |
319 | Passed by reference. |
320 | |
321 | Flags bitfields such as page flags, gfp_flags: |
322 | |
323 | %pGp referenced|uptodate|lru|active|private |
324 | %pGg GFP_USER|GFP_DMA32|GFP_NOWARN |
325 | %pGv read|exec|mayread|maywrite|mayexec|denywrite |
326 | |
327 | For printing flags bitfields as a collection of symbolic constants that |
328 | would construct the value. The type of flags is given by the third |
329 | character. Currently supported are [p]age flags, [v]ma_flags (both |
330 | expect unsigned long *) and [g]fp_flags (expects gfp_t *). The flag |
331 | names and print order depends on the particular type. |
332 | |
333 | Note that this format should not be used directly in TP_printk() part |
334 | of a tracepoint. Instead, use the show_*_flags() functions from |
335 | <trace/events/mmflags.h>. |
336 | |
337 | Passed by reference. |
338 | |
339 | Network device features: |
340 | |
341 | %pNF 0x000000000000c000 |
342 | |
343 | For printing netdev_features_t. |
344 | |
345 | Passed by reference. |
346 | |
347 | If you add other %p extensions, please extend lib/test_printf.c with |
348 | one or more test cases, if at all feasible. |
349 | |
350 | |
351 | Thank you for your cooperation and attention. |
352 | |
353 | |
354 | By Randy Dunlap <rdunlap@infradead.org> and |
355 | Andrew Murray <amurray@mpc-data.co.uk> |
356 |