printk.c 84.3 KB
Newer Older
Linus Torvalds's avatar
Linus Torvalds committed
1
2
3
4
5
6
7
8
9
10
11
12
/*
 *  linux/kernel/printk.c
 *
 *  Copyright (C) 1991, 1992  Linus Torvalds
 *
 * Modified to make sys_syslog() more flexible: added commands to
 * return the last 4k of kernel messages, regardless of whether
 * they've been read or not.  Added option to suppress kernel printk's
 * to the console.  Added hook for sending the console messages
 * elsewhere, in preparation for a serial line console (someday).
 * Ted Ts'o, 2/11/93.
 * Modified for sysctl support, 1/8/97, Chris Horn.
13
 * Fixed SMP synchronization, 08/08/99, Manfred Spraul
14
 *     manfred@colorfullife.com
Linus Torvalds's avatar
Linus Torvalds committed
15
 * Rewrote bits to get rid of console_lock
16
 *	01Mar01 Andrew Morton
Linus Torvalds's avatar
Linus Torvalds committed
17
18
19
20
21
22
23
24
 */

#include <linux/kernel.h>
#include <linux/mm.h>
#include <linux/tty.h>
#include <linux/tty_driver.h>
#include <linux/console.h>
#include <linux/init.h>
Randy Dunlap's avatar
Randy Dunlap committed
25
26
#include <linux/jiffies.h>
#include <linux/nmi.h>
Linus Torvalds's avatar
Linus Torvalds committed
27
#include <linux/module.h>
Jan Engelhardt's avatar
Jan Engelhardt committed
28
#include <linux/moduleparam.h>
Linus Torvalds's avatar
Linus Torvalds committed
29
30
31
32
#include <linux/delay.h>
#include <linux/smp.h>
#include <linux/security.h>
#include <linux/bootmem.h>
33
#include <linux/memblock.h>
Linus Torvalds's avatar
Linus Torvalds committed
34
#include <linux/syscalls.h>
35
#include <linux/kexec.h>
36
#include <linux/kdb.h>
37
#include <linux/ratelimit.h>
38
#include <linux/kmsg_dump.h>
39
#include <linux/syslog.h>
40
41
#include <linux/cpu.h>
#include <linux/notifier.h>
42
#include <linux/rculist.h>
43
#include <linux/poll.h>
44
#include <linux/irq_work.h>
45
#include <linux/utsname.h>
Alex Elder's avatar
Alex Elder committed
46
#include <linux/ctype.h>
47
#include <linux/uio.h>
Linus Torvalds's avatar
Linus Torvalds committed
48
49

#include <asm/uaccess.h>
50
#include <asm/sections.h>
Linus Torvalds's avatar
Linus Torvalds committed
51

52
53
54
#define CREATE_TRACE_POINTS
#include <trace/events/printk.h>

Joe Perches's avatar
Joe Perches committed
55
#include "console_cmdline.h"
56
#include "braille.h"
57
#include "internal.h"
Joe Perches's avatar
Joe Perches committed
58

Linus Torvalds's avatar
Linus Torvalds committed
59
int console_printk[4] = {
60
	CONSOLE_LOGLEVEL_DEFAULT,	/* console_loglevel */
61
	MESSAGE_LOGLEVEL_DEFAULT,	/* default_message_loglevel */
62
63
	CONSOLE_LOGLEVEL_MIN,		/* minimum_console_loglevel */
	CONSOLE_LOGLEVEL_DEFAULT,	/* default_console_loglevel */
Linus Torvalds's avatar
Linus Torvalds committed
64
65
66
};

/*
67
 * Low level drivers may need that to know if they can schedule in
Linus Torvalds's avatar
Linus Torvalds committed
68
69
70
71
72
73
74
75
76
77
 * their unblank() callback or not. So let's export it.
 */
int oops_in_progress;
EXPORT_SYMBOL(oops_in_progress);

/*
 * console_sem protects the console_drivers list, and also
 * provides serialisation for access to the entire console
 * driver system.
 */
78
static DEFINE_SEMAPHORE(console_sem);
Linus Torvalds's avatar
Linus Torvalds committed
79
struct console *console_drivers;
Ingo Molnar's avatar
Ingo Molnar committed
80
81
EXPORT_SYMBOL_GPL(console_drivers);

82
83
84
85
86
87
#ifdef CONFIG_LOCKDEP
static struct lockdep_map console_lock_dep_map = {
	.name = "console_lock"
};
#endif

88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
enum devkmsg_log_bits {
	__DEVKMSG_LOG_BIT_ON = 0,
	__DEVKMSG_LOG_BIT_OFF,
	__DEVKMSG_LOG_BIT_LOCK,
};

enum devkmsg_log_masks {
	DEVKMSG_LOG_MASK_ON             = BIT(__DEVKMSG_LOG_BIT_ON),
	DEVKMSG_LOG_MASK_OFF            = BIT(__DEVKMSG_LOG_BIT_OFF),
	DEVKMSG_LOG_MASK_LOCK           = BIT(__DEVKMSG_LOG_BIT_LOCK),
};

/* Keep both the 'on' and 'off' bits clear, i.e. ratelimit by default: */
#define DEVKMSG_LOG_MASK_DEFAULT	0

static unsigned int __read_mostly devkmsg_log = DEVKMSG_LOG_MASK_DEFAULT;

static int __control_devkmsg(char *str)
{
	if (!str)
		return -EINVAL;

	if (!strncmp(str, "on", 2)) {
		devkmsg_log = DEVKMSG_LOG_MASK_ON;
		return 2;
	} else if (!strncmp(str, "off", 3)) {
		devkmsg_log = DEVKMSG_LOG_MASK_OFF;
		return 3;
	} else if (!strncmp(str, "ratelimit", 9)) {
		devkmsg_log = DEVKMSG_LOG_MASK_DEFAULT;
		return 9;
	}
	return -EINVAL;
}

static int __init control_devkmsg(char *str)
{
	if (__control_devkmsg(str) < 0)
		return 1;

	/*
	 * Set sysctl string accordingly:
	 */
	if (devkmsg_log == DEVKMSG_LOG_MASK_ON) {
		memset(devkmsg_log_str, 0, DEVKMSG_STR_MAX_SIZE);
		strncpy(devkmsg_log_str, "on", 2);
	} else if (devkmsg_log == DEVKMSG_LOG_MASK_OFF) {
		memset(devkmsg_log_str, 0, DEVKMSG_STR_MAX_SIZE);
		strncpy(devkmsg_log_str, "off", 3);
	}
	/* else "ratelimit" which is set by default. */

	/*
	 * Sysctl cannot change it anymore. The kernel command line setting of
	 * this parameter is to force the setting to be permanent throughout the
	 * runtime of the system. This is a precation measure against userspace
	 * trying to be a smarta** and attempting to change it up on us.
	 */
	devkmsg_log |= DEVKMSG_LOG_MASK_LOCK;

	return 0;
}
__setup("printk.devkmsg=", control_devkmsg);

char devkmsg_log_str[DEVKMSG_STR_MAX_SIZE] = "ratelimit";

int devkmsg_sysctl_set_loglvl(struct ctl_table *table, int write,
			      void __user *buffer, size_t *lenp, loff_t *ppos)
{
	char old_str[DEVKMSG_STR_MAX_SIZE];
	unsigned int old;
	int err;

	if (write) {
		if (devkmsg_log & DEVKMSG_LOG_MASK_LOCK)
			return -EINVAL;

		old = devkmsg_log;
		strncpy(old_str, devkmsg_log_str, DEVKMSG_STR_MAX_SIZE);
	}

	err = proc_dostring(table, write, buffer, lenp, ppos);
	if (err)
		return err;

	if (write) {
		err = __control_devkmsg(devkmsg_log_str);

		/*
		 * Do not accept an unknown string OR a known string with
		 * trailing crap...
		 */
		if (err < 0 || (err + 1 != *lenp)) {

			/* ... and restore old setting. */
			devkmsg_log = old;
			strncpy(devkmsg_log_str, old_str, DEVKMSG_STR_MAX_SIZE);

			return -EINVAL;
		}
	}

	return 0;
}

193
194
195
196
197
198
199
200
201
202
203
204
/*
 * Number of registered extended console drivers.
 *
 * If extended consoles are present, in-kernel cont reassembly is disabled
 * and each fragment is stored as a separate log entry with proper
 * continuation flag so that every emitted message has full metadata.  This
 * doesn't change the result for regular consoles or /proc/kmsg.  For
 * /dev/kmsg, as long as the reader concatenates messages according to
 * consecutive continuation flags, the end result should be the same too.
 */
static int nr_ext_console_drivers;

205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
/*
 * Helper macros to handle lockdep when locking/unlocking console_sem. We use
 * macros instead of functions so that _RET_IP_ contains useful information.
 */
#define down_console_sem() do { \
	down(&console_sem);\
	mutex_acquire(&console_lock_dep_map, 0, 0, _RET_IP_);\
} while (0)

static int __down_trylock_console_sem(unsigned long ip)
{
	if (down_trylock(&console_sem))
		return 1;
	mutex_acquire(&console_lock_dep_map, 0, 1, ip);
	return 0;
}
#define down_trylock_console_sem() __down_trylock_console_sem(_RET_IP_)

#define up_console_sem() do { \
	mutex_release(&console_lock_dep_map, 1, _RET_IP_);\
	up(&console_sem);\
} while (0)

Linus Torvalds's avatar
Linus Torvalds committed
228
229
230
231
/*
 * This is used for debugging the mess that is the VT code by
 * keeping track if we have the console semaphore held. It's
 * definitely not the perfect debug tool (we don't know if _WE_
Alex Elder's avatar
Alex Elder committed
232
233
234
 * hold it and are racing, but it helps tracking those weird code
 * paths in the console code where we end up in places I want
 * locked without the console sempahore held).
Linus Torvalds's avatar
Linus Torvalds committed
235
 */
236
static int console_locked, console_suspended;
Linus Torvalds's avatar
Linus Torvalds committed
237

238
239
240
241
242
/*
 * If exclusive_console is non-NULL then only this console is to be printed to.
 */
static struct console *exclusive_console;

Linus Torvalds's avatar
Linus Torvalds committed
243
244
245
246
247
248
249
/*
 *	Array of consoles built from command line options (console=)
 */

#define MAX_CMDLINECONSOLES 8

static struct console_cmdline console_cmdline[MAX_CMDLINECONSOLES];
Joe Perches's avatar
Joe Perches committed
250

Linus Torvalds's avatar
Linus Torvalds committed
251
252
static int selected_console = -1;
static int preferred_console = -1;
253
254
int console_set_on_cmdline;
EXPORT_SYMBOL(console_set_on_cmdline);
Linus Torvalds's avatar
Linus Torvalds committed
255

256
257
258
259
260
261
262
263
264
265
266
#ifdef CONFIG_OF
static bool of_specified_console;

void console_set_by_of(void)
{
	of_specified_console = true;
}
#else
# define of_specified_console false
#endif

Linus Torvalds's avatar
Linus Torvalds committed
267
268
269
/* Flag: console code may call schedule() */
static int console_may_schedule;

270
271
272
273
274
275
/*
 * The printk log buffer consists of a chain of concatenated variable
 * length records. Every record starts with a record header, containing
 * the overall length of the record.
 *
 * The heads to the first and last entry in the buffer, as well as the
Alex Elder's avatar
Alex Elder committed
276
277
 * sequence numbers of these entries are maintained when messages are
 * stored.
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
 *
 * If the heads indicate available messages, the length in the header
 * tells the start next message. A length == 0 for the next message
 * indicates a wrap-around to the beginning of the buffer.
 *
 * Every record carries the monotonic timestamp in microseconds, as well as
 * the standard userspace syslog level and syslog facility. The usual
 * kernel messages use LOG_KERN; userspace-injected messages always carry
 * a matching syslog facility, by default LOG_USER. The origin of every
 * message can be reliably determined that way.
 *
 * The human readable log message directly follows the message header. The
 * length of the message text is stored in the header, the stored message
 * is not terminated.
 *
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
 * Optionally, a message can carry a dictionary of properties (key/value pairs),
 * to provide userspace with a machine-readable message context.
 *
 * Examples for well-defined, commonly used property names are:
 *   DEVICE=b12:8               device identifier
 *                                b12:8         block dev_t
 *                                c127:3        char dev_t
 *                                n8            netdev ifindex
 *                                +sound:card0  subsystem:devname
 *   SUBSYSTEM=pci              driver-core subsystem name
 *
 * Valid characters in property names are [a-zA-Z0-9.-_]. The plain text value
 * follows directly after a '=' character. Every property is terminated by
 * a '\0' character. The last property is not terminated.
 *
 * Example of a message structure:
 *   0000  ff 8f 00 00 00 00 00 00      monotonic time in nsec
 *   0008  34 00                        record is 52 bytes long
 *   000a        0b 00                  text is 11 bytes long
 *   000c              1f 00            dictionary is 23 bytes long
 *   000e                    03 00      LOG_KERN (facility) LOG_ERR (level)
 *   0010  69 74 27 73 20 61 20 6c      "it's a l"
 *         69 6e 65                     "ine"
 *   001b           44 45 56 49 43      "DEVIC"
 *         45 3d 62 38 3a 32 00 44      "E=b8:2\0D"
 *         52 49 56 45 52 3d 62 75      "RIVER=bu"
 *         67                           "g"
 *   0032     00 00 00                  padding to next message header
 *
322
 * The 'struct printk_log' buffer header must never be directly exported to
323
324
325
326
 * userspace, it is a kernel-private implementation detail that might
 * need to be changed in the future, when the requirements change.
 *
 * /dev/kmsg exports the structured data in the following line format:
327
328
329
330
 *   "<level>,<sequnum>,<timestamp>,<contflag>[,additional_values, ... ];<message text>\n"
 *
 * Users of the export format should ignore possible additional values
 * separated by ',', and find the message after the ';' character.
331
332
333
334
 *
 * The optional key/value pairs are attached as continuation lines starting
 * with a space character and terminated by a newline. All possible
 * non-prinatable characters are escaped in the "\xff" notation.
335
336
 */

337
enum log_flags {
338
339
340
341
	LOG_NOCONS	= 1,	/* already flushed, do not print to console */
	LOG_NEWLINE	= 2,	/* text ended with a newline */
	LOG_PREFIX	= 4,	/* text started with a prefix */
	LOG_CONT	= 8,	/* text is a fragment of a continuation line */
342
343
};

344
struct printk_log {
345
346
347
348
	u64 ts_nsec;		/* timestamp in nanoseconds */
	u16 len;		/* length of entire record */
	u16 text_len;		/* length of text buffer */
	u16 dict_len;		/* length of dictionary buffer */
349
350
351
	u8 facility;		/* syslog facility */
	u8 flags:5;		/* internal record flags */
	u8 level:3;		/* syslog level */
352
353
354
355
356
}
#ifdef CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS
__packed __aligned(4)
#endif
;
357
358

/*
359
360
361
 * The logbuf_lock protects kmsg buffer, indices, counters.  This can be taken
 * within the scheduler's rq lock. It must be released before calling
 * console_unlock() or anything else that might wake up a process.
362
 */
363
DEFINE_RAW_SPINLOCK(logbuf_lock);
364

365
#ifdef CONFIG_PRINTK
366
DECLARE_WAIT_QUEUE_HEAD(log_wait);
367
368
369
/* the next printk record to read by syslog(READ) or /proc/kmsg */
static u64 syslog_seq;
static u32 syslog_idx;
370
static enum log_flags syslog_prev;
371
static size_t syslog_partial;
372
373
374
375
376
377
378
379
380

/* index and sequence number of the first record stored in the buffer */
static u64 log_first_seq;
static u32 log_first_idx;

/* index and sequence number of the next record to store in the buffer */
static u64 log_next_seq;
static u32 log_next_idx;

381
382
383
384
385
/* the next printk record to write to the console */
static u64 console_seq;
static u32 console_idx;
static enum log_flags console_prev;

386
387
388
389
/* the next printk record to read after the last 'clear' command */
static u64 clear_seq;
static u32 clear_idx;

390
#define PREFIX_MAX		32
Alex Elder's avatar
Alex Elder committed
391
#define LOG_LINE_MAX		(1024 - PREFIX_MAX)
392

393
394
395
#define LOG_LEVEL(v)		((v) & 0x07)
#define LOG_FACILITY(v)		((v) >> 3 & 0xff)

396
/* record buffer */
397
#define LOG_ALIGN __alignof__(struct printk_log)
398
#define __LOG_BUF_LEN (1 << CONFIG_LOG_BUF_SHIFT)
399
static char __log_buf[__LOG_BUF_LEN] __aligned(LOG_ALIGN);
400
401
402
static char *log_buf = __log_buf;
static u32 log_buf_len = __LOG_BUF_LEN;

403
404
405
406
407
408
409
410
411
412
413
414
/* Return log buffer address */
char *log_buf_addr_get(void)
{
	return log_buf;
}

/* Return log buffer size */
u32 log_buf_len_get(void)
{
	return log_buf_len;
}

415
/* human readable text of the record */
416
static char *log_text(const struct printk_log *msg)
417
{
418
	return (char *)msg + sizeof(struct printk_log);
419
420
421
}

/* optional key/value pair dictionary attached to the record */
422
static char *log_dict(const struct printk_log *msg)
423
{
424
	return (char *)msg + sizeof(struct printk_log) + msg->text_len;
425
426
427
}

/* get record by index; idx must point to valid msg */
428
static struct printk_log *log_from_idx(u32 idx)
429
{
430
	struct printk_log *msg = (struct printk_log *)(log_buf + idx);
431
432
433
434
435
436

	/*
	 * A length == 0 record is the end of buffer marker. Wrap around and
	 * read the message at the start of the buffer.
	 */
	if (!msg->len)
437
		return (struct printk_log *)log_buf;
438
439
440
441
442
443
	return msg;
}

/* get next record; idx must point to valid msg */
static u32 log_next(u32 idx)
{
444
	struct printk_log *msg = (struct printk_log *)(log_buf + idx);
445
446
447
448
449
450
451
452

	/* length == 0 indicates the end of the buffer; wrap */
	/*
	 * A length == 0 record is the end of buffer marker. Wrap around and
	 * read the message at the start of the buffer as *this* one, and
	 * return the one after that.
	 */
	if (!msg->len) {
453
		msg = (struct printk_log *)log_buf;
454
455
456
457
458
		return msg->len;
	}
	return idx + msg->len;
}

Petr Mladek's avatar
Petr Mladek committed
459
460
461
462
463
464
465
466
467
468
/*
 * Check whether there is enough free space for the given message.
 *
 * The same values of first_idx and next_idx mean that the buffer
 * is either empty or full.
 *
 * If the buffer is empty, we must respect the position of the indexes.
 * They cannot be reset to the beginning of the buffer.
 */
static int logbuf_has_space(u32 msg_size, bool empty)
469
470
471
{
	u32 free;

Petr Mladek's avatar
Petr Mladek committed
472
	if (log_next_idx > log_first_idx || empty)
473
474
475
476
477
478
479
480
481
482
483
		free = max(log_buf_len - log_next_idx, log_first_idx);
	else
		free = log_first_idx - log_next_idx;

	/*
	 * We need space also for an empty header that signalizes wrapping
	 * of the buffer.
	 */
	return free >= msg_size + sizeof(struct printk_log);
}

Petr Mladek's avatar
Petr Mladek committed
484
static int log_make_free_space(u32 msg_size)
485
{
486
487
	while (log_first_seq < log_next_seq &&
	       !logbuf_has_space(msg_size, false)) {
Alex Elder's avatar
Alex Elder committed
488
		/* drop old messages until we have enough contiguous space */
489
490
491
		log_first_idx = log_next(log_first_idx);
		log_first_seq++;
	}
Petr Mladek's avatar
Petr Mladek committed
492

493
494
495
496
497
	if (clear_seq < log_first_seq) {
		clear_seq = log_first_seq;
		clear_idx = log_first_idx;
	}

Petr Mladek's avatar
Petr Mladek committed
498
	/* sequence numbers are equal, so the log buffer is empty */
499
	if (logbuf_has_space(msg_size, log_first_seq == log_next_seq))
Petr Mladek's avatar
Petr Mladek committed
500
501
502
		return 0;

	return -ENOMEM;
503
504
}

505
506
507
508
509
510
511
512
513
514
515
516
/* compute the message size including the padding bytes */
static u32 msg_used_size(u16 text_len, u16 dict_len, u32 *pad_len)
{
	u32 size;

	size = sizeof(struct printk_log) + text_len + dict_len;
	*pad_len = (-size) & (LOG_ALIGN - 1);
	size += *pad_len;

	return size;
}

Petr Mladek's avatar
Petr Mladek committed
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
/*
 * Define how much of the log buffer we could take at maximum. The value
 * must be greater than two. Note that only half of the buffer is available
 * when the index points to the middle.
 */
#define MAX_LOG_TAKE_PART 4
static const char trunc_msg[] = "<truncated>";

static u32 truncate_msg(u16 *text_len, u16 *trunc_msg_len,
			u16 *dict_len, u32 *pad_len)
{
	/*
	 * The message should not take the whole buffer. Otherwise, it might
	 * get removed too soon.
	 */
	u32 max_text_len = log_buf_len / MAX_LOG_TAKE_PART;
	if (*text_len > max_text_len)
		*text_len = max_text_len;
	/* enable the warning message */
	*trunc_msg_len = strlen(trunc_msg);
	/* disable the "dict" completely */
	*dict_len = 0;
	/* compute the size again, count also the warning message */
	return msg_used_size(*text_len + *trunc_msg_len, 0, pad_len);
}

543
/* insert record into the buffer, discard old ones, update heads */
544
545
546
547
static int log_store(int facility, int level,
		     enum log_flags flags, u64 ts_nsec,
		     const char *dict, u16 dict_len,
		     const char *text, u16 text_len)
548
{
549
	struct printk_log *msg;
550
	u32 size, pad_len;
Petr Mladek's avatar
Petr Mladek committed
551
	u16 trunc_msg_len = 0;
552
553

	/* number of '\0' padding bytes to next message */
554
	size = msg_used_size(text_len, dict_len, &pad_len);
555

Petr Mladek's avatar
Petr Mladek committed
556
557
558
559
560
561
	if (log_make_free_space(size)) {
		/* truncate the message if it is too long for empty buffer */
		size = truncate_msg(&text_len, &trunc_msg_len,
				    &dict_len, &pad_len);
		/* survive when the log buffer is too small for trunc_msg */
		if (log_make_free_space(size))
562
			return 0;
Petr Mladek's avatar
Petr Mladek committed
563
	}
564

565
	if (log_next_idx + size + sizeof(struct printk_log) > log_buf_len) {
566
567
568
569
570
		/*
		 * This message + an additional empty header does not fit
		 * at the end of the buffer. Add an empty header with len == 0
		 * to signify a wrap around.
		 */
571
		memset(log_buf + log_next_idx, 0, sizeof(struct printk_log));
572
573
574
575
		log_next_idx = 0;
	}

	/* fill message */
576
	msg = (struct printk_log *)(log_buf + log_next_idx);
577
578
	memcpy(log_text(msg), text, text_len);
	msg->text_len = text_len;
Petr Mladek's avatar
Petr Mladek committed
579
580
581
582
	if (trunc_msg_len) {
		memcpy(log_text(msg) + text_len, trunc_msg, trunc_msg_len);
		msg->text_len += trunc_msg_len;
	}
583
584
	memcpy(log_dict(msg), dict, dict_len);
	msg->dict_len = dict_len;
585
586
587
588
589
590
591
	msg->facility = facility;
	msg->level = level & 7;
	msg->flags = flags & 0x1f;
	if (ts_nsec > 0)
		msg->ts_nsec = ts_nsec;
	else
		msg->ts_nsec = local_clock();
592
	memset(log_dict(msg) + dict_len, 0, pad_len);
593
	msg->len = size;
594
595
596
597

	/* insert message */
	log_next_idx += msg->len;
	log_next_seq++;
598
599

	return msg->text_len;
600
}
601

Alex Elder's avatar
Alex Elder committed
602
int dmesg_restrict = IS_ENABLED(CONFIG_SECURITY_DMESG_RESTRICT);
603
604
605
606
607
608
609
610
611
612
613
614
615

static int syslog_action_restricted(int type)
{
	if (dmesg_restrict)
		return 1;
	/*
	 * Unless restricted, we allow "read all" and "get buffer size"
	 * for everybody.
	 */
	return type != SYSLOG_ACTION_READ_ALL &&
	       type != SYSLOG_ACTION_SIZE_BUFFER;
}

616
int check_syslog_permissions(int type, int source)
617
618
619
620
621
{
	/*
	 * If this is from /proc/kmsg and we've already opened it, then we've
	 * already done the capabilities checks at open time.
	 */
622
	if (source == SYSLOG_FROM_PROC && type != SYSLOG_ACTION_OPEN)
623
		goto ok;
624
625
626

	if (syslog_action_restricted(type)) {
		if (capable(CAP_SYSLOG))
627
			goto ok;
628
629
630
631
632
633
634
635
636
		/*
		 * For historical reasons, accept CAP_SYS_ADMIN too, with
		 * a warning.
		 */
		if (capable(CAP_SYS_ADMIN)) {
			pr_warn_once("%s (%d): Attempt to access syslog with "
				     "CAP_SYS_ADMIN but no CAP_SYSLOG "
				     "(deprecated).\n",
				 current->comm, task_pid_nr(current));
637
			goto ok;
638
639
640
		}
		return -EPERM;
	}
641
ok:
642
643
	return security_syslog(type);
}
Geliang Tang's avatar
Geliang Tang committed
644
EXPORT_SYMBOL_GPL(check_syslog_permissions);
645

646
647
648
649
650
static void append_char(char **pp, char *e, char c)
{
	if (*pp < e)
		*(*pp)++ = c;
}
651

652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
static ssize_t msg_print_ext_header(char *buf, size_t size,
				    struct printk_log *msg, u64 seq,
				    enum log_flags prev_flags)
{
	u64 ts_usec = msg->ts_nsec;
	char cont = '-';

	do_div(ts_usec, 1000);

	/*
	 * If we couldn't merge continuation line fragments during the print,
	 * export the stored flags to allow an optional external merge of the
	 * records. Merging the records isn't always neccessarily correct, like
	 * when we hit a race during printing. In most cases though, it produces
	 * better readable output. 'c' in the record flags mark the first
	 * fragment of a line, '+' the following.
	 */
669
670
	if (msg->flags & LOG_CONT)
		cont = (prev_flags & LOG_CONT) ? '+' : 'c';
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723

	return scnprintf(buf, size, "%u,%llu,%llu,%c;",
		       (msg->facility << 3) | msg->level, seq, ts_usec, cont);
}

static ssize_t msg_print_ext_body(char *buf, size_t size,
				  char *dict, size_t dict_len,
				  char *text, size_t text_len)
{
	char *p = buf, *e = buf + size;
	size_t i;

	/* escape non-printable characters */
	for (i = 0; i < text_len; i++) {
		unsigned char c = text[i];

		if (c < ' ' || c >= 127 || c == '\\')
			p += scnprintf(p, e - p, "\\x%02x", c);
		else
			append_char(&p, e, c);
	}
	append_char(&p, e, '\n');

	if (dict_len) {
		bool line = true;

		for (i = 0; i < dict_len; i++) {
			unsigned char c = dict[i];

			if (line) {
				append_char(&p, e, ' ');
				line = false;
			}

			if (c == '\0') {
				append_char(&p, e, '\n');
				line = true;
				continue;
			}

			if (c < ' ' || c >= 127 || c == '\\') {
				p += scnprintf(p, e - p, "\\x%02x", c);
				continue;
			}

			append_char(&p, e, c);
		}
		append_char(&p, e, '\n');
	}

	return p - buf;
}

724
725
726
727
/* /dev/kmsg - userspace message inject/listen interface */
struct devkmsg_user {
	u64 seq;
	u32 idx;
728
	enum log_flags prev;
729
	struct ratelimit_state rs;
730
	struct mutex lock;
731
	char buf[CONSOLE_EXT_LOG_MAX];
732
733
};

Al Viro's avatar
Al Viro committed
734
static ssize_t devkmsg_write(struct kiocb *iocb, struct iov_iter *from)
735
736
737
738
{
	char *buf, *line;
	int level = default_message_loglevel;
	int facility = 1;	/* LOG_USER */
739
740
	struct file *file = iocb->ki_filp;
	struct devkmsg_user *user = file->private_data;
Christoph Hellwig's avatar
Christoph Hellwig committed
741
	size_t len = iov_iter_count(from);
742
743
	ssize_t ret = len;

744
	if (!user || len > LOG_LINE_MAX)
745
		return -EINVAL;
746
747
748
749
750
751
752
753
754
755
756

	/* Ignore when user logging is disabled. */
	if (devkmsg_log & DEVKMSG_LOG_MASK_OFF)
		return len;

	/* Ratelimit when not explicitly enabled. */
	if (!(devkmsg_log & DEVKMSG_LOG_MASK_ON)) {
		if (!___ratelimit(&user->rs, current->comm))
			return ret;
	}

757
758
759
760
	buf = kmalloc(len+1, GFP_KERNEL);
	if (buf == NULL)
		return -ENOMEM;

Al Viro's avatar
Al Viro committed
761
762
763
764
	buf[len] = '\0';
	if (copy_from_iter(buf, len, from) != len) {
		kfree(buf);
		return -EFAULT;
765
766
767
768
769
770
771
772
773
774
775
776
777
778
	}

	/*
	 * Extract and skip the syslog prefix <[0-9]*>. Coming from userspace
	 * the decimal value represents 32bit, the lower 3 bit are the log
	 * level, the rest are the log facility.
	 *
	 * If no prefix or no userspace facility is specified, we
	 * enforce LOG_USER, to be able to reliably distinguish
	 * kernel-generated messages from userspace-injected ones.
	 */
	line = buf;
	if (line[0] == '<') {
		char *endp = NULL;
779
		unsigned int u;
780

781
		u = simple_strtoul(line + 1, &endp, 10);
782
		if (endp && endp[0] == '>') {
783
784
785
			level = LOG_LEVEL(u);
			if (LOG_FACILITY(u) != 0)
				facility = LOG_FACILITY(u);
786
787
788
789
790
791
792
793
794
795
796
			endp++;
			len -= endp - line;
			line = endp;
		}
	}

	printk_emit(facility, level, NULL, 0, "%s", line);
	kfree(buf);
	return ret;
}

797
798
static void cont_flush(void);

799
800
801
802
static ssize_t devkmsg_read(struct file *file, char __user *buf,
			    size_t count, loff_t *ppos)
{
	struct devkmsg_user *user = file->private_data;
803
	struct printk_log *msg;
804
805
806
807
808
809
	size_t len;
	ssize_t ret;

	if (!user)
		return -EBADF;

810
811
812
	ret = mutex_lock_interruptible(&user->lock);
	if (ret)
		return ret;
813
	raw_spin_lock_irq(&logbuf_lock);
814
	cont_flush();
815
816
817
	while (user->seq == log_next_seq) {
		if (file->f_flags & O_NONBLOCK) {
			ret = -EAGAIN;
818
			raw_spin_unlock_irq(&logbuf_lock);
819
820
821
			goto out;
		}

822
		raw_spin_unlock_irq(&logbuf_lock);
823
824
825
826
		ret = wait_event_interruptible(log_wait,
					       user->seq != log_next_seq);
		if (ret)
			goto out;
827
		raw_spin_lock_irq(&logbuf_lock);
828
829
830
831
832
833
834
	}

	if (user->seq < log_first_seq) {
		/* our last seen message is gone, return error and reset */
		user->idx = log_first_idx;
		user->seq = log_first_seq;
		ret = -EPIPE;
835
		raw_spin_unlock_irq(&logbuf_lock);
836
837
838
839
		goto out;
	}

	msg = log_from_idx(user->idx);
840
841
842
843
844
	len = msg_print_ext_header(user->buf, sizeof(user->buf),
				   msg, user->seq, user->prev);
	len += msg_print_ext_body(user->buf + len, sizeof(user->buf) - len,
				  log_dict(msg), msg->dict_len,
				  log_text(msg), msg->text_len);
845
846

	user->prev = msg->flags;
847
848
	user->idx = log_next(user->idx);
	user->seq++;
849
	raw_spin_unlock_irq(&logbuf_lock);
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875

	if (len > count) {
		ret = -EINVAL;
		goto out;
	}

	if (copy_to_user(buf, user->buf, len)) {
		ret = -EFAULT;
		goto out;
	}
	ret = len;
out:
	mutex_unlock(&user->lock);
	return ret;
}

static loff_t devkmsg_llseek(struct file *file, loff_t offset, int whence)
{
	struct devkmsg_user *user = file->private_data;
	loff_t ret = 0;

	if (!user)
		return -EBADF;
	if (offset)
		return -ESPIPE;

876
	raw_spin_lock_irq(&logbuf_lock);
877
	cont_flush();
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
	switch (whence) {
	case SEEK_SET:
		/* the first record */
		user->idx = log_first_idx;
		user->seq = log_first_seq;
		break;
	case SEEK_DATA:
		/*
		 * The first record after the last SYSLOG_ACTION_CLEAR,
		 * like issued by 'dmesg -c'. Reading /dev/kmsg itself
		 * changes no global state, and does not clear anything.
		 */
		user->idx = clear_idx;
		user->seq = clear_seq;
		break;
	case SEEK_END:
		/* after the last record */
		user->idx = log_next_idx;
		user->seq = log_next_seq;
		break;
	default:
		ret = -EINVAL;
	}
901
	raw_spin_unlock_irq(&logbuf_lock);
902
903
904
905
906
907
908
909
910
911
912
913
914
	return ret;
}

static unsigned int devkmsg_poll(struct file *file, poll_table *wait)
{
	struct devkmsg_user *user = file->private_data;
	int ret = 0;

	if (!user)
		return POLLERR|POLLNVAL;

	poll_wait(file, &log_wait, wait);

915
	raw_spin_lock_irq(&logbuf_lock);
916
	cont_flush();
917
918
919
920
	if (user->seq < log_next_seq) {
		/* return error when data has vanished underneath us */
		if (user->seq < log_first_seq)
			ret = POLLIN|POLLRDNORM|POLLERR|POLLPRI;
921
922
		else
			ret = POLLIN|POLLRDNORM;
923
	}
924
	raw_spin_unlock_irq(&logbuf_lock);
925
926
927
928
929
930
931
932
933

	return ret;
}

static int devkmsg_open(struct inode *inode, struct file *file)
{
	struct devkmsg_user *user;
	int err;

934
935
	if (devkmsg_log & DEVKMSG_LOG_MASK_OFF)
		return -EPERM;
936

937
938
939
940
941
942
943
	/* write-only does not need any file context */
	if ((file->f_flags & O_ACCMODE) != O_WRONLY) {
		err = check_syslog_permissions(SYSLOG_ACTION_READ_ALL,
					       SYSLOG_FROM_READER);
		if (err)
			return err;
	}
944
945
946
947
948

	user = kmalloc(sizeof(struct devkmsg_user), GFP_KERNEL);
	if (!user)
		return -ENOMEM;

949
950
951
	ratelimit_default_init(&user->rs);
	ratelimit_set_flags(&user->rs, RATELIMIT_MSG_ON_RELEASE);

952
953
	mutex_init(&user->lock);

954
	raw_spin_lock_irq(&logbuf_lock);
955
956
	user->idx = log_first_idx;
	user->seq = log_first_seq;
957
	raw_spin_unlock_irq(&logbuf_lock);
958
959
960
961
962
963
964
965
966
967
968
969

	file->private_data = user;
	return 0;
}

static int devkmsg_release(struct inode *inode, struct file *file)
{
	struct devkmsg_user *user = file->private_data;

	if (!user)
		return 0;

970
971
	ratelimit_state_exit(&user->rs);

972
973
974
975
976
977
978
979
	mutex_destroy(&user->lock);
	kfree(user);
	return 0;
}

const struct file_operations kmsg_fops = {
	.open = devkmsg_open,
	.read = devkmsg_read,
Al Viro's avatar
Al Viro committed
980
	.write_iter = devkmsg_write,
981
982
983
984
985
	.llseek = devkmsg_llseek,
	.poll = devkmsg_poll,
	.release = devkmsg_release,
};

986
#ifdef CONFIG_KEXEC_CORE
987
/*
988
 * This appends the listed symbols to /proc/vmcore
989
 *
990
 * /proc/vmcore is used by various utilities, like crash and makedumpfile to
991
992
993
994
995
996
997
998
 * obtain access to symbols that are otherwise very difficult to locate.  These
 * symbols are specifically used so that utilities can access and extract the
 * dmesg log from a vmcore file after a crash.
 */
void log_buf_kexec_setup(void)
{
	VMCOREINFO_SYMBOL(log_buf);
	VMCOREINFO_SYMBOL(log_buf_len);
999
	VMCOREINFO_SYMBOL(log_first_idx);
1000
	VMCOREINFO_SYMBOL(clear_idx);
1001
	VMCOREINFO_SYMBOL(log_next_idx);
1002
	/*
1003
	 * Export struct printk_log size and field offsets. User space tools can
1004
1005
	 * parse it and detect any changes to structure down the line.
	 */
1006
1007
1008
1009
1010
	VMCOREINFO_STRUCT_SIZE(printk_log);
	VMCOREINFO_OFFSET(printk_log, ts_nsec);
	VMCOREINFO_OFFSET(printk_log, len);
	VMCOREINFO_OFFSET(printk_log, text_len);
	VMCOREINFO_OFFSET(printk_log, dict_len);
1011
1012
1013
}
#endif

1014
1015
1016
/* requested log_buf_len from kernel cmdline */
static unsigned long __initdata new_log_buf_len;

1017
1018
/* we practice scaling the ring buffer by powers of 2 */
static void __init log_buf_len_update(unsigned size)
Linus Torvalds's avatar
Linus Torvalds committed
1019
1020
1021
{
	if (size)
		size = roundup_pow_of_two(size);
1022
1023
	if (size > log_buf_len)
		new_log_buf_len = size;
1024
1025
1026
1027
1028
1029
1030
1031
}

/* save requested log_buf_len since it's too early to process it */
static int __init log_buf_len_setup(char *str)
{
	unsigned size = memparse(str, &str);

	log_buf_len_update(size);
1032
1033

	return 0;
Linus Torvalds's avatar
Linus Torvalds committed
1034
}
1035
1036
early_param("log_buf_len", log_buf_len_setup);

1037
1038
1039
#ifdef CONFIG_SMP
#define __LOG_CPU_MAX_BUF_LEN (1 << CONFIG_LOG_CPU_MAX_BUF_SHIFT)

1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
static void __init log_buf_add_cpu(void)
{
	unsigned int cpu_extra;

	/*
	 * archs should set up cpu_possible_bits properly with
	 * set_cpu_possible() after setup_arch() but just in
	 * case lets ensure this is valid.
	 */
	if (num_possible_cpus() == 1)
		return;

	cpu_extra = (num_possible_cpus() - 1) * __LOG_CPU_MAX_BUF_LEN;

	/* by default this will only continue through for large > 64 CPUs */
	if (cpu_extra <= __LOG_BUF_LEN / 2)
		return;

	pr_info("log_buf_len individual max cpu contribution: %d bytes\n",
		__LOG_CPU_MAX_BUF_LEN);
	pr_info("log_buf_len total cpu_extra contributions: %d bytes\n",
		cpu_extra);
	pr_info("log_buf_len min size: %d bytes\n", __LOG_BUF_LEN);

	log_buf_len_update(cpu_extra + __LOG_BUF_LEN);
}
1066
1067
1068
#else /* !CONFIG_SMP */
static inline void log_buf_add_cpu(void) {}
#endif /* CONFIG_SMP */
1069

1070
1071
1072
1073
1074
1075
void __init setup_log_buf(int early)
{
	unsigned long flags;
	char *new_log_buf;
	int free;

1076
1077
1078
1079
1080
1081
	if (log_buf != __log_buf)
		return;

	if (!early && !new_log_buf_len)
		log_buf_add_cpu();

1082
1083
	if (!new_log_buf_len)
		return;
Linus Torvalds's avatar
Linus Torvalds committed
1084

1085
	if (early) {
1086
		new_log_buf =
1087
			memblock_virt_alloc(new_log_buf_len, LOG_ALIGN);
1088
	} else {
1089
1090
		new_log_buf = memblock_virt_alloc_nopanic(new_log_buf_len,
							  LOG_ALIGN);
1091
1092
1093
1094
1095
1096
1097
1098
	}

	if (unlikely(!new_log_buf)) {
		pr_err("log_buf_len: %ld bytes not available\n",
			new_log_buf_len);
		return;
	}

1099
	raw_spin_lock_irqsave(&logbuf_lock, flags);
1100
1101
1102
	log_buf_len = new_log_buf_len;
	log_buf = new_log_buf;
	new_log_buf_len = 0;
1103
1104
	free = __LOG_BUF_LEN - log_next_idx;
	memcpy(log_buf, __log_buf, __LOG_BUF_LEN);
1105
	raw_spin_unlock_irqrestore(&logbuf_lock, flags);
1106

1107
	pr_info("log_buf_len: %d bytes\n", log_buf_len);
1108
1109
1110
	pr_info("early log buf free: %d(%d%%)\n",
		free, (free * 100) / __LOG_BUF_LEN);
}
Linus Torvalds's avatar
Linus Torvalds committed
1111

1112
1113
1114
1115
static bool __read_mostly ignore_loglevel;

static int __init ignore_loglevel_setup(char *str)
{
1116
	ignore_loglevel = true;
1117
	pr_info("debug: ignoring loglevel setting.\n");
1118
1119
1120
1121
1122
1123

	return 0;
}

early_param("ignore_loglevel", ignore_loglevel_setup);
module_param(ignore_loglevel, bool, S_IRUGO | S_IWUSR);
1124
1125
MODULE_PARM_DESC(ignore_loglevel,
		 "ignore loglevel setting (prints all kernel messages to the console)");
1126

1127
1128
1129
1130
1131
static bool suppress_message_printing(int level)
{
	return (level >= console_loglevel && !ignore_loglevel);
}

Randy Dunlap's avatar
Randy Dunlap committed
1132
1133
#ifdef CONFIG_BOOT_PRINTK_DELAY

1134
static int boot_delay; /* msecs delay after each printk during bootup */
1135
static unsigned long long loops_per_msec;	/* based on boot_delay */
Randy Dunlap's avatar
Randy Dunlap committed
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147

static int __init boot_delay_setup(char *str)
{
	unsigned long lpj;

	lpj = preset_lpj ? preset_lpj : 1000000;	/* some guess */
	loops_per_msec = (unsigned long long)lpj / 1000 * HZ;

	get_option(&str, &boot_delay);
	if (boot_delay > 10 * 1000)
		boot_delay = 0;

1148
1149
1150
	pr_debug("boot_delay: %u, preset_lpj: %ld, lpj: %lu, "
		"HZ: %d, loops_per_msec: %llu\n",
		boot_delay, preset_lpj, lpj, HZ, loops_per_msec);
1151
	return 0;
Randy Dunlap's avatar
Randy Dunlap committed
1152
}
1153
early_param("boot_delay", boot_delay_setup);
Randy Dunlap's avatar
Randy Dunlap committed
1154

1155
static void boot_delay_msec(int level)
Randy Dunlap's avatar
Randy Dunlap committed
1156
1157
1158
1159
{
	unsigned long long k;
	unsigned long timeout;

1160
	if ((boot_delay == 0 || system_state != SYSTEM_BOOTING)
1161
		|| suppress_message_printing(level)) {
Randy Dunlap's avatar
Randy Dunlap committed
1162
		return;
1163
	}
Randy Dunlap's avatar
Randy Dunlap committed
1164

1165
	k = (unsigned long long)loops_per_msec * boot_delay;
Randy Dunlap's avatar
Randy Dunlap committed
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181

	timeout = jiffies + msecs_to_jiffies(boot_delay);
	while (k) {
		k--;
		cpu_relax();
		/*
		 * use (volatile) jiffies to prevent
		 * compiler reduction; loop termination via jiffies
		 * is secondary and may or may not happen.
		 */
		if (time_after(jiffies, timeout))
			break;
		touch_nmi_watchdog();
	}
}
#else
1182
static inline void boot_delay_msec(int level)
Randy Dunlap's avatar
Randy Dunlap committed
1183
1184
1185
1186
{
}
#endif

Alex Elder's avatar
Alex Elder committed
1187
static bool printk_time = IS_ENABLED(CONFIG_PRINTK_TIME);
1188
1189
module_param_named(time, printk_time, bool, S_IRUGO | S_IWUSR);

1190
1191
1192
1193
1194
1195
1196
static size_t print_time(u64 ts, char *buf)
{
	unsigned long rem_nsec;

	if (!printk_time)
		return 0;

1197
1198
	rem_nsec = do_div(ts, 1000000000);

1199
	if (!buf)
1200
		return snprintf(NULL, 0, "[%5lu.000000] ", (unsigned long)ts);
1201
1202
1203
1204
1205

	return sprintf(buf, "[%5lu.%06lu] ",
		       (unsigned long)ts, rem_nsec / 1000);
}

1206
static size_t print_prefix(const struct printk_log *msg, bool syslog, char *buf)
1207
{
1208
	size_t len = 0;
1209
	unsigned int prefix = (msg->facility << 3) | msg->level;
1210

1211
1212
	if (syslog) {
		if (buf) {
1213
			len += sprintf(buf, "<%u>", prefix);
1214
1215
		} else {
			len += 3;
1216
1217
1218
1219
1220
			if (prefix > 999)
				len += 3;
			else if (prefix > 99)
				len += 2;
			else if (prefix > 9)
1221
1222
1223
				len++;
		}
	}
1224

1225
	len += print_time(msg->ts_nsec, buf ? buf + len : NULL);
1226
	return len;
1227
1228
}

1229
static size_t msg_print_text(const struct printk_log *msg, enum log_flags prev,
1230
			     bool syslog, char *buf, size_t size)
1231
{
1232
1233
	const char *text = log_text(msg);
	size_t text_size = msg->text_len;
1234
1235
	bool prefix = true;
	bool newline = true;
1236
1237
	size_t len = 0;

1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
	if ((prev & LOG_CONT) && !(msg->flags & LOG_PREFIX))
		prefix = false;

	if (msg->flags & LOG_CONT) {
		if ((prev & LOG_CONT) && !(prev & LOG_NEWLINE))
			prefix = false;

		if (!(msg->flags & LOG_NEWLINE))
			newline = false;
	}

1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
	do {
		const char *next = memchr(text, '\n', text_size);
		size_t text_len;

		if (next) {
			text_len = next - text;
			next++;
			text_size -= next - text;
		} else {
			text_len = text_size;
		}
1260

1261
1262
		if (buf) {
			if (print_prefix(msg, syslog, NULL) +
1263
			    text_len + 1 >= size - len)
1264
				break;
1265

1266
1267
			if (prefix)
				len += print_prefix(msg, syslog, buf + len);
1268
1269
			memcpy(buf + len, text, text_len);
			len += text_len;
1270
1271
			if (next || newline)
				buf[len++] = '\n';
1272
1273
		} else {
			/* SYSLOG_ACTION_* buffer size only calculation */
1274
1275
1276
1277
1278
			if (prefix)
				len += print_prefix(msg, syslog, NULL);
			len += text_len;
			if (next || newline)
				len++;
1279
		}
1280

1281
		prefix = true;
1282
1283
		text = next;
	} while (text);
1284
1285
1286
1287
1288
1289
1290

	return len;
}

static int syslog_print(char __user *buf, int size)
{
	char *text;
1291
	struct printk_log *msg;
1292
	int len = 0;
1293

1294
	text = kmalloc(LOG_LINE_MAX + PREFIX_MAX, GFP_KERNEL);
1295
1296
1297
	if (!text)
		return -ENOMEM;

1298
1299
	while (size > 0) {
		size_t n;
1300
		size_t skip;
1301
1302

		raw_spin_lock_irq(&logbuf_lock);
1303
		cont_flush();
1304
1305
1306
1307
		if (syslog_seq < log_first_seq) {
			/* messages are gone, move to first one */
			syslog_seq = log_first_seq;
			syslog_idx = log_first_idx;
1308
			syslog_prev = 0;
1309
			syslog_partial = 0;
1310
1311
1312
1313
1314
		}
		if (syslog_seq == log_next_seq) {
			raw_spin_unlock_irq(&logbuf_lock);
			break;
		}
1315
1316

		skip = syslog_partial;
1317
		msg = log_from_idx(syslog_idx);
1318
1319
		n = msg_print_text(msg, syslog_prev, true, text,
				   LOG_LINE_MAX + PREFIX_MAX);
1320
1321
		if (n - syslog_partial <= size) {
			/* message fits into buffer, move forward */
1322
1323
			syslog_idx = log_next(syslog_idx);
			syslog_seq++;
1324
			syslog_prev = msg->flags;
1325
1326
1327
1328
1329
1330
			n -= syslog_partial;
			syslog_partial = 0;
		} else if (!len){
			/* partial read(), remember position */
			n = size;
			syslog_partial += n;
1331
1332
1333
1334
1335
1336
1337
		} else
			n = 0;
		raw_spin_unlock_irq(&logbuf_lock);

		if (!n)
			break;

1338
		if (copy_to_user(buf, text + skip, n)) {
1339
1340
1341
1342
			if (!len)
				len = -EFAULT;
			break;
		}
1343
1344
1345
1346

		len += n;
		size -= n;
		buf += n;
1347
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
	}

	kfree(text);
	return len;
}

static int syslog_print_all(char __user *buf, int size, bool clear)
{
	char *text;
	int len = 0;

1358
	text = kmalloc(LOG_LINE_MAX + PREFIX_MAX, GFP_KERNEL);
1359
1360
1361
1362
	if (!text)
		return -ENOMEM;

	raw_spin_lock_irq(&logbuf_lock);
1363
	cont_flush();
1364
1365
1366
1367
	if (buf) {
		u64 next_seq;
		u64 seq;<